 Perfect. Welcome everyone to the McLean lecture series. We're very excited to be ending our spring quarter of lectures with Yana Gallin and so I'm going to go ahead and introduce her and then just remind you that we are taking a break. We finished all of our fall quarter, all of our spring quarter and we're moving into our, I'm sorry, all of our winter quarter and we're going to move into our spring quarter starting back on March 22nd with Dr. Anna Volerman and then followed by Dr. Carl Street. But let me introduce Yana Gallin so we can get right to her lecture today. Yana Gallin is an assistant professor at the University of Chicago Harris School of Public Policy. She received a PhD in economics from Northwestern University in 2016. She's a labor economist studying the gender wage gap. Her research focuses on understanding the sources of the gender pay gap preferences, discrimination or productivity. She is also interested in the impact of family friendly policies on the labor market, particularly looking at indirect or unanticipated effects of policy reforms. Many of her projects use the Danish registry data linking workers and firms. Her current work focuses on the savings and career decisions of household before they know their preference concerning child rearing. If you take a look at her website, there are a number of interesting papers, a few of which she's going to be talking about today. And one of them was does information affect homophily, which I need to learn what that is, and also informed choices, gender gaps in career advice. So I'm really looking forward to hearing her talk and welcome Yana Gallin. Okay, thank you so much for that introduction. So I will caveat and say that I'm used to talking to economists. We have, you know, our own norms that include interrupting whenever you have a question and kind of having more of a conversation. I like it. So I welcome you to do that, especially because, you know, I don't have great perspective on, you know, when I'm using terminology that like come up with that that people might not be familiar with. And that probably happens a lot on these slides. So, so please ask and interrupt me anytime with questions, not even just, you know, definition kind of questions, any kind of questions that you have about what what claims I'm making. So what is this talk going to be about? I'm going to actually focus on papers that are part of kind of a series of work that I have with a colleague, Melanie Wasserman. And the question that motivated us was like, do people talk differently to men and women? And how does that affect their perceptions of the world and their choices? So when we're thinking about advice, in particular career advice, or maybe specialty choice for med students, there are people who have a lot of information, that's usually who's giving all of this advice, and people who have almost none. So unless you actually do a job, you really don't know a lot about how it's done. I know that was true even for me, as somebody who went on to be a professor, you know, you spend a lot of time around professors in college and graduate school, but remarkable information revelation once you actually get the job. So there's a lot that I didn't know, even though I was surrounded by people doing that job. It's magnified, I think, in other fields where, you know, you don't even see somebody doing that job unless your parent was an investment banker, you really have no idea what it's like to be an investment banker. So how do you make decisions about things that you don't know much about? Well, you gather some information, and maybe you go online, you can learn something there. But a lot of times what people are encouraged to do is to ask someone. And we focus on this part of information acquisition. So people asking about information about careers, who do they seek out, students themselves, and then what do we tell them when they seek us out? And the questions that we ask in particular are, is mentoring or advice giving that features homophily, so, you know, women matching with women, men matching with men. Is that good? Does it improve outcomes or might there be some problems with it? And why does it occur? And then do people tell male and female students the same thing about jobs? And should they, if they don't? So why might they be telling people different things? And what might be some concerns? And maybe we'll have a nice conversation afterward about the ethics of that, which usually isn't a columnist we don't talk about, actually. We just talk about, you know, whether it described the situation. So this first paper that I'm going to talk about is called Inform Choices, Gender, Gaps, and Career Advice. And as I said, it's joined with Melanie Wasserman at UCLA Anderson. So we know the information matters a lot for educational and occupational choices. When an experimenter sits down with undergraduate students and tells them about the returns to getting an economics degree rather than a public policy degree, it actually does change their choice. And so little bits of information we know have a strong impact on occupational choice. This is also true for job choice. And that's actually not usually how we get information. So it's not often that an experimenter sits down with undergrads and tells them about, you know, hey, I did this calculation. And these are the returns to getting a particular type of degree. Actually, people talk to each other privately. And these informal and private conversations are difficult to observe. They may not be accessible to everyone. But also they may depend on the characteristics of those involved in the conversation. And it's easy to imagine why that's good. But it's also kind of easy to imagine how it could be problematic. So what we ask in this paper is informal information that people give about careers gender-blind or is gender something that people incorporate into their decision of what to tell someone. So our methodology is going to actually be to conduct a large-scale field experiment, where we recruit undergraduate students who are actually interested in learning about a specific set of careers that I'll talk about. And we give them a format type of a way to ask the questions that they want to ask. And they ask real professionals these questions. And then they kind of report back to us what their answers they received. And we analyze those. And the research question that we have is, does student gender affect the information that they receive about careers? So we're going to focus on career attributes in our experimental design that are known to affect female labor market choices. Those are work-life balance and competitive culture in the workplace. So we'll have code whether students hear about those attributes when they ask broadly about the career. And then we'll also ask specifically, I'm concerned about work-life balance. Do you think this is a valid concern? And then we estimate average differences by student gender. So the question is, do male and female students who are asking exactly the same question get different responses or they get different response rates? And then we also can, because they're real students who are actually interested in the answers to these questions, you might think, oh, maybe there are a few people out there who are very biased in their information provision and they tell female students totally different things than they tell male students. But it doesn't matter because students never ask those people for information. So we also are able to ask students whether they who they want to talk to. So we have them rank the professionals that they would most want to get information from. And when we incorporate these student preferences, that difference between the information that male and female students get remains. So it's not the case that the experienced gaps in information are not real or they're different than the experimental ones that we observe. Should I click on the question in the chat? Okay, cool. What we see in terms of results is that, what we see in terms of experimental results is that female students receive more information about work-life balance concerns, actually a lot more. So when asking broadly about the pros and cons of a career, female students are told twice as often about work-life balance concerns compared to male students. And then we anticipated this, so we also had some questions in the study where they asked about work-life balance specifically. And when they say something like, you know, I'm concerned about work-life balance, do you think this is a valid concern? When they do that, male students just don't get responses at the same rate as female students. So professionals engage a lot more with female students on that question. And that doesn't change when we incorporate student preferences for professionals. So it's not the case that female students are going to people who give them more information about work-life balance or less. It's about the same. The very suggestive, I'll show some very suggestive evidence that actually even in this study with not that many students actually, that's why it's very suggestive, female students are more deterred from their preferred career path and that this is partially explained by information on work-life balance. So in particular, when female students received information about work-life balance, about their preferred career path, they express less interest in that career path at the end of the study. But this is very suggestive because we didn't have that many students. So our experimental design is focused on students sending messages to professionals on an online professional networking platform. It's a big one you might be able to guess. The fields that we recruited students from of interest are law. So students want to go into law in the future, management, consulting, data science, and finance. These are chosen because they're the top fields of interest for economics majors at the university where we did the study. And we kind of had to make the questions make sense for what we were specifically studying, although I think it would be interesting to study also bio majors and them gathering information about what to study in med school, what kind of med school to go to. But that's not here. So this is about econ majors and the kinds of careers that they go into. The sample is 10,000 professionals who graduated from top colleges similar to the college that the students went to who work in the student city and who work in one of these four fields. And then when we when we're setting it up, we recruited 100 college students who are interested in learning about these fields. Each student contacts 100 professionals 100 times 100 is that 10,000. So every professional received just one question. So we never see the same professional, you know, answering to multiple different people, they only receive one question. But it's completely random whether they're asked by a male or female student. And so we kind of compare those differences and attribute any difference in response to the gender of the student. Each professional receives a message, like I said, and then what we randomize is really whether the professional receives a message from a male or female student and the message that they receive. Because of just the distribution of what the fields of the professionals were, we had in this city, we had 13 data scientists, 28 finance professionals 33 lawyers and 26 management consultants that each student contacted. And each professional received one of four different types of message. One message was a broad message which asked about I'm interested in management consulting. Can you tell me about the advantages and challenges of that career? So it's a very broad question. It doesn't hint in any way at any particular attributes. And it's actually the wording of it is based on the wording suggested by career planning services for informational interviews when students are cold contacting professionals. So this is exactly what they're recommended to do. And the tone of the message is just the same as what what they're advised to ask. So in that sense, it's kind of realistic. Then we also had specific messages. So students asked about whether work life balance is a concern. And the reason that we did that is because you might think that if we see gender differences in responses to the broad message, maybe it's because professionals believe that female students care more about work life balance and male students don't. And as soon as they're told that that's not true, then okay, then that gender difference goes away. So here we're trying to kind of fix what the student is interested in learning about and want to see if there are still gender differences and how professionals interact with male and female students. And then we had a specific message which asked about whether competitive culture is a concern. So we're really focused on these two attributes in the study. In the messages, we'll have they'll talk about a lot of stuff that's not this, but this is what's kind of pre registered and what we're focused on. We also have a factual message which sadly, you know, this arm of the experiment is underpowered. So I'm just going to caveat that. But but we also asked lawyers only about available hours requirements and big law, which is like a number that you can look up. And so it kind of the answer does not differ by gender, unlike maybe any of these other questions. So like the truth doesn't differ by gender. And and we thought we would get more responses to that question than we did. So we're a little bit underpowered there just as a caveat. All messages emphasize the student is not seeking a job. So we're really trying to focus on this on this information gathering aspect and not anything to do with recruitment. You know, you might be worried that people in these fields in particular in finance are trying to recruit female students because they have a diversity problem. So we're trying to separate from that. This is not going away. Click on the chat. Got it. Okay, thanks. So so when we when we collect the data, what we see is a lot of different stuff. So first, you know, in terms of how the students conducted the experiment, they were asked to they're recruited based on their interest in these fields and wanting to know the answer to the questions that we're asking. And then they're asked to go on this professional networking platform and modify their profile to eliminate any useful information. All students were using the same profile picture, and it was like a picture of, you know, a vague person in the distance and like in a place that's very symbolic for the university. And you can't tell whether it's a male or female or anything about the person really, but there's a picture there. And then they include their true, you know, major, which was almost all econ their true their true graduation year and their name with, you know, just the first letter of their last name is their default. So were you relying on name to kind of indicate what the students gender is? And there isn't much else that a professional can see about a student because we asked them to eliminate all of their like work experience and things like that. And then students rank professionals. So they tell us, you know, of the list of people that you want me to contact here, the ones in order that I'm most interested in talking to so that we have a sense of like who who is realistically, you know, a person that students would want to talk to. And then students send messages. So they, regardless of whether they, you know, rank them number one or number 100 students are asked to send messages to all of the professionals on their list. And they didn't always do that. So there were a handful of cases of I know that person, you know, that kind of thing. And so then in that case, they didn't send the message. But almost all of the messages were sent. And they provided us with the responses they received. And then after about three weeks, we said, okay, if we haven't received a response by now, it's done. And then students filled out a post study survey, where they're asked about, you know, their preferences over careers. And these responses include a lot of information. So first week, see the response rate, whether a person responded at all, we see the career attributes they mentioned when they answer the broad question. And we can do some natural language processing, you know, we can study the length of the message, we can study the sentiment of the messages that generally positive or negative. And we can look at whether there's differential word usage for male and female messages. So there's like infinite number of things that you can examine about these messages, because some of them are quite long, actually, and detailed. But we're really in this study going to focus on just whether they mentioned work life balance, and the response rates, and competitive culture. And in the post study survey, we see student career plans. So we can link, you know, whether receiving more information of one type or another seem to be correlated with student plans. But this part isn't really what the experiment was designed around. So in a way, I think that what our focus is is thinking about getting a snapshot into these informal conversations. And I think there are a lot more informal conversations that are not happening on this one professional networking platform. And I'd almost be surprised if this like initial conversation on the professional networking platform had a huge impact on students lives. But our point is to show that, you know, this stuff happens, and maybe it happens more broadly. And in that important conversation that actually does, you know, determine your career choice, it may be happening there too. So that's all to say this isn't an information intervention on the student side, it's really just testing whether professionals are differentiating what they're saying to people with different sounding needs. And students from Raking Professionals, we can look at whether they're selecting people who are more or less differentiating information based on gender or not. Okay, I'm not going to talk about this too much, I guess it's maybe if, well, I'll mention it. So in economics and sociology and lots of other fields, there's a lot of work done using correspondence studies. So researchers send fictitious resumes to employers and see whether callback rates differ by the name, usually indicating race of the person who is on the resume. So all the qualifications are the same, all the information about the person is the same, and then they just put a different name on the resume. That's a correspondence study. And one criticism of those is that they do involve deception and we don't like to use deception in economics. You know, they're not real people behind them, and you are kind of wasting the time of the firm or the HR person who's reading the resumes. In our setting, that would be like much more concerning, I think, because they're real professionals who like have work to do and if we were using fake students, they would be responding to someone who, you know, they would be wasting their time and that that might be unethical. So our study recruits real students who actually want to know the answer to these questions, and they're the ones sending messages to professionals. So when a professional responds, you know, it is part of a study, but it's also doing what they wanted to do. So they're helping someone by providing them information about a career that they're interested in. So we think the study does respect professionals' time. There's a cost to incorporating real students rather than using fake students, which is that we do give up some control over the characteristics of the students. So in these resume studies, it's exactly the same. The only thing that changes is the name. Here, you know, male and female students might differ a little bit and they do on other dimensions, but we think that our setting allows us to mitigate those concerns. It's online profiles. So like I said, we restrict students' profiles to have minimal information. And the minimal information that's on there, we can use this like controls and our regression. So we can compare only students who are graduating in the same year, only students who are economists, econ majors, to econ majors, public policy majors, public policy majors. And then, you know, we have a lot more information about the students actually. So you might be worried that professionals can infer things like race or ethnicity from the names as well, or that they can go online and find some of the students and learn a lot more about the students than what we have in the profiles. And it's clear that occasionally they did that. Like they would, you know, say, oh, I see you're on the volleyball team or whatever. So, so, you know, this wasn't, this wasn't, it doesn't, we're not having perfect control over the attributes, but we can include controls for what we know about the students like their GPA and whether they're a member of a sports team and things like that from a background survey that we have. And then we can restrict to students with no online presence. So there are a lot of first years in the study who they just don't have, like you cannot find them online. We tried, or they have a very common name. And, and it's, you know, unclear who the person is. So when we restrict analysis to those students who have no online presence, we find the same results, which is comforting that it's not like this other stuff that's driving everything. And then when we restrict analysis to students, we throughout restrict analysis to students whose names clearly convey their, their actual gender. So there are lots of, not lots, but of the 100 students, we end up with 76 students who have non-gender ambiguous names or whose names are like consistent with what they predict in terms of their gender. So for example, like a Cody who's a girl would be dropped because if somebody, it's not clear what the professional would know about the gender of that student. So if they look them up, they'll see a picture. If they don't look them up, they'll assume it's male. So, so we, we are only using names that strongly signal a gender, and it's actually the gender of the student. Okay. The, the other benefit of using real students is, like I said, we can, we can ask students who they want to talk to and see whether gender differences in how professionals respond to messages are there also when we, when we restrict to students, to professionals whose students want to talk to. Okay. Any questions so far? So, so there, we went deep into the pool of people in these, in these fields who live in the city that the students were, so we have 10,000 people who are in these jobs and they're about 10,000 people in these, so that have profiles on this site that show up when we search for them. So that's kind of how they're chosen, like whether we could see their profile and, and we couldn't have done many more than 10,000, we kind of ran out. So, so there are people who list themselves as management consultants working in finance or data science or law who live in the area that the students are asking questions in. But I will say that because of the way that the professional networking program works, they are more likely to be alumni probably then, then otherwise of the university where the students are doing the study. Okay. So, I'll say that there is another way that we could have done this, which is for lawyers, you can see a lot of times email addresses. And that's just not true for a lot of these other jobs. So we can't find the list, like a list of emails for all, you know, investment bankers in Chicago, it's just, oh, sorry, I should have said Chicago in the city that the study is happening and it's challenging. So, so that was the, like a way that we could contact a lot of people. But an alternative would have been to, for some field, for some jobs, I think it would have been possible to just go to the websites of large law firms, for example, and contact everybody on them. Okay. Then when we, like, if there are no more questions, I'll go on to talking about the results. So one important question you may have is, like, do people actually respond to these messages? Yes, students receive between 4 and 21 responses to 100 messages that they send. So it's not huge, but it's about similar to like the callback rate when we do these audit studies. And correspondence studies. And, you know, a lot of, there was a lot more response among alumni of the university actually. So I'll say that there is like a sense in which it's a bit strange to be receiving a random message from somebody who have no connection to, but it's extremely not strange to be receiving such a message from an alumni of your university. And among the, like when we study that group only, we see larger patterns, like more exaggerated information differentiation by gender than in the broader population. Because maybe they're actually more thinking about what they want to say and more engaged with the students in general. The overall response rate is 12%, a little bit less to the broad question. When we see responses to that question, they're typically longer, like it's an open-ended question. So maybe more burdensome to answer. The specific questions had a little bit higher response rates. But, you know, we're not breaking the 15% response rate barrier on average for any question. When we study, we'll talk about the broad question first. So students asking, I'm interested in management consulting. Can you tell me about the advantages and challenges of this career? First, you might be concerned with analyzing, like, what's inside the responses if there's really different response rates by student gender. So, you know, maybe the people who are answering male students are the ones who really want to talk to anyone. And then after that, we get a different group of people responding even more to female students. And it's hard to interpret what, whether it's selection that's driving the estimates or whether it's driving the differences or whether it's like actual differences in the content for a given professional. So what's comforting is that there don't seem to be differences in the response rates to male and female students. You know, there's a one percentage point difference in the response rate, like slightly higher for female students. If we, like, include all messages, including the ones who didn't respond to results, don't change. So these response rates difference aren't driving it. What's also interesting and what people really want to know is, like, you know, are female students only getting responses from female professionals? And that's not true. Actually, there's really not statistically significant gender differences in any of these characteristics of who responds to male students compared to who responds to female students, especially not gender actually of the professionals. And maybe you see something a little bit in the alumni, the propensity of alumni is higher to respond to male students, but it's not statistically significant, but it's a reasonably sized difference. So, you know, it's not the case that female students are only hearing from female professionals and male students are only hearing from male professionals. They're pretty much hearing from the same people, but the things that they're hearing are different. So female students, like you said, get more than twice as many. So the interpretation of this, a male student hears about work-life balance, 6.7% of the time that he gets a response, and a female student hears about work-life balance, an additional 8.7% of the time, and that's statistically significant and different from zero. The columns one, two, and three are different from each other because they're adding more controls. So column one doesn't add any controls for student race or ethnicity. Actually, when you do that, the differences by gender go down a little bit, but not different from one another in a meaningful way. And then when you add all the controls that we can think of kind of for students like their GPA, stuff that the professionals could even see, we don't see much difference from the original estimate. There's still a huge difference in the propensity of professionals to mention work-life balance to female students compared to male students. And if you're interested, I can click on this heterogeneity slide. So one thing that you might ask is are professionals mentioning work-life balance to male versus female students at different rates based on their own gender? Actually, if anything, female professionals are mentioning work-life balance more to female students compared to male students, but also male professionals do it. These two aren't statistic. I mean, there's a big difference, but there's few messages. They're not statistically different from one another. And when we look at other things like whether they respond to the work-life balance question, there's really no difference between male and female professionals on those margin. Like I said, one thing that you do see is different from one another is whether alumni respond differently than non-alumni. Alumni are more likely to gender differentiate in their mentions of work-life balance to female students. They're actually driving the whole estimate. Both do it in the responses to specific work-life balance. So when we look at the pattern in this broad question, we see that professionals are bringing up work-life balance more to female students than male students. And why might that be? That might be because they think female students care about work-life balance more. So maybe this is just like they think that they're trying to do their best job possible. And I don't think that that's really different than their motivation. They're trying to do the best job possible to answer the question for the student. But what's surprising is that there's still this difference in how professionals respond to male and female students when we look at specific questions about work-life balance. So when students ask, I'm interested in management consulting, but I'm a little bit concerned about work-life balance, do you think this is a valid concern? Female students receive almost 30% more responses to that question than male students. Maybe when male students ask, professionals think, oh, that's not the type of person that should have the job. And they don't engage with the students. When female students ask, they do engage with them. Given this difference, again, in response rates, it's hard to analyze the content. But it's not clear that there's a difference in the intensive margin. So they say about the same thing once they answer the question to male and female students, but they're just much less likely to answer the questions for male students. So here's an example of a piece of some of the responses. So in the broad question, a lawyer might say a career in law opens many doors, but it also offers long hours, hard work, firm deadlines, many challenges. In finance, they say the challenges can be the hours depending on the area. And this I point out because it was a common theme, especially for female students to hear about depending on what you do. So depending on the career path you take, this is a concern. And that caveat is there in a lot of the fields, including law and, well, law finance. So it's not usually that they say, it's horrible all around. There's always kind of like a path, but they are much more likely to bring this up to female students in the broad question. And when they're responding to the work life balance question, here's an example of some of the responses. It's definitely a valid concern at a large law firm, your schedule will be outside of your control. You won't have any evenings, weekends, vacations. And house is usually better, but it's still very demanding. Management consulting, yes, 68 hours of work a week, little predictability, weekends are usually open, though. And I'll say that these are like snid bits of responses. We had a lot of responses that went on for like pages and pages. So very informative, very detailed, you know, and like, people are trying to help students and a lot of the students found the responses like extremely helpful. So, and followed up with professionals after our study was over. So we, you know, if you're not comfortable with these anecdotes that I just gave, I'll share also that that we had students evaluate kind of more objectively, not the students receiving the messages, but an outside set of students evaluate the content and students evaluations suggest that, you know, only 3% of the mentions of work life balance in the broad question make students less concerned. You know, 22% of them are consistent with what they would have expected anyway. Like that would be something like, you know, oh, you know, work life balance is tough, you know, maybe that is very vague in general, but a lot of the responses were specific and detailed that this was like a real problem. And those 76% of messages were reported to make students more concerned about work life balance. And student evaluations and responses to the specific question were less extreme, but 51. So most of the messages made students more concerned, but also almost 20% made them less concerned, because like I said, a lot of those responses included caveats, like, you know, you can find ways to make it work. What's interesting is that when we, when we think about this other question that we had workplace culture, we don't see any gender differences. And this was kind of surprising, I think, to a lot of people because there's a big emphasis on female students or women being versed to competitive jobs and competitive pay schemes and competitive jobs in general. So we asked specifically about competitive culture in the workplace. And actually no almost no one mentioned that in the broad questions about these jobs. There were a couple of times in like three times, I think that people asked specifically brought up, you know, the culture was terrible, you know, people were trying to mess up my code when I left the desk or something like that. That was extremely uncommon. So then we broaden our definition of culture to like any mention of culture at all. And we don't see any gender differences in that it was very rare, I think there were like five messages that brought up gender specific issues at all. So so they're not talking about, you know, sexual harassment in the workplace or anything in these messages. And maybe that's because we're only analyzing the first message. So I think that that's like a caveat to this whole thing. Maybe a lot of the reason that we don't see anyone talking about culture or differentiating by gender, any of their responses is that is that these are only the first messages that we're looking at. And that might not be something that you write, you know, on an online professional networking platform to people, you don't record those kinds of things and you don't talk about them to strangers. So it's very uncommon. We don't see to mention competitive culture or to mention female specific issues. We talk about culture very broadly, you know, it's more common like 12.8% of messages, talk about that but not differentially between male and female students. And we don't see any difference in response rates to male and female students when they ask about competitive culture. And I'll say that there most of the answers were like, no, that's not true. So so they disagreed with that characterization of their job for the most part. And we're encouraging people that you shouldn't be concerned about it. And in the factual question, so this was only asked to lawyers, a subset of the lawyers. The question was what are the minimum billable hours requirements and bill big law. And a lot of the responses that go to this website, you can look it up. That's true. There's like an answer to this question and you can look it up. This is like a contractual agreement. But many of the responses didn't. And one other thing that I'll say is female students were more substantially more likely to receive a response almost twice as likely. But like I said, we're underpowered. So these aren't statistically significant once we add controls. And I shouldn't talk about it, but I find it interesting and thought-provoking. So we had about a 10% response rate to this question overall. You can see that we asked only 300 people. So we have about 30 messages that I'm talking about now. Everything's super underpowered. But in those 30 messages, it's interesting that female students were told like a standard deviation higher hours. And it was significantly different depending on the controls that you include. But it's only 30. So I think that we'd have to go back and do this to really make the case for that. But what we see there just descriptively is that there are differences in the responses that they give also about the hours on the intensive margin. And then does this information matter for career choice? So this part I think is very suggestive. And like I said, the experiment wasn't on the students. So there's not randomization of this stuff. Female students may in general be dissuaded from careers. And that could explain some of what we see. Male students, actually, none of the male students were dissuaded. So hint that that is true. Female students are maybe more sensitive in general to any information. But what we see is that that difference is explained a lot by whether the student actually received a work-life balance mention or received a response to the specific work-life balance question. So the propensity of female students to be dissuaded from their preferred career falls a lot when we include these variables. That's not true when we look at another variable like workplace culture. And when we add a lot of message controls, it stays the same. So this is very suggestive, but it suggests that work-life balance does impact students' interest in their career. So those female students who didn't hear about work-life balance, usually in a negative light, were not less likely to be dissuaded than those who did. And what's really interesting is that among students at this elite university who are majoring in economics, which is not an easy major, they're not interested in talking, female students are not interested in talking about work-life balance with professionals. So it's not the case that the information that they were receiving was actually catering to what they wanted to talk about. So we asked not the students in the study, but similar students majoring in the same thing at the university, the same university in a brief survey, just like if you had 15 minutes with a professional, how would you allocate your time across these topics? What would you want to talk about? And what we see is that looking at the top row female and work-life balance, the teal category, it's almost the last thing. It's the second to last thing that they want to talk about in terms of time allocation. Male students actually want to talk about a lot more. So they want to spend 14% of their time on it. It's like, I think the fourth most popular kind of topic among male students. So this could be because female students have heard a lot about it. They feel like they know maybe they're always kind of warned about work-life balance. And so they don't feel like they need more information on this. And male students, you know, even though they seem to want information on work-life balance, they're not getting it either when they ask broadly about the career or when they specifically ask, you know, I'm interested in learning about work-life balance. Can you tell me about it? They're not getting responses to those questions at the same rate that female students are. So just to summarize, overall professionals emphasize work-life balance substantially more to female than male students. Female students get twice as many mentions in the broad question and almost 30% more responses when they specifically ask about work-life balance. We don't see anything for workplace culture, you know, either specific competitive culture question or broadly, do they mention anything about culture in the broad question? We don't see any gender differences. We have some suggestive, very suggestive information that this information actually matters and affects people's perceptions of, you know, whether they find their career still attractive. And I will say also that in a lot of other studies, you find that students' career choices are very sensitive, like actual choices are very sensitive to information provision. So there are studies that suggest that, you know, when you have and have been confirmed that when you have a, you know, engaging speaker come up and talk about their job to like a room full of economics majors, that speaker is female and she has a cool job, you get a huge number more female students majoring in economics. So, and like having long run different, I think, life outcomes. So it's surprising, I think, how easy it is to change students' choices about their major. And in that context, maybe, you know, it's not surprising that even these few messages seem to affect their choices a little bit. I think that what we haven't done is show that if you, you know, could study all of the informal conversations that people have and could record that and see what they're talking about with that impact, would there be a gender difference in those? This is the one kind of conversation that we could get a handle on, you know, controlling for all the characteristics and really say like there seems to be a causal impact of gender in these conversations or perceptions of gender. And we think that it probably extrapolates, but that's what we're missing, like we don't have data on all of these informal conversations and whether they affect people's choices. My guess is that they probably do even more so than the formal stuff. So why do professionals emphasize work-life balance more? We're all to female students. One hypothesis that you might have had is just in general, they want to talk to female students more. So they, female students hear more about everything because professionals like to talk to them. So we don't see that we don't see that female students get longer messages and they also don't get differential responses compared to male students on anything that isn't about work-life balance. So they do get more messages when they ask about work-life balance or when they ask the factual question about billable hours, but they don't get more messages in other questions that aren't about work-life balance specifically. And then another possibility is that female students and male students have different concerns. They believe that female students want to hear about work-life balance and they're just trying to tell them what they want to know about even if they're wrong. So we see that actually male students want to talk more about work-life balance, like I said. So this is an incorrect prior, but maybe this is what's explaining the pattern. So we don't think that's the full explanation because when the student says, I want to talk about work-life balance, can we talk about it? There's still different engagement with male and female students on the topic. You can say, oh, maybe there's a better way that they could have said that and then it would have eliminated the difference, but I think that suggests that there's maybe a problem that it's hard to know for students what they can say to eliminate these gender disparities. So they specifically say, I want to talk about this and it doesn't equalize what they hear. There could be some way of equalizing it, but I think that we don't know what that is and students are unlikely to know what that is. By the way, I'll say students actually don't know, like 50% of students don't know or don't believe that there is such a gender difference. So I think it's unlikely that they're able to kind of correct for this behavior. They also, professionals may believe that there is a gender-specific true answer to the question. So the reason that they respond differently to work-life balance questions from female students is because they know that it's a big deal for female students but not for male students, so they aren't so worried about responding to those questions. So that's not true in the factual question. There is the same answer for everyone, but this is kind of a gray cross out. It should be gray. We find some suggestive evidence that even in the factual question, there's a difference in the response rates to male and female students and also in the content of the answer. So overall, it seems like for whatever reason, professionals think that work-life balance is more important to discuss with female students than male students, even though male students want to discuss it more and even when they specifically ask about it. And then one thing that we can ask is like, are students able to correct this by just like talking to the right people? So are there professionals out there that don't do this and are those the ones that everybody's going to? Or conversely, you know, if female students really want this information, are they going to people who emphasize it more? So when we incorporate student preferences for professionals, we actually don't see any difference in the results. So the professionals that students most want to talk to are similarly differentiating information by gender compared to the professionals the students don't want to talk to as much. So I won't show those slides, but that's true. So overall, we studied, you know, does student gender affect the information that they receive about careers? We found that responses do seem to differ by student gender in particular when we're thinking about work-life balance questions. And, you know, the methodological innovations are kind of to incorporate real people rather than using fake profiles and like chatbots or something to ask these questions. We respect professionals' time. They're answering real people who are interested in the answer. And we can look at, you know, what student preferences are across professionals and how they affect our conclusions. Why does information matter? Why does the gender differences matter outside of dissuasion? I think that a potential consideration is people don't know what they don't know. So our focus was really on work-life balance and workplace culture. Those are big things that people do think about, I think, in general when they're thinking about jobs, but imagine for this thought exercise because I think this extends more broadly than just those topics. The thought exercise, you know, male students have never thought about work-life balance before because nobody's ever talking about it with them or sharing with them that it's like an issue. And by the way, some of the responses that we saw on that it's not specifically about having children like in management consulting, this work-life balance stuff, like people describe it as alienating from your friends because they can't understand what you do, they don't understand like that you're like literally not home more than a few days a week and stuff like that. So the responses are about, I would say, deeper things than like, you know, you have to travel a lot that's hard. And, you know, male students never think about that as a career attribute. They get to the job and they're like, oh, I'm learning about this now that I'm here and if somebody had told me about it, I would have made potentially a different choice because it was important to me. They never thought about it because nobody brought it up. So I think that that's and that's hard to fix in the same way that you would fix like if you know that work-life balance is an issue and you're a female student and you know that people like exaggerate this difference, you can kind of correct for that in your head because you're like, okay, I know they're talking about, you know, 80 hours a week, whatever that probably means 60, people like to tell me that things are harder than they are and you can adjust for, you know, what you think is the truth if you know that there's bias. That's different if you don't know about the attribute in the first place. So I think that that's a concern that my information is kind of special. And then, and then female students, you know, are seem to be dissuaded from their preferred career, partially explained by this emphasis on work-life balance. And the example in which I think that this is clearly not optimal is thinking about the management consulting case. So those who don't know management consulting is a job. And as they say in the messages that you have usually for like two years or three years, and then you find another company that you want to work for, like it's a meet and greet with a bunch of different companies. And you get to know a lot of people. It's like a great, you know, first introduction to the business world or whatever, but nobody keeps very few people keep that job for longer than a few years. There's a huge turnover and that's like expected and normal. So you might say, you know, these 30 year old management consultants that we're contacting with questions about management consulting are like very selected. They stayed in the job until they were 30 years old. And maybe they're observing a lot of tensions with family life and having that job. And if they're telling female students, you know, about that because they want to give them the full truth about the job, like what does that have to do with with anything kind of like these students are almost certainly not going to keep that job until they're 30. And, and they are not concerned about like, you know, incorporating family life with, with their career when they're 22 years old, for the most part. So I would say that's broadly a characterization of how people do the management consulting career, you know, they're not thinking about having children when they're young, they're like investing heavily in the job. And these, and we see that among students, they don't want to talk about work life balance in this, in this group, they don't care about it. So it's strange to emphasize it, even if it is true that there's like this, you know, really difficult thing that you're dealing with when you're 30, 35 years old. What that has to do with the decision of a 22 year old is unclear for this job in particular, that you're not going to stay in past a few years, for the most part. So anyway, that's just my spiel on, on the end of my consulting in particular, I think that's less true for other jobs. So, so I can pause there. Does anybody have any questions? That's my kind of summary of our work on whether people talk the same way to male and female students about jobs. Like I said, I think it's much bigger kind of than, than the setting we were able to isolate it in, but, but that's our setting people asking on a professional networking platform about career information. Yep. So I have more slides though, but okay, yeah. Our Zoom audience can, you know, type questions in. So my question is, you know, as a professional who gets emails about advice, like what is your advice for the professional who's giving the advice? Yeah, if I got like a broad question or a work-life balance question, or like a specific, I can't remember what the specific like truth or factual question was. Yeah. So I think it's hard to answer, but I thought I've introspected a lot, and I think that's all I can tell you is my introspection. So my overall takeaway is that like people give advice pretty thoughtlessly. And I think in a way that's not appreciated by the advice seekers. So when people, when undergrads come to talk to me about, you know, what they should, what classes to take, or like whether they should go to grad school, I think about it a little bit in an individualized way. But for the most part, I have some general thoughts about that. And I give them my general thoughts, and I spend just very little mental energy on that conversation, because there's a lot of them, my thoughts of other things to do, you know, this is just, you know, easy to give this kind of advice. And I think for the student, it actually really matters. Like if I reflect, I had a few conversations like that in my life, and they were super important in determining what I did. So there's like this really big disparity, I think, and how important it is for the person that's getting the advice, and how much effort the person who's giving the advice puts in on average. I'm not saying everybody's like that, but I think that that is true for a lot of people. And I think that that means, you know, and I've talked to other faculty about this, they say like, you know, now that I think about it, you know, maybe when I talk to a female student, I tend to tell them something like, you know, you don't have to do like, you know, econ, which is very mathy, you can do like public policy, PhD, you know. So like these subtle things that for whatever reason, like you're saying that, and you haven't thought about it deeply. And then when you reflect on it, you're like, why did I say that to them? I just kind of thought like that fit for them. We know kind of that people have these perceptions of what, what gender differences are in terms of preferences. So, you know, female women like caring jobs, or they like, you know, things with a lot of connection to, to, you know, doctor work and patients and stuff like that. And those differences by gender are very exaggerated compared to what the gender differences actually are. So I think that people have some things in their mind. And we know that from like psychology studies that the gender differences that people believe are there are larger, much larger than what's actually there. People exaggerate these disparity or differences in personalities. And when you're trying to give like personalized advice, you also might rely on a lot of those kind of, you know, exaggerated beliefs. So I think that that does go on, I think in the background of a lot of people's minds, and especially is exaggerated when they don't think that much about the advice that they're giving. So I guess my only advice to the advice givers is that, you know, if you reflect on it, like why am I telling this person this thing? Is it because they really like, you know, would I be telling them that if they were looking a little bit different? Or if I had, like why am I basing this idea that they should go into public policy rather than economics? No, in our example, I don't know what that would be. There's a question from the Dean of a medical school, Dr. Aurora, should we avoid mentioning work-life balance? It seems like there were concerns when it was mentioned, which gave me pause. Thanks for such a thought, for such thought-provoking data. Yeah, so I don't know. I think that it really depends on, okay, I'll just say in my experience doing grad school, like before me two times, right before me two times, there's like no talk about this stuff at all. There's no emphasis on, you know, gender specific differences and people's life experiences or experiences in the career. It was very uncommon. And maybe it was just the place I went, but I think it was ubiquitous. And then, you know, me too happened in the econ had like a specific paper that was on like this chat board, econ job rumors that that was very popular and highlighted the kind of misogyny on that board, which, you know, was populated by a bunch of professors, maybe lots of grad students for sure. And, and, you know, we have this big culture problem in economics and that became like very salient. And now if you talk to grad students, like, what are they have a lot of meetings, they have a lot of like, you know, events for female students, the events are like talking about culture extensively, they're talking about work life balance extensively. And my only question, my only caveat is like, there are lots of good things about the job too. And are we even talking about those more with female students, I'm not sure, like, it's very, very, very, I think disproportionately focusing on these issues, which we should talk about. But again, sometimes, you know, you're crowding out other important aspects of the job. So I think it depends on the setting a lot, like I think everything it depends. But, but I think that there's, it may be true that you want to talk about work life balance, but you also should keep in mind that when you do that you project, you know, some negative perceptions about the job, and maybe you could balance that out with what's good about the job and why might you not only want to focus on the negatives when people are making decisions, because if they only hear these negatives, then overall it's like a negative perception that they have. Good. I think there are villains in the room. So I think we'll stop the formal part of it and then let you guys come down and ask questions more personally. So thank you so much for a great talk. It was tons of good data. I was fascinated. I think our audience was too. Yeah. So let me, I'm going to stop the share and then I'm going to mute and ask for an honest