 Bonjour, mon égent Tirol et je suis très ravi de vous accueillir, chers étudiants, chers collègues, chers lycéens et professeurs des lycées fermants, rive gauche aux années Saint-Joseph Lassalle. Avant que je présente Marianne, je veux dire qu'il y a des écouteurs, des casques si vous voulez en profiter. Après la conférence de Marianne, il y aura des questions-réponses et des micros se baladeront dans les allées donc si vous voulez poser des questions, ce sera le bon moment. Jean-Jacques Laffont Price est awardé chaque année. L'économie internationale a fait une contribution à la recherche de théorique et hôpico. Le prix de cette année va à Marianne Bertrand, professeur de l'économie de l'Université de Chicago à la Buse School of Business. Marianne est une économiste appliquée avec l'intérêt de l'économie laborale, de la finance corporelle et de la développement. Elle a eu une carrière stélaire qui a étendu un état d'extrême large l'université de la recherche. Mais elle a lancé consistant, elle a focussé sur des primes qui impactent directement les commandes communes, qui targetent les injustices sociales et les policiers pour les solutions pratiques. Elle a travaillé sur diverses formes de discrimination, ratio, gender, etc. Et divers contextes, la judiciaire, l'éducation, les promotions, et aussi l'involvement, la conscience et la discrimination intérieure. L'économie et la société en général sont en besoin d'une talentueuse de rigoureuses et originaux, comme Marianne. Elle a été un model de rôle, un mentor et un champion pour les jeunes femmes. Nous espérons que nous voulons entrer dans cet état. Elle adopte des insights de comportements sociologiques et psychologiques. Elle aussi s'étend. Elle a montré, par exemple, que l'advertissement de l'éducation d'éducation s'appelait plus d'intrusion que de raisons. C'est une chose d'intrusion pour cela. C'est une autre chose de démonstration. Marianne est parfaitement à la maison dans ce qu'elle raconte dans les économies du jour d'aujourd'hui, qui favorisent le réalisme et les contextes d'un grand série et des modèles mathématiques, qui sont bien aussi, mais... Elle s'exhale à ça. Cette approche des économies mathématiques a été essentielle à notre carrière de recherche. Muchoche qui a été évoquée à mesurer les inequalities, à comprendre leur cause et à étudier les solutions, à aider les recherches et les policiers à s'occuper de leurs efforts, ce qui sera plus efficace en fermant la gaffe. Pour mesurer les équities raciales dans le marché du laboratoire du U.S., elle utilise un expérience de field ingenieux pour démonstrer les employés du travail du origine africain américain, je veux dire les noms des noms africains américains, je veux dire, elle a reçu 50% de quelques callbacks pour les interviews et les noms de l'exercice. Ce travail a depuis été élevé dans beaucoup de pays, et pourtant aussi en France avec des noms africains-africains. Elle a travaillé beaucoup sur la glace, et en fait, elle parlerait de la genderé et de l'équalité aujourd'hui. Comme vous savez, les femmes ont toujours un temps difficile d'acheter les positions d'exercice dans les corporations et les secteurs publics. Il y a, bien sûr, beaucoup de raisons pour cela, mais Marianne a été la leader en réfléchissant sur ses causes, gender gaps in education, psychological attributes, demand for flexible time allocation, outright discrimination against them, et l'efficacité of alternative policies, family-friendly work environment, gender-neutral parental leave, affirmative action. Elle a été évident from a 2003 low in Norway where the boats were mandated to have at least 40% of women that has been extended in many places in Europe. And with some exciting news or intriguing news, the women were hired afterwards to be on board, actually, were at least as qualified as a predecessors. So that's good news, but at the same time women who didn't make it on board actually didn't benefit from a spillover effect. Employers and the family are not, and the parents are not the only culprit, unfortunately. Marianne also has a paper which is fascinating and disturbing on the gender identity norm among partners. I guess you are going to talk about that as well, where men don't like women to make more money than they do. And that shows very clearly in the statistics and also in the career path of women and their labour market participation. Development, as you know, the Nobel Prize which will be awarded in a few days to Abidjid Banerj, yesterday Dufour and Michael Kramer recognizes their careful use of randomized controlled trials to pinpoint policy improvements to help the world's poorest. Marianne has also been a leading contributor in this field and she has worked extensively on discrimination and corruption in the U.S. and India, for example. She works as a co-chair of the Chicago Booth Restanding Center for Social Sector Innovation, as chair for the Poverty Lab at the University of Chicago Urban Labs and as a director on the board of directors of the J-PAL Poverty Action Lab. She has also worked on corporate finance and then I'll stop there because you want to listen to her and not to me. She has made major contribution to a study of corporate finance, COPE and incentives, for example. She has studied the adoption of takeover laws in various states in the U.S. And she has found out that majors may not be empire builders but rather be have a preference for a quiet life. So for example, if you are protected through takeover laws you close fewer firms but you also create fewer ones and also you pay your wages to your employees. She also works on U.S. firms and now they use tax exempt charitable donations to the big politicians. Corporate social responsibility is of course a big theme ERTSC. Actually on Friday we had a conference on this specific topic just last Friday and that's going to be one of the theme of the conference which is going to take place on Thursday and Friday at the TSC Sustainable Finance Centre. Marianne unfortunately won't be able to attend physically thanks to the event we know for Thursday but she will be there virtually so she will be taking part of the conference and talking about the role of corporate philanthropy on the political process. Finally, Marianne as you will expect has received many prestigious awards including the American Economic Association 2004 LN Bennett Research Prize and the Society of Labour Economies 2012 Rosen Prize. She says she doesn't like the weight of expectation that comes with receiving a prize. Sorry, Marianne. We are doing a bad trick to you but you have to get used to it because it's not the last one and the heavy bag of medals doesn't appear in any case to be slowing out down and you'll have to live with expectation. There's clearly much more to come from Marianne not least her lecture on gender inequality today I'm delighted to welcome her to Toulouse and deeply honoured to introduce her as this were this year's winners of the Jean-Jacques Lafon Prize. Thank you, Marianne. Thank you so much. Thank you so much. I want to introduce you. Great. Thank you so much for being here and thanks a lot Jean for the really, really kind introduction. So this is a very broad title and this is not working. I can use my fingers. Yes. Is there any way to get this to work maybe? Okay, great. So this is a broad title and I guess this is going to be fairly broad as well. My goal is really to just kind of walk you through through what I view as being kind of the most pressing remaining issues in terms of improving gender equality in the labour market. So I'm going to be wide more than deep on any single topic. Now before I get started it's really important to realize that we obviously live in a moment where concerns about rising income inequalities are everywhere you know in the US in Europe and if you look at the whole agenda surrounding issues of inequality gender provides the most optimistic lines. We've made tremendous progress in the developing world. There's really an echo. Do you hear it as well? Yeah. Should I try? Okay, should I try? It's really hard to read. All right. So there's going to be more pictures than words. All right, I'm going to keep on going. So I'm going to talk. There's going to be two parts of the lecture. I'm going to first take a worldwide perspective and focus on issues of differences in labour force participation between man and woman. That's really the main thing that we can study systematically throughout the world where finer data on income is just simply not available. But then I'm going to focus on the more developed world which you know will stand for the OECD and there I'm going to dig deeper into other factors than just labour force participation but also talk about issues related to pay and the general gap in earnings which remains quite high especially at the top of the income distribution. Now a recurring theme throughout most of the lecture is going to be that beliefs about generals are A or D key force that are still holding women back today. Even though as we move to different places the nature of these beliefs is going to be quite different from quite hostile to more benevolent in some context. Now before moving into the data I think it's worth to remember kind of why we care about these questions and from my perspective there's really two reasons as to why I care about this topic. The first one is an argument about fairness and justice kind of rolls you an argument that it seems that we would like to live in a society where everyone has got the same opportunity. So the same argument as to what we care about social mobility is an argument as to why I care about issues related to gender equality but there's another argument a more utilitarian argument which is that if you think about men and women being born the same distribution of talents it has to be that a society is not going to be at its frontier in terms of efficiency if it has so much in bounds between the genders in terms of being in the labour force and being successful in the labour force. So you can really both view this as like a question that is about justice and fairness but also a question about efficiency. Now in the background of all of this I think we live in a moment where at least I perceive a wave of renewed conservatism gender conservatism I think there's a public intellectual I don't know how prevalent he is in France he certainly is very much so in Scandinavia that's Jordan Peterson that he's been bringing back some essentially views about the difference between the gender kind of going back to the point that the reason why women are not succeeding as much as men in the labour force is because they're just different the idea that gender roles is not something socially constructed but something that is biologically determined So I'm also going to spend some time during the lecture to talk about this in a sense how much do we know about the extent of gender differences in trade, in psychological attributes in preferences and how much do we know about whether the beliefs we have about these gender differences are or are not accurate. So I'll do some of that and that's going to be in a sense more psychology than economics. So let me get going So hopefully these are going to be big enough So as I said I'm going to start with the world and this is kind of what the world looks like today what it looks like in terms of what we can really study in terms of women in the labour market which is just a rate of female labour force participation every point on this graph is one country in the world today big and small and what you can see there is that there is what is now well known between economic development which is on the X axis and the rate of labour force participation by women. You can see exactly the same picture if you don't simply look at female labour force participation but look at the ratio of female to male labour force participation. So what is behind this relationship I think the common interpretation of this relationship is that it has to do with structural change that when economies are poor they start with a very large agricultural sector think about these poor economies everyone works on the farm the separation between work and home is trivial and we have lots of women in labour force in this context because work and home are essentially the same place now as economies develop they typically first start to grow their manufacturing sector and then when they are even richer they become more focused on the service sector. Now the story about female labour force participation and structural change essentially when you start very poor it's all happening in the farm the woman is involved everybody is involved in fact we have lots of children, high fertility and every ounce is on deck now as economies become richer work moves towards the factories the factories are further away from the home it's harder for the mothers to free yourself from the children it's going to be more physically intensive and women don't have often the strength to do that and then finally as we move forward with this process of development we've moved to easier jobs jobs that do not require as much physical strength and we've also moved to societies where fertility rates have dropped and women can more easily free themselves from the home to start working. So that's the typical kind of explanation behind the U-shaped relationship now if you look at the data today there's really little evidence for that story what I'm plotting here is the relationship between female labour force participation changing female labour force participation over the last 25 years and changing the size of the agricultural sector changing female labour force participation and changing the size of the manufacturing sector changing female labour force participation and changing the size of the service sector you would expect to see patterns here in addition to the U-shaped relationship in fact you essentially see a lot of flat lines the story doesn't seem to match the data today now there are lots of reasons as to why you may not match the data today that it may have been a very good story for the past but not for the present why because things have changed so first one thing that may have helped women back before in terms of moving towards manufacturing is that they were very poorly educated and we've made a huge amount of progress in educating girls in the U.S. and work like the one that Esther and Abidjid have been doing has been an important part of this story so there's been improvement in female ligations and maybe that has changed the dynamic that existed in the past it's also true that fertility has been declining throughout the world so women may have an easier time now kind of like moving straight out of agriculture into manufacturing because the size of the home and the amount of home production these are very possible explanations now what I'm going to do instead of those explanations as to why the U-shape doesn't fit the story for the U-shape doesn't fit is just try to draw to your attention to I think one important underlying forces behind the U-shape and I'm going to do that by adding another set of data to the World Bank data I've used so far it's data that we have on social media so in particular there's a data set that collects for not every countries in the world but quite a few countries a set of measures of social attitudes in particular it has questions about gender attitudes now one such question which I'm going to use multiple times throughout the lecture is this one people are asked in the survey representative samples of individuals are asked whether when jobs are scarce men have more of a right to these jobs than women do what I'm going to do is classify the world in terms of what share of respondents would answer yes to this question and we do that very roughly by kind of isolated countries where less than 25% of people would agree such as France from countries where between 25% of people would agree and countries where more than 50% of people would answer yes to this question and if I do that well this is kind of what you observe one is before the three categories is simply looking at the fraction of people that would say yes to this question the higher this fraction the more sexist the places and you can see that there's a clear kind of leadership here that places that are more sexist have lower rates of female labor force participation but more importantly for my point is this this is kind of what's behind that you shape that I started with and I separate these countries into the three groups that I discussed before so what it looks like is just not so much of a U shape but more like three straight lines with one line that's well below the other ones, the blue lines the high sexist country lines and behind this is kind of why we see the U shape the U shape relationship so it's not when you look at the world this way it looks like one driving forces of why labor force participation is still so low in some countries has got to do with these social norms so what's also true about these places is that those more sexist places not only do they start with much lower rate of labor force participation but the extent to which labor force participation by women has been going up has been much smaller in those places than everywhere else in the world by the way we can see hidden behind this picture is that you can see that every dot is essentially above zero so there's really an optimistic message there which is that essentially throughout the world labor force participation by women is higher today than it was 25 years ago but the extent to which we've made progress has been much much lower in those countries that already started with lower rates of female labor force participation so I think I do a lot of micro work this is macro data and I'm not comfortable with it and there's a lot of other explanation what may have behind these blue dots and why these places are different behind norms I think another way to kind of see this is to focus on a few examples and one example that I find particularly important is the one of India because India is a third of the world population may live in India or a quarter and if you look at India on this graph India is actually one of the very rare country where female labor force participation has been on the decline it is going down it is lower today than it was 10 years ago so there's a way to look at India and really again show the importance of the social norms so why is it nice to go within a country go within a country where the environment is given where the schooling is given the return to schooling is given the industrial structure is given now how can I see differences in norms within a place like India I think there's two ways you can easily do that one is to look at religious groups and one is to look at castes within the hidden news so this is what India looks like this is data that only covers the last five years I believe and whatever what you have on the y-axis is female labor force participation and it is as a function of the income of the household the per capita income of the household zeroing in any labor force participation of the woman and then you have three lines one that separates Muslims one for Muslims, one for Hindus and one for Christians and you can see that this is the same country the same economy, the same forces but you have substantial differences in female labor force participation between these three groups which is the more conservative gender conservatives are going to be the Muslim and the Hindus and the Christians are you know somewhat less so you can also see these pictures strikingly if you just look at the Hindus and separate different castes so the strengths of the gender norms are going to be typically much stronger among upper caste individuals than lower caste individuals so what I did there is just taking a different shape between female labor force participation and log household income per capita but separating starting from the the highest status, upper caste of the backward caste, schedule caste and schedule tribe and these lines essentially kind of stack on top of each other the way you would expect so it's pretty remarkable than among upper caste Hindus whatever the household income is female labor force participation between these two groups never gets much beyond 16% so I think visually this tells you that in a place like India these social norms are really really an important force now what I want to do is just say a few things about policy in places like that so what how do you get change how do you change this and how do you change these norms and I think one thing that I feel I have learned from reading the economic history literature is that you need big shocks for change so Claudia Gordon who is one of my hero has written a lot about the transformation of work for women people like Rickle Fernandez have also made really important contribution and one thing that you learn from their work is that an advance such as World War II was really important in moving the US economy at least kind of forward in terms of more equality in the workplace men were sent to war that mechanically moved women into work we needed women to fill in these jobs and that really was important I think it changes the political economy it changes the norms we have evidence that men that had a mother that worked during the war were also more likely to have a working wife because they were exposed to women that worked and got comfortable with it so if you take that logic about the World War II talk and think about applying it to places like India and some of the other blue countries I discussed before there is really an important sense that we need job creation in those places and you know what I try to do over there and I'm sorry again if the visuals are not so great but I try to kind of look over a 10 year period at the relationship between the change in the man's unemployment rate and the ratio of female to male in the workforce again separating my low sexism mid sexism countries from the high sexism countries and what you can see is that the slope of this relationship is different in the high sexism place what it looks like is that in those high sexism places whenever the unemployment rate for men goes up women make less headway in terms of further equalizing the rate of labor force participation with that of men that's I think one visual way you can think about the importance of creating jobs it's only when there will be jobs that a lot of the reforms that may be needed to make the economy more amenable the woman will happen absent that I think it's hard to imagine how you get this place rolling absent putting in place maybe some quotas but I don't think we know much about how quotas would work for women in places there's so much resistance to having women in the workforce I mean another way to say all this again is to talk about India which I think is fascinating because of the decline in female labor force participation and if you read about what's happening in India over the last kind of decade or so there's a lot of discussion of jobless growth in India it's not so clear why there's some signs that automation is happening faster than people expected there's a sense that India may not be as competitive with like even poor countries like Vietnam and may not be attracting much of the manufacturing from the developed world there's also some discussion that India's growth is actually not as high as it claims so people are making up statistics whichever story it is India is not creating job and literally when jobs are scarce men come first and you can see this strikingly figure, this is looking by education level one least educated seven more educated the unemployment rate of men and women in India the female unemployment rate I remember very few women are in the workforce so it's like less than 20% of the population is in the workforce and the females are looking for jobs cannot find them the female employment rate is five fold the men unemployment rate especially among the heavily educated now move on there's also some research that suggests well maybe I'm being too pessimistic maybe these places can change without kind of drastic job creation event that just described so here's an example of one paper a very recent paper that I like it's a paper that was done in Saudi Arabia which falls very much in the blue line it's about like India in terms of female labor force participation so what did they do in this paper well they documented a phenomenon that people would call pleuristic ignorance so what is pleuristic ignorance well it's a situation where people privately reject a particular norm but they believe incorrectly that others accept this norm and because they don't want to feel stigmatized they also end up following the norm themselves ok so this is first on the left side this is a survey this is a graph that tabs white we'll take a bunch of men in Saudi Arabia which are typically the guardians of their wife and then ask them do you think it is ok for wife to work outside of the home ok if I ask a large group of men I can do well this is a percentage of men in Saudi Arabia that agree with the idea it's ok for wife to work outside of the home and then you can ask them to guess what fraction of men in the room would think it's ok for your wife to work outside of the home and then you can look at the difference between those two right the share that I guess the share of men I guess would agree the actual share that would say it's ok and this is what the first graph tabulates and what it shows you is that most men are wrong most men believe that a much higher share of men in the society would be disagreeing with the view of a working woman inside the home they put it this way sounds like lots of men are ok with their wife working outside of the home but they believe that their neighbors are not ok with it because of that maybe their wife don't work so what the second graph indicates is that it sounds like simple information intervention can actually be successful in changing behavior what if I look at these men that have these incorrect beliefs about how many of their male neighbors disagree with the idea of having their wife working outside the home and I simply tell them well you know what you thought the share was 60% but in fact it's only 20% just give them this information and then kind of track some behaviors which they do here by looking at whether or not these men sign up their wives for some kind of job agency and the evidence suggests again the group that's relevant is the group over here where the wedge is negative that when you tell these guys your beliefs went correct they start changing their behavior so whether beyond signing up your wife or job agency whether he says any impact really impacted in the long term that's hard to determine I think there's going to be much more work being needed but what's nice about something like this is that it's very easy it's very cheap it's just an information intervention and it's hard to see what kind of adverse consequences it would have so more information I think ok so before I move to the developed world which you know will take me more time I just want to take a quick detour and talk about life satisfaction and I'm going to show you the connection of this to my first part very quickly so this is another thing that we can study throughout the world and what you're seeing here is essentially I've population weighted the country so you can see where India and China are I think that's all I can guess and what I report on the y-axis is an index of life satisfaction for men and women so representative samples of individuals are being asked how good do you feel about your life at this point in time and this is data that covers about the last 10 years and there are really two things you can see here Angus Ditten has shown the same data site that there is a very strong positive relationship between level of economic development and level of life satisfaction in the data so again what does it mean and it's not obvious it should have happened this way but it means that people over there in Togo when they think about their happiness they are thinking global they are not just looking at their society they understand that there are people up there in Denmark and things are better there so higher income means higher level of life satisfaction now if you look at the gender gap in life satisfaction it is very small you can infer that from the size of the difference between the two lines and the slope of the line and accepting very poor places look like women report higher level of life satisfaction than men you see about the same if you look at expectation of how good life will be in the future except that the slope of the line is less steady but again women look more optimistic about their life in the next five years now the reason I'm bringing this forward despite just the interest of looking at this data is that if you now go back to my three sets of countries mid sexism, high sexism, low sexism and you compare the gender gap in life satisfaction in these three sets of countries what you find is that women report higher level of life satisfaction compared to men in the more sexist places it's exactly the opposite of what I presume most of you must have had in the back of your mind before I got into this data you can do this by just using the three sexism measures I described you can also do this by trying to isolate parts of the world where women labour force participation controlling for lots of other stuff which is what I do in the bottom panel you see the same story so it's exactly the opposite of what I think most of us would have expected now I don't want to do too much of this but I think it's still an important fact so first it's reminiscent of things that we've seen in the developed world where we can do this kind of analysis within countries over time and where there's evidence that female life satisfaction is being going down it's an absolute term and in a relative term compared to men now that's the interpretation of this fact is tricky and there are many and I'm not going to be able to show them out so one is like when they ask these questions compared to whom are they comparing themselves to women in Denmark are they comparing themselves to women in the US are they comparing themselves to other women within the countries or other men it's not clear so who the comparison group is might be important what's also important to realize is that these measures of less satisfaction even though I think there's a lot of enthusiasm about them and how they could be placed on measures of GDP they are very imperfect in capturing the thing that matters in life they don't capture things like agency and control and maybe that's the important part that's missing it could very well be also that the fact that there's limited freedom for these women in those countries also does not allow them to be put into this box now as I said I don't have the explanation but I think it's still relevant when you think about the political economy of change in those countries the perception you may have in your back of your mind of like pantobes demand of women really wanting change because they know how much better the life could be if there were more including the workforce is not the situation that's on the ground right you could do the same and I'm kind of looking quickly through this by looking at something else which is experiential well-being so Danny Kahneman was really influential in helping us think about other ways to measure well-being besides asking people how happy are you which is essentially tell me about your day yesterday did you have positive feelings did you have negative feelings so we can use the same data and look at gender gap in life satisfaction positive emotions, negative emotions here strikingly I don't actually never seen this finding before women actually report lower level of experiential utility than men mainly because they report more negative effect than men do but again my goal is really more to tell you about how these varies with the sexism level and again in this case you find I think the non-intuitive results is that the gender gap in experiential utility is smaller, women are more relatively happier in the more sexist places than the less sexist same message I think it's important, I don't think it can be ignored and I think it's important as we're going to think about policy alright so that's kind of the first part of the lecture, what I want to do now is just talk about the rich world which in my context is going to be the OECD so I'm going to give you a few facts first show you what the world looks like today and then I'm going to talk about what I think are the most important factors to keep in mind as we think about the future and further moving forward with gender equality I want to talk about educational choice and I know there's lots of young people in the room so I practically happy to talk about educational choice I want to talk about the motherhood penalty I want to talk about especially in light of this motherhood penalty how we design family policies and tax policies when I get to that stage it's going to be the right time to take a little digression on kind of the nature of the remaining gender stereotypes in all societies today not Saudi Arabia, not India, but all societies today and then especially in time permit I want to talk about sexual harassment so I I think this is a topic that has taken a lot of the oxygen over the discussion about gender over the last few years it's not a topic that economists address it's not a topic for which there's much data so I just want to tell you a little bit about what I think we can understand and not understand about this ok so I'll finish with that ok so number one so this is the rich world today this is every OECD country in 2014 and this is the average gap in labour force participation which is about 13% across the OECD and the average gender pay gap which is the gender pay gap among people that work full time so this does not include the idea that more women and men work back time the average pay gap is about 15% so think about it is that for every dollar that a man earn on average across the OECD women earn 85 cents ok and the US is kind of like you know in the middle of the air the better part to be is that low quadrant over there where France belongs so from an international from an OECD perspective actually France is doing quite well it is very much in the company of your usual suspects which are mainly Scandinavian countries as I mentioned before I mean when you think about those fairly high rates of female labour force participation in the OECD that hides a lot of part time work for reasons that you know I think will become very clear later it's not the case though that the countries that manage to get lots of women in the workforce only do that by getting just higher share strong relationship between labour force participation and the gap in labour force participation and the share of women working part time but this is a phenomenon that I think is understood but like look at a place like the Netherlands 40% 45% I think part time so this is nearly the norm in some OECD countries alright and then this is progress there's a progress that I discussed there's a change, there's only 10 years of data but even in a short time period you can see that essentially all of the countries have managed to both reduce the gender gap in compensation and increasing the rate of female labour force participation with some clear success stories and as I said just a few countries that have not succeeded have not made as much progress the US is over there and it's very well known that US experience over the last in fact 20 years has been a much more negative one for women than it has been in lots of European countries we have not made much progress either on closing the gender gap in earnings or closing the gender gap in labour force participation alright and this is a change in part time work again no clear relationship which says that lots of countries have managed to get more women workforce not simply by creating more part time work alright the other big fact about the world is that those fairly low gender gaps in compensation hide a glass ceiling right which is that if you start looking at the earnings distribution of men and women what you see clearly is that the higher up you go into the income distribution the less represented women are so I can do this with data but maybe we can celebrate Tama Piketty and do it with the kind of pictures he can draw because he has tax data so this is a US over 20 year period plus more than 20 years and this is the share of women that work essentially kind of nearly 50% and then this is the share of women that have earnings top 10% of the income distribution equal representation would be about 50% today it's only 25% and the higher up you go into the income distribution the fewer you find women to be ok that's really the glass ceiling phenomenon in the picture and recently Piketty and Carthers have produced essentially the same picture for France this is your glass ceiling and your glass ceiling actually look quite similar to the US glass ceiling so the projections if you extend those lines is that women would achieve parity in the top 0 1% of the income distribution in France by 21 44 which is fine but it's still a long time ago a long time to go so these are the main facts now I want to talk as I said about a few things the first one is educational choice because I still believe educational choice as an important thing that we need to address to think about making further progress so what do we know about education I think there's a tendency to say that women have no disadvantage compared to men when it comes to education in fact the quick line now is to say that women are doing better they've reversed the gender gap in education women are more educated than men and in fact that is true so this is again the same OECD and it's the share of men with a bachelor's degree compare au more PhD included versus the share of women with a bachelor degree or more and this is the 45 degree line we can see essentially every country except Japan is on the side of women in education than men do now what's I think more problematic is that behind this success story they remain very large differences in the kind of things women study compared to guys so this is business and law and there you've got countries on each side of the 45 degree lines but this is I think the problem this is STEM this is the share of males that study a STEM field in tertiary education versus the share of women that do STEM education in tertiary education and every country is way way below the 45 degree line in fact there's a positive relationship but it's very small you can see the same when you start looking within the STEM and look at something like that's software engineers or look at kind of your regular engineers every country is strikingly below this 45 degree lines that means they're going to be areas to compensate and you know they are the ones that you would expect which is they are much more women following with this liberal arts path ok so why is this problematic I think it's problematic for you know kind of the reason that the kind of fields that you can find more women are not well paying field and that's very straight forward to you know to see I think the economist has been running a story in my Twitter feed over the last week kind of showing you these graphs better than mine this is my attempt to look at if you look by birth score hoard in the US and for every birth score hoard of male and female try to do the expected earnings simply based on the field through the gender gap in earnings what you can see is that whether you look at the mean gender gap in earnings just determined by field of study or gender gap at the 80th percentile 90th percentile we've made progress from the 1950s to about the 1970s birth score hoard in getting women entering better paying field but that progress has nearly stopped since the 1970s birth score hoard now the other reason why this is important is kind of fields that women seem to shy away from which are the STEM fields actually really good field for them so I can see the danger of what I say because I think no one has a sense as to what the labor market is going to look like in 10 years or 20 years but today these are really good field for women these are fields where the earnings are high this is a graph from Claudia Golding's presidential address a few years back the high earnings field they are the triangle in the green and also fields where the gender differences in earnings between a female software engineer and a male software engineer is very small that's in contrast with the kind of fields I have my students which is business which are very well paying fields but fields where also women really struggle to achieve the same level earnings as men so this imbalance is really important now what's behind this imbalance I think we need to understand so the typical story is this one which is the one that Larry Summers made famous when he said something that upset quite a few people which is women are just not with a math so this is a PISA data PISA data site is an international data site standardized testing across many places in the world so you can really use this data and dig in understanding gender differences you can see that indeed there are a few exceptions there are some countries where women do better on the PISA math test than men but the majority of countries test here below this 45 degree line that's in contrast with this which is science testing in science looks fairly balanced between the gender and reading where there every country is above the 45 degree lines which means that women do substantially better than guys at the reading testing ok so how important is this so I think the first thing too that I want to stress is that the gender gap in math based on all that we know doesn't have much to do with nature but it's also something that's been socialized right so we have now multiple studies that tell us that this is one study that was conducted 10 years ago that essentially established a correlation between gender gap in math within the country and gender attitudes in that country in places you have more gender conservative norms the gender gap in math is bigger in places with these gender norms are much more neutral this essentially no gender gap in math right so this has a lot of social cultural influences embedded into it there's another recent study which I really like unique but also shows the power of social norms in affecting math testing she contrasts eastern and western Europe so whatever you may dislike about socialism there's a lot they really tried really hard to get women included and tried to indoctrinate you know kind of more gender equality and you still see this in the PISA testing for these ex socialist countries today gender gap in math is substantially smaller in eastern Germany than in eastern Europe than western Europe Claudia also shows if you look within Germany women do better today in eastern Germany than they do in western Germany so point number one those differences in math they are importantly related to norms point number two is you know even if they are still gender differences in math how big are they why you know Google and Facebook cannot find software engineer now there's a standard way to measure gender differences in attributes in psychology that's called the coins statistic so what is that is taking the difference in the means for men and women and dividing that by the pool standard deviation of values for men and women so this is a useful statistics just to give you a sense of what the statistics mean for d value of point one there will be about a 92% overlap between two the male and the female distribution for d value of point 20 there will be an 85% overlap between the distribution so small and moderate values of these statistics are consistent with gender differences in mean but not large ones when you look at them visually essentially these distribution are overlapping one another ok all of the work that's been done recently on the gender gap in mass suggests that the magnitude of this gender gap is indeed within the range where coin would say this is essentially gender similarity these are very very small differences this is a meta analysis that was done in the US across all grades you can see the d's ranging from point zero one in fact minus point zero one to point zero six those differences remain true when you focus on the harder more complex mass scales and the data from other countries developed countries that make the same point so even if there's a gender gap in mass it's very very small there's essentially overlap mainly overlap between these two distributions that may explain why you know we only have males that are Maud Zuckerberg and Larry Page but not why we can't find female software engineer ok what I find interesting is this particular study which I believe come from PSE and I just discovered so if it's not about massability then what is it about why these women specifically stay away from these fields so one is just the point about role models and that's kind of if you think about the struggle that we have in economics right now you know we think that's an important factor that young women are not finding enough female role models in economics and that's why they're staying away from the field it might well be that it's a matter of preferences but you know kind of this is the messy economics part Jean like I don't believe that preferences are given I believe preferences are valuable and they also reflection of norms now what I think is another very autogonal explanation is that it might very well be a matter of comparative advantage so this is a study that was just published that again uses a piece of data and then look at the relationship between how well students are doing by the silent mass and their intention to study mass in the future these are 15 year old kids and we can see for girls and for boys there's a positive relationship the better you add mass the higher your interest in studying mass in the future but at every design male distribution is way above the female distribution at every level of massability guys want to study more mass that's the puzzle that we're trying to explain now what's remarkable is the last picture which does what it's not looking at how will you do in mass but it's looking at ranking students based on their comparative advantage in mass versus reading and now when you start ranking girls bears on the comparative advantage in mass versus reading you see that these two lines are much closer to one another so I think what's behind this and I have a 12 year old and I can see this happening is essentially the following my 12 year old is really good at mass but she's a really talented writer and she thinks about herself I think very naturally what am I the best at and she's better at reading what's behind this is potentially I think a really important explanation if this is a big part of it it also suggests some levers it suggests that the role of counselors is really important I can tell my daughter that mass pays better than English but maybe now all parents can't do that I have a comparative advantage but it suggests an important role for counseling or at least making sure that students understand that some carrier paths have different implications for earnings different implications for career volatility so I think it's a positive story because this might very well make a big difference alright next topic I want to talk about motherhood penalties I think this is the other really really important thing that you should get out of this lecture if this is a topic that's new to you I think at this point in time we're really in a situation where we understand that the old story the career family balance is in fact the bulk of what's going on when it comes to explaining gender differences in earnings these are now one, two, three so I think these are six studies in total that six countries and what do they do they track male earnings and female earnings and then they look at what happened to the earnings of male and female after the important event of the birth of the first child you become a father, you become a mother and what these pictures show is essentially that across all of these countries becoming a father has no implication for your earnings becoming a mother is dramatic long lasting implication for your earnings you have a huge motherhood penalty I think visually you can see that the size of this penalty across all places it is much smaller in Denmark than it is in Germany but all of these pictures show the same pattern no penalty for becoming a father huge penalties for becoming a mother I love this picture because in a sense I feel like this is what probably we should be aspiring of if you think about a country that has made huge headway to a place like Denmark and this is what has happened to the gender gap in earnings in Denmark over time where the authors tried to figure out how much of the gender gap in earnings can they explain because of this motherhood penalty how much of it they can explain because of the other things I talked about before educational imbalances and how much they cannot explain which you can label whatever you want maybe like just basic discrimination and essentially the picture says that the gender gap in earnings has been going down in Norway but the not really all the same but if you look at today the gender gap that remains which is about a 20% gender gap in earnings this includes part time work is solely due to the motherhood penalty Denmark has fixed educational imbalances it has gotten rid of this residual let's call it like pure discrimination taste based discrimination and the main issue they still face right now is fixing this motherhood penalty and the size of this mother penalty seems as large today as it was 30 years ago so what drives a motherhood penalty I think we know a lot about that as well and some of it is obvious one of it is just labor supply the most obvious one is just some women leave the workforce meaning the US where we have very little government support for child care if you don't earn a lot you might as well leave the workforce to take care of your child but it's more than that it's not just leaving the workforce and having zero earnings, it's slowing down it's working few hours, it's working part time it's moving from the private sector to the public sector and it's also, if you look at especially those highest status occupation in the economy a penalty you face because you need flexibility and work doesn't give it to you to work on this topic but really showing that in those high status occupation like business and law even taking a small amount of time off is going to have big penalties in terms of your wage and being willing to work 100 hours a week is going to have huge benefits in terms of your wage rate now all these things that drive this penalty more constrained on job search for mothers again, kind of some of the most exciting work here comes from France who's been using French data and documenting a fact that shouldn't surprise us but women, when they search for job are looking at a more reduced option set and why? because they are not as able or willing to commute as men are probably because they need work to be closer to the home so Thomas shows that in the context of France there's a big difference between men and women in terms of the lengths of the commute they have that translate into for women than for men so all of this matters all of this though is going to be modulated by family policy tax policy and norms so let me talk about before I do alright so what do we know about family policies in light of all of this what does it mean to design good family policies if the objective is greater gender equality right we have maternity leave with different kind of replacement rates and lots of variations throughout the world the US is down there there's no federal maternity leave mandate in the US, a few states have them but we are very much an outlier if you at family belief policies they are not substantially related to gender norms it's interesting we think about where these policies come from the reflection of the views that people have if you look across countries it is not the case that there is a systematic relationship between gender attitudes and the lengths of maternity leave I don't think that is that surprising because if you think about maternity leave policies they really have two goals one is to help women work but the other one is also a very gender conservative one which is let's have children right so these two forces seem to come to balance in terms of driving the political economy of these policies now this one is really strongly related to gender norms places that are more gender progressive have government spend much more money on early child education and child care and that I think again makes sense because the political economy here is what the goal of policies like government spending on child care and early childhood is really to get women to work there is no ambiguity here and you see that in the cross in the cross sectional data now what do we know about whether these policies are good or bad for women well this is maternity leave and the two objectives we have which is to get more women to work force and women to be closer to parity with men in terms of earnings there is mild positive relationship between the lengths of maternity leave and getting more women to work force but there is also evidence that longer maternity leave in fact increase the gender gap in full time employment the gender gap in earnings sorry for people working full time if you look at government spending on early child care and child education there is no ambiguity in the cross sectional data this spending is good is that it gets more women to work and it also reduces the gender gap in earnings all this is just the cross sectional data but I think these cross sectional facts very much hold true when you try to do the more expensive work of exploding all the changes in policies that have happened among these 20, 30 countries over time I think the main takeaway is that I'm not forcing you to read the slide is that you there's a point at which too long of maternity leave is going to be bad for women particularly the earnings and particularly for the higher educated women I think that should make total sense is that if an employer has got a guarantee to rehire someone like 3 years down the road he might even not hire this woman in the first place so maternity leave policies cannot go beyond the 9 months, 12 months before they become fairly detrimental to women the opposite of the objective now when it comes to like government spending on early childhood there's no ambiguity the more we spend across countries and change it in spending across countries over time the more women begin to workforce and the smaller the gender gap in earnings so I think this is important again the US does very poorly on kind of spending on children and we are in a moment where we have this whole discussion about social mobility in the US as well and if I read my colleagues that think a lot about social mobility they say well child allowance is really the way to go a location familiale, we don't have any such thing in the US that could be an equalizer in terms of helping poor kids well I would say that there's a double you know it could be a double win I think policies like that spending on children could be good for social mobility and also good for more equality in the labor market alright so this is the daddy quotas this is the other policy that Jean briefly mentioned they are much rarer I think they exactly hit at the right point which is that we want more balance between the gender in the role that men and women play in the home what do we know about the daddy quotas that exist in all kind of places like Scandinavia we know that men take them but they never take more than them and we have I think too little evidence so far to assess whether or not these daddy quotas help women but at the core they remain very short in terms of their lengths so it may be asking too much to see big changes the other thing that I'll say about this which I think again is this point about when I looked at this graph I was surprised to see Korea and Japan as being the most aggressive countries in terms of the lengths if they are reserving no men takes up these policies in Korea and Japan which means that even if they are good policies you're going to have to do something to encourage take up particular places where the norms go in the opposite direction all right I think good I'm going to talk about this because I think tax policy is also important and so it's something that I'm not a public economist I totally under-appreciated before reading a series of paper recently in the particular aspect of tax policy that I think I want to discuss is the issue of joint taxation of household income right so this is work that Alexander Beek and Nicola Fuchsundlund have been doing over the last few years here's a story this is the US this is Germany and this is Sweden this is depending on how many hours you work with assuming that your earnings your wage rate is essentially the average wage rate in each of these countries the tax rate that you're going to be facing ok so that's the first panel on the left the picture is what well the US has got lower taxes lower tax rates on average right that's why the the bold line is below and then you've got Sweden and Germany that kind of look fairly similar to their tax profile and tax progressivity now this now let's go to the picture the picture on the right so what is that well that's the tax rate that a wife would be facing in each of these countries assuming that she's married to a men in that country who works about the average number of hours than men work what you see well a very different picture in particular while in terms of progressivity the US, sorry Germany and Sweden look the same when you look at the tax rate that the wives would be facing which would work on top of her husband well now Germany is way up there much higher marginal tax rates and the US and Sweden what is that well the reason is that in a place like Sweden you have separate taxation so when I work it doesn't matter whether my husband or wife work in terms of like the tax rate I'm going to be facing a place like Germany has got this form of joint taxation system which is that when the wife typically the orthogonal learner works on top of her husband the tax rate that she's going to face it's not about children but you can see why it relates to children because you get married even in a system of joint taxation gosh, you know kind of staying at home alone is boring so I might as well work even if I face a high marginal tax rate because there's kind of complementary between me and my husband taking leisure together but once the children come into the picture this kind of complementary and then the wife disappears and these disincentive coming from tax system become much more important I think so that's a piece that I think I was not appreciating before doing this work but I think we need to discuss more and again that's an easy one because we can change this system we can change the way we tax families so that's a message of the paper they do a lot of work I don't know kind of whether you know where France sits here so this is a sense of like how much tax you face women in a particular country whether you are single or whether you are married and this is a 45 degree line so a place like Sweden it's separate taxation so it is on a 45 degree line doesn't matter whether I'm married or not but then you've got places like Germany, my country, Belgium where this strong disincentive for the second order coming from the joint taxation system France is also a country where these disincentive effects are relevant so this is important and is fixable, addressable via policy I want to make sure this time for questions so I want to skip a few things and I want to stick to stock because I've given you a lot of facts but I also want to do a little bit of reflection so I believe that it's very clear that mothers today in all societies still bear the entirety of the labour market at the cost of giving birth and maybe in the most progressive countries you can think about maybe that's really the only thing that's left separating men and women in terms of labour market success I think it's also clear that the design of family policies tax policies are going to matter and these ways to make this motherhood penalty bigger or smaller depending on how you design those policies now what does it mean to live in a society where this you know the child kind of penalty is the key driver of gender inequality this is clearly not Saudi Arabia or India I think it's really kind of saying that in the most in places like Saudi Arabia in India we're talking about very hostile belief against women the idea that women don't belong in the workplace in all societies we're talking about they're more benevolent seems to be beliefs of the kind women are going to do better taking care of the kids or maybe women have a preference for taking care of kids so what I want to do in the next few slides is just try to tell you a bit more about how much we know about this are these beliefs accurate is it true that women are better at caring for our children do we have the right beliefs about this or are they inaccurate beliefs so I'm going to do a little digression to talk about our understanding of gender stereotypes in the rich world today alright so in economics how do we think about discrimination I think we have we've been stuck with our two workhorse model one is the Bikarian state taste based discrimination I don't like women I don't want to work with them I don't think that fits right men like women a lot and then we have this model which is statistical discrimination it's a model discrimination that is based on beliefs but rational beliefs I'm not going to hire this african-american person because I know that african-american are less educated than white people on average and I know that correctly and so I'm discriminating but there's a reason for it it's rational and maybe from a philosophical perspective there's a problem with this now when you look at the psychology literature and how they think about discrimination their models are so much richer than ours right and in particular they have also this belief based model of discrimination but they are much more willing to accept the idea that the beliefs that are driving you know all the discrimination might be inaccurate ok so this literature in psychology and I'm going to tell you a little bit about what we know the first one is number one are there systematic differences between men and women women better at caring or are they better at taking risks or all this stuff number one and number two are the beliefs about these gender differences accurate so here's my reading of this literature so this is a long slide and you can read it yourself right there's a lot of meta-analysis and if you're interested in this I suggest reading Hyde 2014 which is essentially let's take all the studies that have been done on there which are measuring hundreds of traits as I said you know empathy but you know casual sex or risk taking you know really really hundreds of things that we think matter and let's figure out what is the gender gap on all these things and the overwhelming conclusion of this work is that for the vast majority of traits those D statistics which are the metrics are very small that means what that over a lot of traits they are gender differences but the gender differences that exist are so small compared to the within gender variation that exist so that's I think the overwhelming conclusion of this work that does not do one study at a time but really tries to take stock of all that we have studying and learning now they are exceptions and you know I written the code over there and among the things that systematically differ between the gender with these of values of like up to 0.8 so that means this distribution are kind of really apart from one another includes things that I think we may associate with the idea of taking care like you know agreeableness, dendermidenness interest in things versus people women have got more interest in people than having things so they might be here some roots for beliefs about women having an advantage in caregiving or maybe even preferences for caregiving alright so now comes the second question fact number one most of the gender differences in traits are very small, there are a few exceptions question number two I believe that we have about these gender differences correct but you can study that and again there is a large literature in psychology on this and I would say this literature is mixed you know just like us economists psychologists have got ideological battles so you can read some articles that say one thing and other that say another thing I think the main takeaways is that the evidence is mixed as to whether or not our beliefs are correct now the one study that I want to point to attention to which to me makes a lot of sense in making sense of these results is a study by my colleague Nick Epley, who's a psychologist and here's the idea behind Nick Epley result look at, if I were to ask you right now, what's typical of a man what's typical of a woman what are the things that you really associate with a man the things that you really associate with a woman we don't have the time to do this real life but Nick did this subjects in the experiment and that's what he found and I'm sure you cannot read it but the places where men and women are most judged to be most different based on is empathy women are expected when I think about women, I think about empathy when I think about a guy, I think about not empathy, I think about aggressive assertive and when I think about a woman I also think about womb when I think about a woman I don't think about happy happiness is not a trade associated with women so now the next part of the exercise because I think we are getting short of time is they hypothesize the following is that beliefs are going to be much more likely to be encart on those traits that we associate you know immediately associate with a gender so the idea is is that when I think about women I think about womb and I think about empathic and this is an area where beliefs are going to exaggerate gender differences on the other hand when I think about women I don't think about happy or unhappy so this is going to be a trade where beliefs are going to be fairly accurate in terms of explaining predicting gender differences and this is exactly what they find in this study so these are measures of empathy a lot of these are eye test where you kind of look at whether people are paying attention to faces baron coin most important on this and the way you would read this again maybe it's too small but let me tell you in words is that women are more empathic than guys this is an area where these statistics is actually fairly high women are more empathic empathic but beliefs about how much more empathic women are than guys exaggerate the extent of these gender differences right you now find the same without for happiness where the beliefs are actually fairly accurate so what's the bottom line of all of this is that women care but they probably don't care as much as you think right but it also says that there's a kernel of truth behind you know those beliefs that are inaccurate all of this I think is quite important why am I stressing this because I'm trying to understand why we are stuck with this so why do I what do I do with all of this and these are going to be my few last slides and I'm sorry sexual harassment I won't have time for but I'm going to conclude with this I think it's important to understand that stereotypes are not just descriptive but they're also prescriptive and now the psychologists tell us very clearly right descriptive I mean this is how people will behave I predicted prescriptive this is how I expect people to behave so whenever I believe that women are caring even exactly believe that women are caring I expect them to be to be caring those prescriptive you know kind of the prescriptive nature of the beliefs make it particularly difficult to undo the stereotypes because you expect this behavior you receive this behavior then you observe the behavior this pattern of self-fulfilling prophecy that makes this kind of a vicious side so why are those remaining beliefs, inaccurate beliefs particularly difficult to undo I think it's a really good question can have multiple explanations for it first because they're prescriptive you said that you expect me to be caring I'm going to be caring otherwise I may face some kind of backlash that one view but I think the other thing that we discussed they're actually positive stereotypes it's not like women are bad or lazy these are positive stereotypes women have empathy and they care and these might be difficult to undo because at the core they are much less objectionable than negative stereotypes so what do we do with all of this and these are going to be my last few slides so that there's 5 minutes for question I think that as researchers we pay we play a really important role right I think that we get because of publication bias because of the way journals work because of what the media wants to talk about when it wants to talk about our research we don't see all of data in all the facts and we remember and we publicize the studies that I do not I've done it myself I've talked about the research on gender gap and competitiveness there's been 10 papers that one paper I quote that can't really replicate this systematically so I think as researchers we really, if we take our work seriously face responsibility to just do more meta-analysis and our profession is really not doing a good job in incentivizing us in doing replication and summary work of that the second thing related is just the role of the media the internet advertisers in reinforcing those beliefs the UK just passed a new rule that says that if you are a media agency the kind of advertising you can show on UK television cannot be stereotypical I think these things are important and is obviously roles of parents and researchers the last thing I would say I think this agenda also tells us that there's a role for organizational practices I've talked much less about what happens in the workplace but I think it's important to realize that these inaccurate stereotypes matter not just in terms of how husbands and wife decide on allocation of time between the home and the workplace but they also matter at work so this role in cognitive literature suggests that women get treated differently and are not perceived to have the traits that associate with leadership maybe because they are perceived as being too nice so at the core I think we need to educate much more you know, kind of human resource managers evaluators about the prescriptive and descriptive nature of stereotypes we need to make sure that people that make important decisions within organization are doing so within enough time to make a decision again this is usual in psychology that says that we're going to rely on stereotypes much more when we feel rushed stressed so creating a good environment for people to be evaluated hiring decisions to be made is going to be important we also know that one way to avoid stereotypical thinking is to just try to get people to think about the individual and not just the category the kind of she's not a good fit and that's my explanation as to why I'm not hiring this woman should not be something that is accepted by organization organization would have much clearer criteria in place that force the evaluators the recruiter to explain why she's not a good fit rather than just stating that fact I mean at the end I guess what I was going to say is that I've done lots of work on how quotas and how quotas can help women in workplace I think it's just forcing diversity in organization I'm skeptical as to this being the agent of change and I like to flip it rather than this diversity inclusion agenda that we have thinking more about how we create inclusive organization and getting diversity is not kind of that is probably a better way to think about it so I'm going to stop there too long, there should be time for question so I won't tell you much about sexual harassment but in a snapshot I think that we can study it I think there's data and I think it is sometimes predictable organizational practices, organizational dysfunction the way we write incentive contracts for people, the way we pay people the amount of control we gave to people at work all these things seems to be correlates at least of sexual harassment so it should not be a subject that we stay aware of I think it is related to many of the things that we studied I'll stop there