 Thank you for a very interesting presentation. So I have two quick thoughts. One is, you're identifying that it's probably the bargaining power channel that is having a role to play. So do you see changes happening in terms of the expenditure or the decision making that women have when it comes to expenses on the food expenditure, schooling expenditure, and other things, right? Because that would substantiate that this is really the bargaining power that is leading to this differential impact. And the second thing was you are finding this the impact coming from rural areas, right? And in rural areas, most of the women are engaged in these family farms where they directly contribute to the production, which can be used for self-consumption and directly aids to the, you know, the households died to diversity in that sense. And that's what I find in one of my papers that both types of labor market engagement as well as home engagement can have an impact. So I was interested in understanding what is really going in the Indonesian context, right? Because you're finding that it's when they are doing this. So is there any trade-off that they are facing in terms of taking labor market employment compared to working on the family farm? Hi, thank you for a very interesting presentation. Just a quick, very quick, small question. You included age of the child as a control, but you didn't comment anything on the results. And I thought it was like positive, and maybe as the, I don't know, if you have any hypothesis about this, if this is as the child grows up, the maternal employment effect is larger because maybe for smaller kids, one of the two forces is rebelling, or if you have thought about it, or just to comment on that. Thank you. Okay, so thanks for the question. So first point, bargaining power and expenditure. So our measure of bargaining power is basically, I didn't have the time to talk about this, is basically the expenses that mothers make on food, routine expensive purchases, clothes, clothes and children clothes. So it also includes, in a way, expenditure for education. So that's basically how we capture this. The other questions about rural, the effect on rural areas. In my view, it's largely captured by the fact that actually most of these women are kind of in the informal sector. So if, for example, we look at the descriptive statistics, I think I have it here. So most of them are kind of self-employed and family worker. We have 15% are a family worker. So it's largely comes from this type of employment, which is most we spread, of course, in rural areas. It's also true that in rural areas, we might observe, but we don't can observe this in the data, kind of also larger networks of child care, informal child care provision. So the neighbor, which you don't, cannot have so much in town, you know, in urban areas. So this might bring the effect that we see, that is only in rural and not in urban areas. And age, so that's basically just the effect of age on child development, so because it's not interacted with the maternal employment. So it basically just means that as the child grows, then it's, which is consistent with the height for age, then the height is, you know, is improved or with schooling even better. So it's, yeah. And not caring much about what other people's perception is, right? So maybe you could, you know, shed some light on that. And the final comment is about the reverse causality here, right? Because it's possible that, you know, you're marrying them early, so if there are girls who are not married early, they tend to be harassed more. And that's what is leading to this relationship that you are capturing here. Thank you. Thank you very much for the very interesting presentation. I think she asked the first, the question I had about who is asking, whose perception your caption has already been asked, but I also wanted to find out what, I don't know if I missed it during your presentation, but what are these specific questions about, you know, about harassment? And then I was wondering if you get to capture, you get to look at the differences in effects when you look at actually lived experiences if people actually report that they have been harassed. Was there anything like that in the data so that you can actually look at the effects for the lived experiences versus the perception to see, you know, if there's anything going on. And I also wanted to find out if you've considered, maybe, I guess there are several questions about, you know, perception. So maybe if you can create an index out of it and see, you know, what you find instead of maybe looking at one after the other. Thanks. Yes, so these are very good comments. I agree with you that whose perception is it? It matters the most. So I think you, I have to, I haven't explored that, but you would probably agree that generally when these surveys, the surveyors go to the households, it's the head of the household that they try to interact with. So I'm guessing maybe that will be the case. But yeah, I will be, it will be interesting to check that who is that. And your other point was related to removing that own perception of household from that variable. So you think that that's not a right thing to do. Yeah, yeah, actually that's what the paper is trying to do that, not looking at the actual crime rate, but the perception of households, how they feel, because we are trying to explode the chastity mechanism. Yeah, so that's a valid point. So maybe I think that first when I started this, I did it without removing. The results are the same, but maybe I should include it and put it in an appendix or something. And your point about, what was your point, sorry? Yeah, so that the question is how frequently unmarried girls are harassed in the community, in your locality, so that's the question. And all other questions related to the crime, like theft, robbery, so they are the same. And the answers are never frequently sometimes. So that's how it is. So is there a way to capture a particular household has somebody, maybe a female or a male who has actually experienced some form of harassment? Because that could be a strong indicator. Is it not just like, you know, in this household, somebody was there anything like that? Yeah, I don't think we have that, yeah, but that's a good point, we don't have it. Hi, sir, really nice presentation and a lot of interesting insights, but at the same time some, you know, some which are not very, you know, settling also. So I would highlight my concerns. And but first I want to start with the fact that the wealth shocks, right? So if I understood it correctly, you're talking about whether shocks. Not just whether shocks, it's all negative possible shocks. It's a lot of disasters. Business, not sales, family, debt, loss, income, it's a bunch of them, but I mean, we might have to emphasize the pressure source for some floods. Flux, right? Yeah, so they are weather shocks and they generally lead to income shocks, not wealth shocks, right? And probably that is why you're not able to capture any significant changes when it comes to schooling or early marriages, right? Because these are temporary shocks. Today you have drought this year, next year you might not have it. So keeping the climate change aspect aside, but that's probably one of the explanations why you don't find a lot there. Also the school expenses, right? They're not very high anyways. And if there are such shocks, the returns to being in the labor market are anyway lower. So you would rather prefer to stay in school where you don't have to pay a lot instead of going to, you know, leaving school and adding to the job market. So that is probably one story you can explore there. The second thing that I found a bit difficult was the negative coefficient of the bright price, right? So if there is a shock to income, it makes perfect sense that you don't have dowry to pay. But when it comes to bright price, that is the amount that you are going to be receiving if you marry your daughter early, right? So that should have had a positive coefficient, even if it's not significant, but you're finding a negative significant coefficient. That worries me a lot, right? So what is really going on then? Because the story that you're trying to build that doesn't, you know, it doesn't really reconcile with the results you're finding between dowry and the bright price. So that's where the difference with Pakistan comes in. So with bright price, yeah, that's true that it would come in from the husband's side, but a lot of these marriages are within the community and I think a very large number of people who are married with their cousins, right? So it's not just that it's a different family, it could be the same family as well, like it could be their cousin, like their hala son or something, right? So the effect, I think the shock, if it affects, so it would be the whole neighborhood basically. So it's not just that the family here itself is not able to pay dowry, at the same time the groom's family will not be able to offer a higher bright price as well. So from the both sides, it would be a decrease and it would be, yes, it would be more of mutual agreement to postpone the marriages then. So I should have added that, we had a lot of cousin marriages, I think about 80% in this sample. And again, these are first cousins, so like very close relatives. So I would think that that's the mechanism here, it's not just the bride's family or also the groom's family, they would decide with them. And, yes. If you actually have a distance, right, if they are within the same community versus outside the community? Yeah. Because that's how marriages started, right? Exactly. We ensure the estate can be in shocks. So in my data, I can track where they go and they stay in the same villages basically. I should add that to the paper, I don't have enough space left, but that should, I should put that. I was wondering whether your framework allows for dual labor markets in terms of formal versus informal employment and to which extent your elasticity might differ precisely if you take this into account. So I don't know whether you have done it and what are your thoughts on that. And the second question is, I wonder as well how well you capture what's going on at the top of the distribution given that you use survey data, have you wondered what might be the effect of use, not mean data to look at what's going on in terms of a gender gap within your model? Very nice paper. Two questions. One, I want, maybe you could elaborate a little bit more how you think about imperfect sustainability between men and women. And you're thinking about within what you call an abstract occupation, women and men are doing fundamentally different jobs, different things, and that's why they're imperfect substitutes. And the second is, I'm just, this is a small curiosity, but why did you, I mean in 94 Mexico suffered humongous labor market crisis, the tequila crisis, right? So, aren't you making your life very difficult by starting in 89? Why do you gain, I mean, I presume that chalk changed the labor market in very dramatic ways. Last, the city of labor supply responds to wages, right? And for the skilled unskilled for women, we were seeing that the elasticity is high compared to that of men. So, there are two reasons for which this could arise. One possibly could be that they are engaged in different types of skilled work, right? And if that is the case, then it comes back to the point that was made back there, right? That they might not be substitutable after all, right? So, how do you like, do you find something happening there in terms of the substitutability and the elasticity in terms of the interaction of these two parameters? Great, so these are great questions. Thank you. About the formula and formal, I think the model is, it's reaching out that we can potentially integrate formula and informal. One of the elements that, again, we've been thinking a lot, is how wage determination is done is through, we're essentially perfectly functioning market. So, wages equal marginal productivities and we like to move into another type of determination. That would be interesting if we move in the direction of formula and formal workers. But I can see ways in which we could extend the model because it's really flexible to think about that. That was like the first, about the top of the distribution, we cannot say a lot. So, we use survey data. We do some exercises using census data, but that's not gonna really, it's not gonna help you a lot. So, yeah, if you want, you have to be careful about, if you're thinking about the wage gap at the top, you have to be careful of that, although we are always dealing with labor income, right? We're not dealing with capital income or the sort of thing that might be important at the top. That's the second, that's a tough question about imperfect substitutability, but it's a very good question, I've been thinking a lot about. Personally, I do not think this is technical substitutability. I do not think that this is just some things that male female workers can do or cannot do. What I think is that if you break these occupations, if you really break these occupation at higher digits, what you're gonna observe is very strong occupational segregation. And you observe it in the data once you decompose that. So what is happening is that females and male are simply working in different occupations once you really go and desegregate that. And when they're entering those occupations, females are competing with females, not necessarily with male. Is this, the market is very segregated, particularly at the lower end of the distribution. So if you go at the bottom, that happens if you go at the top in high-skill occupations, in occupation of people that have tertiary occupations, the mix is much, it's much, I would quote unquote, better. So I think in substitutability in that sense, but that is capturing the model in terms of a technical parameter. And we've had, yeah, I've had some struggles when I present this just to not give the wrong kind of, the wrong interpretation to that, but I think that is the case in terms of Mexico. So we use kind of, when I was looking at the gaps in that period, we don't take exactly the moment, but when I was looking at the gaps of the distribution, we don't take exactly 94, 93, 94. So we take before and everything over here, estimated over the 20 years, 20 so much years that we have of the data, but I agree. We have a huge shock over there in 93, 94. Now, this is a model that is nothing, specific nothing particular to Mexico. This is a model that we can apply. And what we're trying to do is, maybe I'll pitch it over here. So we have a wage page for this and we're trying to do is just automatize the model so it can apply in other circumstances because we believe that it could be an interesting model just to think about these issues in other contexts. But I agree with Julianne in that sense. And I don't know. And the last is about the wage elasticities of labor supply. I don't have, the only thing that I could say in that regard is that our estimated elasticities are very close to what you observe in other estimates of the literature, but you would really have to, not only that, but the pattern of becoming less elastic over time that happens between in on-scale and in-scale. It's also something that it's observed and the fact that essentially, mail labor supplies basically in elastic. So it's very consistent as to exactly why I wouldn't, again, I could, we don't have evidence on that. But I believe it's a super interesting pattern that you're observing. So it's kind of converging more to mail in that sense. So that's, thank you. Thank you.