 Good afternoon, happy to be here presenting this paper. We're in a tight schedule, so I'm going to go right to it. This is joint work with Sonia Balotra at Warwick and Fanhuaca University of Houston, and I'm here based at Los Andes University. So I'm going to try to kind of motivate this female labor for participation, one of the most salient features of the labor market past century. We're going to be focusing on the empirical part in Latin America, and this phenomenon has been exactly the same in Latin America in a different timing. So if you look between 1990 and 2012, female labor for participation increased by about 35%. At least in that period, it was the steepest increase of any region in the world, although we have seen that FLFP has been decelerating in recent years. And at the same time that you see this increase in female labor supply, you also observe chart changes in the weight structure, and maybe you have seen some of the papers that have been shown over here, particularly in Latin America, so educational premiums are changing, experienced premiums are changing, the gender wage gap is changing, right? So the main question that we want to ask in this paper is, it's what's the relation between these two phenomena? So what is the impact of this massive change in the size and composition of labor force on the wage structure more generally, and specifically on the gender wage gap? So how FLFP affects the wage structure, that's what we're going to try to do. So kind of the starting point for us is that these are two forces that are going to reinforce each other, so what is happening with wages is going to affect participation decisions, and what is happening to participation decision is also going to affect the wage structure. Kind of that's our starting point. If you go to the literature, you can pretty much, you can find at least four factors that are going to be linking these two forces, the participation and the wage structure, two of them coming from the demand side, and at least two of them coming from the supply side. So if you think about the demand side channel, maybe the key one is going to be structural change. So the economy, the composition of employment and production of the economy is changing. And if you, mostly moving from, for example, from agricultural to industry to service sectors. And if you have, for example, that if women have a higher tendency to work in service occupations of the service sectors, and you have that these sectors or these occupations are expanding, you can think of this as a positive demand shock on female labor, right? The same goes with non-neutral technical change. So some occupations are being replaced, some occupations are being complimented to the extent that, for example, women are more prone to being some of these occupations, you're going to see demand, you're going to see changes on the demand side of the economy. So that's one of the components. The other is related to imperfect substitutability between male and female labor. To the extent that you allow for that possibility that not necessarily because of technical reasons it could be or because of cultural reasons or because of perception that employers have about female labor, you could have imperfect substitutability between these two. And I'm going to show you in a while how that, that imperfect substitutability can affect what it's going, what it's happening in terms of the wage structure. So those are like two demand side mechanisms. On the supply side, you have demographic compositional changes. So for example, increasing in educational attainment, people with higher education tend to participate more. We see convergence in educational attainment and that's going to affect not only participation decision but it's also going to affect what's happening with relative wages. And in the last, it's a very large literature on what we're calling over here, non-wage supply shifters. So it's everything that is related, for example, for fertility, marital status, social norms, home production technologies. The shifters that are moving what is happening in terms of labor supply that are also potentially affecting relative wages through, as we're going to set through these mechanisms. So what we want to do in the paper is there's, there's extensive literature in each of these channels. What we want to do in the literature is try to model these channels within a unifying framework where what it's specific about this unifying framework is when we're going to allow labor supply to respond what it's happening to wages and we're going to allow wages or the wage structure to respond what it's happening with the labor market and with the labor supply. So that's what essentially what, that's what the paper is about. So we have a full model, but before we go to the full model, what I want to do is take initially a toy model, like a simplified version, and try to pin down exactly the mechanism how these things are going to look like in the model. So it's going to be easier to understand. So we're going to start with a simplified situation where we have three occupations. You can think one of these occupations is home production. So not paid occupations. And we're going to divide the population by gender, male and female. So I'm going to be talking every time I say gender, it's sex at birth, right? And we're going to divide in these two types by gender, female and male. And the model is going to have a supply side and in that supply side, we're going to have is that workers are going to choose these occupations based on either preferences, demographics, these labor supply shifters that we were discussing before and wages. And where we're going to arrive in terms of the supply side of the model, I haven't told you how we're going to arrive these, but we'll get there. We're going to arrive at the determination of labor supply in a given occupation for a given gender in a moment T. It's going to depend on the total population of that type at that moment in time, times the probability that someone of that type is going to choose the particular occupation that probability is given by this function F. So here we start seeing parameters that are going to be important. One of these parameters is going to be these side parameters that are going to give me a sense of what is the wage elasticity of labor supply for that occupation. So that's what we're saying that labor supply might be responsive to what's happening to wages. And on the other hand, this is also going to depend on some set or a vector of several variables that are going to shift the supply side, right? So that's the supply side of the model and the demand side of the model, we're going to have a standard need assess production function and in the traditional kind of classical result when you use this production function, you're going to be able to decompose relative wages of two types of workers. In these times, for example, male and female workers, male being K, female workers F. You're going to be able to separate these in two components. One component is going to be over here that is going to depend on these alpha shares. These alpha share parameters are going to give me the intensity in which the inputs are being used. And these are going to be able to bury in time capturing, for example, shift things that are happening on the demand side, structural change, non-neutral technical change. And the other component that you can separate these is in terms of the impact of what is happening in terms of the relative quantities within this occupation. And the impact of the relative quantities is going to depend essentially on what's happened with the elasticity of substitution between male and female labor, at least in that occupation. So if you have perfect elasticity, these things are going to go to infinity. There's going to be no effect, but if elasticity is not perfect, if they're imperfectly substitutable, then what is happening in terms of relative supplies is going to affect relative wages. So we have these two components and we're going to put them together and we're going to have as an equilibrium model in which wages and labor supplies are going to endogenously adjust to resolve the model. Now, these are the two key equations. So the termination of labor supply and what's happening to relative wages. And over here I come back to the four channel that I was discussing before. So we're able to capture these four channels through the simple framework in the following sense. So for example, you take some supply shifters that is within this vector of observables, for example, fertility rates. So fertility rates might affect the probability of choosing, for example, a market occupation. This is going to change labor supply. Labor supply is going to affect relative wages through this imperfect elasticity mechanism, but that in turn is going to feed back into what's happening in terms of labor supply through the decision on wages. So the idea, and you can do this with a different channel. So this is going to be the channel of supply shifters. This is going to be the channel that it's working through demographics and this is going to be the channel that it's working through the demand side. So we have, and again, the two key parameters over here are these wage elasticities of labor supplies and this elasticity of substitution between male and female workers. So that's what we have. What we want to do in the model is essentially try to estimate these parameters, right, to estimate the parameters and there's going to be, some of these parameters are going to be of interest by themselves, but we want to do some counterfactual analysis. For example, if we fix fertility rates at, you know, exogenous at some level, we allow the model to run the equilibrium at the end where we're going to try to capture it's what's happening in general to the wage structure and what's happening with labor supply. So that's basically what we're doing in the simplified framework that we have over here. We're going to do this in the Mexican context. So we're going to use Mexican data. Let me just, sometimes you have to try to motivate why some countries, I don't think over here is the case, but you see a very sharp increase in female labor for participation starting from a very low level. You start in 1990, around 35% and in 2014, you're at 60%. So we're talking about a really massive increase or incorporation of women into the labor market. And the other thing that has been drawing our attention is what's happening with the gender wage gaps in Mexico over these periods. So what I have over here, if you want, is how the gender wage gap is changing at each percentile of the distribution. So changing between 1989 and 2014, and this would be at the 5th percentile, 10th percentile, so on and so forth, until the 95th percentile. And this is very particular and very peculiar because what you observe is that at the lower tail of the distribution, what you observe is increasing gender wage gap, at the upper tail of the distribution, what you observe is converges. And I'm going to have to move quite fast. So one of the elements that we're thinking about is one of the reasons that this might be happening is through the force that female labor for participation is exerting downward pressure to the wages of women that are in occupations that are in the lower end of the wage distribution. That's if you want one of the hypotheses that we have in the work. Okay, we use, I'm just not going to move to the data. Let me just say this, what we do in the full model is take this initial model that I was discussing and we're going to allow to have different types. So we're not going to have only male and female labor, but we're going to have male and female labor also separated, depending on the educational attainment that they have. We're going to organize this one as the CS model that has three layers. I'm not going to go through the layers, but the general sense of this is exactly the same. In each layer, you're going to have these alpha parameters that are going to capture in the demand side and these row parameters that are going to capture in the elasticity. And in terms of the supply side, we're going to have is a random utility model where the choice of occupation conditional on your type, it's going to depend some non-pecuniary words, pecuniary words, this is the side parameter that I was discussing, and some IID take shock, and the utility of home production is going to depend on some preference parameters and these are the supply shifters that we are, in this case, we're going to take as exogenous. So this is one of the things that we would like to incorporate, incorporate these as endogenous decisions, but that's what we have. So probability of having a child under the age of five being marriage appliance, so on and so forth. So depending on the distribution on this take shock, you can characterize the probability of choosing each of the occupations and in general you're going to arrive at the same equation that I was showing you before, labor supply is going to depend on total population of those types than the probability of choosing this and the model closes in terms of wages, equal marginal productivity, so that's kind of another limitation of the model because we don't have really space for discrimination or for bargaining power in the determination of wages and the equilibrium is going to be just the man equal supply. So I'm definitely not going to have time to talk about identification, but we spend a painful amount of time discussing identification in the model. If anybody's interested, I'm more than happy to discuss this after the presentation. Okay, so let me at least to give you kind of main takeaways given the amount of time that we have. So I was talking about some parameters that were important for us, one of those parameters is this elasticity of substitution between male and female labor. What we find is some evidence that it's essentially it's not perfect elasticity. Moreover, if you like, these wage elasticities can vary depending on the occupations. We separate occupation in three abstract manual and routine following all the literature. And we find that these elasticities are much lower manual occupations that tend to be located in the lower part of the income distribution. So that figure that I show you in terms of how the wage gap was evolving, this is consistent with this idea that potentially female labor supply increases in female labor for participation, particularly at that part of the distribution is exerting higher pressure. We see more substitutability in kind of analytical abstract type of occupation, which kind of we believe that makes sense. So this is one of the elements, one of the key parameters. In terms of the demand side, we find something that it's also, we're observing the literature. Demand side trends appear to be working in a way that is favorable towards women. This is related to how structural change might be thought of as this positive demand shock to female labor. We observe it for college educated, high school educated workers. And this is another thing that it's also very interesting and it's related to the side parameters. So a lot of this literature studies only the demand side or the supply side. When people are studying only the demand side, essentially what they're saying is that labor supply is inelastic. That's kind of the underlying assumption. We do observe, this is a measure of this elasticity. We do observe that it's inelastic for males, but we do not observe the same for women. We do search for responses of women to what's happening, to wages. And another thing that it's interesting, this is females, depending where they're skilled, another thing that we're interesting is over time this elasticity is becoming lower. So in a sense it's becoming more inelastic over time, which is also consistent some results in the literature. Am I done? Okay, so I run very fast with that. I wanted to show you some of the compositions, but we're definitely not gonna have time. What I'm gonna make, I'm gonna try to also make the point. So first point is this point about the elasticities, how we're capturing and potentially not perfectly substitutable. The other is what's happening on the demand side. The other is about weight elasticity of labor supply being higher for women. And the last thing that I wanna capture is the importance that we give to thinking about this in terms of general equilibrium. So, I'll give you an example. So we have in the model, the model predicts convergence in terms of female labor participation of about 20 percentage points, right? So female is increasing, males are staying more or less the same, there's convergence about that. One of the exercises that we did is suppose you fix the availability of appliances that affect female labor for participation, especially among low-skill workers, you fix it about in the 1999 level. What happens over here is that this convergence is much lower, right? It's about 62% lower. Now this is the case if we shut down the general equilibrium component. That is, if we shut down that side, we don't allow labor supply to respond to wages. But if we run the full model in general equilibrium, we find that this negative effect is mostly attenuated because you have less women entering the labor market that implies less pressure on wages, less downward pressure on wages. But if that is the case through the wage mechanism, more women are entering the labor market and that generates an attenuation effect. So moving from the partial equilibrium that we've seen to the general equilibrium, it's fundamental, we make the case that it's fundamental in different aspects over here. And I think I'm pretty much...