 So well thank you all for being here today in this very last session of the conferences. It's Saturday so we are really glad that you made it. So well the paper I'm presenting today is a co-authored paper with Rachel. As the title clearly points out we are looking at ethnic inequalities and the educational and labor market outcomes in Guatemala. Let's move quickly to the objective of the paper. So basically what we want is to understand horizontal inequalities but also indigenous politics in the country. Why do we think that Guatemala is important case to illustrate these two phenomena? Well first we have two main reasons. First Guatemala has the second largest indigenous population by percentage in Latin America and a very well documented history of horizontal inequality and violent state-sponsored oppression. So we'll go through it later on. The second reason is that comparing to other Latin American countries and notably with Bolivia what we find is that there is a weakness of indigenous political mobilization at the national level in the country. So basically what we're saying is if you take the case of Bolivia and the case of Guatemala both of them with a high share of indigenous population and if you look at the political sphere you will realize that on one hand Bolivia with 60% of indigenous population since 2005 or 2006 if I'm not mistaken has elected an indigenous person as president of the country and the indigenous movement also has a relative good presence at the Kroger's level. However if you take the case of Guatemala with 40 between 40 and 60% depending on the source you are looking at of indigenous population we'll realize that they haven't had an indigenous president but also at the Congress level they have a very low very weak political representation. So basically the point that we're trying to make here at the argument of the paper is that you see the economic literature heavily relies on the indigenous non-indigenous divide when looking at inequalities. So basically we create the indigenous group if it were just one a very one homogeneous group. On the other hand you have the anthropological or more political science literature that points out to a huge diversity a huge heterogeneity within the indigenous population and take this argument to say that is this fragmentation within the indigenous group of the country that explained to a certain extent this weak political indigenous mobilization in the country. So the argument we're trying to advance here is that it's of course this cultural difference within the indigenous population but also the socioeconomic difference within the indigenous population that explain may explain this weak political indigenous mobilization. So let me give you some context for the for the country itself. So Guatemala it's a small country in Central America it's home of some 50 million people as I have said before 40 percent between 40 and 60 percent of the population that depends pretty much on the sources you are looking at it's has self-identified as indigenous. Within this indigenous population we find 21 language ethnic groups most of them Maya descent but however different etno-linguistic groups right in terms of socioeconomic inequality this has been well documented in previous war between the indigenous and non-indigenous divide you will find that indigenous people are the poorest of the country they are mainly rural and they are fine in elementary corporations and in the agricultural sector mostly. When looking at the more political sphere or the political representation in the country we are going to take the case of the 2011 elections because that's the year we're using for the analysis but what we find is that out of the 150 seats available at the Congress only 22 were won by the indigenous leaders. This role speaking is around 14 percent of the seats available for a country where half of the population is indigenous right. So there's a last point that I wanted to make here is that historically the indigenous population in Guatemala has been marginalized and well all over the periods right. Since the colonial periods they have served as basically a very cheap hand labor for the government then with the state with the independence and with the coming of the liberal state that also were forced to work or serve as a work cheap hand labor in the coffee plantations right and what an important point during the history in Guatemala is that there was a civil war that lasted 36 years. It starts in the 60s and end up in 1996 with the signing of the peace agreement. What is important about it is that the moment the war was over and we start counting out the victims of this war we realize that 83 percent of the victims of the war were Maya were indigenous. So during this war period there were discriminating massacres against the Maya population sponsored by the government. So that serve as a context for the country itself and what we want to show here it's basically the socioeconomic differences within the indigenous population. So we're going to split the sample in two. Basically we work with two years 2000 and 2011 and here what you see is the is the educational outcomes of the different ethnic children groups ethnic groups of children in the right. So the ladino that you see here here so the ladino so the first column basically let's focus on primary complete. So this is basically an indicator variable telling us whether the child has finished primary school by the age of 15 years has successfully complete primary school. And when we see that the ladino which is the majority group at let's see 2011 74 percent of them has successfully complete primary school. If we take indigenous group as a whole we see that the average is 63. However what is interesting for us is to see that when we study each of these groups separately you we find a huge difference among them right. So for instance the Kechi children they have been they are worse off than the others right. So just 50 percent of them have complete primary education by the age of 15. And there is a huge difference of course with the Kechi children 73 percent have already complete primary education by the age of 15. So these two are quite similar between the ladino and the Kechi. The Kechi is doing much better than the other indigenous groups in the country right. So again what what we want to point out here is that there exists important socioeconomic difference within the indigenous populations that have been haven't been taken into account in the economic literature so far. When we look at the earnings so for us labor market outcomes will relied on earnings and earnings distributions of the different groups and what we see here is that so this is for 2000 this is for 2011 so the red line is the earnings distribution of the ladino the majority group of the population. The solid black line is the distribution if we take the indigenous group as one and all the others represent the main four biggest ethnic groups in the country right. So of course there's a trend towards greater equality in education as well and labor market outcomes so the distribution here is more centered at the mean it's less dispersed of course. However we see that the indigenous population together as a whole or separately but each ethnic group is always sorry is always falling behind the ladino population right. So this was for just to show some some statistical differences between the groups. So now what are we actually doing on the on the paper is quite simple so we're gonna work with two household representative service so we work with the household service of 2000 and 2011 and what we're gonna do is we're gonna look at three different educational outcomes human capital outcomes so these ones are exactly the same as that we saw in the previous slide so we care about whether the child is enrolled in the school and but is also attending the school or was attending the school at the time of the survey where that will be the first outcome the second outcome will be whether the child has entered a school the formal school system at the official age official and create in a watermelon seven years old so it will be one if you have entered before or until well out for some of your main error to eight years old let's say and zero otherwise right so this is about late entrance and then we look at the school competition so basically here we want to know if children aged 15 years old by the age of 15 they have successfully complete primary school or not. So the models are pretty straightforward pretty standard so we use probit models these are indicator variables and then for the labor market outcomes what we're doing we focus again on earnings differentials so we're gonna see the different earnings distributions of the of each of the indigenous groups at the means so we use what hack a decomposition and then we also want to know whether there are some differences across the the whole earnings distribution so I quickly move to the to the results so here for educational outcomes so first panel is the data for 2000 2011 all of these three are the outcome variables and here what we're showing is simply the effect of the variable of interest the variable of interest is a dummy variable equal to one if the child is an indigenous zero if it's ladino again for the kitchen here or any of the others is equal to one if the child is a kitchen zero if it's ladino so it's always with reference to the majority group to the ladino children right so for instance obviously two two two important things to notice between 2000 and 2011 all these significant differences have already mostly disappear however what we care about is here so let's let's look at a school attendance a school enrollment so if we take the indigenous group if it were just one group one homogeneous group we see that there are no significant differences between ladino children and indigenous children these difference have disappeared however if we do it disaggregated although this is true for most of the other groups we still see that the catchical children are still four percent less likely to be enrolled in the school in 2000 so if we do this for all the other outcome variables let's focus on primary completed we see that at the mean the indigenous group together they are still four percent less likely to have completed primary school however if we split it we have cases like the mom children who are eight percent less likely to have completed primary school so it's double the the mean and then you have the case of the catchical for instance who is who has traditionally traditionally been doing better than all the other ethnic groups in the country right so once again this this difference here although it seems small they are very important for public policy analysis because if you take it just if it were just one group then you are missing all these nuances between the among the different groups so very quickly that the results for the earnings distribution so I'm presenting only the wajaka decomposition so this is the difference in the gap the earnings gap at the mean we have done it for the whole quantile thing but the results are more moralized the same and for the point that we want to make here is first we see so this is in Ketzal's and is in the lock so let's don't focus on the magnitude let's just focus on the fact that there is a positive gap right for all the groups the indigenous groups compared with the ladino and also of course trend towards greater equality the gap has been reduced between 2000 and 2011 and what is important here is the part of explain and unexplained part of the gap so the explain part of the gap is basically what is due to the difference in endowments between the different population is always one indigenous group versus the ladino groups right and then what we see is about the unexplained part and explain it part is what it's usually attributed to discrimination in the labor market right so what we see is of course that again if we take the indigenous group there are huge difference within the group itself right so let's take the case of the Kichigal where the unexplained part of the gap is more than half of the gap itself right it's more than 50% of the gap so somehow there is some kind of labor market discrimination for this specific group the Kichigal when we look at the man however we see that the entire gap is already explained by difference in endowments and this again is a huge it's very important for public policy analysis right so basically these are the results that we have the point that we wanted to make is first the indigenous non-indigenous divide that is highly found in economic literature obscures meaningful diversity within the indigenous population and second one we have showed that there's so important socioeconomic difference within the population this fact together with the cultural difference and ethnolinguistic difference that have already been pointed out in the most political science and anthropological literature these facts together may explain the weakness of the indigenous mobilization in the country good that's all thank you