 So this is a paper joint with Maurizio Bussola from World Bank and Fabian Herzl with a PhD student at the Paris School of Economics. And so this is an empirical paper which we produce, we provide some estimates of inequity opportunity in some South Asian countries, okay? So motivation. So just to say one thing on this. So there are three main reasons one can be interested in inequity opportunity. One is purely normative, you know, there was a big discussion in the 90s and even before among philosophers, political philosophers and normative economists about what is social justice and why we should move when thinking about social justice from the space of achievements to the space of opportunities. So from, let's say, a well-pharistic approach to a non-well-pharistic approach in which the process by which a given distribution is obtained is relevant. So this is an intrinsic reason if you want why equity opportunity is fair. The other is instrumental, okay? Equity opportunity is interesting because in equity opportunity can be a buyer to economic growth. There have been different ways in which this has been specified and declined. There are different arguments in equity opportunity can go against a good, inefficient allocation of talents, can be against the good incentives in a market economy. But, you know, and then there is also another argument which is that in equity opportunity, the equity opportunity is not only the conception that is proposed by philosophers and economists, but it's the conception that is very popular among normal citizens, okay? So we should take into account this. And so now the real question is, now are all these arguments valid for poor, I mean, no rich, no Western countries? I say that because most of the leisure show both the normative leisure show and also the leisure show which is trying to elicit the conception of justice of individuals is based on most of it on Western rich countries. So the problem is, is this relevant for poor countries? Our answer is yes, okay? I don't have time here to say why, but in the paper we discuss this, okay? So I think it should be discussed, okay? Why? Because some important economists as you know, pose this important question. Why we should be interested in equity opportunity in a country where there is an enormous level of poverty and extreme poverty, why we should be interested in measuring equity opportunity and not just deprivation and poverty and total inequality? So there are some arguments. I think this issue should be addressed. Yes, there was a session here, I think here in which this thing was discussed. Now, this is the region we studied in those of the countries, Afghanistan, Bangladesh, Bhutan, India, Nepal, Pakistan, and Sri Lanka. Now I try to speed up the presentation. Now what we do, so the main effort here was to work on the data, okay? To obtain the estimates and to obtain comparable estimates, so comparable across countries. Then to capture trends, we estimated equity opportunity across birth cost, okay, in different countries. And then we try to understand what is the contribution of the different circumstances to equity opportunity by sharply decomposition. And we try to exploit the co-resident people to infer the parental characteristics. And we discuss the co-resident bias in this context, okay? Now, a preview of the results. The court-based analysis reveals trends which are hidden in construction analysis. You will see that this is very clear. We basically measure equity opportunity into two dimensions, education and consumption. We explore also the income, but at the end, we prefer the household consumption, so education and household consumption. And what are the basic results that there is an reduction of equity opportunity in years of education in almost all the countries. And within heterogeneity in improvements and levels, there is not a clear improvement in equity opportunity for consumption, okay? And our conclusion, I mean, a provision conclusion for the co-resident data is that the co-resident data are only representative for narrow age group of general population. If you want to apply the conclusion you have for the co-resident data to the full population, there are really big differences, so you can't really do that. I mean, there is big bias. We try to quantify that. I will not present much here, but we do it in the paper. So we already have seen the model equity opportunity by Chico, so I go quick here. And so what we do, we do here a parametric. So we just measure in equity opportunity in the ex-ante approach. We do both parametric, which for the continuous variable will be linear model. And we also have here some dichotomous variable, which is electricity. And there we use probit analysis. And for the measure of inequality, we use the genie for the continuous and the dissimilarity for the binary variables, okay? And we compare the results we have with two different approaches, the parametric and the machine learning. I will show not all the results we have, but I mean, most of it. Now, these are all the surveys we use, okay? For all the countries. So I'm not going to, and for all the countries, we have, for most of the countries, we have different years here, okay? Of the same survey. So for the cost analysis, what we do, we pull together all the surveys, okay? And then we identify different costs on the interval of five to 10 years, okay? And then we follow those costs. So we are able to see the trend. Basically, we do that for the educational variable. That it's easier because for the education, you don't have life cycle effect that you have for the consumption, for the income, okay? So the analysis is easier. Okay, so the outcomes for the individual level you have here, so education and electricity. You also have individual level of income, but we will not show these results. At the household level, you have consumption per capita for countries, you have income per capita. And of course, you know, by taking the household level, you underestimate the gender dimension. And so we take into account that when we measure the contribution of the circumstances to integrate opportunity. As for the circumstances, what do we have? We have gender, as I said, but which is underestimated given that for the consumption, we use household level consumption. We have your benefit of residence as a proxy of your benefit of birth. We have geographical vision of residence as a proxy for original birth. Now we have some statistics to see how big is the migration from one region to the other is not that big. So our results are quite robust in that sense. Then we have a variable which we call demographic group. I will show you later. And then we have parental education. Now, for most of the countries, we have gender, urbanity, geographical region, demographic group. For some countries, for countries specifically, we have also parental education, okay? So we have basically two different set of circumstances, a limited one and an extended one, okay? So the extended one, we have parental education. And then we try to measure in a greater opportunity with the extended circumstances, also for the countries for which that differential is not present in the survey by using co-resident data, okay? So let me show this. So this is basically for each circumstance, you have the country for which the circumstance is available in the data. And here for the parental education, you have the four countries for which the parental education is in the survey, which are Bangladesh, Bhutan, India, and Nepal. And these are the other countries, so Afghanistan, Pakistan, and Sri Lanka, for which only the co-residents analysis will allow you to use the parental education, sorry. So let me just spend a few words here. So the demographic group variable is a composite variable, which is also different from country to country, okay? So here there was a kind of compromise, okay? Which is, in the words of Chico, it's certainly arbitrary. But you know, arbitrary doesn't mean that they cannot be justified with a reasoning. Then the reasoning should be compelling, it should be in a sense convincing. So what we have basically here, we have a variable which takes very different meanings in different countries. So for instance, in India, you have a variable with six categories in which you mix the caste and the religion. And for the Pakistan, you have a variable which basically depends on the language for Sri Lanka, the ethnicity. So this comes from the specificity of each country, okay? So the compromise solution we have is that we have a model with the same number of circumstances, so covariates. But that covariate, one of those covariates, so what we call demographic group, has very different meanings for the different countries, okay? Reflecting the different social and cultural structure of the different countries. That was in a sense, you know, the compromise solution we had. And okay, let me show you the first results. So first of all, we start with education, then we go to consumption, education. This is the full inequality. Sorry, the picture is not really focused, but anyway. So this is the total inequality in education. This is inequality in years of education. So the genie of the years of education. As you expect, the values are very high. You know, the years of education is a distribution with a few values, so it's very easy that the genie increases. And you have a clear decreasing trend for all the countries. Here are the birth co-authors, okay? So we identified the trend because we have the different birth co-authors. So from those born in the 60s to those born in the second part of the 90s, okay? So for each birth co-authors, you have inequality in the years of education. And you have that inequality decreases for all the countries, okay? And in a sense, this is confirmed when we look at the portion of inequality in education which is explained by circumstances. So also inequality of opportunity in education has been decreasing for all the countries from the oldest to the youngest co-authors, okay? So there is a decreasing trend. It's the strongest decrease in some countries like Bangladesh and the Bhutan. And here you have also India, unless pronunciated in some countries like Sri Lanka and Nepal, but the decreasing trend is everywhere. Now, this is with the limited circumstances, okay? So we take everything but the parental background. If we add the parental background and here for all the countries, so the co-resident sample, you see here the triangle is inequality of opportunity adding the parental background. The safe call is inequality of opportunity without the parental background. As you expect inequality of opportunity increase when you increase the circumstances increase but increases of a big amount in most of the countries. And you can appreciate where the parental education plays a big role in the different countries. And this is also true when we do for the non-co-resident data. So for the few countries, which are Bangladesh, Bhutan, India and Nepal, for which we have parental education information in the survey. So we don't have to resort to co-resident data, okay? Now, if instead of taking the co-authors, we had taken just the cross-section, so to see what is inequality of opportunity in the different years in which the country has been observed, we wouldn't observe that decreasing trend, okay? So, yeah, that says something to our way to measure the trend. And as I said, for the education, the working with age groups is quite safe because for income and consumption is more complicated. So you have to decompose the life cycle and the cost effect is more complicated. But, can you tell me? Yeah, okay. I was expecting something. Literally, share is consistent with what we saw with education, okay? So it's an increasing trend and also the dissimilarity index says the same thing. So I will not spend time here. So let's go to the consumption. So here you have the total inequality in consumption, okay, for the different countries. So we start from, let's say, let's look Bangladesh from the oldest co-authors, 60 to the youngest co-authors, 90s, okay? And you see that there is a decrease in total inequality in consumption. So this decrease is common to almost all the countries we observe. So for different co-authors, you have an inequality decreasing trend in the four co-authors we observe, okay? And when we look at the portion of inequality that is explained by Sivkunsansi's inequality opportunity, this trend is confirmed in some countries. It's not that here in some other countries. But one thing which is, I find also interesting here, if you compare the previous slide with the second one, is that the ranking of countries in terms of inequality and the ranking of countries in terms of inequality opportunity is not the same. It's not the same at all. Here you see that India is the country with the lowest inequality in consumption and is among the countries with the highest inequality opportunity, okay? And this is also true for Afghanistan. It's among the countries with the lowest inequality, but among the countries with highest inequality opportunity. So the ranking's changed quite differently. And so it changed the ranks of the countries, it changed the trend over the co-authors. Okay, these are the results I summarized here. And yeah. Let me go. So the contribution of safe consensus to inequality opportunity, first education and then consumption. For education you see that you have the sub-region, the region and you've been, which all together is basically the location of individuals plays the biggest role. So explain most of it. So most of the inequality opportunity we observe with this data is due to the location of individuals. So sub-region, region and you've been. Here we don't have the parental background, okay? Because this is the common set of circumstances, common to all the countries. And we have a smaller role of the gender. What is important is that when you go from the oldest co-authors to the youngest co-authors, yeah, you have a decrease in the role of gender, you have an increase in the role of the location of the region. So the regional inequalities basically is increasing in explaining opportunity inequality across co-authors in most of the countries. What about the consumption? For the consumption you have again the same results in terms of magnitude of the location variables with respect to the rest, but you don't have that clear trend in the increasing role of the location that you had for the education or stuff. Okay, the last thing I want to show is the relationship between total inequality and inequality opportunity. Now, we also seen I think yesterday some graphs in which we plot total inequality and inequality opportunity. If we do it with education, we have what we expect now. We've seen many of these graphs in the lecture show. So there is an increasing relationship between total inequality and the relative inequality opportunity, okay? So those countries in which there is higher inequality are also those countries in which the portion inequality explained by circumstances is increasing. And if you do it with consumption, we don't have any evidence of that, okay? So we just have, yeah, we don't have any, this is just a correlation, but we don't have any kind of correlation. And okay, I think I will stop here. It's zero. Thank you. Good. And yeah. Yeah.