 I basically pursued the topic of the conference applying for the Brazilian case. I will divide my presentation in three parts. The first one gives an overview. The second one looks at trends, causes, and impacts of inequality. Try to look at different surveys in Brazil to address some measurement issues on inequality. I look at then some interactions between growth, perception, and assets with inequality. I present some recent research we have done on residential capital inequality and inequality data from personal income tax. That's something that's just starting in Brazil. We don't have a lot of data on that, but something is starting to come out. And finally, I discuss the public policies agenda on inequality, discussing a specific case of Bolsa Família, which is a conditional cash transfers program in Brazil. So this is my scripts. So let me start with a big picture of inequality in Brazil and in other countries. So I borrowed here a graph from Branko Milotovic, which compares inequality in each country with world inequality. Here is the US curve, which 60% of the world population is below. The poorest in the US have 60% of the world population poorer than it. It's divided by vintiles. Here is India. Here is China. Here is Russia. So we can say using any poverty line or wealth line that US is richer than Russia. Russia is richer than China. So US is richer than China and India and so forth. So where is Brazil? Here is the Brazilian curve. So basically Brazil is everywhere. The poorest of the poor in Brazil is as poor as the poorest in India while the richest in Brazil are not very far from the top incomes in both Russia and the US. So Brazil is very close to the 45 degree line. So Brazil is a good picture of world inequality. We know that from the per capita GDP that on average Brazilian GDP is not very far from the world GDP purchase power parity-wise. But the distribution is also relatively similar. If we were to draw a graph with the other bricks, South Africa, I think would be quite similar. A little more inequality, same level of income. So Brazil is a good picture of inequality in the world of income distribution in the world. And here is another graph from Branko Milotović and the dots are taken from Brazil. So the line is the trend of inequality between countries weighted by population. So it's inequality between countries. It doesn't take into account inequality between countries. We see after the 1990s a fall of inequality between countries, given the growth of China. And then in the 2000, a further decline in this line because of also India growth. So inequality between countries is falling. And the dots show the trends of inequality within Brazil. That's the Gini coefficient. Sorry. So inequality went up in the 60s during our economic miracle. Went up a little bit more, but not a lot. Until 1990s started to fall. And there is a sharp fall in the 2000s that I will detail here from 2001 onwards. The point I would like to make here is that not only the level of inequality in Brazil captured by the Gini is similar to the world inequality in the beginning of the millennium or now, but the movement is also similar. So the story we've been told worldwide of the Chinese and Indian emerging from poverty into a new middle class or into better living conditions is similar to the one you observe within Brazil. People from the rural areas, from the suburbs, from the favelas, they are going up in a similar way. So the power I would like to make is not only the picture of Brazil is similar to the world, but the change both in GDP per capita is similar. Inequality also similar trend between countries, not the whole inequality in the world. But Brazilian inequality is still very high in spite of this fall. It's number 18 in 155 countries. So it's still a bad picture in terms of inequality. So let me show you a little bit of the Brazilian data. This is the inequality within municipalities in Brazil in 2000, the Gini. That's the inequality in 2010. So basically, let me show you again, 2000, 2010. So basically, inequality fell in 80% of municipalities in Brazil during this decade. This is the map of poverty in the world, extreme poverty, using a $1.25 a day poverty, extreme poverty line. So Brazil is an intermediary color, which gives the idea of Brazil's kind of a mean of the world in a sense. And here is the inequality within Brazil. So Brazil has all colors in it, has poor country colors, rich country colors. So although the mean is close to the world, there is a big diversity, a lot of inequality. It's still in Brazil. That's 2010 data. So this is the first message, this parallel between Brazil and the world income distribution and distribution of poverty within the country. So let's look at the movement of extreme poverty series. So extreme poverty in Brazil fell 69% from 2002 to 2012. It is 10 years, 69%. And this is the Brazilian main target. I'm not talking about the target of the Ministry of Social Development. I'm talking about the target of the whole country. So the main target today is to overcome poverty until 2014, which is today. I mean, this year. And it started in 2010. So this is the Brazil main target, is to overcome poverty, extreme poverty. And what we learn about the Brazilian experience about the combination of inequality and growth, that Brazil is pursuing what you may call a middle path, where half of poverty fall is explained by growth, and the other half is explained by inequality, by distribution. So Brazil is kind of balanced these two parts, growth and inequality, which I like to call a middle path in these 2002 to 2012. The 2012 is the last national household survey available, the PINADGE, which I know is here in the wider data set. And in two weeks' time, we will have the 2013 data available. So this process of fall of inequality with growth, this combination led to an increase in a new middle class in Brazil, which is not like a US or European type of middle class with two dogs, two cars. It's more like a poor, but it's a worldwide middle class, where about 42 million Brazilians joined these new middle class and another 11 million joined what you may call a traditional middle class. So 53 million people in this process from 2003 to 2012 joined these middle class income-based. If you look at the map of Brazil, we see this transformation of traditional new middle class share in the population, 93 to 95, a first increase associated with stabilization. We had very high inflation in Brazil, then was stabilization. Then 95 to 2003, not much change. It was a period of international crisis. Then the period of social inclusion. And now would be the Aquarii 12 projection. We are nearly there. 2012, we have already the data. We are more or less on track in spite of the deceleration in growth in Brazil. So basically, three quarters of Brazilian population from center-west and southeast downwards, at least three quarters in all states, are in these middle class. And as we can see, 93, like 20 years before, the map had a different color. So it gives an idea of the transformation. One data that I like is when you ask people about a serious bundle of goods and services, if they bought it in the last three years, and if it was the first time they ever bought these items in their lives. So this gives from durable goods, on-house car, motorcycle, travel by plane, domestic, international, cell phones with or without internet. So basically, one third of the people who bought tickets, domestic tickets for travel, their travel in these last three years, did it for the first time in their lives. While 64% of the people who purchased a technical course was the first time in their lives, and three quarters of the people who bought their own house formerly was also there. So this gives an idea of this rising new middle class, but also of inequality fall. When you first experience these purchase of goods and services, and some of them are actually investments in terms of vocational education, on-house, not only consumption. So let's go a little bit deeper into the movement of inequality. This is basically the same graph I showed before in a different way, the same data. So these are the long-run trends of inequality in Brazil. Follows pretty much a Kuznet type of process. You've inverted, you shape it. So in the 60s, we had this big rising inequality. It was a period of fast growth. It was our so-called economic miracle in Brazil. Brazil grew at the time something like 13% per year during six years in a row. On average, poverty fell, of course, a lot because of this high growth in spite of this rising inequality. Then inequality went up a little bit more on the 70s, on the 80s. The 90s had a mild fall, and then a big fall from 2001 onwards. So this is the main object of my talk here. What's behind this inequality fall and what are the consequences? So just to give you an idea, the Gini is not, of course, a general way of looking at inequality. It gives a lot of weight to the top of the distribution, not so much weight to the bottom part. But what it means, the Gini fall from 59 to 52. What's the big deal? 52, of course, is very high still. This gives us a better idea of the size of the distribution change. We divided the population in vintiles. So the 5% poorest grew in real per capita terms in this 11-year period, 2001 to 2012, 138%, while the 5% richest grew 26%. So the poorest is growing 550% faster than the 5% richest. So this gives, I think, a better idea of, so Brazil is a country at many years ago, a professor at Marbaixa. Brazil is maybe called Belindia, a mixture of a rich Belgian in the middle of a poor India, big and poor India. So this would be a description of Brazil. I would say today the name still applies, not only because inequality is big, but the poor part of Brazil, as we can see, is growing as fast as India is growing today while the rich part of Brazil is more as growing slowly as most European countries are. So I think in the movement it's still. But in the last survey, we see the genie stop falling. And this is the change of income distribution in real per capita terms by vintiles again. So inequality stopped falling mostly because of this rise in the top 5%. While the rest of the distribution is behaving quite nicely, especially the bottom 5%, which grew 20% in a single year, this may be an impact of the Bolsa Familia upgrades we had in 2011. We'll discuss that later. So the question is, has inequality stopped falling in Brazil? Given this last data, so I brought some data from a household survey just in the metro areas, which only cover labor income, but our household per capita labor earnings. So this is basically the trend of the genie in this period. We can see here a stabilization in 2010, 12, sorry, consistent with the PNG data of inequality. But then there was this back to the downward trend. And in fact, if we get the data from the last six months, moving average, 12 months moving average, we see the genie in Brazil falling at 0.1 percentage points every month like a clock. So inequality is not only falling in the last six months, but falling as fast as this big drop 10 years before 2004. So inequality now is falling very fast if we take these at face value. These are, of course, only metropolitan areas, only labor income. But there is an expansion of Bolsa Família, which would reinforce these movements. This is the Toyota trend, same thing. Of course, the Toyota is more sensitive. So in the genie, we have like a 9 percentage point falling in 10 years and 11 years in the tile, like 20 percentage points in the same period. So this is one idea of the fall of inequality in Brazil. So let's go into the reasons behind the fall of inequality. This gives the growth of the percentage difference in some of the productive attributes of the population, also using the metropolitan areas data. So what we basically see is people having better access to productive attributes, like at least incomplete tertiary education. There is a 19 percentage increase in this period, or people who have a professional course, or people who are formal, contribute to social security or employees, which are formal and so forth. So the first part of the story is access to basic productive attributes. But the other side of the story is that people with these better productive attributes are getting lower income rise. So we have a composition effect. People having access to better productive assets. But the return to these productive assets or attributes is falling. So this is a particular combination, which is not happening worldwide. It's happening in Latin America. There is some work from Nora Lustig. And I heard a new paper, a new book that will come out here. Which showed a similar difference. So inequality, most of the within countries, and most of the countries in the world, is rising. Because returns to attributes are rising faster than the increase in access to these attributes. In Latin America and Brazil in particular, it's falling because the two movements go in the same direction. Normally they cancel out each other to some extent. And so what we basically see when we look at the data is that people with personal attributes associated with exclusion, like women, black people, people from who live in the peripheries in Brazil, which normally are poor people, young people, or people who live in big and large and normally poor households, they are growing above the mean. So what I really like when I look at Brazilian data is that when we normally look at this traditional excluded group, they are doing better in a continuous way than the average, the mean of the population. This is the human face of the Brazilian inequality change. Just one word I know Professor Schorach is here, which has shown many years ago that inequality, especially inequality that we measure in Brazil, which inequality of monthly incomes when it's falling may be due to not only to fall of inequality across people, but also a fall of risk in people's income. So in Brazil what is happening, we have some longitudinal data, short run longitudinal data. What's happening from, so we are seeing the probability of people who are below the median to cross the median upwards and downwards. So basically in 2002, 2003, 26 out of 100 people would cross the median from top to the bottom part. Now it's 13, so went to a half, while the people who crossed the median from bottom upwards was 16, now it's 27. So I don't know if there is a fall or rise in the risk of the population, but the good risk, which is the opportunity to rise has been going up and the bad risk of going down has gone down. And when I find interesting when we run some models, analyzing this process for different groups of the population, what happens is the following. Excluded groups like illiterate people, people who live in the peripheries and so forth, they have still higher risk than the rest of the population. But their risk also fell faster than their included groups counterparts. So in Brazil, good risk is rising for traditionally excluded group, more proportionately than the rest of the population. That's the only advantage of having very high inequality that we have in Brazil, is that we can see this transformation going on. But it's only the beginning of it. So let me look a little bit on interactions between inequality and other topics. So look at inequality across individuals, across horizontal inequality, across excluded groups. Interaction with prosperity, with growth. And to have an idea of this interaction, we don't look only at GDP growth, but also at household survey income or consumption growth. This is a point made by the commission that on the progress of nation by Amartya Sen, Joe Stiglitz, and Fito Si. They made this point. Also look at assets distribution, not only income flows. And one fourth dimension is look at the sensibility of people, how people perceive these change in flows and stocks, not what's happening, how people see that. So this slide shows, I think, a very important feature of the Brazilian characteristic performance after 2000. And first is GDP growth per capita GDP growth. GDP didn't grow that much between 2003 and 2012. It was only a 28% cumulative real per capita growth. When we look at the household survey, the growth is faster, it's 52%. So there is a difference between GDP and national household survey growth. People, now the growth that people perceive in their homes is different from what national accounts see. And when we look at the data, we see that the nominal series are exactly overlapped. What is very different are the deflaters. So GDP implicit deflater has exactly 24 percentage points, more inflation than consumer price indices. And that's what explained that, not nominal series. And this is, of course, the Brazilian case is very different if you look for national accounts, even in the growth part or national household surveys. So there is a big, big difference. Every year, when the same institution, the Brazilian Institute for Geography, releases the national accounts data, normally economists are quite depressed. And then normally two weeks later, when released the national household surveys, the social scientists are very happy. So there is a contradiction. And this is explained by the GDP, the deflator part. And you know, economy is a mean to achieve better life. So I'm a social scientist, too. So if you put a little bit of inequality in the picture, we see, first, the green line is the median growth. And then the purple line is the 10% poorest growth. So basically, the 10% poorest are growing four times faster than GDP, median three times faster than GDP, in this period, the mean twice faster than GDP. So the performance of Brazil depends on what you're looking at and who you are looking at. If you're looking at the bottom part, you have an idea of an emerging country. If you look at the GDP part, it's not an emerging country. So this is the. So let's look at different parts of the distribution. So there is four parts. I'm looking at the mean, the bottom 40%, and then the extremes, the top five and the bottom 10%. So I'm trying to look what explains this growth, of total growth. So this is a methodologist's work done with Nanaki Kakaoani, who was a senior researcher here for a while. He sent his regards. He says, I think in the 80s or something like that. So basically, the total mean grew 3.6 real per capita terms per year, while the bottom 10 grew 6.8 per year. In both labor income grows something like 2.7, 3%. The difference in the bottom 10% is something called Bolsa Família. Bolsa Família makes quite a lot of difference in the bottom 10%, almost double the growth rate. But if we look at the bottom 40%, it makes a difference, but not just 0.8. So it's a methodologist like a legal type of methodology where you add different parts of the growth rates. And so in the Bolsa Família makes a difference in the bottom 10%, also in the bottom 40%, not at all in the mean. Social security is the second most important income sources in the mean. But the point I would like to make is the fact that even though the Bolsa Família plays an important role in the bottom 10%, labor income is growing more in the bottom part than in the mean. So labor income, which gives an idea that sustainable growth is not like Bolsa Família, you cannot expand it indefinitely. So here we'll go very fast. A lot of details here. So the top part we already saw, I'm comparing mean with the share, the bottom 40%. The point I would like to make is what explains the growth of labor income is not so much quantity of labor. The quantity of labor is actually falling in the bottom 40%, but increases in wages, in average per capita remuneration. And this is explained especially by a little bit by hourly wage, by education level, but mostly by education advance. So what is behind the change of income distribution in Brazil is basically labor income and what's behind labor income is education. This is what you may call an education bonus. So the income of the bottom 40% should be growing at 3% a year just because of the increase of education if everything else is constant. And for example, demographic bonus in Brazil is 0.5. So the education bonus is six times more important than demographic bonus in explaining growth in the bottom part. But even in the whole distribution, it's three times more important. So education is what is behind this transformation. The main factor is labor, but behind it education. So this gives an idea of the trends of mean and genie of years of schooling. This is still a bad picture, but a good movement, falling in equality and increasing the mean. But the levels are still low, although better than 10, 20 years ago. This is the proficiency in math. So Brazil, among all countries observed in 2003 and 2012, Brazil had the lowest proficiency in math in 2003. When we look at the change, 2003 to 2012, this was the highest change. So the same story applies. Still bad picture. We have just like four countries behind us in the pizza, but it was a big movement. So of course, the advantage is when you have bad social indicators that you can advance them faster, but doesn't mean that you will. So look at the Human Development Index by Municipality. This is a work done by Ipeia. This is the red part is the municipalities with very low Human Development Index. So to give a more structural picture. So 41% of Brazilian cities had very low 5,500 Brazilian municipalities had very low Human Development Index in 2000. In 2010, 0.6. So this gives an idea of the change. From 41 to 0.6 in 10 years, life expectancy rose something like 3. something years in these 10 years. So there's a relatively deep transformation. When people look at the Brazilian case and say, well, you have Bolsa Família, this is not the major part. There is a structure transformation going on. And Bolsa Família helps in this structure transformation as well. But this explained by education, by health, not only by income. So let me have a quick look at residential capital. Residential capital in Brazil and other countries is 50%. So I'm looking a little bit at the distribution of assets. It's about 50% of families' wealth, physical wealth. As Piketty's book shows, shows very... So what I would like to show for me this was a surprise is that there is a big rise in the values in the residential part and residential capital in the bottom part of the distribution in Brazil from 2003 to 2012. The red is the mean, 37%, which is lower than the mean income growth. But if I look in rural areas, people who live in cities but far away from their work or people who live in precarious homes, these home values went up. And normally, poor people live on their own homes. Precarious homes, but their own homes. And so there is this increase in the value of residential capital. And if I look in the characteristics of the people of the location, I see, for example, people who have low income, low family income, residential capital rose 90% while the mean 37%. So what I would like to show, to stress here, is the change in the distribution of residential capital in Brazil, which is also within households, there was an increase in durable goods. The increase in the coverage of public services outside homes didn't go at the same pace. So homes and what happened? Within homes, there was quite a big improvement, but not so much a public services supply. This picture gives us the difference between the lowest curves of residential capital and income, the top ones between 2003 and 2012. They cross each other, but they are more or less on the same level, the change in inequality. And the bottom part shows us that the fall of inequality in residential capital from 2009 onwards was faster than income. So there is a structural part of these not only incomes, but also capital, not only human capital, but also residential capital. There is, I think, an interesting aspect on popular housing programs is the fact that smaller homes gives us a higher yield than bigger or more expensive properties. If I have money to buy $100,000 home, I get a better rent out if I have $250,000. So there is not only equity and sustainability concerns, but also efficiency concerns. So we look very fast in the personal income tax record. A lot of people have been looking. Thomas Piketty, Tony Atkinson, among others, we don't have this data available in Brazil. So this result just came out now in August. So the first point I would like to stress is the following. If we are having a problem, the fact that national household surveys is growing more, is growing faster than GDP, the problem here is much bigger. Net incomes according to personal income tax is growing 50% in five years. Now I'm using 2007 to 2012. That's the data I have. So in five years, it grew 10% a year, basically, or 50% in five years. The PNG growth, the national household, was 23 on per capita terms. So it's growing much faster. Part of it is explained by the fact that one is total, one is per capita. If I make PNG total, it goes up to 29. What I think is behind that, and I have many doubts about that, is the fact that in Brazil, there's a big movement towards formality. In 10 years, formality grew 10 percentage points from 50% to 60%. So people are entering the income tax. The number of people who declare income taxes didn't change in this period. But if I incorporate the two movements, the movements of population size and the formality, just applying simply, the PNG income growth would be 53. So maybe there is no contradiction. 53 against 50, and the fact that the people that don't pay taxes' income is going faster. So this would fit the two parts. So the point here is personal income tax, total income, not only taxable income is increasing quite a lot in Brazil, and part of it is in formality. It is because of falling formality. We don't have yet the top incomes measure. OK, but if we divide a population in income in the three groups, traditional classical groups, capitalists, henchiers, and workers, what happened in this period? So what happened basically was a rise in the capitalist share, 1% point, a rise in the workers' share in 2% points, and a fall of henchiers' shares in 3% points in this period. And this is more or less consistent with national accounts. So there is this net total income. We don't have the micro data for income tax available yet. We have, I don't know if I would call a secret room, but a room that allows you to process secret data. It preserves people identity. So we have that in a payer. We are working on that. We don't have the results. But we have a few tabulations. First one, a very nice one, which splits people by their occupation, main occupation. They declare an income taxes. So we can calculate the between groups in equality. So what happened is these between groups in equality fell from a genie from 0.37 to 0.28 in this period. So it's a big effect and a big change on a big effect. So this would be a sign. I don't know what's total inequality there, but maximum one. So it's a big transformation because it's the genie. 22% fall. The second one, I also have tabulations from the states of people. And when I look the transformation of the genie of different inequality measures in this five-year period, I see a big change, 31% or 15% in the genie, but on a small basis. It doesn't explain a lot of inequality. But the between groups component, the between states component, but there is a reasonable fall. So these are some initial evidence on. So let's look a little bit on sensibility. I show you, I know there is a world ranking of happiness from Gallup and other data sets. And Mexico is doing very well in present happiness as Nordic countries do traditionally better. Traditionally well. I use here different evidence on. Not like life satisfaction in five years time. This is the world map according to the Gallup world also Brazil, Mexico as well, South Africa. We are in the group with US, Canada, Australia, New Zealand, Nordic countries of very high future life satisfaction. The question is give a great your life in five years time. In fact, Brazil is number one on that. Not only in a single year is every time the survey went to the field, we get the first place. So that's why Brazil is known the country of the future. Give a great your life in five years time. Our average grade is 8.8 on average, which may be not a good sign. We are too optimists about our future, perhaps not good for education, not good for savings, but that's how we are. And inequality, although we are still very equal, is also very low. Everyone give a very high grade poor people, rich people to their future life satisfaction. So it's the second lowest genie just behind Belgium. What's the problem with Brazil is the fact that when people rate their lives in the future, give a very high grade, but when people rate the country in the future, give a very low grade. So there is some inconsistence. Everyone is above the mean, which is an impossibility. So, but I think capture a little bit of the Brazilian problems. We have a collective problems. We rate the country much worse than everyone thinks, oh, my life will be great in five years time, but the country, oh well. The country, so I think the two, the optimist part and with a lot of equality between expectations, but also a consensus that the country won't do so well. Okay, so, so I think here a sensitive issue about demonstrations, okay. We had demonstrations, big demonstrations in Brazil in 2013, just during the Confederations Cup. This, of course, for those who were in government, big challenge. And so I think this maybe help us to understand a bit the political process behind that. So I have four columns. The first one, if we'll participate in demonstration, the second one, you didn't, but you wanted to. The third one, you didn't want to participate, but you agree, and the last one, you disapprove. So let's look at the extremes. So as expected, people who joined the demonstrations, which were like 3.5% of the population, were young people, males, compared extremes. With higher education, much higher income, like household nearly twice, heads income or per capita income, 40% higher. So what I'm saying here is that people who participate in demonstrations are not the, perhaps in all demonstrations tend to be like that, are the top parts, especially in education. And if I run a model to select what explains more the participation in protest, the main variable is access to social networks. Okay, so there's a new technology. But there are other evidence, as in the previous table, that people who did participate are those who are better off and look at the data that I just showed you, are those who didn't do so well in the previous year. You know, like people in the Northeast, which is the poorest area of Brazil, they tend to participate less. And for the people who work are more prone to participate in the demonstrations. So it's different from the Wall Street, Occupy Wall Street movement, where was a protest against inequality by, I believe, mostly unemployed people in Brazil, was people who were working, highly educated, but any inequality in Brazil, as we've seen fall, of course there's many causes behind such a big social movement. First one, new technology, social networks. There's also variables related with use of public services. Or if you rate badly, public services is behind the demonstrations. It's also part of the demonstration, young people, but I believe there is also a, I cannot reject the fact that the movement income distribution, people who are in the top part, didn't do so well, were, did join the demonstrations more strongly than those who benefited more. This is so, it would be consistent. And two questions, qualitative questions. If you think that incomes of the people who are poorer than you improve it more, you are more likely to join the protest. And if you're against Bolsa Familia, also now reinforces your participation. So, this is the, and if you take part in the Confrederation Cities, which is a kind of, no, they were getting public resources but they perhaps wanted the public resources for education, for health, not for building stadiums. So, this is the, so let me enter very briefly in the Bolsa Familia part. So, this is present and past happiness, life satisfaction of people with and without different programs. So, the point here is that Bolsa Familia beneficiaries had lower past happiness and before and now they are relatively close to the mean. So, there is a reducing inequality of life satisfaction. And when I compare it, people I see that in fact, Bolsa Familia beneficiaries had a bigger jump in between present and previous happiness. So, very briefly I think my time is running out. So, these are the concentration curves of different programs. Bolsa Familia is this different curve that's like in another planet compared to other incomes. So, it's very well-targeted, the concentration. So, Bolsa Familia reaches 25% of Brazilian population at a cost of 0.5% of GDP. And why it affects people because a well-target program. It goes to the very poor. So, 0.5% of GDP makes a difference in lives of the one quarter of a Brazilian population. It also has some prosperity effects through channels of, you know, you make the wheels of the economy turn if there is idle capacity, so if we calculate the multiplier effects of different programs because Bolsa Familia go to the poorest that tend to spend most of their income or higher share of the income, the multiplier effect is 1.7, while other programs is much less. So, if there is idle capacity, this may be a problem for monetary policy when the government wants to hold. If Bolsa Familia, the economy, if Bolsa Familia is expanding, it acts in the other direction. So, just looking very briefly at some characteristics of Bolsa Familia. This is the income distribution. This is how Bolsa Familia works now. Basically, the benefits are proportional to the income gap using the extreme poverty line. So, if you are poorer, you get bigger benefits. And now we have something we may call social federalism where different states and cities are building programs using the Bolsa Familia infrastructure, tailor-made programs. So, in Rio City and the state of Rio, we have such a program where they use not a $1.25 a day, but a $2 a day poverty line. They also filled the gap and they don't use the clear income, but they use a permanent income estimate, a chronic, you know, taking into account assets, not only reported income. So, just to give you an idea, this is the card of the typical card of a state program. So, we have both the state and the federal. Without Bolsa Familia, extreme poverty would rise 36%. So, this is more or less the size of the Bolsa Familia impact. Without the indirect effects, okay? And this is a simulation of the impact of Bolsa Familia according to age. So, there's a big effect and there is a recent change where you give, you increase the number of children, are eligible to the benefits, you increase the value, et cetera. So, if everything, that's the potential. So, there's a big change in poverty among children in Brazil. We know this is like that from previous data and this is a simulation, okay? So, we have still to check. So, I know my time is running out. So, basically we have different, a new agenda of changes in Bolsa Familia. So, Bolsa Familia's infrastructure has a very good administrative record that covers one-third of the Brazilian population. So, it's a census, an operational census of the poor and on top of that, you can construct very quickly, very fast, different upgrades. That's the nice. So, the Bolsa Familia is only the beginning of a new generation of policies towards the poor. So, for example, we use now in Brazil the $1 a day poverty line, $1.25, $1.25 a day poverty line. That's interesting because it gives us a neutral ground for states, cities, and federal government to cooperate from different political parties across different mandates. So, the emphasis have been recently in some states parades engagement. Having a conditionality, which is a means conditionality. Early childhood education. Incentives to perform, especially aligning teachers, parents, and students incentives. The Bolsa Familia models play a key role. The 93% of the cases were the ones who get the benefit in their hands. The main message I would like to give you is that once you have a system established, you can construct programs very in an industrial way, affecting millions of people, a big share of the population, very in a very fast way. I'm talking about matters of weeks until the money or the conditionality is established. So, this is the... So, we have a big agenda, don't have time to go. There is the direct effect on resources. There is a direct effect on public services. Children of the families who are poor, they are first in line to have access to nurseries, to early childhood type of policies. So, there is fear, the budget constraint, the direct effect, the well-being direct effect, production function type of effect. That's not how clear this effect is operating, we don't know yet. And here, there's a new agenda on giving more decent-parked markets to the beneficiaries of Bolsa Familia through financial education, through better regulation, consumer protection. So, there is a... Here, I have a book, I'll leave a few summaries in English of a book we just published on Bolsa Familia, so this might interest you, gives you the details of this. So, last slide. So, what we have learned first, the four interaction questions on the quality of development. First one, inequality in Brazil, income inequality is falling, is still falling. Second one, GDP is growing less than household surveys on the mean. So, Brazil has been doing better according to surveys than to national accounts. That's a key difference. In different ways, it's perceived by the people, people in their life satisfaction, although they're more critical about public services, the quality of public services. To some extent, this process is sustainable because we see rising stocks on education, also labor incomes is rising, human development index, housing. So, it's a change in the basis of the asset distribution. So, this is what we make always structural, a deep transformation. I think people tend to look at the Brazilian case and we Brazilians tend to look at and say, well, this is not such so structure because Bolsa Familia is behind it. Bolsa Familia plays a role, but the major role is playing by education, by labor, and by other assets, life expectancy, housing. So, this is why I'm positive about the perspectives. And perhaps the biggest test of sustainability is the fact that Brazilian, in the last three years, Brazil started to grow much less than before and social indicators, including income-based indicators, are still doing well. So, there is a big difference. And so, this is a sustainability test. Of course, big challenges ahead. We don't know what's going to happen. So, this gives you a map of the databases used in the presentation in Brazil. And here, I prepare a final slide with different links to papers related to the Brazilian case that I participated. And also, here's some databases which might be useful for having a more detailed look at the Brazilian case. Okay, thank you very much.