 Let me first say how happy I am to be here in the ILO. I mean, I worked, as Tony mentioned, in several agencies, UNCTED, WIDER, UNICEF, Economic Commission for Europe, briefly, very briefly, for CEPAL. And I think that my life has been blessed, because much of the economic thinking in the United Nations at large originates from these agencies. And I remember that while I was working 76, 77 in the Economic Commission for Europe, I was spending most of my time here with Rolf Wouter, because the ILO was then the leading agency in the development debate. And then, in a way, one could write a book, and then I moved to UNICEF. And then UNICEF, basically, became a very important actor in the development debate. So one could write a book on the rise and fall of the leading position of United Nations agencies in the development debate. But the fact remains that in the UN, there are five or six agencies which are producing original thinking. And thank you, Isabel. Thank you to the Director General of ILO for hosting this conference. And I think that it is very nice to see many friends here. And in this way, I feel like I'm in my own family. And in this family, I would like to bring in a little bit of optimism. And now the debate is dominated by Thomas Piketty's book, which says that inequality is rising everywhere. And this is not true. It's rising, basically, in the OECD group. And this is how the country is mainly analyzing. And in order to, if you want to change things, well, you can start from first principles. Or you can look around and see if there is anyone who has done things otherwise. And Latin America represents, in a way, a good example of how policy could start to be built. And nothing terribly radical, but things have changed. And then this is how they've changed. You see, the income inequality in Latin America was, in the early 1980s, the highest or the second highest in the world together with one part of Africa, with the Eastern and Southern Africa. And during the period of, during the lost decade, the inequality kept rising again already from high levels. And then during the augmented Washington consensus, which is the 90s, more or less, it rose again. And the turning point is around 2002. And then we see that this is the average for the region, but it's not an average with huge variance. Actually, it's an average with low and declining variance. And so we see that from about 2002, you see inequality starts falling. And it falls very much. It falls basically to, in about eight years, it goes back to the level of the early 1980s. And the question is what causes that? But before we do that, I mean, not all Latin America experienced a decline, but 15 countries out of 18. So basically, the vast majority of these countries, Central America had a much smaller decline in inequality, South America, and particularly the Southern Corn, a much bigger one. So you go from Nicaragua, which had an increase in inequality of two points during the period 2002, 2010. And then you go to Argentina, where inequality falls by nine point, and Brazil as well. So two very large economies experience a decline in inequality. Now, one could say, well, there is a debate which says, well, that has been, inequality has been declining in Latin America during a period of high commodity prices, easy access to finance, an increased number of Latinos work moving to Spain, to Europe, to the United States, and so on and so forth. So basically, it's a cyclical effect. And the question is, is it cyclical? So what we did, we plotted also the changes in inequality between the year of the crisis, which is 2008 and 2012. And basically, we see that inequality keeps going down. Perhaps there is a slowdown in 2012, but this is a panel of 11 countries. They are the largest, but they are not all of them. And then if you take other measures, basically, you see that inequality goes down as well. And so it looks like that the decline in inequality in Latin America is more structural than cyclical. Now, during the recent conference in Helsinki, I think it was a month ago or two months ago, there was a Marcellon area was the Brazilian minister for strategic affairs. And basically, it gave us the data which are updated until July 2014, so until four months ago, three months ago. And that means this is wage inequality. It's not income inequality. But then it concerns only five or six of the main cities, but it's quite telling. So it looks like that even up to 2014 for the largest economy of the region, inequality is falling. Now, one could object, but you know, Tony Atkinson, Thomas Piketty and Manuel Saez and others, they are saying, well, these are data based on household budget surveys. And we know well that the distribution you obtain from this information is truncated. The top people often they opt out from the survey. They saw, so if I'm a rich man or a rich woman, I can say, no, no, no, I'm not obliged to take part into the survey. And the way now people will go around that, including Piketty, Tony Atkinson, Facundo Alvaredo and others, I mean, basically is to reconstruct the top tail of the distribution using tax returns. And fortunately, we have data for three countries and for the last 10 years, which are Uruguay, Colombia and Argentina, which show that you see the dotted line is the genie corrected by the information concerning the tax returns. And you see that it is higher, quite a bit like in the case of Argentina, five or six points. But you see that the trend is not altered, you see. So there is not a change in trend, even if you use the genie distribution corrected for the share of the top 1% derived from tax returns. So when we find this information, so there is a level effect, but there is not a trend effect. Now, so we are fairly sure that what we found tends to be, stands to prove of different data sources. Now, since, I mean, there's been quite a big debate in a new book which we did in honor of Sir Richard Jolly, there is one chapter by Robert Wade. And the chapter says, the title of the chapter is, why is income inequality neglected so much in the current debate? And then he quotes four or five people like Martin Feldstein and Margaret Thatcher, I mean, some of the neoliberal people, including a Dutch economist who is basically more English than Dutch now, Boijter. And Boijter says, I really am really sorry about people being poor, but I don't care about inequality. So there is a debate at what the target is, is that you have to reduce poverty. And then if you do that through unequal policies which are unequalizing, it's also worrying. In the case of Latin America, this is the proof that Boijter is wrong. And if you take what has happened more or less during the last 10 years, you see that the dark bar represents the decline in poverty, which is due to redistributive policies. And the lighter bar is the amount which is due to growth. So if you are in a country with very fast growth, but rising inequality, poverty will not fall too much. And so here during this, let's say during the last decade, I mean it varies from period to period, from country to country, but basically 40% of the poverty decline is due to improvements in the income distribution. So even if you want to take a narrow perspective of focusing on poverty, then inequality is a reduction in equality is a sanction. Now the big question is what explains the inequality in Latin America 2002, 2012, and then if all the countries are like Brazil, then 2014. Now I think that there are different views. I mean if you talk to like Ricardo Hausmann who is a well-known Latin American commentator, he says luck, you know, in the income inequality has fallen because the market was characterized by high commodity prices and all that, you know. Other people they say, well, no, it's not really luck, but it's growth because Latin America during the last decade more or less had a rate of growth which I think is, Fredo, correct me if I'm not wrong, but it's slightly higher than even the period of the income substituting industrialization. So it's around four, five percent. But five percent growth is not necessarily leading as we shall see to falling inequality. And the third explanation is public policies. And this is the one which basically I tend to favor and I try to demonstrate why this is true. Now if you want to be technical, I think it is quite complicated to prove convincingly what you have to say. So if I have to go to the meeting of the hard-nosed economist, this is what I have to say. Now, what type of information can you use? Well, you can use micro-surveys. So household budget surveys. But the household budget surveys, they provide some information. Provide information basically individual characteristics. So family characteristics and all that. Don't tell anything about exchange rates, public spending, they tell you hardly anything on policies. But they provide you accurate information on what happens at the micro level. So what we did, we decomposed point-to-point two surveys between let's say 2000 and 2010 for six countries, which you will see that I mentioned. So in this case you are able to determine whether the decline in the genie or the rise in the genie, I think we selected also a country with rising inequality is due to what? Changes in the distribution of wages, change in the distribution of capital income, change in the distribution of transfers, and so on and so forth. And then if, since these incomes are distributed in a very unequal way, then of course if you have a fall in the capital share, if you have a rise in the capital share, everything else being equal, this will be unequalizing, it's quite obvious, you know. But then in the end then we don't get the political economy that both my predecessors have mentioned there. And so I find many of these decompositions done by people of not my age, normally they're about 27 year old or something like that, sorry for the younger one. So basically they ignore the political economy. And then when you want to get high scores and doing your PhD thesis in American universities, what you have to do. But then that does not help you much in terms of deciding what public policy should be. So then we link the information coming to that today, what we call the underlying causes of income inequality based on economic theory, political economy, pundit regression and so on and so forth on this. Now then of course we compare the results of the two, they fit and moralize they fit. Now the decomposition is done this way, I don't want to bother you, but there are some algorithm like one is by Branko Milanovic which was cited by Isabel in which the changes in the total gene is equal to the changes in the concentration coefficients of the different sources of income, labor income, capital income, transfer income, remittances income and changes in their own total share and total output. So in a way we combine the micro distributions with changes in labor share versus capital share. And so it seems complicated, but it's quite simple. Or we can use the Lerman and the Tzaki which is very similar. So you see we basically rely on six sources of income, UW is Unskilled Wages, SW is Skilled Wages, R are Rends, then RK are Capital Profits basically, then TR are Transfers and RE Remittances. And then it's good that we have data also on that. And what does it come out of all that? Now here I plotted the data for two out of the six countries and then we took Chile and Ecuador. So and we have the 90s in which inequality rose and then the 2000 in which inequality fell. And we wrote the political regime which was in power and then the period exactly considered. And then we saw that in Chile during the liberal period inequality actually was Christian Democrat. Inequality rose a little bit, but it fell much more during the second decade. And then we saw that the absolute changes in not in the total genie, but in the wage genie basically behave in a similar way. It rises during the first period then it falls during the second period. And we calculated again what was the skill premium. So the ratio of the wages of skilled workers to unskilled workers. And so we see that the point that Richard mentioned that and which was accepted by the IMF that the skilled wages basically rise faster than unskilled wages during the first period but they fall massively during the second period. Then we look at whether there is a change in the urban rural wage. Now in Chile, Chile has a very small proportion of labor in the agricultural sector, so it's not important, but in Ecuador it is. Then we look at changes in the genies of the other income sources, capital income, public transfers and remittances. And we see that in the case of Ecuador that remittances are equalizing, which is not to be taken for granted because all the literature so far has say that they were unequalizing. So basically this is what we have done for each one of these countries. And this again is a similar decomposition done by two of the authors, Caveman and Mauricio Maurizio. And then basically it shows that the decline in total gene inequality, 73% is due to improvement in labor inequality, 24% in the case of Argentina. 24% to improvement in the distribution of pensions during this period, public cash transfers are unequalizing and other non-labor incomes are equalizing. So this has been done for about six countries. Now if you take the case of Brazil, it is interesting to see that the story of the Bolsa scholar, Bolsa Familia, is quite relevant because the Brazilian basically say that this increase in transfers like Bolsa Familia and Bolsa Scholar and all that explains about one-fifth of the total decline of a huge decline in income inequality. So like three points of genie would have gone down because of a small transfer because the Bolsa Familia is very small. I mean altogether it's about less than 1% of GDP. So it looks like a miracle. So we have always been slightly unconvinced. But anyhow, this other table confirms the results of the prior table. What are the overall decomposition? Well, as I mentioned, is labor income falls. And then we say, well, in parallel to the decline of labor income inequality, there is a decline in the skilled premium. That we analyze it as closely as we could, but that is a rather complicated story. Well, it could be to the stagnant demand for skilled labor. Basically you have to build a supply demand of skilled labor and skilled labor. And in a way, one could argue that there has been a rapid increase in income inequality because of the major improvements in technology. I mean adoption sophisticated technology during the 1990s which somehow levels off during the 2000. This could be an explanation. Now, then the second there is an increase in demand of unskilled laborers. Well, because why? Well, because there are policies increasing the demand for this type of labor, particularly in the trader sector, which means agriculture. And if you look at some of this country, basically international commodity prices, they rose. And so the wages in agriculture, they rose because demand for labor rose in agriculture. Some other people they say that the inequality, I mean the decline of the skilled premium has to do with the worsening quality of the higher education. So that the people who come out with a degree, they receive a lower wage in the market because the universities are not as good as they were before, you know? Now, the one which we tend to consider is the one which explains the most in our view is the one in red. So there is a rising supply of skilled labor due to higher spending on education in the 2000 and also in the 1990s, you know? And we'll come back to that. Now, the other point is that Latin America has gone through a very rapid demographic transition and an increase in immigration. So that if I have a greater proportion of young men and young women going to second and tertiary education, if some of them they migrate. And the birth cores are smaller than the supply of unskilled labor is declining. And finally, institutional factors. I think that minimum wages. And in Latin America there is this, I mean the World Bank has produced interesting data showing that in Latin America this lighthouse effect works. So if I increase minimum wages in the former sector, this spreads also to some degree to the former sector. So this is the thing. Now every model as any one model, this supply demand for skilled labor and skilled labor shows to obtain in the end of the skilled premium. It's very, very difficult to do because you don't really know what is the demand. You know what is the equilibrium point. You know the employment, but not the demand. So demand can be proxied in some ways, but it's not simply meant. You don't have intentions of employers because and so you have to use some proxies. Now the interesting finding is that this seems to be going back to the, Rolf, going back to the 70s and 80s, that also the urban wage gap is an important factor in reducing totally inequality. And that only in countries where, I mean there is still a sizable supply where the agriculture is still important. Then we have the transfers. Now why should the transfer reduce income inequality? Well, because they are well targeted, but also because they've been financed with increased taxation. And I think that what we'll see later is that tax-to-GDP ratio have gone up quite a bit and because they are better targeted. Now the other point is that remittances. You know remittances in the literature normally are treated as unequalizing. So here we have the case of El Salvador, which is a country with a huge migration, out migration, one third of the Salvadorian are in Los Angeles. And you see that the red line is the trend in the genetic coefficient before migration, excluding remittances. And the other one including remittances. And you see that over time the distance between the two arises from three points to six points. So which means that if the narrows go to Los Angeles to sell coca, probably this is equalizing in the country of origin. Now, how about capital income? And now we know very little about that. And this one of the weakest point. And then, and I think that the debate could be richer dyes in this. I mean, we know for instance that the Brazilians holding assets abroad are, I mean, they own something like about 200 billions of dollars. And this is money which has been earned in good part in Brazil, belongs to Brazilian, but they don't pay taxes or very little taxes in offshore places. And so if you include the income generated by that, we are not entirely sure what would come up. And even the data, even the approach in which you correct the distribution of income by means of the tax returns, they don't declare it in Brazil. So one part of the capital income may be escaping to offshore centers. And so if you calculate the Gini distribution by nationality, that may be different from the one which we found. So that is still to be clarified. Now, so this is more or less what we find out with the decomposition. And this is it, but why wage income goes up, why this goes down. And so we found statistical data which are coming closer to our story. But then we don't know what are the political economy factors. So we say luck. So these are the five possible explanations. We'll go straight to them. Now, is luck equalizing? Well, if you look at the term of trade laws, remittances laws, financial bonanza improved. Now, what are the distributive effects if you have an increase in the copper prices in Chile or Peru? Well, copper is produced in the sector with an extremely high capital concentration. So if you take the distribution, if you take the production function, it's very unequal because to dig mineral into the ground basically you need a lot of machines and a little bit of labor. And the labor which is employed in the mining sector is normally highly qualified labor. So ex ante is unequalizing. Now, how about finance? Was this financial bonanza improving access to finance by the small guys? No, because there was no changes in the institutional approaches in the financial sector. So basically, and then for the remittances perhaps because we saw it, remittances in the six countries, the remittances in three of them are unequalizing and three of them are equalizing. So it depends where you are. So the direct effects tends to be unequalizing. But then you have an indirect effect because if you have terms of trade effect, I mean you have an income effect in the national economy. The balance of payment improves. You can import all the technology you want. You may create more jobs, you distribute more wages and therefore there is a growth effect which you will see a little later. But overall ex ante we discussed, we basically argue, I argue that the financial bonanza if anything should be unequalizing because of the distribution of assets in the sectors which are produced in the Scots basically. Now, people they argue, yeah, but then if you get a high bonanza, governments receive more revenue. And this is true. So we plotted for all countries for 2003 and 2007, the relation between terms of trade on the horizontal axis and the tax revenue to GDP on the vertical axis. And you see there is a very, very little relation. So the correlation coefficient is 0.18. So this is for 1990, 2007. So if you take the period 2003, 2007, it's very small, 0.18. So no correlation. So if you restrict the sample to the eight main commodity exporters and only to the years 2003 and 2007, yes there is so the correlation coefficient rises to 63 which if I remember correctly my statistics basically is not an extremely high, but it's not tremendously high. So even the fact that the Latin American countries have earned more revenues which is important for many respects is only in part and only in a few of the countries can be attributed to commodity prices. Now, is it growth? So we say, well, international conditions, not too convincing that like Ricardo Hausmann argues that the source of decline of inequality comes from a good global environment. Is it the domestic growth which they had? Now, here basically we look at growth and we see bilaterally like in the chart that Isabel showed that we see that the correlation are low, they ask where there is a totally, practically all the variance is unexplained. But exactly does growth lead to inequality? The clients look at China. China's had until a year ago. Basically a very fast rate of growth and a very fast rate of income inequality. So it's not the rate of growth which matters, but it's the pattern of growth which is what we were saying when Rolf and I we were wearing shorts, you know, here in the ILO. So those who argue that growth is this miracle that solves all the solutions, I mean all the inequality problems basically falls. It depends on the pattern of growth. And so growth looks also an unlikely thing. Now, the liberal policy changes so the third solution. And I think that we should look at the political economy. So what we call the politics of policy changes. If you had a sort of a very conservative regime would they have introduced policies which made more sense from the point of view of the poor? Unlikely, not because they are bad people but because they reflect the interest of different social classes and perhaps they are closer ties to the international institution, international financial sector. So what do we have in Latin America in the 80s? I mean I lived shortly in Chile in 1976 with General Pinochet. And most of Latin America, Argentina, even the highly civilized Uruguay they were military dictatorship, you know, Brazil and so on and so forth. So in the 80s and 90s there is a gradual return to democracy. Now one rule which is a very strange rule when you come out of an authoritarian regime be it the sort of a Russian communism or military dictatorship, the political, the democratic political regimes which imagine are never center-left, never. I mean, so there's Tandik and Kandawira, our friend from unrest. Basically, so we have identified the sort of a transition period which may last six, seven, eight, 10 years. I eat the period in which you introduce neoliberal policies, then you get fed up with them. You find out the distribution worsen. You find out that only a few people benefit. You find out that employment does not rise very much. And then if you have democracy then you vote against these people. And this is what has happened. Now in Latin America there is a survey, an ongoing survey which is called the Latino Barometer which is a private institution based in Chile which collects information for a significant sample of the population of the region. And basically it is very clear that people, they were against the privatization of telephones and water that they thought all the economic policies were benefiting only a part of the country. All this shift to the left. It's not that everybody has become Marxist or whatever. But basically, so there is no ideological realignment but there is the fact that people saw their own pockets, I mean their own wallet, shrink during the 80s and during the 90s. Now if you have a regime which is more favorable ideologically and also because the social classes which express them basically they want this type of things then you will introduce policies which will be imperfect, they will not complete but they will go in this direction. And if you look at the history of political regime in Latin America this is the one. You see that until, I mean the turning point is late 90s and in this case is 1998. The red ones are the center left regime. The blue one are the center right regimes. And you see there is a cross and this is quite striking. And the center countries have remained more or less the same. So if you go back, we see that in the end the inequality fell less but also in a conservative regime. So there is a sort of a spill over effect also in this type of countries. Now what is the story, what do they do? Well they basically adopted a new micro-economic regime. So they abandoned fixed specs which is the first thing. Second they had them, they were in a way like the Washington Consensus that the very prudent budgetary policies, they even had surplus, not primary surplus, it's a surplus for the region as a whole. They had the prudent monetary policy and so they say but how do they finance this increase in expenditure? Well, because they tax more. So basically the tax GDP ratio rose by at least three, four points in the region and in Brazil nine points. So Brazil now taxes more than in the UK, more than in Spain, more than. And so basically the government has been collecting money and I think this has been a very important feature of their own new experiment. Now if you raise more money through the budget basically it can increase public expenditure in a non-inflationary way. And then there was a massive increase in social spending particularly education and housing I believe. Then the policies, I mean if you abandon, I mean if you look at the standard macro policy in the neoliberal regime, this is very prosyclical. So when things go bad you cut, which is completely crazy including in Europe now. So we have to follow austerity in a period in which the economies are depressed. You should follow austerity in periods when the economy expands. But in periods in which the economy is depressed, I mean if you take basic ISLM or ASAD basically then you have to use an expansionary policy. Now I mentioned exchange rate which has been very much better. Then the other point which in a way is tries to bring some smile on the face of my pessimistic English friend here at the head table. I mean basically there has been a lot of work being done in better provincial regulation of domestic banks. And now why? The question is, well we don't really know but there is a sort of a committee of central bankers which reviews and so many, many, many, many better regulation of the domestic bank in financial sector. Why? Well my guess is that they suffer from so many, so many crisis, so many financial crisis I make. That basically, I mean it made sense, I mean they paid the very high prices. I mean there are some guys in, I'm sure in the ILO but also in the World Bank who started the cost of fixing the financial crisis of Ecuador, Argentina, some sort of, I mean it was immense and spread over several years. So perhaps with the change in political regime they started trying to fix the problem. Now there has been also a change in one of the standard dependencies of the region which is the dependence on foreign capital. Well I mean basically if you take the financial assets I mean the deposit in central banks they've gone up and foreign debt owed by the country as a client. So the country finally they came to have positive net financial assets towards the rest of the world. Now what happens to the labor market? Well I mean I think here there are other things which are less easy to document but there is a very interesting chapter by Kaifman and Maurizio. Well first of all I mean if you look at the 80s and 90s there has been a massive informalization on labor which is what happens here. So the fear of the permanent secular stagnation I mean I think it is an underconsumption I mean this has clearly happened there but there has been an effort to correct that. So the share of labor which was not covered by contract not covered by social security basically started has been deliberately being attacked in the region. So it's still very high extremely high and Latin American formalization is also into the formal sector of the economy. So you have a large corporation who basically are labor without giving them a contract. Now again rising the number of work is covered by collective contracts. In Argentina they were telling me that during the Menem government there were four people which were in charge of inspection against informal employment, four. And so how could they cover a country so big? So now there are 300s perhaps there's still too few. Then the other thing which tends to be equalizing is the recent realization of wage bargaining. Then there's been a massive increase in minimum wages which I'll show you in a moment. And an increase in social pensions in many parts of the region. So some of the efforts that Isabel was mentioning in terms of social protection and the most obvious case are perhaps Argentina but Bolivia, Bolivia basically no social protection. And they provided the $20 universal pension to everyone over 70 years of age. And all the evaluation they show that these $20 they represent something like 50 or 60% of their own cash income. So it's very relevant intervention. Now how about minimum wages that I think I die low this I'm sure I'll die low data. Basically if you look at the one in yellow you see that in Brazil the real, so this is inflation corrected, minimum wage rises from 114 to 182. So by 60, 70% in Argentina it quadruples. In Uruguay is more than doubles and so on and so forth. Now Mexico, Mexico no lo hace bien. Basically Mexico basically has a decline in the real minimum wages. Mexico in fact has a decline in income inequality but not as pronounced as let's say in the southern corner. Now what happens to tax policy? This is what happens. So this is a tax GDP ratio in the three periods. So you see that the new liberal tax reform that Isabel was mentioning before basically you see that tax GDP falls if you are in a period of crisis tax GDP falls because there is tax buoyancy negative tax buoyancy it's normal. And then you see that when the economy recovers basically you go back to the long-term rate which is about 15, 16% but then during the 2008 this goes up by three points which is quite a bit. And is it equitable this increase in taxation? Well it is and there is one of these leaflets here is the abstract of one of the chapter of this book which I contributed to prepare. And so if you calculate the Reynolds-Molensky index which is the difference in genie distribution before and after. So now when is negative this means is basically is regressive taxation taxes more the poor than the rich. And so we did that on very detailed studies for the 1990s and then for the 2000. And then you see that for instance in the case of Argentina taxation was regressive in the 1990s and becomes progressive in the 2000. And the change is huge it's almost four genie points which means the tax system tends to become more progressive. And if you see that in the 2000 there is only Honduras and Salvador which they still have a regressive taxation because they rely mainly on consumption taxes value added tax and next sizes. Now one of my PhD who is the co-author of this chapter Bruno Martirano did these thesis on the new tax reform in Uruguay where they reintroduced income tax. Uruguay which normally people they think is the Switzerland of Latin America or whatever else basically they completely abolished income tax during the 1990s. And basically it shows that basically there is no efficiency effect I mean because the idea is okay introduce taxes and people will no longer go to work the firms will no longer go to invest it's not true basically. So there has been a deliberate effort raising taxes and I think that the low taxation Latin America was one of the main factors of macroeconomic instability in the past. Now then what you do with this money? Well you increase public spending. You increase public spending on what? Well mainly in the social sector and you know there are people who says well we're born too far perhaps you should invest a little more in infrastructure. And I think that if you look at that an interesting part of this an important part of this expenditure went into expanding cyclotron education. This is enrollment in cyclotron rates so the blue bar they give you the increase in enrollment rates. So the first one Uruguay 1992 2005. Now Uruguay had already a very high enrollment rate in secondary so the increase is only five points but if you look at the last one which is Brazil the enrollment rate increase in cyclotron education is more than 30 points. Now what are the red bars? The red bar are the Q5 Q1 ratio so is the ratio of the enrollment rate of the children belonging to the top 20% of the distribution to the children belonging to the top to the bottom 20% of the distribution. So what do you see? Well you see that in most cases Q5 Q1 goes down. Though there still are a few three seven but in 11 it goes down. So which means that the increase in enrollment is basically taking place among the children of the poor. So there has been a large increase in educational enrollment among the in secondary educational enrollment among the children of the poor since the 1990. And if you take the data in the bottom part of the table you see that average spending on education for child 014 including kindergarten in real terms. PPP so basically it gets multiplied by five. And if you take the share of GDP which tells you a little even more it goes up 2.8 to 4.4. Now it's not so simple to attribute that to policies so Gasparini and Cruces in this volume basically they've developed an algorithm and they basically conclude that 50% is due to growth. So this increase in public expenditure because the government spends 2% of GDP but GDP keeps growing so you spend more money. Then second is that there is a court effect. So the number of children entering secondary education is basically slowing down while well because of birth rate they've fallen 15 years before but at least 33% is due to policy. So which means that at least one part of this increase is driven by the desire to put the children of the poor into school. Now if I increase public spending on education particularly among the poor basically I can compute on using Barron Lee data on the years of education of the labor force and other source I can compute the gene of education. So how nicely or how poorly distributed is the human capital stock among people? Among workers in particular. So you see that basically everywhere that there is a massive decline in the gene inequality which means there is a massive decline in the distribution of years of education of the labor force. And so that varies from of course you see that the biggest decline are in countries where the enrollment rates were very low so like El Salvador which is the sharpest decline. But also Brazil. So in Brazil was a very segregated type of education not many people in Mexico. So basically the very large increase in education second education particularly improved the distribution of human capital which is a reproducible capital which unlike the land which is not reproducible. Now if you look at social assistance that the story is more or less well known there has been an extension in social insurance there has been an increase in conditional cash transfers. Now there's a big debate. Iowa has been very unconvinced by the Brazilians by particularly by the, what is the name? Now his name escapes me. But the argument that with 0.5 pais de barro by pais de barros who is now is the minister of minister of deputy minister for social affairs. So his own views and his PhD is top economists. So the question is that if I take a distribution of income I take the Lawrence curve and then I give a tax and I give 0.5% to the bottom part. Well the distribution will go out a little will go in a little bit. But not much. So how could it be that with 1% of GDP I save the reduce so massively inequality? Well this is what Marcelo Neri says. Now much of this money is given to people which are living in outlying areas, rural areas, small cities where the economy is very little monetized. So the multiplier effect. So if I give one Rajesh to a poor family in a small village basically the local GDP will increase by 1.78 points. And then they calculated that the multiplier effect for all the other types of social transfer and they see that for instance if you give it to the public servants the effect is very small or smaller than one. So now finally I think that we should say what did the new policy model in Latin America did not do? Well first of all there has been no major distribution of assets when I was spending time here with my cop against Rolf Wouter and so on and so forth. I mean there was also Keith Griffin here and I was very much influenced by his thinking, by a book which I now have given to my PhD students called Poverty in Latin American and in Southeast Asia. And so Latin America still has the latifundia so it's still very high land concentration and when the leaders came to power Lula in Brazil in Paraguay, Lugo in Bolivia, Evo in Guatemala I forgot which one, they all basically promised to redistribute the land and I've been told this country to look into that in particular. None of them delivered, none of them has delivered. And there are many arguments which have been given where no other population is urban and then basically it's politically very difficult to do. And now mines, gas, oil, and our face Bolivia is an exception but also in that case very little nationalization. There is a nationalization of YFP in Argentina but actually with compensation. So basically you get the mine, you get the oil but then you give them five billions. Now access to credit and finance and for small orders. And then university education. University education remains still skewed and we've seen the recommendation. Now industrial policy. I think that the country has been abandoning with an open trade regime. I mean I'm more assertive than Richard. I think that all the data I've seen they tend to show that by large trade liberalization in the periods in which trade liberalization is unequalizing. And there is a paper by somebody in American academia I mean very, very clear on that. So if you do that basically you're doing the deindustrialize. And then I think that one of the problem is that rather than progressing with their own industrial revolution the region, I mean Argentina is a highly developed country. Brazil the same, Chile the same, and so on and so forth. So rather than witnessing a sort of further increase in the industrial base then you see that there's been a move towards financial services, return to agriculture and then reduce dependence on foreign finance. Now I think that Stephanie Griffith Jones taught us that countries with basically, which basically finance a larger proportion of their own investments with domestic resources are more stable. That is quite obvious. So there has been a lot of changes in the banking sector, good changes in the banking sector. In Brazil there has been an increased role of the Banco Nacional de Desembolvimento while all the private sector was not lending. But again, we are far from becoming able to finance our investments entirely with our domestic savings or at least a large proportion. So altogether this has not been a radical model. This has been as it is defined in the literature the social democratization of the Latin America. Now with under aggression analysis I will save you that but basically confirms what I told you. Now let's look at the problems. This is what still needs to be done. Then now tertiary education, university education. Now changes in tertiary enrollment rates you see that they have increased. But you also see that the Q5, Q1 ratios increase everywhere. And if you look at the political elections in Chile, the progressive regime, I mean there was this student, very young students, like 23, leading the campaign of the opposition, basically saying to the center left candidate and certainly to the center right candidate that basically you have to expand access to tertiary education. Because if we do have further technological shocks, basically we will go back as before. So the wages of the skilled laborers will go up, the wages of the skilled laborers will remain where they are. And when I was in Chile in January, and basically I'm told that the government has targeted the 3% increase in taxation, basically on increasing corporate income tax, mainly to finance broader access to education. In addition to that you have to finance the strengthening of public secondary school, which are of lower quality. In Italy it's the opposite. In Italy, if you go to a private school, it means you don't want to study. But in Chile, basically in the last part of Latin America, the private schools, they prepare better the students. And then entering the university is a bi-competition. I mean, I don't know whether it's true everywhere, but in the last part. The second problem is that if you look at that, if you draw a sort of a, if you calculate what should be the taxation norm in relation to GDP per capita, and this is done on all countries, not only the Latinos. You see that with the exception of Argentina, Brazil and Nicaragua, all the others are below the international norm. So we still have a need to increase taxation, and we have done a little bit of regression analysis, basically showing that if you increase direct taxation on GDP by one point, Gini falls by 0.9 to 1.2 points. And if you reduce excises by one point, this Gini will improve by, because they have a very bad incidence by 0.9 points. That's it. So lots of still needs to be done, but I think that we have a sort of an anti-picketing antidote. Thank you. Okay, thank you very much. Sure.