 Thank you, Rachel. Hello, my name is Elise Crozer. I'm a PhD candidate as was just said at the University of Cambridge and The presentation has a rather dramatic Title and picture. Let me just tell you that this is a photograph from my living room in Mexico City and Mexico is a very interesting case for Inequality study especially Mexico City because it has the richest man in the world living there. So it has Arguably one of the highest levels of inequality in the world That's where I live when I'm not in Cambridge. So forgive me if my examples are a bit biased sometimes towards the Mexican case Besides that my analysis is on world inequality, I'll tell you what I mean by that in a second The main point maybe that I'd like to make is that If we measure inequality but not Know where it actually lies within the income distribution. It's very very difficult to Improve its levels. So I think that we should or I argue for Having something like a poverty line for inequality as well an inequality line a threshold a cut-off number If we know below which or above which level inequality should be it's easier to target it and to address it I'll show you some data that I think supports this case Right I created some kind of a roadmap that we should follow That's closely related to the paper that that I handed in it's not chronologically the same but you'll Follow the content is very similar First I say that to address inequality effectively we need to know where to locate it and I'll explain what I mean by that Because inequality is mostly located on a world level in the extremes of the distribution that is in the top and the bottom parts and mostly In the top part. I'll show you the data for that The indicators that we use to measure inequality and that's mostly the genie today Are not able to detect these changes in the tails Sufficiently which means that we might underestimate inequality or depending on what kind of inequality want to measure We might not get it right and that is a problem in trying to ameliorate it so I think that if we make explicit the actual concentration at the very top and We offer for example threshold of maximum inequality like an inequality line That could help to curb inequality So this is what I want to do and the last part presenting an alternative measure It's called the palm of volume to I'll tell you why Just before I move into the actual paper some concepts I look at wealth income inequality and that's the shares that different parts of the population hold in Income so it's relative inequality in this paper I compare income inequality in Two different with two different data sets the first one is Milano rich's wealth income development with the benchmark 2008 sir And because it disaggregates data for the top 5% so the top vent all or for 5 percentage Groups throughout the population and I use that data set because it has 116 countries compared that's over 90% of population and GDP. So it roughly approximates wealth Dynamics and then I use a sub-sample of 41 countries from the list database that disaggregates data to the top 1% of the income population There are some mythological problems about that we can discuss about that later if you want But it's the largest sample that's available to disaggregate top 1% income earners I Also use some part of the paper to look at distribution over time I will not go into so much detail here because of time limitations, but we can discuss that later as well And that's for a sub-sample of 25 countries okay, so where is Inequality located in the actual distribution across countries turns out that inequality is mostly defined in the extremes of the distribution Because the population parts between the fifth the cell and the ninth the cell hold roughly 50% of Total income throughout the countries in the world and I'll show you a graph in a second Well, we can see that That's a quite curious Discovery that Gabriel Palma Cambridge professor might some years ago It's very stable over over time as well to some degree and especially across countries so The differences in the level secured by the top Ventile top 5% of the population range from almost 39% of total income in South Africa to 11.5% in Slovenia so differences are vast Slovenia Has the top 5% owning less than a third of what top 5% in South Africa owns and in the bottom 40% of the population is quite similar So in South Africa for example 5.5% of total income goes to the bottom 40% of the distribution While in Slovenia, which is the most equal country according to most measures It's 25.5% so it's five times that And this contrasts very much with the relative homogeneity in the upper middle groups Where the difference for the 55% of the middle? Are 20% of income difference between China which has the largest upper middle group And the Central African Republic which has the lowest Okay, so to understand this better I show you the graph The blue part is the share of the bottom 40% of the population and you have the 116 countries lined up so you can see that There is some difference in The blue part there's especially large difference in the upper the yellow and the green parts Which is the top 10% of the distribution of income earners The red part in the middle That's the middle of 50% it's really fairly stable across countries So that's the one very interesting aspect the second very interesting aspect is in the lower part You have the inversion of the of the distribution you see that while the yellow part is The 19th ventral so it's not the top 5% but the next 5% is Incredibly stable across countries is almost the same Where really the difference lies is in the green part, which is the top 5% of income earners So it's very very different from the more equal countries to the left towards the more unequal ones to the right A different way of looking at this more explicit is This graph where you have the 19th ventral which is Vental which is the yellow line is very very stable across countries and Blue and the green lines is the 10 top 10 and top 5% respectively Move along a very similar pattern and this indicates that most of the inequality within the top share is defined by the top 5% rather than the top 10% so Although the top 10% is Very unequal or or because it is very unequal We can see that even within this that's all there is large inequality and top part of the top That's all is much more unequal or is much richer proportionally than the rest Another measure, I don't have data or I don't have a graph for this now But the top the cell has the highest genie if you compare different decels Genies with each other So what happens if we look at smaller ranges top 1%? I'm not sure you can see it until the back, but the the dark line the blue line is the top 1% of income earners for a set of 41 countries and you can see that the range is from 3.1% again in Slovenia versus 13.5% of totally income in Colombia So you have the top five countries are Well except for South Africa, which is in the top always is Latin American countries and that's quite expected However, what is less expected is that if you look at the middle Here you have This one that's Belgium actually so according to this Ranking some countries that we traditionally expect that are much more equal All of a sudden are fairly high up in the ranking you have in the middle here. You have Denmark You have Finland this one is Finland You have others kind in even countries around here So if we rank countries according to the top 1% of the of the shared their population holds the Dynamics or the ranking changes Compared to the to the genie coefficient I'll tell you in a second why this is very interesting and problematic This is as I said the development of Inequality over time. I know it's very small It's not so essential for you to see all the details It's just to show you that there are some divergent Experiences between countries. So it's not as clear is not that in all countries. It's falling or rising. However in most countries The top share is rising As we can see in this graph The blue line is the top 1% around 1990 not exactly the same year but roughly 15 years of difference minimum is between the two measurements Data is very limited for this kind of study. So These are the only countries that have comparable data over time from the list data set And as I said the top 1% in blue is around 1990 and the top 1% share in red is Around 2010 so we can see that in most countries the top 1% share is actually Increasing and I know Minister Neri gave a very optimistic presentation in the morning about Brazil and I'm very happy about the Brazilian case. Unfortunately, it's not the same in all countries and maybe I'm a bit more skeptical But from this day, we can see that even though there is important exceptions like the Mexican case the lowest Which saw a large decrease in the share of the top 1% also. It was very very high. So There is some some reason for optimism, but in most countries the top share is increasing Okay, so why why is this a problem and why can we not measure this properly? If we look at the development of The Gini coefficient over time in the Mexican case because that's the one I know best we can see that from around the 60s to 2012 latest year for which this information the line is almost stable So that would mean that there was very little development and inequality If we look at different indicators and here I use the Palma We can see that there is very large movements. It's a red line So there has been a lot of things going on in the income distribution the palm of what it measures I'm not sure whether you're familiar with it is the Ratio of the top 10% over the bottom 40% and there is good reasons to use that as an indicator that are explained by Professor Palma or by Co-op and Sumner in their papers last year you can look at them You can ask me about that later, but they're not going to details about why this is a preferable measure Besides the fact that it does capture changes in the extremes of the distribution, which is why it doesn't look as flat as the Genie curve here Since I From the data that I just showed you there is a large differences in the extremes If we want to measure inequality properly, we should look at the extremes and we should have an indicator that can Display changes in the extreme. So the palm is a and an interesting proposal compared to the genie if we're interested in measures at the top Most people are in fact interested in what happens in concentration at the top and bottom rather than smaller changes in the middle for which the genie is More adequate. So an example besides the Mexican case to not make it too biased Of the problems of using the genie to measure concentration is that for instance, Portugal and Sierra Leone both have a genie of 34.4 in 2010 if we use the Palma measure for example to capture inequality it Becomes 1.38 for Portugal and 1.73 for Sierra Leone Which means that there's 20 position ranks between the two countries. So it becomes a very very different picture To understand what it means that the Palma becomes 138 or 173 I'll tell you what it actually measures it measures the share of income That the top 10% holds compared to the bottom percent So if it's over one it means that the top 10% own more in total income than the bottom 40% of the population In most countries actually they do so To come back to this graph because I only told you half of it in the last slide We can see that it's counterintuitive that Belgium or Denmark or other Scandinavian countries are in the middle of the of the income ranking rather than at the bottom where we usually locate them with the the genie coefficient and this is explained by the difference in the shares of Top 5% income, which is the red one or Top 10% of income earners with the share they hold If we now measure inequality according to the share of the top 10% we will get a different ranking Then if we measure it according to the top 1% so yes, the Palma is an improvement over the genie But it's probably not enough if we want to look at the very high concentration at the very top Especially for countries with high inequality like the Latin American ones which Expectedly are in the top, but also if we care about concentration in high income Low inequality countries according to the genie So an alternative that I propose in the paper is Palma volume 2 Or volume 3 which is the share of or the the ratio of the top 5% share to the bottom 40 or the top 1% share Of the over the bottom 40 you can see that they behave somewhat differently. This is the volume 3 The red one is the volume 2 and the blue one is the conventional Palma You can see that South Africa is a bit off the chart because it's inequalities too high But especially in the middle range the top 1% measure gives very different results than do the Traditional indicates if we had the genie here it would have a completely different measurement as well so If we care about the concentration at the very top There is a case for using Ratios that can display and can capture inequalities at the very top better They're more explicit about that and they're more intuitive in the calculation than the genie Maybe you were wondering how much is too much since that was the title of the paper of course we're not going to an ethical discussion which is very related to cultural and country preferences and Nates to be on India's idiosyncratic discussion in each country, but the virtue of the Palma volume 2 that I calculated for the global for not for the global distribution for the international distribution so Across countries is that on a global level. It's One the measure that means that top 5% own as much as the bottom 40% in share And that is a very easy threshold for technical reasons if we want not for ethical reasons, but we can probably argue that more than If the top 5% own more share than the bottom 40% that could really be too much so if we look at the actual countries and the differences between their incomes it's The the Palma volume 2 is point four or five times of the income of the Lower parts in Slovenia in South Africa. It's just over five over seven Sorry, so that means that the top 5% in South Africa own 56 times the share of total income that a person in the bottom 40% Can hope to have so that's really absurd difference and absurd inequality So might be too much I'm not going to details with policymaking as a result to change inequality, but this is kind of the plus part But of course is not enough to just fix an indicator and a threshold. We need to go into policy I'll just quickly point your attention to the fact that inequality market-based in Mexico and in Denmark for example, which is Much more equal country according to the genie according to the 1% depends is fairly fairly similar into those in the 10 However, the market income, sorry the disposable income is very different. So obviously the importance Lies in policymaking. This is a very good news Because that means that actually something can be done about inequality, but to do that I argue that it is important to have a threshold to know What we have to actually target so the conclusions that I derived from the Data that I found is that The income concentration at the top is not only higher than expected, but it also depends on the indicators that use Inequality becomes much higher than expected in different countries These levels are unlikely to be in the interest of the majority of people so it's definitely too much We need an indicator that is explicit about This too much so that explicitly states the concentration at the top And the palm is an alternative to do that is to be complemented with other measures But we definitely need to fix an objective to know what we actually want to approach much like the extreme poverty lines To have an extreme inequality line that we know we don't want to overstep This is not The only but definitely the first step to go towards an inequality that we'd prefer to have rather than the one we have Thank you