 It was my really pleasure to be invited to this conference. Like, thanks for inviting me. So I just gave some responses to the two presentations. I think that my comments just have some general... Issues also have some specific issues. The first is that I would like to say there's really two excellent presentations indeed. The presentations discuss the two very important issues in the world distribution studies indeed. The first is that the global, the world's distribution. So how to measure precisely the global, the world's inequality in the world and its dynamic change over time. That's so important indeed. So the second is that you see in inheritance and in inheritance taxes, whether they reduce or increase world's inequality in the countries. So because we have some belief in inheritance will increase the world's inequality, something like that. But Daniel's presentation gave us some new evidence that in some countries it may not be correct. So I think that two important issues is so important for studying the world's distribution. The third one is that some results from two presentations are so exciting. When I read that, I really learned a lot from that. It's also so significantly important indeed. From James' presentation, we know a lot of the results. So that is, you see, 50%, the bottom 50% of adults has a worse share, the less than 1%, less than 1%. So while top 5% have a share greater than 75%, top 1%, greater than 48%, the Gini coefficient in 2014 is so high. You see close to 0.911, so called incidence indeed. So also the James' presentation arrest another very important issue. So the pension, you see, should be encroached as a component of wealth, household wealth, you see. But the impact of the pension on the wealth inequality is a difference from one country to another. That will reduce wealth inequality. For example, in Australia, but in Greece, but in Canada, something like that. That's so different from one country to another. Also James' arrest the questions, also important issues that really we need some better data. For example, for the up-dead income distribution, also we need the better data for collecting the sampling and the measurement errors, something like that. There's a lot of work to do to get precise the figures for the wealth distribution, either in country or in the world. So from the Daniel, the presentations, I will have some impressions that the data are so unique and so valuable indeed. There's a so large sample of data, you see. Also he mentioned two very, very important, very, very interesting results. I just mentioned that. So I based on the data from Sweden, you see. Inheritance unequally distributed, that is expected. But inheritance reduce the wealth inequality. That is, I did not impact it. Also inheritance taxes play the minus role, but increase the wealth inequality. Increase the wealth inequality. So I think that result even are consistent with some studies in other countries, but I don't know that result can be found some evidence in developing countries. I was thinking about that. Whether if we have the data from China, whether we can get some result, I'm not sure perhaps the result are opposite indeed. The rich people will have more inheritance and pass down to their children. So based on this result, so I have some question that is, even Daniel mentioned that the data include both year errors and non-errors. Inherited and not yet inherited the people. But this impact are only among errors, something like that. There seems no very, very strong evidence show whether you see inheritance will reduce wealth inequality. Between inherited and non-inherited. Second, the question is that, that you do not really understand the why in the Swedish societies. The less wealthy people just receive more inheritance than the wealthy people. There are some mechanisms, maybe some policy. So that is my general comments on that. Just come to some specific issues, mainly related to the GIMS presentations. The first is that how to get the wealth data for top 10 indeed. The GIMS mentioned several channels to do that. We agree with that. So when the method is left, just use the information from Forbes and the other rich list, something like that. But I don't know if the data from Forbes are reliable or not in the other countries. But as I know that the information is not so reliable for China. When the Forbes published some number of billionaires in China, the Chinese people don't believe that. Even some rich people don't believe that. They say, I'm not in the Forbes, I think my assets are much more than that. So we should think about the information from Forbes can be used for the different countries. My second concern is about that. I think the GIMS did a very good job to re-estimate the share of top 10 or top 20% of rich people by correcting the figure from the household surveys that is. For example, for the data in 2002 in China, the share of the top 20 or top 10 is underestimated. But how to correct it? So from the report, the share is much higher than that from the household surveys. That means that big difference is based on some estimations. Because I do not know what is the method to use that. According to my understanding or our data, I think perhaps using the estimate, the share of the top 10, top 20, the rich people according to some method may have some overall estimating problem. So for example in 2002, just because at that time China just started the housing reforms. That means the housing assets were limited. Also you see the number of billionaires was quite small. As you see GIMS presentation in 2001, only one people is on the list of folks. So also the size of private firm was small that. So that means there are not so large number of rich people at that time. So that we should think about how can you see estimate the share of the top 10 or top 20 people. So also GIMS mentioned our new data set. That is probably I just talked a little bit more about the new data set. So we collect the new data set with a lot of information on the household wealth. Like housing assets and financial assets, production assets such as. So we did some rough estimations about that. So GIMS mentioned the figure I told him is about 0.65. This figure may be underestimated. We will do re-estimations, try to get more precise figures. So the figure is between 0.65 to 0.7, something like that. That depends on whether we should consider. You see, include some top 10 information, consider do some correction for some provides, something like that. So yeah, I will tell the GIMS and also audience when we get some new data and kind of find some figure at least. Okay, thank you. Our last speaker this morning is Subbu, Subramanian from Chennai and again being very much involved in our work on wealth over the last 10 years. Thank you very much. I will try and keep this brief. I would like to concentrate on just three issues and I will deal with each of these as briefly as I possibly can because I can see that we are getting on close to lunchtime. The first issue will have to do with some general remarks on the subject of enquiry. The second will have to do with a few quick headline points about wealth inequality in India. And the third will be on a specific issue relating to the measurement of inequality. In regard to some general comments on the subject of enquiry, it is very welcome to notice that the subject of the distribution of wealth as opposed to the distribution of income or of consumption expenditure is beginning to acquire a certain salience in the field and I think this has a great deal to do with the initiative which wider undertook in 2006 with the publication of the book Personal Assets from a Global Perspective. I mean the sort of attention which say the work of Milanovic justly has attracted on the global distribution of income is something which has not quite been paralleled by work on the distribution of wealth and not least because I think there has not been a great deal of applied or empirical work on the subject which is a little surprising considering the importance of endowments in the scheme of things in theoretical economics. After all the second of fundamental theorem of welfare economics has much to do with the role of wealth or of endowments in general in assessing the goodness of end states of affairs. In particular if you recall the second theorem says that any desired allocation can be sustained as a Pareto optimum for an appropriate redistribution of initial endowments and then there is a work of people like Amartya Sain who have pointed to the centrality of endowments in determining entitlements and indeed if you go back to the 1970s following on the work on distributional justice by Nozick there was a resurgence of interest again in the role of endowments in securing notions of both efficiency and fairness as reflected in the work of people like Hal Varian. Of course Amartya has a much longer history going back to the work of people like Thomas Payne in his rights of man when he talked about estate duty. It is very welcome to see this whole subject making a reappearance in its new empirical incarnation in terms of both general treatments of the distribution of wealth and in particular of the impact of inheritances on the distribution of wealth. Having said that I will move on very quickly to some headline points relating to the distribution of wealth in India. As Jim pointed out we have decennial estimates which are available through surveys on the distribution of household assets in India and we have data going back to the 1950s in fact but if we look at the last four decades in 81, 91, 2002, 03 and the latest set of data which is available for the year 2009, 11, sorry, 2012, 13 what we notice is the following. For capital levels of wealth have increased tremendously in India although India is still a wealth wise very poor country in the national scheme, in the international global scene. The well-adult wealth holding in India and I'm talking about assets not of net wealth, net worth is of the order of about 12,000 US dollars whereas for Sweden which is number one on the list is 237,000 dollars so there is a huge distinction between a relatively poor country like India and the better off countries. But even so if welfare is an increasing function of mean levels of wealth holding and a declining function of inequality then it's particularly unfortunate that in India the distribution of wealth has been increasing quite rapidly. But more on the average level of household assets between 1981 and 2011, 12 in rural India mean per capita household assets have increased by a factor of 4 and in urban India by a factor of 9. The gross rates in per capita income or rather per household wealth holdings have been of the order of 3.8% between 81 and 91 for rural India which went down to 2.4% between 91 and 2002 and has gone up considerably to 8.2% for the latest round of data between 2002 and 2012-13. In urban India these rates are 6%, 4% and a whopping 13%. So what we have is that India is a comparatively poor country that there's been a very sharp increase in wealth per household over the last decade and indeed conventional measures of inequality suggest that there has also been a huge increase in the distribution of wealth in terms of standard indices such as the Gini coefficient and there has also been a large increase in the intersectional distribution of wealth. The ratio of urban to rural per household wealth has increased from rough parity in 1981, 1.03 to 2.27 in the latest round of survey data that we have. Many of the predictable features of a decomposition of wealth by caste or by occupational status confirm our worst suspicions that while overall inequality is increased both within group and between group inequality levels increased with the first far outstripping the second in terms of both caste and occupation. In terms of caste, the forward castes have done clearly much, much better than the scheduled castes and tribes and in terms of occupation, the self-employed category has done vastly better than casual and organized labour. In terms of the composition of wealth in India it is still land and buildings which account for the lion's share of wealth as opposed to financial assets which have a fairly dominant role to play in more advanced economies. This could have something to do with underreporting of financial wealth but then so is underreporting especially in the value of wealth a feature of both land and buildings. So it's still real estate which accounts for physical real estate which accounts for a bulk of wealth in India and it's a question of time before the bubble burst and one wonders what the harvest will be when that actually happens. So that's briefly about wealth inequality in India. Let me just wind up with one point on measurement. I recall in one of his papers Tony Shorok had mentioned that we have a tendency to stick with certain conventional ways of looking at measurement for reasons of what he called inertia and network effects and this is no more in evidence than in the measurement of inequality in general. We tend to stick very often with relative measures of inequality so we have the old work cause the Gini coefficient or the coefficient of variation. Now why this is of some salience is something that I discovered by simply reinventing the wheel when I was looking at the inclusiveness of the growth process in India and I found that one way in which one could approach the problem was to think of the distribution of the product of growth in an equitable manner analogously to problems in the optimal allocation of an anti-poverty fund or in analogy with the Talmudic bankruptcy problem where you have different patterns of egalitarian distribution of a given stock among competing contenders for them and I found that if I resorted to the least egalitarian of possible egalitarian divisions of the product of growth namely an equal distribution across quantiles then the results on inclusiveness of growth which I got were very different from the standard results which were to be effected in India inequality in the distribution of consumption is not a major problem because most relative indices of inequality indicate a rough stationarity in the level of inequality and this is also very much the case with wealth where in fact there was actually a decline in the genie coefficient of inequality and so far as urban India was concerned. So this takes us back to a very old problem. I mean relative indices of inequality invoke the scale invariance action namely the notion that inequality is invariant with respect to equal proportionate increases of income or of wealth whereas in a problem which has been known from at least the 1970s largely due to the efforts of Kristof Kom there is a rival to the scale invariance property which is the translation invariance property which says that inequality invariance is preserved not by equal proportionate increases in wealth or consumption or income but by equal absolute increases in wealth or income. Now it is arguable as indeed Kom has argued that in the presence of wealth and the presence of growth relative measures tend to be rightest in their ideological orientation and translation invariant indices, absolute indices tend to be leftest and there is a case for intermediate or central measures. Now as it happens these measures do exist but much of it has been localized to a somewhat arcane discussion of these issues in the theoretical literature on inequality measurement. But if we regard the composability of an inequality index as an important property and also the property of unit consistency namely the requirement that the inequality index must rank distributions in the same way irrespective of the units in which income is measured or wealth is measured. So if these properties are important then virtually the only index which is available intermediate or centrist index which is available for performing this job is an index due to a person called Kritsha. The Kritsha index as it happens has a very nice interpretation it is simply the product of the coefficient of variation and the standard deviation. So it's a product of a well known relative index and of an absolute index. And why this is of some importance apart from the fact that I mean at a theoretical level this might engage the attention of those who have a fancy for this sort of thing is that on the ground it makes a huge difference. Now if we're going to be whether or not inequality is good is a normative issue and there are those who hold that it has very deleterious consequences for the efficiency, for the prospect of conflict for any intrinsic notion of fairness that you might have and indeed also instrumentally on health outcomes. Now these are surely matters that can and should be debated but very often these issues are thrown out of the ambit of inquiry or debate by virtue of the fact that the diagnosis is to the effect that there has been little or no change in inequality especially in a country like India. And this is because of this exclusive reliance on relative measures of inequality. And even if you don't go in for an absolute measure but if you were to confine yourself to centrist measures such as an intermediate genie coefficient or an intermediate crickshaw measure then we find that the increases in wealth inequality in India of an order which my colleague Jayaraj has been working on this costs the graph to actually skip the page. So there have been huge increases in inequality once we take account of absolute differences and I suspect that this is also going to have implications for your own findings on the impact of both taxation and of inheritance on wealth inequality. If you were to use not a relative measure but a centrist measure then I think that would be more in conformity with ordinary intuition in these matters. Namely that wealth inequalities would have been seen to have increased with inheritances and with a progressive and complexation. So that is all I have to say. Thank you very much indeed. Okay, well thank you all the speakers for some very interesting comments. I have to say I think this is very much a topic for the future. I mean wealth has not been very prominent in the past but I can be reasonably sure that it will become much more important in the future and for some of the reasons which the speakers have talked about I just want to mention one thing about which Danu didn't bring out but I think is important in inheritance. It's not so much just what the impact is on wealth inequality but it's also on the intergenerational aspects. I think that's what's the main one of the driving reasons why we're interested and we think inheritance is so important and this is tied up with the fact that the wealth to income ratios have been rising throughout the world and particularly quite strongly in developing countries. That means that wealth relative to income is becoming more important and the fact that it's easier to pass on wealth assets are easier to pass on than income is that it raises these concerns not just about the fact that inequality is going up but it is going to become more persistent through time and that the old-fashioned ideas about that people can simply do wealth themselves and this question about whether or not children will be able to find their way in life and do well irrespective of their family's background so I think is the issue about whether that's going to be challenged in the future and I think that's one of the driving concerns and it's not easy perhaps to articulate that or even to predict what's going to happen and do analysis on that but I think that is what is very much behind some of the concerns about growth of wealth and wealth inequality. We are a little bit late. I know we started late and I think we have a little time for some questions so if there are people that would like to raise questions I think I'd rather do that rather than have the speakers respond immediately. Do I have some takers? Yes, please. Excellent. Is there anyone else that wants to ask questions that we can... Yeah, I'm Julian, Chinese Academy of Social Sciences. I learned a lot from the presentation and responses comments. It's very good. Just one question to James Davis that because I didn't read your paper I don't know how do you deal with the land for example in China although the agricultural land does not belong to the household but household in fact have the land if it rent or leased to others those households could get continuously their rents so I don't know how did you deal with the land your calculation perhaps in 2002 data the distribution of land is relatively even but now the land distribution is very uneven I don't know how do you deal with. Thank you. Okay, thank you. Is there anyone else that wants to... I see we're losing people and I think lunch is beckoning. Let's have one more comment question and then perhaps let the speakers have a last few comments and then we'll have to just deal with questions outside the hall. Thank you for having the opportunity to ask a question and thank you for the excellent presentations. First place my name is Adrian Perils from the Finnish Metallurgical Institute doing climate impact assessment there so my background is economics and not metallurgy. I have a question for the... related to the presentation of Mr Wildenström I was wondering about sort of the portability of his results and also in fact the durability of his results in that sense how important are for example changes in household formation over time to what is happening with these heritage transfers so if households start to have less or more children or if couples are getting older but for example the woman is getting much older than the man for example which has happened for some time how that kind of processes are influencing how this is unfolding. Thank you. Okay, well let me then... I think perhaps who should we start... then you'll start I think and then we'll reverse the... Jim then finish. So now I'd just like to thank the commenters for their comments and I think in the interest of time I think the last question from the floor on the role of family formation at least what we looked at was whether the rich had more children whether that could explain the patterns and in fact it did not in the Swedish case but of course over time demographics would matter a lot and we see an increasing degree of assortative mating in the Western world I think so that could mean that at the household level these inequalities may even become magnified when we take that perspective into account but this is a big discussion and I settle here. Yes, I just have one quick comment actually on Daniel's paper which I think is very interesting I think it would also if it is conceivably possible or could be done interesting to look at the impact of inheritance on the distribution of lifetime wealth in the sense of the present value of people's lifetime labor earnings plus their inheritance so if inheritance actually equalized the distribution of people's lifetime wealth that would be even more striking just in reply to the question that was asked about land just for clarification for people in general the urban land is not included in these estimates because it's not owned in any sense by the urban dwellers it belongs to the state agricultural land there is a kind of a well-developed tradition going back to the 1980s as I understand it in China of how to estimate the value of the agricultural land there's not a market for the land so you have to estimate the value and the imputed value is based on the present value of the net income that can be earned on that land but the question that was asked about the impact of the land being leased out I don't know the answer and actually I thought perhaps we could turn to Professor Li Shi to see if he has a comment on that Yeah, I think that is one of the good questions for China in rural areas the parents do not own the land but when we calculate the land values as a part of the wealth we just use the information of the agricultural value market value of agricultural products we suppose the land assets which will produce returns something like that then we use 5% you see that long-term interest rate that imputed the total value of the land something like that I think yes, Jim is correct Okay, well I think I'll bring this session to an end the one comment I make is really that 10 years ago if we had this discussion we would have been just scrabbling about to get any sort of coherent view about wealth holdings in the world and what was going on in terms of wealth and equality trends now as we sit down today I think we've learnt a huge amount in the last 10 years and the whole issue has been transformed clearly we do have problems just getting still getting the data into a form where it's up to the standard that we at least I wouldn't say expect but hope to get in other areas and I'm thinking if we had this session again in 10 years time I suspect we would again be quite transformed in terms of our understanding and our appreciation and some of the topics which we've been at least grappling with here will be a lot clearer so if you like watch this space because I think this is one of the topics which is really going to be given a lot of attention in the next 10 years and I think it is very important not just amongst economists but also amongst the wider public this is a real important issue for social policy around the world so I want to thank the speakers for coming along and talking today and if there's any issues which people want to raise do so and just try and grab us outside as well thank you very much