 Okay everybody, thank you for coming. I think anybody knows my work is probably suspecting I'm going to say that what's next is a poverty measurement but I'm not. We face very unequal challenges ahead in terms of particularly in research but also very much on policy and inequality, addressing inequality is in my view the greater challenge going forward. There are two main reasons. One is that we have this powerful instrument of economic growth for absolute poverty reduction. I emphasize absolute poverty reduction but inequality not so clear. In fact, roughly half the time in growing economies inequality increases after time it decreases. Much less powerful for inequality. There's also while there's broad agreement today that we want to eliminate poverty that's not the case for inequality. We don't have that kind of consensus. Anything like that kind of consensus. So my question today is should and can that change? This is all an extension of an op-ed I wrote in Le Monde last year and basically addressing that question. Okay, the motivational challenge. Why should we care about inequality? I'm going to review that and the policy challenge. What do we do about it? What are the constraints? What do we know already? What do we need to know going forward? Why do we care? First thing, let's get this cleared up. Zero income inequality is definitely not the goal. I don't think anybody could defend that very easily. We have all kinds of straw man arguments about class warfare, socialism and so on. I think we have to agree on a couple of things that heterogeneity and the ability of people to turn income into welfare is evident. That alone would mean even within a utilitarian framework that alone would mean that we don't want zero income inequality. We also recognize negative incentive effects of very low earnings inequality. They get exaggerated but they're there. The concern really is about high inequality, countries that have very high levels of inequality or are getting there. Ethical arguments, the ethical objections are pretty clear. Utilitarianism, as I mentioned already, a host of rights-based arguments. The fight between utilitarian rights-based arguments goes back 200 years, you probably know, and it's alive and well today. Issues of fairness in process, unequal opportunities, unequal outcomes. Unequal outcomes when they foster unequal opportunities in the future. Specific inequalities all carry weight. What's really new is our knowledge of and further work needed but our knowledge of the cost of inequality in specific dimensions. But we've learned about the costs associated with credit and labor market failure, land market failure too, particularly hurting poor and middle-class people. We've learned a lot about the political pediments to reform, economic reform and provision of public goods associated with certain forms of inequality and the emergence of both left-wing and right-wing popularism as part of that story. Social costs of inequality are getting more attention. Accumulated knowledge here I think pointing to a lot of concerns. It's not unambiguous. Some areas less clear than others. We also, the fundamental thing, I guess from the point of view of poverty and social welfare more broadly is how much harder it is to reduce poverty, promote social welfare in high inequality countries. Another point, although I said zero inequality is not the goal, certain kinds of inequality, zero is the goal. Certain between group inequalities we want to be zero. Certain within group inequalities we don't. And that's quite important to note. So the between group inequalities have a political salience and a motivator for action that within group inequalities are not. Ravi has pointed out a few times how standard inequality decompositions ignore this fact. Between group inequality can matter a lot more than we think in our standard economic measurements. Inequality of opportunity is another important area. We've seen arguments here going back to Aristotle. Aristotle's distributive justice, which is not the way we think of distributive justice today if you read Aristotle, but the idea of equality in the declaration of the rights of man and the citizen, which motto of France and the French Revolution. We also learned a lot about the lost output associated with inequality of opportunity. Counter-arguments galore and I'm just going to summarize a few of them and what I see is the case. Arthur Lewis. Development must be inegalitarian because it does not start at every part of the economy at the same time. Yes, it's so intuitive that statement and very clear, but there are some qualifications. We know that the distribution of the gains from growth depends heavily on the initial distribution of endowments. It depends on the pattern of growth. There's all kinds of clear qualifications there and important feedback effects, as I've mentioned. John Rawls. Rising inequality is acceptable as long as the poorest are making progress. I hear this so often that poverty trumps inequality, trumps the word. I've got to find another word. Anyway, really a dubious claim in my view, although I'm a great fan of Rawls. If there's any philosopher I turn to, it's Rawls. But a number of things, the dynamics of this, that high inequality can store future poverty deduction, even though things look fine now. Another issue we've learned something about and the measurement issues do a lot of measurement, but if we don't agree on who the poorest are, then this kind of perspective calls really for much broader distributional perspective, distributional assessment. Arthur Young. You may not have heard of Arthur Young, but he was the first agricultural economist writing in 1771. This epitomized captures the incentive argument beautifully. Everyone but an idiot knows that the lower classes must be kept poor. They will never be industrious. You've heard that today, haven't you? Some version of that. It's still around. It was a majority view probably 200 years ago. It's a minority view now. But the accumulated evidence Arthur Young, people like David Ricciardo were the greatest economist of his time and the greatest exaggerator of his time. They had no evidence. We have the evidence now. We've learned a lot about incentive effects on labor supply, welfare programs and so on. Accumulated evidence from people, many people, Bob Moffat, Ravi and his colleagues, Abhijit Banerjee and others. All of that is true and clear. Incentive effects at moderate marginal tax rates are pretty small. They're not zero, but it could exaggerate it. On the other hand, incentive effects at 100% marginal tax rates benefit withdraw rates in social policies that imply 100% marginal tax rates on poor people are really dumb. That's a bad idea. Perfect targeting is a horrible idea. How can we do better? One aspect is better measures, measures to inform public debate. I often go back to the issue of inequality of what that Amartya Sen has come back to often and reminded us often. We really do need to demonstrate the welfare relevance of the measures we use. We don't do that enough. I think welfare is the anchor. Some concept of welfare is the anchor for everything we should do in poverty inequality measurement. I've always thought that the money metric approach is preferable. We've got a lot of advances in the money metric approach. I'm not a fan of multidimensional poverty inequality indices, but we already can have that discussion. It's a good discussion. I see much more scope for better money metrics of welfare. The fact that you measure poverty inequality in a money space doesn't mean you think money is all that matters to people's welfare. That's nonsense. It's how you measure money, the money metric of welfare, how you map from welfare to money. Obviously, that's going to be contingent on many factors. The axioms of poverty inequality measurement, some are pretty clearly acceptable, others are not. Scale independence axioms. I mentioned the talk yesterday. Decomposability we've mentioned. Another area is the evident lack of attention to what's happening to the poorest, and yet we see in social policy all the time emphasis on leave no one behind. You see that expression, read that expression 100 times in the sustainable development goals, for example, and in social policy generally. But we don't measure that. We don't measure the floor. We measure the portion of people below some poverty line or a distribution sensitive poverty measure, but we don't look at how the poorest are doing. When we do, in fact, it's looking not looking so good that progress against absolute poverty is not coming with raising much progress in raising the floor, the lower bound. Not the biological floor, that should be below the actual floor, but the actual lower bound of the distribution of consumption income. Yeah, thanks. Better evidence we need on all of this in terms of policy, the measurement concerns that I've touched on already. What does the evidence suggest? It's not real good. Roughly a third of the poorest, about the poorest quintile globally and not receiving any help from any direct form of direct intervention, including interventions that claim to be helping them. We've got roughly one billion people living in poverty. We've got roughly one billion people receiving some form of help today in the developing world. It's just a different billion. They don't overlap that much. Current social protection policies, this is broad picture, but in globally current social protection policies are raising the floor of about 50 cents a day. Almost all of that is through social insurance. Social assistance is raising the floor in the world today by about 1.5 cents a day on average. Spend more on social policy. You have more impact on the lowest level of living, on the poorest people in the world. It's largely the variance is largely to do with how much you spend. It's not about how you spend, but it should be a lot to do with how you spend. Challenging constraints and here I just want to mention going forward, we know a lot of the constraints pretty well, and I mentioned that the obvious ones here, budgets, incentive effects, political economy, but there's one not so sexy constraint, information that I think is absolutely crucial and is under undervalued. There's some recent evidence of a series of papers that I've done with Kate Brown and Dominic Van Der Waal is here where we've been trying to really quantify the information constraint, working particularly in sub-Saharan Africa, but it's pretty alarming. Here's one little gem from it. Even with a budget sufficient to eliminate poverty with perfect targeting, not something a policy I recommend as I said already, but even if we had enough money to eliminate poverty in sub-Saharan Africa, with the information we actually do have, the targeting methods we actually do use, we're going to leave about three quarters of the poor untouched. Our data, our instruments for targeting, and all the information requirements and incentive effects you put all together, it's not looking real good, particularly in low-income countries. It's kind of a cruel irony, the poor of the country, less effective we are today, and that's a huge challenge going forward. There is a new role I believe for redistributive interventions, but there's got to be, I think much smarter than the ones we talk about now, there are no magic bullets here, you're going to have to adapt, you're going to have to worry about all the concerns I've talked about, senior effects, behavior responses, behavior responses are on one side of the debate, behavior responses are greatly exaggerated, as I mentioned, the example of Ricardo. On the other side of the debate, Ricardo's representatives of Ricardo are life's day saying the same thing, and the other side, the behavioral effects are entirely ignored, incentive effects are ignored, and that doesn't make a lot of sense, it's a middle ground that's missing. One option I keep emphasizing, I'm not a unconditional fan of any policy, but I think going forward we're going to have to think more seriously about untargeted, more universal provision in poor countries. When this country was very poor, as poor as probably any country in sub-Saharan Africa today, social policy emerged also, particularly in Sweden, nearby, and they didn't start with finally targeted administratively costly sophisticated interventions, exactly the things we're recommending in low-income countries today, they started much more universal, basic health and education. They were less ambitious because they knew they couldn't do it, and yet we're telling low-income countries all over the world that the canon should be finally targeting their interventions, it doesn't make a lot of sense. So UBI, universal basic income, should be the benchmark. I think in many countries we can do better than universal basic income, and I think it should be universal full basic income, not just income, including imputed values for in-kind services, so we fold in the issues about universal health care and education, that should be the benchmark. If we can't do better than that, fine. In most low-income, middle-income countries today, we aren't doing better than universal basic income. But universal basic income would do a better job. Now, if you're asking me, should you do a UBI and say, France? I'd say no, it doesn't make sense, but in a lot of countries it makes more sense. So keep it on the menu. These are my conclusions, I think I've said all this, we need well informed debates on the intrinsic and instrumental case. We need sensible policy interventions and proposals for both pro-pro-growth and redistribution. Stop focusing on targeting as the objective in all of this. Focus on the distributional outcomes and of course policy relevant monitoring and evaluation, learning from our mistakes. Thank you.