 seen a lot of progress in the era, but there is so much still that remains to be done. And even more importantly, as Professor Valium will address, also thinking about what can we actually do to reduce poverty and eventually get rid of poverty altogether. To some extent, this is about helping other people. A lot of it is, of course, people helping themselves. But it's also the more fortunate people helping other people get out of poverty. And I think this is an extremely important theme on a much wider scale than just linked to poverty today in the world. Quite fittingly, this Friday we opened an exhibition outside here in the atrium about the Great Famine. This was the Great Famine, not in Africa or some of the countries that we used to talk about. This was the Great Famine in Samara, Russia in the 1920s. This was also a time when the Swedish Red Cross put together one of our first humanitarian relief efforts. That was really the starting point for that kind of intervention and helping people in great need not so far away from us. But still, again, it really tells me and shows me and it's an inspiration when it comes to thinking about how we can actually help people in greater need than ourselves at times of troubles. And I think I would welcome all of you to actually take a look at some of the pictures out there. It is a very strong reminder of what some people have to suffer even in the world today. And I think it's also an inspiration in terms of helping others in the world today. And I think this is something that serves as some inspiration also when I'm involved in developing economics and helping others. So I invite all of you to that. But now back to today's discussion. I will hand over soon to Lena from Sida. But before I do this, since I'm sort of your local host here at the school, I would like us all to just extend a warm hand of welcome to all the guests from near and far. And thank you again to Finn, to Martin, to everyone involved in organizing. So warm welcome to the school. Thank you. Thank you, Torben, for that. And on behalf of Sida and also of Sweden, I would also like to warmly welcome all of you to this lecture. And especially, of course, warm welcome to Professor Martin Ravallion. And Sweden and Sida, we have supported UNU wider even since it started. That is now 30 years. And we think it's important that in this era of global challenges and, as you may know, now we have important goals for sustainability. There is increasing need to understand and analyze the mechanisms for policy option and for economic understanding more than ever before. And we, Sida, we support wider as a node for research, especially in the era of development economics and engaged in important but also challenging task of linking research to policymaking and also contributing to capacity development in low income countries. But we also support wider as a source of knowledge for our own work at Sida. And so many of wider research areas are of direct relevance for us as an aid agency. And of course, this relevance is today's lecture is a very clear example of this relevance. Sida, as Sorbjörn correctly pointed out, the overriding objective for Swedish development assistance is to promote better living conditions for people living in poverty and oppression. And the task is still very big. And the task of Sida to fulfill with the objective of poverty alleviation is of course to fill this with concrete action. And within this, we need to understand the multidimensionality of poverty and understanding what interventions are possible and efficient to counter poverty. So this understanding and this analysis is vital for us to do a good job. And in recent years, we have been increasingly aware that providing direct cash to poor people has a potential to fill financial gaps. And it has the potential to be a tool for empowerment and also to promote market development. But we are also aware that it can be problematic and that successful implementation depends very much on details, context, circumstances. So we need to develop our understanding and skills in this area to understand where it works and what it takes to make it work for poor people. So I'm looking forward to this, not only to have a better understanding, but also I hope that Professor Martin Revellian will really challenge us in this area so that we can do a better job to support poverty alleviation. So warm welcome to all of you and thank you. And over to Fien. Thank you Torbjörn for being both a great host and for a warm welcome. It's appreciated. And thank you Lena and Cedar for support. Without that support, our work would not be possible. Now dear colleagues, friends, it is really a great honor and a pleasure for me on behalf of the organizers to introduce to you Professor Martin Revellian. He's the 2016 wider annual lecturer and I guess I can't sort of resist of making the point that this is the 20th wider annual lecture. Following up on last year's lecture, Amartja Sen, you will actually find his lecture outside as well as on the wider website together with the previous 19 lectures. Martin is inaugural Edmonti Revelliani, Professor of Economics at Georgetown University. And previously he was the director of the research department of the World Bank where he worked for 24 years from 1988. Now Martin has researched extensively on poverty in developing countries and policies for fighting poverty and in 1990 he proposed what has become known as the one dollar a day poverty line and since then and his colleagues both within the bank and elsewhere have monitored progress against global poverty reduction by this and other measures. There are very few people in the world who has advised so many governments in all regions of the world and its national agencies as Martin. He's also the author of five books and 200 papers in scholarly journals and edited volumes. Now his new book Just Out, The Economics of Poverty has just been published by Oxford University Press and as one sort of looks at the wise words that have been added here, I for one take note of the fact that the chair of the wider board, Professor Ravi Kanbua says the following. This book is a tour de force covering history of thought, analytical tools and policy issues. It provides an indispensable introduction to the economics of poverty. Martin Ravallian is a global leader in the field of poverty analysis. His book would prove to be of lasting value not only to students but also to seasoned researchers and policy analysts. I strongly recommend you to have a look, take a flyer if you find it outside. We should have sufficient numbers and then get on reading it. It's a wealth of information and analysis. Martin is also a senior fellow of the Bureau of Research and Economic Analysis of Development. He's a research associate of the NBER. He's a non-resident fellow at the Center for Global Development and he's president of Society for the Study of Economic Inequality. In 2011 he received the John Kenneth Galbraith Award from the American Agricultural and Applied Economics Association and in 2016 received the Frontiers of Knowledge Award from the BBVA Foundation in Madrid. Much more could be said but I for one would like to sort of also add that Martin is a prolific teacher. I mean he is taught economics, for example the LSE, the ANU, Princeton and Paris School of Economics. So Martin's importance for the profession goes much beyond just the research. It goes on to this very important work of trying to work on building capacity of economic analysis across the world. Now during his lecture Martin will discuss the economic and political issues of the use of direct interventions such as cash transfers and in-kind contributions against poverty. There's much hope for these interventions but also much frustration. The performance in reality has often fallen short of policymakers' expectations. The lecture will review the evidence on both the successes and failures based on the many impact evaluations that have been done over the last 15 years. I think the topic could not be more pertinent. Urs has also been referred to both by Torbjörn and Lene. Martin we all very much look forward to your lecture. The floor is yours. Thanks very much Finn. It's really an honour to be giving this wider lecture and be honoured to be giving any wider lecture but the 20th in particular. And it's great to be back in Stockholm and thank you for coming. When Finn and I first discussed her topic for this and some options he very much favoured a focus on an important set of policy issues but a critical focus. That's what I'm going to give you today. The importance of being critical here is a natural thing for any researcher to do. We're never quite happy with anything. But the point of being critical here is to move the discussion further and to mobilise thinking in ways that are positive for poverty reduction more broadly, to recognise the limitations of our knowledge as well as applaud our successes when we've had them. The topic today, direct interventions against poverty is particularly salient because we're seeing across the world we're seeing developing countries taking up these kinds of interventions in a big way. When we look at the record against poverty reduction it's a mixed record. I mean if we focus just on absolute poverty in this picture I've given you here. Absolute poverty by a dollar a day, the numbers of people who are poor by that measure are declining appreciably over the last 30 years but also seeing rising numbers of people who are relatively poor meaning they're poor by the standards of the country they live in. So absolutely poor here means you're poor by a global international standard, you judge people by the same real consumption level in different countries and relatively poor means that you're no longer poor by that standard but you're poor by a standard typical of the country you live in and obviously richer countries have higher poverty lines. The average poverty line is developing world is around a dollar a day, a dollar 25 a day at 2005 prices, a dollar 90 a day at 2011 prices, somewhere around that number but across the world we see big differences and the highest poverty line I've found in the world is in Luxembourg and it's $43 a day at roughly the same average income the poverty line in the United States is $13 a day. You've got to realize that there are differences, the reasons for those differences and that they reflect aspects of poverty that aren't included in this dollar 25 a day line anchored to the poverty lines found in the poorest countries. But another aspect of this is important the continuing exposure to risk, poor countries even in rich countries, employment shocks, health shocks, agrochromatic shocks, a host of things that everyday life for poor people across the world is a risky venture and we're not just talking about the kinds of global crises that you get attention in the media, we're talking about crises in everyday life. Further, a realization that economic growth is a wonderful thing, I mean it's very hard to get sustained poverty reduction without economic growth and that's what's delivered a lot of that progress against extreme absolute poverty in the world and we should recognize that. But we also recognize that it hasn't done much for relative poverty, it's a mixed blessing from the point of view of inequality, we're seeing inequality rising in half the countries in the world, growing countries and inequality falling in the other half. The way inequality is conventionally measured, if we focus instead on absolute inequality, the absolute gaps between the rich and the poor, that's rising in virtually every growing country in the world. Absolute inequality, not relative inequality which is about ratios, absolute inequality about differences, absolute inequality is rising everywhere. Pretty much, there are very few exceptions. We're also seeing new evidence that the poorest are being left behind in this process. We've got to applaud the success and the success I show you here is undeniable at least against absolute poverty but we've also got to recognize that the poorest of the world are not participating as much in this progress as we would like. If you try to measure the average living standards of the poorest people in the world, they're barely above biological minima and they've stayed hovering above biological minima over that 30 year period. Our success is there are fewer people living near that level but in terms of raising the floor, we're making much less success and that's an important objective for social policy as well. We also see losers and gainers, we see churning in this process of growth with absolute poverty reduction in the developing world, churning meaning there are winners and losers, some people gaining, some people losing. When we focus just on the average, we miss that. We miss that churning going on. There's actually very little we can do in economic policy which will not have losers and we've got to realize that. There too is another role for social policy. Now we're seeing this new attention and it's quite dramatic really since the new millennium about 2000, there's new attention to direct interventions and it's formulated I think quite well in the SDG 1.3, implementing, I've given you implementing, to implement nationally appropriate social protection systems and measures for all, including flaws, I had absolutely nothing to do with writing this but I applaud this including flaws. I think that was my influence and by 2030 achieves the substantial coverage of the poor and the vulnerable. But how are you actually going to do that? Developing countries since 2000 have been adopting, turning to social protection policies being a big way. If you go back to pre-1990, it was really quite hard to find a developing country with anything you'd reasonably call a social protection policy or social assistance or what I'm calling direct interventions today. That's changed dramatically. In fact now it's hard to find a country that doesn't have something. However ineffective and weak but something and the growth rate is extraordinary. That growth rate in that last point here is of 3.5 percentage points per annum in the coverage rate, the percentage of the population covered by social protection and developing world. We're seeing an explosion of interest and spending and real resources going into these policies. So remember we've got one billion people and absolutely poor. We've also got one billion people who are in some form of social protection. The problem is it's a different one billion. These two curves, these two sets show you this intersection. I'm giving you a rough stylization. We've got one billion people living in absolute poverty in the world. We've got one billion people getting some form of social protection now but it's the intersection of these two sets is quite limited. I'm going to show you that in a more rigorous way in a moment. But interestingly the intersection gets bigger as countries develop. When we look at the poorest countries in the world it's very little intersection. As you move to richer countries it gets bigger and bigger. I'm sure in Sweden it's quite large and it's not complete. I'm sure in Sweden there's also further work to do. This is a graph of the coverage rate. This is something I compiled from a data site of the World Bank on the coverage rate of social protection policies in the developing world. So the vertical axis here, safety net coverage for the poorest quintile, that's the blue line here, and safety net coverage for the whole population, the red line. And I plotted that against GDP per capita for the year of the survey. A couple of observations here. Really two main things I want to point to. One is look at how low the average coverage rates are in the poorest countries. Average coverage rates for the population of the whole and average coverage rates for the poorest 20%. But second point, look at what happens to the two curves as GDP increases. You move to higher levels of GDP. They diverge. So richer countries have greater coverage and as I said before they tend to be relatively better at assuring that coverage is reaching poor people. But here's our challenge, these countries. Now there's a variance. Those are averages, but you can see there's a scatter around that. And we see that there are some poor countries that are doing much better in terms of coverage of the poorest 20%. The one billion is roughly the poorest 20%, right? Much better. So there's a variance in performance. Another interesting thing, if we think about today's rich countries and we go back to when they were as poor as today's poor countries and you don't even have to go back about 200 years and in some cases only 100 years will do it. I'm not sure about Sweden, I should be but I'm not. But for Western Europe if you go back 200 years you're looking at a region of the world that was as poor as sub-Saharan Africa today. We have reasonably objective measures now that data is not perfect but it suggests that that's true. In the period where today's rich world made all that progress against extreme absolute poverty we saw a rising floor. We saw if that progress in lifting people out of poverty came with a rising floor in a way that we haven't seen in the developing world. Interestingly the developing world is accelerating out of extreme poverty but it's doing it in a rather different way. And over a shorter period than today's rich world did. It was doing it a different way and the difference is a lot to do with the effectiveness of social policy, the topic today. OK, that was the prelude. The lecture is now going to critically review the policy issues in this class of interventions. I'm going to first try to define the role of the interventions in a broad sort of way. I'm then going to look at economic arguments, foreign against, and I'm going to give you three case studies. I pick one from England, but England when it was poor by today's standards. Then I'm going to pick one of the two largest social protection programs in the world today in India and China. Three case studies, England's poor laws, India's National Rural Employment Guarantee Scheme and China's Debail Program. And those two programs in India and China orders a magnitude larger than anything else in terms of population coverage. Now, when you're India and China, it's not too difficult to do that. China has the largest social protection program in the world in terms of people coverage. Of course, 1.3 billion people. That's not so difficult. OK, the themes for this talk, this really summarizes. I hope you don't leave after this slide, but this is really the whole talk. There are no sort of magic headlines here, but a series of issues that I think we have to take much more seriously. Protection and promotion both are valued, but it's unclear if governments get the balance right. Protection and promotion, I'm going to define what those mean. I'm going to use those as the twin objectives of social policy. But are governments balancing those objectives in an appropriate way? And the reasons to think that government failures are a problem here. Another, there is a region of trade-off between protection and promotion. You can protect people too much from the point of view of their promotion, but you can also exaggerate that trade-off. And I think that trade-off is seriously exaggerated by many critics of social policy. So I want to recognize the problem of incentives, and you can't scratch any economists, you get a problem of incentives. So as an economist, I've got to take that seriously. But I also think that some critics of social policy, and generally not serious economists either, do exaggerate the problem. Incentive effects are important, but there are many other constraints which are far more important to achieving better social policy in poor countries. And remember, this is about achieving social protection in the poorest places in the world. I'm going to argue also that targeting is a fetish. It is something we've got to rid ourselves of in thinking about the objectives of social policy. It is one thing that sometimes helps, sometimes not, but it's become an obsession in thinking about this class of interventions in the developing world. Another point, see, information technology can help. There are smarter social policies we can construct now in poor countries, and we've got to realize the potential of new technology there. And finally, evaluation and monitoring are hugely important. That sounds like a cliche, but it's really important. This last point is actually just as important, adapting to evidence. We see a lot of evidence and we see a lot of research, but we don't see nearly enough adaptation to that knowledge. Okay, the policy problem. Twin goals of protection and promotion. Let's first think, why do we have poverty? Well, in a nutshell, these two slides give you the answer. First, we've got to realize that there can be poverty and inequality in a fully competitive market economy. The ideal, the kind of textbook market economy we learn about or we teach our students can generate unequal, have unequal endowments. We can have problems of low productivity, lack of marketable skills, social exclusion, geographic isolation, debilitating disease, environmental degradation, are all entirely consistent with an efficient market economy. The trouble is, we don't have an efficient market economy. We've got all kinds of market and government failures that in themselves generate poverty. Prominently in those is the very idea of the dynamic poverty trap arising from threshold effects. All kinds of ways in which there are minimum thresholds in terms of wealth in particular, human and physical wealth, which you need to reach if you're going to be productive. Sitting here in this room, listening to me, you're probably burning an average of... Well, my students at Georgetown, when I tell them this, they usually come up with a figure of 1,300 calories per person per day, but looking at this audience, I'll give it a little bit extra, actually, or maybe... I don't take it wrong, let's say 1,400. That's what you need to do absolutely nothing. Just listening to me doesn't burn many calories. I hope it doesn't, if anybody's getting upset, but it doesn't. You need to get to that magic number, it's called basal metabolic rate, to be productive, to work. Physical labor requires you get there, and that's a threshold effect. But there are other threshold effects too, minimum levels of education. My students at Georgetown know full well, and students everywhere in the world know that you can't get a job with epsilon of education. You've got to get as a minimum threshold of education. If you're an entrepreneur trying to start a self-employment activity in a poor country, you've realized that you may need as a minimum threshold of capital, physical capital you need to start that enterprise. And we understand, from a more theoretical literature in economics, we understand full well, that these thresholds, under plausible conditions, can generate poverty traps. Situations, low level attractors, they're called in economic situations in which people are stuck. There's a large, a large increment to their wealth is needed to escape poverty. By the same token, that also means that there are people who are maybe comfortably well off, they're not in a poverty trap, but they can be vulnerable to large negative shocks. They can be thrown into destitution if the shock is large enough. That's those are implications as the existence of threshold effects. There are other problems too. I call the site as titled market and governmental phase for important reason. These are market phase, but the existence of a market failure implies a government failure. The role of governments is to get rid of market failures. So if market failures persist, governments are failing. Geographic poverty traps, very important. External effects of living in poor areas. Given your own endowments, your productivity may be lower than it would otherwise be, given poor infrastructure, negative externalities associated with living in a poor area. Also with incomplete markets, we have uninsured risk and that uninsured risk is spilling over into production decisions everywhere in the world today. Kids are being taken out of school because of risk. All kinds of opportunities for investment are being foregone. Crime is often a manifestation of this. So in that context, thinking about those as the causes of poverty, let's think about policy. Two types of anti-poverty policies and I make this distinction between protection and promotion was influenced by a book by Jean Dres and N'Matthew Sen, Hunger and Public Action, which came out of a wider research project, as I recall, the first wider research project, which I was involved in back in early 1980s. It came out of it through a sequence of time, but they make this distinction between promotion and protection. I think I'm defining it in another paper more rigorously, but I owe it to them. Protection policy simply means things that are short-term palliatives for assuring that the consumptions don't fall below some critical level. Promotion means things that either get people out of traps or raise their productivity. You can be poor, but not trapped. You're at a nice, what we call a stable, preferred, long run equilibrium. It may be low, though, it may be of low level of wealth, and your problem is productivity. There are some people, the problem is a trap. They really need a significant boost to their wealth if they're to get onto a trajectory to their own preferred long run equilibrium. So promotion policy is either about allowing poor people to break out of a trap or raising productivity. Now, protection. When I started writing this book, you'll see that the first third of it or so is a history of thought about poverty, anti-poverty policy, and I learned a lot from doing that, writing that part of the book. One thing I learned is that people have been talking about protection, the protection aspect of social policy goes way back. Ancient Asia in both ancient China, ancient India. This guy here is Kottilya, the first economist, in my reckoning, 300 BC, India, what is now called India. And Kottilya wrote a book, The Science of Statecraft. It's a nice way to define economics. And he wrote in this book all the things that the government needs to do, and that included social protection. He had stuff on monetary policy, it was an amazing book. I mean, it's crude by modern standards, but this is 300 BC. But other examples in writings in China and in ancient Europe. Promotion is a new idea. This really started, probably, my reckoning's late 18th century. So we're talking 250 years, a period when promotional policies started to become important. But the debates, they took ages. The debates on promotional policy in Europe, it took ages. Struggles for those policies, the labor movement was involved, individuals were losing their freedom or worse in the struggle for promotional anti-poverty policies in today's rich world. With economic development, we do see an increasing emphasis on promotion. Protection tends to dominate in poor countries. And the key question I want to pose to you today is does protection keep people poor? The kind of argument that we hear about from critics of social policy. So I don't think there's any doubt that that's possible. You can protect people too much from the point of view of long-term poverty reduction. So such a trade-off can exist. I'm not gonna deny that possibility. We think of it as like this. We've got promotion as an objective, protection as an objective, and this is some isosocial welfare contour. This is just saying that along this point, I'm indifferent in terms of the protection of promotion. So I allow substitution between them. And we've got a set of possibilities for the economy and the existence of this promotion protection trade-off can be thought of as thinking that the tangents to this function are linear or negatively sloped. There is some trade-off. We can protect people at a cost to their ability to promote them. I think that's likely, but I also think it has been exaggerated. When we look at the debates on social policy, I'll come back to the debate on England's poor laws, we see that exaggeration from some of the most prominent thinkers of the time. And I think we see it today too. Difference probably is that there was the majority of thought back then and I think a minority of thought right now. But also when people debated social policy 200 years ago in place like Sweden, they really did not have evidence on these things. They didn't know about labor supply responses. I'll give you a quote soon from David Riccardo, which is striking here, but we now know a lot about this, particularly in the US, the country I live in. The bulk of the evidence in the US is now pretty convincing that labor supply responses in particular through work effort to protection policies, they exist, but they're modest. You will still increase people's real incomes, you'll raise their welfare through social policy transfers and so on, even allowing for the labor supply responses. They're there, but they're not huge. More evidence, and this is a direction for research, more evidence is needed for developing countries on precisely these labor supply responses. Because in developing countries we have a big difference, the informal sector. The informal sector means that the choice margin becomes very much affected by the opportunities to work in the informal sector. When the informal sector is relatively non-existent, a very small part of the economy, it's not an option for most people when you have a large informal sector, and that's gonna change things a little bit, and we need more research on that, consistent with the kind of work that's been done in the US. But I think you can also see situations, and a large body of researchers consistent with this, which suggests that the trade-off may not exist for countries when they're very exposed to risk, very vulnerable. Here are the way I've drawn that. When we've got people who have very little protection, so I mean by vulnerable, they're very exposed to downside risk, for them, more protection is actually perfectly consistent with more promotion. It relieves constraints on their choice, there can constraints that are coming from market and non-market institutions. In the short run, in an unemployed economy, it can also mean a gain in aggregate effective demand. This was the point made in chapter 22, the General Theory by John Maynard Keynes, that in an unemployed economy, the whole calculus of incentive effects that are associated with redistribution, the whole calculus flips. Now we're going to have poor people with larger, high marginal parentheses to consume, they're going to consume more out of redistribution than rich people, and that's going to increase when there's a demand of constraint that's going to increase aggregate effective demand in the economy and increase output. But even in a fully employed economy, we're now realizing there's a host of ways in which this trade-off ceases to exist and may even become a win-win situation. You can have more protection with more promotion. I've mentioned some of them, credit market failures in combination with diminishing returns to capital can generate that. Political economy and polarized societies, we tend to get dysfunctional states, not that I'm thinking of a country I live in at all. Multiple equilibria, poverty traps, I've already mentioned, protection from large negative shocks may be crucial for sustained promotion. When that threshold exists, you may be well above it now, but if it exists, you are vulnerable. In some sense, everybody's vulnerable to large negative contractions. There are also neglected constraints. I've said incentives need, you have to take incentives seriously, but we also have to be careful about them. We need evidence and we need to base our assessments on the evidence, not on ideology. But there's a whole lot of other constraints that may be far more important in practice, particularly in developing countries, and information is a huge one. The poor in developing countries are not easily identified. Our information here is quite weak. Proxy means tests, which have become a technology for social policy across the developing world now, proxy means tests, in my experience, are pretty poor proxies. They really don't identify poor people terribly well, because there's errors of inclusion, errors of exclusion. Better social protection policy in poor countries is going to require better data and better mechanisms for revealing the data. Mechanism design is going to have to take data revelation much more seriously. Administration, weak states get poor service provision, poor governance, all of this is, people worry about corruption in developing countries a lot, but the other side of the coin to corruption is weak states. You tend not to get corruption when you've got administrative capability in a country. Political economy, I've always loved this expression from Larry Summers in the early 1990s. It's not, I don't agree with everything Larry says, but I love this one. Programs are poor programs and not sure. In other words, that if you focus and target interventions too much on poor people, you'll lose middle class support and you won't have a sustainable agenda for poverty reduction. Political economy, and I'm going to come back to the constraints on political economy at a number of times. Another set of constraints, which aren't well understood, are about flexibility. It does appear to be really difficult to achieve state contingent transfers in poor countries. The success stories are all about targeting interventions to poor people but not in a very flexible way. In other words, the transfers don't adjust to changing circumstances. It gets hard for somebody to get to shock. It becomes hard to get onto the program. Somebody has a windfall gain. They'll still be on the program. There's a stickiness we see repeatedly. That can come from a number of reasons. Fiscal stresses to governments that can be very hard to accommodate large covariate risks in social policy. Participant capture, it seems to be a recurrent problem and it generates a stickiness. And I'll come back to participant capture. Moral hazard at local level. This is something that really is not at all well appreciated but I think I'm really raising it. Well, the first thing, it's not the first to raise it. Moral hazard problems. Local governments can expect the centre to help in a crisis and I'm going to illustrate this in the case studies. This is really important because if you're decentralising social policy as we tend to do in developing countries and we don't really have much option to that often. When you decentralise social policy to local governments that think that the centre is going to help in a crisis, you're going to have a moral hazard problem. The local implementing agents may well undervalue protection relative to the centre. And that's going to be an issue later. Administrative capacity for better social safety nets is going to require the kind of infrastructure that we don't have in most countries but we're starting to get. I mean, India is a famous example. We've got now, I think, 900 million people with a biometric identifier in India. This is amazing. You get, when we get to full population coverage in India of this biometric identifier, the social policy options are going to change dramatically and that's a very positive thing. There are costs along the way and the concerns about the privacy implications but, on the whole, I think this is good news. There are many challenges here in making a pro-poor social net politically sustainable and I've mentioned some cases already but we have to worry about this. State and contingent social safety nets that ensure have a broader base of support than the current set of beneficiaries. Protection, comprehensive protection, the kind of, you've got in the kind of Nordic welfare model. Comprehensive protection is an important lesson for developing for poor countries too because comprehensive protection means that people who don't today or maybe in a normal circumstance aren't going to turn to these programs but the comprehensiveness of the insurance will give it, will let them support it. We saw this in Maharashtra in the early 70s and the famine that hit there and there was a very successful intervention by the Maharashtra government introducing the Maharashtra Employment Guarantee Scheme which was scaled up nationally in 2005 we're gonna come back to that as a case study but that program illustrates well you've got a large intervention of that sort which was feasible politically because you had a broad class of, much broader class of people in the society a much broader segment of people in the society who saw the benefits of the existence of the scheme because they were exposed to risk and they realized it. They knew, they didn't know about the, talk about things like poverty traps but they knew there was a downside vulnerability a large enough shock could throw them into destitution even though today normally they're not turning these programs. So achieving that political broad base of political support is often about providing effective insurance for everybody. The universality of protection can be very important to the sustainability including the scope for promotion. Community based approaches have a lot of promise but also they must recognize that they come with some difficulties too just like we have participant capture or a more familiar concept is elite capture in community based interventions and we've learned that that happens but again I don't think we should exaggerate it we've also learned that community satisfaction is actually crucial also crucial for the sustainability of social policies. Communities in poor places have to feel that this is right and fair policy and that's crucial to their support and you will need their support. Public information campaigns and other set of issues that I take very seriously and I've written writing on a lot now a number of issues here but most importantly I think this last point the credibility of any new social safety and putting any new social safety in place it's credible if you introduce it before you get rid of the old policy. Nine out of ten times that's not what happens. People get rid of the old policy introduce the new policy and there's a credibility problem. People don't trust the government big surprise and because of that you'll get resistance you get rights in the street when you take away the broad based subsidies on normal goods that we see in most developing countries because they don't trust the government to supply an alternative. They may fully accept and be happy with that alternative but it's not there yet. So very important start the new policy have it in place before the old one is eliminated deal with the cost of that. Okay improving the trade-off social policies that try to protect and promote and this is really the new direction that a lot of thinking is taking. Here the message is striking. Your incentives which have been so emphasized by critics of social policy that social policies make poor people lazy and dependent and they don't want to save they may be if your mouth is arguing that they have too many children. All of these adverse incentive effects is also positive incentive effects and we can use incentives to develop better policies. All kinds of ways in which the mech what economists call the mechanism design can be really important. The incentives you build into the social policy can be very crucial to its success in both protection and promotion. All kinds of examples of that self targeting is one of them subsidies on inferior goods. The more you can obviously the more you can move subsidies from normal goods to inferior goods the better. Normal goods meaning goods which for which demand increases with income and inferior goods meaning goods which demand decreases with income. So you don't have to be an economist to realize that that's an important direction for social policy reform. If you have inferior goods that's probably in the problem in practice finding inferior goods. Workfare is another example where you build mechanisms into the design work requirements that do two things essentially they discourage non-poor people from participating and they encourage poor people to not participate when they don't need the program. So you build incentives into the design of the program to assure that it achieves better protection and promotion. Other efforts to improve the terms of that trade off this term social investment has been become quite important in the literature and all of these are around the introduction of explicit conditionalities. We're gonna talk about the most prominent example but there are others conditional cash transfers. This is an idea that in various forms I found a hint of the idea quite a explicit hint. Hints a bit is not quite the right word. In Adam Smith the wealth of nations 1780. He understood the idea of a conditional cash transfer. He talked about tuition subsidies targeted to poor families. But it didn't happen. It wasn't policy for a long, long period maybe in selected places some universities might introduce it and so on but it wasn't a comprehensive social policy for a long time. Some countries, my own country Australia introduced equivalent of a thing of a conditional cash transfer in the 1940s I believe or 30s. But there are other examples in very importantly in the developing world a couple of initiatives Bangladesh's food for education program at about the same time progressor in Mexico now called Opportunity Artists actually it's called something else now it changes its name regularly but same basic program. Bolsares Scholar in Brazil. The research on these interventions if anybody ever asks you for an example in which research on social policies influence social policy this is a big time. Those policies have been scaled up to 30 or 40 countries on the basis largely on the basis of the research. That research has pointed to very important benefits to poor households from these programs. The sustainability of the programs depended crucially on the research findings and the research findings have generally been quite positive. But now our concerns since this is a critical discussion I'm generally a supporter of CCTs I'm not sure about the C and that's one of the concerns, the conditions. There are concerns about inflexibility we're not seeing the kind of state contingent insurance that we would think is ideal for any social protection policy we're not seeing that in CCTs we're seeing a degree of stickiness in terms of who gets is on the program and who's not. We're also concerned about the conditions do they really change behavior in a positive way and some evidence that in fact you can get very similar outcomes without the conditions because those conditions can be costly to poor people. You are telling poor people that you've got to keep your kid in school longer than you would have liked to that means that current you're telling them to change their inter-temporal trade-off. If you take the kid out of school you've got an income gain. So you're telling them you're going to take an income loss now and we're incentivizing you to take that income loss i.e. keep your kid in school longer. Now the problem, part of the problem with that is it's paternalistic. How do we know? The poor families are making the wrong choices here. I think like many economists I'm very irritated by paternalism whenever I see it. You've got to make the case and I don't think advocates of CCTs often have made that case. And this last point, is the problem really on the demand side? Maybe you're incentivizing people, kids to stay in school when there's no teacher there, there's no textbooks, they don't learn anything. All the problems of education and maybe also healthcare are on the supply side. Whereas the policy assumes that that's all working fine the reality is very different. Another area for improvement making work fare more productive the work fare programs we see across the developing world asset creation, it rarely gets the kind of emphasis I think it should get including in India which loves these programs and there's an important scope for change there. But also trade-offs. If we try to make work fare more productive meaning that you get in a work fare program you're required to work for the transfer payments that are made. The mechanism design that the incentives building to that are intended explicitly to assure that non-poor people will not take up that work and also that poor people will leave the program when it's not needed both sides of the incentive argument are present but asset creation may be hugely important asset creation for poor people may be undervalued. The way of making work fare more appropriate for promotion is to obviously assure better asset creation. That emphasis has been lacking. Targeting, before we get to the case studies and I just want to say a bit about this the motivation for targeting comes out of the limitations perceived limitations of untargeted interventions and we see that in many ways. The classic untargeted intervention is a basic income guarantee and part of the problem why we haven't seen basic incomes basic income is where you give a transfer payment to everybody of a fixed amount whether poor or not. So I love telling my Georgetown students about the referendum that's coming in Switzerland this year you heard about this, they're going to vote for a CCT, it's not a CCT a unconditional transfer payment a basic income guarantee, no conditions of $2,500 US dollar equivalent per person per month. You get that for doing in nothing. My students at Georgetown love this idea. $2,500 a month and actually you get it from year nought just built up in an account the idea is that by the time you're 18 you'll get this big lump sum transfer. And now all of social policies are folded under this it's not like a new thing that's added on top of the bill of the old policies it replaces everything unemployment insurance, all of that goes. Some people call it citizenship income that's the right of citizenship in the country. It's actually quite an attractive idea and I think it has a lot to say, a lot to go for it but one of the arguments people make for basic income guarantees is dangerously wrong. That is that it doesn't involve any incentive effects well it's obviously wrong on the income effects because there's an income effect and that's got to think of that as an incentive effect it's right in terms of substitution obviously if you get the same transfer payment whatever you do there's nothing you can do to change your transfer payment so that it doesn't provide any incentive to do anything good or bad, all right? So it's neutral from that point of view but don't forget the income effect but finally you've got to finance the thing and they're all incentive effects everything else comes into the picture it doesn't go away you go back to exactly the same set of problems that we've always had. So it's a little bit deceptive I think the arguments that advocates of bigs basic income guarantees make but it should be on the agenda and something to think about as a counterfactual for some existing policies for example. In fact, if you think about a conditional sorry a basic income guarantee financed by progressive income tax that is formally identical to Milton Friedman's negative income tax part of the reason why bigs are so current in social policy thinking across the world now is that the left and the right agree very easily on them Milton Friedman's negative income tax is a favorite social policy of the right across certainly in America but it has always been for a long time but it also gets a lot of support from the left although people on the left generally hate it when I point out that it's formally identical to Milton Friedman's negative income tax. Other concerns, responsiveness to shocks it's not going to be very good at protection. The financial affordable basic income may come at a cost to other things in terms of promotion out of poverty. Other examples of untargeted programs across the developing world we still got untargeted subsidies attached to normal goods. Obviously meaning that the subsidy goes to the non poor the bulk of it goes if it's a normal good you consume more of it as your income rises so you get more of the subsidy as your income rises and they're still common. And no way I can see any justification for this. I think more we replace those sites of policies with smarter policies either targeted to inferior goods or even basic income guarantees the better but the limitations of this whole set of policies the untargeted policies leads to the arguments for better targeting and the idea is very simple if you think about this picture we've got incomes of the pth percentile so here on the horizontal axis percentiles of the population and the vertical axis the income of each percentile so if it's the 50th percentile then the point on the vertical axis is the median. All you're doing with a perfect targeting the other extreme to universal programs all you're doing is filling these poverty gaps exactly. Here's the poverty line you fill the gap all of those gaps and you calculate the aggregate poverty gap and that's your policy and that's called perfect targeting. Any country that does that policy the rightist described is asking for serious trouble. Why? Well as I said you know we can exaggerate incentive effects but don't ignore them. What would such a policy imply for the incentives of poor people to get out of poverty? If that's literally what you do, you're doing then you're imposing a 100% marginal tax rate on poor people. What's the marginal tax rate on the rich in Sweden? Now anybody? 55, 60 so way lower than 100. Why do I say that? Because if that is the policy if that's perfect targeting if that's supposed to be the ideal that everybody's aiming for then if your income rises by 100 rupees your transfer payment will fall by 100 rupees if you're poor. There you're facing a 100% marginal tax rate. So why would you work if you don't like work? Most poor people I've talked to don't like work they would stop. So immediately the cost of this program will rise to this area. There's no point working and it will actually get larger because you'll attract some people who are above the poverty line, you'll attract them to they'll see a trade-off. Oh I work, if I stop working my income will fall to this amount my income will be a little bit less than it was before but I don't have to work and of course if you don't like work if you value both consumption and leisure then you're gonna stop. Now that's just a classic argument that's an extreme version of the classic argument that social policy critics have always made against social policies and I said don't exaggerate that argument but also although the evidence suggests incentive effects aren't large when we get to 100% marginal tax rates we're talking about it's unbelievable to me that incentive effects won't bite seriously. So this is a bad idea as a policy. Budget constraints on the other hand create another problem. We see a bias across all social policies that I've seen toward inclusion errors. In other words they're trying to minimize what, areas of inclusion error if you're making an error in social policy in targeting it essentially means that you're incorrectly classifying a person as poor. Error meaning that somebody's getting the program who shouldn't so that means that somebody who's not poor is getting the program. Now budgets determined by inclusion errors by minimizing inclusion errors you'll reduce the cost of the program and give you an illustration from England's old poor laws. Exclusion errors on the other hand are no less important and arguably more important to the effectiveness of social policies in practice. Exclusion errors are situations where you're incompletely covering poor people. You're leaving out, crucially leaving out large numbers of poor people from your policy. We're seeing repeatedly a bias towards minimizing inclusion errors and ignoring exclusion errors and I'm not the first to point this out but it's alarming. Okay, that's the review of all the issues and I point it to what I think are some of the key things to think about, partly to think about in terms of research going forward but it's all very dry. I wanted to give you some examples from real life programs. Now we could talk about dozens and dozens of programs but I'm going to take three. England's old poor laws. Well the Elizabethan poor laws were really a very intelligent policy. Elizabeth I in England came up with this idea or her advisors came up with this idea that they would instruct the parishes to deal with their poverty problem. They created a law that said that each parish had to deal with its poverty problem, had to raise the money but it was mandatory. They had to provide transfer payments conditional on certain events, unemployment, sickness, death of the breadwinner, a host of specific disabilities, illness, whatever, a host of specific disabilities, each one they had to deal with. But they had to deal with it. See what's a smart policy for the central government. It doesn't come with a big bill for the central government but that comes with a few problems along the way. One of them is horizontal inequity. If you're a rich parish, poor people in a rich parish are going to do much better in this program than in poor people in a poor parish for obvious reasons. But no obvious attempt at promotion. This was really a protection policy and every time I look through the legislation, I look through what was happening, very little emphasis on promotion. Essentially what they're trying to do is deal with state, it may provide state contingent transfers to deal with serious events which were downside risks facing poor people. Now this policy had been going on for 300 years. Some economic historians, people like Sola and others have I think convincingly argued that England's old poor laws were an important part of social stability in England. You know, toward the late 18th century, the aristocracy in England was very afraid that the French Revolution would spill across the English Channel and I think one reason was the old poor laws. But the old poor laws were financed. Who's financing the policy? Well, and the parishes have got a what's their revenue? It's going to be property taxes. So the land holding classes were financing all of this and they were getting more and more worried about it. Resistance emerged and economists started to help them with this and here I'll let you include this included David Ricardo. David Ricardo the undoubtedly the greatest economist of his time but also possibly the greatest exaggerator of his time. This is David Ricardo on the poor laws. It is quite in the natural order of things that the fund for the maintenance of the poor should progressively increase until it's absorbed all the net revenue of the country. I don't know if you're an economist or not but it's really hard to write down a model or a set of a behavioral model would give that kind of property. It's really very hard. Amazing exaggeration. But of course, he didn't have any data and nobody knew about labor supply responses or responses and savings decisions and so on. The behavioral, the economics was not there to inform these things so you could say anything you liked. The reforms were dramatic. There were these things, work houses, which already existed. They started on the 16th century. I think they started in Amsterdam. The central idea was that if you're poor, if you want relief, you have to agree to confine yourself to a work house. You would work there, you're basically incarcerated. You'd get support, you get your meals and so on. Typically be separated from your family and you're confined in this place. They're trying to control your behavior, make sure you do all the right things, you don't do any of the wrong things with the money and resources but then they feed you and that was it. The idea was explicitly self-targeting. If you look at the arguments made, they thought this would be magic. They would reduce the cost of the old poor laws that risen by 1830 to 2.5% of England's GDP. By about 1830, I guess the aristocracy was no longer worried that the French Revolution would spill across the channel. Napoleon had been defeated and everything looked okay. So then they made their assaults on the old poor laws, introduced the work houses. Huge contraction in the cost of the program went from 2.5% of GDP to 1% of GDP within 10 years. That's Peter Linder's calculation. Huge success, but what have we done here? How did we get it from 2.5% to 1% by discouraging poor people from going to the work houses? I don't believe that was achieved by excluding ineligible non-poor people. I think poor people just saw that they didn't want to be exposed to this. It was too high a price. They respected their freedom, they respected they wanted to be with their families. And people realized this. There was a social movement erupted at the same time in debate. Charles Dickens, I'm sure you've heard of Charles Dickens. If you haven't read all of the twists, just read the first three chapters. It's a great, it's a social novel, a great description of the work houses and what life is like in the work houses. This is my favorite picture of all of the twists from an old movie where he's asking for more gruel. Can I have more please? But others too, Benjamin Disraeli, the head of the Tory party, was a staunch critic, Florence Nightingale, who criticized the health conditions in the work house. The debate that erupted over this policy, it really erupted immediately. So what does it illustrate? What I've just described is almost textbook, the kind of policy debate. The work houses are an extreme thing, but the essence of the policy debate is exactly what experts from developed, from rich countries are telling governments in poor countries, finer targeting, make sure that the non-poor are not getting benefits in the program. That targeting, finer targeting by reducing areas of inclusion, that's common across the board. And we're missing something. We're missing really important things. We're missing certain costs on participants. Actually, I think if you're a genuine utilitarian, you'd think it was very ambiguous what you're doing through work houses. And I think the same point is also true of a lot of social policies in current developing countries. We have to think more comprehensively about the costs we are imposing on poor people of these policies. And that's the important message here. The emphasis on areas of inclusion undermined the new poor laws even as a protection policy. Second case study, the largest work program in human history in these employment guarantee schemes. Here's an effort, the economics, the kind of equations of the policy are pretty much exactly the same as the work houses. There's no work house, but a work requirement is imposed. You receive assistance from the state if you agree to work. And the particular form of that guarantee is that up to 100 days of work for each household per year, unskilled work in public works projects and rural areas, the work is provided on demand. It pays a peace rate such that you would earn the statutory minimum wage rate in agriculture from this work. Gives equal wages for men and women for the same work. That's the policy. It was a very successful policy politically. People argue quite incredibly that Congress party won the national election because of this policy. It had to form a coalition government with the left, but it was back into power and nobody really expected it. They came up with this policy as a smart way of winning that election, but of course a lot of commitment to the terms of that policy. And essentially what they're aiming to do, the motivation is compelling. They're aiming to provide insurance to poor people, meaning that when they need work, they can get it. They ask for it. They're trying to empower poor people to actually ask for something which is really crucial to their rationale. But the problem is the reality. When we look at this scheme, it doesn't do, particularly in poor places in India, it doesn't do what it's saying it does. It doesn't guarantee employment. We see lots of rationing on the scheme and we see a limited impact on poverty when we factor in all the costs, again, it's similar to the arguments about the new poor laws in England in the 1830s, when we factor in all of the costs of the program, including the foregone opportunities. People join this program, but they give up something. They gain, obviously, but they give up something. You've got to think about what that is, too. Then we've got non-wage costs incurred in supervision and in setting up the local bureaucracies for doing this program. My own calculations in Bihar, one of the poorest places in India, poorest states of India, suggests that this program, when you factor in all the costs, the effect of this program isn't actually no different to a basic income guarantee. If I took the same gross budget and just gave it to everybody, whether they're poor or not, I would have actually slightly larger impact on current poverty. That just tells me two things. One, that those costs are significant. But two, that unless the scheme is actually producing productive assets, it is not a cost-effective intervention relative to the options. And that actually doesn't just include a basic income guarantee. I also modeled an option where I used the same gross budget and I gave it to everybody with a rationing card in my sample data. A rationing card is issued by the government. It's called a BPL card. BPL stands for Below Poverty Line. So I do a simulation where I take the same gross budget of this program and I give an equal amount to everybody with a BPL card. That had more impact on poverty, on poverty than the current program. That's very disappointing news. Well, why? We've got to probe a bit more further than today to understand what's going on. And then the process, we learn a little bit more about why there's so much poverty in places like Bihar. And a scheme like this doesn't change that fact. The performance issues are all about the capacity of local states to deliver on programs like this. What we see in practice and the poor of the state of India, the less effective it is at implementing this scheme. And you see this in this graph? On the vertical axis, we have here the share of households who are rationed in the scheme. That means that they asked for work, but they didn't get it. Remember, under this scheme, it's demand determined. If you want work, you go to the local village leader to say, I need work. I need it badly. Please give it to me. He's got to organize a project. Some digging a tube wall or whatever it is. He's got to organize a project to employ you. Under the terms of the scheme, under these conditions. So you've got to actually act upon, you've got to act upon this. You've got to take an action. If you, and you maybe with local officials, tell us you're not too bad. I don't have any more money anymore, or you've got to come back. Come back next year. When we look to poor states of India, we see this repeatedly. They are rationed heavily. Lots of people who want work are not getting it under the scheme. And the poorer your state on the horizontal axis here, I have the head count index of poverty. The poverty rate, percentage of people living below the official poverty line. That's dramatic. This is a simple performance indicator. This is an employment guarantee scheme. The idea is that guarantees employment gives employment to people who want it. So that's a performance issue that is striking. This scheme works less well in poor places. Yet it is an anti-poverty scheme. Why is that? Because places like Bihar do not have the capacity to implement this scheme. And that is manifest in rampant corruption on the scheme and a lot of dysfunctionality. And in the end, a lot of poor people who want work but can't get it. How are you going to change that? Two things, they're also the causes of the problem. Poor information is one of them. People, I demonstrate this in a survey we did of 3,000 households in Bihar, a very, very carefully constructed survey. It's not just my observations from field work. But this is an observation from field work. This is a total qualitative observation. This is the public information about the program in a poor area of southern Bihar that I was working in for the field work. It's not an easy area to work. I mean, I tried to go there two years earlier and I was told I'd be kidnapped or killed. It was just a matter of which. It's a violent place. It's not a very safe place. But this is the public information program. It actually, Enrega is the scheme here. And this guy here is holding a job card. He's got this big smile. He's happy he's got a job card, all right? This is, now, one problem with this, that public information program is you have to be literate for a start. 50% of the population are illiterate, adult population. It's clearly not very effective. That poor information means that many poor people are unaware of their rights under the scheme. So we decided to try and change that. We made a movie. It's called B. Reg's The Movie. We decided on the basis of focal groups, we decided that we needed an intervention that would capture people. We really sucked them in. So he wrote a script with a love interest. We have everything in this story, a fictional story. We're pulling all the strings we can to try to teach people their rights under the law to replace this awful banner with something that would work. And this is a screenshot from the movie. But if you go to my website, you can see the whole movie. You'll have to understand Hindi or the Bihari version of Hindi. But you can go watch this 25-minute movie. So this guy comes back from a partner. He's a rickshaw-wielder and partner. He's coming back to see, lives in the village. He's coming back to see his family. He hasn't seen his little daughter for ages and has seen his wife. And we, you know, the expression of the violin and you know, it's all going, there's no violin. We couldn't actually afford a song and dance act, but we did just about everything short of a song and dance act. And repeatedly through this, we're just telling people, these are your rights. Under the law, you can do this, right? Now we found the movie really worked in terms of raising awareness. But it didn't work in terms of changing the reality. People became much more aware of their rights. We did this, we took the movie to a randomly selected 40 villages in our survey and we used 100 villages as the controls and we looked at the effects of the movie on people's knowledge. We gave them a quiz on the terms of the scheme and we showed significant impacts on their knowledge. So great, but it didn't change the reality. They were still being rationed. We found a bit of an increase in the wage rate. They were a bit less exploited. The wage rate had typically got about a 10, 20% cut off the top from a local official, a little pricey pays for his help. You reduced that, so you get a little bit better about extra five rupees a day if you're already on the scheme, but not much else. Because of the second problem, the supply responses in poor places. This is coming down to two, essentially two problems. One is local costs, local cost of delivery. All kinds of costs here that are not always evident, but if you go and look and you see them, local administrative possibilities in poor areas, skilled labor is really scarce. Part of being a poor area is that skilled labor is scarce. And that actually means to administer a program like this I found in partner in the sector of rural development, I found there may have been others, but I could find only one person who was capable of doing the Excel spreadsheet to report on the program back to Delhi. Administrative capacity is weak. That comes with being a poor area. And that means that there are huge costs to implement a scheme like this, that one person was very busy, as you can imagine. But there's another set of problems that people try to reduce corruption on this program in a situation, in a naive way. They try to do it by increasing the marginal cost of being corrupt. How do you do that? You increase the penalties in this part of India. We're here often here, expression millionaire Mukia. A millionaire Mukia is a local village leader who's a millionaire in Rupees, but out of the scheme. Out of the money he's making from the scheme, okay? But millionaire Mukia is, if you're not careful, you could end up in jail. The government is a surveillance mechanism. There are fines, there is a probability of being caught. All of that is increasing the marginal cost of being corrupt. But you're facing a rising marginal cost of being corrupt. Why? It's a complicated scheme. The more you employ people, the more projects you open up, the higher the cost to you. You've got to bring more people into your network. You've got to share the bribes across a larger set of people. You may have to bring in the district engineer, the sub-punch, a bunch of people. And you get outside your comfort zone. As a local leader, you're comfortable with a certain number of people, you trust them. But as you expand the program, more and more people have got to be brought into it. The way we say that is you've got a rising marginal cost of corruption in the program. If you've got a rising marginal cost of corruption in the program, then everything the government is doing to reduce corruption is creating the rationing and the poor performance in poor areas. You see that very simply. Here's the marginal cost of corruption. What the government is doing is shifting up that cost. This is the level of employment on the scheme. This is cost or benefit. We think so, decreasing marginal benefit or constant marginal benefit. So what does the government doing, tightening up in the scheme, increasing the penalties, is going to reduce the amount of employment provided by the scheme. It's going to contract the scheme. It's going to increase the amount of rationing. The scheme is going to, and that's why. Bihar is way up in the top of that diagram before. Precisely because the government is fighting corruption on the scheme because it's so worried about corruption in the scheme in Bihar, in the millionaires. The way you fight corruption on the scheme is make it impossible for poor people to be exploited this way. You've got to change fundamental things in the society, in these villages that, and if you don't change those things, this is not going to work. So my conclusion on Enregar, at least in the parts of India where Enregar is needed the most, poor states of India, Bihar in particular, but elsewhere, my conclusion is that it's well under the frontier. Going back to my trade-off argument, promotion and protection, Enregar's sort of well underperforming from both points of view. With others I've written a book, not this book, earlier book last year, with very specific recommendations, a whole lot of very detailed things that have to be done to fix this program. And I argue if you don't fix, do those things, dump it. Turn it all into a basic income guarantee or just use the BPL cards and give out the cash. Last, China's Debail Program. Very quickly here, I just have to mention Debail because it loads the largest cash transfer program in the world. We know very little about it. There's only me and about five people in the world who seem to be researching Debail. Partly problems of access to data, problems of working in China, it's not the easiest place to work. From some points of view. But it's a great program, a very interesting program. Why it's really interesting to an economist is it appears to be a very bad idea. What Debail is doing is literally perfect targeting. That's what it claims. The legislation is trying to do this. Everybody gets, if you're, this is the Debail line. Remember my drawing before? Now we have the Debail poverty line and it's a different line in each city. So they learned from, I have no evidence that they'd read anything about the debate in the 1830s in England. But they certainly seem to have learned from the old Paul laws in England that the parishes deal with their poverty problem themselves. Here it's let each city set its Debail line and they had to bring everybody up to that line. They had to fill all the poverty gaps. So if the Debail line is such and such in your city, which is determined by your city, the Garcetti government had to give everybody with income below that line exactly enough to fill the poverty gap. So when I first heard about this, it first appeared in the late 1990s, I thought, my goodness, they're gonna create a poverty trap, a public based poverty trap. Research continued and the story unfolds a bit like the story about the National Rural Employment Guarantee Scheme in India. The reality is very different. First thing we find, areas of inclusion are modest. This is the best, to my knowledge, the best targeted in the narrow sense of reducing areas of inclusion, making sure that non-poor people don't get the benefits. This is the best targeted program in the developing world. There's a compilation by Cody, Grosh and Hodnot of various social programs and they measure targeting performance in terms of reducing inclusion areas. So the share of benefits going to poor people across all these programs. And Bihar, the Debar program is just way better than any of them. So excellent targeting performance. Really administrative capability is China, administrative capability is excellent, but coverage is a huge problem. First observation. Now, okay, what do we do with that? I was asked to advise the Ministry of Civil Affairs in China and what do we do? My advice is, well, if it's a poverty trap, don't expand the program. Coverage is a big problem. Obviously, we would think to expand the program, to expand coverage, coverage of poor people is not reaching enough of the target population. But if it's a poverty trap, the last thing you wanna do is expand it. I mean, do something else, but don't expand the thing, you're just creating an even bigger poverty trap with 100% marginal tax rates on poor people, higher than the tax rates on any rich people. But is it a poverty trap? When we turn to the evidence, it all looks very different. The benefit withdrawal rate on paper is 100%. The reality, much, much lower. This is quite a difficult thing to work out and this is the source, this is where I document the evidence in a paper last year, but the benefit withdrawal rates on this scheme are not 100%, they're actually more like five to 20%. What is happening? Well, it's remarkable. Local, and we saw this in the field, local implementers of the scheme are recognizing the incentive problem and they're smoothing the benefits at a daily basis. We saw this in qualitative work. If a poor household would got a job, they would not cut their debaubed benefits, they'd say, the local official would tell me this, oh no, we shouldn't do that. They didn't talk about incentives, they didn't talk about marginal tax rates, but essentially they recognized the incentive problem and they said, well, let's smooth the benefits, we'll keep them in the program for another three months or so. This was local implementers who understood incentives and were essentially implementing a different program to the central government's program. Another way of thinking about this is we have participant capture in the program. Local officials, they've got a client base, the people in the program, and they stay in the program. It's not necessarily a bad thing, but it is participant capture. It lacks the flexibility. People getting onto the program when they need it, people leaving the program when they should. It's the other extreme. 100% marginal tax rate is too high, but the marginal tax rate on this program, the benefit with draw rate, same thing, marginal tax rate, benefit with draw rate, the same thing, but the benefit with draw rate is too low. The program is not providing, the program is biased towards promotion, away from protection. If anything, the center is the opposite. Central government is biased towards protection, away from promotion. Now you can think of why that might be the case, central governments are going to bear significant political costs of crises, and this is China, which suffered the worst famines in human history because of policy failures, and the great leap forward. They know the history too. Crisis are bad news. Instability, keeping the commerce party in power is crucial, obviously. So they are probably biased towards protection. Local administrators have a bias towards promotion, probably realizing that the center will come in and help. Moral hazard is maybe part of the story. So if anything, we're going to go back to our characterization of the promotion protection trade-off. Debail program, I'd argue, we have a local, the actual program, what's actually implemented by local officials, and we have the center's legislation, what it's supposed to do, and this is the, and I'd argue, the optimum sum between the two. And I'm not just guessing at that, there's actually research by good economists, there was a very important paper by Ravik Kanpur and others on working out what the optimal tax rate should be on a program like this, and it's somewhere around 50 to 70%. It's certainly not 100%, but it's certainly larger, higher than the current program. Conclusions then, two lessons from all of this. These are lessons for social policy reform and the ones that I carry forward and from this, preparing this lecture. First, focus on protection promotion, not finer targeting. Let's just put this in its place. Targeting can help, sure. Sometimes, maybe, but keep your eye on the ball. What's the objective here? It's not finer targeting. It's reducing poverty, and they are not the same thing. There are all kinds of costs of, in reality, that come into the story, and there are information problems, asymmetric information, all kinds of measurement areas in our data, or when we say that something's well targeted, we're using data that's not designed for that purpose and measurement areas are huge. A lot of problems there. As I've emphasized, a trade-off can be expected. Look for it. You design the program badly. You will have a trade-off. Your face could face as a very severe trade-off, but it's often exaggerated by critics. Second law is lessen. Strive to improve that trade-off. Strive to achieve more protection for given promotion. I think here the social investment approach is encouraging, but again, it's not a magic bullet. We've got to think about that in a comprehensive way. Challenges here involve achieving greater flexibility. That is a recurrent problem. We're not seeing a lot of progress there. And also, we're seeing a lot of programs which really are too complex for the setting we're talking about. Part of being a poor place is, as I've emphasized, administrative capabilities are weak. When you try to do a complex program, you'll have perverse effects. That rising marginal cost of corruption, the fact that fighting corruption is actually making the program work less well and in a narrow sense of assuring that people get work when they need it, that's an important lesson. It's a lesson about tailoring social policies to administrative capabilities in the context. You've got to have another development agenda which is to expand administrative capabilities, making governments more capable in poor places. But in the short run, at least don't try to imagine. They can do all kinds of stuff that they can't do. And finally, monitoring and evaluating. And monitoring and evaluating is what gave me all of the research that I talked about. Well, that doesn't include the old Paul Laws debate, but in terms of the research that's been done on social policies, it's been hugely important and influential going forward. You can read more about this in the book. I'm not advertising the book, but it's a nice book. I think so. Thank you for your attention. Thanks a lot, Martin, for this tour de force. We decided that we would try to start the Q&A session with two brief introductions or comments. One from Anika Sundein, who's chief economist of CEDA and the second from Toni Addison, who is chief economist and deputy director of WIDA. So, Anika, over to you. Thank you very much, Finn. And thank you very much for the invitation to come here today. And thank you very much for an interesting and very rich lecture, Martin. I think it's clear from what we heard this afternoon that social protection is absolutely crucial for combating poverty and for redistribution. And in particular, in the changing of poverty. Poverty today is much more about inequality, the distribution of resources rather than the absolute lack of resources. We also see a core of chronically poor people, also for whom very little has happened, also in countries with rapid economic growth. So it's central in our efforts of leaving no one behind and as recognized by the SDGs. But I think it's also important to underscore how these programs, if they are well-designed, are part of the social contract in a society. And in that way, are integral to nation-building democracy and social cohesion. If we look at developing countries today, they're expanding their social protections programs rapidly. Sub-Saharan Africa is just one example. And many donors like CEDA are involved in these efforts. And Sweden supports programs in several countries in Sub-Saharan Africa. If this is gonna be successful, if these programs are gonna be successful long-term, however, country ownership is absolutely crucial. It's the country itself that has to design and implement their programs. And there I think Professor Revalio makes a very important point and that's the one about administrative capacity. We see that many of these programs are complicated and we have to think about that when they are implemented. What type of programs can and should be implemented? And I think we as donors have reason to reflect here. If we then look at what kind of programs are implemented in particular in Sub-Saharan Africa, it is the conditional cash transfers that we've just heard about. And they are a model on the programs that were introduced widely in Latin America in the 1990s and the 2000s. These programs, and as research have shown, have been successful in reducing poverty, at least in the short run. But going back to the issue of country ownership, I think it's very important to be careful when you take a model that's been successful in a certain context and then transplant it in another context. You really have to think about the country where you are when you implement a social protection program. When you design these programs and think about the various components, I also think it's important to recognize that these are not just singular programs, but they really are systems. And that goes to the fact that they have to meet several objectives. It's about poverty alleviation, about protection. It's about protecting against risks that we all face during life, unemployment, sickness, disability, work injury, old age, and they have to help build capacity for increased productivity, health, and education. So the programs that we see expanding widely now, the cash transfer programs, I would argue, is just one piece of the puzzle. They build on the idea that they are targeted and they should be temporary, albeit for a fairly long period of time, and it's only the poorest that should receive the benefits. And as Professor Valiant clearly pointed out, there are a lot of problems with targeting. It's difficult, it's expensive, it's complicated, and it often misses the mark. I would add another argument to why we shouldn't focus so much on target and what are the characteristics of the people that these programs are aimed for? There are some groups that will never be able to support themselves through work, children, the elderly, the disabled. These are also the people who are at risk for poverty and are the target of the cash transfer programs. So instead of thinking about designing the programs more or targeting them on income, I would say that given that these are the groups that are at risk for poverty, there's an argument for permanent social protection programs such as social pensions and child allowances. We heard that today's rich countries appear to have done much better than today's poor countries to reach the poor when today's rich countries were poor. My hypothesis is that this is because universal programs were part of policies when we went from being poor to rich. Sweden is a clear example of that. In the beginning of the 20th century, Sweden was one of the poorest countries in Europe. The first social protection program that was introduced here was a social pension in 1913. That was then followed by child allowances 30 years later and then successively the programs were expanded. Programs that are targeted at the poor are also easy to cut. Universal institutions that encompass a majority of the population tend to reduce levels of poverty inequality much better because they don't have the exclusion problems and they have support by the middle class. But of course, if we look and particularly if we look in Sub-Saharan Africa today, we know that tax bases are small and fiscal constraints prohibit countries from introducing universal schemes. For example, looking at the young population in Africa today, it would be difficult to introduce universal child allowances. Social pensions could be easier. So targeting may be a way to start but if you are gonna target on income or wealth, you have to make the construction simple and transparent. Not just because of administrative capability but also because of it's much easier to understand. Simple targeting is a way for beneficiaries to understand why they receive the benefits and how much they receive. And it also makes, paves the way for having a transparent and functioning grievance procedure so you can complain why didn't I get the benefits when I'm eligible. The second point that I wanna talk about is of course the conditions in the conditional cash transfer programs that they impose on households. If you are to receive the transfer you need to send your kids to school, you need to go to health checkups. And as we heard in Latin America, this has had success in health and education outcomes at least in the short run. But here I would argue that policy makers try to solve problems that may originate somewhere else by imposing conditions on the poor. If you ask people living in poverty what they want, what do you want for your children? The answer you will get is education. They want education for the children to be able to make a better lives for themselves. But in the countries and the countries where CEDA work, the obstacles are many. There are high fees to go to school, the schools are of poor quality, there are few schools and they are far away. So the question if these supply side problems will be fixed by imposing conditions on the poor, one argument is that if you do this, the poor could have a voice and it will put pressures on governments to expand health and education. But I think this is actually questionable given that people living in poverty almost always lack power and voice. So I think instead social services need to be part of the social protection system and ensure equal access and that's much more efficient than imposing these conditions. My final point is about the long term perspective that social protection really is long term and that it takes time to build. I said that Sweden started with social pensions in the beginning of the 20th century. It took 60 years until our social protection program was fully expanded to what it is today. I also think it's crucial to think about the design of the programs and financing together. They go hand in hand. Tax revenues are necessary for these systems to be sustainable long term and as countries expand social protection programs, they also need to expand their domestic resource mobilization. They need to think about how they're gonna finance these programs. The link between social protection, the link between the design of the program and the financing is also crucial because if you do that right, you can build a willingness to pay taxes and get something that's also sustainable. And of course the link between the labor market and these programs are crucial to create incentives to work. Thank you. Well this was a tremendous tour de force by Martin. The first example that Martin gave us, the first empirical example from England had personally tremendous resonance for me because I come from the coal mining area of the north of England and my ancestors, the family history, the family legends are all about actually moving in and out of poverty depending on the cycle in terms of coal production and avoiding the workhouse in Victorian England. And certainly there's a family legend of a young woman having to go into the workhouse in the early part of the 19th century and actually being separated from the rest of a family because this is what the workhouses did to deter you and make sure that you didn't go into the workhouse unless you were in a very desperate condition. There's also actually a family legend about another member of the family who tried to solve his poverty problem by stealing a sheep and stealing a sheep in early 19th century England was a hanging offense and he was convicted and he was hung. If he had stolen a pocket handkerchief or something smaller, he'd have been transported off to Australia, Martin's homeland. So in some ways, this sort of set of examples, you know, illustrates the kind of political dynamics around poverty, the seeking of control over poor people but also as Martin said in the case of India, the expression of poor people's rights, their right to demand a better future for themselves and for their children. But as Martin showed, it's a very complex set of factors that you have to get together to produce an effective social protection program. In some ways, we live at the most exciting time ever in terms of poverty policy. As Martin said, we've never seen so many people move out of poverty. We still have deep poverty across the world but we have lots of exciting new technologies, some of them coming from new technology itself to replace the old mechanisms and old ways of doing things. So for example, the make work programs, the public works programs that India has had for many years, Martin mentioned now the move over to biometric identifiers, the precursor now to moving to more effective and cost effective forms of financial transfer. One issue of course with public works programs is the quality of the infrastructure that's constructed which is sometimes not a very good quality. A very good comparison might be for example with Ethiopia which Martin didn't have time to mention where there's a very large public works program reaching now I think nearly 10 million people which has done a reasonable job in terms of creating infrastructure potentially better than that of India. But we've now got better ways and means to think about these issues. But it's still a difficult mix because there are of course all kinds of macroeconomic dimensions. The very macroeconomic shock that might make social protection essential to keep people out of poverty could be the same macroeconomic shock that reduces fiscal space, that reduces public revenues available for that social protection. Clearly as we move towards basic pensions, more universal systems of social protection it has to be within a macroeconomic framework that preserves those over time and sustains them. So these are really very, very big issues. Now Martin is very well known as an innovative quantitative economist particularly in the use of household data and other data. And something that's very much come out for me in this lecture but also in Martin's book is the importance of data. And it might be nice during the discussion to have a little bit more sense about what kinds of data we do need in the world. Over 30 years we've seen much more data created particularly household level data, better measures of poverty. But I think none of us feel as researchers that we really have the data that we sufficiently need sometimes to inform policy. I have one, and this is the point at which I end, I have one question for Martin. And the question is, many young researchers are thinking about careers in economics. And if they want to go into economics, obviously one of the great areas they could go into is development economics and they could go into the economics of poverty. So my question to Martin is if you were recommending to a young researcher now what should they go out? What should they commit their future to doing? Perhaps in the form of a PhD or some other piece of work. What would your recommendation be to that young person, that young man or woman going out from Stockholm to work in India or the Congo in terms of research impact? Pieces of research that would really change policy for the better and for poor people. Thank you. If I could invite Martin, Annika, to join Tony and maybe I could just mention that I am deeply interested in that data point. Some of you may know that at wider we hold sort of two conferences a year and we had worked quite a lot on developing a concept paper for a data for development conference to be held in September this year. And just two weeks before we were going to announce it the World Bank came up and said we're gonna do the ABCDE conference this year in June on data for development. So there seems to be at least a shared concern about this. Obviously we were then changing so we are now going to respond to crisis in September instead but the data point is a very valid one. Martin, do you have any sort of immediate reaction to Annika and Tony? Well I think it might be efficient to collect questions. I mean, I'm hoping there's lots of questions. I'm gonna start over here on the left hand side if you could sort of indicate if you wanna intervene, have a question and then sort of move over. Please do introduce yourself if you have any questions. Okay, yes please. Thank you. My name is Gunnel Axelsson-Nykan and I'm from Church of Sweden International Department working on development issues. I would echo very much of what Annika said about especially about targeting in the more universal approach and really saying as Scandinavian and Sweden listening to your arguments about targeting and the downsides of that is quite, I mean, we all have it from growing up and the political debate and all about marginal effects et cetera and we know all from our personal life the benefits of inclusion systems. However, we can see that it is the more and you also reflect that it's the more targeted system that are now not only in Latin America but also in Africa and Asia often promoted. And we can see that at least up until now I think it has been mostly diffed and World Bank that has been promoting these kinds of programs where many of the experts actually have the Anglo-Saxon welfare model in their DNA to say. Whereas we can see that the experiences from the universal models as we have in Scandinavia hasn't been that influential in giving a device to developing countries. Of course, this is easy for us to say, yes, it's good for targeting. It's not as comfortable maybe to say the same thing in the United States because it inevitably goes back to your own society and your own system. So my question would be, would you think it's possible for institutions like Difford and the World Bank to actually promote more universal models? Okay, yeah, please, yeah. My name is Mia Honabranzen and I'm the CEO of the Center for Business and Policy Studies. You said in the beginning of your talk the need for adaptation of policy in relation to evidence that comes from research and the type of evaluations or programs that you have described to us. I would like to ask the problem of adaptation of policies, what is the reason in your mind behind that is it the fact that people are unaware of the research and the evidence and lack of expertise in actually then adapting policies or is it more maybe the type of questions or values and ideas about models and that is behind that you don't believe in the evidence, so to speak or kind of have a higher type of goal with your choice of model. Okay, I think here in the front with Tobian and then over here afterwards. First, Martin, thank you for your excellent presentation and overview, I think this is a perfect place for this. This is a school and this was a fantastic lecture so it's also an inspiration to all of the teachers and students here. I have two questions, two different issues. You mentioned political economy a couple of times in your talk but I was wondering how you think of these different programs in different political circumstances. I could see big differences in terms of how much of populism you may fall for as a politician under certain political regimes and how that may be different in others. So under what sort of political circumstances would you promote certain policies or avoid them? And I'm thinking now in terms of also interest of SIDA and other donors, where should we push for these kind of programs under what political sort of umbrella should that be done? The other one that I didn't really think you touched so much on and I think it's also interesting and I think Annika was hinting at this is the trade-off you mentioned here was sort of within the realm of having transfers but really another trade-off is different items in the government's budget and really is this the most cost efficient way over time to reduce poverty or would sort of education or healthcare or something else be more relevant? And one of the issues here I think it's extremely hard also when we now say we should evaluate things and see what works. Evaluating school reforms and poverty takes much longer than maybe evaluating some of these programs per se. So how do we actually figure out what is the most cost effective way of reducing poverty at the moment? Over here and I think we will pass over to Marty. Also, my name is John Joyce. I'm the Chief Economist at the Stockholm International Water Institute. Thank you Martin for a very innovative lecture. I want to pick up on Tony's point about data and picking up on what Martin said during his lecture on data not designed for purpose, measurement errors, et cetera. And as an economist we always like to use the quantitative side of things but I'm just wondering how we've used qualitative data from an ethnographic perspective and how much economists now need to start working more seriously on qualitative data in order to reduce specification bias and mis-specification, et cetera. Thank you. Okay, thanks. You want to take a turn, Martin, a time response to this? Sure, these are all good comments and I can't do justice to all of them but Annika's comment really, I like this way you talked about it and I'm not sure, is that really how it happened? I mean, I know this a little bit about the Swedish history but if it was the case that the rich countries started with universal schemes and progressed to more targeted ones, it would be very nice for my argument. So I like it but is that really, has that got any external validity beyond Sweden? I'm not sure but I've got to look into that because it nicely emphasizes my point that currently all this emphasis on fine targeting in poor countries is running against that historical process and the historical process makes a lot of sense starting universal and you develop capabilities, you get more finely targeted and you don't overdo it because you've still got to be politically sustainable. That would be a very nice story but so I'm going to look into that. I like the way you described it, it fits very well with how I think about it I guess but I have to look into it. Data, who was the point of, I didn't write down then, who was talking about, yeah, you're talking qualitative but somebody else on data, I just can't, oh yeah, right, okay, it was Tony. We're two different, very different questions and it doesn't separate. So Tony, and your two questions. We all know how important data is. I think there's two kinds of ways that data matter here. One is for doing the kinds of research and evaluation of what works and what doesn't and another is in how data is used in policy and I think I have to be careful about those differences. I see an enormously important role for high quality surveys, more from time to specific programs, building the capability for survey data collection in poor places and I think that's happened in a big way. I mean, you often hear this characterization that socioeconomic data is poorer in poorer places and probably that's true but it is a bit exaggerated. I mean, you get some really good data in poor countries but data for informing research and doing evaluation. When we start to take that data to social policy then we get into a lot of trouble because then we've got to extrapolate to much larger samples. Then I start to worry. I mean, this proxy means test that's so popular I'm doing a study now for Africa where we're across every country that we have data in Africa we're just systematically going through to figure out under the ideal conditions how well does the proxy means test really work? Now, in the audience, a proxy means test is exactly what it sounds like. You don't observe income, that's the data problem so in the survey, do a good measurement of income but when you go to scaling up to the policy you've got to impute incomes under less than ideal circumstances for the purpose of implementing the policy. They're very different things. A proxy means test is an attempt to use a relatively small number of variables to predict somebody's income based on the correlates you found in the survey but you take that model and you scale it up to the country as a whole and you implement a policy. Even under excellent conditions on the survey what you get in the end is really, sorry, it's very poor indeed. The inclusion, exclusion errors, it's horrible I mean in many countries what we're finding it's just a little bit better than tossing a coin. If you really want to reach for the undernourished women and children as part of your anti-poverty policy this is not going to work and part of dissatisfaction with social policy finally target social policy in poor countries from the community perspective is they see it. They just see how badly it works, right? Which is part of the motivation for this study to really do a careful job of seeing how well it works and I tell you when I look at it it doesn't look good and we're basing our social policy in poor places on this technology. So very important distinction the data you use for the surveys and the evaluation versus how you scale up to the social policy based on correlates you find in that data. You can have a great survey but the correlates are still so weak the information base for fine targeted programs is very poor and you're not going to reach the groups that you want to reach as effective as you might like. It's very subjective though, right? Another little observation two people can look at the same thing I see this repeatedly and come to different conclusions toward late in this study that we're doing in Africa we haven't written anything yet but we've been massive number crunching. Late in the study I started looking at what people who defend proxy means testing which once included me by the way I mean until I started really looking at it seriously I look at what the arguments they make and they occasionally do something quite similar to what I'm doing and they look at it and say oh look how good it is and I look at it and say look how bad it is and it's the same thing is the glass half full or half empty there is its perception but obviously people are defending it this technology is taking over and you're seeing it all over the world now in developing countries this proxy means test technology and I really worry if that's going to be able to deliver again back to the point that you may be much better off with the universal program and somebody raised the really important issue of funding these programs I think it was Annika remember what I was doing with that Nrega I'm taking the budget of Nrega they've already got a budget I'm taking that and I'm allocating it uniformly they're not spending more money than they've already decided to spend I don't think they're spending enough that's another point and that goes back to the very important point that was raised by I think Professor Baker your second comment I think items on the budget, the trade-offs between things this is really neglected and it's silos they've got social protection silos we've got human development silos, education they've got worse and worse in the sense that they're not communicating with each other and very few people in development now are asking those questions about the trade-offs should we be building bridges or should we be building schools and we're not informing we're not telling countries about they want to know the answer to your question as well but we can't tell them because our research is in these silos and this is a really big problem I'm not seeing enough of that kind of broad cross sectoral work and that's one of the many recommendations I'd give back to Tony's second question many recommendations I'd give to PhD students those bigger trade-offs between the line ministries if you like and it's part of the politics of it when we deal with countries we deal specifically with the ministry we're tied to that ministry we used to have much more capability I think for the cross ministerial debates and discussions than we have today sadly I think the old planning commissions the legacy of the sort of control economies that were common in developing countries well the control economy was pretty much a mistake we know that market driven economies are much better producing growth and all of that fine I think that's pretty well established but the planning ministries all got axed with very few exceptions some stayed but they were doing different things we didn't we threw the baby out of the bathwater the baby was the capability for making those broad cross sectoral prioritizations in public finance I think we've got diminished capability in that area that's not helped other questions how to get more universal programs this is the person from church of Sweden very first question you know I'm a little bit uncomfortable with this Anglo-Saxon versus Scandinavian model characterization but okay that's not a big issue I can live with that Anglo-Saxon side but I'm thinking I don't think that just Scandinavians or Nordics or whatever who are talking about this this goes way way back certainly lots of discussion of this in the 18th century there was anyways a lot we could say there but that's more history of thought stuff on your the more much more important question how do we get World Bank and DFID well I think DFID is pretty much following the bank and how do we get them to talk about the scope for universality by me and you I mean discussions like this you know I've tried for 24 years to get the bank and the fund to think differently about targeting and I will acknowledge it's one of my greatest failures it's so embedded target the thing finally you do any you know spend money on poor people but target it really well in situations where you don't have the capability to do that credibly it's really you're just in the end it's poor people who are not getting the benefits and the help they should be getting because in the end you have to judge it by impacts on poverty not by targeting it's a mantra that has been very hard people will you know smart people will obviously recognize my arguments and agree but it's become and in fact again the silo of social protection the industry of social protection has not helped either so I don't have an answer to that I think we just got to keep that that's that thing I keep saying that targeting is not the objective it's poverty reduction now we can debate about what that means but I think we've got a pretty clear idea and we can talk about specifics but keep your eye on the objective in terms of people's welfare and how you get there and the more we can emphasize that the better what have I left out of me is comments on the I think a lot of it is the politics of capture and bureaucratic inertia I think it's the main reason we don't see more adaptation to inform evidence that in programs become embedded and they have a lobby group in the government and they have a constituency and they're fighting to keep their program all the time and we see that repeatedly and that's not just a problem of developing countries everywhere we see that and that kind of entrenched support and a bureaucratic inertia and protection of your favorite program is the main reason I think and we just have to call that out and a lot of this as a researcher I always think what I'm doing and other researchers are doing is really doing just that calling it out informing a public debate with evidence to try to move it on a bit and that's really the main point of my lecture on the John's point about the qualitative data I've never ever thought you could only study poverty and inform anti-poverty policy with qualitative data I find it very limiting but equally well I've never thought well that's an exaggeration I probably once thought you could do it with only quantitative data but I've now come to the view you need both you know but you don't need it in the way it's often used qualitative work after quantitative work not before or in parallel with so I'd really like to see more economists doing this this basically means do your quantitative work every study I've done almost everyone I've had puzzles what we do in economics and empirical research is very black box like we have these black boxes we don't see inside them with our tools the worst case of course the randomized control trial is it's a complete black box about how you got the outcomes that you got but everything it's not just a problem with RCTs a lot of what we do in economics it's not very revealing about the processes empirical economics it's not very revealing about the processes that generate impacts and I find actually talking to people I think I'm much better at work and I always try to get people to have the reverse comparative advantages to work with me but I think you can learn a lot about processes of how things actually happen by talking to people and using subjective data in intelligent ways as well recognize it's subjective but you can still learn from it I think have I left anybody out I think I've covered most of the questions that was pretty all array of the questions yeah please here in the center and then here my name is Odi I'm a PhD student from South Africa so China's been quite successful in reducing poverty and I think if you look at the numbers the global numbers a big chunk of that is coming from China can we attribute all of that to growth and if not how much can we attribute to growth and where do we attribute the rest okay here in the front my name is Rudra Kupamshana I'm a former diplomat and a former banker originally a scientist my question I want to introduce another kind of silo that is integrating many countries in the emerging markets and I'm thinking of traditional social structures like clans and so on the Indian case system you could call an inflation rigidity in the Indian labor market and my question is since these social structures frequently seem to be as poverty traps how do you eliminate poverty in countries where the strength of the social structures exceed that of the state possibly looking at the fiscal revenue relative to GDP thank you okay other questions okay here over the back ah sorry let's go and then we take the other one thank you for a very interesting lecture my name is Christina I'm from the trade unions I have two questions the first question is in the beginning of your lecture you mentioned the fact that rich countries of today were better at targeting the poor when they were poor and it would be interesting if you could elaborate on that and explain why was that the second question is the factor of a weakening bargaining power of the trade unions and the lower level of unionization in the world right now and how that would affect the possibilities of protection and promotion and in some countries like in Sweden it is actually the two-part model that sets many of the protection and promotion policies so how would you see that thank you thanks I did not be in the back thank you professor my name is Damir I'm an economist working at Stockholm Peace Research Institute the question goes partly to your case studies you mentioned in China case that knowledge through the movie you provided didn't really transfer to action so the question is what helps to move people from being knowledgeable being informed changed attitudes in moving to action change in behavior okay thank you I think we will pass over to you Martin and then Tony and Anika if you have any okay I'll try to go and order this time rather than chopping around the South African PhD student okay so there's two levels of answer to your question first is a kind of almost a mathematical level you can think of a change in poverty as determined by growth in part and changes in distribution okay in a narrow sense all of China's poverty reduction has been due to growth rather than changes in distribution because distribution has got worse relative inequality in China has been on a trend increase over 30 40 years not sort of all the time increasing not monotonic as we say but other periods falling but the trend upward has been dramatic in fact if you just extrapolate the trend in China's inequality back over the last 30 years it's going to be the same inequality of Brazil within 10 years but I don't think that'll happen and I actually think it's topped out I don't think there will be limits because there's changes in the Chinese economy particularly coming from the absorption of rural labor China started from a situation of massive it's a classic Lewis model we had massive labor surpluses wage rates were no upward pressure on wage rates for a long period that upward pressure is now emerging wage rates are rising at the low end in China the capital is actually starting to ex-leave China looking for lower wages elsewhere and that's a very good process great good news because real wages are poor people who were poor are rising and that's great going forward I think that'll put a there are other things that involve too but that's one of the big economic changes it's going to put a break on the rising inequality but there are other forces that are working the opposite direction so we'll see what happens that's the easy answer now they go up to go deeper it gets more complicated I'd point to a number of things that were really crucial then you've got to ask what's the term of the growth what's the term of the inequality what's under the surface there are a number of things I actually think the sequencing of economic reform China's one of the few developing countries that maybe got that sequence roughly right it started with agriculture and rural development moved on to a basic industry it moved up a kind of sensible ladder in its policies this is a highly interventionist economy amongst developing economies with a lot of administrative capability they can do a lot they can also make big mistakes as they often have some of the greatest mistakes in human history like the Great Leap Forward Revolution the sequencing of reform from Deng Xiaoping's initial impetus for reform in 1978 focused initially on agriculture and rural development and stayed with that until mid to late 80s and then started to move on to industry and that sequencing was roughly right I actually think they moved on too quickly I think they could have had an even more dramatic poverty reduction the absolute poverty reduction is the fastest and greatest largest in terms of the sheer numbers we've ever seen could have done better if it had stayed with a lot of the agriculture and rural gender and the grand reform agenda was left undone by the time the industrialization pushed the TV's the export led manufacturing growth and by the time that started to take over in the trajectory for poverty reduction the bulk of the work was still done by rural poverty reduction in China another aspect that they got right I think which is a very important lesson the first is important lesson for sub-Saharan Africa because very few countries in sub-Saharan Africa have got that sequencing right they've tried to they've gone straight for industry they've neglected agriculture they've been to their peril and that has not helped the poor so second thing is evidence based policymaking we don't think actually politics was driving the first as well but it's the second especially the politics here was that you had a potential food crisis in China in the late 1970s that they really had to address a food security problem or massive but the second very important thing was in the the way they learnt the way they tailored their sequencing reform and the specific reforms to evidence now Deng Xiaoping again he said seeking truth from facts that was a political reaction to the culture revolution to ideological based policymaking and it was so important they didn't have the tools to do good evidence based policymaking yet but they had the culture of evidence based policymaking as a reaction to now essentially and that was really important so you saw all the time they were looking experimenting the program I talked about today that emerged out of an experiment in Shanghai and all these experiments going on all over China they still are and they're looking at them and learning and that seems to work let's do that when I look at the quality of their evidence you certainly don't see randomised control trials but maybe now the quality is not always there but the culture is there and that's really important and that's another lesson for sub-Saharan Africa so there are deeper answers to your question there are other things I could say let's start the question about social structures this is really hard and the caste system in India particularly has been just so hard to change and I don't have an answer I see it all the time now all the time we're worried about social exclusion everywhere from Bihar through to Brussels and we're seeing it all the time and this is partly why I emphasise the importance of thinking about relative poverty as well as absolute poverty you've got to look at both and that's really about the concept that sociologists have and going back to people like Peter Townsend and Britain social exclusion is hugely important as a concept in my view there are debates about that some economists are not so fond of the idea but I like it and I can't it gives me a way of understanding a lot of problems and the lack of inclusion associated with the rigidities in India's labour market associated with the caste system is just horrible it's horrible for the information transmission for learning about public programmes not as well as for the labour market I don't have any simple answers you know in India's been officially fighting the caste system since India has been a country since independence in 1947 the struggle has been going on but it's still there it's alive and well what do you do? I don't know there's a nice topic to add to the really tough questions that Tony would like to give PhD students and it needs a lot of focusing though before you can turn it into a researchable question but certainly we have lots of research on the implications of caste and clan social structures the costs of those, the economic costs of those and the cost for effective policy making as well but in terms of research on how to change it I don't know Christina's question rich countries I think the answer, rich countries doing better targeting when they were poor I think the answer to the previous discussion about universality but I put that out as a hypothesis that where was Christina? oh yeah sorry I put that out as more a hypothesis that rich countries were more successful in social policy early on because they better tailored what they did to the capabilities they had administratively and the constraints they faced politically but that's my contention but I can't prove it but I think that was important I think the decline of unisation is a big concern I worry a lot about internal labour markets I worry about dual labour markets I worry about the problems of labour market rigidity I don't think there's a hell of a lot we can always do about that but I think making I can see the potential for gains to poor people from making labour markets work better more effectively but I don't think unions need to be against that so I think a lot of the if the union movement was more sympathetic with the arguments that people like me make about labour market flexibility and the scope for poverty reduction we could make more progress but labour movement has been largely a hugely important progressive force in the history of fighting against poverty but not always and we should recognise that too and the point about labour market flexibility is one example I think I can give you others too but by and large my answer is that yes it's a big concern and I think it will make it harder probably to achieve the kind of social consensus that we need particularly around I think we have a social consensus around poverty reduction I don't think we have anywhere now a social consensus around inequality reduction we just have continuing debate and I think the labour movement what's left of it in the United States could be hugely important here but achieving that social consensus will be about a lot more than that it will be about trying to unpack this concept of inequality it's what Barack Obama asked a bunch of people who invited to the White House last year people working on US inequality how do I talk about inequality in America without being accused of class warfare of instigating class warfare and that's a hugely important question how do we get that discussion about inequality to be much more focused on the costs of inequality much more focused on what aspects of inequality are really damaging a generalized discussion is just not going to get the social consensus but I think the labour movement could be a very important part of that Damien's question on the from this just a little clarification was about India, not China but that's not important and the lesson was that there are two things are broken people's knowledge about their rights poor people don't know what is due to them and the public sector is not responsive to their needs and both things come together to mean they don't get what they're entitled to and that's poor service delivery which we see all the time service delivery to poor people is generally poor and it's the combination of the lack of knowledge and the lack of a responsive supply side but the two are kind of working together the lack of knowledge is often part of the reason why the supply side is unresponsive local leaders don't share information because it reduces their power they have all kinds of rents that they get out of that process and you've got to cut it somewhere to unravel it and what this study the movie did is tell me the whole point of it was to see whether just changing knowledge would make better service delivery is it just about telling people their rights under the law and to say they know and the answer is no it isn't that the supply side also needs to change so the knot is not unraveled that way you're going to have to have both they're going to have to know what is due to them and the supply side has got to be more effective in delivering it okay last question Martin in your all the work that you've done and all the countries that you've visited what has most surprised you for good or bad what has most surprised me Annika any other additional last question no I was just thinking about the point about looking at it as systems and if you can elaborate because I think another reason why Sweden has been successful transforming from Port de Rich was that during that period we saw it as a system and having protection programs that also supported labour force participation and you know enabling both women and men in work by introducing child care and so on it took some time but really thinking about it as getting people onto the labour market so any thoughts on that Martin again an interesting thought but I don't have any comment more than what I've said what surprised me the most oh goodness what do you think oh going back something you never would have expected that theoretical economics would not have predicted that oh oh goodness are there no surprises in the world but I'm looking for the number one surprise you asked me not did anything surprise you you asked a much harder question what was the most surprising thing the time because I've seen a lot of time comparing China and India and my prior I started in India obviously and I've been working in India for 35 years but I came to China in early 1990s and and in India there was a lot of focus on China people are really very interested in China and you get a discussion all the time so and I was going to China as more sort of somebody I don't think I could ever be an honorary Indian citizen but I would love to think of myself like that a bit as somebody coming from an Indian perspective and China was a big surprise I was surprised at the administrative capability I was surprised at the but more subtly I was surprised at the to get things wrong too and I would ask myself if I was reborn in kind of Hindu style as reborn as a poor person in a village in India or a poor person is a village in China which would it be and after about the first few trips in China and working in the field in southwest China and Sashi in the lowest plateau region the answer is China and why I guess the number one surprise because I'd be freer here we've got a country that's not a democracy India is a democracy the largest democracy in the world and yet I'm saying that if I was reborn poor in an Indian village versus a Chinese village I'd rather be in a Chinese village because I'd be freer because I saw poor people in Chinese villages and they'd get up for their rights and when they were trampled on they would fight you couldn't do to them what I saw being done to poor people in Indian villages so that I guess number one surprise that when you think about freedom and democracy there's many levels of it sure China is not a democracy at the highest level of that term are more subtle and finer grained interpretations of what that means okay I said in my opening remarks that Martin is a global leader in the field of poverty analysis I certainly think we got that demonstrated today both in in terms of depth I'd like to say thank you very much Martin for a magnificent wider annual lecture I'd like also to say thank you very much for your collaboration and to CEDA for support and I'd like to thank for the engagement in the lecture you're all invited to a reception outside where the discussion can continue and you can engage on the questions that you did not ask but I would like to ask you to join me in a big applause for Martin thank you very much Martin