 Good morning. Thank you very much for the invitation to participate in the panel and the conference and also to Miguel who organized this panel. As you can see, I have a very grand title, Social Assistant in the 21st Century. As you know, this is about lower middle-income countries, not high-income countries, although high-income countries will get a mention as I move on my presentation. I wanted to say that normally when we discuss social protection issues, discussions tend to get normative very quickly. By normative I mean what ought to happen or what ought to be there in lower middle-income countries. But this presentation is really, I have taken out all the normative side of it and I'm going to focus really on the positive side, that is what is there in terms of social assistance. As Miguel indicated, one of the limitations we had for describing what is there in terms of social assistance is the absence of reliable, harmonized data sets on social assistance. But that is changing very quickly. And Miguel mentioned the SAPI database. I also have assisted database which is called Social Assistance in Lower Middle-Income Countries. They are sisters if you have a look at them, they are very close together. And of course there is the ASPAR database from the World Bank. So the database for discussing these issues has improved and hopefully will continue to improve. And so I'm going to based on that information to really have a focus on what there is as opposed to what there should be there, right? The presentation, I got three main aspects of it. The first is to show you the large expansion of social assistance provision in lower middle-income countries. I think that probably is not very surprising to you but with the global data that we have, we're able to provide much more detail on that as you will see. Now, I'm interested in these from the perspective of the institutions, welfare institutions that are emerging in lower middle-income countries. So I'm much more focused in terms of institutions as opposed to particular policies or programs. I'm not going to say very much about outcomes either. I'm just interested as a researcher looking ahead what kind of welfare institutions are going to be in lower middle-income countries. So that is my interest. And the interesting thing about this is that if we look at what there is now, it's likely that social assistance will play a prominent role. Whereas in high-income countries, the social insurance plays a prominent role and social assistance is residual. If we look at what we have now, it will appear that the welfare institutions are going to be quite different. So that's the second aspect of my presentation. And then, of course, I'm interested in what are the things that we should be looking into and perhaps we're not looking into. We have research gaps. And I got three that I want to discuss very briefly at the end of my presentation. First of all, why is it that there is this expansion of social assistance in lower middle-income countries? We know, for example, that programs work well. We know, to some extent, the context in which programs work better than others. But we don't know why there has been this expansion. So causal explanations of this expansion of social assistance are really at an early stage. The second thing I want to discuss is the relative weight of protection and social investment. This is discussed quite extensively in high-income countries. And perhaps we should pay more attention to it in lower middle-income countries. And the third issue, which is also in my view and the research, is the politics of vertical distribution, as opposed to horizontal distribution. We spend a lot of time discussing horizontal distribution, much less so vertical distribution. But if we want to understand the expansion of social assistance, we need to understand what is the politics behind these developments. So that is the structure of my presentation. I'm going to start with some definitions. I know that this might look very didactic to you. And perhaps you know this really well. But I really need to do this for the rest of my presentation. So apologize in advance if this is really very well known to you. I'm sorry to repeat this. I'm also a teacher so that I do this as a natural course of action. You know, the first thing you do, you start with this. So perhaps it's also a professional bias on my part. But it's also important because in the context of international development, we are very loose without definition. So I want to get this much more precise. Now, this is normally how traditionally, for example, the ILO has looked at social policy. You have on the one hand basic service provision, which is education, health, housing. Those are mainly services in kind, transfers in kind. And then you have social protection, which are transfers in cash. And there are three types, which I'm going to probably reduce to two. On the one hand you have social insurance, which consists of contributory programs that address life cycle risks as well as employment risk and employment-related risks. And you have social assistance, which is budget finance, is rules-based. I'm going to place a lot of stress on this, particularly in the context of identifying which programs are social assistance and which programs are not. This is really important, what is rules-based and what is not rules-based. And the address poverty and vulnerability. In the traditional ILO perspective, because there is also something called labor market policy, passive labor market policy consists of workers' rights. They are still there in most countries in terms of legislation. But the active labor market policy increasingly is feeding into either social insurance or social assistance. In countries that have those two kind of components of social protection, there are some, for example, training programs within social insurance. And there is increasingly a very high proportion of active labor market policies which are integrated with social assistance programs, particularly CCD. So perhaps the third one is going to drop out at some point. It's really important to keep this distinction in mind, conceptually, because social insurance is about horizontal-raised distribution. So you pay contributions to your pension fund today, and when you retire you draw the benefits from it. So it's horizontal-raised distribution across the life cycle. Whereas social assistance is vertical-raised distribution. It's from the better off to the worse off. So conceptually, it's really important to keep this distinction in mind. Now, because I suspect many of you working in international development, it's really important also to make a distinction between emergency assistance and social assistance. I have in the table a large number of distinctions that you could come up at least could be even longer. Emergency assistance and social assistance are very different. But in the work of the international, particularly the UN police, the two get merged simply because of operational reasons. The World Bank focuses on safety nets because that is the work that they do. And so they don't make the distinction between social assistance and emergency assistance. They tend them together. The same for the IMF. It is confusing sometimes because if you look at the most recent World Bank publications like the State of Safety Nets, they refer to safety nets or social assistance as if they were the same thing. But in fact, they are quite different. Now, I'm not going to tell you through a list, but there are a few that are really important. The first one, for example, emergency assistance deals with misfortune. Social assistance addresses poverty and vulnerability, which is mainly related to the economic system, to the consequence of the economic system. In most cases, capitalism. So there is a substantial difference in there. If you look at the driving principle behind it, emergency assistance driving principle is the Samaritan principle. You don't ask when someone is in emergency or in cases of humanitarian assistance. You don't ask for the socioeconomic status of the people. You just help them. But on the other hand, social assistance is built around a citizenship principle in which societies have a commitment to maintaining a minimum living standard. The other one, which is important, is that if you look at the actors involved, in emergency assistance, you have NGOs, charities, UN emergency services, and increasingly in social assistance, you have ministries of social development or agencies that are specifically charged in low and middle income countries with implementing social assistance. The last one is also important. The main objectives of emergency assistance are consumption and recovery, whereas in the case of social assistance, you have consumption and social investment. So there are very significant differences between the two. In what I'm going to discuss, and in the database that we have collected, we make this distinction very clearly. We focus solely on social assistance for the reasons that I discussed. Okay, so if you take a global view of low and middle income countries and we use the datasets that we have collected, this is more or less what we have. So you have on the one hand, as Miguel has already shown, an increase in the number of programs. So you go from 89 typically social assistance programs in 2000 to 224 in 2015. In some cases, there is some kind of inflation here because there are some very small programs, particularly in South Africa, in the way in which some countries do social assistance, they distinguish between different groups that are vulnerable. So you have one program for people who are with disabilities. You have another program for all the people. You have another program for children and so on. So there might be some kind of inflation there, but you can see very clearly in terms of the beneficiaries, in terms of the reach of the programs, that increases from 95 million in 2000 to 842 million in 2015. This is probably an underestimate because we don't have data for all countries for all years. So it's likely that if we did have perfect data, this would be much more like a billion or thereabouts. The other point to make about this figure is that it includes both direct beneficiaries and indirect beneficiaries. For example, where you have a social pension, it comes to the older person, to the pensioner, but it's unlikely that the pensioner is not going to share that transfer with other members of the family. In fact, I have done fieldwork on this issue in Brazil and South Africa and the majority of respondents said that they share most of their social pension with the rest of the household. It's really hard to think about the situation where you have a family who is reached by social assistance and someone would say, you are getting a social pension, this is your dinner and the rest of you, that's your dinner. It's really hard to think about that. So it's important to think about social assistance in terms of all the direct and indirect beneficiaries. And this figure of 842 million covers both of them. If you look at it by region, there are kind of different trends. So the expansion is not even across all regions, as you can see from there. The top is South Asia. These are just raw numbers. Incidentally, it's really hard to normalize this for the reasons that Martin Rebellion was discussing this morning. The beneficiaries of social assistance might not be necessarily people who are in extreme poverty. So it's really difficult to normalize. So it's better to present the actual raw figures. But as you can see, there has been a very large expansion in South Asia. A lot of it is to do with perhaps the National Rural Employment Guarantee scheme in India. The big expansion will be 2005 and 2010. But there is also an expansion in Latin America, although perhaps a level enough by 2015. Then it becomes much more difficult. South Africa and Africa has been an expansion in terms of the number of programs, but in terms of the number of beneficiaries that is not as strong. There are very mixed trends in Europe and Central Asia. As you can see, the lines move a lot there. So the expansion is not uneven, and even within the regions it's also not uneven because different countries within the region might have different conditions. Now, one question we need to discuss is what are the types of social assistance programs? And you have usually from say the World Bank or the other UN agencies a distinction between different instruments on the basis of the functions. So you have public works, you have conditional transfers and conditional transfers. In this database, we are looking at them from a much more hopefully conceptual perspective, which is to look at the understanding of poverty which underpins these programs. So you have programs which are only transferring income, and the assumption there is that poverty is to do with consumption primarily, and that by transferring income you are going to improve the consumption of these families. There are other types of programs that combine transfers of income and assets, either human assets or community assets, and for the understand poverty has to do with productivity deficits, not just consumption deficits but productivity deficits too. Conditional cash transfer programs, for example, they are employment guarantee programs are engaged in improving community assets. And then you have other types of programs that plays a very strong emphasis on inclusion. These programs tend to have quite a lot of intermediation. Beneficial is not simply receiving a transfer in cash every month, but there is also someone that is in touch with the households and follow up with what the households are doing. Intermediation is important. Chile Solidaria, for example, black challenge in the front to support the reaction is also another example. So inclusion plays a much role as assets and consumption. If you look at those ones, and this is very interesting, this is what you get in terms of the actual reach of social assistance. The blue ones are pure income transfers, they transfer income to improve consumption. The pink ones are employment guarantees, the green ones are conditional transfers, and the top break ones are programs that aim to improve the inclusion of families. So they are more integrated, they are multi-dimensional, they have much more in the way of intermediation. I'll come back to that point later on. So that is what there is, right? Now, if you put it in a global context, and here I'm going to use data from Aspar because we didn't collect data on social insurance. Aspar is quite useful because they base their estimates of beneficiaries of social assistance and social insurance on the basis of household surveys. At the same time, that has some difficulties, but there are some issues with it, which I don't have the time to go into, but I'll be quite happy to discuss later on. Now, this is what is interesting. The red ones are the households in the household surveys that receive nothing, no social insurance, no social assistance. The third that Martin Ravellian was discussing this morning is his presentation. The green ones are households that receive only social assistance. And as you can see, the green one is very, very, very significant in most regions there. The orange ones are the ones that receive only social insurance, and then you have the blue ones that receive both. Now, the kind of takeaway from this is that basically social assistance, in terms of the number of beneficiaries, in terms of the reach of the program, is the dominant component of social protection. Of course, if you look at it in terms of budgets, the opposite is the case. Social insurance is much more significant than social assistance. But if you look at it in terms of the number of beneficiaries, then social assistance is important. Right, I'm going to miss the next one because it will take me a bit of time. Right, yes, I don't... Okay, so very quickly about the research gaps. So it's likely that on the basis of what we have seen so far in the 21st century, it's likely that the welfare institutions' social protection in low-income countries is going to be based around the core of social assistance, not social insurance. That is an anomaly if you look at it from the perspective of high-income countries, particularly European countries. So we need to explain why is it that this is happening in low-income countries. There are a lot of explanations that you see in the list there. Unfortunately, these are all environmental explanations. So you have the democratization processes that facilitate social demand, economic growth that facilitate fiscal space, left coalition governments, for example, in Latin America in the 1990s, sorry, in the 2000s, international contexts, millennial development goals, sustainable development goals. The lack of dynamism and failure of the social insurance model, which is particularly acute in Eastern and Central Europe, the former socialist countries, and also in Latin America. But notice that all these are not... It's not something that happened in 2000. It's something that has been there before. There are kind of instances, phases where these conditions have been there in the past. A lot of research has gone into looking at social assistance as an electoral tool. Just the quick response that I would give is that the more we look at it, the more that it isn't an electoral tool. This is why the rule space is politicians that want to use social assistance as an electoral tool would be much better of putting hospitals in place or schools in place than giving private transfers. And not only private transfers, but where they cannot control who gets it because there are rules that prevents them from manipulating that. Very quickly. Two, what is the relative weight of social investment and protection? Again, as you can see from my picture there, conditional income transfers and integrated anti-poority programs are a very significant part in terms of the reach of social assistance. Clearly, there is a much greater interest in social investment. And we need to know a bit better what is the relative weight of these two. There is a lot more to discuss. Again, this is something that has been discussed a lot in the context of European countries. In 2013, the European Union passed a social investment package that is pushing European countries towards greater social investment. And what is interesting is done in terms of social assistance in low-income countries, is happening for low-income groups as opposed to better-off groups. And the last one, perhaps this is the most important, we don't know very much about the politics of vertical distribution. The politics of horizontal distribution is really straightforward. Why is it that people have social insurance? Simple, because they face similar risks. And by pulling risks, they are going to be better off. And this is why you have social insurance. This is self-interest that propels people to put forward to implement and belong to social insurance programs. Why would better-off people pay taxes to reduce poverty? Groups which perhaps they are never going to belong to. That is a different question altogether. And I have here social contracts, left coalitions, transvestite and electoral tools as potential avenues where we need to try to work out why is it that, what is the politics behind the expansion of social assistance. And I leave you with my conclusions because I have run out of time. Many thanks. Thank you.