 Okay, so hopefully I can sort of pick up from where Tony left off, and what I want to do is try and examine kind of the literature that we have and what we already know about locations and neighborhoods in urbanizing Africa and kind of focus on the microeconomic evidence that we have. Before I start, I just wanted to do kind of three framing points of motivation. Firstly, just focus on rural migration, rural urban migration, and kind of highlight that as part of the framework that we need to think of. Secondly, look at what we know about neighborhoods and economic outcomes and the evidence that's coming out of developed countries there. And then talk about the rise of welfare spending in African cities and the move of welfare spending to those cities, which we need to think about in the context of these other two points. And so hopefully that generates a few questions for urban research, and then I want to talk about just two examples of work that I've been doing in Addis Ababa to try and look at these kinds of questions. So the work I'm presenting is focused on urban Ethiopia, but hopefully give some examples of what we do know and what kind of research we can do to answer these kind of questions. So the process of urbanization has already been discussed and underscored. I want to highlight something that's sort of implicit in what Tony was saying. It's the extent of migration from rural areas to urban areas. So many African cities, Addis Ababa is an example of that, have gone through the demographic transition. Fertility rates are incredibly low in a lot of African cities, and yet the cities are growing at an enormous rate, and that's of course because of migrants from rural areas. So I've been looking at the Addis Ababa, the Ethiopian census data, and that data tells me that Addis has gained 25% of its population in migrants over the last 10 years. And so the growth of cities is being driven by high fertility rates, but high fertility rates in rural areas and then migration to cities. And that's something we really need to think about, that we're dealing with urban populations that have just arrived, that are new to the city, and that has to be considered. And a lot of the migrants in these countries are young people. They often move alone and we're not talking about families moving together, although there is a lot of that as well. And it's young people looking for jobs, kind of high productivity jobs, and ambitious young people. Secondly, a lot of you will probably be aware of the new research coming from Raj Chedi and co-authors from the United States. And it's a long way from Africa, but it's really showing us just how important where you grow up, so the location of where you live in terms of your outcomes and income mobility. I think this research is really important, and it tells us something about how important mobility is. So being able to move to high productivity areas is an important part of how people improve their living standards. And we don't know very much about what this looks like in developing countries at all. What we do know is about urbanization. So we do know that there's special inequality between rural areas and urban areas, and we know that if people move from rural areas to urban areas, their incomes often grow, well, incomes are much higher, their access to services are improved. And we know that there are barriers to them moving, so mobility is constrained from moving from rural areas to urban areas, and people would move more if they could. So we know that, but we don't know much about what happens within cities. So once people arrive in the cities, once new migrants arrive there, where do they live? How are they housed? What neighborhoods are they living in, and what impact does where people live when they arrive in cities to impact their outcomes? And so there's a lack of knowledge of the role of networks, neighborhoods, access to public services, and so on for individuals participating in labor markets. There is some evidence, of course, on slums, and that's the focus of a lot of urban research in developing countries, trying to understand what the impacts are of growing up in a slum, how to improve the situation in slums through tenure security programs and slum upgrading. But in light of what I've been saying up until now, it seems that new slums are going to be located increasingly further away from cities, so we're no longer just dealing with kind of the inner city slum, people, residents who've been there for generations and generations, those people are under threat as well, their rents are going up, they often are forced to relocate because of development projects going on in the center of cities, and those are a set of issues that need to be addressed, but we have a separate set of issues which are these kind of upwardly mobile, young people, migrants to cities who are looking for places to live. And in my work in Ethiopia so far, I see an enormous amount of mobility of households and individuals in the cities, and the rental market is actually working quite well there, and people are relocating for jobs, they're relocating to be closer to the places that they work when they find jobs and to look for jobs, and that's an important part of how cities work and function. So just quickly, I don't want to go into too much detail here, it's not the focus, but there does seem to be an increasing move towards urban spending and welfare spending. So we've seen a lot of success from work fair programs and rural poverty alleviation schemes, and now increasingly there's pressure on governments to roll these out in cities. And with the kind of rise of unconditional cash transfers and how effective they've been seen to be, there's more and more of this sort of activity going on. And certainly for the urban citizenry, there's a lot of pressure on governments to start delivering these things, especially when you're living in a very unequal society and you see the wealth all around you, there's a lot of pressure on governments to start rolling out more welfare in cities, and that's starting to happen. One area is social safety nets, and the expansion of these in urban areas, and we need to think about are they mitigating the risks of migration, are they providing the right kind of safety nets, and how are they influencing rural urban labour markets and labour supply. And the other one I'm going to focus on is housing, and the kind of large scale housing projects that are being built, almost every day I seem to hear about a new country in Africa that's starting to build public housing for its people, and I just want to talk a bit about how those sorts of housing projects are designed, are they being built in the right areas, and are they actually aiding people to get to work and live close to work. So that's a bit of a research agenda, I've discussed what I think we need to be looking at already, we don't have much evidence on this at all, almost all of these questions that I've been discussing, neighbourhoods, social networks in local areas, how labour supply is impacted by welfare spending and social grants, all these sorts of things, we know a lot about that in rural areas, and we don't know much in urban areas. And it's for a number of reasons, I think one is the types of data that we have, we have data at a low enough level that we can look at different neighbourhoods within cities, the other thing is that urban fabrics are continuous, so it's hard to delineate one area from another and look at the impacts of one on the other without there being enormous spillover effects from neighbourhood to neighbourhood, these kind of general equilibrium effects within cities, which are hard to look at, especially because we only have small samples of cities, and so they're hard to kind of factor in. And as a result, there isn't a lot of evidence out there and we need this kind of more desegregated data, but there's also I think a role for RCTs to try and deliver evidence where we can't quite do it. So that's the kind of thing that I want to be talking about here, which are two randomised experiments, one's a control trial and one is actually a government experiment. Both of them look at a suburb and they both look at questions of where people live and ask the question, does it matter where people live? So the first one, I'm going to look at the impacts of reducing transport costs on job search in Addis Ababa, so if we allow people to travel to work, does that improve their participation in the labour market and secondly the impacts of government housing in the same city. So here are some maps, similar to Tony's maps of developed country cities, this is urban density in Addis Ababa and it is actually quite a monocentric city, those tall towers being the residential areas in the centre of town. This is kind of a heat map of where people live, the darker blue places being the higher sites of urban density and that orange ring is kind of the median distance at which people live from the city. The city is growing enormously as I showed you early on and increasingly people are being located outside of that ring, further and further to the periphery. The median distance now is I think five kilometres from the centre and it's getting further and further as the city sprawls and grows. The firms that we have in a study we're doing with firms in Addis Ababa and we're looking at how they do recruitment in a separate project, we've mapped them and these are all the big employers in the city, they're really, really big ones and they are clustered in the centre apart from a few manufacturing firms that are on the kind of southern road down there. And so you can kind of map the heat of where jobs are, heat map of where employment opportunities are. And those are for all kinds of jobs in all kinds of industries. If you focus on the kind of highly skilled jobs, you end up with a much more concentrated set of locations. So those are the kind of really good high skilled jobs that young motivated people are looking for. And this is where the information on employment opportunities is found in the city. So these are public vacancy boards. This is where you go if you're looking for a job in Ethiopia. You go and check these boards and look for vacancies that you can then apply to. You then also have to go on and make the applications, go and talk to people, go for interviews and these sorts of things. And the premise underlying this research is that process is expensive because transport costs are very high for people who are living a long way from the centre of the city. And so that creates a barrier to those sorts of individuals investing in job search optimally. So I'm not going to go through in detail the paper that I have on this which has kind of a theoretical model of cash constraints and a whole long RCT that we did. But to summarise it very quickly, we did this control trial where we gave our transport subsidies for 12 weeks to job seekers in Addis Ababa. So it's a small scale study. It's not something that we're thinking of as sort of a scale that policy. But what it does do is it tells us something about how job search works and what the constraints are for young people in the setting. And what happened was when we gave people these transport subsidies and we lowered search costs, they searched more frequently for jobs. So that tells us something about how they were constrained and they ended up with better employment opportunities at the end of the study. They had more permanent jobs by significant margin. And so because the subsidies only brought people from the periphery to the centre, it only brought them to where other people were already living, that we can kind of take as causal evidence that this distance in cities is generating frictions and it's leading to inequality of access to opportunities. So people who are living on the outskirts have an unequal opportunity to access employment and are being locked out of labour markets. And given that that's going to be the majority of new people in cities, that's something to be concerned about. Just a couple of other points that come out of the research and I'm just going to go through these very quickly. We did a phone call survey where we phone people every week and a lot of interesting stuff comes out of that about urban livelihoods. So young people's outcomes we found to be surprisingly volatile. Employment isn't a yes-no thing in this setting. People move in and out of all sorts of different types of jobs. Construction work for men and domestic work for women and they switch in and out of jobs all the time. They also switch out of job search in different types of job search. They're experimenting with different things. So their lives are quite volatile and it's something to kind of take into account. The fact that these subsidies had some kind of effect indicating that just by giving someone a bit of cash that helps them to buffer against that sort of volatility. And what we found was that the subsidies actually reduced participation in temporary work for a while. So these kinds of inferior forms of daily labour, there was reduced participation in that, suggesting that people need those sources of income just to keep searching for work. So there you can see how over time the control group in our sample gave up job search over time. They became discouraged and stopped, whereas the individuals who had the transport subsidies were less likely to become discouraged. So there's some power in having this kind of deep data, this kind of high-frequency data of how young people live their lives. Some quick things on targeting and take-up, which I'm not going to go into too much detail. But one thing about this policy was that it was, the take-up was conditional on you arriving to collect the money, so not everyone got the money, only people who wanted to participate in the study. And so that in some sense provides a self-targeting. So very quickly, implications of the study, I think there are two main things that come out of it. One is for urban planning, which we've been hearing about already, talking about the need to make cities denser, create more housing in the centres that the people are closer to the employment opportunities, and to reduce transport costs. And there's been an enormous expansion of the transport network in Addis recently with a light rail project, and it's going to be really interesting to see what that does to labour markets, and we need new methods. As I was discussing earlier, how do you evaluate a project like that? How do you evaluate a whole new railway line in an urban setting? And that's a challenge in a project going forward. And secondly, it tells us something about how young people are cash constrained, and there's a role there perhaps for welfare programmes, unemployment insurance and things like that, which are definitely being spoken about in the urban context, but important to think about how they're designed to support people who are particularly vulnerable and unemployed. Okay, so that's that. I've gone through it very, very quickly, and we can talk about more details later. I want to just quickly talk about a new project that I'm starting on, on housing in Ethiopia, which is on very similar themes. So this is a picture, an aerial photograph of a public housing project on the outskirts of Addis Ababa. So each of these are a five-story building, a five-story housing project that the government's been building to address the shortage of housing in the city that so many African countries face, which is why we see so much in formal housing. But there's a question about whether you upgrade existing housing stock, upgrade slums, or whether you relocate people to new housing projects. South Africa is the key example of relocation, where the government's built three million houses over the last 15 years. That project's been widely criticised, despite the fact that it's delivered so much housing at scale because of where the houses are located. So they're right on the periphery of cities, far away from jobs. And that's considered now to be a design failure on a momentous sort of scale. Ethiopia is going down a similar path with some variations, and it's going to be important to see how that plays out. So these are four or five condominium story houses. So there is that density element that the South African housing project didn't have. What these are subsidised mortgages, ownership is transferred to the residents who are going to move in. They're building this at an enormous pace. More than 50,000 units per year in a city, you know, of four million people. It's going to have a huge impact on the urban fabric and where people are housed in the city. And what's interesting about it is the government's using a lottery to give out the housing because there's so much demand. So I think about 500,000 to 700,000 people are registered, are waiting for this project, waiting for housing. And on these waiting lists and the government's using a lottery to try and fairly distribute the opportunities. This is again a map of population density where people live. And this is kind of the new master plan that the government's aiming to put in place, some new roads you can see there. But then also the kind of dark blue splodges there. Those are those are the housing projects. And I can tell you that those are very, very far away from where people currently live. They're in the completely yellow areas, which means that there's nobody living out there right now. So right on the periphery. And you could see that on those aerial photos that I showed you. So in a sense, this lottery provides a natural experiment, maybe something like many of you would have heard of, the moving to opportunity experiment in the US where people moved to different neighborhoods and you look at the impact of that. This may be something where it's a little bit in reverse. So this is moving people further away from local neighborhoods. But to new communities, we don't know how those communities are going to develop. If you go and have a look at these housing projects, they do seem to be incredibly vibrant economic places. And so it's not clear whether they're going to be places of opportunity or lack of. So what the study is doing is we're tracking a sample, firstly, of a thousand households who won a lottery recently in March. So they were told by the government that they're going to be moving into housing within the next year. They haven't moved in yet. And what we did was we sampled 200 different corridors. And in each corridor, there are five households that are going to move in across nine different housing sites in four different neighborhoods of the city. OK. And so what that means is we've got we've got groups of five people who are going to be neighbors in the future. They're going to be living in very close proximity, although they don't just yet. And that's an interesting part of the study design. And then secondly, we've got this group of a thousand households who are still waiting for housing. They were eligible at the time of the last lottery, but they didn't they didn't win. And we've matched them to the winners on key observables so that they're directly comparable. And at the time of the March lottery, there was basically a 50. If you're in my sample, there was a 50 50 chance that you won the housing. So we're going to be able to look at the the impact of winning housing by looking at this treatment group. This control group do stand to win housing in the future. And so we're going to be able to tell when the next lottery happens, if there is if it's definitely a fair lottery and if there's kind of balance across the winners and the losers to make sure that there isn't manipulation in the process. OK, running out of time. So just just quickly, I'm going to skip over discussing the financing of this and how it works, but we can talk about that a bit later. What I want to do is just to look at firstly the impact of the physical housing. How does it impact people's lives, their their assets and health savings? What happens to communities when you're moved away from the community that you lived in and how do new ones develop in these housing projects? And based on the new work coming out of moving to opportunity, it's going to be really interesting to see in the long run what happens to to children that grow up in different settings. There are no public services in these areas yet. No schools and hospitals and so on. Those are going to be built. Hopefully by the time people move in, but it's not it's not clear that they're going to be good access to services in the short run. Secondly, we're going to be able to look at where households move. There's random variation in the different places that people are going to end up. So we can we have some some people that are moving into housing projects that are very close to the centre of the city and some that are going to be further away and we can use that natural variation to compare outcomes for people that are closer or further. And finally, hopefully you can do some interesting work on these corridors of people. So we have people that are randomly grouped together in groups of five. They're going to have to interact intimately over the the rest of their lives. And what's going to happen there? You're going to be grouped together with people that you didn't choose to be grouped together with. And how does that affect social cohesion in urban areas? How people in how people deal with strangers with people of different backgrounds, ethnicities and languages? And so hopefully exploiting this random corridor allocation, we're going to produce some evidence on that. OK, that was a very, very quick overview. And I'm going to turn it over to the next speaker.