 Ie, felly mae genny'n ddondeig ymwydiad iw Ligicol. Ond yna yn amlwg amser y blodl yw eu bobl wedi'i gweld'i unrhyw gweithio expertaf o unrhyw ar gyfer Ylwbr mewn i'r bwynt yn gweithio iawn. Yn y ffrwng ar yw'r amlwg, bwysig rydyn ni'n ymddangos. Ie wnaeth amlwg ymgwrdd wedi gweithio i fy ngwrs ar gweithio arno. Mae cyfrifio bydd iawn ymweld iawn o'r cyfrifio data yma. Gwyerdd cyfrifio, gwair hynny, gan certyn i'r cyfrifio sydd gyffredig, ymgyrch o'r cyfrifio cyfrifio, gyfrifio yn y cyfrifio, cyfrifio sydd gyfrifio, yn cyfrifio cplaygu. Ond o'n gwirio i'r cyfrifio cyfrifio. O'n gwirio i wychidol yn cyfrifio cyfrifio cyfrifio, Y nhw'n ddod, y sydd yn ei ddweud i wneud, ymanylch, y cam, ychwanegu y ddataeth o ddano'n ysgol, y ddod yn iawn i gwerthu digwydd, ac dyna'n meddwl'u gyd deithas. Da, rwy'n meddwl a'i ddod. Felly yma'r ddataeth aogaeth gynllunion arloed. Felly mae'r ysgol yma, ydych chi'n gwrthod yw'r ysbun yw'r tynnu ymddangos. Yr wyf yn ymgyrchol yn ymddangos cyffredig, a'r cyfnod yn cyllid, ond ydych chi'n gweithio'r arddangos. Felly, byddwn i'n gweithio, yr ysbydd ar y project, a'r cyfnod ymgyrchol yw'r gweithio. Felly, byddwn i'n gweithio, mae'n gweithio i chi'n gweithio i gwrthodol ar gyfer y ffindig, yn cyfnod ymgyrchol. Yn lighterau, Y dechrau'r cael sy'n gweithio, y gallai amser, y dyfodol sy'n cael ei ei wneud serfio'r cyffredinfa, mae wedi'u'n meddwl yn dod i gael ar y cyffredinfa'r tael gan ti'n rhoi'r cyffredinfa'r cyffredinfa'r cyffredinfa'r cyffredinfa. But a chynghwitaidd a chynghwitaidd a chynghwitaidd y gweithio ac hynny ychydig yn gwneud hynny i gael ar y model, a yna gan ddwy ar y cyffredinfa cyffredinfa'r cyffredinfa'r cyffredinfa'r cyffredinfa. It's going to be a rather conceptual issues given talk. OK, that's the plan. Let me start by the scale of the challenge, just that quantitative motivation for thinking about this. You see the numbers that I put on the bottom of that slide. This is just focusing on Africa. Africa is probably about a third of a way through its urbanisation process in proportionate terms, not in absolute terms. So just doing the arithmetic on that, sub-Saharan Africa, we're talking about another half a billion people also entering African cities over the next 25, 30 years, and obviously population growth and the urbanisation process going on. Just to drive that number home to you, that's a third of a million a week entering cities, or it's saying that Africa will have to build one of the numbers twice as much urban capacity over the next 25 or 30 years as it's built previously over the last 100 years. Imagine you're African cities now and over the next 25, 30 years it's got to build twice that over again. Obviously city structures last, buildings last a long time. It's really rather important to get this right. It's a huge challenge. I find it hard to think of more pressing urgent problems in development economics and housing, doing things to house that extra half a billion people and get them jobs functioning in cities over these coming decades. That's the scale of the challenge. I said I'm going to talk about issues. I really want to start talking about urban potential. How we think cities, what cities can deliver how they should work. I'm going to start not talking about development in particular, but talking about urban potential and then lying Africa up against it a little bit. What do we know about cities? I've put successful cities there. I think I probably mean most cities in the developed world more or less successful. First thing we know about them is that they have high productivity. I've put that 5% number there. In the cross section, as you double size, productivity goes up by about 5%, which doesn't sound like a big number until you think of a city of 100,000 and a city of 8 million or something and that's an awful lot of dubblings. 40%, 50% productivity advantages to being in cities. That's come out of the econometrics. It's fairly robust facts that come out of looking across cities, come out of looking at firm data, come out of looking at individual data, come out of econometrics where you've actually tracked, where you have individual fixed effects. They've actually tracked people through their lives and moved to cities and then moved out again. That sort of ballpark number is pretty robust for developed countries. For developing countries, much less work has been done, data issues, and in so far as work has been done, it's probably a weaker relationship for reasons that I will come on to later on. It should be the same relationship. It should be that productivity message I think should apply. So cities are productive places. Cities are places where new things happen. So again, developed countries, we know cities where R&D takes place, innovation, patents, big cities are active in that. For developing countries, possibly more important, cities are where new things start up. They're incubators. If you're going to set up a firm doing something new, do you do it in a small village or do you do it in a city? Well, you probably do it in the city. So cities have those fundamentally important attributes of high productivity and where new things happen. Why? What's special about cities? Well, Jonathan's already said, ultimately it's Adam Smith. It's a scale in specialisation that drives it. Let me unpick that a little bit very briefly. I mean, it's sort of big markets. It's an intensity of economic interaction. So if you think about labour markets, obviously a city has a big, thick labour market, so firms looking for particular types of workers, you can match skills better. Probably most importantly, the incentives to acquire very specialist skills are much greater in big cities than elsewhere. If you're going to specialise in something or other, you're only going to do it in a place where you know there are lots of potential people who want to buy that skill. So that's got to be a city. So labour markets, product markets, you get scale and competition. Now that trade-off between firm size and intensity of competition is relaxed, the bigger the market. You get forwards and backwards linkages, good old fashioned arguments from development economics. If you've got all the intermediate suppliers nearby that sort of intensity, those linkages are strong. You get knowledge spillovers. And also perhaps you get economies of scale just in electricity, utilities, public service provision, all that stuff. So fundamental mechanisms supporting those productivity numbers I gave you. The mechanisms really work through connectivity. This sort of density of intensity of action. So lots of workers are close to lots of firms. I had a number somewhere, I think in London, I forget the exact number, it's something like 2.5 million workers within 40 minutes by public transport of 2.5 million jobs or something. So that worker-firm interaction is really intense than consumers to firms. And perhaps most importantly, firms, firms. So you've got your intermediates, all that. But of course, let's think, if it's connectivity driving it, then what do you need for connectivity? Well, it's either one or both of transport and density. So you need transport infrastructure evidently in a city. A city is a transport infrastructure intensive. And that requires public investment. Evidently, be it public transport or road transport, very massive public investment is a key ingredient. And the other ingredient is density in the sense of residential accommodation, business accommodation being close together. Not just sprawling out flat, but being close together. I'll say more about that. I mean, it's absolutely a crucial point on where Africa's not doing terribly well in constructing urban fabrics that support this density. So you can see the line of argument I'm getting here. You need transport infrastructure density to generate the interaction, the connectivity that supports the productivity and so on and so on. A little bit more on density. Density is one aspect of just efficient land use in a city. If you're trying to get to the centre to work, then land values will be high and an efficient allocation will mean there's an incentive to build tall and that supports density. The textbook Monocentric City is a city where there are these rent gradients. So land prices rent is high in the centre. That's the incentive to build tall to be dense and then it tails off. When I say textbook monocentric density, people tend to think that's just the textbook, but at this point I always like to actually show a couple of pictures for rather successful cities. They are monocentric. That's employment density in London, New York and Hong Kong. You get about 150,000 workers per square kilometre. Fantastic density in the centre. Each of those cities, you can spend hours looking at these maps, there's canary wharf and Croydon. Residential density, less spiked, but again that density. London much less dense than New York and obviously Hong Kong as you all know, enormously high residential densities. This message of density, efficient land use is really important. One other set of pictures on that. These are different cities on the horizontal axis, different distance from the centre on the vertical density. The top row are Asian cities and they're all to the same scale. So Shanghai in the centre there, fantastically high central density. The middle row are European cities. So they've all got the gradient. A little bit hollowed out in the very centre just as our message is residential. They've all got the gradient. The bottom is American cities. Clearly levels of density depend on the transport technology when the city was built, but they've all got that sort of shape. So the density message, I'm trying to hammer over to you a bit. These are the same figures for some non-market, shall I say, cities. Johannesburg obviously had a particular history when the centre density occurred, residential density, way out of the centre, in Moscow and in Brasilia. So density matters for connectivity, for productivity. Density can be achieved by a well-functioning market or in the absence of that other controls cutting in. It's not achieved, and I would argue with considerable costs. Right. That was a run-through, I think. Everything I wanted to say. No, one other thing on urban potential before we actually start talking about Africa and development and things. So cities, density, delivering productivity, delivering productivity, great. But cities are a balance between that on the one hand and they are high cost places on the other hand. Okay. Commuting costs, land values, capital costs of building, dense buildings, the transport infrastructure or whatever. So I see cities as this balance between potentially delivering productivity but also being high cost. So there's a balance there. Now of course in the successful city, that balance works out such that the productivity offsets the cost. And indeed there's some surplus in between and that surplus is typically, or much of it, are typically capitalised up into land values in the city. So I want you to have this sort of model of the successful city in your head. One other thing before beginning to talk about Africa. A corollary of what I just said is that cities, they may be expensive but they can be completely self-financing. That surplus between the productivity benefit and the high cost is capitalised in land values. And land values are a very good thing to tax. Land values are a great thing to tax ethically because if I own a bit of land and its value goes up, I haven't done anything to generate that. It's the city effect that's generated this. So in terms of fairness it's great. It's non-distortionary if you can tax the land value, not the structure on it. It's relatively easy to collect. And there are these Henry George theorems saying that the tax is sufficient to finance all the infrastructure. So that's the model while we're in the hand of your mind. How am I doing? Another? John may have gone. Five, six, five, six. Okay, let's talk about Africa. It's not going to be a great empirical set of stuff but some remarks. What do we know about urbanisation in Africa so far? First, it is occurring early in the sense of urbanisation levels relative to a per capita income. And that is sort of inherently problematic. If you want to build decent housing it's quite expensive. We can talk later about actual numbers on that. Maybe there are some people who argue Africa simply can't afford it. I don't agree with that. But Africa is urbanising, I was starting to urbanise it, historically unusually low income levels. Second fact, Africa has got massive excess primacy. That is to say, the largest city in each country is way larger relative to the second and third than is typical. Okay, I think lots of historical reasons for that on the role of governments in attracting anyway. But there is that primacy issue so that the usual urban hierarchy hasn't evolved in most African countries and I think we'll have to. Above all, investment levels have been too low to support the transport infrastructure required and the more general utilities, public services, that whole city infrastructure. Investment levels have been too low to support that on the public investment side and also on the private investment side. Private residential investment levels have been too low. Obviously, when you look at the housing stock in many African cities, it's just horribly bifocated. There's high income, decent housing and low income shacks of one form or another. Terrible under investment and that bifocation and informality. But of course, a corollary of those two lacks of investment, lack of investment in both those areas, is a failure to deliver the connectivity. If you're not actually investing in moderately high rise buildings by which I mean five story, not 50 story, if housing is just sprawling out, then you're not delivering density. If transport systems haven't been invested in, that's the substitute, you're not delivering that either. So the driver of the sources of growth and productivity in a city are absent. So this failure to invest is other stylised facts, obviously, informality in the housing sector. Also informality in employment. So a failure to attract firms. So a very few stylised facts there. But let me try and put those together with that theory, with the model of the successful city that I just outlined. I really think there's a danger that we're in a world of multiple-equilibria poverty traps, a cumulative causation, the sort of world Steglitz was talking a little bit about it yesterday, but at the urban level. So let me give you three sort of mechanisms that I think each of them have a cumulative causation element in them, and collectively are really, really dangerous. If you think about attracting firms and jobs to cities. Well, okay, African cities are in a situation where land use is inefficient, you haven't got the density, you've got sprawl instead, you haven't got the transport infrastructure. And inefficiency, by definition, means relatively high cost. So these are cities that have low real incomes, but nevertheless have high nominal wages and high cost cities. So do investors come in? Does footloose manufacturing come in? Well, obviously not. These are horribly high cost. They're low real income, but high nominal income are cities because land use in the city is inefficient. The city is a high cost, okay? So put that together and you're failing to attract manufacturing. And there's this empirical literature on, I mean by Doug Dolin and others, on urbanisation without industrialisation. So African cities, particularly those in resource rich countries, urbanising but just not organising the city in a way, having some Dutch disease from the resources, putting that together just not attracting. There's a lot of high productivity activities you need. A second element to this, structures, buildings have long lives, therefore construction decisions are forward looking, therefore expectations matter. So it's easy to tell a story, write down a model, whatever, where there are self-fulfitting expectations. You expect your city to be slow growing, not attracting manufacturing. Therefore you think rents will be pretty low in the future. If rents are pretty low in the future, you're not going to, it's not worthwhile building tall, you'll just build low. So the city will sprawl and be high cost and inefficient and won't attract the growth. And again, we're in a cumulative causation story. Another possible cumulative causation story here, public finance. I've said that in principle tax the land values, it's enough to pay for everything. But there's a timing issue there, obviously. The land tax comes in in the future, but you need the money up front. Well, in China you can fix that, by essentially expropriating the land and selling it off right away. But there are problems here. So weak tax systems, poor expectations, inability for government to capture expected future appreciation means that you're not getting the private finance. In those cumulative causation mechanisms together, I think real poverty trap, low-level equilibrium stories, and real worry that existing African cities are rather caught. Okay, final slide. Final, good. Policy. Now, I'm not going to, obviously you don't have time to go through a great full policy diagnostic here, but a couple of points. Obviously when you look at the residential market, there are blatant egregious market failures. We said there was under investment. Well sure, if property rights are insecure, if finance mortgage markets are not available, if regulations are inappropriate, lots of African countries have got very low floor area ratios, very large plot sizes, so regulations are simply ignored. So there are things that can be fixed there. The structure, yeah, public policy response, obviously I won't say any more. Similarly on the commercial side. So there's a policy agenda that I've said and I've sketched here very loosely, that obviously the projects will, as the empirical work goes on, get much more diagnostic, drill down and say much more specific things other than just on one slide there. So just to conclude, quantitatively an absolutely enormous challenge here, a real policy challenge and quite a difficult one, because it's across so many areas of government. You've got to get the city mayor doing it, you've got to get the finance ministry doing it, you've got legal stuff to do on property rights, you've got to get industry moving in it. So it's cutting right across many areas of government. So you really need kind of presidential role.