 Thank you, James, for the introduction. I can't help but maybe add one point on the peaking of the youth bulge and sort of an insight I kind of got not so long ago, but we talked a lot about the demographic dividend. There's some recent work pointing out that actual demographic dividend is a lot related to the education of the new cohorts coming in. So it's not so much that they're more working-age people, which is sort of in an accounting sense, but that it's actually that these new youth is also better educated, and that if you bring that into the share of how much it explains of the actual contribution of the demographic dividend, that that's a big part of it. I think it's something we need to take into account also thinking about Africa. We're on the one hand, yes, enrollment rates have gone up quite a bit, but we also know that the quality of that education has left something to be desired, so I think just as a caveat. With that sort of moving into this presentation, so basically it's starting from this observation or kind of the question which has been high on the agenda, sort of panicking almost like Africa's youth is leaving agricultural mass and sort of what do we do? So I think it's useful to clarify that this year a little bit, James started with it, but I think it's sort of two, why is this a problem? Why do we care about this? So one argument could be sort of from a food perspective that yeah, we have a growing world population, also Africa is a growing population, so we need to produce food. If people don't want to farm, who's going to produce the food? So maybe that is the concern. Or the other way of painting the concern is that if they're all leaving agriculture, where are they going to be employed? That's sort of from the employment side. But then at the same time, an exit out of agriculture is what has been observed all across the world throughout history, so that might actually be a good thing. Often it is seen as a sign of a good thing of positive development, doesn't have to be. There's no automatic equation, there is structural transformation that is a sign of development, so you could use it as a diagnostic. It could also be driven by push out of agriculture and then sort of you get an urbanization of poverty if you want. But the notion that as countries get richer, less people will be in agriculture, is normal because our preferences change. This is simply Engels's law. If we spend less on food as a share of a total income, then there will be fewer people producing that. So it's actually a normal sign. Now it's also true that that structural transformation often happens through youth, so we would expect as people leave out of agriculture it tends to be the younger who tend to leave more agriculture than it will be the older people. So as I said, but there is a problem if it's structural transformation without agricultural productivity growth. So this sort of starts to frame the issue, what is the level, first of all, what are the data? What is the level of exit? And is this excessive? Is this normal or if it's normal then it may not justify a youth specific agenda. This is sort of a normal process which we observe as countries develop. So in this work, which is jointly with Amparo Palacios Lopez and Eugene Maiga, so we started to look at is youth leaving agriculture? So the numbers James put up was sort of without age differentiation. So how do you actually look at that? You can take cross sections, is that enough? What are the problems with that, et cetera? Then the next question is it leaving agriculture more than the adults? So is there something specific? And then the next question is, if it's leaving more is that an excessive? Is that what you would expect in a structural transformation sense? So you kind of want to layer the questions a little bit. And then if there is time, there may not be time, so look at the timekeeper. Also where would people be employed and some recent work by David Shirley, which I think was supposed to be here today, I can talk a little bit to that. This basically work drawing on six African countries with basically repeated or cross like panel data, cross sexual data. So to start with, so one simple way of looking at this question is simply to compare. We have the population employed in agriculture, we just differentiate by age. In this case, youth is broadly defined up to 35-year-old, 20 to 21 to 35-year-old. Compare them with the 35 to 60-year-olds and it's youth less employed than the other group. And basically this is the number of hours per week work that's coming from the LMS ISA data. And you see that basically youth is less employed in agriculture, especially in Nigeria, that's where the difference is the biggest. Otherwise it's up to two hours, an hour and a half. You also see that sort of in the poorer countries there is already less of a difference. But this is of course a bit problematic because there may simply be a life cycle effect. It may be that youth, before they go into agriculture, they actually try some other things. The land may not be ready, they may not be able to take over from the father or in the household. So they get into agriculture at a later stage. So if we observe that all around, then this is just simple normal, it's nothing that there is exiting agriculture in a way. So the number of factors, so lambda A here, is the participation of youth, the number of hours work in agriculture of an individual. So it's a life cycle effect. There's a country effect, sort of some countries may have richer agriculture than others, so there may be more people employed in agriculture. There is sort of a, when you observe there may be some specificity, sort of a year specific effect. And then we sort of want to net net some of the, we want to net out the life cycle effect in essence. And sort of we are gonna, let me just illustrate this with an example. So this is, here we have Vietnam where you could argue that early on in the 80s, they're sort of at the beginning of their structural transformation, now they're a bit further along. So this is 2009, so we do the simple cross country or cross sectional comparison about the share of people employed in agriculture across, compared between the older and the younger age group. And basically you get the difference between 46 and 58%. So 12%, youth is 12% less employed in agriculture. But as I said, one could argue this is simply a life cycle effect. So one way to deal with that is to sort of look at this over time. And so we have also in 1989 where you see early around was a higher share. And there the difference was basically 3%. So early on in the late 80s the young were 67% of the young were employed in agriculture and 71% of the older age group was employed in agriculture. So only a difference of 3%. So you could say, look, that difference, let's take that as sort of the life cycle effect in a situation where there's little structural transformation where one is sort of at the beginning of that structural transformation. You can also do it differently. So when you say, we basically compare youth across time. So then you get a decline by 21%. The older age group also exit agriculture by 13%. So basically it's that difference in difference which gives you an idea of sort of the age related effect, exit of youth. So the youth in this case is 8% more leaving agriculture than controlling for sort of survey design issues as they come with these surveys as well as life cycle effect. But even then the next question is, is this excessive? Is this what you would expect over this type of period in a regular transformation process? If it's not, then what specific things need to be done? So we apply this, so I just illustrated it for Vietnam but sort of across these five countries, five African countries from the LMS ISA data. So we basically over a 10 year period, it's a little bit of extrapolation. We have two years, two survey years, and I just to normalize them across the different surveys so we expanded it to 10 years. So basically yes, youth is leaving agriculture but so are basically the older age group. So if you look at the orange part, then net specific youth related age effect is very little. And so of this notion, there's nothing like an accelerated exit of youth out of agriculture. Youth are leaving agriculture just as the older age groups are leaving agriculture according to this data. So basically if we see why is youth leaving agriculture, a lot of it is related to age-related attributes. So they tend to be, especially education is a major one. So here we under the first column, this is just linking youth, youth dummy, linking it to the number of hours worked in agriculture and then you do youth dummy controlling for a lot of the other characteristics like education, household size, land, et cetera. And it's sort of, once you control for it, then basically that age effect disappears to a large extent. So in two, three countries it becomes insignificant. Actually in Niger it goes up. So the youngsters are moving into agriculture according to this analysis. But it's most pronounced in Nigeria where in a way you wouldn't be so surprised. That's also the county first ahead in the structural transformation. So that their youth is more rapidly leaving agriculture is less of a surprise, I would argue. So this is basically then full results with all the different control variables. And if you look across these different countries, what you see is, so you control for age, but that it's really the factors which matter in terms of what determines whether you're in agriculture or not. So education reduces whether you're in agriculture if you have more land per capita people tend to stay long. So nothing surprising. These are sort of engendered. In this case, this sort of feminization of agriculture. Not quite the feminization, but what you do see here is that male are more involved in agriculture than female, which is sort of consistent with some other work we have done here. I wouldn't call it feminization. So this is not a sign of feminization. It's not whether women are now less involved in agriculture that there's a decline. But at the level level, male are definitely contributing more controlling for all these factors to agriculture, at least in terms of number of hours work than female. So there's less sort of in terms of employment in agriculture, less uniform patterns in terms of whether one's close to the market, the agroecological zones, et cetera. So it's really this education and farm size, which again makes sense. So if you kind of then try to see whether what sort of contributing, what sort of explaining this age specific effect, and basically we kind of decompose that. And what you find is that, as I said, the age specific attributes is education and gender, which sort of explain whatever age effect you find. So if youth is declining, if youth is leaving agriculture, it has a lot to do with age, sorry, it has a lot to do with education, and there's also some links with gender. But so it's nothing like there is something specific about being young that you leave agriculture. So then I think the next question sort of a corollance and some of you were in a previous session on the agriculture technology. So there's a bit of repetition here, but maybe another audience. So given that I still have a little bit of time, I'll go walk through it. So okay, where in agriculture might people then be employed? So James has mentioned the food system where basically I have on the farm, which is still the majority, but then there's also a good chunk of the farm. So I think David Shirley and his colleagues have done some interesting work trying to sort of in an accounting sense, trying to get a handle of it. Most recent work here is looking at that within agriculture. So across which crops can we expect the people potentially to be employed? What crops would they be producing? So basically what we have here is, so what they did is they took the Tanzania household survey and from that they kind of estimated how much of the different foods people would be demanding assuming one year ahead, assuming the income growth which has been observed. So they basically estimated the income elasticities of the different staples. Mace, rice, pulses, as well as fruits and vegetables, et cetera. Then they also looked at, if you now look at the production patterns and how much labor one needs to produce one kilogram of each of these different crops. So you could then sort of map the two and say this is the expected demand. We assume that imports shares remain the same. So we just project one year ahead. What type, where will the people, how many labor days would be needed to cater to that demand and across which crops? And then on top of that, they kind of overlaid this with, if you look at these employment or labor outputs by farm size, so assume kind of you keep the same shares that the different farm sizes contribute, how much would then be, how much employment would be created in a large farm, in a small farm, et cetera. So that's sort of how they laid this out. It's an accounting exercise, but it's sort of useful to see there may be areas with high labor productivity. So if there is a growth, think fruits and vegetables. There is what you see here. Fruits and vegetables have high labor productivity, meaning few labor days per kilogram of output. So if you take the reverse, that means that you get a lot of output per labor day. Many of the staples, so they have lower labor productivity, rises sort of in between. This is for one hectare. And here you have it for, this is the average. So if you compare these two, you also see that as farm sizes increase, the labor productivity increases. There's also more capital intensive production patterns. So it's actually what you would expect. This is just simple productivity, labor productivity, sort of it's not total factor productivity or basically if you use capital, that sort of is attributed now to your labor productivity in this sort of simple calculation. So what that means is if, for example, incomes increase, there is dietary diversity. So people move, eat a bit more fruits and vegetables. But if you have high labor productivity, you will proportionately employ less people because you can actually produce a lot with one person. If you have a crop where you say, look, you need a lot of people to actually produce the crop. So if you get a volume, then you'll need more labor to meet that volume. And sort of these dynamics give you some interesting patterns. Maybe one last thing, sort of you see that a lot of these things should be moved up here. But a lot of the staples is sort of most of them, these are the shares of the total production. So the total production for wheat and rice and how much is produced by these different farm sizes. So the staples are produced mostly on the smaller farms. So oil seeds, this one here, that's basically produced more on slightly larger farms. And cash crops are sort of the medium size they're in between here. So with that, they then kind of expand or map it into the number of labor days which are associated. So the next year is the amount of additional food in each of these food crops that will be demanded by the population. And what you see here is that it results, especially wheat and rice, that's where most labor days will be needed. Assuming the production technology remains the same. You see that these don't create, so these other grains, the other sort of other staples except from wheat and rice, they don't really produce so much for additional income per grower. So they need to create more jobs but it's not like you're gonna get more income per job. So the key point, one of the key points they take away is on the one hand, vegetables, these are the number of man-hours created. They gain a lot per grower but it's sort of a small slice of the farmer groups of the farmers who will benefit from that. Think poverty reduction, then you would try to get many people to benefit even though it's not huge amounts of benefits but sort of gradually increasing in their incomes. And it's sort of what you get out of this simulation the key point that staples will be important for absorbing that labor. Now if you change the production structure and sort of say, look, let's assume, and they did some scenario simulations that we shift that slightly across farm sizes that slightly more of the demand will be catered to or will be produced by larger farmers than very rapidly this additional labor demand disappears because they're gonna use more machines, et cetera. So you kind of start to get a handle on what crops are important, where will people be employed and what then also to invest in can give you some signs. So to conclude, is African youth leaving agriculture? Clearly youth engagement in agriculture has dramatically inclined and if you just look at that in a cross-sectional sense then you would say, oh, people are leaving agriculture on mass, this is a big problem. But according to the data, so have adults. So in a way, this comes back to the first question which James asked, is this more structural transformation issue sort of in general, if you get agriculture to be more productive in your poor paying then that may be the issue much more than whether it's youth or older people who are leaving agriculture. And as an interesting anecdote, it seems sort of just got some, it's not proof, but sort of an interesting observation that the fact that in Ethiopia one observes that a number of people that find high turnover in some of the manufacturing jobs, you also see that actually earnings in agriculture have gone up quite a bit and that the advantages of working in one versus the other may have become less. So that's why sort of, as I said, it's an anecdote. So I would argue no clear signs of an accelerated or an excessive decline in youth engagement in agriculture today, at least based on these data. And that the age correlated attributes are more important than the fact that you're young or not young itself. And then finally as a last point that where people will be employed within agriculture that staple crops will remain an important place. Actually staple crops on smallholder farms will remain an important place for employment. Thank you.