 So this is the Burkina Faso case study, and that's joint work with Gordon sitting here and Odo. You can see from the title, it's a paper about demographic forces. I'm not a population pessimist, I say it upfront, but there's a demographic issue here in this country case study. So let me start with this figure. What you see here is the evolution of GDP per capita, which turns between 1994 and 2011. You see that of all, that's quite a steady, sustained growth, right? There's, I mean, pushed in the beginning by the devaluation of the CFA Frank, and then there are few interruptions due to droughts, good prices and so on. But in total, it's quite sustained. So you would expect quite substantial poverty reduction. So if you look at the poverty head counter, these are these red boxes. You see that in the beginning it's increasing. Well, that's mainly due to this drought, because the 1998 survey was done right after this drought. So it's really the effect of that. Then to 2003, it's declining quite substantially, but also not too impressive. I mean, it's, after all, it's almost 10 years. And then if you look further to 2009, you see a mild decrease in poverty. So compared to GDP, that's not so impressive. And indeed, if you compute the gross elasticity of poverty, you find a value of minus 0.5. So this is, say, by international standards, this is not a lot. There's, of course, a huge variation across countries in this elasticity, but we are definitely at the lower tail of that distribution. And so the question is a bit, why do we see this low elasticity? So what I do now is not so much explaining how we computed these poverty figures. So if you're interested in that, then I invite you to have a look into the paper. But I try now to explain, you know, this development we see here. So let me start with the story up front, and then I try to provide the empirical evidence or to convince you of the things I put here on the slide. So what we argue is that economic growth mainly had two sources. The first is a massive migration from rural areas to urban areas. So people leaving the countryside and entering the urban neighborhoods, mainly the informal sector, all of the informal sector. So it's a compositional effect. And the second one is agriculture growth. And that's mainly driven by land expansion, you would see. It's not so much due to a productivity increase, but I, again, I get back to this. Okay, what we say is somehow both sources are not really sustainable, right? You would somehow expect some sort of structural shift, but that's not what we see, right? And given this stagnation and almost no increase in productivity in the agriculture sector, we experience quite a tremendous increase in food prices. So the gap between the consumer price index and the prices of the main food crops that are consumed by households, and in particular, of all the poor households, is increasing over time. And that has quite a significant impact on the purchasing power of the poor. And you would see that this food price inflation is even so important that we have some suggestive evidence, it's certainly not causal evidence, but suggestive evidence that there are health effects. We see even over some periods, increasing child mortality. Good, so let me start with sectoral growth. So this is sectoral GDP in the aggregate, so not in per capita terms, for the three main sectors, primary, secondary and tertiary. You see growth in these three, in particular in the tertiary sector. And also paid in the primary and secondary. But of course what we are interested in is how does it look like in per capita terms. So what we assume here is that, well, the primary sector that's, of course, first of all, rural and the secondary and tertiary sector that is first of all, urban. So you see the development of the population over time in, I mean, on the national level, that is the last line, and rural and urban merits. So we have an average population growth rate of about 2% in rural areas. But in urban areas, we have, particularly more recently, we have growth rates of 7%. So also fertility is, of course, higher in rural areas, but given this high migration, you have a huge increase of the population in urban merits. So we compute now the per capita terms, blind these growth rates for the primary and secondary and tertiary sector, and that's a picture of what you get. So there's no growth anymore in the secondary and tertiary sector. It's even, particularly in the secondary sector, it's rather declining. And there's a little bit of growth in the primary sector, and you can see that better here. So this is, somehow, in terms of growth rates. So you see growth in the agricultural sector, but stagnation, if not negative, grows in these two other sectors. Yeah, so that means, I mean, these people shifted to the urban areas, and of course those who made the shift, they increased their income, but those who have always been there, there's no change in their income. Well, it's really a stagnation. Here you see the urban patterns, yeah, that's confirming that again. If you look at the urban area, there's no ingress in the public sector and no formal wage employment, yeah. It's all about, still, I mean, over time in the informal sector, in the senior areas there's, of course, also a bit of agriculture. And in the rural area, it's, of course, first of all, agriculture, and there's a bit of switching, you know, between food crop and cattle crop production that is very much tied to the development of the producer price for corn. It's relatively easy in fact to switch to cotton production. But it's also declined more recently because the price is not any more so interesting. Good. So let me go to the agriculture sector. So I said there was at least some growth in the agriculture sector. So what has been the source of this? So here you see data from a very nice agriculture survey that is done every year in Lopina Faso. If you look here over time, you see that the production that you have in the middle, that is first of all due to an expansion of land. Yeah, and there's only very little increase in land productivity. It's about one percent, the annual rate in the food crop sector and in the cotton sector, it's also just one percent. But the land expansion is 7.6 percent. So all the clothes is generated through an expansion of land. So again, this is not really the standard. So here's the problem. We need some improvement in technology to increase that productivity. And you see that it's hitting limits because land is becoming scarce or the land quality is deteriorating because there was a lot of use of fertilizer in the cotton sector. You see the cotton price that was very interesting after the devaluation in the late 90s and early 2000s but then said this is not anymore so interesting than it was the time before. Good. So what about food prices? So I said there is this lack of productivity increase and all this leads to a steady increase in food prices. So look at these figures. The bold line is the consumer price index and then you have the three main food crops consumed by households and mice. And you can see I mean it's of course quite volatile but over time if you look at the trend line this gap between the CPI and the prices of the food crops is really increasing every year so it becomes more and more expensive to pay for your food consumption. And you see around the CPI the price of rice so you could argue well is that not a substitute? No it's not but the price per calories is far too high so people can't substitute to rice at least not in rural areas. So what does this food price inflation do? Well it increases somehow the share of income that has to be spent on food crops. Again households have very limited possibilities to substitute and there's also no effect through the production side because most of these households are net consumers not net producers. So there's no benefit that comes from some other selling series on the market. Very little. So here's an exercise we have done in the paper on the JDE where we somehow applied this data value decomposition where you decompose a change in poverty into a gross component and a redistribution component and we added a third component we call it the poverty line component Murray presented this yesterday for those who attended the session. So what you see here is the poverty effect of a differential in the inflation of the poverty line relative to the CPI. So it comes from a change in relative prices and the fact that the poor consume a different bottle than the urban household that is underlying the CPI. And if you look over this period in 1994 to 2003 you see that gross redistribution had reduced poverty but then there is this huge poverty line effect. So a large part of the potential poverty reduction was offset by an increase of the prices that are relevant to the poor and that is continuing to continue. What I said in the beginning that we even had some evidence that this has health effects and that's somehow what you see here so that's computed with demographic and health service so we have wasting, stunting of infant mortality and under 5 mortality and for all these figures you see even an increase in the 90s whereas in many other countries at the time it started to decline and only very modest decline in the 2000s. So it could be linked to this increase in food prices. It's more and more expensive somehow to make the basic needs in terms of food. So this is really important. So I come to my conclusion so if you look over this entire period then somehow we see we have this doubling of the population size between 1985 and 2010. We have today 6.5 million people below the poverty line that is even a million more than in 1994 and what we argue is if you do not see very soon a structural shift in this economy so a higher productivity in the agricultural sector and some change also in the urban labor market with higher productivity drops then it will be very hard not only to feed the population but also somehow to sow these 0.3 to 0.5 million men in government that enter each year for the labor force. So there's really something wrong. Of course the next question is so why had this structure transformation not taken place? Is it political economy? There are a lot of donors on Bokina and they all intervene and somehow try to have influence from the priorities. It is a lack of commitment so these are all interesting questions probably hard to generalize but this is somehow the next step I think. And then I have a final slide sort of a reaction to these two famous papers by Pinkowski and Salai Grotin on the one hand and then the young paper on the other. So what would they have found with their methodology in the Bokina Faso case? Would they have found the same or something different? Well for the Pinkowski and Salai Grotin case I mean they work with GDP per capita and then they take the distribution out of the genie. So they would have found huge reduction in poverty. Why? Because there was a cross of GDP in the aggregate and if we use the say the nominal genie coefficient so enduring that relative price has changed. Because the genie is invariant to the general inflation rate. You would have gotten a very significant poverty reduction. What about Jan? Jan relies on the asset index. Well assets are also increasing here but what he assumes is that the income elasticity of asset demand is constant. Well if you have that significant change in the relative price it's very unlikely that this elasticity is constant, right? You would have found something different and I think both would not be credible. That's it. Thank you.