 So, we probably have about 20 minutes or so for questions, so can I invite people to make concise questions and before the question, can you just tell us, can you just introduce yourself briefly? Thanks, Paul Mosley, University of Sheffield. I'd like to invite Arnor in the first instance to maybe develop his answer to the question what we need to do, but I'd also like to ask all the presenters if they would mind doing the same thing in the light of the differences which the book reveals between the countries which managed to reduce poverty a lot and the ones which didn't. And if they will to do that in the light of the fact that Africa now is very distinctly the region of the world where poverty reduction performance is worst, and possibly to be provocative I could suggest a simple hypothesis which some people think explains this huge difference between Africa and the rest of the world and indeed between different regions of Africa which is that the ones which did reduce poverty did something serious about the development of smallholder agriculture and the ones which didn't were not serious in their smallholder agriculture policies and as a consequence weren't able to increase yields and therefore and that spread to the rest of the economy and therefore poverty didn't reduce. Okay, thank you. Yuka and then Eric. Yes, I'm Yuka Birkhtila from University of Helsinki at UNU Weider. Thank you for the very nice presentations. I have a couple of questions first to David on the, and I would like to ask why your numbers on poverty differ a bit from those reported by the World Bank, can you give us some pointers on the reasons for the difference, and I know I should know this because I've seen the code, but I don't remember now that I mean what was it exactly that the utility consistency did for the consumption basket, was it in fact something similar to what the Rwandan statistical authority I believe then did on purpose when they changed the basket to contain items that have a cheap, no, okay, so my mistake then remind me, remind me maybe somebody else in the audience on what it actually did, and then I very much liked them, Marike, your talk and the idea of using these mixed methods. I'd like to follow up on Arnes' point on whether we, is there a reason to worry about the reliability of the findings in the, for example, DHS surveys? David, if I understood it, you use some kind of a leased cost bundle. So you start with a caloric requirements, let's say 2,000 calories per day, per male equivalent or what have you, and then you look at the leased cost diet that gets you the 2,000. So my question is the actual diet that is consumed by the households may be very different from the leased cost diet. So you did say something about utility consistent, so I'd like some clarification at this point. Secondly, I was very surprised at the very low genie coefficient for Ethiopia. I mean it's in the order of Scandinavia, it's much lower than the US, much lower than the UK, much lower than most European countries. Is this real? I mean if it is, it's a very good sign, but I raise this as an issue. Third point, and this has implications for AERC. The last date that you have is 2011. We are now in 2018. In the meantime, there's been steady growth in Ethiopia. The conditions now I would speculate are much better than they were in 2011. Now I understand you had no choice because the 216 survey was not yet available. But I think it has implications for AERC. If AERC could try to speed up the analysis of existing surveys, facilitate the analysis of existing surveys, and perhaps convince countries in the World Bank to conduct more surveys, this would give us a closer picture of the present situation. So we don't always have to deal with past trends. Then for Mareke, the more she talked, the more I was reminded of the, what I call, the Heisenberg Bohr principle in quantum physics, that if you observed a phenomenon, you essentially distort the phenomenon. This is true in quantum physics. And I don't know if there's an answer to it, but clearly with a questionnaire, if you ask a household a number of questions, they may give you answers either to please you or to avoid having to pay taxes. You don't necessarily get a true picture of reality. And I wondered if perhaps you had some ways of not solving this problem, but of making it less acute. Thank you. We could take in a couple of questions if people have lemma. And yes, look, yeah, lemma. Just a couple of questions on Rwanda, the qualitative field work. I was just wondering, how does one go about validating qualitative field work? That's one question. The second is, you know, the voice and accountability came up in Rwanda. And that's also endemic to Ethiopia. You didn't mention that. And I was just wondering what the role of this state led kind of guided development would have. And also now we're actually going through some changes, right? So I think if you can actually opine on that, that would be useful. And I was even thinking about the data, 2005 shock. Have you seen any like changes as a result of the 2005 shock? I was actually looking at some numbers. And I think where I could see is some kind of a disaggregate regions. You know, I was kind of impressed to see the kind of time variation in the Tigra, Orban, the rural, and then Oromia, you know, and then the current movement actually was in a state mostly by the Oromia. So I think the political economy dimensions need to be looked at. And the question that was raised in Rwanda is just sustainable. And that's the same question that applies with Ethiopia as well. Thank you, Luc Christiansen from the World Bank. Just following up on Lemma, maybe one change you saw in the data is which struck me when they did this aggregation by the regions is the fact that poverty went down a lot in Addis between 2000 to 2005 from 30 to 10 percent or something. And then it stayed at 10 percent after 2005. What was one of the major criticisms of the during the elections in 2005 was that they're all catering to Addis. So the government must have taken some of that message and kind of went even further away or invested more in the rural areas. But that doesn't aside. Second point to me, an obvious thing to sort of cross check would be to see what's happening in agriculture between 2011 and 2013. Now, I noted that on your rainfall on average, there was a little bit lower. That may not be enough to explain it, but sort of to look at the consistency. It seems to me it's an obvious thing to do. What's happening to yields, maybe linking to Eric's point there. You could also use satellite data, some other ways of looking at it and just sort of survey data might be a way to sort of cross check that a bit. Then the flip side, I really like your point, David, at the end. So you say, look, what may have contributed to this poverty reduction in Ethiopia? And what's sort of interesting to see, you're really describing a package. You're talking about technological change. You're talking about agronomic knowledge, extension agents. That's sort of on the production side. Then you talk about infrastructure, basically talking on the market side. So you're kind of working on both sides. And on top of that, you add something on risk, which is very critical for agriculture in order to facilitate some of that adoption. So while that kind of package working on these three fronts may not be coming together all the time in each of these villages, clearly Ethiopia has been working on all these three fronts at the same time. I may want to add sort of a fourth element as a question. We also know that Ethiopia has been investing in trying to attract labor-intensive manufacturing. I'm not so sure how much of that was already in place. I think it started after 2005, so to what extent it was already there in 2011. So you really get this kind of, we're producing, we're increasing staple crop productivity, which is what Ethiopia has done. They may start to come to a point where you can get an accelerated release in a productive way of labor out of agriculture. Then that has to go somewhere. Part of that might be absorbed in the rural non-farm economy, in the secondary towns, sort of to higher domestic consumption, construction, non-traded goods and services, et cetera. But part of it can also be absorbed by the labor-intensive export-oriented manufacturing. Flowers is not, this is like manufacturing. So flowers is one, but there is a leather industry, et cetera. So that kind of other angle of it, is that something you have looked at or sort of to what extent it contributed, et cetera. May that have come a bit after 2011 is sort of maybe more important now. Thank you. I've been wondering in the context of inequality for Ethiopia, whether the case of expatriates has been considered. There are many expatriates now returning to the country as they reach retirement age or otherwise with a better economic situation. Now, they have clearly a very high, much higher level of savings and resources as compared to the general population. Perhaps this has also contributed to inequality, but many of them are also using these funds for investments in small-scale industry. So has there been a consideration of both aspects and their implication for inequality? Thank you. Thank you. You're going to... Thank you. Thank you, Ade. I think, Eric, I take your point. I think it would be very interesting if we were to update this data. For example, I think there has been some surveys in 2018 and maybe you're going to see some surveys coming up. It would be interesting maybe to contrast to see what has really happened. Thank you very much for that. But I think in the Kenyan case, I still didn't get it. Why is it so uninspiring? That was your comment. Because I think, Germano, you could have given us a set of factors. Why is it so uninspiring? And in fact, to tell you the truth, if I go out in Kenyan... I'm a Kenyan, so I go out in different rural parts and I can tell you that the institution is actually desperate. You can see it. So, essentially, you don't even need data to compute that. You can actually collaborate that story about the uninspiring. But there has to be a problem. But that also means, one time in 2007, Kenyan was growing at around 7%. And one of the questions I used to get from the journalist is that, can you feel it? Where is this growth? It's the same question you can't see. If we can show, at least I can... It's evident about Kenyan the way it is. But if you go to Ethiopia, can you also see this kind of change, that pubert reduction? Is it feasible? It got to different parts of Ethiopia. It's the same question I would like to see. Can you see it in terms of those many years of transformative changes and pubert reduction? Then, can you see it? Is it feasible? That's perhaps an interesting question to say. Okay, thank you. If you can be brief, we'll allow one last question. One last question. Yes, my name is Holger Hansen from the University of Copenhagen. It's a very simple comment, I would rather say, on Rwanda. I think we are left a little bit confused about the Rwandan situation. And I would ask, isn't there a space or time or money to do an extra survey and get a new set of data from Rwanda? I think it's needed because it's a very important case of a developmental state and how it performs and so on. So I think we would need to be updated. And then I could ask you, will you equal the data you have from Philip Ranschens with the ones from David Booze and Mutepi? They seem to be very different. And Philip Ranschens is very well known for his, if I may say rather, subjective view on Rwandan. Thank you. Okay, thank you. Okay, let's come to... I mean, I don't want to encroach in our tea break too much. So let's come to responses from the different presenters. Can we go in the same order that you presented? Germano, do you want to respond first? David, then Mareka, and then Arne. Thank you very much, Andy. So just an observation on Kenyan case, why it is not inspiring in terms of, okay, we have seen high growth rate, but very little reduction in poverty, even inequality. Actually, inequality has come down a little bit, but I think the answer is pattern because, okay, in my presentation, I left out some information. The information I left out, which also Andy did not have, is the results from the 2015 survey, which shows that even during the period where growth was not very high between 2008 and 2015, during that period poverty came down by 10% from 46 to 36%. Those are the 6% from the recent survey, 2015. So actually, something has been happening in Kenya, which is positive. The other observation about Kenya is when you walk around in Kenya, and you want to walk around in Ethiopia. Actually, from my perspective, you think that Kenya is doing better than Ethiopia. Okay, also when you go to other countries like our neighboring Tanzania, when you also walk around, you feel the same. I don't talk, I just look. When also in the... So this is qualitative data. When researchers come to Kenya during ARC meetings, they feel that actually Kenya is doing very well. Okay, anyway, part of the reason is an inspiring part. This information was actually missing. A bunch of questions, so I'll do my best here in two or three minutes. Okay, so the first question, what do we do from here? So I'll take my perspective as someone who worked on Madagascar and Ethiopia and contrast the two and have two points that come out of this. The first is political stability. Madagascar has been repeatedly hit by shocks, political shocks, and this has really kept them from moving forward, which gets kind of a question of how is this sustainable for Ethiopia? So that's a valid question, I think. But then the second is what Luke describes as a package of policies that are geared toward infrastructure, enabling agriculture, extension, and the like. And so that's my sense of the importance of agriculture, but also thinking about this as a package. Now for some of the questions about the poverty numbers compared to the World Bank numbers, it's the poverty line. The consumption aggregates are effectively the same in nominal terms. So it gets to the poverty line, which gets to the question of what is the consumption bundle? So the consumption bundle is based on the caloric requirements in each of the spatial domains and based on the consumption patterns in the spatial domains. The confusion that I apparently have led out there was with the revealed preference tests. So the idea is that if relative prices change from a base level, then you would expect households to change some of the composition of the bundle. But don't you use the word lease costs? Does that imply something? Yes, the idea is that lease costs along the utility, along the indifference curve. And so the entropy method is meant to adjust those consumption patterns as little as possible so that you maintain utility consistency. And so I hope that explains that. Low inequality, it strikes me as being quite low as well. Though if you think about the majority of the population are in rural areas, inequality is not that large among the rural population. And I didn't get a chance to talk about it, but it's mostly within region inequality that describes national inequality. Excuse me, between urban and rural. No, it's yes, yes, but most overall, and when you break down the tile inequality index, 80% of that is 85% of that is due to within region inequality. So in terms of the diaspora, your point is certainly a valid one. It's something that we weren't able to identify that with the data, but it's certainly something that the introduction of investment that Arnie was talking about, the diaspora has been an important part of that. My understanding is one. Yeah, yes, thank you. So thank you for all the valid points and the questions. I will try to address each of the questions, but be very brief for the sake of time. So first on land, it's true, Rwanda has been very active on the front of land and agriculture. So they have implemented land consolidation programs, crop specialization programs. They have distributed seeds, fertilizer, implemented land reform. There are some critics who say that this could have been better designed and implemented with more bottom-up strategies rather than top-down. So it's not ideal, but it's probably better than doing nothing, which is the case in, for instance, Congo, a neighbor of Rwanda. Then on the DHS, is it reliable? Well, the DHS method is fixed because it has to be comparable across countries and the implementation is tightly managed by macro-international. So in general, it's considered as reliable. Also, of course, the numbers flowing out of the DHS are rather positive and so would not never be contested by the regime. And I'm not so surprised that indicators for health and education are positive because these are areas where you can quite easily implement top-down all kinds of measures, vaccinations work at all time, at all places. We know that, but in the area of agriculture and livelihoods and economic transformation, it's much more difficult to work top-down and figure out what works for whom and where. So I'm not so surprised that on the educational and health front, Rwanda is performing well because they are well organized. So you can go a long way just by that. And then I got a bunch of questions on the reliability of qualitative information. So quantum physics, yes, I'm also worried about the distortion and the reliability of qualitative information. I feel personally very uncomfortable working with focus groups because there you can easily cherry-pick information flowing out from those focus groups. There is a danger for confirmation bias, certainly because many of the Rwandan scholars have turned into activists as well and are also very negative on that front. So I'm very wary about that kind of information and those studies. Concerning the life histories, I have a bit of a different view. First of all, there are many life histories. The respondents of those life histories were randomly selected in a number of communes in Rwanda. The answers and the narratives of change and the rankings are embedded in the life history. So there is a consistency check. It's not just an answer that you give to a question. It's embedded in a story and so there has to be some consistency. Also, there is an overlapping period in the life histories. The respondents were visited in 2007 and 2011 and in 2011 the respondents were asked to start telling their story in 2000. So we have an overlapping period, 2000, 2007, that we can use to check for instance for recall bias. So although not perfect, I feel more confident working with these life histories. Collecting objective data in Rwanda, yes, I would very much like to do that, but I've been working on Rwanda since 2002. I've collected my own data in Rwanda. Since a couple of years, I'm just passing through Rwanda with a transit visa on my way to Congo, but I would not think of starting a data collection process in Rwanda because it's so tightly controlled and it's impossible to do independent research. So yes, it would be ideal, but it's not very realistic at this point. Arne, did you want to? Just a quick point to Paul Mosley's comment. If you read Andy's summary of what was learned from these studies, his sort of strategic point is that agriculture has been important if you're going to achieve success here. And it reminded me also that I actually was in Taiwan in 1990 and did a paper with Paul Collier on the importance of agriculture for industrial development. There was a whole conference on that, so it's part of the book. And I also talked to the people on the countryside in Rwanda, the administrators and the policymakers there, asked about their poverty strategy. And we never had one, we always focused on productivity. And maybe that's also important. You need strategies, you need policies to implement. And of course, we heard in the morning now still this plan about how much policy we need. That makes me slightly worried. Africa seems to me not deficient in policies, but in implementation of policies. So you can't do too complicated things, but the focus on agriculture I think is an important and good start anyway in this context. And about Eric then. By the way, Eric was the opponent of my thesis as well, he was everywhere. He was worried about the inequality figures in Ethiopia. And luckily now they're up, so they're up at 0.39 according to world development indicators anyway, which is up there with India and these guys. So they are no more, no more unequal than Sweden. Thank you very much, Arne. OK, I think we should go for the tea break. But first of all, let's thank the presenters, discussant and the audience questions.