 I'm Tuy Lecturer from the Diplomatic Academy of Vietnam. I'd like to put forward a question to Dr. Sam Jones, please. So Dr. Jones, with your expertise and experience, I would like to seek your view. Would you support the view that when a class is overcrowded, if we are halving the class size and reducing the syllabus and the teaching hours by the same proportion, we could actually achieve the learning efficiency in terms of skills and knowledge, acquirement, retainment, applicability, and utilization. Thank you. Heady behind you? Yeah, I'm sorry. I have a question for the last speaker because I was elsewhere. I'm sorry about the others. This question really raises the issue of whether we are measuring. I mean, as much as I enjoyed your presentation, I myself have been searching for these kinds of solutions. Your point about this high association between total factor productivity and labor productivity, I believe that makes sense because Danny Roderick makes the point, capital accumulation is the basis of that in East Asia, now that you've included India. But my concern is something that I used to discuss with Sanjay Lal himself, are two things in relation to the use other classifications by users, rising stars, falling stars, loss of opportunity, and so on, from the same classifications. Now, when you had this graph on TFP on one side, the level, and then technological sophistication, one of the problems with trade data, I believe, and I think this is recognized by the late Sanjay Lal, is the fact that we are classifying anyone who exports an item like microchips. And these firms could be really very low-end firms, like the ones in Philippines. My feeling is because you have Malaysia on top of the curve that you have and Korea below, my problem is it's probably taking care of that, missing something. Simply because Korea, which had a poor capital income, last standard of Malaysia in 1969, only in 79, it overtook and came down a bit because Park Jung-hee was shorted in 79, 80, it got back, and now it's about four times higher, with 50 million people against 30 million people. I'm not sure whether it's telling the same thing, simply because trade data tends to give us the whole thing. The link I have with that is, say in a field that I said it could extensively, and I think here value chains help a lot, the drivers. If you're dealing with, say, consumer electronics, or even industrial electronics, they are far less technology-intensive than microchips, and that microchip is a component. So that's why we tend to have this notion that countries that import components, the logical argument we make is, they are in low-end sort of areas, and therefore the ones that assemble finished goods. Can I assembly? Yes, I agree, the drivers are the producers. So I'm just wondering whether you might take into account these things. Thank you. My first question to Mr. Wu has been more or less asked already, because I was wondering why Mauritius was so high up, South Korea so much down, I said, what is happening? So maybe the answer is here, but you respond. Secondly, Mr. Wu, I wonder whether you could interrogate the conclusions you make, which I think are very good, but many of them would require interrogating the TFP concept itself, because you referred to many processes at work, which I thought maybe either in further research or whether you have already done it, whether you have linked these conclusions to the process at work, over and beyond TFP. Secondly, to Professor Nobe. Yes, it is quite interesting to hear that Kaizen is actually from below, but now I wanted you to reflect on the fact that in Singapore, Rikwai-Yu came from above. In Ethiopia, it's coming from above. In Tanzania, we see the Kaizen unit is not expanding enough. Could it be that in the initial stage, initiative must come from above? I leave that to you to reflect on. To Sam Jones, people who ask questions and when they missed the presentations, I'll risk to comment on what you said, even if I'm not here, judging from what we have discussed before. For the case of Tanzania, we find quite variations within East Africa, by countries, and I think there are good reasons for that, I think. Now, the question is whether you have examined variations within the country. We have large variations of performance between schools in the same environment, and one of the factors which has accounted for that variation is the administration, the quality of the headmaster in making sure that teachers are teaching, hours of learning are observed, and the performance of school boards, ensuring that performance of teachers is actually taking place, and the basic supplies are available in the same environment. So I wonder whether you observed that, to be able to say from the differentials, can we learn something about the possibility of improving learning outcomes without necessarily having large budgetary implications? The last point is does learning outcomes matter? Because I think this is an important area, and in my country, Tanzania, we are moving that direction to try to measure outcomes. But do they matter? Because as you were hinting, learning outcomes in India are quite low. But does it seem to hold back development? So I wonder whether you have done some study to show whether learning outcomes matter? Because only when we say education matters, the measures we have seen are in terms of enrollment, achievement in secondary school, university, but now you are coming with a good concept of learning outcomes. Does it matter? I think we'll have the panel answer these questions, so we'll start with Sam. That's okay. Yeah, okay. Thank you very much for the questions. I'll try and be reasonably brief. So with respect to class size versus syllabus and what effects these can have on learning, I think from talking generally from the literature, I mean, what we find with respect to class size are probably two main findings. So first of all, there's a nonlinear effect of reductions in class size. So with very, very large classes, so we're talking 60 plus students, we do tend to find some small gains in learning outcomes when you reduce the class size. But once you get down to what might be called international norms, so let's say below 40 students per class in a primary school, there's very little evidence that reductions of class size systematically improve learning. And indeed in places like the US, for example, although a lot of money has been placed in reducing class sizes, there's very little evidence that this has been systematically useful in improving test scores. But of course in the context of 90 plus students that you find in some areas of East Africa, I think we would generally consider that to be just too much. So what does matter for learning seems to be teacher quality. That's a very difficult thing to really define, but having high quality teachers, highly motivated teachers with an appropriate incentive structure seems to be far more fundamental than let's just say school inputs that you can count. So that's probably what we do know. So I think that the question about the syllabus is quite interesting actually. I don't know of a lot of literature on that, but what we do know for example in East Africa and many African countries that they face the challenge of multiple local languages. Typically, often the language of instruction is not a language that they speak at home. And even the syllabus in these countries does not support some of these local languages even though they're taught. So for example in somewhere like Kenya, rural primary schools can start teaching in a local language for the first few years of education, but there's no syllabus available whatsoever to support that. So these are questions that I think could be researched to a certain extent, but I don't have any specific information about that. With respect to do learning outcomes matter, I think we can answer that in two ways. First of all, from a human rights perspective, I would say yes, right? I think it's reasonable purely if we see education as a merit good that can be useful in terms of building better citizens, creating dialogue around things in terms of developing information and about health. For example, there's substantial evidence that we need education to support health outcomes and so on and so forth. Does it matter at a macroeconomic level for growth? I think that would be something that one could look at. We do know from a macro standpoint, I mentioned it actually in the presentation, that what matters for growth are test scores and not the quality of, and not the quantity of education. So we do know that from cross-country variation, it'd be interesting to see whether we see that from the microevidence within country variation, but what we find, for example in Africa, we don't really have very good information about, for example, GDP growth at the local level to see that, but that could be something to look at. The last point I'll make is that, yes, we have looked a little bit at variation within regions, but one of the challenges we face with this data is that the information we have at the school level isn't very good quality, but that's something we can look at a bit more. Thank you, Sami, for asking questions. Yes, Ethiopia is very top down, but I hope the spread of Kaizen in Tanzania will be more bottom up. For example, in your country, the hospital sector and the health sector is adapting Kaizen very rapidly, and then people can feel the good impact of the adoption. So I think the awareness may increase from that kind of activities. But at the same time, when the demand is being created, you have to have increased supply. So then the Tanzania Kaizen unit is making great effort, but I think a little bit more push is needed. So two months ago, you helped me a lot to talk to your president, and then your vice president came to my university the next month. So I hope they understand the importance of that kind of human resource development, and then a little more help, because until the market is really created for training and consulting services, some kind of government intervention would be needed. Demand side, the government can boost awareness through the media campaign, but also helping the health sector in the case of your country, and at the same time, the supply side, you have to nurture a greater number of trainers, maybe trainers of trainers training. Yes, thank you very much. We've got 30 minutes. Three minutes, okay. Okay, thank you for the question and comments, but let's first talk about the TFP versus the index of technological sophistication numbers. But then if you recall that how it is actually constructed, then using traded data at the three digital levels. So obviously, for example, China is in terms of a score actually highly ranked like a 3.75, because they tend to export a big chunk of high-tech nature of our products. But there's obviously caveats. I didn't have time to talk about it when you construct the data. But at the three digital level, then even the same high-tech categories, you could actually contain highly sophisticated products such as mobile phones. But at the same time, much, much simpler like some plastic, the phone shell, the case could be also categorized under the same high-tech classification. So obviously you cannot tell the differences in much finer differences. So that's kind of one of the problem. But then this is more the second aspect is much more a significant problem because rapid rise of international production network like a global value chain and on bundling these production stages across the different locations, different countries. Therefore, aggregate the trade statistics doesn't tell much about us to us. What is the process? What production stage has been involved? So obviously there is a great difficulties in precisely capturing the magnitude, either technology or TFP, which I'm gonna turn it in a minute. But I guess that rather than these detailed differences, rather if you look at the pictures and clearly there is a tendency of technology and then this sophistication of the export tend to go together. But then second point though, this value chains. Since I'm documenting more macroeconomic level, the data, so I didn't show anything, but there is a huge amount of micro data evidence pointing into the positive impact, backward linkages or forward linkages within this global value chains. So again, that could be a very important element at the micro level, but since we're focusing on the macro aspect, so there was not maybe, it's not there, but clearly I presume I believe it's happening there. But then the second regarding Mauritius compared to Korea, if you remember how actually we construct the TFP, although including Bob Solo and then many other economists believe TFP is a good measure of technological progress, but in the end it is obtained as a residuals after you are counting for the contribution from the capital stock, human capital and labor, but then in the case of Korea, if you look at it, human capital measures for instance, then in terms of maybe PhD, I heard being a Korean, I heard that elsewhere though, Korea has a highest number of PhD per capita. So if you take the numbers when you consider the TFP, then much less left to be captured by the TFP, and then Mauritius opposite much low level of capital stock and then human capital, then given the rapid growth we already know and then no one can dispute, so obviously bigger chunk will be attributed to the TFP contribution. So that kind of things I think well known in the literature, so I think that's kind of a cautious you know, the tales we have to think of, but then the final point regarding our processing export. This is much again a bit harder because as I said, even more difficult now because of a rise of international production network, so we have to think about it, what production stage is actually contributing, how much in terms of technological upgrading, but then in the China context, interesting enough, this processing export dominated by the foreign affiliate, foreign invested companies, in other words, FDI. I show you some of the strong evidence at the macro level, FDI and then capital, the equipment import having a very strong impact on the TFP growth. So to the extent that this processing export also engaged by these foreign affiliates, then on top of this value chains, backward link with this and forward link with this positive impact on the TFP growth, but on top of that, clearly this might have additional impact meanings export process could have technological learning components, but just very quick, yeah, one final observation though in the literature, I guess that they could not solve the issues whether you simply export more than you learn more or simply because you are actually better producing something, therefore are you becoming an exporter and my sense of the reading the recent literature is actually pointing into the second, meaning in general highly productive companies, they tend to become exporter rather than simply trying the export and an export and you becoming better not because of it, thank you. Sorry we're late for coffee but we'll just give the speakers a round of applause and conclude the session. So thank you all for your contributions.