 Hi everyone, my name is Yuxi Wuyan and I'm an Associate Professor of Economics from the University of Tulsa in Tulsa, Oklahoma, the United States. My co-author is Professor Eric Thorbacke. He has been a mentor of mine since my days as a PhD student at Cornell. So our paper is really just to explore the poverty growth dynamics. So there is a vast empirical literature exploring the link from poverty, from growth to poverty reduction. I mean, depending on the regression analysis frameworks adopted, we have at least two strands of substrands of literature exploring the empirical links from growth to poverty reduction. So adopting a regression framework, most notably Dollar Crate 2002, and they have an update in 2016 where they explored a number of world economies and then they found out that the income of the poor grows at the same rate as the overall average income in countries studied, you know, from in two studies covering different time periods. The most recent follow-up covered all the way from 1990s all the way to 2016. Now in that strand of literature, they regress poverty reduction rate on economic growth rate and a number of other initial conditions and control variables deemed relevant for poverty reduction. Another strand of literature is Professor Thorb mentioned, the analytical framework based on the identity model. First way back to the debt and rebellion in 1992, Kequani 1993, and then popularized by Professor Boogie Nhan in 2003 where he proposed an analytical identity where poverty reduction is due to a change in main income and also a change in distribution, the so-called identity approach. Now following using that approach, we also have a lot of literature mainly by Professor Bosu and some others where both of the two strands of literature despite of the different approaches they used came to the conclusion that economic growth is of central importance in terms of in words of dollar and credit is of central importance for poverty reduction. Now however, two things, so abundant literature on the link from growth to poverty reduction but we do not have as many, nearly as many empirical or research on the reversal link from poverty reduction to economic growth. We have some conceptual work way back in 2006, World Bank Perry and his co-authors and then Professor Nisaki and Thorbaka under the auspices of UNU wider had a number of studies also trying to explore the reversal link from poverty reduction to income growth where you can imagine a number of channels through which high initial poverty and slow poverty reduction rate might hinder or dull subsequent economic growth. Empirical wise, we have some more recent literature low peasant seven and Professor Revelli and also the book by Professor McKay and Thorb which you just mentioned, 16 countries in Sub-Saharan Africa, 75% of the population where they painted a more nuanced picture and then highlighted that growth remains important but then we might also have a potential where growth is not always effective in reducing poverty and then on the other hand you know poverty reduction could be significant in some of the countries even though the economic growth is not significant. So more recently Professor Thorbaka and I and also Dr. Shumelis who you know is not able to be here today, we co-authored a couple paper also trying to explore the reversal link from poverty and poverty reduction to economic growth. Now as I already mentioned these studies suggested a more nuanced growth poverty dynamic that asks for more exploration and more empirical studies. So the research questions that we have had in this one research are the following. First of all we want to ask does poverty, we look at both the initial level of poverty and also poverty reduction rate, we wonder does poverty affect growth, do we have empirical evidence, we have had some not many not much and we also could conceptually imagine poverty having an impact on subsequent growth but then do we have empirical evidence that's question number one where we adopted this model which is actually really based on the identity approach where you have poverty reduction rate as a function of the income growth rate and then you have initial, oh this is a typo, income growth rate, the change in inequality interacting with the initial inequality and also interacting with initial relative income and then also related to the change in inequality and its interaction terms with the initial level of inequality and also initial relative income. So this is the model that we adopted for this one research being aware of its limitations. The second question we are asking is really the abundant literature is already supporting but we want to just make sure. So does faster economic growth bring about faster poverty reduction? So we look at both from income growth to poverty reduction and also from poverty reduction and initial poverty to income growth. Finally we compare these results between sub-Saharan African countries and also the entire developing world as a whole to see if there is a difference in terms of the poverty growth dynamics in Africa versus in the entire developing world as a whole. So the data we have to rely on the only international comparison has to come from the World Bank. So they have some 1600 unique country year observations from the surveys that they conducted in different sub-periods between 1981 and 2018. These are panel data, however, they are highly unbalanced so much so that they do not allow us to run really panel regression. We tried to fix the effects, we tried the dynamic generalized method of moments and the results are highly sensitive to even just a little tweak of the data code or to even just a little bit change of specification. So we end of the day, we had to resort to, despite of being aware of the parallel of cross-sectional analysis, we had to resort to a cross-sectional analysis where we reduced the highly unbalanced World Bank panel data into a cross-sectional data consisting of 129 what we call less developed countries or developing countries including 1444 in sub-Saharan Africa from the same period of 1981 to 2008. Now each country obviously has only one growth spell and one spell during which poverty reduces which is consisting of the very first survey year and the very last one we look at the annual average rates. This is a compromise because of data limitation and we very much look forward to, you know, learn about potential fixes of the thing. Now I'll skip all these two because I'm aware I only have 15 minutes. This is the entire sample which we call all the sample including 129 countries from the developing world including 1444 from sub-Saharan Africa. The average growth spell is 21 years for the entire sample even though we have some countries which are, you know, having much longer or much shorter growth spell. And for sub-Saharan Africa the growth spell is roughly 18 years. Now we also, because of how we choose our regression procedure, we also have a so-called GMM sample where we ran, you know, generalized method of moments regressions which cause for not just two but three time periods. You need, you know, the very first and the very last and then the middle one. The middle survey year and the final survey year will make our growth and reduction, poverty reduction spell but then, you know, the middle one will be IV instrumented by the very first year. So that's how you have generalized method of moments. Allow me to jump. It's really just, you know, essentially this is just a standard IV estimator. The thing is it allows, it improves the efficiency of a standard IV when you have the presence of heteroscedasticity of unknown form. So that's the two regression procedures we have used for this analysis. I mean, it's really limited by the data that we have in hand. And of course cross-sectional analysis really does not allow for any, you know, satisfactory causal identification. And I'm looking forward to suggestions for improvement. And also Professor Tarp in the beginning, his analysis of the analytical framework and its limitations already gave me some ideas. Now the findings here are, first of all, I mean, those are extremely small, I apologize, but then the idea here is we do find some empirical evidence suggesting that faster poverty reduction does lead to faster subsequent growth for the entire developing world, but especially in Sub-Saharan Africa. So faster poverty reduction, the impact on growth in Sub-Saharan Africa is roughly, you know, twice of the impact of poverty reduction on growth in the entire developing world as a whole. So this is confirmed with some empirical analysis despite of all the limitations. Now the second finding here is we looked at not only the impact of poverty reduction rate on subsequent growth rate, we also look at the initial level of poverty on growth. And we find that high initial poverty weakens the effectiveness of growth in reducing poverty. So that is consistent with what had already been found in previous literature. So we are able to confirm. So the first two findings are basically answering our questions about the link from poverty to growth. For one, high initial poverty weakens the effectiveness of growth in reducing poverty. For another, faster poverty reduction rate to begin with has a positive impact on subsequent growth, especially in Sub-Saharan Africa. The third finding is really just to confirm what we have already learned. We have a little bit of an interesting nuance. So in general, we have been able to confirm that faster economic growth does bring about faster poverty reduction. However, it's not so much in Sub-Saharan Africa. So in the developing world as a whole, yes indeed, we have faster growth leads to faster poverty reduction. However, in Sub-Saharan Africa, poverty reduction is not responding to growth as much perhaps related to what Professor Tarp just mentioned. Some people, you know, they improved, but they still do not cross the line. So here we have only presented OLS results, but GMM estimates are consistent and it was omitted for parts and many. Now the last finding is we are able to, because we used the identity model, so we are able to estimate the effectiveness of growth in reducing poverty, considering the initial conditions in terms of initial relative income and also initial genie. We are able to estimate the growth elasticity for poverty reduction, where we noticed that the growth elasticity for the entire developing world as a whole during the studied period is roughly two, whereas it's only half, I mean 1.14 in Sub-Saharan Africa, again confirming what we have already you know found out in the previous finding. I mean confirming that. So the implications therefore is, so for one, I mean I will be short, the implication is we probably should also focus on poverty reduction in the first place. By attacking poverty in the first place, we might be able to improve subsequent growth which would in turn, you know, further reduce poverty more effectively and therefore you have a virtuous circle or spiral take-holds and that's the title of the article. That's all we have. Thank you so much. I hope I stay within the time limit.