 We are trying to assess the impact of growth on poverty and we are using the composition analysis and this is the method that was suggested by Latin Ravalion but generally the problem that we want to investigate in this one is the assumed negative relationship between growth and poverty has not taken place in Tanzania. We are seeing that economic growth is increasing but at the same time the reduction in poverty is not as people have expected it and that's what has motivated us to undertake this study. In other words we are trying to decompose the change in poverty into two key aspects. These are the growth effects and the redistribution effects to see how much each of this has contributed to the change in poverty and that's mainly the problem that we want to investigate. So the main objective of this undertaking is to assess the impact of growth on poverty in Tanzania and how exactly do we do that? It's by quantifying the contribution of growth into the poverty change and the contribution of inequality. And the methodology as I said we are using or we are following what has been suggested by Latin Ravalion 2002 and that is called decomposition analysis. In other words we are decomposing the change in poverty let's say in this period and the period before and we decompose it into three main elements. One is the growth effect the other one is the composition of the redistribution effects and then we have what we call the residual. In other words this methodology is like you are answering two main questions. For example you ask the first question is what would have happened to poverty if you hold let's say in a quality constant and you change the growth you change the income, the average income and again the second one is like again you ask the other way around now what would have been the poverty levels if the distribution of income had changed without the change in the main income. So this is basically the methodology and the measure of poverty that we are using is what has been suggested by what you call the EFGT that means it is encompassing the head count poverty, the poverty gap and the severity poverty and this is the FGT measure that we use to decompose to see in any given change in poverty how much of it will be contributed by the growth effect how much will be contributed by the redistribution and then also netting out the residual. So for this particular poster we only concentrated on the first measure of poverty and that is the head count poverty and we have decomposed the change in head count poverty by using the basic needs poverty line which is a bit higher than the food poverty line and as I said before the findings shows that for example when we use the basic needs poverty line and it is the same question that we are asking for example what would have happened if what would have happened to the poverty levels if growth has changed while we keep the average income while we keep the inequality constant inequality as measured by the Lorenz curve and you can see this the first finding shows that if really inequality was constant and then it's only growth that has changed then you are saying that the poverty index would have decreased by 29.7% by using the HBS the first dataset that is the HBS 2001 and also the same change would have been 44.94% if we are using the HBS 2007 as a base year but the actual statistics shows that the change in poverty is only 2.23% and again if now we ask the other question for example it's only the growth that has changed but at the same time the inequality has remained constant then we are saying that poverty would have increased by 42.7% for the base year 1 if we are using the HBS 2001 and also it would have increased poverty by 27.04% if we are using the HBS 2007 as a base year I mean the composition analysis on the change of poverty we are using 2 datasets the first one is the HBS household budget survey that was undertaken in 2000 and then we have the second one that was undertaken in 2007 so that's why we are trying to decompose the change in poverty using these 2 datasets if you look at the when we are using the food poverty line again we are almost seeing the same figures that if really the inequality was constant then the change in poverty would have been the change in poverty contributed by growth effects would have been 16.8% for the first dataset and then it would have been 42.87% if we are using the second dataset which is quite different from what is stipulated by the actual statistics and therefore generally what we are saying is as a conclusion is that we really said that in Tanzania inequality has hardly changed we have constant inequality over the time but if that was the constant inequality was there then that means that we should have seen such a reduction in poverty which has not taken place and therefore our general conclusion is either we need to relook at the inequality statistics that has been given because they are not giving us the exact estimates that are given by the decomposition analysis and in general lastly we simply say that our findings here really emphasise the role of inequality when it comes to poverty reduction if inequality would have been controlled then it would have much more effects when it comes to poverty reduction so generally that is the presentation on this particular post