Dear Shanta, I read your blog entry on agriculture and found it very interesting. But if we go even further, would agriculture productivity increase significantly, would poverty reduce significantly? I am not really sure. Indeed, there are growing evidence that in rural areas, people go out from poverty by working outside agriculture. Most farms in Africa are around 1 hectare for at least 6-10 people. Let's assume production would be multiplied by 4-5 times, would it be enough to go out from poverty (assuming farmers would sell at the "town or capital" price despite the fact they would not fully load a truck)? In most cases, when you produce cash crop with a low ratio value per kilo, it may be actually difficult to exit poverty. You most probably recall S.Dercon's exercise of tracking people from rural areas in Tanzania and what he found out is that the best way to go out from poverty is to... move/migrate. Then, is it really worth to subsidize agriculture when we know that no matter the amount of investments, they will probably not move out of poverty (unless land consolidation/land reform happens)? As long as agricultural production and productivity do not increase dramatically change (large plots for a small numer of farmers), poverty could probably prevail in many areas in SSA. It may be difficult to transplant the trend, which happened in Europe or in the US but, like you know, more than 50 years ago, employment agriculture, for example, in France, was more than 20% with average plots of a couple of hectares and limited mechanization ; now, this employment is around 5% with much larger plots and mechanized agriculture (it has probably some impact on environment). But, if the objective is poverty reduction, increased productivity on a one hectare plot may not be the solution... education for migration (to small towns/cities) may be more appropriate. Would not the WDR 2009 on economic geography be the answer to the WDR 2008?... To fight poverty, the best way could be to move out from agriculture by moving at some stage to small agglomerations? Agriculture is not an "economic" activity as such; it is usually the way to survive by prodiving food in rural areas. That is also why, the surplus to sell will never be large enough on a one hectare plot because the first function of the plot is to provide food (and sell the not-consumed part). We increasingly see that when people in rural areas start to combine food (from the plot) and incomes coming from other activities, such as services, it is how they are more prone to move out from poverty.