 It's a challenge to talk about foreign aid at wider. You know from the selectivity that most of the people in the room will know more about aid than you do, or than I do in this case. So I will refrain from giving the great overview that Sam gave and instead give what could be considered a case study. I know that economists hate case studies, but I've recently drifted into interdisciplinary work and I've learned that it actually can provide insights sometimes, even though it is mainly gossip. So I'll try to give you a case study of the debate and how it involves in foreign aid in this discussion of aid. First of all, looking at some numbers, trying to get back here, this is the sexual distribution of aid commitments from the bilateral donors from 1970 to 2010 and you will notice that my scale is off by 100. But this should be in 2013, US dollars and then 100 million dollars. What you should note here is that we have four main sectors and the first sector on economic infrastructure is transport, communication and energy. So we've been talking a lot about this, this is infrastructure building. Justin Lin alluded to this, Durkan alluded to it and also Steve O'Connell talked about the importance of infrastructure, the importance of energy and power in their analysis of the need for aid. We also have production sectors, agriculture, industry mining, construction and then trade and tourism. So these are what's called the economic sectors, as I said. In total, it's about 30 billion US dollars and we have the social infrastructure where we had the largest increase, but in there we see water supply and sanitation. And finally, on top of what I'm not showing, what we have is these multi-sectoral allocations, that's budget support. But what you see from this is that we actually spent a substantial amount of money on capital investment, infrastructure or machines or whatever. As Justin Lin talked about yesterday, the World Bank is also now changing or has been into infrastructure. It's now the single largest business line of the World Bank Group. It represents 43% of total assistance from the group to low and middle income countries. So you could say that infrastructure and by that capital in total is certainly on the agenda in the aid community and the way we give aid. Now, some analysis would then think about what's the return? Sam talked about the return to aid in their simulation study. This is a study by a Casselian fire in quarterly journal of economics in 2007. And what they did was a calibration of the return. This is a broader scope. They're thinking what is actually the return to investment in reproducible capital. So they subtract in some of these natural capital and they correct for some price differences between rich and poor countries. Moving from a very large difference in the return to capital from rich countries and poor countries from 27 here to an 11%, moving down correcting as they do for these differences of natural capital and prices. They find that actually in the group of rich countries the return to capital is about 8%, 8.4%, whereas for poor countries it's lower. And they rationalize virtually all of the cross-country variation in capital per worker without appealing to international capital market frictions. So capital markets are working in that situation. They're simply a lower return to capital in the poor countries relative to the rich countries. In the conclusion, Casselian fire state that as a result increased rate flows to developing countries are unlikely to have much impact on capital stocks and output unless they're accompanied by a return to financial repression and in particular to an effective ban on capital outflows in these countries. Even in that case increased rate flows would be a move towards inefficiency and not increased efficiency in the international allocation of capital. So in that perspective the call for more capital or infrastructure, whatever investment in that perspective this is actually not efficient as such, right? If we consider infrastructure as being part of this broad capital concept and this is what they do in their conclusion. This is a bit interesting because it takes us back the 30 years. It takes us back to the mostly micro-macro paradox because the World Bank in 2003 and 2004 were still producing economic returns, economic rates of returns on their projects. So these are projects in different sectors. Whoa, this goes wrong. These are returns to projects in the different sectors and you will see the economic rate of return in the next final column and then the revised economic rate of return. For different sectors you find very respectable rates of returns but again project-based. So these are the assessments by the World Bank on their projects only the ones where they actually do compute an internal rate of return. When you think of this in terms of country and incomes then we see that in the lower income countries we have a return of about 19% and 20% for lower middle and upper middle income whereas if you go to the high income countries the returns to these projects is much, much lower. So we're back, you can think of in terms of having this paradox where we still find in the early 2000s, still find high returns on our investments, project investments but at the aggregate level we find low returns and capital flight would be induced according to Kasselian Fyre. Basically when you think of what they're doing and what you can't be doing there are three approaches as Kasselian Fyre mentioned to computing aggregate model of product of capital. One is to look at interest rates. This is hampered a bit by other restrictions in the markets. You could also do regression analysis of GDP growth on changes in capital which would be more or less investment but it could be a structure like a growth accounting formulation or you could calibrate based on national account statistics and other data. Kasselian Fyre using the third method by calibration and this is why they have the country by country observations. Together with Carl Johan Delgar he and I have been working on estimating the return to aid investments using regression. So we start out from this growth accounting formulation and allow aid capital to be different from domestic capital if you like. So making this distinction between if you like infrastructure investment funded by aid and machinery or whatever private capital as part of the domestic capital. So we think we allow the marginal products of the two types of capital to differ within countries and then we estimate the returns instead of calibrating these. The problem is of course we do not observe the aid investments. We do observe what has been allocated but not what has been built as such. The same problem goes for the accumulated investment used to create the capital stock in general but for aid it's virtually impossible to get a good guess of the capital label. So this is why we think of this as an unknown parameter and you can think of the share of aid investment in the share of total aid. It's on the margin but what you find that the return to aid capital is of course a function of this. So if all aid is invested we estimate a return just below 20%. If half of the aid is invested we estimate a return slightly above 20%. What you get from this is that this is not an important unknown parameter. It's not truly important to know with great precision how large a fraction of aid we actually spend on capital as such. We get close to these 20% as you will see from these regressions. This is the overall average from 94 countries over 40 years. It's in the wider paper mentioned here. As you have to do with regression work you have to do a lot of different estimations, different regressions. So we run six different regressions on two different data sets. What we get is 20% gross return. This is the gross return we get so we have to deduct the depreciation on capital. The gross return in the neighborhood of 20% does not seem to be completely off. You may think of this as a bit high but then again the depreciation rate could be fairly high in some of these countries as well. So thinking of 10-12% as a return or a macro level may not be totally wrong. What we get from this is the 20% and we also get this is in good accord with the median World Bank projects. These 19-20% is what we get from the average across 94 countries, different data sets, different periods. So we think of this correspondence as supporting our results. We would also claim that this is not necessarily in contradiction with the overall return on aid being more or less the same in rich and poor countries because we are discussing different kinds of investments. Aid is not funding private sector machinery or capital equipment. It is funding infrastructure. And it should be... It's not difficult to think of a situation where aid has a huge return whereas private investment has a low return because that's exactly what happens when you have poor infrastructure and it could be when costs are high. So we may actually find a negative correlation between aid investments and private investment in this situation. So it could be that the overall return as calibrated by Kasselian and Fira is precise for this overall thing. But it has nothing to do with the return to aid. So the claim that aid inflows are just going to rush out of the country but capital outflows seems not well substantiated in this case. We can get some support for this in another reason studied by Pablo Selyer who's here at Eva Soonersen. They have been looking at foreign aid and foreign direct investments. And what they find is that if aid is directed towards infrastructure, human capital, whatever, there's actually crowding in terms of attracting foreign direct investments. If aid is for the competing sectors, there's a crowding out. So it's fairly clear in our mind at least that aid investment do have a positive economic return. Thank you very much.