 Okay, when I prepared this talk about capabilities, you can see it says the short version, I actually thought we had more time. So this has been boiled down and it focuses on the qualitative data that we collected during the Learn to Compete project that John talked about just before. When I say qualitative, normally I work with a lot of data. So we're talking about that we interviewed around 100 multinational companies and around 200 linked domestic companies to those multinationals. I will come back to the reason why we did that. The main reason is that we would like to look at whether FDI is actually building domestic firm capabilities. That was what we looked at. And when we looked into the literature, we saw that a lot of these studies were using what we call an indirect approach. They were studying whether the share of multinational companies in a sector is actually increasing the average productivity levels of domestic firms in that sector. Nothing about whether they are directly linked or whether it's just what I call externalities that is actually generating these spillover effects. And we tried with this approach, I think it's quite novel. I have not seen it before, at least, to trace these domestic firms and link them to these multinationals to see whether we can actually disentangle some of these FDI spillover patterns to domestic firms. So what we actually have here is that the common feature of FDI spillover literature is that they measure these indirect effects. So basically the studies, they look at the links between the increase of the presence of M&E's on the improvements of domestic firm productivity. That can be in horizontal aspects, so firms in the same sector as the multinational, but it could also be in the vertical dimensions. So whether it's upstream or downstream spillovers. What we found in the literature, there is a couple of papers in the management literature. But in the economics literature, we do not see a lot of distinction between what we call indirect spillover and externality effects from what we call the direct effects. So how do we actually see the firms that engage actively with the multinational? Are they gaining more in terms of productivity by interacting directly with the firm, or multinational, than a firm that does not interact with the multinational in the same sector. So controlling for these aspects. We have a lot of literature stating, and I've just cited some of the people that most of you know, I hope, that Arrow he emphasized early on in 69, that knowledge diffusion is not an automatic process, and it actually is dependent on the direct relationships between firms. That is actually also confirmed indirectly by Hirschman in his book from 58, that the lower the absorptive capacity is in a country. So you can also refer to the Houseman Hidalgo lack of economic complexity, the lower the probability of actually experiencing these indirect spillovers. And that is actually also what we see in the papers in the publications that in countries with a lot of economic complexity in a lot of Asian economies, we see a lot of spillover effects in the studies, whereas in Africa we see lower spillover effects in the estimates. This is Hirschman would say that this was what he wrote in 58. So nothing new about this, but we don't know much about these direct effects. So what we do is that we try to disentangle the direct from the indirect effects on these domestic firms. And this is the novel approach. And we've asked these simple questions, interviewing 102 FDIs, and we link those 102 FDIs to the 226 domestic firms. Let me just from the outset say, just describe what we're actually doing. So everybody knows that this has been not an easy process. We actually asked a lot of the, all of the investment promotion agencies in the countries that we went to, asked them to identify the 15 most important M&E's in the country. Then we went to these M&E's and asked them, these M&E's, you can see this is the case for Vietnam. And then we asked them to provide their suppliers, the domestic suppliers that they have, the domestic customers they have, and the competitors they have. You can see here in the Vietnamese case, they have a lot of domestic suppliers, domestic customers and so forth. Let me just show you the picture for Kenya. This is how it looks for Kenya. Asking M&E's, you can see there's not a lot of what we call Hirschman linkages. So there is not a lot of economic complexity. They are not trading with the M&E's, they are not trading as much with the domestic firms as we see in the Vietnamese case. They have lower economic complexity as such. Here, because some of you would state those that know Kenya well, they will say this cannot be true. But this is the updated version where we include information whether M&E's, they trade or engage and to act with other M&E's. So you can see in, for example, the firm number one in Nairobi doing packaging is actually engaging with a lot of other M&E's as suppliers but also as customers. Those are in the red. So there is a lot of engagement in the Kenyan case between firms but it's between M&E's and not so much between M&E's and domestic firms. And this is a general picture we see in Africa compared to Asia where the domestic firm is engaging more with M&E's than we see in the African case. What would that suggest? Indirectly or directly, we were actually thinking about Hirschman's book. We would say that if the ISB lovers are going to be less pronounced in the African setting than in the Asian setting. Given time, we also do a lot of other things in the papers. We're actually looking at whether a lot of literature is actually discussing whether if the ISB lovers only work if it's producing intermediates for other domestic firms that you only learn via the intermediary process. And we see for example in Kenya that it's only, I cannot see here, over 110% I guess, that is of 3% is actually for intermediates and most of it is final goods production. That could suggest that spill overs are also in the Kenyan case less likely to occur between M&E's and domestic firms. In the Vietnamese case we can see that 61% of the engagement between domestic firms and FDI firms or M&E's is actually through the intermediary process. And that increases the scope for learning. The same goes for when we look at whether they trade. We talked a little bit about this as well. But we actually find in these processes that spillover effects are less likely to occur through trade patterns than through FDI patterns. So direct engagement through FDI, so presence of M&E's in the country. Spillover effects and building capabilities is more likely to occur when the firm is present in the country compared to when they trade across borders. That is also what we found very significantly in the study. So if we go to the summary of the findings, we have some papers on this. Let me just so I can make sure that I actually get all of it. Sorry about this. So we find, first of all, that there are very few linkages between M&E's and domestic firms in sub-Saharan Africa compared to Asia. But something that we didn't present here, that when these linkages are present in the African setting, they are more likely to lead to direct spillover effects. So where both the domestic firm and the M&E state that they have transferred knowledge to the other firm, also the domestic firm. I don't know if John is going to come in to talk about this. But what we also found is it's not technology. It is not machinery. It is management skills that is transferred. It's logistical processes that are transferred between these firms, M&E's and domestic firms. So it's not so much physical technology that we actually see upgrades. A large part of the direct vertical linkages is done through formal contractual arrangements. That is also important, especially for the IPAs that we interviewed. So when we see these transfers actually happening, they are actually stated, stipulated in a contract. And that was a big surprise to us because a lot of us are thinking about these processes as automatic. But if we think about what Arrow said in 69, they are not automatic. They actually have been thinking about this and observed this for a long time. And this is also what we see in our study, that it's not an automatic process. There is room for policy in this aspect. Then we also have that direct knowledge transfers are likely to occur through FDI as compared to trade. And this is consistent with the view that tacit knowledge transfers are more likely to occur through localized linkages. There is a lot of theory on this. And linking to the Hausmann and Hidalgo work, we see clearly that lack of economic complexity in African industry makes direct linkages an important player. They cannot rely on indirect linkages as much as we see in Asia. We shouldn't expect that FDI have similar effects. It's not the same mechanisms that is at play in the African setting as in the Asian context. Good. Thank you.