 Maa, maa pepa nukes at export spilovas and the idea is to find out if domestic farms land from foreign old farms in both Kenya and Malaysia. And the outline of my presentation takes that approach, I start by giving the motivation, then the research objectives, the results, data and how I do the estimations. In terms of the motivation, I start by noting that these two countries have, they share more or less a similar history because after independence they obtain independence from the British. Then they adopted similar development strategies immediately after independence. But at the same time they also have many differences. And one of the differences of course has to do with the initial conditions, that it's possible to argue that the initial conditions at independence are not the same and therefore there's no need for comparison. But a more compelling reason why this comparison is necessary is because after 2002 when in Kenya when the new government took over power, the elites started developing a lot of interest on what was happening in Southeast Asia and Malaysia was one of the countries that Kenyans tend to compare themselves with. The argument is of course is that in the 60s they were doing better than Malaysia but all of a sudden Malaysia has been able to industrialize and Kenya has experienced stagnated growth over the years. So what has happened is that as a result there have been a lot of engagements between the two governments where Kenyans have gone to Malaysia to try and learn how they have been able to industrialize. And one of the fruits or the outcomes of this collaboration has been was the establishment of the free trade zones in Kenya after the Penang model in Malaysia. There was the establishment of the Kenyan National Economic Social Council which oversees the development of the country through the implementation of vision 2030 which collaborates with the National Economic Action Council in Malaysia and vision 2030 was also developed after Malaysia's vision 2020. So there is a lot of interest on what is happening in Malaysia and therefore this so in the process of learning I thought with this kind of comparison our policy makers can learn something about what happened in the manufacturing sector in Malaysia over the years. Now one of the arguments in literature is that the reason why Malaysia has been able to do very well is to do with their ability to not only attract foreign direct investment but also to use it in a productive manner. So and scholars like Jomo Rashia have talked about that and FDI was mainly went to the non-research resource rich sectors and it was mainly export driven starting from the 70s. So Kenya has also managed to attract FDI but for some reason Kenya has not been able to, the argument is that Kenya has not been able to take advantage of the FDI benefits and therefore the manufacturing sector has stagnated. Now when you think about FDI and the benefits that can accrue from FDI my study focuses on the indirect effects and export spillovas, one of them and according to Bloomstorm and Coco the definition is these are the benefits that would accrue to domestic farms in host countries through foreign owned farms export operations and which may pave way for local farms to enter the same export market. So there is a lot of learning that takes place so that foreign farms can also be able to start joining the market and this may be because of probably transport infrastructure or information that these farms get to know about new export markets as a result of interacting with foreign owned farms. Now what has happened in literature which of course is one of the contributions of this paper is that most of the studies that have looked at this issue tend to use random effects where the estimate a probit model and which captures the decision of farms to export so farms decision to export so it's either 01 probit estimations is that you can't do the fixed effects estimations because of the incidental parameter problem so what we try to do in this paper is we use a linear probability fixed effects model to do our estimations because of course once we do random effects model you tend to ignore the individual specific effects then of course for these two countries there is very limited literature and studies done in this area so it was also one of the comparing reasons why we looked at this. Now so we go to the research questions the first one then is do foreign farms influence the learning to export by domestic farms in Kenya and Malaysia and yes what are the transmission channels again a lot of literature does not look at the transmission channels so we also take advantage of the dataset we have to look at the transmission channels now I start by giving the results and the main result is that we find very limited evidence of exports pilovas in Kenya from foreign direct investment this well this was to a large extent surprising because Kenya has a reasonable level of FDI in Kenya and FDI tend to be very vocal so you'd expect that because of the presence of FDI for a long time they had learning also been taking place but based on our estimations there is very limited evidence of learning then the next thing that we see is that in Malaysia indeed there is evidence of export pilovas and these positive pilovas are mainly through backward linkages but you also have negative pilovas from what I'm referring to competition and information channels now I will discuss how I do the estimations and the proxies that I use for that as we progress then what you see in Malaysia is that the concentration of this pilovas the negative pilovas is actually in farms that have low productivity so it is the farms that struggle really that end up having suffering from the presence of FDI now we also find that this was our robustness test so we try to find out if indeed the relationship can also be the other way around that you can have domestic farms or foreign owned farms learning from domestic farms and we do not find evidence of that so indeed the learning effect takes place from foreign direct investment to domestic farms and not the other way around now we also try to speculate although the people does not really look at why this is why we have this kind of effects pilovas one of the reasons why we have positive pilovas in Malaysia we attribute that to the labour content requirements which have been enforced for quite some time and therefore seems like that kind of policy has paid off and in Kenya we do not have this so in terms of the data that we use we obtain some data from it Kenya from the Ministry of Industrialization the government used to collect this data awhile back but after 2005 they stopped collecting the data so we were fortunate enough to get a good sample of farms for both foreign farms and domestic farms then in Malaysia again the dataset we had was from 2000 up to 2006 again after 2006 the Malaysian government stopped collecting stopped releasing this data so I was fortunate enough when I was doing the study to access this dataset and what you see from this although most of the details are in the paper what you see from this is that the number of farms in Malaysia is way much more than in Kenya so there has been a lot of growth over the years but then we also have a reasonable number of foreign farms local farms in both countries so in terms of the variables that I consider I will focus mainly on the spillover variables we have RDF this one RDF which captures demonstration effects and this is defined as the share of expenditure by foreign farms in each sector the total expenditure of RDF in that sector then we have the the competition effects proxy which is the ratio of foreign owned farms share in employment in a sector to employment in the sector and the idea here is that if this will give us the importance or the significance of the foreign sector in the domestic market so if the sector is very important then you'd expect competition pressures going to the domestic farms then we also have FEX which is the proxy for information spillovers which is captured by the ratio of foreign owned farms in export sectors to total exports in sector J now the idea here is that we are looking at the export activity so if domestic farms are in sectors that have a high concentration of foreign owned farms engaged in export activity then you expect that they will tend to learn in the process then we also have the last proxy of the spillovers which is the domestic which is for backward linkages and that is the share of the cost of direct raw materials from domestic farms by foreign owned by foreign farms in a sector to the total cost of the direct raw materials so that captures the backward linkages and then of course we use the other variables we have other variables that we have control variables that we use how do we do the estimations okay fine so the empirical strategy that we follow is that we first estimate the determinants of farms decision to export in both countries so for all the farms then secondly we look at we do another estimation to find out whether foreign owned farms affect the decision of domestic farms to export and through which channels and then lastly we interact productivity measure with this spillover variables to try and see whether to try and see if because the literature suggests that the most productive farms could easily get into the export market and therefore you'd expect them to have a high concentration of spillovers and lastly we for the robustness test we estimate whether indeed the relationship can be reversed so whether domestic farms would influence foreign owned farms and Chia I would wish to stop here thank you very much for your attention