 So I'm going to present on migration, income pooling, as well as food deprivation. So I think we are all familiar with the context that migration is the defining feature of our century. Along with it we have got families or households getting specially dispaced. So we have got households at origin and then we have got the migrant households. Similarly giving us two sets of households that are dependent on each other. Technology is on the rise meaning that these households are now able to maintain and share decision making across distance as if they were together. So this background is now standing against quite a number of economic approaches. The first one being that economies traditionally view a household as an independent separate unit which does not in any way relate to the other. So it emphasizes on co-residents as well as eating from the same pot to define a household. So as a result the households model that have been developed up to date they don't take care of this special dimension. Also the interdependency is underplayed and finally the remittances are still not incorporated in the income of these models. The second one, second approach is on income pooling and I may have to define how I use income pooling in this context. So if household consumption is independent of who brings money into the households then we have income pooling. In other words the expenditure outcomes will be the same in spite of who receives the money. So that's how I use income pooling. I don't use income pooling in the sense of economic psychology where let's say a man and a woman would bring money into the same account or into the same pot. I am interested in the expenditure outcomes of that man. For instance if a man is to receive 100 dollars and then they are at lipid to use it for the household or for themselves so we see those expenditure outcomes. The same as for a woman if we take that 100 from a man and we give it to a woman then we say will the expenditure outcomes be the same. If the answer is yes so we have income pooling. So you can tell that income pooling is a stringent approach in that in a household where people live together within the same roof their expenditure outcomes can definitely differ. So economists have been studying income pooling in the same household but I'm extending this research to now study income pooling in what I would call geographically separate households which are dependent. So it's geographically stretched households. So we want to see if the income of the migrant that comes to this house in terms of remittance is used in the same way as the income of the household of origin. So that will be my income pooling in this instance. So the formal models are especially led by Beka who believes in the unit household model which equates to income pooling. Then we have got alternative models that have been coming up like the collective household models which actually do not equate a household into an income pooling unit. So this paper I've already mentioned a bit what I'm doing but I think I'll emphasize that on top of creating this geographically stretched household model I provide its testable empirical as well as policy implications and then I use those implications. I use the data from Zimbabwe Bulawayo to test the implications. In essence I'll look at the determinants of remittances. I'll look at income pooling from the migrants remittances as well as income generated at household of origin. And then finally I look at the impact of migration at the household of origin in the context of food deprivation. So I wouldn't want to dwell much on this but what I try to do is to extend the model by Beka which was extended as well by SIN to include the component of the migrant household. So to do this I maximise the utility of this household which I think has this component of consumption of purchased goods from the market, self produced activities or self produced goods and then finally the stock of time that the household spend together. So I maximise that subject to an income constraint that I model after a household member is migrated. So with that I find that the model gives almost the same outcome as the other economic theories. But what it goes on to do it now gives us the household income at origin which will be different than it would have been if it did not have a household migrant. So in short that is the model that I try to build. Then what are the implications derived from that model? The first one I think which is a bit common to us all is that the household with a migrant is expected to have higher income exports to mitigate food deprivation. And that's what I want to test. There are two falsifiable conditions that emanate from the model. The first one is that the migrants must be remitting definitely without remittance. I see a situation whereby the household at origin will be worse off compared to a situation if it did not have a migrant household. Households would have lost labour to most of the households in this context of Zimbabwe. They would sell an asset at times a cow or something to make sure that a particular person goes abroad so that they remit. So if this labour is lost and then there is no remittances this household may actually be worse off. Then the second one is that the remittances must actually be used for the benefit of those left behind. Now what do I mean by that? Some migrants send money to build their mansions, their houses which does not at times benefit the household of origin in terms of getting food. Or securing their food. So that's the other condition. Then the social policy implication is that a blanket social policy that excludes migrant households from social assistance may be prejudiced in that. If it's true that some migrants don't remit then a policy like Steven Devereux they prove in Zimbabwe that NGOs would ask a particular household to say do you have a migrant or not. If the answer is yes then this household does not receive social assistance. So should we have such a blanket social policy that just says if a household is a migrant then it's better off that it may be prejudiced because some households do not necessarily receive remittances as I will show later. Then the last one does not necessarily relate to what I'm presenting so I'll skip it. Let me then quickly get to our data. So we collected data in Zimbabwe on 300 households. So this data is structured as follows. I think what's important for me to quickly talk about is the fact that, sorry, is the fact that migrant households like a household with a lot of migrants, those households with the migrants have got less self production activities. Then the lower the number of migrants the higher the number of self productive activities which in a way just from a pictorial point of view demonstrate the entrepreneurial shield that may take place in the households which may stop labour migrating. Then I look at the household characteristics, the households with migrants they have a lower wage at origin compared to the households without migrants. Yet the households with migrants have consumption that is almost the same as that without migrants. I see households with migrants managing to reduce food deprivation and this is significantly different from households without migrants just at descriptive levels. I then look at the individual characteristics of the migrants, slightly below half they remit both cash and non cash remittances. There are more females in migration than males and most of them are actually in South Africa. So the first thing that I try to establish is the determinants of remittances. I use a logistic regression model to do that and then sorry the figures may not appear but what I see most if I look at the full model by full model I'm combining the local migrants as well as the international migrants. Migrant age I see it matters right across my full model as well as my restricted models where I look at migrants within Zimbabwe and migrants outside Zimbabwe. Education is also another factor as well as having a child at the household of origin. To my surprise I thought that gender would have been a factor but in this case it's not and I think the presenter Julie when she presented earlier on she also found the same result. So I was comforted a bit where the remittance you know gender is not necessarily a factor in the context of Zimbabwe. So I like that finding. However I'm going to show later on where gender seems to play a role on income pooling. So for income pooling I look at the partial derivatives of money spent from remittances on particular sets of goods like what I would call sustenance goods where I look at cooking oil etc. As well as clothing and education and I compare to the partial derivative of the money at household of origin. If the two are equal then we will claim that there is income pooling. Alright so these are the results. I will try to now speed up by five minutes. So what I find from these results is that on frequent expenditures females with children at the household of origin they pool their income together with that at the household of origin. There are two implications for this result. The first one it could be that females usually have a dedicated role of buying frequent food types like cooking oil etc yet men they dwell a lot on clothing etc which is reflected this side as well. But that is very important in that if we are to look at food deprivation then a migrant household with a female could particularly do better. I now split the sample into international migrants as well as local migrants. For international migrants the results are the same as the previous line but for migrants with things mobile I see that all migrants pool their income perhaps maybe of proximity. I go on to look at the impact of migration on food deprivation and I use a switching regression model to do this. Right with that I find that migrant households become a bit better in terms of they reduce food deprivation by 45 percentage points and they do better than non migrant households who in their own right without a migrant they reduce food deprivation by 8.9 percentage points. However I see that non migrant households would have done much better if they had a migrant such that they would be better compared to migrant households by 25 percentage points. So let me quickly conclude I see three factors that matter for my for remittance in Zimbabwe that is the age education and having a child at the household of origin. Then the second thing that I see is that gender matters for pooling income on frequent and low cost purchases that characterize food patterns of poor households. The third one is that on high value as well as infrequent purchases like education and clothing men also seem to pool income with households of origin. I see a situation whereby if a household is a migrant who is not above 30 years not educated as well as not having a child at the household of origin that household may actually be worse off compared to non migrant households to an extent that a social policy that is blanket may miss this point and with that I thank you.