 Hi everybody, thanks for being here today and for the session it's really great to have you. Exactly as Lucia said, I'll be talking today about multi-dimensional poverty and two studies that we did actually, which is creating an index for five different countries in sub-Saharan Africa based around gender and the context of forced displacement, and then analyzing those results at the household level for individuals as well as doing intra-household inequality measurement, which is a new exciting methodology we've been practicing at OFI using the MPI. So to kind of get into that then, the roadmap for this session is to be giving the kind of background and motivation for why we've studied what we've studied as well as a little bit of information about the MPI methodology for those of us who might be less familiar with it, as well as talking about the data, the MPI structure that we used, and then presenting the results. So before we get into the results, let's do a quick and kind of simple methodological overview so we're all on the same page when we discuss what the MPI is showing. So sub-Saharan Africa is one of the most conflict-affected regions in the world, with one-third of the total number of global conflicts taking place in that region. That's due to both conflict as well as climate change and various different crises associated with that. So the five countries that we studied were Ethiopia, Nigeria, Somalia, South Sudan, and Sudan. Now all five of these countries included in the study are some of the most conflict-affected in the region and record some of the highest numbers of conflict-induced displaced populations. And so what this context and what this study is bringing in studying those places is that descriptive analysis of forced displacement and multi-dimensional poverty, which is something that hasn't really been done on this scale before, analyzing the differences in the level and composition based on displacement status and gender, comparing multi-dimensional poverty measurement to income poverty measurement, as well as discussing some of the implications for policy and institutional settings. So now the MPI methodology. So what is multi-dimensional poverty? So multi-dimensional poverty measurement is a strategy that tries to focus the ordinary person, their beings, their doings, their capabilities in the discussion of poverty. So there's been increased recognition for this style of measurement over the last five to ten years with the SDGs, goal number one, talking about target 1.2, eradicating poverty in all its forms and dimensions, and importantly, multi-dimensional poverty is not a replacement of monetary poverty but a compliment to it. We see, I'll discuss this a little bit later, we compare some of our results by both measures, and you see that actually we often find mismatches about the populations we're studying. So it's important to remember that both are tools that have a lot of predictive and explanatory power in describing the realities that people experience in their ordinary lives. So what is an MPI? So it's a counting based multi-dimensional poverty measure. There are three key statistics that characterize every index. So the MPI is the product of what is the H, which is the incidence or the head count ratio, so the proportion of the population who are multi-dimensionally poor multiplied by the intensity or the A, which is the average share of the weighted indicators in which multi-dimensionally poor people are deprived. So I can go answer more questions on this later if you want but we have a lot of results so I'll kind of get into that now. So turning to our studies data and MPI structure. So as I said, five countries that we're studying here, importantly in Ethiopia and Sudan we're actually studying two refugee populations whereas the other three countries are focused more on internally displaced people. And also to know in Nigeria it's actually the six northeastern states of Nigeria that we're looking at. And the reason we used these particular data is because they actually were, the sample design was representative for both IDPs and the host community to allow that kind of comparative poverty analysis that we do. And so here's the structure of our MPI. So if you look on the left, there's the dimensions, then all of these dimensions are equally weighted and we looked at, we included four of them. So education, health, living standards, and financial security. You'll see within that we have plenty of different indicators, including school attendance, pregnancy care, drinking water, sanitation, cooking fuel, legal ID. And some of these are more kind of standard MPI indicators that are been reproduced in lots of different MPIs around the world. And some of them we've created or chosen specifically for their usage in the context of forced displacement. For example, bank account, legal ID, early marriage, these are issues and arenas that are particularly important for migrant communities. So let's get into the household level results of what our MPI showed us for the five different countries. Now what you can see is that in all countries, except for Nigeria, we saw a significant difference in poverty between the hosts and the IDPs, with IDPs experiencing higher rates of poverty. And if you look again at Ethiopia and Sudan, which are the two-camp samples, that difference is really huge. And what's interesting about those two examples is that actually in Ethiopia, the refugee population makes up more of the sample. So they're 88%, but the population share of IDPs in Sudan was much smaller than their host peers, and still their poverty was much higher. So this table now shows the breakdown of the poor population by the intensity of their poverty, ranging from 50 to 100%. 50% was the poverty line, the poverty cutoff that we used in our MPI. And again, you see some real ranges here. So in Sudan, for example, 72% of the poor in host communities, and 62% of the poor IDPs are in the lowest intensity band, relatively close to the poverty line. Whereas in Somalia, the country with the highest average intensity across both groups, the distribution is much wider. So there's again, this is the type of methodological tool that the MPI can do is give these kinds of different breakdowns based on population shares and different indicators to show how poor the poor are. And if we look at the MPI for male and female-headed households by country, again, you can kind of see these differences playing out. So in Ethiopia, South Sudan, Sudan, female-headed households are more common among the forcibly displaced population. And for other aspects of household compositions, which I won't present today, but we did do the analysis on. So age of the household head, household size, that sort of thing. There are, can we see? Ah, sorry, there are not hugely significant differences, which is to say we controlled for them. And it seems that actually what's driving these differences is poverty. So one of the key features of the MPI methodology is that you can break each of the indicators down into something called censored head count ratios. A censored head count ratio is essentially just a way of saying the percentage of the population who are multidimensionally poor and simultaneously deprived in that indicator. So this is for Sudan, and you can see for all of our different indicators the differences in bi-displacement status between non-IDPs and IDPs. So not only are we showing you the overall, this is how poor people are in that country, but we can actually break down in what ways, in what arenas, in what indicators, the poor are poor. And again, we can see these censored head count ratios located only for the refugee population in Ethiopia. And again, we're looking at gender of the household head. So you can see there are some pretty big differences among headships specifically for refugees in Ethiopia. And as promised, here's some of the differences between the poverty measures with income poverty versus multidimensional poverty. So each bar is representing the whole population of each country. And the bottom kind of yellow beige bit of the bar is the proportion of the population covered by both measures, whereas the top blue-green of the bar is the difference. So you can see there, again, there's some variation. So in Somalia, it is about 50% where the measures are covering the same population, whereas in the Northeastern Nigeria sample, it's like 28%. So there really are, there are two different measures showing us different things. So looking at the individual level results, and so the individual and the inter-household level results that I'm presenting today is some of the innovative work that we've been doing as part of these two, as part of this research series, because the MPI has traditionally been looked at from more of a household lens, and we're trying to break that barrier down to really understand the experience of individuals. So yeah, so explaining the divergent experience of individuals in the household. So here is, for the five countries, the percentage of the population who are multi-dimensionally poor by gender and by displacement status. So as you can see, though, in Ethiopia and Somalia, women and girls who are forcibly displaced are experiencing a significantly higher rate of multi-dimensional poverty on average than they're not in displaced counterparts. And so you can see, again, this kind of variation. Some of it is significant, some of it isn't. In Northeast Nigeria in particular, it's the sample size means that some of the findings that we had weren't necessarily, that we didn't see the kind of significant differences that we thought that we might. But in Somalia and Sudan and Ethiopia, the differences are much larger. And again, this graph is showing the same key statistics, but now we're zooming into the child population, so anyone under 18 years. And actually, if I flip back and forth, you can see that they look pretty similar. So now let's talk about comparing the gender differences and the deprivation. So those are those indicators that we discussed. And assessing the gender gaps in each of the individual indicators. And so the indicators that had information for which we could do this type of analysis, i.e. from the individual and not necessarily just the household level information, were spanning three of the four dimensions. So education, health, and financial security. So I won't show this for every single country, I have it, but it's probably far too much information. So I'll focus on Ethiopia and Sudan, which are, again, the camp samples. And so looking at Ethiopia, right, so you've got the indicators on the left-hand column, male, female, and the difference for both refugees and hosts, and the all households, so the whole population on the top of the table, and then the latter half of the table is the MPI poor. In Ethiopia, you can see that the education deprivations are really high with around eight and 10 children, sorry, poor children, not attending school, regardless of displacement status. But interestingly, the gender gaps in school attendance, so that's the differences column, are significant only for the refugee community. So this type of, and also the other thing that I would note here, because this was one very consistent finding we saw across all five of our countries, is that the early marriage difference, the gender gaps was pretty much always significant, and it was always worse for refugee and IDP populations. And a key feature of this study, and looking at the five different- I'm so happy you've got trees. Two minutes left. Thank you. A key feature of the study that we see is variation, variation, variation, where the context of each country matter a lot, but the one finding that was consistent was around early marriage. And so again, here's the example for Sudan that kind of shows similar thing, gender gaps in education and financial security were only significant among the wider population, but again, but the deprivation levels in education for Sudan were surprisingly low. So again, context matters a lot. Key takeaways, variation, early marriage. So lastly, the inter-household results. So we focus on four, the four individual level indicators that we had information for. So school attendance, primary school completion, unemployment and legal ID. And we're looking here at both accumulated and contemporaneous disadvantages. So accumulated disadvantage would be primary school completion. We might be looking at people who didn't have education in their prior country, where school attendance is current because is the child in school, are they not in school? And we saw some pretty interesting things. We saw that in all countries but Sudan, there was a significant relationship for school-aged children's experience of inter-household inequality and their displacement status. But because of the sample size that meant the only Nigeria and Somalia, those results were robust. And again, what you can see is that we're really just seeing a tons of variation for each of these different contexts. But that we all, but we did notice that most of the results for the gender, most of the results for inter-household inequality were being driven by the gender variable rather than the displacement variable. We also ran an interaction term with gender and displacement to figure that out. But you can see the main differences here. And again, with financial security, it's interesting because displacement can boost female employment rates if women in the household are taking on traditionally male tasks or the norms around paid work are changing. But actually inter-household inequality and unemployment was relatively insignificant in our study. So if we could do that. Yeah, and so those are the key takeaways. So thanks everyone. Thank you.