 Good afternoon or good evening everyone. This is lockdowns, food security and the role of external support. We have five exciting presentations, hopefully a very lively Q&A, but also very little time. So therefore we go to Al-Muqsid Akim who will present his work on the role of remittances in Nigeria. Thank you for having me and my name is Al-Muqsid Akim and I'm happy to be here today to present our paper titled, Do Remittances Unexpectedly Ensure Against COVID-19 Implementation and Shock on Food and Security. So this is a joint work with the firm AIVOG and Jeffrey Couton. Put it in another way, the basic question that we are trying to address in this paper is to look at whether remittances can ensure households during this current COVID-19 shock. The question we are arising here is related to a large body of literature looking at the insurance role of remittances in this shock. There are many studies providing evidence of this insurance role of remittances. Most of them show that remittances can ensure households by allowing them to smooth their consumption during shocks like idiosyncratic shocks like death or covariate shock such as precipitation rate or temperature shock. So in this literature what is interesting is there are two underlying mechanisms that are highlighted and one of these mechanisms is what we hear, exposed mechanism. This is a mechanism that operated before or during or when the shock happened. Basically, for example, when the household experienced a shock, the household may receive remittances that may help to cope the shock. But in this particular context of COVID-19, this mechanism is unlikely to happen because the COVID-19 affect also the migrant in the destination country. So the remittances is likely to decrease and then we are not expected this mechanism to operate. This exposed mechanism can be opposed as ex-ante mechanism. This ex-ante mechanism is a mechanism that may operate before the shock happens. If you consider the particular context of COVID-19, for example, the remittances that the household may receive remittances before the COVID-19 shock like the previous year, one year before. And in that case, for example, the household may use part of these remittances to invest in productive activities or to access financial services such as savings and credits. So this is the mechanism that we are trying to look at in this paper. So now let's jump to how we do that. So basically we use six webs of household surveys that include two webs before the COVID-19 and four webs during the COVID-19. And then these data allow us to implement a difference in different strategy. And the main thing that we found is that if you look at this graph, first that following the shock, and we look at our outcome of interest here is food insecurity. And after the shock, we see that there is a significant increase in food insecurity. And what is interesting to note is that this increase in food insecurity persist over time. And we found that remittances can mitigate this adverse shock on food insecurity. But it is interesting to note that this mitigating effect operates only at the early stage of the shock. Here you can see that the mitigating effect is significant only when the shock happened at zero, at time zero, and at time one. But from time two, it becomes unsignificant. So the question is about, because earlier we talked about the ex-ante mechanism. And in this paper, we proposed a formal way to test this ex-ante mechanism by looking at whether access capital can amplify the mitigating effect of remittances. And in this table, you can see that households that have access to capital have higher mitigation effect of remittances than those who don't have access. So quickly, we draw from these findings two policy implications. The first one is that remittances are an important private element of social protection that's worth considering, especially in this context of COVID-19, where government around the world are trying to revisit and to rethink the social protection strategies. So it makes sense to include measures that channel remittances toward increasing household capital. And a second policy implication is that the remittance-provided protection that we found here has to be considered as complementary of existing social protection systems, because we saw that it is effective only at short-term. Over the time, it is not effective. Thank you for your attention. Thank you so much. Moxie, it was very interesting and also thank you for keeping time. I just want to remind everyone that there is a Q&A tab. You can use that to ask questions, and we'll have a Q&A session after all the presentations. And if you want, we can also invite you to the stage to ask your question using the video. So next up is Kibrom Tafere, who is going to present on the role of safety net during the pandemic in Ethiopia. Hi, everyone. Thanks for attending my talk. The title of my presentation is COVID-19 and food security in Ethiopia through social protection programs. This is joint work with Kibrom Abay, Pushprahana, and John Hovman. The COVID-19 pandemic has assisted food security and social protection systems, and its impact has been unprecedented, according to a recent World Bank report. In 110 and 150 million people are projected to fall into extreme poverty by 2020. And the World Food Program has similar alarming projections. It's estimated that the number of people facing food-acute food insecurity is likely to double due to COVID-19. This circumstance, social protection programs can be an important role in reducing food insecurity and protecting against the situation. In fact, since the outbreak of the pandemic, more than 200 countries have implemented some form of social protection measure. However, we know a little about the effectiveness of these programs. That's where our people come in. So that's two simple questions. One, what's the impact of COVID-19 on also food security? Two, are social protection programs affecting these impacts? To answer these questions, we focus on Ethiopia's flagship social protection program called Productive Systemic Program. From now on, I refer to it as just PSMP. It's a large social protection program with approximately 8 million beneficiaries. It is targeted geographically at the districts with chronic food insecurity and within these districts at households who are food insecure and lack assets and have limited alternative source of income. For this study, we used two rounds of data. The first round was conducted in August 2019. In 1988, we were at us where the nutrition-sensitive component of PSMP was supposed to operate. We conducted follow-up in June 2020 and we managed to cover approximately 60% of our regional households. At the baseline survey, the August 2019 was face-to-face, whereas the follow-up survey was a phone survey. As a result, we were only able to reach households with access to telephone and this would have implication to our estimates and I'll talk about them in a bit. So to our identification strategies, we use standard differences in different strategies where we compare food security outcomes of PSMP beneficiaries and non-beneficiaries before and after COVID-19. As I mentioned earlier, because our follow-up survey was a phone survey and we were only able to reach households with access to telephone, our sample is by design, a selected sample. This would introduce bias in our estimates unless we address it. So to deal with this problem, we construct sampling rates which we used to adjust our estimates. So going to our results, the COVID-19 pandemic increased the likelihood of food insecurity by 12 percentage points for non-beneficiaries, but participation in PSMP reduces the likelihood of food insecurity by 9 percentage points. That is, PSMP beneficiaries are 9 percentage points less likely to be food insecure. So in terms of the raw measure of food gap, which is the number of months in which a household had difficulties to satisfy its food needs in the last 12 months has gone up by approximately half a month for non-beneficiaries, whereas this number is less by 0.3 months for PSMP beneficiaries. So these results suggest that the participation in PSMP plays an important protective role. So in terms of heterogeneity, we find that the effects of PSMP, the protective role of PSMP, is higher in poor areas for poor households in households who live in remote areas. To conclude, we find that the security deteriorated in the aftermath of the COVID-19 pandemic. The QPS productive safety net program mitigated the impacts of the pandemic on food security. And we also find the protective role of the PSMP is higher for poor households in households who live in remote areas. Thank you very much. Thank you. Thank you, Kibran. Next we have Saif Aras, who will be presenting on the impacts of the pandemic on Rohingya refugees in Bangladesh. Hello. I am Saif Aras. I am currently working as a research fellow in James B. Grant School of Public Health University. Today I am going to present a paper named impact of COVID-19 on households living income and food security on if the amount that is possible disgrace my nationals and its adjacent host community in Bangladesh. As we already know that many of us know that if women are the mostly majority ethnic group who lived in Myanmar in centuries, they have faced, they have faced persecution for decades. The largest influx happened in August 2017 when almost 750,000 FDM in Bangladesh. Now these large communities currently living in Bangladesh who are largely dependent on community at 8. This FDM community is surrounded by a minority of less than half million of Bangladeshi who are residing in nearby villages. They are known as host community. One of the poorest population groups are these host community with a poverty rate less than 32%. Since the arrival of FDM in separate episodes of tension has been raised including different kinds of health-free, competitive labor market, rise of living expense and etc. Next comes a new crisis, COVID-19. And after the lockdown was initiated both the FDM and host community faced massive challenges in their lives. So the objective of this study is to assess the impacts of COVID-19 on households income, food security and to identify coping mechanism among most vulnerable group in both two communities. We followed a sequential, exploratory mixed method study design using both qualitative and quantitative techniques. For qualitative, we conducted 59 qualitative assessment including IDIs, FTDS and KIS. And for quantitative, we conducted 2032 interviews. For both method, the target population was most vulnerable group. Who are these most vulnerable group? After several discussions with different kinds of stakeholders and literature review, different literature group found pregnant and lactating mothers and innocent people with disabilities and the people seeing a female-headed household head without income or no income are the most vulnerable group in these two communities. And we found those who are mostly the key bread earners in these two communities who are the informal job earners. Females were the mostly engaged in managing household chores. Over half of the FDM and adolescents were married. Around 19% FDM and adolescents reported that they had never gone to school. And also 16% of most community reported that they have dropped from their school. We also found that single female-headed households are much higher in FDM and camps than the host community. That is the almost 25%. To understand about the COVID-19 impact on household income, we asked the respondents about their job and household income status during pre-pandemic and pandemic period. And we found 54% FDM and respondents reported no change in their income because members of FDM and neighborhood had regular job before the pandemic. In the contrary, almost 70% of households of host community reported a decrease in their income because they are mostly informal workers. They had small business which was highly affected by the COVID-19. As a coping strategist, FDM and host community, host survey households are mostly mentioned about their reducing their expenditure. They also reported about taking loans from different sources to manage their household expenses. In terms of relief, only 45% FDM and reported receiving food which was not sufficient to meet their needs. In case of food security, almost 65% FDM and households reported running out of food due to lack of money in the past nine months of data collection. Almost half of the surveyed households mentioned that they could manage only rice on a neighbor's one to three times in a month. The picture is almost same in the host community. The quantitative assessment also found that the families and persons who solely depend on ration are suffered most. As a coping strategist of food shortage, most of the respondents reported about borrowing food, eating less. Only 9% FDM and households can stop food for the crisis period. And it is also reported by the host community that they have to starve sometimes. On the particular analysis on single female headed households, we found 50% of these vulnerable households had to explain food consumption reduction. The picture is almost similar in the host community. Appealing loans is the most reported coping strategist. Only 2% can manage to save money. And this is because as they aren't have limited access to the employment and earning opportunities as they are women. So we can conclude that those who are involved in the formal sector in these two communities were suffered most. Female headed households experienced the old scenario. So we are recommending that focus to skill development and training should be provided to female headed households. And targeted food and economic stipends are needed in both FDM and host communities. Thank you all for listening to me and also thank you for giving me the opportunity to present the important findings. Thank you. Thank you very much Saifa. Next we are going to get a presentation from Krithika Singh who is studying Desiree in compliance with the COVID pandemic measures in India. Hi, good morning everybody. Thank you for joining me for this session. Today I'm here to discuss my research work which was understanding varying compliance with the COVID-19 measures within India. I used a mixed methods approach to identify and examine the positive applyer in the Indian provinces of Kerala, Maharashtra and New Delhi. Hi everyone, I am Krithika Singh. I recently graduated from the University of Birmingham. I'm deeply passionate about working in social development sector. I wish to be part of research and projects that work to enhance the state-society relationship. So the aim of my project was, my research was to understand why a country like India could not elicit compliance. Moreover, the non-compliance arose to a level that it became the largest migrant crisis the country had witnessed since its independence in 1947. I also set out to understand why within the same country there were different areas that could have elicited different levels of compliance by the citizens. This was the methodology I used for my research. In the phase one, I studied the legitimacy of the country at center and within the provinces and what degree of compliance they could have elicited at that point of time. So to understand that, I used a framework in which I analyzed the state-society relationship, the quality of service delivery, perception of procedural justice and public trust. So I understood the legitimacy and the estimated compliance through these four segments. In the phase two, using a qualitative research methodology, I studied and identified the trends of non-compliance that were observed in Maharashtra and New Delhi. And in the last step, I compared these with the positive bias that is Kerala. So these were my findings. Firstly, a legitimate state cannot fully elicit compliance. So for instance, in India, it is a highly legitimate state. The party isn't powered through popular votes. Even in the surveys that have been circulated in the initial phases of the lockdown, it could be seen that the people are highly welcoming of the rules that are being set out by the government. However, soon within few weeks, the scenario completely changed and the largest migrant crisis surface. So this just goes on to show that legitimacy cannot, can only partially elicit compliance by the citizens. Secondly, I understood that state legitimacy depends on the services and the perception of the procedural justice of the state. So in one of the states that is New Delhi, while going through my data sets, I came across an interview in which one of the ladies was saying that to be able to access these services that by the government, they have to be part of a certain circle. They have to be part of certain class. And if they're not, they will not get those services. And so this just goes on to show that even if the state, Delhi is a highly legitimate state. The people have selected that government time and again. And they're highly approving of the leader. However, they did not, since the access to the services were poor, they did not trust in the state and they did not comply with the rules that were being set down by the state. Thirdly, I came, I understood that trust is the key factor that leads to public action and citizens compliance with the state. And trust is not an absolute concept. It is fluid and needs to be instated time and again. Going back to my example of Delhi, one of the major problems with the state was that the government had to do a non-compliant section of society for the minority groups. And so these minority groups, although had selected the same government, their experience with the government over the past few months had been rather harsh. So the same minority groups had been subjected to a Citizenship Amendment Act and subjected to the communal riots. So once they had these kind of experiences with the state, which kind of tarnished their relationship with the state, and they did not comply with the rules that the state had installed. So Kerala, the positive outlier, so Kerala, what they did right was that they had a robust service delivery system. Their budget allocations have always been higher than any other state, even if their GDP is much lesser than those states. Then they had an agile response to the emergency because of their experiences with the previous epidemics. Most important and the greatest contrast I saw with the other states was the welfare and the degree to which they had provided services to the people. Two examples I'd like to share here. One was that while the helpline numbers were getting engaged in Delhi, Kerala offered help in nine different languages for its migrant workers. And even the community kitchens were set up. They were offering food to the migrants from their home, from their villages, so that they felt at home in Kerala away from their house. So this was how they, Kerala was able to install and instill trust within the migrants and they were able to, you know, elicit high levels of compliance. Thank you so much for listening to me. This was all that I could present in these six minutes. If you have any other queries or questions, please do look through my document. Or you may also contact me. And if you have any constructive criticism, I'm also really looking forward to hearing from you about that. Thank you so much for joining me today. Thank you. Thank you so much, Kritika. And thanks also for reminding us about the question and answer feature. There's a tab there. You can post your questions there and we'll address them after the next presentation, which is about localised effects of the pandemic in Uzbekistan. And this is by William Seitz. Hi, everyone. My name is Will Seitz. It's great to be here with you today. So I'm going to talk about dynamically identifying community-level COVID-19 impact risks. This was some work that we did in Uzbekistan with co-authors who you can see the names here. But the situation was, you know, early in the pandemic in Uzbekistan pretty severe. You know, society faced an unprecedented shock. But we knew that, you know, not all needs were the same, but there were particular groups of people who were at elevated risk and who were in need of more support from government and other partners in Uzbekistan. There was quite a bit of information that was available that was being collected at the local level in small neighbourhoods called Mahalans. But that information was systematically available to policymakers in a way that could be acted upon. And at the same time, official surveys were put on hold because of the risks of collecting. So what we did was we undertook an effort to consolidate all of this information, all this administrative data that was being collected at the local level in 9,120 neighbourhoods. And we matched it with a survey that was ongoing at the time of listening to citizens of Uzbekistan survey, which was not disrupted by the pandemic because it was a home-based panel survey. And then we used smaller estimation techniques to fill in some of the gaps from the administrative data. So looking at indicators or measures that were not directly observed and sort of imputing them at the local level using smaller estimation techniques. Then, you know, with a combination of these different sources, we were able to create a database and a summary index of particular risk factors by these local communities. So for each of these 9,120 communities across the country. And we then, you know, together with our government partners made this available to policymakers for the response effort. So as I said, in Uzbekistan, you know, early in the pandemic, the crisis was pretty severe. Due to lockdowns and, you know, the risks that were posed to the public, you know, employment took, you know, a sharp downward trend early in the pandemic. So a decline of more than 40 percentage points. You know, a number of households without a number who was working. Remittances are a really large share of income for many low-income households in Uzbekistan and remittances fell by more than half early in the pandemic. And the government was rolling out mitigation efforts, you know, targeted social assistance and so on. But, you know, resources were limited. And they wanted to make sure that they were targeting the resources to the people who were most in need, you know, from the impacts of the crisis. So here you can see graphically, you know, big shocks to, you know, people working in employment. So basically there's, you know, these neighborhoods in Uzbekistan collect information about local conditions. And this information was typically aggregated to the level of the rayon. So there's about 190 of those. And sometimes aggregated to the level of the region. So there's 14 of those. But there wasn't a systematic way that this was all being collected and sort of given to central decision-making bodies on a very regular basis. And so at the time of the outbreak, this information was not available to policymakers. So what we did was we collected all that information systematically, that made sure that it was all, you know, using consistent definitions and so on. And then we matched it with a high frequency survey data that we had collected using this survey that I mentioned before. So on that basis we did small area estimation for those indicators which weren't observed directly in the local level. So here we're talking about, you know, income remittances, these sorts of indicators, using what's called the fair area area level approach. I don't have enough time to go through all the details, but we can discuss it if people have questions about that. And then on that basis, sort of using the alkyr foster method, we created a community level risk index, looking at, you know, high risk areas, trying to identify which places were in particular need of government support. So here's the index. It's split across six dimensions with indicators under each one. So the indicators within each dimension are equally weighted at the beginning of the process. And then we reweight them based on feedback from a stakeholder survey. And that yields the final index. So this is the Mahala level index. So for all 9,120 Mahalas aggregated to the Rayyan level, just above that. And you can see that there's some concentration factors across the country. This also allowed us to look at, you know, particular indicators that were of interest to the counterparts in a disaggregated way. So for instance, in March 2020, this was, you know, at the Rayyan level, the number of households, the share of households where no household number was working in the past seven days. You can see there's a couple of clusters of that here. Moving to April, you can see that there was a huge decline in reported employment, but that it wasn't that there were some, again, clusters of areas where it was particularly, you know, it was particularly strongly disrupted. And this happened, especially in highly populated dense areas. Recovery happened at different paces and different places. So you can see here by May, much of the country was back to work, but this was largely in areas that were rural, agricultural areas. And urban areas remained under lockdown for a long time. Whereas by June, most places had relaxed, but again, there were some clusters of areas where the risk factors were higher in lockdowns. You can see something very similar for remittances. So this is remittance prevalence before the outbreak. So in March 2020, you can see that there's some areas that rely much more on remittances than others. Moving to April, there was a sharp decline in remittances. And it recovered, but in different ways. So this is where people had typically been receiving remittances before. And so the composition of remittances changed in the immediate aftermath of the crisis in May. And then in June, the full sort of spectrum of remittance flows began sort of recovering throughout the whole country. You can see here. So this information was available. We made this available with our partners to, you know, to governments and policymakers. And we put it on the dashboard so that policymakers could make decisions either on dimensions on particular indicators or the overall index as needed. So thank you so much for this opportunity to tell you a bit about our work. I hope we get a chance to discuss questions and maybe a little bit more detail. I'm very grateful for that. Thank you. Okay. Excellent. Thanks a lot. So it's time now for the Q&A to put their video on. And we already have a couple of interesting questions here. So let's start. There's actually a question from one presenter to another. So let's start with that. So Kritika, Sifa is asking, have you conducted any survey for this project? You're muted. I'm sorry. I did this research last year in between May to June. I'm sorry, August. So I was in UK doing my masters. So it was logged on. I couldn't go back into India, although I had full plans to. So I used all the online data available. So that means interviews, tweets, or calls to my friends, families, their friends. However, I could get in contact over online like phone or Skype or anything like that. I couldn't personally go for any surveys or anything like that. It was a desk-based dissertation. Thank you. I hope that helps. Yes, in a sense. Yeah, thanks a lot, Kritika, and good questions. It kind of highlights the difficulties that everyone is having as researchers during this pandemic. So there is a question to William from Arthur who asks, I missed how the risk index was established. Could you please explain how that is calculated? Sure. Thank you so much for the question. So basically it's a combination of indicators from two different sources. So the first source is administrative data that was being collected by local communities throughout the country. And the second source was a monthly panel survey that we had in the field before the crisis began, and it was active throughout the crisis in Uzbekistan, and it's still ongoing now. So the way that we filled in the information for those indicators that weren't universally available for all of these local communities was a technique called small area estimation. So we identified particularly important indicators for the crisis. So some of them were on food security. Some of them were related to remittances. Some of them were related to income. And then we put the sort of the universe of all of the questions that we had available in a survey that we sent out to various groups that were involved in the response effort. So it was including governments, stakeholders, and international organizations, local NGOs, and they sort of responded about which indicators they thought were the most important to include in this risk index. So then we did the selection on the basis of this feedback, and then the weights for them in the index were all using the responses recorded in this sort of stakeholder survey. So the approach is just a basic alkyra foster sort of, like, you know, if you're familiar with an MPI type of multidimensional poverty index kind of indicator, it uses the same approach, but the unit of analysis is the the the local community instead of an individual or a household. So that was in a nutshell sort of how we how we did the risk index. Thank you very much. So I have a question for Saifa. So I found your research very very inspiring. So thanks a lot for presenting that to us. I was wondering how if you've been following this community for a long time and how this recent events in Myanmar are affecting the situation there and refugee camps if you have any any follow up studies on that? Actually, it was a baseline study. We conducted this study in from November to January. After that, the baseline study was stopped. But this is a as I mentioned in my presentation that this is a large project. This presentation is actually partial findings of a large project. So we have a plan to make a midline study than the end line. So we guess we can, as you said that the Myanmar recent incidents, we can follow up these things in our midline, but not in this study. This is a baseline study. Okay, thank you. We'll stay tuned for those results as well. So there are questions from Rodrigo Olivier. So I was wondering if he wants to ask them live. Seems to be that there is a lot of questions. So I'm wondering if he can appear online or I can also ask this in case he's not available. Okay, so why don't why didn't I ask? So there's been a couple of questions and answers already between Rodrigo and Al-Muqsid. So just to repeat the last question that is there. So Rodrigo is raising a concern that perhaps the more vulnerable households have higher probability of receiving remittances, but also stop working at the same time. So how do you deal with this in your analysis? He's asking. Yeah, thank you. Thank you, Rodrigo, for this question. I mean, we hear this question. In this paper, we focus on remittances received before the pandemic happened. And then we make sure that when the households stop working, we make sure of the let's say the exogeneity between past remittances and the likelihood to stop working because of the pandemic. Because the underlying assumption is that past remittances is not effective. There is no way to think that past remittances can be linked to the COVID-19 can be affected. Remittances can be affected by the COVID-19. So this is how we basically our main identification strategy considering past remittances instead of current remittances that may be correlated with the current family. Thanks. Let me ask a follow-up question, a little bit related. But I was wondering, in your view, there are international and domestic remittances, right? So I'm guessing that the international remittances are larger, but maybe less frequent whereas the domestic remittances are maybe more frequent, but less in terms of quantity. So how do you see which type of remittances have played a bigger role in this setup? So in this paper, I didn't have time to go deeper in the result, but we split off the remittances according to the origin. So we do find that international remittances have higher impact, mitigation impact than domestic remittances. Okay. Thank you very much. I think we have lost Kibrom. I also had some questions for him, but let me return to William. So I was wondering if you've been doing follow-up work on this and maybe the time period you looked at this is quite a short term. So I'm wondering if there's been a kind of bounce back to the pre-pandemic levels in terms of these incomes and employment and so forth. So if you've been doing any follow-up work, it would be nice to hear about. Yeah. So the survey that I was referring to is monthly and it's been going on, it started in 2018 and it's going up until now. So we sort of have a real-time or near real-time sort of discussion of these results with our counterparts. And basically, what we found was that the bounce back was pretty quick. So it was just a couple of months before the labor market was sort of performing closer to where it was in the previous year, but that it took a very long time to get back to exactly the level that it had reached in 2019 at the same time. And for some subgroups, it hasn't yet fully recovered. So particularly among those who were self-employed before, there was a pretty quick bounce back among wage employed or formal workers and a much slower recovery for the self-employed. There are also some particular sectors that were much harder hit than others. So anything linked to tourism or things like hotels, restaurants, these had a much harder time sort of in the recovery than has been the case of, say, manufacturing or agriculture. So yeah, there's some compositional effects, but largely speaking, most of the recovery in the labor market has occurred. At least it's gotten back to where it was in 2019. The remittances are a different story. So there again, there's a question of sort of the composition of what's happened. So in the initial downturn, what happened was most of the migrants that we're referring to here are international migrants and they're working in Russia. And the ruble quickly lost value early in the pandemic. And so the value of remittances plummeted and people weren't sending back very money during that period of time. But when the ruble strengthened sort of in May, June of 2020, remittances started flowing again and then recovered their level and went past it actually as a sort of form of social protection, I think in the same spirit as some of the other materials that have been presented today. So we found something very similar, but the challenge has been that no new migrants can go abroad, right? So everybody who was abroad can continue sending remittances back and even those who had some of work disruptions while they were abroad continued to send remittances back, maybe digging a bit deeper to help our family back home. But because of the lockdowns, the number of migrants going abroad is much lower than it was before. And this has had a large negative impact on poverty reduction since that time. So this is a big headwind in recovery. We have a bunch on food security issues as well, but I've been talking a lot already. So maybe we can get to that if somebody has particular questions. Okay, yeah, I think we're running out of time a little bit, but I see that Kibrom is now back with us. So let me ask a question for him as well. So your context is rural ETOP, right? And the PS&P operates, has been operating there for a long time in these areas. So I'm just kind of curious about what do you think are the kind of channels for these income shocks? You know, the World Bank and other organizations have been predicting that this pandemic is severely affecting urban areas, maybe a little bit less the rural area. So kind what's your views about how the PS&P households are affected by through the agricultural channel or something else? Yeah, thanks for the question. And sorry, I dropped earlier again, so I'm having issues with my connection. So yeah, the pandemic can affect rural households in at least one of three ways, the way we see it. One is what we call the direct channel, that is because someone in the household contracts the virus. And as a result, income drops, household income drops. And this in the rural context would mean hired labor. So households would make incomes from agricultural income from hired labor would lose their incomes. And the second channel would be the indirect channel, even if one does not contract the virus, the fear of contracting the virus alone would induce one to withdraw their labor supply from the marketplace. And as a result, income drops. And related to this is the issue of restrictions, government restrictions. As a result of the range of restrictions that were put in place, people cannot just access markets. So I have not discussed this here, but in our data, we have a lot of qualitative questions where we ask households how they were affected. And roughly 25% of households say it's because of closure of markets, because of government restrictions, they cannot access market, they cannot sell their products. And you see this in the variety of items people consume. So in rural areas, you see the consumption of perishable goods going up. The reason is because people cannot access markets, so the goods they would otherwise sell out in the marketplace they consume. So it is just, for the most part, the channel is through this marketplace, restrictions and because of kind of withdrawn labor, labor supply, household income drops. That's how we see this working. And of course, there is the, so just briefly to touch on the value chains, additional value chains. So there was pretty much a collapse of it for a bit, but in the Ethiopian context, it was just reversed pretty quickly. So it bounced very quickly, didn't last too long. So we'll have the three channels we see. Thank you very much. And so we have basically one minute left. So it's time for me to thank the presenters. Very exciting work. I'll be following up what you guys will be doing in this front in the future. And thanks a lot for our audience as well. Some of you have to wake up very early for this, including the presenters. So thank you very much and enjoy the conference. Thank you. Thanks, everyone.