 Welcome everyone to this session on Peace and Conflict. We are running alongside a series of other parallel sessions, so we'll give a slow introduction so that we allow people to join in. In these sessions, as you've seen in the program, we have 45 minutes and we have three presentations. It does look like there's five speakers, but they're joined. They go with us. And the presentations are going to be around the relationship between pandemics and violent conflict. So we'll have all presentations are prerecorded. So thank you very much for everyone for having done that work beforehand, which helped us manage the time given all the parallel sessions going on at the same time. We have presentations by Reynard Sexton, who unfortunately will not be able to join us today. The discussion is clashing with a lot of teaching and Labor Day in the US and a series of other issues. So Reynard has sent us this prerecorded presentation. Then we have Collette Salemi presenting joint work with Jeff Bloom and Ada Gonzalez Torres, who's presenting joint work with Elena Esposito, who will try to also join us a bit later. So we will start and I'll introduce the speakers as one at a time. So the way we sort of thought about running these will have the three presentations, one after the other, and then we'll leave time at the end for Q&A. And may I ask everyone, if during the presentations or afterwards you have any questions, how would you please add them to the Q&A tab and not to the chat tab because it's slightly easy to manage because the chat has a lot going on there as well. So if you are to the Q&A, I will be monitoring them and we'll be asking the questions on your behalf. If by some chance we have a bit more time at the end, I may unmute a few people but let's see how it goes so far in a warning that did not happen because of lack of time, but we'll try. So Evo will be kindly uploading the presentations and the first one is by Renard Sexton. Renard is an assistant professor in the Department of Political Science at Emory University. He works on the relationship between conflict and development and the presentation that we're going to upload from him is looking at how government capacity improves the ability of governments to improve pandemic risk reporting in conflict areas. So Evo, if you don't mind loading up the presentation. Thank you very much. Hi everybody. Thanks for giving me the chance to present this paper to you all. And I'm sorry that I can't be here live. So you'll just have to settle for the recorded version. So today I'm going to be talking to you about our paper, sustained government engagement improves subsequent pandemic risk reporting in conflict zones. This paper was recently published in the APSR. And so we'll be talking about some of our results and how we carried out this study. It's a joint project with Dotonheim from Florida State and Nico Ravenelia from UC San Diego. And we have a link here to see the published version of the paper. So going back to 2004, there was the massive Indian Ocean tsunami. And this, this disaster was really transformative in this, in this region, big impacts throughout Southeast Asia. And we look at two different areas that were both impacted by the 2004 tsunami that were at that time experiencing conflict. What's interesting is that the disaster pushed into very different directions. So in the context of Ache, the ongoing violence between Chinese insurgents and the Indonesian state, the tsunami experience and the buildup afterwards or the reconstruction process afterwards is widely considered to have pushed towards peaceful resolution of that conflict and, you know, transformation of that conflict towards peace. In contrast, in Sri Lanka, in the ongoing fight between the Tamil Tigers and the Sri Lankan state, the conflict or the disaster, the tsunami disaster is regarded to have pushed the conflict even more deeply into conflict. So what we found is that these crises responded to a disaster in two very different directions. So the conflict could become worse or become better as a result of the disaster response. And when we think about disasters, you know, these we think of as quote unquote natural disasters, you know, hundreds of millions of people are affected every year. And many of the deaths that occur as a result of natural disasters occur in conflict affected areas. And so, you know, a couple of the things that we're trying to understand here is there is this issue in conflict zones where governments really struggle to communicate with insurgent groups, communities in these conflict affected areas are often skeptical of government. And so there's a real lack of information flow and slow responses to these disasters. So what we wanted to find out is how can governments improve community participation in the wake of these large crisis situations. And our thought is that the lack of cooperation with governments sponsored crisis response is a symptom of larger issues of trust. And as a result, citizens and especially local leaders are just not very incentivized or not very willing to share information to governments even when crises take place because of the lack of legitimacy. But the literature to date suggested by delivering economic services and improving things, you know, governments can make local communities trust them more. And so there's this kind of counterinsurgency literature that suggests that this is possible. Now the question is, would this actually hold for emergency scenarios, crisis scenarios? It's possible that a crisis such as the COVID-19 pandemic could totally swamp out and overwhelm any confidence building efforts. On the other hand, it could be that it generates a sort of rally around the flag effect that makes people all want to cooperate. So the context that we're going to look at for this is the conflict affected region of Bicol region in the Philippines, where there's been a communist insurgency ongoing since the 1960s. And this zone, this map here shows where Bicol region fits within the Philippines. And over the last five or 10 years, which percentage of local villages have had a strong NPA presence. So you can see that there's been quite a bit in these zones, especially in the farther flung regions. Now, we were carrying out an RCT trying to build relationships between government services and these conflict affected countries when the COVID-19 pandemic hit. And so what we wanted to understand is as this crisis strikes the Philippines in March, lockdown is announced by the president's office. We want to understand whether the intervention that we were carrying out called Ushoptayo, which was increasing service provision in these conflict affected areas, would it actually build trust and help to respond to this crisis more effectively? And, you know, even today, frankly, information sharing is really important. But in the early days, sharing information was really, really valuable as government tried to determine how to respond to this pandemic crisis. So the program that we were implementing, Ushoptayo, was a combination of the Philippine National Police, the Region 5 from Bicol Region, regional headquarters, the Philippines Department of Social Welfare and Development, and other civilian agencies that provide services to local communities. The idea was, on a monthly basis, these agencies would convene with local village leaders in these conflict affected areas to try and figure out what's, to try and match villages with services that these government agencies could provide. And the idea was to build on some of the best practices that have been learned from the literature about what appears to work when it comes to these kind of community-oriented service delivery efforts in conflict affected areas. So they were supposed to be small. They're supposed to be feasible. They're supposed to have repeated interactions. They're supposed to have a lot of work done, a lot of work done, a lot of work done driven. And just to give you an example of some of the things that these villages were able to receive through the Ushoptayo program. They were all new services for these villages, but they're programs that pre-existed. So these were long-standing government efforts that were just not making their way to these conflict affected areas. And some examples are getting job training, getting culture, getting assistance in starting your own small businesses or graduating from high school. A really successful one was a security guard program in which high school graduates from far-flung conflict affected Barangays were able to get licensed to become security guards and then go essentially be mall security in some of the more urban areas, which was a really good job. It's well-paid and you can send money back home. So we did one full year of testing, starting in 2018, before we even rolled out the program in 2019. So to give you a sense of the region, as I showed you, we did randomization, both at the municipality level and then at the village level. So within all of Bicol region, we excluded municipalities that were very, very remote or were too affected by conflict to be included, which only excluded a relatively small number. And we ended up with a total of 80 municipalities. We then randomized half of those municipalities to receive Ushoptayo and those ones are in dark blue here. And then the other 40 to be in the control. Then within each of these municipalities, we randomized five villages to receive Ushoptayo and five to be in our control group. Just to zoom in so you can see. So within all of these treated municipalities, you see that there are five in red and five in white. And that those show you the treatment and control villages. So altogether we had 200 treated villages and 600 control villages. We carried out a baseline survey back in 2019. We got information from all the village copy funds and also the, the youth chairs. We included various endorsement experiments to understand people's attitudes towards conflict, even though those are very sensitive. So you don't want to ask them directly. But you do not want to ask them directly. So that's it. And so we wanted to do some general analysis and experiments to gather that information. Now in 2020, after about six months of operating, the Ushoptayo program was halted because of the quarantine order because of the pandemic. Now the interagency region five task force on COVID. Which had the PNPD, SWD, all of our partner agencies. allocate, you know, limited testing resources and other public health resources. Now in these conflict affected areas, they really had a difficult time being able to reach the population. So they sent out a community health survey by a text message to the leaders of all the communities in the conflict affected regions. And these are the three questions that they were asking. Basically, did they have high risk individuals? Had people visited Metro Manila recently? That's where all the transmission of infection was taking place. And finally, are there anybody people with symptoms? And as of March 2020, these are the best information or the best questions that the public health authorities could ask at a time based on what they knew. And the PNP, our partner, collected this information on behalf of the task force in the conflict affected areas. And so what we wanted to know, we asked the PNP, could they share aggregate data, which they're happy to do? We wanted to know whether copy tons from these conflict affected areas actually responded to this request for information, which was so critical for them, you know, taking decisions. And we found that overall, about 53% of village leaders responded to this request from the task force to share those three pieces of information about COVID risk. What was interesting though, is that in our treated areas, those places that were receiving services and participating in USEPTIA, 61% of village leaders responded. Whereas in control areas, only 51%. And this is to summarize the effects. So we see at the village level about a 10 percentage point increase or about 20% treatment effect from our intervention, the USEPTIA intervention, which had been going on for about six months. This chart here summarizes that the first and second coefficients use the village level randomization. The third one here uses the municipality level randomization. And you can see the effects are quite consistent, which is good to see. Also, we did a check of spillovers by comparing control barangays, control villages, in the control areas versus control barangays in the treated areas. Now we would expect there to be no difference between them if there are no spillovers. So we're happy here to see that, in fact, there weren't any spillovers. So this helps us to feel confident that our randomization actually worked and there was not interference between our units. So to understand what was driving these effects, which are really quite important, that USEPTIA seems to have incentivized these village leaders in conflict affected areas to participate and collaborate with the government in dealing with this crisis. Which kinds of people did this matter the most for? And we looked at four different characteristics at baseline. The support for the rebels, how much they trusted the government, whether the government has capacity, and lastly, whether you need political connections to access services. So first, using an endorsement experiment at the municipal level, we looked at municipalities that were positively disposed towards the rebels, versus those that were negatively or neutrally disposed towards the rebels. And we found is that the biggest treatment effect, that is where USEPTIA pushed people the most to share with the government, probably to control, was in these places that at baseline were quite positively disposed towards the rebels. Similarly, or sort of additionally, we found that it was place that had medium levels of trust in the government. It wasn't that they totally trusted the government or totally didn't trust the government, instead the effects were primarily in these zones, which had middle range trust in the government, where the effects were the strongest. We also found that the effect was driven primarily by Barangays, where at baseline, the village leaders did not believe that the government has the capacity to meet their needs. The places that already believed that the government has capacity, there was no treatment effect. And lastly, we found a small difference in that places that believed that you needed to be connected to get services, rather those that didn't believe that you needed connections, we found the treatment effect was a bit stronger. So we have evidence from these regressions and also some additional stuff from the paper that really was updating of beliefs by these conflict affected villages, the leaders of those conflict affected villages, regarding the capacity of the government to meet their needs. In earlier years, the NPA, the rebels really emphasized the villages that the government can't help you, they can't meet your needs, don't rely on them, rely on us instead. So people had a relatively low baseline belief that the government could deliver for them. And it's actually in those places where people change their behavior the most to cooperate on the on the COVID information collection as a result of who's up dial. We also rule out a bunch of other mechanisms in the paper. So it's not because of a change in security. It's not because of capture of the program by the rebels. We also have some evidence that, you know, the overall evidence here is that investing in this soft public services, these relatively small programs really can help when crises occur. And it's really important to note that this program seems to have worked by convincing those people that were most skeptical of baseline. And, you know, certainly as of when this paper was published at the beginning of 2021, relatively little experimental work had been done when it comes to social science on how COVID-19 mitigation can work, especially in the developing world, and especially in conflict affected areas. So with that, I'll close and say thanks so much for listening to this. As I mentioned, the paper has been published, but the experiment is ongoing. So feedback is really, really welcome. We're hoping to actually collect, we've continued running Usubtayo as a remote program. That's had a lot of challenges, especially because these conflict affected areas don't have great cell phone access in some cases. And so it's been tricky to do that. We're still going to be collecting data for the next year or so. So any feedback will be really welcome. We also think that there's some relevant policy implications. So any and all comments are welcome. So thanks very much. I appreciate your time. And it's great to be here. Okay, perfect. And like I mentioned before, any questions, please add them to the Q&A. And if they are directed to Reinhard, please do feel free to still add them on. And I will pass it on to him because as you heard, we would really appreciate that. So now we move to Colette Salemi, who's presenting a paper with Jeff Bloom. Colette is a microeconomist and she's a PhD candidate at the University of Minnesota. She's working on the nexus between conflict migration and natural resources in developing countries. And Colette and Jeff have a paper and well, it's actually an update of an earlier paper on the effects of COVID on a series of conflict patterns. So if you don't mind loading up this presentation too. Hello, my name is Jeff Bloom and I am a research economist at the USDA's Economic Research Service. And my name is Colette Salemi. I'm a PhD student in Applied Economics at the University of Minnesota. In this presentation, we will be sharing an update on our descriptive work on conflict events and the COVID-19 pandemic. The relationship between the COVID-19 pandemic and conflict is theoretically ambiguous. On the one hand, the pandemic has reduced local incomes, and in turn, the opportunity cost of engaging in violence. This could increase conflict. On the other hand, the pandemic has reduced the value of some natural and physical resources. This could reduce conflict. In addition, disruptions to food supply chains, especially at the onset of the pandemic, could lead to higher levels of conflict. In this presentation, we will be updating our descriptive analysis from our paper, COVID-19 and Conflict, published earlier this year in World Development. We specifically document time series trends in different types of intergroup conflict and we discuss five quantitative case studies in India, Syria, Libya, Lebanon, and Chile. We use data from the Armed Conflict Event and Location Data Project from July 2019 to, in this update, July 2021. We analyze daily counts of violent events, including battles, bombings, explosions, remote violence and violence against civilians, as well as demonstration events, including protests and riots. After an initial dip in conflict events in the days immediately following the World Health Organization's declaration of COVID-19 as a pandemic, daily counts of conflict have at least entirely rebounded. In fact, in 2021, the daily counts of conflict may be higher than they were prior to the pandemic. While violent events do seem to be declining, it is hard to attribute this decline to COVID-19 as the trends appear to predate the onset of the pandemic. Here we show the trends over time for battles, remote violence and bombings, and violence against civilians. This initial dip in conflict events is even more dramatic when we look specifically at trends in protests. After a very steep but brief dip in mid-2020, protests events bounced back and almost doubled their pre-pandemic daily count. This is consistent with the idea that the initial national lockdown slowed protest events, but frustrations with political leadership may have increased since the initial months of the pandemic. We do not see any noticeable trend in riots, which are much more scarce of an event in accolade data. In the initial months of the pandemic, India implemented one of the world's most strictly enforced national lockdowns. This lockdown strained millions of migrant workers in urban areas with little access to food or social support. We see a short-term reduction in all types of conflict events, and this corresponds to a short-term fall in protests. There is also a sharp spike in violence against civilians and riots corresponding to the lockdown. Conflict trends in India have mostly rebounded to pre-pandemic levels by July 2021. Syria is within a decade-long civil war. In March 2020, Turkey and Russia negotiated a ceasefire agreement over the Idlib Governorate of Syria. This ceasefire was unrelated to the COVID-19 pandemic. It appears that the ceasefire led to a sharp reduction in all violent events. Through the summer of 2021, conflict has remained at relatively lower levels than in past years. But bombings have continued in the Idlib Governorate in 2021, with numerous actors violating the terms of the still active ceasefire agreement. In Libya, the pandemic overlapped with the war between the Government of National Accord, or the GNA, and the Libyan National Army, the LNA. In April 2020, the LNA called for a pandemic-motivated ceasefire, but the GNA refused. In late April 2020, Libya imposed a 24-hour lockdown and then 10 days of curfews. After a small dip in conflict in early 2020, we see a steep rise through March 2020. This increase persists through the WHO's declaration of COVID-19 as a pandemic. Conflict peaked and eventually declined as the GNA victory sent the LNA to retreat. In October 2020, the two factions signed a permanent ceasefire, and violence has remained low ever since. In late 2019, protests motivated by government corruption and a deep financial crisis erupted throughout Lebanon. The country implemented a national lockdown on March 15, 2020. The lockdown does not seem to have led to a reduction in demonstrations. Daily counts were already falling in the preceding months. Despite the lockdown, there was an increase in demonstrations in March through June 2020. Protests and riots spiked again in the spring of 2021, as the Lebanese pounds value hit historic lows. Recent protests also took place around the anniversary of the 2020 Beirut port explosion. In 2019, Chile experienced a dramatic escalation of civil protests. Ultimately motivated by increasing economic inequality and political representation, daily accounts of these demonstrations were declining in early 2020, but began intensifying in March 2020. On March 13, the Chilean government banned public gatherings of more than 500 people. This led to a sharp decline in both riots and protests after the national ban, which have yet to return to pre-pandemic levels. Across all Acolyte countries, we see a short-term decline in intergroup conflict in the initial months of the COVID-19 pandemic. This trend is more due to a U-shaped trend in protests than any other systematic trend in violent conflict. We also document critical heterogeneity across types of conflict and countries. This heterogeneity is important, but makes causal inference tricky. How do we deal with confounding events? What is an appropriate control group? In many contexts, we need more data on income in COVID-19 prevalence, but some progress has been made. Berman et al., in a 2021 article, finds similar results to our descriptive trends, but using much more sophisticated estimation approach to deal with the challenges of endogeneity in the relationship between COVID-19 and conflict worldwide. Thank you so much for attending our talk. We look forward to your questions and comments during the discussion. Thank you very much, Collette. Again, any questions, please add them to the Q&A. I had the pleasure of seeing this paper being presented last year, and it's really good to see the update, and it's a really valuable work. But now we're going to go back in time a little bit, and we have Ada Gonzalez Torres, who is going to talk about a previous virus outbreak and the impact on violent conflict, talking about the Ebola virus in West Africa. Ada is an assistant professor in the economics department at the Ben Gurion University of Negev in Israel, and her research focuses on the social aspects of epidemic outbreaks and the causes and consequences of violent conflict. So absolutely ideal for the panel that we put together here. Ada also has a pre-recorded presentation, and is with us here in the panel and available for questions afterwards. Thank you very much and if you don't mind again. I'm presenting a paper on epidemics and conflicts based on evidence from the Ebola outbreak in Western Africa. This is co-authored with Elena Esposito, the University of Lausanne. We study whether epidemics lead to conflict, taking the case of the 2014-15 Ebola outbreak. In particular, we study civil violence, riots, protests, and violence against civilians. We also study what drives this effect related to trust and the provision of public goods. We also provide evidence of long-run impacts on violence. The West African Ebola outbreak in a nutshell. The 2014-15 Ebola epidemic is the largest in the history of the disease. It caused around 30,000 infections and over 11,000 deaths. It hit Guinea, Liberia, and Sierra Leone for the first time. You can see here the timeline, it's one epidemic. It hits for the first time in early 2014 and ends around 2015. For the few cases, it's still recorded in 2016. The case fatality rate is of 40 percent and the median age of death is around 30. The overall epidemic burden is 500 deaths per million in the population. This is led by Liberia with 1,100 deaths per million people. This burden is similar in magnitude to the countries that were most hit by COVID, but it's orders of magnitude higher compared to the COVID burden in Africa so far for a similar time period. We focus on Ebola infections. This is the relevant measure because Ebola is very rarely asymptomatic. It's a brutal disease. People may die within one to two weeks since symptom onset. Ebola survivors on average have moderate to severe long-run consequences, including loss of sight, premature death, and other consequences. You can only transmit Ebola if you already have symptoms and it's transmitted only through body fluids, so it's not airborne. Here's a snapshot of the data. Ebola hit for the first time in Guinea at the intersection of Sierra Leone and Liberia. You can see in June 2014, it was already spreading and there are a few similar violence events. One month after, Ebola spreads more and we see more civil violence events, especially in places with Ebola cases. Now, whether this is causal or not, this is the main question of this paper. In general, if you want to study the impact of an epidemic on civil violence, the main identification issue is that violence could exacerbate the spread of disease and weak institutions or poverty can facilitate the spread of both. We address this using an empirical strategy that is a difference in difference design combined with an instrumental variable strategy. So we're going to look at the incidence of conflict in places that later on are going to be more or less hit by Ebola, pre- and post-epidemic. The identification lies on the fact that the timing of the first Ebola case is random. It's the contagion between a bat and a human being and all other cases are human to human contagions, largely driven by the geographic distance to the first case and an apparel trans assumption that we're going to test empirically. We also propose an IV for the overall epidemic burden to address this possibility that there's post-treatment selection into high Ebola areas being driven by social unrest or other time-varying confounders. The IV is the geographic distance to the first case, interacted by a post-treatment dummy. It has an advantage that, well, it predicts Ebola, but also we can test whether it predicts conflict before Ebola starts and it does not. And it is a geometric distance to the first Ebola case as opposed to actual travel routes or transport systems which may be related to conflict or variables related to conflict. The estimating equation is a difference in differences and continuous treatments where Ebola total is either the number of Ebola cases measured at the end or a dearly quarters and conflict is conflict in your early quarters. And this is a post-treatment dummy. We're going to show you results for flexible differences designed where instead of the post-treatment dummy, we have a time dummy for each yearly quarter before and after the epidemic hits. And this will allow us to look graphically at the prior trans assumption. Here you have the main results. So you see here the coefficient for each quarter time dummy before and after the epidemic hits. You can see that there's no difference in conflict incidents in places that later on are going to be more or less hit by Ebola before the epidemic starts. Once the epidemic hits, we see a significant rise in civil violence driven by places with high incidence of Ebola. And this is starting especially in the second half of 2014 which is when Sierra Leone and Liberia get hit which are driving the results. The regression results suggest that a one standard deviation increase in Ebola incidence leads to an increase by 8 to 16% standard deviation in conflict during the Ebola epidemic. Back of the envelope calculation suggests that the overall increase in conflict due to the epidemic is about 40% in conflict incidents from a baseline of 100 civil violence events in a year. Here you have the regression results. You see the over less coefficient less coefficient which was written down in the last slide. You can also see the reduced from effect of the distance to the epicenter and the fact that we have us from first stage and no pre-trans. The impact of the epidemic on civil violence is a robust to different specifications and tests. Here I only lay out a few and I want to draw your attention to the first. So we also run a high frequency panel with fixed effects at bi-weekly level and this addresses the concern that newspaper reporters move into high Ebola areas in order to report conflict. It's very, very unlikely they can do this in two weeks and also they don't know with that accuracy the number of Ebola cases live. So once we've established that the epidemic led to civil violence we want to know why this is the case and when we expect this to be the case. A standard model of conflict with predicting ambiguous effect of the epidemic on civil violence because there are some aspects of the epidemic that lower the opportunity costs of rioting. For instance the negative income shock and there are other aspects that actually lower the benefits of engaging in violence as for instance the fact that you have increased risk of contagion. So in theory we expect an ambiguous effect. However we suggest in this paper that there is something that we do know that an epidemic will change systematically and that is the demands for the state. In particular an epidemic increases the demand for a state to hold contagion and the state can do this in many ways. It can provide hospitals to treat patients, can enforce quarantines to avoid new contagions or vaccines with the same purpose. Whether what the state does is proceed as good or bad will depend on trust in institutions and on what the state eventually does or does not do. So hospitals for instance may be good or neutral independently of trust in institutions versus the other measures may be more dependent on how much you trust your state. To study this we're going to look at heterogeneous effect of the epidemic on conflict by pre-existing levels of trust and we're going to run an event study design based on the provisional public goods. So for the first just looking at the impact of the Ebola epidemic on conflict incidents in places with below versus above mean level of trust what we see is that the effects entirely driven by places with low trust and there is no impact of the epidemic on conflict in places with high trust. In terms of the policies we see that Ebola treatment units which are hospitals they we observe a lowering of conflict in places that get to the Ebola treatment units after they arrive and we see a rise in conflict in places that get area of locates or discipline teams in places where they arise. When we split this in low and high trust areas what we find is that establishment of Ebola treatment units or hospitals to treat patients lowers trust both in both areas but the effect is significant in places with no trust. This is consistent with the model in which in high trust areas there's only one equilibrium of no conflict versus in low trust areas there are two potential equilibria with conflict and without conflict. When the epidemic hits and the state reacts by establishing hospitals you're more likely to go to the low conflict equilibrium. For district quarantines we have divergent effects in high trust areas and low trust areas in places with high trust. People trust the state that these coercive measures are helping in halting the epidemic there is if anything there's a lowering of conflict in low trust areas on the contrary these measures are interpreted as coercive and you see a rise in civil violence in these places. Finally we also look at long-run impacts on conflict and we find that the places that are going to be more hit by Ebola in those 2014-15 outbreak see no difference in conflict in the years prior to the pandemic outbreak and the however they see a rise in conflict both during the outbreak and years after the outbreak and it's suggesting that there are long-run effects of this epidemic conflict relationship. To conclude this paper suggests that epidemics may act as a critical juncture when they arrive there's a raising demand for public goods and for response from the state now depending on what the state does and depending on underlying levels of trust this leads to multiple equilibria possibly leading to civil violence and this has long-run consequences for affected areas. Policy implications of these findings is that building trust is crucial to halting the epidemic and break cycles of civil violence. Thank you very much and also to your co-author so we have a few questions in the Q&A but before let me just give some of my taking so these are three really important papers and they're some of my favorite papers in this very small area of research but growing and I presume it will keep growing sadly and I think there are two main takeaways. I mean the papers clearly show that something that is fundamentally a public health adverse shock has political implications and even everything that's been shown across all these different case studies it doesn't it doesn't bode well because we can only assume that some of these trends will carry on as the pandemic develops. The other important takeaway I think that's shown in all the papers is this idea that crisis responses are very very important because they crisis are interlinked so something that has happened in the past will have implications for future ways of dealing with pandemics and this kind of brings home this kind of cycle of government intervention and so forth and in particular what seems to be shown both in the first paper and the third paper is that trust and socially political trust are the driving factors whereby one shock has implications for a second shock happening much later so these are very very important results and really I think should raise a number of questions but I have a few but let me first go to the Q&A and ask the questions that are there. There seems to be two questions for Colette. This is by Adwesha and the question asks the conflict seem to be very different in nature varying from demonstrations to violent conflict have you analyzed the data from two separate types of conflicts separately and that there's a related question which touches upon the issue many demonstrations were actually anti lockdown have you been able to separate these demonstrations from other political demonstrations with a show ongoing serious problems in these various countries so the underlying social tensions that might have been there even without the pandemic I assume. Yes wonderful questions the first one I might need a bit of clarification because we did try to disentangle and I think you know our presentation tried to convey the cases in which we were looking explicitly at protests which often in our data means protests that were predominantly peaceful rioting or versus riots which are often coded as less peaceful protesting and destruction of property so we're trying to as well as on the other hand different types of violent events so we disaggregated as much as we cope with the data that we have from Akled. The challenge with disaggregating further to your point about the type of protest is often that information isn't complete within our data so I think you're totally right it'll be very interesting to just separately look at just trends in the anti lockdown protests when did they happen. I think we may have lost Colette with just me. Okay let's maybe Eva you could contact Colette and see if we can solve the technical problems in the meantime well let me move to Ada. Ada I actually had a question about your paper obviously looking at Ebola very important results. What do you think are the lessons for the current COVID-19 pandemic so Colette you back again we're just moving to Ada and then I'll give you the chance to go back is that okay. Yes I don't know what's happening I never lost your audio but I know I disappeared so please continue. Okay let me ask Ada the question then and then we come back to you so Ada I was asking what do you think are the lessons for COVID-19 and whether your findings may provide some explanations for the pattern that Colette and Jeff are talking about in their paper and then after that we'll go back to Colette maybe she wants to add. Yeah thank you so much for the question I'm also actually curious to hear what also Colette thinks about this but so we do think that this can be applied to other pandemics I did try to convey like also the differences between like the two types of disease so Ebola like struck much more quickly and it hits mostly younger populations especially like the most in terms of deaths and so on obviously COVID hits of like all ages but like these could be differences that that may matter the fatality rate was really high but but but on the other hand also I think that the main maybe the main policy finding is is really this difference between a low and high trust areas which which can be applied also in the case of COVID in the sense that we expect mostly events of civil violence mostly in places with low trust and really building trust with between the state and civilians is key also not only to to avoid the spread of violence but also to ensure the the adoption of certain measures that can be perceived as coercive or or that maybe maybe misunderstood. Sorry I was muted that's great. Colette do you want to go back to your answer and also maybe you have some thoughts about what I just mentioned? Sure so yeah and apologies again for the internet issues. I to completely agree that there is a case to be made for doing a separate analysis just looking at anti lockdown protests and looking at protests they're explicitly in response to grievances around governance at this time because that is anecdotally what we what we've been reading a lot about it's mostly a challenge of of of actually coding that in the conflict event data that's that's really where the challenge lies but I completely agree it's something that's super interesting it just was a little bit too much work for this current project. Great yeah yeah that's that's understandable okay I don't see any more questions I don't know if anyone has anything to add we actually over time but I am told it's not a big issue no okay so let me bring this panel to an end and thank you both for joining us and thank you for the participants for joining us as well if anyone is interested in more conflict related issues we'll have a coffee break at 6 30 p.m Helsinki time I will not attempt to move that in terms to tell you what that is in other time zones and we also have on Wednesday at 17 50 so 5 50 p.m Helsinki time we have an interview with Chris Blackman as well for those interested in conflict so I hope to see you either later or on Wednesday take care and thanks very much Colette and Ada for joining us thank you so much for having me to present nice to meet you all nice to see you too