 Hello, and thank you very much for having me here. I'm really feeling very honored to be invited here as a discussant. However, I have to say I hope I give justice to all the three presentation because I'm an anthropologist by training and come from a completely different field. Nevertheless, I feel that we had here three very exciting and mind-expanding inputs, papers. And I'm very happy to carve out number 34 issues I would like to discuss where I would try to bring the three presentations in connection with each other. The first, I think, what we have always to consider is this strong narrative we have on climate change and its negative repercussion impacts on society. And I think this is a narrative which is very strong in our mind, which we all more or less believe. I think, Tim, basically, in the morning keynote, you also mentioned this again that this is particular for academics something we always have in mind. The interesting fact is that we hardly, we then change to research. We hardly see that we can observe this in research. Or put differently, what we observe in research is that the data does not provide us with the evidence that climate change has on conflicts or negative repercussions in society. And the reasons, therefore, are manyfold. And we see this very often. There's one researcher who is making one argument and the other one is coming directly with the counterargument. I think one issue is particularly if we're talking about conflict-prone countries is that we have hardly data of, really, the countries where we see violent conflicts. It's not by accident that Gracia mentioned that you're doing fieldwork in the Philippines, Kenya, Senegal. There are always the typical countries where we are working. But what about countries like Afghanistan, Iraq, Somalia? Mentioned Somalia, I think was really brilliant. South Sudan. For most of these countries, we hardly have any data. So I think we have directly one of the main problems is that for most of the conflict, violent conflict countries, the data is missing, particularly if you're then thinking about future research in these countries. Second, also mentioned by Gracia, I think that we have a high complexity of factors, this non-linearity, and this strong contextualization of conflicts, which makes it very difficult to understand what has really an impact on conflict situations. If you just read the paper by Katherine March of 2019, she shows that it's all about poor development and bad governance, and that climate change hardly has any impact on conflict situations. So it depends very much on which indicators you're using, which variables you're using to understand the complexity. And of course, climate change is a long-term process which we can hardly understand the impacts if you have a short-term process of a violent conflict of about eight to 10 years. So this is, I think, one problem we are facing here. And this brings me, of course, to the very stimulating and also provocative talk of Marco and Gabriel. Thank you very much for your paper, which just tries us to show this direct correlation. I think it was really very fascinating. But of course, it also directly provokes a kind of objection. I think that Marco, you're also more looking for these questions about where it works and where not. And of course, you're just thinking about countries' case studies where it's not working, where it's working, and I think this is something I'm sure that some of you will come back with your own cases in mind. I was also asking my question, do you follow more kind of a relational approach of the average deviations of temperature? And I asked myself, to what extent you can identify certain tipping points. They can really say that they say beyond 35 degrees or so on, you have changes. And of course, I think there will be the main critique of a kind of a deterministic view. And I think that this will be really something where, in particular, if you're just following more my anthropological view, you have case studies very much on the local level, which will come up with other results, but doesn't mean that your research is not right. Second point I would like to make is that I find it very fascinating, although scrolling through the three papers here, is the question of terminology. On the one side, we are talking about climate change. And on the other side, we are often talking about conflict, about security, about peace. But these are three completely different terms. And I think we have really to understand what we are talking about. And I would like to unpack the three terms in the next five minutes, because I think that this really matters. If you talk about conflicts, the main argument is that natural resource scarcity leads to more conflicts. I think this is the overall idea we always have in mind. In particular, that climate shocks, droughts, floods, and so on lead to conflict. This is also what you try, Garcia, to show in your studies. And of course, what we see is that particularly on the communal level, on the local levels, countries which are exposed to drastic climate change are the ones where we see most of the violent conflict. So this could be an argument which would bolster the idea or this interconnection between climate change and conflict. However, there are also many, many counter examples. And this is particularly that with resource scarcity, we observe a much stronger cooperation. We'll just give you an example from my own research in Northern Kenya where I'm doing studies on pastoralism. And in Northern Kenya, the same with Somalia, we observed no extensive droughts for the last three years. And what we can observe there is that the cattle rates, the cattle wrestling is dropping down over the last three years. Because this is the main form of violence you can observe here. The reason is that in times of drought, the pastoralists are starting to share their pastures, are cooperating together, because if you have a cattle which are under poor conditions, you can, they're even more burdened than you can sell them or do something with them. So there we can observe that only in rainy seasons, the cattle rates are increasing because then the cattle are made of use for the communities. So you see it's more or less the opposite, that more or less in situations of droughts, the conflicts or the violence is dropping. And I think there are many arguments, particularly the really focusing on the local levels where you see that our ideas we have in mind with climate shocks that the opposite observes. Let me come to the second term, security. I think here, I think we have to be very clear about what we are talking, because security is a very vague term. And I think what I learned here, it's important to differentiate between this kind of everyday communal, local security. You can go back to all this nice livelihood approaches on the one side and on the other side, the national, international security debate we have. And for sure, what you can observe currently is a circuitization of the climate change debate. And hereby that it becomes a very strong part of a national agenda, international agenda. And if you follow my argument on conflicts, where we talk about cooperation, what we see currently in the circuitization debate is separation, that national interests are coupled with the idea of climate change. So it's not about cooperation. So more or less a fueling of conflicts as soon as climate change enters the security or the national security debate. Interesting enough is that we observe in the climate change debate that security becomes increasingly important. This was very fascinated Emily by your talk using the term of soft security. You could also think about why not using the term of peace, but you brought in the term of soft security which reminded me about the debate 20 years ago or 25 years ago on human security. It's the idea somehow to widen the security approach to expand security beyond hard security. And on the other side, to keep the balance between the nice, shiny, soft, human, soft, human idea rather than the hard security and to bring both debates together. But I think we all have to be really very serious about to what extent this works out in the end or to what extent you bring together two debates which should be separated. I think it's more or less an open question. The final one is, of course, on peace. And of course, we have a lot of debate on environmental peace building. And I think here again, as Emily has shown us, the whole idea of cooperation and trust is here key. But I think you also addressed rightly the question, how serious are we about this process? How serious are we about that we create an environment where this kind of peace building can take place? Again, some examples of my research from Kenya where I did a lot of research on NRT, the Northern Range and Trust Fund. It's a huge NGO with a high number of conservancies where they tried to bridge environmental peace making with conservancy and issues. What we see there is that in some areas, peace building approach between different communities works that here, possible, is brought together. We also see the opposite. We see many areas where the NRT is fueling existing conflicts and where we even see a militarization of conflicts. So I think we can see both. So I think it's very difficult here to really to say that peace building works in one or the other way or put in a nutshell, not all with all good things coming along together. I will leave it here. I'm really looking forward for this debate and John, I think it's now your turn to take over. Thank you very much. So everyone, you've heard from our speakers. You've heard from our discussant. Now we would like to hear from you. So it's time to have some questions. What we'll do, I think as the Norma gathers, we're gonna pick one side of the room and work our way across. Usually we seem to start on that side which seems a little mean. So this time we'll start over here and we'll start with the gentleman at the back, the gentleman in the middle and then the gentleman at the front for the first round. Thank you very much. Very exciting presentations. My question would be for Marco. So to examine whether you tested if your models hold, if instead of the deviation from the monthly temperature you do it for the deviation from the weekly or bi-weekly to week temperature. Thank you. Yeah, hello. I'm Darryl Sequera. Referring to the second speaker which who related rise in temperature to rise of intention to disagree or become more conflict ridden. I was a bit sort of concerned about this relationship between the two because actually what we imagine it seems to me and I'm not a psychologist but what it seemed we're really measuring is the psychological status of the community when the temperature rises. And so of course I know from personal experience when it's hot you tend to get a little bit more irritated and so on. And the first thing is you might fight with your close relatives or friends. But then you make up later on and you put it down to heat. But actually the consequence which apparently is seen to be as due to a temperature rise is actually a change in psychological status of the person and that is brought about by several factors in my opinion. So for example, when you talk about the impact of climate change we might only look at temperature but what we're also seeing what we ought to be also be considering is that in these arid areas that we talk about for example in Africa and other parts of the world we are actually at the moment at the point where there has been an accumulated impact of misuse of land, deforestation, over grazing, over population of livestock and so on. Over population of people I mean and also of livestock and so on. And so the consequence has been that partly due to this mentality of humans even in rainforest and better forested areas climate change and of course the use of fossil fuel and so on, climate change has come about and the places where it is impacting most is in the more marginal and arid areas of the planet. But even- Sorry to cut you off but that's actually in itself an excellent question so maybe we'll move to the third question if you don't mind, thank you. Thank you. My name is Daniel Watson I'm from University of Sussex and Akled as well. So I noticed the first couple of speakers used Akled data. I do wanna put a cautionary note that we would not support claims about climate change and conflict or climate change and riots being closely related. We do not see that in the evidence so I need to put that there first. So I think I'm a bit, I would like to echo some of Conrad's comments. So it seemed like you guys were using quite a Malthusian perspective to understand the relationship between climate and conflict. Generally building on Thomas Malthus's largely discredited ideas about environmental degradation, leading to scarcity, leading to population movement or migration as we call it today and societal collapse. Now that has at the time did not have much evidence to support it. It doesn't have much evidence now and I would again urge some caution. I just wanna make a couple of comments if I may. So Marco on some of the accounts you're building on a tradition of research where there was some research that says that when it rains less in Africa and Africans get hotter they fight more and this has often been interpreted as quite racist. I would be quite cautious about having this kind of biological determinism argument for riots particularly because in Africa the reason there's a deficit of trust can be for a whole host of factors economic volatility at the moment, democratic reversal and rise in coups can be some of those reasons. I'm aware that I'm probably over time so if there's another round I might ask another one. Thank you. So looking at this first set of questions Marco it looks like you're clearly on the hot seat but one of the things that comes up in our third question is this notion of in some sense responsiveness to climatic shocks in terms of how institutions might adapt or change in ways which might either exacerbate these climatic events or mitigate these and I think Grazia, you and Emeli actually in some sense gave us different examples of those. So I will let Marco go first and then if you guys wouldn't mind speaking to the last point that would be very helpful. Okay. Thank you so much for super interesting questions really wrecking my brain once you were talking to come up with something smart to say about them. About the weekly temperature anomalies that's a great point. No, we haven't done that. That's relatively easy to do. My guess would be that the results should be relatively stable just because so for example now like the average temperature in the first week of May is relatively similar to the average temperature in the second week of May. So in that sense, you wouldn't find massive temperature changes from one week to the other but that's definitely something that we could look at. We haven't looked at that yet. So it's a great point. Thank you. Sorry, best thing I can say is we will look at that. Then the second point, sorry, I was writing I wasn't checking the newspaper. I was actually writing down about psychological status. Yes. Yeah, that is unfortunately, it's kind of a trade-off because we really, I think it's a disciplined thing. I think economists, we just care a lot about kind of the identification of causal effects. That's kind of what makes us happy and what we are respected for kind of. That's where the literature is going. So in that sense, I think I'm guilty of trying to please the masses. So there's a clear trade-off of kind of looking at short-term deviations, which I think you're right. I mean, in a certain way, it's really hard to generalize. I mean, does it one day just because one day it's hotter does that necessarily mean that long-term climate change will do that? I don't know. So in that sense, it's a trade-off but on the other hand, if we're looking at long-run temperature changes, then it would be very hard to disentangle kind of the pure effect of climate of anything that you said, like operating via institutions, et cetera. So I think we just came down, we prefer than not being able to necessarily generalize it for the long-term but we are pretty confident that we can identify temperature change. The other thing I thought was an excellent point was maybe to distinguish between kind of areas like arid areas or areas where there's been misuse. I think in the protest data, we could, it would be really interesting to do that. So one thing we did is we looked at kind of cells. It's about 100 by 100 kilometers and we overlaid those with kind of GDP of the country and we found that the effect of temperature is much stronger on protests if the country's poor. So which kind of really fits with the attitudinal stuff. Plus the other thing we overlaid it with some sort of ethnicity maps and the effect is much stronger if more than one ethnicity lives in that cell. So which you kind of think those are the cells where conflict might be more likely to arise. But I think your suggestion could be done very easily by looking at kind of how arid the land is. So again, great points, thank you. Thanks a lot Dan and I come from University of Sussex as well so great to see colleagues. All right, yeah, I'm totally with you on your point in the sense that we don't claim that climate is a direct impact on conflict. There has been quite a lot of attempts and literature. I mean quite a few very, very good scholars have tried and attempted to find a direct link, causal link between climate and conflict. But our assumption and our point of departure is that we assume we think that climate has a role in these nexus. We know very clearly, I mean, CGR has been selling this for, I don't know, 50 years that climate is an impact on the socio-economic systems on food insecurity, socio-economic risks and insecurities that are fundamentally, this is the other part of the literature, recognized as drivers of conflict. We know that there is a connection between food insecurity, forced migration, inequality and the security side of the story. How do we put these two things together? There is inevitably a question, a research question there in understanding whether climate is a threat, a multiplier, it could really increase the likelihood of conflict that we find evidence for that. There is also a question of data. You know, the usual suspects of the data that we've been using are extracted from different sources. We can now be at the session just now thinking about how accurate and how relevant are some source of data, some type of data for the type of research questions that we ask. So we should definitely improve on that in trying to find small scale type of tensions that conflict don't necessarily end up in fatalities or they don't end up in larger scale conflicts that are more related to natural resources aspects. And then there are situations where climate conflicts are not related. I mean, at least we don't see the relations. In Sahel, for instance, that there has been quite, the Sahel has always been exposed to conflict, always been exposed to climate impacts. But not every drought leads to conflict. So what are these mitigating factors? And one mistake that we should really avoid and that we are trying to really push for this message is that we can not really only rely on what the data are telling us. The good tell us is that there's climate exposure in one area, but then conflict happen kilometers away from it. So there is high heterogeneity in terms of the occurrence of these two different, but very, very important phenomena. So we need to really integrate and use our best methods and approaches to integrate qualitative and quantitative type of approaches. And also we need to consider the other aspects. So some specific contextual factors that could mitigate the impact of climate and conflict. And as I said, not every drought leads to conflict in different areas. So instead of looking for generalizable statement, external validity of this research, probably this is the case, this type of research, where we don't want to do that. We want to be as contextual as possible, as localized as possible as current we're saying. In Kenya, in some instances, peace building strategy work, in some other instances, maybe the next village, this don't work. That doesn't mean that climate doesn't exist or conflict do not exist. And also the final point on the importance of the risk of conflict. Even if we don't observe conflict, it doesn't mean that the risk of conflict doesn't exist. And in fact, in the previous session, at the panel discussion, some of these were saying that rather than focusing on the incidence and number of conflict, we should look at the propensity or the risk of tensions or conflict occurring. There are situations where areas where they're controlled by armed group and they're relatively peaceful. But it doesn't mean that they are actually, there is peace and there will be peace forever. So these are my points. Thanks. Thank you. Well, would you like to add? Yeah. Yes. It's what I would like to also comment here is relation to these risks of the conflicts. Like for example, Europe now, the psychologists, they are working together with the youth researchers because what they have recognized is that when the youth, for example, they are so concerned now in Europe that what is the future? What is gonna be the day future? What is the future opportunities? And that actually creates the risk. That creates the risk of the conflicts, but it's also health risk. It's also the risk for our education system and so on so. And these type of climate related risks, we also need to understand that those are not necessarily the conflicts now now, but if the youth, for example, they do not believe that they have a future. It's creates the various type of risks in our society. Thank you. So we have time for a second round. We'll start in the middle here. Nothing will do the gentleman in the blue shirts, the lady behind you, and then we'll go to the back of the room. Hi, I'm Rodrigo Livera from Unwider. Thank you for the great presentations. I have two questions for Marco. So, Marco, the first one is, why don't you present or estimate results comparing instead of the difference between the average temperature? What don't you look to the standard deviation? So the deviation from the historical average, what do I mean? Because most of the papers in the rainfall literature look to this measure, and I think what these literature try to find or try to investigate are extreme events, right? So I'm not sure if when you look to one, two, three degrees from the historical average is in fact the extreme event, but if you look to the deviation from the historical and you can look to one or two or three standard deviations, you can have, in my opinion, better measure. And the second one is about the country level analysis because, for example, just taking one country, Mozambique, for example, this is more than 2,000 long kilometers country. So do you have any information or any analysis about the sub-national level because it's difficult to believe that there be an effect in votes for all the country, but if you have data for sub-national levels and changes in the temperature at the sub-national levels and also election data, maybe this can help in presenting this. Thank you. Hello, I'm Margarita Scorcina from the University of Florence. Thank you so much for the presentation, very interesting. And I have a few questions for Grazia, to Grazia. I really like the framework that you used. And I was wondering why you used the DHS data because we know that they are cross-sectional data. So I was wondering why, instead of some other longitudinal data, for instance, LSMS data. And also, if you think there could be some selection bias because we know that in areas where there are conflict, it's more difficult to collect data. So maybe this could create some selection bias. And also, if you look at the medium effect versus long-term effects of climate change on conflict. And the last point, if you also compare like ongoing conflicts versus new conflicts. Thank you so much. Thank you. My name is Jackson from Conflict Research Network in West Africa. I want to find that, as much as I understand the fact that the literature argues climate change has a multiply effect on conflict. And you had data sets on Nigeria. In the case of Nigeria, where we have multiple drivers of conflict, for instance, ethnic conflict, political conflict, as well as a national resource-based conflict. And the data on temperature and the incidences of conflict. I'm a bit worried. So at what point would you tell if the occurrences of conflict during the hot period is because it's climate change? Reason being that did you collect data when temperature was cold to compare when the temperature was hot? Because in Nigeria, we have lots of conflicts and crises all year long, from the Northeast to the Northwest to North Central to Southeast and South South. And these conflicts, ethnic, religious, political, protest, and it goes on all year long. So how do we know it's because of climate change? Thank you. So I think we'll give Marco a little extra time to work out some answers to these questions. So, Grazia, would you mind going first? Thank you. Thanks a lot. Thanks, Marguerite, for the great questions. We actually use on the data, the HS data. So we use both DHS and LSS data, depending on the country. Obviously, LSS is not available for all the countries and on that we have also LSS for Nigeria. So we've done the analysis for LSS data and it's pretty much consistent. So we find similar evidence with the advantage of having the agricultural production also in the LSS. So we also add another layer of the analysis to that. So that's a great point. In terms of the selection bias, so it's more difficult to collect data in conflict affected areas and so on, I mean, yes, absolutely, there could be. But we are looking at slightly different types of conflict that might not really affect the collection of these type of data. So we're looking at small scale conflict that they really reflect the relationship between groups that are affected by natural resources, carcities or related and food security. So we don't really look at countries where there is civil war or et cetera, which is something that we are going to do, by the way, with the new initiative that will be launched the next year. So looking at the Afghanistan and the Iraq potentially, I mean, we're still deciding, but yeah, we will be looking, but that is tricky, yes. So we need to be innovative and creative in a way we actually get the right data because at the end of the day it's all about which data we get. We, in our analysis, we look at, we try to consider the temporal heterogeneity of these phenomena. So we consider the climate anomalies and the temperature and the rainfalls that are happening before, other indices, that are happening before the observations, before the interviews of the households. So the climate is considered a previous lag, like minus one. Then for the conflict, we look at those conflicts that are happening after the interviews. So in the quarter or in the semester or in nine months after the household has been interviewed. And in order to also compensate, so the offset, the fact that we don't have panel data in DHS, we use a system, we identify grids using a special techniques, special approach, so we identify grids within the country and then we sort of reconstruct the panel through that. Hope I've answered your question. Thanks a lot. Thank you for that, your question about the standard deviation versus the kind of level. It is not something we've done. I know it's done a lot with rainfall because I think the standard deviation, rainfall is usually kind of the ideas that that affects agricultural output and then that might affect another thing. We, the reason why we just looked at levels because we kind of thought that our paper is already a bit, might be a little bit controversial. So we wanted to stick to the methodology that other people have used. So we kind of set everything like that grid. I actually downloaded that from another paper. The idea is that we try to replicate everything exactly like previous studies to kind of have a bit of a stronger argument. That being said, I think it would be really cool to look at whether the effect of kind of anomalies varies from kind of more volatile areas to non-volatile areas. That would be really cool kind of to distinguish, maybe it's stronger, you know, if you're already used to kind of changing climate, which might be answering a little bit your question as well because you could combine these kind of short-term fluctuations with kind of longer term trends. So I think that's a really cool real possibility to do. By the way, we do include rain as a variable. We find zero effect, which we've actually found pretty reassuring because we're really trying to measure this kind of direct effect. About the sub-national level, so we have two research assistants that worked a lot to digitize all the data and it was so much work just to do it on the country level. Actually, the data are there for sub-county level. So I think it took them already two months. I think it would have taken them a year and the idea is that this is not really a paper only on elections, but it's just a thing. But I have to say, I'm not gonna tell this to my two research assistants. Apparently these data already exists. I think they did everything which was already there, but I'll definitely have a look because you're right. And so it's just a broad brush. The idea was that the main analysis is on intentions to vote and it's just corroborating it with some sort of incidences. But thank you, it's a great idea. So let's see if we have... Yes, this is proposal for Rada, for you then. It's actually, once we talk about the risk and climate change, there is the one more layer of the risk and possible conflicts. Because once there is a lot of now the funding directly, for example, the climate mitigation actions, but also the offsetting. These, like a lot of political scientists, but also the studies related to political ecology, it's also concerned that what type of conflicts, what type of risk, these type of mitigation actions, offsetting actions. Because these are also changing the political setting. These are also the relation to land use. These are also related to ethnicity. And it's actually same time once we try to do the offsetting and through also the donor agencies, try to carry out the mitigation actions in various countries. We actually can cause the more conflicts. We can also be quite harsh throughout these, the blueprint projects to the local realities. So maybe that could be the one layer in your studies, which could be actually integrated there. Yeah, thanks for that. In fact, I mean, one of the main challenges that we have in researching on this climate security next is actually the research gap that we want to fill is precisely this. So we are aware that climate adaptation and mitigation can create maladaptation. So really affecting all those potential drivers of conflicts of what do we do about that? So we really need to make sure that our climate action becomes sensitive to conflict, to peace and security, and really embracing these concepts whenever we design the way we design the climate action. And really making sure that we don't really change the equilibrium, the balance, or we don't do harm. And in fact, I mean, the objective should be that while we want to increase climate resilience, we should also be able to contribute to stability. And this is quite straightforward in the way we're thinking about it, because if we don't have stability and fragility in the countries where we work on, then we can't really achieve climate resilience. So there has to be a combination of both. Thanks. So I would love to take more questions, but I've already been shown the yellow time card. And for those of you who are football fans will know that once you get a yellow card or red card is often not that far behind. So I want to end by with two points. One of which is that one of the things I really liked about this session were the deep hug complementarities that existed between what at surface looked like very disparate presentations. I think starting off with Grazia's presentation, emphasizing the notion of how complicated these linkages are. Oftentimes, particularly in the popular press, the argument seemed to go is that climate change will lead to conflict over resources, conflict over access to resources, and that will then spill over to conflict. And I think what you did in a very nice way is to suggest that things are much more complex than that. Marco then does an effective deep dive on one of these mechanisms. And I thought the results which were particularly interesting were the results about trust, the way in which, in fact, these extreme temperature events may make pro-social behavior more difficult, which then has implications both for the likelihood of conflict but also for the ability to negotiate conflicts over things like access to resources, which then takes us very neatly into Urmeli's presentation who talks about responses which we could develop to actually anticipate these potentially more conflictual environments in the future. And of course, this then leads back very nicely to some of the things we were talking about this morning in Tembezli's keynote address. So with all that, I would like to thank again all our panelists for their very rich presentations and discussions. Thank you very much.