 Good to see you all this afternoon. As Carlos Benson, I'm at the University of Oklahoma and today I'm really happy to present my ongoing work on the linkage between National Disasters and economic inequality. I should mention this caveat that this is very descriptive in nature the project so for those of us who are really used to looking into individual level data It's going to be a little bit different But I'm really hoping to get a lot of good feedback from you all and kind of work on this I'm also thankful to you and you wider for funding this proposal earlier on and I got some really good feedback earlier this summer as well, so the motivation behind working on this project comes from this stylized fact that forest fires are ongoing phenomenon in a lot of countries in fact, there's a study lot of studies actually indicate that fires affect an annual average of 19.8 million hectares of land globally and so At a micro level that has been an influx of studies that looks into what kind of adaptive actions Economic essence undertake to mitigate the negative effects of these natural disasters so in this study I'm mostly focusing on the incidence of forest fires and It's an interesting avenue for me to work on this because rather than looking to individual level accents I'm focused on cross-country level Measures and that's where the measure of economic inequality that Carlos mentioned which is coming from the weak database comes in really handy to look Into how are these natural disasters? Influencing different measures of economic inequality, so there's two primary objectives in the study first I'm interested in estimating the relationship between The intensity of forest fires and economic inequality. I'm able to take advantage of Variation across countries over time looking into satellite data on fire events And I'll talk a little bit about that in terms of measuring the intensity of fires I'm Exclusively using two different indicators one is the number of fire events reported in a given country during a given year The other one is the use of fire radiative power Which is essentially capturing the amount of energy that gets released in a specific fire event To major inequality, I'm focused mostly on Zini coefficient the paper Looks into other measures of economic inequality too, but for this presentation I'm mostly presenting results on Zini coefficient and then I also break down the estimates across different subsamples and hopefully that will have More policy implications as we think about what these estimates are really telling us so where I'm hoping to get more Feedback on at this point on this project is pinning down the mechanisms So I've kind of thought about you know few mechanisms that I'm implementing in this project But I'm not really finding a clear-cut analysis at that front. So we'll be good to get Feedback on that front. So let me just give you a brief preview of the results in a sense I saw that Zini index in ruler areas increases by roughly 13.72 percent and 22.02 percent for every additional unit increase in the number of fire events and The fire radiative power so to capture intensity. I'm looking into the measure of fire radiative power I also break down the sample into different groups, and I find that this effect is prominent across Upper middle-income countries and those belonging to East Asia and Pacific region in terms of the contribution To the best of my knowledge is the first article to be doing this analysis on a global scale where I'm able to Quantify the linkage between incidents of fires and economic inequality There's an influx of literature There's a rich literature looking into how natural disasters influence our economic level outcomes at the individual level and hopefully this contributes to that rich literature as I mentioned most of the references that you see in the paper are Looking into a specific natural experiment setting and kind of focus on individual level outcomes Which is different from this study where I'm looking into cross-country comparison So in terms of the data that two different sources of data I take advantage of the first one is satellite data on active fire incident locations, which is available from NASA's From database. Essentially it has information on the longitude and latitude and When the actual fire happened right so there's information on the brightness of the fire the temperature of the fire the fiery depour Did the fire happen in a given night or a daytime and I eventually aggregate these fire events at the country-year level because the Indicator that I'm interested into major economic inequality is available at the country-year level, right? And that's where world income inequality database comes in which has rich country-level information on several different indicators Namely ZIN index GDP and all other relevant indicators So in terms of the research design essentially it's a simple Specification that I'm using where I'm modeling economic inequality as a function of two different indicators of fire events One is the number of fire events to put it in a given country during the given year That's fire and then FRP is the intensity of fire, right? Essentially looking into what's the average rate of radiant heat that gets produced by a given fire and during the entire year I augment the specification with country fix effects country-year You know quadratic annual time trades as well as your fix effects in different specifications in the paper I also look into different other baseline indicators at the country level I interact them with your indicators and essentially the results that I present today are roughly similar in magnitude So I'm just focused on this baseline specification But for the paper includes other specifications as well that kind of takes advantage of using other baseline country-level information That three caveats in terms of the methodology that I want to point out one In terms of the intensity of fire events fiery to the power is something that's been proposed by a lot of You know folks working in remote sensing literature. So that's there's an influx of paper Using fiery to your power and so that's where this this paper follows the recent literature There are also some limitations also sit a bit satellite data, right? So this satellite data is kind of Looking into one by one kilometer pixel information So the center of the pixel denotes where the fire happened There might be cases where more than one fire event may have happened But maybe the satellite data doesn't capture that in a sense. I'm aggregating them all at the country-year level but that's the caveat that I want to mention here and In terms of interpreting the estimates, right? There are these cross-country differences that that we have to be mindful of so this is a little bit different from Kind of you see a lot of studies looking into administrative beta on income and all of that stuff So this is Something that I'm not really able to look into because every indicator is at the country level So I start the analysis with some basic descriptive Grubs here you can see Colonel Dern City plots on Zini index GDP per capita number of fires Faraday to power across countries of different types I break it down by high income upper middle income lower middle income and low income categories and essentially as we expect there's a massive economic inequality in countries that do not belong to high income Categories here. You can also see that number of fire events is heterogeneous across countries of different income categories Mostly you can see that on an average number of fires Tend to happen in countries that do not belong to high-income category, right? This is just the descriptive analysis, there's a lot of geographical variation in terms of the economic inequality as well as the the The stock of fire events the darker the color The higher the value of Zini index and so you can see that the blue Dots in the map are where we do not really have Information on so that's that's missing here but the point is there's a lot of geographical various and both in economic inequality as well as Fire events. This is again looking into a global map of fiery to the power. You can see that This is essentially log transform So there's some variation going on the darker the color the higher the intensity of fire events that I see from the satellite data before I kind of dive into the Kind of the fix effects specification I also collapsed data at the country level just to see what's the Correlation between these indicators as far as economic is concerned and kind of linking that with a number of fires as well As fiery to power and you can clearly see that as number of fire events There's an increase in economic inequality here So this is one of the key tables in the paper and so what's going on here is you can see I have seven different specifications. So I start with the entire data on column one and then I also kind of focus on Measles that are reported with high quality yearly level You know per capita measure a ruler and urban areas and you can clearly see that column six is where I am finding Surgical significance in that as number of fires and fiery to power increases There's a 13 and 20 point oh two percent is increase in economic inequality However, when I look into the interaction term right as the number of fires increases the The the effect on Zini index is is negative, right? So this is in some sense kind of I interpret this as some story behind creative destruction, right? So there might be a lot of different You know risk mitigating mechanisms countries adopt you can think about you know use of fire risk maps For instance Nepal and India came up with this fire a lot missing, right? So the these other additional channels that may actually play a role in kind of you know reducing economic inequality But I'm not really able to Empiricalities that are in the paper these are again the space on some other studies that have looked into this I also break this down by Income level categories between high upper middle and lower low middle income and Essentially the effect that we are finding on the previous table is being driven by countries in upper middle income Categories right and I don't really see any statistical significance in column one and column three But as you can see that there's a statistical significance consistent with the previous table in column two I also break it down by different categories of income so there's net income net or gross income gross income consumption and market income and Mostly you can see that the effect is driven by net or gross income and and consumption, right? And that's what you see in column two and column four and When I break it down by different reasons so we have to be a little bit careful with the reduction in sample size here So you can see you know column six focus on Saudis that only has 65 observations So I don't really want to make a lot out of what we see on column six But essentially what we're finding here is between column two column six and column seven Fires and fiery repower is actually having a positive effect on on Zini index And the same story behind the the relationship on the interaction Tom is consistent in in in these specific countries and I think that's where I'm kind of interested in figuring out what the channels that are leading to this heterogeneous response across countries is where I really need to be thinking about and Before I wrap up I only have a minute remaining So there are a couple of other things that I've done in the paper which is rather than focusing on the contemporaneous effect I've also looked into dynamic effect, right? So how long do these effects persist, right? And there's a lot of microlabel studies that shows that these socks may have a lasting effect And I see that these effects here last between one and five years and in terms of the mechanisms I'm also interested in finding out. Why is it that we see A change in economic inequality, right? Is it that years of schooling change significantly in response to these events? Is it that? Labor share in agriculture goes down and so when I look into those specific channels I do not see any statistical significant effect at this point So I don't really have a clear story on what's the mechanism that is actually driving these effects And so that's that's where I am essentially with this paper and so any questions or comments would be really welcome Thank you so much