Economic Identities and the Historical Roots of Climate Change Attitudes l Edgard Dewitte

0:00 Okay. Hi everyone. Uh my name is Edit, not Edward. Um I am at Oxford and I'm super happy to be here. It's obviously uh one of the best proud I can have for this paper. Really were looking forward to your comments. It's a paper that circulated the last few years. So I think a few of you have um have seen it already. This summer I did sorry this winter I did some some extra data collections. Um, and I'm really planning to submit this uh during the summer. So, all comments are very welcome. That's the time. Preferably there no than than doing the referee report. So, >> like Raquel, you can threaten your future referees here. >> No, but I'm I'm very I'm very nice.

0:46 Um, okay. So, the title is economic identities and the historical roots of climate change attitudes. Um and the motivation for this paper is um this quite striking or at least I find quite striking fact that about 40% of American declare not believing in climate change. Um and not only the share is pretty high it's been stagnant or very slowly increasing over the last 30 years since it's been surveyed. um despite obviously very increasing awareness and understanding of the issue and the current scholarship that's work working on this is not very able to uh explain this kind of stagnant and stagnant pattern. Why? First because people have looked at climatic events and what they found is that it gives a very kind of

1:32 short uh effect on attitudes. Um and of course it's a it's a very partisan issue uh and partisans should play a role but even within partisan group there's a lot of variation and it wasn't a partisan issue at least way less than now 10 15 years ago right so in a sense we don't really understand what is underlying uh formation of belief about climate change and what this paper is trying to do is that it's trying to shift the attention a bit away from climate change which for many people is still quite invisible um towards what climate change is really questioning which is the use of fosure and fossil fuel on the opposite are very uh they're ubiquitous especially for communities that are built around their production and maybe that's known to some of you but the US has the longest

2:18 and largest history of extraction of fossil fuel by far for the 160 years uh it's been one of the largest producer if not the largest producer and something we're going to see together today is that it's an extremely widespread u production about three4 of US counties uh have seen an oil and gas well drilled in their grounds. And what's been documented mostly uh qualitatively is that of course the industry had major impact economic impacts to the communities in which it developed but also strong cultural impacts and in particular it created something that one could call the economic identities or oil identities and I'm going to be specific what I mean by this but you think of a cultural attachment to the

3:04 resource and you can think of several ways to reach this type of attachment impact the formation of beliefs At least this is the question I'm asking today. I'm asking can long run histories of fossil extraction through in particular this global cultural attachments can impact present day beliefs about the climate. Answer is going to be yes and I'm going to try to convince you of this in three main steps. The first one I'm going to show you that there is a strong causal relationship between a history of extraction and present day beliefs in climate change. And I'm going to do that controlling importantly controlling for present day exposure to the resource which is obviously the major confounder um but also for many other things that

3:49 you could think play a role like a bundling of ideology. Maybe it's not just specific to climate change relief. It's a general ideology. I'm going to try also to kind of look at some question of selection into those areas. So I'm going to really try to convince that the places themselves have an impact on individuals. So I'm going to do that. I should have mentioned that before using data on virtually the universe of oil and gas wells that have been drilled in the US since the 1860s. And now that I've kind of tried to establish that places have an impact on individuals, the question is how are these places different right and here is my argument about fossil fuel identities and I'm going to document that using a bunch of archival data and I'm going to try to also document why those identities arise. And here my argument

4:35 is really going to be about the public good provision that firms uh and and the industry in general has been providing to the to those places and to kind of kind of tie the whole story together. Uh and this is where I'm still working on a little bit. I'm going to try to show that variation in where those identities develop within those highly extractive places in really is what is tied to the variation in police today. So really kind of pushing forward by okay error that I was expecting obviously was one depending economically on oil but that's basically what >> yeah here here this is kind of conditional on oil explosure today. Yes.

5:21 Um, okay. Yes. Sorry, please. >> I have a question. Should we take care of people's pain more? >> No, thank you for the question. It's usually I mean it's it's going straight to my conclusion. Uh, so I'm happy to talk about it now. Um definitely this is not specific to oil and gas both if you think in terms of climate change reliefs right there's there are other industries maybe a bit less in a less salient fashion but there are other industries that are very CO2 oriented that could develop some some some form of attachment right the meat industry the

6:06 car industry aviation right obviously you need to account for those to get the 40% even if I do back of the envelope calculation with my my special case here I get substantive numbers but not the the 40% right so you need to account if you want to think about the big picture um but also obviously not not even thinking about climate change beliefs right what this paper and in a sense it's a good jump to my uh to my literature review of course kind of straightforwardly I contributed to the literature on on on climate change beliefs but the second thing I'm doing is that I'm trying to bring an explanation that is around centered around what I call economic identities which of course is something documented by sociologist before but maybe the econ literature has been overlooked at when they've been thinking about national identity, religious identity etc. Right?

6:53 Um but in a sense attachment to a local industry uh or even the consumption behavior sorry is is really kind of probably impacting a lot of behaviors and beliefs right and I think this is this is why indeed this is a specific case but it's this has a general application for sure. >> Yes. >> Yeah. So um so when you say that it's caused by >> so it's a very good question. So it it is indeed implicit in that I'm tending to actually going towards more actually not talking about beliefs but attitudes I'm still have to think about it. Why? Because it's a it's it's quite hard to really sentangle those different layers right so you can ask a question and the survey I'm using is actually to some extent distinguishing between those those things. um you see that there is a

7:40 small but existing fraction of people that do state climate change is a thing. It it's happening but it's not caused by by humans. Um but this tends to vary very very much the same way as as as as these other beliefs. And of course again go back to your question this is all with the implicit thinking that when you talk about climate change no it's very salient that you cause you talk about fossil fuel generated so human human climate change but it could have different opinions on how important Yeah.

8:31 >> Yeah. Yeah. No, no, 100%. And this is the reason I'm saying it's it's a bit hard to disentangle is both because depending on the survey you use and I'm I'm using different existing surveys and they ask the question differently and sometimes it's more about the importance of the issue and and how much we should do about it and sometimes it's about do you actually believe it. Um but but it's also because you see an impact on both kind of of what of what I'm I'm showing. You see an impact on the propensity of people to say believe in the existence of the issue and on their propensity to think of to what extent we should do something about it or it's important. So this is why exactly to that point I think at least the paper here is more talking about attitudes and it's and it's it's going to be cleaner to have a

9:17 bit of a general vision of the attitudes and not thinking about the exact form of form of beliefs that people have. >> The beliefs are choose to believe something which is which is consistent with what we because there are less information. >> Yeah. Yeah. 100%. This is in a sense that that's that is more where this paper is going to is trying to show that it's not purely related uh and actually way less to what we think but this is not necessarily claim but it's not only related to the economic material interest you have in the resource and

10:03 one thing I can try one result to to to to uh to give you a sense of that is that even in regions where the industry is gone whether the production industry or even I I build a lot of measures of you know the the other downstream few industries that take the inputs uh of the industry even in those regions the history of oil still matters. Places with a longer history have a different attitude. Okay, thank you for the the comments. Uh this is uh just to say well in a sense I'm bringing through this existing literature on the historical roots of attitudes the question of of of climate. Um, and in a sense, what I'm contributing here is hopefully that with this idea of economic identities or

10:49 identities in general, which of course I'm not the only one to to discuss, you kind of understand a little bit better why something persists and something don't, which is uh which is obviously a key existing puzzle of this literature. Okay. Yes. >> Just this exposure to the negative effect of climate change like catastrophes or so. >> Yeah. So I have I have I have at some point I have a control on the the existing not only damage so far but forecasted impact of climate it's it's very auto right the only thing I could say on that is that if you have definitely only if you have a very strong negative impact of fossil fuels which is the like the only thing that we have in the US and that is the oil spill in the Gulf right

11:36 um the BP oil spill then you might have something that is you tempering this impact of of of oil and gas industry >> the more wildfires or so >> I mean sorry it matters but completely autoal to this to this story but what papers have shown is that the wildfires matter temporarily so if you survey people the week after they're going to be yes maybe it's an issue six months after they forgot about it and they're back to the kind of steady state beliefs which I'm trying to explain to you yes >> you say it's historical you mean that they need not no longer be involved that community in oil possibly. >> Yes. >> At all. >> Yes. >> Okay. >> But they are often they are. Of course. >> Well, then it's not historical. >> Yeah. >> Current. I just trying to make a

12:21 differentiation between historical and even, you know, if it's current, you obviously don't want to be blamed for something, but even if it's historical, you may not want your families, etc. or someone who's living in neighbors. You don't want to be blamed for that. I guess I'm putting that because I'm going to push back against identity. I see >> because I see too much identity papers now. And I'm going to say I I don't not I don't know that it's identity could just be guilt. >> No, I think so. I think you're right. I think the the question is of course uh whether we're thinking about the exact psychological mechanism that's at play or whether you're thinking about what is the visible element that is impacting this mechanism. And what I mean by that is that um definitely guilt could be

13:08 part of the story but in this case for instance very few people were actually immediately employed by the industry. What's interesting and it's actually why I'm focusing on oil and gas and and not coal a coal industry is the exact same story uh and and the results are the same but what's interesting about oil and gas is that the fraction of people that are actually immediately employed and involved in the area in the in the industry uh is very small. And so if you're experiencing guilt or some form of you know not willing to be for the industry to be blamed if your family and yourself haven't been immediately involved in the extraction it has to be because you feel an attachment still and this is what I mean by your local identity maybe super happy to take >> it could be I don't know I'm just going

13:53 to throw out things tax revenue what fraction of tax revenue comes from the oil and industry and therefore you know the schools and the hospitals what bears There's a name Exxon did blah blah blah founded this hospital. >> It's Yeah. No, this is exactly my story. So maybe what I should not say is use the word identity. Uh I'm just saying no no but I mean you >> I told her last week she was very excited. That's true. >> But maybe I don't get Rachel. No, but I think what the the reason why I'm flexible in the naming is because I think the naming is not essential to to the methods. And so the um I think an attachment is a prop it's a proper word uh whether it is because you have just a set of positive opinions with respect to the industry uh that as long as they are

14:39 kind of historically rooted and not just because today you have a clear material interest into it uh I'm fine and the public good of those firms is 100% what is going to help building these identities. >> Yes, right. I think one thing that could help you say for now you know it's it's sort of a lot identity story is that it's the social aspect of it. So say if you have someone who was born there and they started with this beliefs and then if well the moment when they moved out of the community and they sort of lost this identity so if you could show that kind of evidence then it speaks to this identity literature that it's really a collective identity. So no, so you're you're completely right. The issue is

15:25 that it's very hard to get panel data with movers surveying surveys before beliefs before and after. Correct. >> You don't need to have a panel. >> Okay. So you think I can just look at the movers and see. Yeah. Yeah. No, you're right. And I have some data on. So I use it to kind of show that in migration is not driving. But you're you're right that out migration studying out migration could help a bit understanding the the mechanics. Yes. True. >> Okay. That's the outline. Uh we've talked a lot already and reviewed some of the results. So maybe I I'll skip some but um very straightforward. These are the three steps that I that I just showed in the introduction. Very quick background so that you fix IDs. Orian gas industry in the US could be characterized by two main features at least two main that are interesting in

16:10 my case. Um first there's a very large and easily accessible endorment. So this is why it's been such a big producer since the mid 19th century and the first one. So you have oil fields um across the entire country and there's an interesting part of the institutional kind of property rights dimension of the US which is that there is private ownership of subsurface minimums. So that means that it's not a big governmentite entity that is taking care of extraction. Uh it means that if you own a plot of land, you own the land below and hence uh what happens is that uh there is a very widespread private exploration in oil. Once oil is found, this is this is what happens, right? It's like immediately you see all plots of land trying to get some oil because in particular because of this idea which

16:57 is the law of capture um which states that any resource that you extract from your ground not just that is lying below because of course with oil how do you know where it's lying? Um any resource that you extract from the ground is is yours. So this is how it show. This is Texas in 1910. But take like if you want a bigger picture, this is my data on oil and gas wells. And you see that it's extremely widespread across the country. Um and what you see here already that I'm showing, which is an important distinction. The red dots are the producing oil and gas wells. The gray ones are those that were drilled but never actually produced commercial quantities. They were intended for it, but they were not successful. And it's a US specific thing, right? It's um it's 60 US is is um here US in 2005 was 9% of

17:45 world production but 16 60% of most of our world. Um so data I use is is I am mostly relying on this um well data why because it is the more comprehensive historical data. You don't get good production figures historically. uh unemployment is uh at the the senior level and of course as I said it's very low employment industry so it doesn't get much of variation but but I I I complete with the the data with a lot of present day local prediction environment data as you can see um for the wells we're going to discuss if needed but for the wells I have a lot of different uh elements production status and operators are going to be quite important and um on surveys. As I mentioned, I use a

18:31 bunch of surveys, but the main one results you're going to see today is mostly on using the CES, which is this big uh survey run in the US still today, but has an interesting question about climate change between 2006 and 2012. Uh, and the survey is great because it's a very large sample and it's covering a lot of the of the of the counties like 96% of counties are I covered. For those that might say be skeptical, I think this crowd is fine with that, but skeptical about surveys and I take your your question in a second. Um, I show that at the county level, those beliefs, they do correlate with some form of behaviors you would expect from people that do not uh have strong attitudes proclimate, right? So, they buy less electric vehicles, they give less to green packs, um, they vote less for

19:19 green belot initiatives at the state level and things like this. Yes. So you're also going to show us like a map for the deposits like oil and maybe also even historical map for that. >> So I I want and the reason is the following of course very intuitively or also when I started this project I was like great the good thing about working on this type of resource is that you can instrument to some extent with the existing resources. The issue is that historically that is not a feasible thing because the actual reserves in terms of accessible reserves depend a lot on the technology. So if you want it's very hard to say no something that is today accessible reserve for oil and gas wasn't the case before the invention of fracking for right and this is the true so you would have kind of a weird

20:05 setup where you need for every existing small variation in technology existing accessible resources so I'm what I'm using instead uh and this is also a good transition to here I'm using another thing so what I'm the variation I'm I'm I'm interested in is the crossount variation not only the presence of oil but also on the length of exposure right this is back to my uh to my argument and the feature I'm going to use because I can use those deposits is the uncertainty of drilling because what I observe as I mentioned is that I observe whether a well is successful or not and in fact for the large part of the century a lot of wells weren't successful because the science to really know whether oil and gas was to be found

20:51 in a certain place was was very was very poor. Right? So when I say a lot is about a third of the forms of human wealth that I have in my in my data are unsuccessful. Um about 173 counties have been drilled unsuccessfully. And this is going to be kind of a straightforward control counties. Um and what's interesting uh I mean I bring a lot of anecdotal evidence to kind of back that up. I also show that to some extent quantitatively with the correlates that I had at the time. Um they can simply question one second. But what you can say is that not only the presence of oil once you've drilled is is unclear, but also how long are you going to be able to extract the missile. So the length of extraction post drilling is also quite quite random and I and I and I kind of

21:37 show that one question if you want to follow up on your question. just curious. I mean because looking at actual whales they do have a lot of problem right because like it's a position to take if you want to draw or not and there's some places you just can't like if you're living in a city in apartment building you obviously cannot dig a home run right it's going to be like >> homesteading county then you get a lot of land >> then that's correlated with many but that's the point I think this is exactly the point of taking only the successful counties as a control because you know they factor what happened is that my these are the counties in my sample the gray are the control and successful ones and the the dark are those that actually had oil and gas um and indeed many of

22:24 that takes away a lot of places like the northeast regime had oil both for kind of practical reasons but also because it's it's not the same type of setup to to drill but I in a sense I want to stress also that oil is such a valuable resource that even if it's below a building in New York, they would people would drill for it. That's >> Yes. >> Guess it's relevant given this map, but like you were describing as as like so decentralized and all these small wells like how is the data collected of unsuccessful on these dry wells? I'm surprised that there's even records. >> Well, the there is records of that because the again because it's a valuable resource um quite quickly states try to organize some form of registers register of where it was drilled. One reason of course is that

23:11 apart from a few places um it's still land that is owned by people right and so there's kind of leases that have to be signed and those kind of official documentation to some extent is well recorded but of course I don't have the absolute universal and success for us pre900 it's it's been actually a lot of so the data source is in part a private uh entity that is collecting this data for the firms but a lot of archival work on my side to make sure that the coverage on this prep period that no one cares about is is decent, right? It's not extremely exhaustive, but it's pretty um because then you can cross the exact number of wells is not important. So you can cross accounts of whether people actually search in an area to my data to see okay is this area related to something.

23:56 >> Yes. >> Just clarifying are there any doesn't work out. >> Yeah. So, so that the completely now you the point is that then you abandon the place you know was like the the the process of getting of trying to drill a well doesn't involve a large settlement that would mean that there was kind of resistance of these unsuccessful communities but but good point um okay so that's my variation that's the same map I show you before but also with the length of histories right and this is uh this is really key and you see that even in kind oil intensive regions from one country to another you have like very different

24:42 length of of extraction meaning number of decades during which there is oil and gas extraction and I'm plotting that in a very straightforward or less where I believe on the on the left hand side the oil exposure on the on the right hand side and then a bunch of controls now and it's actually an open discussion which one I should put in the sense of many of them are going to be bad controls right because they come expos of the treatment of finding oil but they're going be able to rule out a bunch of hypothesis. What's important to note is that I have worked with state fixed effects. So I'm really comparing two counties within the same uh the same states and in particular two counties with the same level or similar level of present day exposure. So here there's a bunch of things I do um and want I mean

25:28 you can look at them if you want but bear in mind that in all setup my my my results are robust. whenever you think of grouping counties because of course it's a it's a it's also a very bad control right the prediction today is correlated to the history um and and it's exposed but if I plug the variable itself in the regression my coefficient is very little impacted um you're going to look at the results but this is true even if I do matching or grouping um and if I focus as I said before on counties with no prediction yes >> do you know where there is a variation in ambient domain whether the the the state was allowing more firms to >> Yeah. So there is there's two form of variation that I mean I don't know which one you're thinking about but yes

26:13 there's first of all a variation in who owns the land and in kind of public land there's been a bunch of rules for instance of who can drill um and kind of unsurprisingly in those places I see less I see an effect but less of an effect um then it's a bit more tricky in terms of the rules that evolve over time in terms of proation of kind of control of the industry um let me defer the answer because that's going to be related to my last part uh I think That's that's relevant. Okay. Results uh baseline uh as I mentioned one in this case one standard deviation of number of decades of exploration which is about 3.2 decades um decreases by 7 percentage point your probability to believe in

27:00 changes to have five time change attitudes. Now this is a county level variable impacting an individual. If you compare it to other determinants, that's what I write here. This is roughly the same magnitude if not more than the variation in the county's share of votes for Republicans at presidential election. So it is pretty sizable I think at least in comparative perspective. And here is where here I add those kind of present day county controls which kind of have a lot I take your question saying which have a lot of different dimensions with the partisansship education income industry share right all of these are obviously to some extent impacted by the treatment itself um but it's kind of interesting that even controlling for them uh you

27:47 still find a pretty distinctive uh effect um and here is when I say controlling for present day pollution being still producing counties and the same in places with no production anymore for the this is no production for the last five years. So it's pretty clear that the industry is gone. Um and and the results are stable. Yes. >> So I wonder if we could say um look at the people who were growing up during uh a period when was very important. So say take the 70s or early 80s. Um and in places that were still truly at the time and if you know people who were growing up around that time to be compared to others age component and whether that left them

28:35 with a stronger sense of attachment. >> Yeah. So I what I do in tell me if that that corresponds is that actually I do something similar uh when thinking about so here we're looking at alternative channels and and the one question is of course migration uh and what I'm looking at is actually the decades of oil that happen after you arrive to a place or after you're born in a place and in the kind of pure selection dimension you know if people would come to these places with this a set of beliefs that would not that are actually not impacted by the experience of the place you would see this showing up and you actually so to some extent yes but I mean I could compare those that are when it's more salient not just whether it's happening or not but not sure I would measure that

29:21 right because you're at 70s 80s very period here is comparing just whether there is been extraction or not after >> I guess I was talking more about like the budget so what's the driver of this >> so you mean one yeah what you mean the exact time period or when you know someone's lifetime is actually important. >> Well, given that a lot of people sort of form their their a lot of their beliefs during say the >> Okay. Yeah. The impressionable. Yes. Yeah. That I find not not I mean no it doesn't matter what time exactly the the thing happens again because if you think about the treatment uh that we're looking at here, it's not a sudden shock uh that you experience. It's a general

30:07 cultural environment or or just some signs in the place you're living in that are reminding you and that are kind of socializing you into into having a positive thinking. Um so so in a sense that makes sense. >> Find your gender differences. uh I find it a bit to be stronger with men the my effects which can be understood in the sense of in general women have much higher or significantly higher beliefs in attitudes towards climate change they suffer interesting gender dimension that I'm not exploiting in this paper knowing how much the kind of in a more qualitative fashion how much the oil industry and the whole adventure has been tied to very masculine type of

30:54 >> but did you interact seconds of exposure gender or >> what I think I did >> yeah so I split the sample between men and women >> um okay here what I want to show especially given that I don't have too much time but I want to show something that I think is pretty um pretty still helping understand it's really specific to the to the story I'm saying so here I'm mentioning I just look at the fracking revolution as kind of a shock in present day production And I think this is kind of that's going back to the thing of it might not be just economic dimensions, right? Because the what I find using the fracking revolution as a nice exogenous shock is that it really impacts belief quite little. Um so it's seem to be even though it's obviously a very strong economic shock, positive

31:41 economic shock. Now here I'm doing the same thing but not this is the my main bible but this is with different attitudes social attitudes uh and same with physical attitudes uh with distribution etc. And you see that it has very little effect so it seems to be specific it's not kind of a big ideology that oil and gas is creating um it seems to be specific to your beliefs about one thing which is which is climate change. Um okay, how are communities uh possible communities different? Well, my answer as I said is uh oil identities which are defined as a sense of individual community attachment toward the resource or uh the industry and which can include

32:27 or or be summarized by a set of positive uh beliefs regarding the contribution of the industry. Very conveniently chosen quotes to kind of back that up. This is a community organizer in a oil and gas community saying big oil. There's no such thing as big oil. It's all our friends and neighbors and people running our towns and supporting us. The oil industry is us. Um there's a lot of very funny anecdotal evidence or at least I find it kind of quirky of festivals around oil and gas of businesses rebranding around it. uh of changes in children vocabulary that don't ask to be swinging on their on their ways but to be pumped like an oil an oil well on the on the way. So, so just signs of an appropriation of the industry um political identity or not

33:14 and there's of course a kind of theoretical reason to believe this is happen. How do I proxy for that quantitatively? I use the case of sports teams in the US which um as you might know is a important element of local cultures right virtually all places in the US have teams sports teams where amateur pro high school etc. Um, and those teams take name. You know the Boston Celtics, you know the Pittsburgh Steelers. Uh, well, what I've learned working on this is that for 40 years when the NFL team of Houston was still in Houston, they moved to Tennessee. Uh, they were the Oilers. They moved to Tennessee. They became the Tennessee Titans because it didn't make that much sense to be the Oilers of Houston of Tennessee. But in fact, if you look at local newspapers pages of like sports

34:00 team results, you see a lot of those oil teams. Um, and it's great because actually newspapers is something we can we can use um and uh and that I scraped to kind of in in a search for those not the main ones which are obviously discussed earlier but for those local uh unimportant if you want teams um and indeed what you find is that after uh after all is found in the place again just controlling for places that were successful or not over time not immediately but over time the probability that you have sports seems that our name orders is um now okay let's try this is a kind of a nice first step into seeing this happening um how have they been created my story or at

34:47 least something I've tried to push and it's also convenient for for the last part of the paper um is to really think about uh oil companies being providers of public goods and if you think it's like it's also a way to tie it to why national identity are rising in part it's related to the state provision of public good, right? Um and here in fact in those places as this as this quote is is is stating this was happening a lot uh and this was often perceived as a positive thing even though of course they are very strategic dimension for the oil and gas industry to do it. They need to release the land to hire the workers. they need to exist with this vast amount of wealth in places that have um that don't have often that much uh that much privilege but but but it's

35:32 perceived as a nice thing as as this person says we've got tons of different oil field companies that have been started by people who grew around here um and they just want what's best for the people that they've been working that been working right whatever they can do for the community they are great community-minded people um and one thing that they do which we can measure of course you cannot capture uh everything but one thing that they do which is kind of nice is that companies organize a lot of social events. They organize picnics. They sponsor conferences and these are all things that are advertised also in the newspapers. And the good thing is as I've mentioned I have the names of the companies extracting even all those small companies uh that are doing the extraction because in many cases these are very small companies. >> This is true for any sector of the economy I would say. >> Correct. you know, New York, you know,

36:19 who finances the museums, you know, finance companies do that, the opera. >> So, yeah. No, no. >> Yeah. So, I think and from, you know, for the same reasons as as as >> I would be super interested to see to what extent that indeed contributes a positive association with the industry um at least in this case. So you see that indeed again this is a bit more like it's not super surprising but you see that this happens way more in places where oil and gases is actually being extracted. Makes sense. These are the companies that are that are doing it. Um and no because I have like one minute left. I will just talk I have other measures of you know local public finances changing. you see that tax revenues are are are

37:05 extremely uh jumping when the oil is found. Local politicians are more likely to come from the industry etc. So kind of I'm documenting a bunch of things. You you also see a lot of places being renamed. So you have petroleum mountain old creek things like this um this is kind of to some extent descriptive. And in the last part what I'm trying to think about is how can I really show that this is the vector of uh of changing beliefs today and here I'm lo using a variation that is in the industry size and what I mean by this um is that uh so let me let's not get distracted by the slide um what you find is that depending on geological conditions you can as an individual drill a well

37:51 pretty easily and have your own oil and gas well in your in your field next to your farm and that's great. you get a certain revenue and some regions have a lot of these very small by small it's like one or two oil wells firms extracting oil gas other regions for several reasons whoever property rights or uh again the ge geological conditions that ask for high invest or higher investment you have firms that have 10 20 30 oiling gas well so bigger firms and these are those that you would expect are doing this kind of publication right these are those that are kind of in the relationship of nursing with the constituency is not just every individual farmer doing it. Um and indeed so these are the measures I

38:38 showed you before. This is the congress members etc. Um and indeed it's in those places where in the firms are bigger that you see more identities as proxy by artists building up and that you see by impact of decades of producing all exploration being twice bigger. So these are places where the company size is above median and you see that again there's not zero effect in other places because that's not the the public provision uh is not the only possible channel of attachment but this kind of seems to be pointing out towards towards >> and again I'll question the identity part then I think they're providing all these public goods and you're being sold an image or maybe an also you're you're

39:26 getting net benefits. Yeah, I don't say you're getting benefits. >> You're getting No, you're getting benefits, but if the industry is gone, you Yeah, it's the image you kind of stick to to some extent, right? Uh but maybe coding that identity is a bit too >> No, no, but you're saying that if it's gone and they're not providing any public goods anymore, then still >> Yes, of course I cannot I can control for the presence of the industry. >> Do you see that in your regressions that it's gone? So I this is here because what I don't I don't know that their current public good provision but what I know is whe and I can define that. So this is the non-producing prices for the last five years. This is true if I look at employment not just production. This is true if I look at downstream industries presence in

40:12 >> yes let me wrap up conclusion slide. uh and actually the conclusion I've said in the beginning so conclusion slide is what I said probably this industry but also other industries might have impacted climate change beliefs and if you think about changing those beliefs you probably want to think not just that the material dimensions that those beliefs are question those statements about the climate are questioning but also the cultural elements that they are questioning and I think in in all of our countries we have uh cases of of environmental regulation threatening some for of cultural developments or um or industries. And in fact, if you look at just the snapshot of beliefs around the world, what I find interesting um is

40:58 that with kind of a continent fixed effects, it is um Venezuela, Norway, uh the Arabic peninsula etc that display the lower. Of course, this is not rooting the present day the kind of economic dimension, but it's kind of one step thinking uh into into into the application of that. Thank you.