good afternoon we are here today to speak about uncertainty concerning climate policies with lars peter hansen uncertainty is the key word of this speech i'd like to make some brief introductory comments uncertainty well it is important that is crystal clear but this pervasiveness has become especially stark in the past few years because of some catastrophic events i listed three things there the financial crisis of 10 years ago you might remember fukushima nuclear catastrophe and what's going on in these months i.e the pandemic these are disasters which have already happened we are here to speak about future disasters uh due to climate change which is in itself a source of uncertainty these recent phenomena have further stressed the importance of uncertainty uncertainty makes uncertain the payoffs of a course of action for example if i see well my harvest will depend on the weather which is uncertain however in general our actions are influenced by uncertain factors what's important in particular in economics and social sciences is that persons when they have to decide they form expectations about the factors that affect the uncertainty for example a climate change for farmers technology uncertainties for entrepreneurs and considering the expectations people try to do their best this is a key difference as compared with particles studied in physics and social sciences we have to do with persons with agents which or who sorry think about the vectors influencing their actions a very important primary factor is that of public policies so people have expectations on public policies for example the policies which might have an impact on climate change such as climate policies and they act and react based on their expectations it is therefore important their economic or social policies or anything involving agents or persons uh take on board that specific feature these are not particles these are thinking beings who interact and so we have to avoid what we called the grida manzonianes useless shouts more practically if we think of healthcare policies in the public sector as we have seen in recent months and environmental policies in the light of climate change such policies should take on board social and natural sciences knowledge for natural sciences when speaking about the pandemic biology is key to understand the contamination for climate physics is important because the climate change is a physical phenomenon first of all and then we should consider social sciences because agents are not persons are not particles they react they think they adjust to policies it is therefore important here that there is a clear dialogue between natural and social sciences and a notion of uncertainty which unfortunately cannot be avoided it's an utopia to think that we can do without that we have uh today a very important speaker last peter hansen a great thinker and many issues and uncertainty in particular he gave very important contributions to understand the impact on public policies and economic phenomena by uncertainty so his work is of fundamental importance both theoretically and from the point of view of econometrics to understand how uncertainty works and that has changed our way of thinking about such phenomena so professor hansen had a key role concerning the way economists approach these fundamental problems here's a david rockefeller distinguished professor at chicago university you know that the school of chicago is the most important school of economics generation after generation many important economies took their degree there and lars is one of the most important members of that school in 2013 he was awarded the nobel prize in economics for his work it's a great pleasure and a honor to have you here speaking about uncertainty concerning climate phenomena over to you professor hansen and thank you very much for being with us thank you very much massimo i very appreciate the kind words and the very nice introduction to what we both agree is a very fundamental and important topic um i'm very sorry today that i actually can't be there in person but but as we all know the events that the surprising events which we've experienced the last several months has changed the way we've been doing activities in the near term so today i want to talk about climate change and how climate change uncertainty can impact how we should think about social evaluation in general and about economic policy i will be showing you some examples of some example calculations from some work i've done with mike barnett and william brock so let me start off with this this quote from mark carney when he was the head of the bank of england uh mark carney described climate change as the tragic strategy the tragedy of the horizon and his gen and he suggests how we address it we need better information um we can build virtuous circles understanding tomorrow's risks better pricing better decisions and the like that is a what grand ambition um and uh if only things were quite so simple so i'm all in on that type of aim i think there's a lot more subtleties to addressing this than might be captured by this quote however so in connection with this quote i'm going to put a different one that's a little bit of a different contrast so mark carney of course is head of the bank of england and this is a warning it comes from a very interesting essay by hayek who is discussing the limits of economics as a science and it's a fascinating essay i don't agree with all aspects of the essay but there's this statement in the essay that kind of i think that is important that we keep in mind thinking about the impact how we how economists impact economic policy even if true scientists should recognize the limits of studying human behavior as long as we have public and the public wants things like that has these expectations then there will be people who either pretend or believe they can do more to meet popular demand and what is really in their power so that's really a key part of what we have to wrestle against here yes we want to deliver on the ambitions that that mark carney lays out but we also have limits to our understanding so we want to be able to do it in a credible way and not in a way that can be harmful if we if if we go in with extreme overconfidence in the life we can we can easily be led to um to false confidences and policies and end up being counterproductive so this is the type of questions i want to be wrestling with here so i'm going to talk about components of what i think of as growth rate uncertainty to the overall economy to the so-called macroeconomy and there's a variety of sources of this um elsewhere i thought about some of these other sources including so-called secular stagnation after the financial crisis macroeconomists were speculating on how long it would take us to recover from it there's debates about how technological progress going forward we had some amazing progress in the last century to what extent will continue more to point today is that if we look at geoscientists um there's lots of evidence evidence of human impact on the environment but what's less known is the quantitative magnitudes and the time scales that are which the precise time scales in which they might play out then economists are very interested in uh taking this these kind of geological scientific insights and to try to make guesses about their subsequent implications for economic and social activity so environmental economists will then make conjectures about how climate change itself can alter economic growth in the future and this gets into important questions about potentially how we might adapt to changes in the environment so i would be focusing really on the latter two of these and and their interactions so let me expose you to some some early thinking about uncertainty so economists most of the work in economic analysis is under a so-called risk model and to put it simply think about flipping coins things to think about rolling dice these are situations where we know probabilities but not outcomes and how far can we push that type of paradigm so knight back in the early part of the 1900s was writing about uncertainty in broader terms and was really pushing us away from this simplified notion of risk and trying to uh suggest this could be very important for a variety of phenomena uh keynes kind of echoed this sentiment in some of his writings as well saying our knowledge of the future is fluctuating vague and uncertain now both of these are i think hit upon some important questions neither of them really gave us the conceptual underpinnings to think about them in very systematic ways so but but but they certainly were put important questions on the table so let me just talk about this risk a little bit more um in its simplest form think about an urn suppose we know it has 25 blue balls and 75 red balls and we pull a ball out at random now what's the chance of getting a red ball well it's three fours we know probabilities we just don't know outcomes so so that's what i i'd like to use the term risk and yeah many other other other economists use the term risk and this is kind of the notion of risk that night had in mind now let me make things a little bit more complicated uncertainty can be ambiguity suppose i make the urn no longer visible i've got some red balls in there i've got some blue balls in there but i don't know what what the fractions are again i can draw things out at random now there's a possibility here of learning if i kind of do repeated draws i put you know draw something out put it back in draw again put it back in tabulate the number of red and blue balls i'm going to get better and better estimates of the fraction of red and blue balls uh this is kind of an illustration of bernoulli's law of large numbers from you know more than 300 years ago um so that's already makes it one level more complicated and the statistical discipline is certainly uh this is a question that this is a problem that's not routinely addressed with that in that literature so there's lots of thinking about this now put in economic dynamics so i so one way to uh picture this is imagine you've got a bunch of these urns it's an earn you draw from today and maybe to earn tomorrow is a little bit different and the earn the next day is a little bit different as well now i can no longer do this draw out of an urn look at tabulate red and blue balls um and figure out what's in a given urn because it earns changing on me even though i might learn i'm learning about a moving target and i think it's getting closer to the type of complexities we have to face when we're thinking about the economic dynamic phenomena all the way from covet to climate change to technological change and the like now we all have some idea about how interest rates might compound yeah we we play out the interest rates of our time in a compound and and that can have you know big implications compound interest rates there's a compounding of uncertainty that can also play out over time and these uncertainties can compound in very different ways it depends on the nature of the uncertainty i've just sketched out two different notions of compounding here about how they might play out one is more like you've got that same urn and every period you just draw from that the ambiguity i talked about and at the end of the day there's just two outcomes a red ball or a blue ball and that's it but now let's go to the bottom diagram here so i draw the first urn and that dictates my next turn and i draw that urn and that dick takes a subsequent error and the urns may differ on you now i'm going to start tabulating a lot more outcomes the compounding the uncertainty is more complicated because there's more things to consider as we play things out over time and i think this is the type of uncertainties that we often face and so i think the bottom one is compounding in a very different but a very important way as distinct from the top one so there's a very eminent italian probabilist different eddie he wrote in this area a wide variety of important areas one was in subjective probabilities and this was the philosophical underpinning of what is commonly known as kind of bayesian type methods in statistics these are known to be very very powerful ways to approach things like model ambiguity and the subjective probability approach basically fills out probabilities you don't know maybe i i have some prior notion of how much how many red and blue balls there are and then is able to apply a lot of the same rules of probability after i fill things out over these uh so-called subjectively of these dimensions in which i don't have prior knowledge and so it's a very elegant theory that uh dfinity develops and justifies but what's important is that defeneti himself understood the limits to that theory and so here i've got a quote from him subjectivists should feel obliged to recognize that any opinion is only vaguely acceptable so it's important not only to know the exact answer an exactly specified problem but what happens you start changing things a little bit how sensitive are things to kind of these changes and things and these subjective inputs we need to make in order to make these probabilistic assessments and i think this is very this is a very important insight and this is uh um one that's and definitely's work under appreciated in my view my colleague bob lucas was i wrote wrote a discussion of a very influential empirical contribution by friedman and schwartz on the history of monetary economics policy and lucas you know was very complimentary of the book but it's but if review ends with this or or ends in part with a quote like this the facts do not speak for themselves they need all the abstract theoretical help they can get so i'm putting this on on here for the following kind of reminder of course we live in this world we do is very data rich well you know there's lots of very important work being done on um uh looking through methods uh computationally tractable methods for looking at large dimensional data sets for finding patterns in those data sets uh machine learning type algorithms and the like and you know these are very interesting important contributions but you also have to have a conceptual underpinning and i talked about this as well in the southern europe reference here um for a lot of questions that economists want to answer it can't just appeal the facts and the facts tell you the answers you need models theories conceptual frameworks this is all this more this is very clear on climate change because we're talking about pushing the environment the work into situations we really haven't experienced in the future right in the past we can't just look at past data and say well here here's uh here's what's going to happen so we also have to rely on these models and these models require some subjective judgments and the like and and uncertainty as well comes into play there so um i thought today i would spare you a lot of formal mathematics and i'm going to substitute a painting in this case this is a very famous it's a very beautiful painting by the tour um so let's talk to those painting for just a moment there's a person on the right thinking that this is a fair card game this person in the middle the dealer somewhat shifty eyes is a person pouring the liquor that is perhaps no uh in on this and then there's a person all the way on the left who has something behind his back some cards so not surprisingly this is called the chi so why am i showing this to you we can build wonderful mathematically elegant models and those models can tell us probabilities just like uh in very complicated sets of circumstances we can go all in on those models say that's the model it tells me probabilities i'm done now i just figure out what the sensible decision is but for a lot of social problems that's not where we're at and so what you have to put when you look at these models you know i i'm a firm model builder right i spend lots of evidence kind of solving community and analyzing models you need to better use them in sensible ways so that and not naively like this card player you need to ask about how might the modeling effort mislead me so you need to use them sensibly and not naively so where's where does uncertainty emerge in all this so this quantitative model building i i think of each model is telling an interesting story it has policy implications it helps me understand the world but it's not a complete description of that uh of this complicated world we live in it's an abstraction or simplification so i like to think of these models as telling quantitative stories and i also like to think of the big multiple models on the table that we should consider there's this quote from st thomas aquinas that i that i quite like that says beware the person of one book and it carries over to models beware the person of one model it's important to look across models so this is like quantitative storytelling but with multiple stories so how does uncertainty come into play well we have this notion of risk a model is going to have these random impulses into it these so-called shocks we call them and that and then from there we figure out they transmit to the economic environment in some dynamic way and then we think we can compute make probabilistic statements that's that's our version of risk there's multiple models on the table multiple stories different implications we want to look across those models and see what their implications are and then there's a third piece each of each of these models is itself an abstraction it's not intended to be a complete description of reality and if it were it wouldn't be very useful to us because it would be far too complicated to to truly help us understand the world better so we understand the world through lenses of models but the models with simplifications how do we take that into account so there's formal ways of doing this and but let me just kind of guide you through a discussion of how they work roughly speaking navigating uncertainty so we want to use probability models but we want to acknowledge that they're misbet specified they're wrong but that should both still think they're insightful this ambiguity as to which among these models is the best one then following the suggestions of dfinity we want to do some we want to do a type of robust approach we want to say look at sensitivity analyses i want to use the model and sense it in in a sensible way i don't want to throw out statistics because i need statistics and probability to bound things if i come into this problem say well look i i don't know anything with full confidence i'm just going to give up then that gets us nowhere so we need we still want to use tools of probability and statistics to help us bound the types of uncertainties now to turn this into decisions there's some there's a theory of uncertainty aversion just like risk aversion we talk about in economics all the time risk conversion risk risk premium asset prices and the like there's a risk of there's a more broad notion of uncertainty aversion you dislike this notion of uncertainty about probabilities of the future you have to take some type of stand on that and the implementation then is to target what uncertainty components have the most adverse consequences some things we don't know much about may not matter to the answers we care about but sums do we can use quantitative methods and uh and formally figure out answers to this and that helps us um approach these uncertain these uncertainties which we face in actual decision making so economists love to think about things in terms of trade-offs so i like to think about there being a type of uncertainty trade-off so as i watched the early stages of covet and the use of mathematical models people were using models in very in different ways and then this led to some confusion in some of the public discourse i believe some people would take models and figure out best guesses what do i think is my best estimate of what could happen another type of calculation could do is figure out well what possible bad things could happen if a as things play out now i could go all in on the best guess but then there's some chance that says something really bad might happen and and i'm and i'm just not prepared for that i could go all in on what's the worst outcome possible and then i maybe let it do nothing i need to be in the middle in between these two trading these things off and that's where this conversion at this aversion comes into play it's a statement of how everybody about how i trade these things off how much attention do we want to pay the best guesses for that versus guesses that lead to bad outcomes and and when people were talking about model outputs they were often getting these two things intermixed and confused both are interesting calculations but the but there's an interesting trait out there but you uh but which is important in in designing proven policies i think it's also important by the way for constructing investment strategies say if you're wanting to know how much to invest in green technologies and like so now let me turn back to climate change let me turn to climate change a little bit more specifically here there's a construct that is used in just in policy discussions called the social cost of carbon in truth it's commonly referred to a policy discussions but its meanings and targets of measurement can differ across applications and it means different things in different settings um in the u.s under the obama administration they would routinely report social constant carbon numbers on web pages but not but but their explanations were not always so coherent for why for why they're reporting or how they construct them but i'm going to use well well-posed versions of these and i'm going to use this as an analytical tool to assess how important uncertainty is so here's how this here's how the social cost of carbon works i can look at the standpoint of markets markets can give me information about how much i want to pay for say fuel fuel at the at at the gas pump but then i can ask how much from a social perspective should should i be paying those two calculations could be very different because carbon emissions alter the climate which in turn impacts economic opportunities and social well-being it gives and that in that impact's not going to be captured by market prices it's called an externality economists have been familiar with externalities for a long time so so this is an externality whenever there's externalities then there's this wedge between a private cost and a social cost and one way to assess climate change is how big is that wedge because that tells us uh it helps us quantify the need for policy interventions now the simplest type of policy intervention is probably politically naive it's you know economists always like like prices and taxes one could one one could uh actually devise the so-called pigouvian tax on carbon emissions that corrupts this uncertainty uh that this may not be the most politically feasible way to go and and so i'm not going to necessarily push this as the pre as practical policy making but i'm going to use this wedge to tell me how important the need for policy is and what the contributions are coming from uncertainty so i so i mentioned before that you know geoscientists have uncertainty about the impact of say emissions on in this case temperature so what i'm showing you is here's the so-called histogram what it does it takes um it takes climate sensitivity estimates from a bunch of models and just tabulates them and so this and so what these what so what happens here is people run simulations across these models they put in changes in emissions and they see what happens to temperature over you know moderate to longer time scales and they measure that sensitivity how how responsive is is our our future temperature to current changes in emissions and we think of alesis as these these responses are really over the course of the next several decades quite durable in nature i mean they're they're not literally permanent but they're but they're very durable and so this is lifted from a paper climate science showing you as i run these model simulations that parameter that climate sensitivity parameter from from emissions to temperature changes shows a lot of uncertainty you know in this case we see a central tendency of about one yeah i'm not going to get into the details of numerical magnitude but but the central tendency is about 1.5 1.7 or something but now one is doesn't look that unreasonable and things you know numbers substantially bigger than two as well so there's a fair fairly substantial range here about about this climate change uncertainty so now let me turn to a simplified characterization of the economic damages i'm using what environmental economists called a damage function here it's a very simplified way to capture the impact of temperature changes on the economic opportunities and so i'm plotting these things out over over temperature changes that might occur in the future and um there's lots of debate over how big these damages are going to be this is something we don't know a lot about um the work of nordhouse uh tracks the blue line quite closely here it's meant to be that the blue line is roughly his his best guesses um marty weitzman was pushing numbers that were substantially larger and part motivated by uncertainty uh rough uh uh that doesn't that's more along the lines of the red line here the high damage line here as well there's lots of recent work in climate science that just wants to put a hard budget constraints and we should only allow temperature to move uh two degrees and for economists that's equivalent to saying that that's what that damage function that just falls off a cliff right it too and we tend to think that things aren't quite that extreme but there's lots of scope about differences here and the way this is plotted out here is the two lines are on top of each other up up to two degrees so we're talking about differences that the way that we really haven't experienced um in terms of the global economy yet and so this is a case where we don't have it's hard to tease out lots of information from data here so what i've tracked here is uh in in our calculations what's called the social cost of carbon so the blue line is the total it's a total social cost of carbon and what happens here is this cost increases as we get more closer as we emit more uh we get closer and closer to the damage region and even though this is off of some hypothetically optimal social policy uh we will eventually get to places where closer to damages and then the costs are then the cost starts to rise on it substantially and emissions go down at the same time that cost increases and that's what the blue line is and so it's gonna and this is the trajectory over the next hundred years now one possibility is i take those that histogram and just put in the median or the centering point i i wait low and high damages 50 50 and just pretend that that that's the damage function and that's the um and that's the climate sensitivity parameter and proceed well um that would be one way to go and i would understate the amount of uh but that would misrepresent the types of uncertainties they're talking about it wouldn't really take into account the distributional aspects of it so what we do is we do that type of comparison what happens if we just went with the center point versus taking account this um these full these different distributional responses now i don't know how to wait high and low damages that histogram is not a statement of probabilities it's just a bunch of uh recorded um differences across models so i can use that histogram as a baseline starting point but it's a bunch but how about if i move those a little bit so what this shows is that if i don't have full confidence in how to weight these things then then i'm really led to a substantial contribution of uncertainty to the social cost of carbon now this of course depends on statements of aversion and stuff like this but all i want to illustrate here is the uncertainty component to this important cost can be very substantial so how about private sector uncertainty so private sectors also faces uncertainty with this fundamental type i've talked about not knowing what the what's what's what the temperature the long-term temperature responses will be to emissions or not knowing what economic damages will be but they are um but they also face what i think of as policy uncertainty and a lot of the uncertainties they're facing in the short run is speculations on what type of government interventions they'll be in the future so like if i'm sitting on a big kind of uh set of carbon reserves and i think there's going to be some policy coming down very very soon that that's going to make those very very costly and that's going to lead me to want to exploit those reserves quickly of course if i think it's down the road always i will behave differently a lot of what we see in the policy responses now have to do with the uh or i'm not sure that the the private sector responses have to do with this policy uncertainty not necessarily the fundamental uncertainty so policy uncertainty is also an important part of this in terms of the the current policy machine now the models i talked about to do those calculations was highly simplified and we want to push it to you know enrich it in a variety of ways thinking about how do we accelerate we want to accelerate the shift from fossil fuels to renewable energy if so how much do we want to subsidize that what are the best ways to do that how much we want to you know feature nature-based solutions like increasing sync capacity and and resilience through our forestry and agriculture and oceans and food systems how much we want to be thinking about trying to advance adaptation helping you know making global efforts to uh to accept the faculty climate change but to uh manage the impacts of it i mean these are all very very important policy questions and uh and the point i want to make here is as we enrich in the analyses this is going to leave in place just going to open up new channels with uncertain consequences now before closing here i want to make one point here and and i want to be very clear there's lots of people in the social arena that say well since we don't know things we shouldn't do stuff that's not what this is this is this says just just the possibility of bad outcomes should be enough to make us want to act today so lots of policy makers are afraid to talk about uncertainty for fear they're just going to say well therefore we shouldn't do anything but that's not the message of decision theory under uncertainty it's the possibilities of bad outcomes um are really enough to make us think hard about the design of policy so with that i think i've talked long enough what i'd like to do is just end with a with a quote from one of my favorite american authors mark twain education is the path from cocky ignorance to miserable uncertainty what i've hoped i've laid out here is uncertainty candidates can in time seem miserable but there are ways to think about it systematically sensibly that can help us proceed in making better decisions in the future thank you very much for your time i'm open to some questions thank you very much there is a question from the audience online that is it's a general question growth growth and uncertainty can the situation in developing countries can be the growth in developing countries can be reconciled with sustainability so is sustainability possible not only in developed countries but also in developing countries i.e uncertainty and growth are they compatible with a sustainable approach um i wonder if you might be able to repeat a little bit the english version of it i was having a hard time with the simultaneous italian it um and english i know it has to do with growth and sustainability but i'm not sure i got the precise question is it possible to have to repeat it compatible with sustainability even in counties that are not yet so developed so if if the the somehow all these pose a challenge in countries they still have to fully develop to sustainability or uh of course it's a very vague question it's a very it's a one million question one million dollar question yes yes at least for sure okay thank you yeah so one thing i did not talk about and i it's an absolutely important question and that's the um the heterogeneous effects of climate change across different regions around the world and and this is um the work in this area has been um uh too limited to date in part because and there's lots of there's scope for a lot more interesting work along these directions because climate change obviously affects different regions and different economies in very different ways and that and and framing a coherent world approach to climate change that involves taking that into account the question about sustainability is an interesting one um i certainly do think we want to approach how we uh that the cost of climate change and to recognize the existing differential situations around um countries and regions and make sure that we kind of distribute costs in some type of sensible and seemingly more efficable way um sustainability is going to be possible to the extent that it's going to require some you know technological thinking and some ways of redoing economic activities and the like uh i i i am cautiously optimistic that um we we can figure out ways to to build sustainable economies over the long run that are that that don't damage our economy uh to require you know multiple efforts and and the like uh i don't think of an easy problem but i do think it's one that um with innovation and and the right incentives that we can't address in sensible ways i agree with professor maranachi we could have a full hour session on this topic and probably still not settle that so fully resolve it oh i can't hear me i can't hear yet yeah well there's aspects of it that were certainly disappointing um it's interesting as you watch there there were attempts by many policymakers to look at model outputs um we were handcuffed however because the models were largely epidemiological models and they didn't really put in the full economic incentive effects that they're important to think about thinking about policy implications and this is because economists and epidemiologists have really not collaborated on what we think of as integrated assessment models in which you really fully integrate the epidemiol epidemiology with the economics there's been lots of very counterproductive discussions about well there is no trade-off we should just care about health and elect but there are trade-offs i mean we can't close down the entire economy certain portions of the economy have to be functioning i mean which um we make we have to make decisions about how we're going to engage in education going forward what type of you know which among the activities are quote essential activities such that that should continue um how do we structure you know once we put policies out there we can you know um how do we make sure that people you know conform with them and and and the like so trade-offs are inherent here and even within the health arena there's there's trade-offs because if you start isolating the population there are lots of evidence that isolation itself can cause psychological and and say health damages um you know even if it's not directly attributed to covalent is indirectly attributed to it and there's just no doubt that that's the others that there's adverse health consequences through isolation so trade-offs are kind of there throughout all this um i do think lots of policy makers we just didn't have the modeling tools to be as helpful as we might be now the ones i didn't like very much were the ones that say well there's heterogeneous model outputs i'll just take i'll just find the model output i like and i'll go with that because and that led to i think something to some counterproductive policies uh the the policies of the u.s at uh at the federal level at times kind of appeared like that and and it just was and that part was quite frustrating um so i do think you know some sensible policy makers really didn't want to try to use modeling outputs but the model that's but but the type of models we had at our disposal were while they had some solid epidemiological underpinnings the the quantifications of them were still kind of filled with unknowns and quite primitive and enter in the inter interconnections between economics and epidemiology we're not we're not appreciated except that they need to be that's yeah that's part of this warning about the the one model issue if i put heterogeneous models out there then there's a danger of policymakers going to look across models and say oh well i like the output from that one so we'll go with that model all in and and that's part of the danger of policy but that's the danger of putting all your weight on one model arias with these uh catastrophes these all these what's happening it will finally motivate people to collaborate across borders or what's your sense uh yeah so yeah obviously it's a very it seems on the surface very very attractive and and and in some sense the right thing to do it is very hard to do well um i've seen such some some partial successes and some big failures if you just kind of imagine that you're going to say well the epidemiologist will build this part of the model the economist will build this partner model and staple them together that those efforts almost you know never are all are never all that productive the ones that are more interesting is when there's an exact dialogue that takes place and the epidemiologists have to understand a little you know something about how the economic work model worked and the economist have to understand to some extent how the epidemiological models work the same things through climate change i mean i um in order to do that well you have to make sure there's understanding going both directions and and not just kind of let my field do my thing in your field do your thing we'll staple them all together that part you know seldom works and and to build this mutual understanding that takes a lot of effort and that takes scholars that are really committed to want to do that so i think it's possible to do it well i think it's hard to do it well i do think that if you do it well there are potential benefits for how science can help society this may also impact education in the sense my call for a reform of education creating people that have the ability to understand both words in a sense so nice poor need for also some change even in the undergraduate education or just more advanced so more flexible programs that people are able to to grasp issues from different from social natural sciences yeah yeah i said i i haven't thought hard about the right way to design such curricula but i i absolutely do think it's that that's the case and you know that the skill set that are just most valuable are the ones that buy the flexibility and and that's the skill set that by the ability to understand what's going on in multiple arenas so i i'm but but then it's but then one can also suffer from superficiality if you have a very superficial knowledge of epidemiology and a very superficial knowledge of economics and you kind of you're probably going to have struggle as well there has to be some there has to be some depth the same time ability to communicate across the areas yeah yeah thank you lars support this illuminating conversation so if there are no other questions i will i thank you very much for all the all the wisdom that you gave us and and let's hope that actual policy making will be responsive to these stimuli scientific stimuli so finally there will be more room for rational policy making indeed let's hope for yes indeed thank you very much for this opportunity thank you thank you very much for the illumination thank you sure so everything okay you