 I'm pleased to meet DriverBody, y Majesty's ring as it's fantastic to see so many of you here. Thank you all ever so much for coming to this seminar which is co-hosted by both the Soccomорошing Environment Institute y Yo-aid. I should introduce myself first I guess First... ..and it's also co-hosted by Yesy. Our environmental sustainable institute. Unfortunately Sue can't be here just now but we have strum crawl also produced. sy'n fath, mae'n rhan o'i gwleidio byddwn yn enw. Felly mae'r thwyddiad byddechrau'i gweld. Felly, mae'n teimlo bod, fel younw, mae'n rhan o'r rwyf ysgol ychydig yw'r rhesymau arweinydd o'i gweld draws ar fy tyw. Mae'n rhaid pethau i hyn. A i fyddechrau, fel yw'r wych yn CWHó, sy'n nhw wedi bod yn rhywf yn gwahodogi arbennig. Mae'r solwyn ar y taethau arda nhw wedi... Felly, ydych chi'n gwneud y fath Yn gyfwledd ar ei wneud, gallant nawr yn cael eu cyflwrach ei stornyddiadau eu ramordeb fel GAud. A ydych nhw'n iawn atёт oedd y traddwl wedi gwneud y cyfle ac rwy'i gwirionedd. Mae'r rai cyfweld arall, ac mae'r llwyffodol yn ffamil yng nghydweithio ar fy nghylo sy'n unigau yng nghylch. Mae'r llwyffodol yn gyflym i'w gweithio o'r llwyffodol. Mae'r llwyffodol yn y Llywodraeth, Sufdaint, Acre ac driver y Swindol. Mae'r llwyffodol yn cyfnogi'r llwyffodol. Just to try and put some of the values in some sort of context, so Denny has a PM 2.5 concentration, annual mean concentration of 1.2 micrograms per metre cube and just so that you've got something to relate that to, York's PM 2.5 annual mean concentration is 12 micrograms per metre cube which is still above the world's organization of 10. So there are really big disparities in terms of the levels of evolution going in and we want to try and make sense of all of that in terms of how do we actually cost the impact associated with our evolution so we can hopefully have a signal to do something about it. So we got a couple of firsts, I think in trying to do this email, one of the perhaps most challenging things is that we are linking up live with Stockholm So we have one of our panel members, Johan Shulun-Shirna, our policy director of SDIU, is in Stockholm. I hope he can hear me now, but I won't expect him to respond to this now. But he will come in as we have the panel debate. I should give you just a quick introduction to how we're going to find structure of this seminar. It might have been him. So we're going to have a first-minute seminar from Nick and I'll give a pop-introduction to Nick in just a minute. We're then going to have a panel discussion with six members, I think, on the panel. Someone from the top over there, but I'll let Karen introduce those people to you as well, because you'll probably forget, by the time we get around to that. And during that discussion, we're hoping that we can get some questions from the audience, as well as questions to inform within the panel themselves. It would be great. So I should introduce Nick Miller. I first met Nick. I think all that five years ago, or something like that. We were working together on the UNEPWMO Black Carbon and Ozone assessment, which came out in 2011, and was quite instrumental in setting this in for further work now on short-lived conduct solutions. And Nick was responsible for the economic valuation of the impacts of SLCPs on three different things, climate change, human health, and one of his favorite ecosystems, and we will not talk about that today. And you were absolutely brilliant in that report, I think, because you helped all of us make some kind of economic sense of the results that you were producing. Time for the background to you. You received your PhD in Environmental and Natural Resources Economics from Yale in 2007. And since then, you've been an associate professor at Middlebury College working in the Department of Economics. And you're also involved in a number of important working groups in the States including the US EPA's panel on the economy-wide modelling of the benefits and costs of environmental regulation. And you've also advised the National Bureau of Early Climate Research. So you have a wealth of experience and knowledge that hopefully we can learn from today. So I will stop there. And without further ado, Nick, the floor is yours. So thank you, Lisa. Thanks for the invitation to talk. It's very much a pleasure to be here. Just in terms of motivation for the topic, economic development and the production of things and the consumption of things produces air pollution. That's not news to anybody. And it's also not news that historically throughout the stages of development that pollution was often produced in an innovative fashion. And so I'm sure we've seen pictures like this, perhaps, not from this particular place. This was taken in the late 18th century. No, 19th century in Pittsburgh where I'm from originally in the process of manufacturing steel. And there's just no attempt here to do it, right? Not even on greater. There's another picture, perhaps more impressive in a sense. Same era, same place, same process, the manufacturing steel. And when that degree of pollution, in the sense of the lack of the data was produced on the scale that it was produced in Pittsburgh and in other places, led to outcomes like this. Now it's not diminished the importance of use of coal and home heating and early automobile use, but here we have roughly middle of the day and mid-warning hours with darkened skies. So when we have outcomes like that, those early episodes generated concern among citizens, right? And in a sense, perhaps indirectly, in government. It's intuitive that such episodes, such degrees of pollution, would potentially have effects on in healthy environment and broader ecosystems and perhaps anti-centrally directly on health. Perhaps what's more important from the policy point of view is certainly today is whether background levels and chronic exposure, long-term exposure to those background levels has an effect as well. And then the question is raised, were or are there demonstrable quantitative, statistically significant effects associated with those exposures starting early on and moving towards the present? Well, there's lots of literature as we stand here today. There's lots of literature that has been published in epidemiology and in other fields that has demonstrated this link. I'm just showing one particular piece that was published relatively early on as far as this literature goes. And that was an article by Lester Leven, he's been assessed in that period in science roughly in 1970. Yeah, published in 1970. And this paper, it's credited with being one of the first. I'm not actually sure if it is the first. That demonstrated in a statistical sense that there was a significant association between exposure to PM-10 and adult mortality rates. Recently, the literature has supported these findings broadly. I'm citing two papers here that focus on the US. These are recent publications, but they are only recent editions of a long-standing set of studies that have adapted in 1980s. And they basically agree with what Leven Sesson found. These particular studies are focusing on PM-3.5, but they're basically showing the same thing. Statistically significant associations between exposure-defined particles and adult mortality rates. Importantly, and I'm going to pick a little bit on the WHO standard, this research, the Leven Sesson research, these papers, and so far as I'm aware, the global burden of disease study don't show a safe level of exposure to these local air pollutants. There's no clear threshold at 10 micrograms per cubic metre or 8 or 15 beneath which we or ecosystems are objectively safe. And so I'm pointing that out in part to co-plan at the WHO, but also from the point of view of thinking critically about policy design and thinking critically about benefit cost analysis, if we don't have a level of air pollution at which risk goes to zero, then the choice of the target is really difficult. Some might say artificial. Okay, so we've established, the literature has established, that epidemiology shows a significant association between human health and air pollution. But critically I think we need to step back and ask, as an economist this might not surprise you that I'm asking this, do we need policy to protect human health insofar as that is affected by air pollution? Or will the decentralized activities and choices of individuals and firms arrive at the right level of air pollution? Now, tongue is firmly in cheek here. We saw the early pictures, right? So you know what the answer is going to be, but I raised this nonetheless. So what does a microeconomist have to say about this? What does the discipline generally have to say about this? Well, we know one way to think about pollution is that these are wastes. They are residuals generated through the process of production of things, steel, or consumption of things, like motor fuel, that we put in our view. Some residuals are priced in markets. Think some, not all, but some solid waste, right? Firm space costs to dispose of used cardboard, individual households, I know at home, and pay a fee for a firm to come and cart away garbage. So some residuals are priced, and, of course, some residuals are not priced, right? CO2 is probably the best example of an un-priced residual, but so are the local air pollutants, some water pollutants in the market. So what? Well, if those residuals are priced, at least in principle, firms and consumers face the cost of disclosure, right? So it's a cost of doing business to a firm to dispose of solid waste, and they recognize that and they adjust accordingly. We then may still have damages. We might still have health effects, but in so far as the prices for those residuals reflect true value, then we're going to have an allocation, an efficient allocation of production of that thing, which is generating those residuals. Far more problematic is the case where residuals are not priced, right? Where costs are hidden, or to an economist, they're external to a market. Right? Because in that case, if costs are truly external, then the classic market mechanism has an efficient tool to allocate the core resources across society to break this down. Crisis for products are inefficiently low, cost of production is inefficiently low, and the market yields an inefficient allocation of resources. For instance, we've earned too much coal to make power because the costs associated with combustion of coal aren't generally fully-faced by farmers. So to most micro-economists, that fact in and of itself justifies policy. It justifies an intervention to correct what we'll see in a minute with the coal market failure. Two other implications that I think are probably less, I'm not sure if they're less well-known, but they're less often thought of, is that measures of output, if we have externality, measures of output may be inaccurate, that is, GDP might not capture the number, right? And in addition, measures of growth in that thing, GDP, GNP, whatever it is, may also be inaccurate. And growth matters both from this sort of public perception of how the economic system is doing. You tend to think, well, GDP was growing at extra cent last quarter, things must be okay. I'm going to invest in this particular asset. Policy makers pay attention to growth rates more than many other metrics of economic health. Macro-economic policy makers, in particular, pay attention to growth rates in setting interest rates, et cetera, et cetera. So the point is, there are broad implications when we expand the notion of externality to be one that affects outputting growth, there are broad implications which further, in my view, justify intervention in markets where externality is going forward. So I asked that question before, I answered it in some sense right away with the pictures on the first slide, but I answered it again from the perspective of microeconomics and I'm willing to bet that there are lots of other perspectives that would also answer that question in the government. So in terms of next steps, that is, we've identified a market failure, that is the presence of pollution, unpriced pollution. We now need to think about if poverty means to correct that market failure, to design policy. And again, from the economist's perspective, the polluters need to face the cost of pollution to adjust their cost of production appropriately, or on the other side of the market, demanders need to face the full price of the product. So determining the cost of pollution requires estimating impacts of pollution because those are the cost. Health effects, species loss, property damage, whatever that may be. And one approach to doing that uses what are called integrated assessment models. And these tools effectively delink emissions, the human activity, all the way through physical environmental modeling to impact and ultimately valuation. And just a citation from some folks on the screen in the room of a recent example of integrated assessment modeling applied to both climate and local areas. So this is useful, hopefully, in a self-evidence sense, but we can also be more specific and say, look, the use of integrated assessment modeling is helpful from the ex-anti-perspective in asking what will it cost to implement a policy and what does society stand to gain in terms of a reduction of damage? It's useful from the perspective of ex-post policy evaluation. How did the policy do? Did it work? Did we generate benefits? What were the costs? How many jobs were lost, et cetera? There are, of course, pitfalls associated with this, as there are in many empirical exercises, the first of which is each stage in the modeling process, as we'll talk about in a minute more specifically, introduces uncertainty, perhaps significant uncertainty. I'll spend a little bit more time talking about the final stage of the integrated assessment model, which is evaluation, converting impacts into their dollar of equivalence, and that can make a lot of folks really upset. It's a very contentious undertaking. I'll try to argue to you and try to convince you that it's worthwhile, all the while recognizing that it is contentious and that there's no one right answer. And then finally, as I mentioned before, in this talk, I'm going to push a little bit more on this idea of integrating benefit-cost analysis into our measures of output and growth, from the perspective of trying to break down the illusion that there's always a zero-sum component to environmental policy. We do environmental policy at the expense of growth before we have maximum growth and we forget about the environment. That final bullet is going to try to break it down again. So I've tackled the introduction. The remaining portions of the talk will consist of some really conceptual model and it's just more of a conceptual discussion of benefit-cost analysis. I'll talk about empirical modeling with integrated assessment and then just a smattering, a touch of results from some recent work and on the conclusions before we get to the panel. So benefit-cost analysis, I went back and sort of looked up literature on benefit-cost and read around a bid and I've found this statement that sums up the rationale fairly effectively and activity is worth doing if benefits exceed the cost. I mean, who can argue with that? It's just, it's got to be true. So in broad terms this seems sensible but I think when we start looking at things a little bit more carefully we need to be more careful about that conclusion and I think most policy makers are going to be clear more careful than just following that guidance. So when we think about winners and losers which is an important implication of benefit-cost analysis, this early literature focused on what's called Caldera Higgs compensation and that very simply means that if there are winners given a policy intervention they can compensate the losers, and therefore in principle society can be made better off given that compensation we have net benefits. Now no one's ever arguing that this happens in practice. It's always a conceptual definition it's always a conceptual criteria. But we do have papers that actually explore and reveal that some economists do care about this judicial concern and in particular thinking about whether benefits accrue to particular income, neighborhoods, races, segments of the income distribution and whether the losers seemingly concentrate in different races, segments of the income distribution etc. And talk and think very carefully about that and evaluate policies according to some metric that pays attention to those distribution issues. So we go from utterly insensitive to these concerns to far more sensitive. My other concern here is that there's an asymmetry in benefit cost analysis which is well known this is something really new but there's an asymmetry in the sense that costs this is cost of policy not cost of pollution cost of policy are very easily monetized we invest in pollution control equipment, we switch fuels we purchase natural gas instead of coal those are monetizable things there are other costs but the bulk often manifest in those categories for environmental policy. In contrast benefits may be partially or quasi monetizable at best and we can think about cases in particular thinking about mortality risk or evaluation of ecosystems or species where it's really problematic and difficult to think about valuation on the benefit side it's therefore quite possible that benefit cost analysis leads to a biased sense of the merits of the policy just based on the ease with which we're going to monitor. So back to this integration between benefit cost analysis and the national accounts the benefits of environmental policy really are reductions in damage reductions in these adverse investments and if those are happening to a market those benefits aren't going to by definition show up in GDP they can't in terms of conventionally measured GDP if they're truly external to the market they can't be done conversely most costs lie within the market boundary again purchasing pollution control equipment that's a purchase from one firm to another buying natural gas instead of buying coal again within the market boundary so broadly when we think about linking environmental policy to GDP we have at least the strong possibility for a biased picture of environmental policy when it's the cost that are going to lie within the market boundary and the benefits that by and large aren't and in light of that I would propose that we need an alternative measure of national output and this is again not a new suggestion it's in the literature I'm just emphasizing it we need a measure that in fact recognises these costs these non-market costs or external costs and that measure would simply be in this case defined as environmental and adjusted value added which would just be GDP less pollution than let's recognize full cost of production on to empirical modelling as I mentioned integrated assessment is a tool that's commonly used in this space it's a tool that connects anthropogenic activity anthropocentric activity emissions broadly to environmental consequences these tools have been around for a long time I think some of the earliest ones were in the late 70s and early 1980s that were focusing on local air pollution impacts but they've since been applied in the climate space as well and we're now getting integrated assessment in water in other areas and recognize that IAMs or integrated assessment models embody what's called a damage function approach and you can see the damage function here in that we're connecting emissions produced by some human activity through physical modelling to ambient concentrations to exposure maybe people maybe other things ecosystems to a link to physical effects because not all exposed persons exhibit a case of asthma or a change in their mortality rate we need some functional relationship that's going to establish that link and then finally the last stage is monetization so within the context of policy evaluation we might have some intervention for the US that might have been passage of the clean air act or a set of amendments to the clean air act and that's of course going to generate costs and I'm going to propose that that's a separate calculation and modelling exercise in integrated assessment so in a sense that happens offline but also recognize that the policy intervention if it does anything right is going to change emissions it's going to modify human activity otherwise it's not fine and when that happens all of these other things change in the model that is to say we have the change in emissions and that gets processed through the physical model which links emissions to concentrations and all these other modules reflect that change if the model is working right so we get all the way through to valuation and it's the change in valuation it's the change in damage that's going to reveal the estimate of benefits right presumably although it's not always the case presumably the policy intervention reduces damages in most cases all of these things fall including damages and therein lies the benefit right so benefit cost analysis then it's just about comparing those two modules however we decide to do that so I'm going to refer to some empirical results in just a few minutes and the model that I use to generate those results and to do that research is the AP2 model which has been used in lots of prior publications it's been used by the US National Academies of Science and EPA and you can see in its basic structure it's very much like the generalized integrated assessment model that I had on the screen so in order to I suppose inform you about the process that led to the results I'm about to show I just thought I'd go through very quickly in terms of data sources to fix ideas in terms of where the results come from in terms of emissions as is standard practice no matter where you are I suppose I'm using emissions data in this case provided by EPA right and so these are measured or estimated emissions for things like power plants and steel manufacturers and estimated emissions for things like vehicles I then use an air quality model to link those emissions to ambient concentration levels across the US I use vital statistics data for exposures in particular county specific population estimates by age group and crop specific crop census data from US Department of Agriculture as well and then I use dose response functions and in particular for PM 2.5 the two dose response functions that I alluded to earlier in the talk form the link between exposure and some estimated change in mortality risk and those are mathematical functions that come from epidemiological papers from peer reviewed research and then finally in terms of valuation I'm going to spend some time talking about because it's the one I probably know more about than the others but also because it's a place where we have in a sense a lot to discuss as we think about gauging the impact from air pollution and therefore the impact from policy so this is the monetization phase where we go from changing risk or changing yield if we're talking about crops to dollars and so I'm going to ask a question that is why do we do this and I'm also going to speed up this on the time guidance I just got so why do we do this number one it enables us to rank damages across pollutants what's a ton of PM 2.5 worth relative to SO2 or CO2 across sources power plant A versus power plant B across industries and across sectors so ranking aggregation across different damage types we're going to look at mortality risks cases of illness crop yields materials depreciation all expressed in different physical units you can't add them together if you want the total impact of a ton or a kilowatt hour you can't add those together unless you have a common metric and third and perhaps most straightforward it allows us to evaluate benefits directly against costs costs are almost always expressed in other terms monetary terms and we need benefits to be monetized in order to make back comparison okay how do we do this if I have my preference we look to markets because markets are a great source of information for the value of things but we don't have markets for things like mortality risks in species the existence of or ecosystems so when we're in a situation where we don't have information for markets we employ non-market valuation techniques of which there are two types revealed preference we look for evidence of people's values for these things and market transactions that are linked to environmental goods and services that is the value of a view in a real estate market for instance or we ask people directly on surveys when thinking about the value for mortality risk we know that these values vary across human characteristics demographic characteristics and I'm focusing here on the value attributed to mortality risk because it is the dominant source of damage monetary damage when you think about local air pollution so differences in value across age this has been estimated by Keith Fascusi and Joe Aldi and they find an inverted U shape with the value maximizing in middle age I'm going to go relatively quickly the literature has found differences in willingness to pay to avoid cancer risk across race types EPA has although it doesn't use in its standard benefit cost analysis an income value of a statistical life elasticity richer people are more willing to pay to avoid mortality risk I'm not going to go to gender the literature hasn't really gone to gender either in a policy framework recognize that with political constraints none of these sources of variation and the value for mortality risks are applied to standard practice that means that our estimates are probably off one way or the other for different groups but it's pretty hard to imagine applying different values across those groups defensively from the point of view of public policy okay so now on to results quickly these are taken from a given 2014 paper of mine we have aggregate damages aggregate damages from air pollution outside of brackets and air pollution in CO2 equivalent emissions this is economy wide for the US from 1999 to 2008 and you can see these are falling they're falling in both cases so here's thinking about benefit cost analysis one reason that they're falling is the end of the clean air the only reason there was a recession coming on in here so here's EVA GDP less that air pollution damage and here's pollution intensity damages were on the order of 6 to 8% of GDP in the US in 1999 and they fell quite a bit even relative to GDP okay now we get to growth rates pollution damages are falling and here's the kick when you think about market rates of growth in the US economy growth is on the order of 1 to 3% those are numbers taken directly from the Bureau of Economic Analysis their annualized rates of growth what's really interesting is that the adjusted indicator GDP less air pollution and climate damage is rising more quickly than the market indicator between half and a third of a percentage point so trust me when that first came out of the computer I re-ran it several times because it was not what I was expecting to see until you look at the picture this is just for the utility sector in the US but this is the market measure of output in real terms it's basically flat pollution damages are falling and therefore the difference between the adjusted indicator the dashed line and the adjusted indicator is shrinking the bottom line is catching up the bottom line is the adjusted indicator and if it's catching up it's got a steeper slope and slope is an indicator for annual growth rate so if GDP is rising while pollution damage is falling that adjusted indicator is going to be more rapidly growing than the market indicator and updated results for 2011 I think I'm just about out of time to say there's lots of heterogeneity within a country in terms of pollution intensity I'm going to pick on West Virginia lots of coal mining lots of coal-fired power generation lots of petrochemical manufacturing damages on the order of a fifth to a third of GDP huge air pollution damages North Dakota similar except add oil and gas extraction 1999 the colors and the meaning in terms of quantitative scale stays the same when I changed to 2008 dramatic reductions in pollution intensity spatially that are concentrated in manufacturing, utilities and oil and gas extraction so on the conclusions costs being careful damages from pollution often manifest outside of markets these are the externalities that suggests to an economist a need for a policy intervention to correct that market failure thus we have benefit cost analysis benefit cost analysis to be done requires damage measurement otherwise there's nothing on the benefit side damage measurement is complex but I would argue feasible and we have results like this that come out of damage measurement that are revealing and in particular thinking about the difference in growth rates between the adjusted indicator and the market indicator is to the integration then of benefit cost analysis with the national accounts and in terms of progress I close with a picture of the same street in Pittsburgh roughly today-ish and not today thank you very much so thank you for coming everybody I think before we kick off with a panel debate does anybody have any questions of clarification or more general questions for Nick Thomas is there been in the middle when you how did you get to your cost what are the assumptions where should I start just give me the main one for example the cost of human life how did you quantify that so what I'm using is the value of the statistical life which is an awful term for what's actually an intuitive trade off that we all make that is the process evaluation here is eliciting a trade off between a small amount of money and a small amount of change in our chance of death sounds crazy until you think about purchasing a smoke detector or a fire extinguisher in your home or a bicycle helmet or bottled water when municipal sources aren't clean this particular estimate is coming from the labor market and the labor market is showing us the rate of trade off between on the job mortality risk and wages an idea that goes back to Adam Smith this idea of compensating differentials is it imperfect? I hope this emphasize that it is because we know that's going to depend on income there's evidence that it depends on race there's evidence that it depends on age right I think the best way we move forward with this approach is applying the uniform value assumption broadly number one number two is the link between inhalation of fine particles and the effect on mortality risk epidemiological evidence as I said has been around since the 60s and 70s that there is a link but it's an association these are not laboratory studies nor should they be to be clear but causation is not crisp but those two are the primary sources of uncertainty and they are the key in linking a mission to ultimately to monetizable damage you haven't given me a number one I like that I can't explain this better but in the UK it's known as a quality a quality adjusted life year I can't give you a quality I can give you a value of a life year that comes from this estimate it would be around 200,000 US dollars I would say your $2,000 if that's a gauge and that would be around 35 to 40 years of age pegged at that because that's what it would be it gives you a call factor yes please I went to Carnegie Mellon so I know what's very well part of the reason the sky is different now is they closed on the steel mill so does any of these costing look at the exporting costs free in my model no an offshoring and outsourcing is a very important part of the story so here part of what's going on here is air pollution policy but part of what's going on here is macroeconomic change and changing the sectoral mix ideally without time constraints without budgetary constraints we would model this globally and then you wouldn't miss that exported piece so your point is a fair one I have not done the decomposition for these results to show what percent is coming from just moving where the production happens let alone the case that air pollution policy may be less stringent where the production is moving in which case who knows how this goes but from a first pass national accounts point of view fairly powerful so you must have known Leicester Lay or you knew of Leicester Lay yeah excellent any other quick questions for Nick just a quick one is it just costs of deaths that's a factor in Twitter so this is a really tricky question standard practice at EPA is to count the morbidity states so short term and long term illness in addition to mortality that is to say the mortality risk the changes in risk what you're seeing here is an attempt to be about double counting in the sense that you would often have say a case of chronic bronchitis that leads to a death or an ER visit that's valuable that then is immediately preceded by a change in death rates I'm not doing that because as I said being careful about not double counting but if you did calculating all of those morbidity costs these damages would be higher not by too much 5 to 10% at least according to EPA's estimates but that's a fair point Can I ask a quick follow up question just in relation to the epidemiological work that you would have talked about just there seems to be a lot of really clever research coming from the US that would suggest that that association I don't usually imagine that it's an association but that it would be perhaps severely biased that it would be severe based perhaps overstatement I just don't have any perspectives on that I think there's a lot of clever stuff that's to use some special designs that would suggest So I'm going to say two things and I'm going to be careful in how I say them One, there's some of that work that has been applied recently in China using regression discontinuity designs that have supported the PM2.5 mortality relationship in at least a quasi-causal way Two, a lot of the literature you're specifically thinking of the econometric literature, right? Because it just supports the research that you've talked about in this study but it's just that it would suggest that the relationship is stronger than perhaps it previously believed it's not that it's coming up and saying that there is no link in fact it's going to say that there's a stronger link but perhaps commonly perceived just in this case So the papers that I'm aware of in the US setting that use quasi-experimental designs don't look at PM2.5 but they look at TSP or they look at PM10 and I'm not an epidemiologist so those of you who are can speak up but I think the mortality signal is much noisier in those coarser particle groupings than PM2.5 So it's not insofar as I'm aware of that literature it's not quite yet an apples to apples comparison in the US Yes, thanks for the talk I was wondering if for the actual policy so when you say taxation you mean like the pavilion tax or do you mean existing taxes on some other product I guess more like the pavilion tax So most of my dissertation work focused on the estimation of what those taxes would look like would look like so these results can be used to inform policy design in that sense and in particular what I was doing was plant specific pavilion taxes for different pollutants and it's a really neat thing to do it's a lot of fun as a researcher I went shortly out there after the EPA and gave a talk about that and found out the hard way that there's just no way that we're going to move towards pavilion taxation number one and number two have it vary by plant or industry work So yes, the apparatus can be used in that way I have doubts about the political feasibility of that approach Loads of non-economists in the ring would you like to tell them what pavilion taxes sure Taxing is either a product or an emission according to the damage per physical unit so if we make a kilowatt hour at Trax power plant that's going to have some damage socially if that plant were taxed according to the monetary value of that impact that would be a no This is all making me very happy I've spoken with my background and it's a while since I've heard about Pigu and Calderdates, it's all marvellous Thank you very much So let's invite our panel to the front and introduce them we have four extra panel members in the room and one online How do we get Johan? I'm here OK, marvellous So let's introduce Johan first Johan Shulin Shiana Is that a sort of attempt at your name Apologies but that's completely wrong Johan is the policy director for the Stockholm Environment Institute and has done lots of research on air pollution policy in particular So welcome, thank you Mike Ashmore come Mike is professor of environmental science in the Stockholm Environment Institute here at York and does research on the impacts of air pollution on human health and ecosystems We have James Lomars James is a research fellow at the Centre for Health Economics in the team for economic evaluation and health technology assessment So he knows all about Qualys Thomas, he can ask James about Qualys as well Helen Weatherly also health economist at the Centre for Health Economics and Helen's worked particularly on public health recently also in the team for economic evaluation and health technology assessment and last but not least Professor Tim Dorran Tim is in the Department of Health Sciences Professor of Health Policy but also has a public health background and an interest in environment and health I should introduce myself as well My name is Karen Bloor As I said I'm a health economist by background but I'm also the university's research champion on health and well-being So I'm here as a complete innocence to air pollution and it's very nice to have you all here So now then how should we do a panel debate Is there anything that you'd like to respond to directly from the talk or any any issues that you think need raising that haven't been raised so far Would anyone like to pitch in with any of those kind of things or should I pick on more specific questions I can see the evidence generation if that's okay You said that we shouldn't be doing lab experiments and obviously there are lab experiments but there's usually rats and then people So we can look directly at that to physical harms of some of these pollutants and again my understanding is that when we look at the epidemiological environmental studies they're not always well linked to those sort of animal models so it's just a case of let's work out the relative risk of having an average exposure of x amount of particulates and a like that to cardiovascular disease Great, work out the traditional risk of death So a lot of things are moving on a little bit So in terms of the explicitly the link between top psychology as far as I'm aware EPA's Scientific Advisory Board is often more apt to use the criteria coming from the cohort studies that are tracking people across time really the early 90s I'm not familiar with the toxicology literature so I can't say whether on a consistent basis for the criteria air pollutants PM10, SO2, NOx etc that those are informing the setting of standards I think that they are much more likely to when thinking about the regulation of toxics right, these sort of other benzines and things like that that are not part of the criteria air pollution I think the other thing is that even if you're looking at short-term effects your main programmes are long-term criteria effects don't further ethical constraints because people have serious asthma or those can't be tested as well so there will be limitations I think what the literature is trying to do and maybe this is how we sit in the UK those cohort studies of general association which I think people initially were somewhat sceptical about in terms of causality and where the toxicology and the detailed studies not of how people would have mechanisms has helped us to develop a mechanistic understanding of how it can be of those small concentrations of high particles could be accelerating the death of the people through cardio-bots given disease so it's not the toxicology setting the standard that's providing some supportive mechanism for the assassination but I think understanding the limitations of those epidemiological studies the cohort studies were important there's a small number of them almost all of them in the States a few outside Europe and North America they're following people from what, 20, 30 years old onwards so they don't capture the full lifetime effects of air pollution which we now believe is from prenatal in me to prenatal through to over a lifetime those things aren't captured so there are real limitations to these studies which have to be taken into account in terms of the decaying by the ocean but presumably there are opportunities I don't want to diss any of the competition but with the natural experiment by the experiments in the US you're looking at gradual and actually quite small reductions of bloom and then modelling that against outcomes we could look at manufacturing of China or India where it's been a sudden sort of explosion and modelling that against health outcomes and that would give us a clearer signal also maybe some of the changes in the other direction like recession right and one of the papers I referenced in response to the earlier question did just that along the Hoare River there was a home heating, no home heating policy I think was the change in the dramatic difference just along that boundary and the estimate in terms of health effect on mortality was actually smaller than the cohort studies from the US and I think part of the intuition behind that was the flattening of the dose response function in high levels of people's lives but it's exploiting just that kind of technique that you're referring to a bigger change there are some really interesting things along from the US I think policy was the introduction of the easy pass to prevent idling toll boots and they found quite marked effect I think on some prenatal on birth health outcomes and on neonatal mortality rates as well right I think that was Walker and Curry so were they looking at residents or looking at people working in the woods residents fascinating can I just ask Helen and James particularly from Nick taking quite a cost benefit type approach which is sort of a traditional economic approach in this kind of area but it's not what we usually do in UK health economics we don't often value benefits in monetary terms I just wondered if you guys in the centre of health economics were approaching this in a cost benefit way or whether you're using cost utility the colleagues that so much mentioned I think as you were talking I was thinking about the US compared to the UK perspective and certainly it's well recognised in the UK in terms of health effectiveness to use a nice that's the nation's care and excellence approach to evaluation we accept these extra well forest approaches where outcomes are measured and health quality of life and value according to social preferences using a national tariff but of course we've got a sort of public sector approach to healthcare provision and I guess that influence is the kind of methods that one uses was wondering more generally in terms of the US perspective perhaps as far as I see it may sit differently very much more private market approach to provision and how that influences the valuation I'm not sure that it influences the valuation so much because remember the metric that I'm using which is industry standard in a sense for lack of a better term is an average value that's applied socially it's just coming from a different it's coming from a different conceptual foundation but it's being applied I think in the same way that is one value for you said a national tariff so it's being applied in that same way so I don't think that the spirit of private versus public is generating a difference in how the value is applied I think the difference is in how the value is estimated right I mean certainly colleagues in the chair I think people will change these things because of what's going on in winning Mr Payne's with more cost benefit analysis of features although perhaps some of our global health team I don't know if they see more of that at the time of winning Mr Payne's pay approach they use yeah I think well I'd probably agree with what you've been saying in the way it's applied is very similar so you have a unit of health we tend to use the quality as has already been mentioned but for our purposes we try and do that as the opportunity cost that arises from public sector spending so it is very much sort of a public sector oriented way of thinking about it and in order to do that it is slightly different how then those values are obtained as well so for the nice threshold that Helen mentioned the idea behind it is essentially that if we were to spend the money on NHS activities elsewhere it would generate this much health so that the new thing coming in or the new policy that's being evaluated needs to be at least as good as that to be accepted so it's kind of like an opportunity cost way of thinking about it and we tend to well we're doing some work at the moment in Centre for Health Economics in York where we try and estimate the relationship between the same qualities and then that informs our opportunity cost rather than going down to reveal preference or state of preference methods and is that the link between expenditure and health outcome made strictly in terms of public sector expenditure or is that gathering information from individual choices across the country well again it's kind of because of the UK I was thinking about the healthcare setting the vast majority of spending in this public sector expenditure you would look at things like purchasing vitamins or things that outside the healthcare system of course the Department of Transport would use a similar approach to yours they would have a statistical value of life rather than using a quality type approach so the quality metric is pretty much focused on the NHS really on the healthcare system outside that they use similar methods OK Yohan, can you hear us all and would you like to add anything? Yeah I can hear you very well so I would like to put into the debate it's my experience of working in developing countries so we recently held a policy dialogue as part of this thing called the Climate and Clean Air Coalition and there are members of that who are from different African, Asian, Latin American countries so what we were asking was what are the science needs that you require and this is often to ministries and environment in order to increase the interest in the issue of air pollution and so they mentioned that the need to estimate the health impacts which has been discussed here but also they definitely talked about the demand for some sort of economic assessment of the impacts of the benefits of taking action because they feel that without the quantitative information they can't motivate policy change so they definitely think it's worth making now in terms of the different cost estimates then of course we can calculate direct costs but we don't always have data from developing countries but at least there's a it's relatively simple when you're comparing the direct costs of mitigation to the direct cost for health but of course that's incomplete and so we then have the BSL approaches to cope with the full cost of air pollution but it becomes more difficult to explain what this means and so if we use the willingness to pay approaches then in developing countries there's often a limited data about willingness to pay but also I feel that policy makers sometimes get a bit lost in what it actually represents when they're comparing the costs of mitigation to the value of the health impacts using these approaches so and then of course the willingness to pay in a country will depend on its GDP so you tend to have lower values and that's what I see as being the controversy that is there and unfortunately the work that Nicholas did in the UNEDWMO assessment didn't make it to the summary for decision makers because of the worry by UNEDWMO that it was going to create more controversy so I guess a question I'd like to ask is how do we provide a simple explanation of when we compare in the cost of mitigation to the benefit value using willingness to pay methods in a way which is easily accessible by decision makers because I feel that sometimes it gets ignored by the complexity of explaining what it actually represents so I was wondering if Nicholas had or others had some guidance on how we can develop these numbers and get them used to influence policy in a very clear way so in terms of the development side that is the development of the numbers I don't have a sort of silver bullet or a solution it's a matter of doing field work in those particular parts of the world that lack the estimates I know China and India are places where folks in my shoes have developed DSL or willingness to pay these estimates and they've done that in a defensible way on the how do we explain the complexity or the intuition behind what the metric is actually representing I think I used it so I'm revealing preference for using this approach to an explanation but we actually lots of folks make these trade-offs all the time in their daily context in their daily lives and finding a country specific way in which that trade-off is made I think would go a long way towards relaxing people and backing them away from the thought that this is about valuing lives which it's not so when I said to the audience when we purchase a smoke detector that's relevant to all of us and I saw a lot of nodding like oh yeah I get that trade-off that's not going to work in all parts of the world but I would argue that there are probably instances in which that trade-off is being made money for risk and if policy makers could focus on that uncovering what those contextual specific trade-offs are going to be pretty helpful that this might be perhaps perhaps Mike as well how in these developing countries is their quality measured how's it modelled, how's it managed we're talking about a vast range of countries with different infrastructure facilities well absolutely but I suppose when you think about all the little monitors around the city like York that we have clearly we're in a completely different part of the world with a different set of technology and priorities maybe just to give you an example my colleagues have just done a small pilot study in formal settlement in Nairobi a huge city rapidly developing very quickly there are one or maybe two stations on a rooftop in the centre of town in the suburban area that's all there is and those by no means capture what's happening to those people those impoverished people subjected to informal waste burning local factory emissions household fuel combustion massive indoor preparation and that actually is not an incident of a problem in developed countries as well I think it is extreme in that sort of setting so I think if you go to Sub-Saharan Africa if you go to Latin America in many countries in Asia the monitoring is very limited so it has not captured the true exposure and therefore the true impacts of often those people whom those health impacts have been most severe and most important so how do we base I don't know what do we base calculations on what do we base estimator models on but perhaps I could input a bit yeah great, go on so we're actually here this is an area where we're trying to to help a number of different countries to use on the modelling front so some countries do have monitoring in cities and we're going to be working at the moment with Ghana and in Accra they have seven permanent monitoring stations so that's quite unusual for West Africa but we're then going to be trialling the development of a tool which can estimate emissions look at atmospheric transfer and estimate concentrated PM2.5 and then you use those concentration response functions that we discussed to estimate premature mortality so we're working with the national government there and in a further 10 countries around the world in Asia, Latin America and Africa to try and give them the tools to start to estimate, give quantitative numbers for the emissions for the impacts and we would like to add valuation to it because there is a demand for it and it's a question of how to do that the best way so there's a real demand for quantitative numbers because they feel they can't motivate policy change without having some numbers about what the problem is and what the impact is causing so I have one anecdotal piece presumably one of the constraints on air pollution monitoring networks is cost the device themselves I was recently on leave at Carnegie Mellon and found out that there's an effort underway to develop a low cost cell phone, smart phone size PM monitors that there's lots of issues associated with them at this point but the idea that we can move from high cost, low number of observation networks to much lower cost higher density networks is perhaps worth mentioning as a target moving forward Absolutely, that's a very important development but we need to be very careful about that accuracy of the term and actually the study I was mentioning got the people in the community themselves to take simple low cost monitors around and understand their own sets of what they experience but I think you mentioned equity and I think that's really important when we address these questions in terms of primary evaluation but of course it's all to do with the greater exposure often of poor communities for greater sensitivity in terms of the health impacts so all of these things need to be captured and it's also about what are the interventions that are important may be for those particular communities it's a completely different set of interventions indoor probably is definitely most cost effective in terms of needed interventions some of the broad national monitoring when you're looking at the big power stations so this issue of scale of how it addresses the impacts of the concerns particularly for the communities is really important I mentioned working with national governments but it's more than just national governments in terms of addressing this problem and quantifying what the benefits are Anyone else like to pick up on equity issues here between communities between generations all these different sort of how do we build any of this into an economic model I think those intergenerational issues are very important actually I didn't get from you talking about how far that's addressed in terms of thinking about the benefits and who benefits and by how much it's not the association it's the avoidable of impacts which of course can change a lot of other societal factors at the same time but air pollution is one of a number of risk factors associated with primitive cardiopathic illness so if we change the air pollution levels now maybe it won't make much difference for our transparency of somebody who's on the cusp as it were but you know what does it mean for a young child not so much here so I was just reading before the talk catching up on some form of grading if you will a senior thesis at Middlebury focusing on human capital formation among young kids and getting them into school and keeping them in school and certainly exposure to air pollution has been associated with missing school and so you know point taken right I hope it wasn't clear enough in saying this is one slice and what I presented is one slice and a perfect slice of an annual impact and there are all these sort of cumulative human capital formation type effects that manifest in ways that as of yet we're not really able to capture which speaks to the intergenerational piece because the human capital piece does last life times okay well I see all kinds of interesting perspectives around the written, certainly from the people I know and that's not all of you but do you have any questions or comments or just thoughts that you'd like to make around this whole area yes please It's spoken from pure ignorance but it seems like this suffers from the same issues as the whole climate change discussion about 20-30 years ago where there's enough models to provide some really interesting information but not enough to provide it consensus for any to go location is there any any movement towards that kind of consensus building model matching kind of behavior for these types of questions so one context that I can employ to provide an answer would be about a decade ago a number of us modeling the impacts of sulphur dioxide from power plants, coal-fired power plants in the US started to see evidence that the cap the aggregate limit on emissions in the US for that pollutant appeared to be far too high according to cost-benefit tools so if you value those tons the ton at the margin was about an order of magnitude more expensive than it was to remove that time my model was suggesting this other models were and you know back to association instead of causation here clearly as I make this common a couple of years later when there were revisions made to the cap in the ballpark of what we were seeing as far as what looked like equating the benefit per ton to the cost per ton so in terms of thinking about consensus across research teams in terms of thinking about consensus across different modeling types and different modeling approaches that's an example where I can't say it caused the change but the change did follow on the heels the publication of those results but I think it's also a question around how far of course we can improve the economic valuation we can improve the pollution model and the effects I guess there's an issue in the question around how far do you actually need to improve it before you demonstrate the refraction so the estimate in the UK is that we have 40,000 additional premature cases of premature mortality a year in the combined effects of PM2.5 and 0.2 that's a huge health burden to my non-economic sort of money working at it so now you thought that's a significant issue of course 20 billion pounds I think in terms of your figures that seems to me to be a huge quality driving you can just imagine what the numbers are in that area so I don't think we actually need to wait for these refinements and these improvements in terms of number of methodology to demonstrate the seriousness of the situation and the potential benefits of action so why aren't policymakers acting? well that's my question thank you I have a very good example on a doorstep it was just an electric new one a new London mayor and on the basis of some figure that says 10,000 people died because of air pollution he will now increase the tax on toxic vehicles so these models do work my question to you really was how did you get this number of 10,000 where does it come from? it comes from an official advisory body committee on the effects of air pollution who basically review the health evidence and do the calculations and work out the methodologies I think that particular number from London comes from public health England and done the calculation borough by borough across London but I am not here because she actually sits on coming up and she might be able to I do, people spend a lot of time organising over various coefficients so they will look at all the data that are available and do meta-views and just try and come up with the best number and quite usually on the committee there is an understanding that you have to really carefully use these coefficients but what the press will go and do and you do these 10,000 to 40,000 people who will die in that 120,000 by 70,000 per pound in a year so that is how you use these coefficients I think it is really to that term mortality because we could say premature mortality corn to cormie, that is a statistical construct or artefact to do the calculations and get the genealogy that very easily gets translated into deaths and people have this public perception that kind of air pollution comes along and somebody is perfectly healthy drops death actually it is one of a vast number of risk factors like obesity and exercise and diet etc etc and smoking which can precipitate a cardiovascular death or a spiritual death a little earlier so that is what we are really saying it is just one of those factors there is also a public health action on all of those other things but relatively different just a 10,000 number does it mean 10,000 life years or what does it actually mean they are an official figure but they could correct me is if you imagine those 29,000 I am talking to the UK figure now 29,000 for being 2.5 and you imagine there is an equal risk across the population I think it is nine months of years of life lost but the problem is we do not know how it is attributed across the population that is the average figure but it could be if it was only those 29,000 people I think the value is 12 years of life lost something like that the problem is translating it to what actually an individual person experiences in terms of potential life lost or risk of death and that of course will depend on whether they exercise what they tell it is not the evolution or what it is how evolution interacts with all those other factors is this important that is partially why there is a big inequality aspect to it all as well because not only are there more greater emissions of pollutants in deprived areas but deprived areas also have high base vulnerability to the conditions that would lead them to so they face a double one if you are from a deprived area in terms of the effects of that quality the inability to engage in a variety of behaviour go inside, turn on air conditioning or whatever it is maybe it just compounds that it is not clear cut in the UK so I think public health England's report was the most polluting area to see what is going on in Westminster in that case it might be levelling out bringing the cardiovascular risk of the welfare of people or something like that Kensingtonshire is a funny place because it is a very wealthy bit and it used to be north Kensingtonshire until I don't have to eat Kensingtonshire but I think in terms of what Sadiq Karn is planning to do I don't think any of these proposals in 12 years ago they wouldn't have worked it's only because they are following on this incremental change so they got the congestion charging in first and that was bloody difficult and then they got the funding from the bikes and you are sort of building on that and I think there is a limit to what people will accept thinking back to those pictures of Pittsburgh we had similar very drastic events in London you can look at the 1940s 1950s studies where you had big smogs and you could immediately see the effect that that had on you've seen big peaks in deaths and big peaks in hospital admissions and people could see it people would not see it you couldn't see down in the street and it was very visible and that means that people want to act on that whereas now it's invisible so you are trying to get across that you are constantly exposed to this invisible it's difficult and you are asking people to take in positions so when they tried to do congestion charging in greater Manchester on the lines of the London one the population didn't swallow it they didn't want it and the local MPs actually within Manchester wanted it but they had to get agreement from those in the leafy suburbs around and we have a little MP in Stockport who all the way through his campaign to get elected said that I'm terrible green I want to do things for the environment I realised he was now responsible for constituency that includes Wilmswell and Bramall and lots of people who drove big SUVs and suddenly he flipped and now he was against it because it was not in his political interest to support this congestion charge which would affect his constituents and from which they would not benefit because it's protecting the lungs of the people in greater Manchester they were going to drive their SUVs through so it's difficult to get you need to get a lot of people to agree with the direction you have to act with them to very often act against their own best interests to do it so it can be very difficult again Yes please So are the estimates of the kind of health damages and therefore the kind of financial costs do we think generally across the panel are we underestimating them currently are we over-estimating them right are we being too conservative just to get a feel so we're talking about the WHO and stuff giving out numbers where are we at Johan why don't you pitch in first once we let this lock go you won't get a word in So do you want me to intervene there Yes why not are we about right or overall underestimating health costs and economic costs Oh well I'm not an economist so I would bat that one back to Nicholas Muller I would like to see more data on the direct costs we're actually going to start doing some stuff with the World Health Organization to try and understand the cost to the health sector in different countries but I do think that it's this the big numbers are always the mortality and then that uses this value of statistical life which is a normal name but you know I don't think that anyone's come up with anything better than that but whether that is a you know how good the numbers are that come from that I'd rather ask an economist so I'll bat that one back to the panel in your Okay well we've got a screen in the room so let's give a go Well since I'm not an economist you won't get an answer from me Let's go with So maybe if I just add something to your audience and then we can let the economists go in So from my point of view I think the values are probably if you think about this in terms of the values in the UK okay if you're asking about the values globally then there's so many vast uncertainties that have been very hard to estimate but my guess is because we're relying on sort of poor population average concentrations we're missing a lot of very highly exposed people who have got a much bigger impact and my other argument is to say some of the underestimates we're only costing or valuing or estimating a limited number of impacts I've already mentioned perinatal and neonatal effects so the evidence of perinatal for effects during pregnancy has become stronger and stronger and low birth weight and premature birth can have vast consequences throughout people's lives those aren't in the picture at all so I would argue because it's a very significant lifetime effect from early childhood or even before that we're not actually capturing including in these estimates we're only capturing part of the picture which I think is the answer it's just worth reiterating the methodological piece are we valuing what we can value accurately and then there's the are we missing stuff and I think that it's probably true I'm just conjecturing that it's probably true that the latter component is in a sense bigger nobody's really looking at ecosystem effects in the monetization space in a way that reasonably reveals a value for a species it's just really hard to imagine doing and we know these effects are there the second thing is this isn't fair what I'm going to say but it's okay there are policy constraints as I talked about so we'll never be able to even if we wanted to in an unconstrained black box where we have the values that vary by race, by age, by income we could do this perfectly you can't secondly there's no ex post revelation of value in this space you can't go back after the policy intervention and say well really what was your willingness to pay back what is your value it just doesn't happen so it's a really important question it's an unanswerable question in a sense to a degree of precision and therefore I think the first piece is the one we need to focus on that is what are we missing and can we do better there I need to come with someone he's monstrously inaccurate but I don't think we're never going to get a very accurate figure on what kind of exposure to particulates is going to reduce so I think what we're arguing about is just how damaging it is and what kind of cost you can put on that by then we've got sophisticated enough that we can put a lower bound on that and say even if we accept a lower bound it's going to be cost effective to come in and do some of these interventions okay there's a question about that again I'm practicing a local pressure group on congestion I'm very interested in the idea of comments on the movement of cars from outside to in one of the levers I think that we want to try and use is the link between congestion that appeared on the monitoring station we want just outside my children's school but I noticed that it doesn't ever really hit the limits of the EU put on it which is one of the levers that we can use against to try to pressure the council into doing something about it so what does the panel think about the temporal variations in these statistics of their spikes that follow the daytime but then don't meet the averages that they trigger the EU warning you should just iron your charm next to it all that that was one big one that was made by Volkswagen back that's a very interesting question I'm discouraged and fascinated by the fact that environmental standards are almost always on the mean they're not always but they're almost always on the mean I'm doing research currently that looks at not the second moment which is the variance but the third which is skewness which is we can have a satisfaction of a standard and not only have days that vary above it but have that one killer day right where you know in Phoenix I'm aware of a reading within the last ten years it was 900 right so those happen and I'm not sure how we interpret them I'm not sure how the genealogists think about those days but I think this is a fascinating and important topic it is and I'm not in a genealogist I have some to add and I think the difficulty is for these current studies you're looking at people's exposure over decades you can't relate but we don't know how important the peaks are as opposed to the mean rates but what you can do in terms of symptoms and people's lung function is to move them quickly from one place to another so there have been studies in London for example where people have been moved from a very congested part of Oxford Street next to lots of diesel buses into Hyde Park into Hyde Park and you can see that effect immediately now of course we can't check from mortality it's like the sort of ethical problems of putting people in the chambers but there is evidence around which suggests that those short-term peaks will have effects on people's lung function and maybe even on their auspices so that's the sort of evidence you have to use but if you compare with the means that are based on these long-term epidemies or studies you just don't know perhaps I could give a smaller introduction yes sure one thing that's interesting is that when we're talking about these guidelines whether they're WHO or EU or whatever is that they are sort of considered to be something that is relatively clean but it doesn't mean to say there are no impacts below those limits so I was talking to Michael Brow who is an epidemiologist in Canada and he's been finding impacts down to two micrographs from eSQ2P and 2.5 which is below the WHO guideline of 10 and I think somebody mentioned there is evidence that there is no threshold to damage by PM2.5 so I think there's one thing about complying with the regulation but it doesn't mean that there are no benefits to go even cleaner and my new evidence is supporting that it's really interesting I mean children here are really sensitive for it but as I've emphasised a couple of times we don't just don't know what the lifetime consequences of this early exposure is yet we don't have that kind of don't we know that from history don't we know that from the kind of Pittsburgh and London smogs and that kind of thing principles and the data is there but I don't know I don't think people have actually haven't a quantitative relationship I'm not sure that's there that's nice PhD of course at those times whoever is monitoring we don't think that's how you reproduce it I can remember wandering through London smogs survived so far but I think there are states in the US where you come to the new school within 200 metres of a freeway in California so there are regardless of the economics there are areas in the US which have recognised this particular exposure and limited where you can build new schools and take measures to reduce the exposure and existing schools so I am going to be a umiologist there's a general problem we have and it's not just environmental health and environmental sciences it's more generally within health services researchers we do tend to look at averages I think this is one of these areas where that is completely inappropriate so there are those of us who actually now look at ranges and look at fluctuations and consider what we call critical periods so for a school a critical period is when you actually expose it's when the kids are going into the school and coming out of the school and that's likely to be when you're actually getting the peaks of pollution because it's commuters and parents dropping them off in the car so you should be looking at those periods when they're most at risk and looking at spikes or looking at levels at those particular times or when they were average so one of the reasons why I asked is that I really picked and came up from Imperial College to see where the intent was and they used an integrative unit so they took a unit of exposure and then added it across people's lifetimes and the results of that were truly frightening and I guess I wasn't I hadn't seen anything about NO2 exposure which of course was NOx exposure which was kind of the equivalent of fright on now so I was just wondering about this NOx exposure but again you shouldn't necessarily if you wait for the council to do something or if you wait for the national government to do something you'll be waiting a long time so I think if you can make the case to governors and there are schools that do walk to school days and all the rest of it to try and protect that population of kids within their school so you don't have to wait for the word to come from on high or for the population to come through I think there's local activism that you can advance which you can address some of those concerned without having to wait for the evidence to be approved and for policy makers to act on it okay well sorry but just thinking of peak loads of air pollution some things like PM are driven by natural occurrences like wildfire is there room in the model to bring in those kinds of things that are natural and aren't really they're external from the market so the most obvious place as I think about your question is the emission in motorways albeit in an imperfect way do both document and include biogenic sources and so there's the classic dust storm wildfire but then there's perhaps more less intuitive sources that is emissions from vegetation that are an important part of PM 2.5 in the southeastern US I'm not sure about in Europe and I mean those are really difficult problems to tackle right we have a hard enough time curtailing our own emissions and our own sort of drive to produce things and consume things but I'd like to say that from a modelling perspective they're there and from a modelling perspective at least in parts of the world the modeler could permit those to understand what is their average effect what share of the burden they're contributing as far as we believe in them indoors it's a good point well I think it's really time for a little drink of some kind so let me welcome Lisa and yeah Lisa okay thanks ever so much everybody thank you in just a minute we were sort of talking earlier who's going to do something up with this and I will do a very brief sum up I think it's been a really interesting discussion thank you so much for all the interaction that we've had people in the audience, the panel with Yohan done a fantastic job in Stockholm it's always difficult when you're offline I mean I think in summary I think we can see that there's lots of difficulties and uncertainties in trying to make these assessments it really looks like there is a problem, a big problem that's still continuing to increase and get worse in particular parts and places around the world and therefore policymakers and I guess ourselves as well in terms of our individual behaviour need to be doing something about this to try and reduce the impact so that's perhaps the kind of main take home message and then by the sounds of it there are lots of additional bits of research PhD studentship projects and everything else that could actually go in and pick apart many of the different sort of detailed aspects that we've talked about here today so that's been really great and we have an opportunity now for people to continue to have these discussions and network outside, we have drinks, we have nibbles so please don't go away stay behind and talk to people unfortunately you're having a hard time to do that so we'll have a drink for you so a proper first up I think, I'm sorry so proper thank you thank you ever so much to Nick for coming all this way and being here and it was such an interesting talk that was fantastic, it's really nice to see you all after five or so years to all our panel members thank you ever so much for coming along a nice and lively debate and asking questions of each other I think was great thanks ever so much as well Karen that was perfectly shared it was wonderful and the people I really shouldn't forget are standing at the back so we've got Sheila who from Yesi who has really made all of this possible thank you ever so much for the help that you put into organising this and Howard who's helped with the other AV people to make sure that we had a connection with Stockholm that was making me a little bit nervous I have to say so that's all worked really well and again big thank you to all of you for being here and coming and contributing so actively and maybe the next one might touch a little bit more on the existence of air pollution food security and food access and things like that would be nice ok thank you very much thank you