 There's a number of works so that was clear from the first session yesterday in terms of how it's currently envisioned and of course there's more ways that it can enter in the current current version. It's a matter of like entering it through productivity growth or potentially through labor force participation rates and labor force growth. And perhaps in some other ways too. But in thinking about the risks, and then in terms of how that where that information actually comes from a lot of that comes from these climate impacts analysis and we saw that in the second panel yesterday where the climate impact analyses are excuse me, I guess the third panel yesterday where the climate impact analyses are based on historical evidence and historical data looking at either granular from a bottom up approach or top down looking at GDP effects of various climate variables. All of this work has been important, but all of it clearly is just first steps in a much broader research program. The other the other thing we learned or recognized or saw is that both uncertainty really matters and also the sectoral and regional differentials. I mean we know very well that at a global scale climate really has going to have extremely different at different levels of, excuse me, at different levels of different regional and different levels of scale. So, and of course it's the case that many of the people who are going to be hurt the worst and many of the regions they're going to be hurt the worst in climate especially in the short term are ones that are ones that had really nothing to do with creating the emissions in the first place. So, that granularity, if we narrow it down to say an individual economy like the US economy which is the main focus here, that granularity. There's there's really interesting features about it some of it washes out. So, or some of it might not be big enough to rise to national levels or some of it might be handled by insurance markets. The aspects of that granularity potentially could be cascading and I think that that's one of the things that we might be talking about today, where, in principle, insurance should be able to handle things like wildfire risks in California and flooding in Florida, but if there's market inefficiencies or market frictions or regulations limiting prices, instead of just prices responding then you might actually just see insurers dropping out and insurers drop out then it goes up to goes to the state. And then you've got sort of this cascading problem of where these risks actually fall. Then they fall on the taxpayer rather than on the reinsurance industry and so how does that fit in through all the financial system and we're going to have some discussions of that today and I believe probably are probably our last panel, and then related impact of physical risks in our last panel. And then another theme that's really important is thinking about the uncertainty of all of this so there's many different ways to think about dealing with the uncertainty. Exactly what's the best way to deal with it is really difficult and it's problem specific. Certainly in terms of formulating policy from the outset thinking of optimal policy optimal policy has to be really thought of as a sequence of optimal policies where you're learning is you're doing and you're taking to account. All of the deep uncertainty that is associated with the climate problem and with how the economy is going to respond. And that twitches does a practical matter. How do we evaluate our really difficult policies that we are complex policies that we implement today. And our first panel is going to be looking at that, at least in at least a couple of the papers are going to be a case study of evaluating the inflation reduction act and, and I guess parts of it, looking at the ii j as well, and seeing how that fits in how those policy evaluations fit in both from a macro perspective, and from a from a carbon emissions perspective. So I think overall this is like a big tour of some of models that are used in the government models of modeling families that are used in in in in academics to think about formulation of policy and models that are used in in the private sector and in, and in other entities to actually guide the guide the construction and development of policy so. So with that, I think we're sort of covering the landscape but it's a really big landscape and it's a really complicated one and the thing that's super interesting about this is not just that is important, but actually that it's evolving in real time I think because we're really learning a lot and the research is really blossoming and hopefully we'll be able to see a little bit more of that today. So, Amy and we're just going to turn this over to you for the next panel. Thank you, Jim. And our first panel will be about some aspects of the energy transmission at transition. The title is projecting economic and financial impacts of a transition to a low greenhouse gas economy. So we have three speakers, three very exciting speakers, Emmanuel Campilio, John Larson and Neil Marotra, and the first two are remote so with that, why don't we start with Emmanuel. Hello everyone, can you hear me. Can you hear me. Yes, yes. Okay, lovely. So, thank you very much for having me. I'm really good at telling you that unfortunately I fell ill a couple of days ago so I'll do my best not to cough during my talk but I won't be the most brilliant speaker today. So I apologize in advance for that. We were asked to develop some considerations about the factors that would be good to incorporate into macro-cognitive models of low carbon transitions and there are several of course and some of them have been already discussed yesterday. Let me put forward for the sake of discussion. The next one, which is transition related expectations. So, transition related expectations by this term I mean everything that has to do with the low carbon transition so future policy implementation, the development of technologies, the possible stranding of physical and financial assets. These are key of course in defining the future transition pathways because they define our investment choices today, and also potential transition related disruptions is whole idea of climate means key moments or greens ones. They often have the roots in the misalignment of expectations compared to reality. So it's very important to have in models. They're also very hard to capture in their full complexity, let's say. Behavioral economics tells us that expectations are messy. They are heterogeneous volatile their subject to cognitive biases. They're forward looking, but we find that planning horizons they're very influenced by what happened just recently. And it's it's hard to capture them and that, you know, to a first approximation the usual way of doing of having them is either the traditional no classical way, which is usually based on rational homogenous forward looking expectations, or the complexity macro way, which allows for heterogeneity more easily, but also relies on purely adaptive expectations, most of the times. However, there are recent lines of work in economics, for instance, the whole diagnostic expectations, or the heterogeneous expectations kind of work that have potential to be incorporated and give new insights for transition So what I will try to do is to very briefly present you one application one attempt at including this heterogeneous expectations approach, looking at transition and policy uncertainty, together with a couple of co-authors. So the motivation is very simple. I don't think I have to explain it that much, but we had several examples of policy reversals in the climate policies fear in recent years. The Australian carbon tax is one example. The complicated relationship of the US with the Paris agreement is another prominent example. And many times, these policy reversals are motivated by concerns that either governments or society or both have around the costs of a low carbon transition. So this was triggered, this was what triggered the Gilles Jean movement, as we all know, but there are many several, several examples also in emerging economies linked to fossil fuel subsidy phase out proposals. All this uncertainty about the policymaker commitment plus a number of other factors create dispersion of expectations. We have actually very few data on this. I hope I will have the time to come back on this. So this is the only survey data that we're aware of concerning policy expectations, it comes from definitive. So what you see here, they asked in 2020 to a bunch of carbon market professionals, what do you expect the EU ETS carbon price to be in 2021, that's the blue line, 22, 23 and 2030. And so we see that there's clearly heterogeneity in expectations that this is also expanding, increasing in psychological time so the more we move in our planning horizon, the more the expectations are dispersed. But we also see that there's an increasing. There's a trend of increasing carbon prices on average. So we tried to capture some of these dimensions by developing a dynamic model, we have two technologies, I will not have the time in 10 minutes to go into any mathematical details of this. We have these two technologies, low carbon and high carbon and firms decide how much to invest in each of them, depending on what they expect the cost differential of the two technologies to be in the future. And these cost expectations are in turn affected by carbon price expectations. So we assume that firms observe the policymaker announcement. This G bar tau that would be the announced growth rate of future carbon price packs. Nowadays, of course, announcements come more in the form of net zero dates, but you can derive the implicit optimal carbon price from that, and we use that. And then the firms evaluate the credibility of the announcement looking at the past track track record of the policymaker in keeping its word. And so we have these two populations, believers and skeptics, and each firm can decide at each moment to switch from one belief system to the other. Policymaker can decide to default on their goals, if they perceive transition risks to be high. And so the actual tax implemented. Can be different from the announced one, depending on the government's commitment with this C parameter. So if a government is fully committed, it will just do what they promised they would do. If instead it's lower than one, then they will give some weight to transition costs concerns. And these two choices, both the degree of trust in the government and the investment choices, they are heterogeneous across firms. And we have these two key parameters this beta and this gamma which we frame as responsiveness parameters to signals, but ultimately they're really linked to the dispersion of these perceptions these expectations. And so we can move from a classical limit, so to speak, setting where both parameters are equal to infinity so expectations and beliefs are homogeneous to the other extreme where basically choices are made at random, when beta and gamma are equal to zero we we never get to this extreme. And then we derive some first analytical results. So let me just guide you through the main points here. What you see here is a study state analysis and above you see the no classical limit so with the homogeneous beliefs and expectations and below you have the heterogeneity case. The takeaway messages are that even if you have homogeneous expectations and beliefs you can still have certain situations in which multiple equilibria arise and this is the case of ambitious but weekly committed policymaker. So this you need to think of a policymaker that tries to trick so to speak firms into decarbonizing by loudly announcing that the decarbonization is coming. And we show that to some extent, under certain conditions, this actually works, even if there is no commitment, but if it doesn't any backfires and the transition completely fails. If we allow for heterogeneity instead we get this situation in terms of study states, and you see that the conditions of existence of of the study states change, but also the nature of the study states change. If we allow for heterogeneous expectations. So in some cases they change for good, meaning, if you look at this area. For instance, we have partial decarbonization with an unambitious but committed policymaker, and this is where we had no transition at all in the classical limit case. In some cases they change for for worse. And I'm referring to this area where we move away from multiple equilibria to go back to a unique study state which is however high carb if the policymaker is particularly ambitious but weekly committed. We have a couple minutes left. Okay, thank you. The numerical results are similar. Again, let me just tell you that what the main points are. What we show is that we calibrate everything to the European Union economy. And even without with a fully committed policymaker we show that heterogeneity affects the speed of transition. So let's move to this chart just to the lack of time but this is essentially what what it says, and this is relevant of course because you reach fully carbonization eventually. But if it's, if the transition is lower this means more emissions more cumulative emissions, more temperature increase more damages. The chart on the right essentially repeats what I already said in the analytical case that is you have certain conditions that trigger a vicious circle of credibility loss. High carbon investments, high perceptions of transition costs and weaker policies, eventually leading to to a failure of the transition. So this was just an example. I just wanted to throw it out, throw it out there, but really there's, I think, a lot more work to do in this area. Some of it has to do with capturing the expectations which we don't fully understand. We don't have enough data we don't really know how they move properly. And there's a number of methods which are quite complimentary to complimentary to one another, like financial markets, like chronometrics or surveys, natural language processing or experiments that could allow us to calibrate the model and also develop more sophisticated models. Speaking of, this was just, you know, one first attempt. One thing that might be relevant for instance in the US case is to have electoral cycles. And finally, and I'll close here. The most important question I think is how to how to manage all these expectations so probably the ideal setting is one in which expectations are all aligned. They're as homogeneous as possible and aligned to the policymaker, which hopefully is aligned in turn to climate science and it's not easy to understand what is the best policy institutional framework to achieve this. Central banks might have something to do with this given that they have to know how of expectations management, but of course this opens the whole discussion on whether they should be doing something about it. Of course they are doing something about it. This is just to say this chart is just to say that they are heavily communicating on climate change. So there's something there to study as well. Let me conclude and happy to come back to any of these points in the Q&A. Thank you. Thank you very much. Our next speaker is John Larson from the rhodium group. Thanks. Can everybody hear me. Yes. Excellent. Buddy, thanks for having me. It's really an honor to be a part of this panel in this workshop today. I'm John Larson. I'm a partner at the rhodium group or an independent research firm that does a lot of work both on the economic impacts of climate change as well as the policy impacts of climate change trends and I'm going to talk about the intersection between macroeconomic assumptions and US greenhouse gas emissions a day with using the inflation reduction act is one policy example to explore as part of the interaction there. Next slide please. The rhodium groups, all of our US focused energy system modeling starts with our version of the national energy modeling system which is a platform created by the energy information administration. We have modified and operate ourselves and maintain our inversion of the model and make a number of additional augmentations and expansions to represent emerging clean technologies and their role in the energy system as well as updated assumptions around markets and policy and macroeconomic outlook. I will admit, you know, the macroeconomic module and NEMS is certainly not as fiscated as some other state of the art components out there but I think at least for this discussion it's still a useful tool for showing kind of rough directional trends on macroeconomics and you know, but also just to know that this is our version of the model versus EIAs. Next slide please. And looking at our current policy forecast for 2022 that we put out last summer before the inflation reduction act came out, we use, and we'll talk more about this in a sec but we use a variety of different assumptions to construct different emission pathway scenarios to reflect the fact that the future is one of the obvious areas of uncertainty is the rate of macroeconomic growth in the United States over time, especially if you're looking out to say 2035, 2040 those effects can compound and be material over time. So just picking out two scenarios from our work, we had a central emissions pathway and a high emissions pathway that we looked at as part of our range in 2022, of which the central emissions pathway assumes 1.9% annual GDP growth rate from 2023 to 2035 it changes after that but just for sake of keeping our timeframes consistent here. Meanwhile, we had a high macroeconomic growth rate that was at 2.3%. The 1.9 is roughly consistent with what the Congressional Budget Office was projecting at the time that we were setting up these scenarios. The 2.3 is certainly more aggressive and reflected EIAs central case for economic growth. And long story short, when you assume a faster growing economy, you will have higher emissions all else equal. On the right hand side is the cumulative net greenhouse gas emissions under both scenarios, holding other assumptions constant and the key point here is over the course of the timeframe between 2023 and 2035 we see almost two and a half billion tons more greenhouse gas emissions go into the atmosphere, simply due to the drivers of macroeconomic assumptions. Next slide please. Now, why do we see that the there's a lot of answers but the primary answer is energy demand. When you assume faster growing economy your industrial sector is producing more things it takes energy to produce things. When you assume a faster growing economy, you have more people commuting going to work, and more freight and goods and air travel, which increases energy demand of the transportation sector you also have more income and with consumers and they are spending that money on a lot of things they're spending energy services. And so you can see here between the two scenarios. You know, in some cases, a couple of quads more energy demand when you're looking across different sectors of the economy, and I should say electric power energy is incorporated into these charts here so that that's a key part of this. And so, you know, the mission, all you know, like I said there's a few other things here I 10 minutes so I'm focusing on the key drivers here but energy demand is goes up with higher macro assumptions and that is the primary driver of emissions increases under higher macro assumptions next slide please. When we looked at the impact of the inflation reduction act we started with a range of assumptions and different scenarios while holding policy constant so either with current policy before the IRA or after the IRA is implemented we hold those policy specs constant and then we had three main drivers of emissions trends that we considered that translate into a low central and high emission scenarios. And one is fossil fuel prices, the higher the fossil fuel prices assumed a lower emissions although it's equal clean technology costs, the lower those costs are the lower emissions will be all out to equal and then again, I already talked about economic growth and you can see how we took over three different emissions across three different scenarios here to paint a range of potential outcomes. And I just say there's links in the slide deck here for reference if anybody wants to dive deeper into all the all to unpack everything under the hood we have documentation there. So keep this in mind when you we go to the next slide. Our total net greenhouse gas emissions the great line is history, the blue and orange ranges are the blue range reflects emissions outcomes without the IRA and then the orange range reflects outcomes with the IRA in place. So very short the IRA does a big makes a big lift, it gets as much as a 10 percentage point relative to 2005 emissions reduction in us emissions compared to no IRA, and getting emissions as low as 42% below 2005 levels by 2030, certainly nowhere the target the US current administration is set for the country of 50 to 52%, but with it in one piece of legislation as a major step forward. But what's what's actually happening there with regard to the emissions profile given what I was just saying about macroeconomic assumptions well the first thing to notice the higher ends of the range reflect higher economic growth assumptions in the first place so that means, you know, all else equal net emissions will be higher with the IRA than with higher macroeconomic assumptions than with moderate macroeconomic growth assumptions, but even then holding macro constant the high end of the range goes from with was 24% below 2005 levels before the IRA and is 32% below with so the IRA still does a lot, even with the high economic growth assumption in place. Next slide please. And it's important to understand why the emission reductions happen the way they do given what we were talking about with macroeconomic growth is worth looking at where the emission reductions happen. This is just from our central scenario. And this is net GHS across, across sectors, and by and large the, the vast majority of emissions reductions across all of our scenarios and in this case in particular, occur in the electric power sector, this makes sense those are where on some of the biggest most lucrative tax credits are it's also one of the most responsive sectors to climate policy. And we do see some emission reductions elsewhere across the energy system, including an industry and transportation but much, much smaller than, in fact, in order to do smaller than what we see in the electric power sector. So with that in mind next slide. And the last substantive slide. The, what you're seeing here, and what the IRA effectively does to get the major drop in emissions relative to know IRA is a shift in investment from the traditional kind of pathway balancing capital and variable costs in the electric power sector, shifting it somewhat over to higher capital and lower fuel dependent so the left hand side here is cumulative electric power resource costs from 2023 to 2035. And the important thing to note here is that system investments go up. This is this is absent subsidies so subsidies are not counted in these values. There's a total investment going into the system in response to keep the lights on but also and serve energy demand. And with the IRA you can see total investment goes up. O&M goes up as well, but you do see that there's nearly $100 billion a little more than that in fuel savings because you are now displacing fossil fuel consumption with new wind and solar generation. And what that what what we find interesting about this is that the rough total investment is about the same between the two policies the way that without IRA in the scenario. Which means what you're really doing here is just directing that money into different types of generation and effectively what you're doing on the right hand side is just dropping the carbon intensity of electric power in the United States substantially compared to a no IRA scenario so you know trends without the IRA did see the emissions intensity of electric power going down to about 2.5, 0.25 tons per megawatt hour. But we see with the IRA that getting as low, you know, getting cut by more than half down to as low as 0.1 tons per megawatt hour by the early 2030s. And we have a rebound I'm sure everybody's wondering what that means up to 0.15 tons per megawatt hour and that is the expiration of the existing nuclear retention tax credit in the IRA and those nuclear plants are no longer economic and fossil fills in behind but long story short though the key takeaways carbon intensity is going down due to shifts in investment across the electric power sector by design from the subsidies. And, and that is effectively getting you the emission reductions you need without any real material shift macroeconomically speaking and why is that last point grounded well it's because if you look at the total investment we're talking about between $3 and 2035 talking about a little less than $2 trillion either way there's no net serial it's substantial net change in electric power investment, and in a $25 trillion economy like today and growing over time that is still relatively a component of the overall economic picture especially when you're taking, you know on average we're talking about like roughly 150 billion dollars a year in investment. And that's just not you know, nothing compared to a $25 trillion economy. So, you know what we, the key takeaway here next slide is that macroeconomic conditions and assumptions around future growth are actually the bigger deal for directional emissions in US energy modeling at least in the time frame we're talking about through the mid 2030s. And that's why the rest of this workshop is so important to get that right, especially as the climate is changing and new risks and disruptions are on the horizon. And at the same time policies do matter and can tackle climate change. But what they're really doing instead of shifting the overall economy is shifting the carbon intensity of energy production and consumption. And that, while important and probably has some sectoral implications from a macroeconomic perspective are not complete shifts to the broader macro picture. And, and that's because the clean scale of clean energy investment, while large is still relatively small compared to the US economy. And, and, and maybe the last point I'll make is that given all this one takeaway for us is that there's plenty of room to meet the need of more to carbonization in the coming decades ahead in the United States. And with regard to investment and other policy action that can shift incentives and behavior over time. And we should not be all that concerned about the macroeconomic implications, because the macroeconomic application trajectory itself is actually more fundamental and more important to the overall economic picture as opposed to clean energy investment. So with that I'll stop. And again, thank you for having me today and really appreciate your time. Thank you. Our last speaker is Neil Marotra from Minneapolis Fed. Thanks so much for having me on the on the panel it's very nice to be here. So, like john I'm going to focus my remarks on the inflation reduction act and some recent work that I've done with Catherine Wolfram and john vice line trying to understand the economic implications of this largest piece of legislation to tackle climate change. So, just, you know, brief review IRA works to subsidize clean energy investment, it has a set of investment and production tax credits for clean power generation. Importantly, for our purposes, those subsidies are uncapped, and they expire only after emissions targets are reached. So when assessing the sort of fiscal and macroeconomic implications of IRA, what really matters is how much take up there will be of of these tax credits. There's a substantial piece of legislation that offers tax credits for households to purchase electric vehicles and to purchase lower carbon options for home heating. And then there's a set of incentives around more nascent technologies like green hydrogen and carbon capture and sequestration. So in my brief remarks today I want to think about what are the implications of IRA for energy markets and what are the macroeconomic implications of these of these provisions. So on the energy market side, our work suggests that that IRA will have a significant effect on power investment a 50% increase in power investment and a sizable reduction in CO2 emissions. I think our numbers are close to what what john is finding in his modeling. We see that this increased power generation does raise the possibility of very low or even negative wholesale electricity prices. And I think that will be relevant for thinking about the supply side effects of of this policy over over time and the policy of subsidizing the subsidizing power generation, and then we see a significantly higher fiscal imprint of the legislation over the next 10 years of around $900 billion relative to what CBO and JCT had originally scored. So over over the next decade. The way we the way we try and look at energy investment is through the Regen model so one of the co authors is at the Edison power Research Institute and the Edison power Research Institute has a has an energy industry equilibrium model called Regen. And so what we do is we run these tax credits through the Regen model to understand what will be the impact of of this policy on power investment. We see is a large increase in clean energy investment a 50% increase relative to baseline. So if you look at the top bar the top bar is renewable power investment over over the past decade it was around 20 gigawatts per year and new capacity additions. Even before IRA, because of cost reductions in in wind and solar, we saw, we were projecting this model projects of a substantial increase in investment in renewable power IRA supercharges that so there's another 50% increase on top of that. So relative to the prior decade we're about, we're projecting about a doubling of investment in solar and wind and battery. It's important to note that our projection is actually on the conservative side. So there are a set of projections out there including sort of some influential analysis from a group at Princeton that argues that the, the IRA will really supercharge investment in power generation. They see that's the bottom bar, they see a five x increase relative to the prior decade in in clean clean power generation. So, this, this investment in clean power generation drives lower carbon emissions, we see a seven percentage point reduction from IRA in US greenhouse gas emissions relative to baseline this is again similar to what the last presentation showed that moves us literally closer to the Biden administration's target but doesn't get us all of the way there. And these, these emissions reductions continue into the into the next decade because, because these tax credits don't necessarily expire at the, at the end of 2030. So, our modeling suggests that IRA will have significant impacts on on electricity prices and it raises the possibility that wholesale electricity prices could be driven to very low levels, or possibly even negative levels. So what this graph is showing is, you know, there are certain times of the day when there's high demand for electricity and certain times of day when there's low demand for electricity. One of the issues with renewable power is that it's intermittent so you solar and wind don't necessarily produce at the times when there's, there's peak demand. So if you sort demand from its highest hour highest peak hours to its lowest peak hours at the low peak hours you're going to have times where there's relatively more supply of electricity relative to demand driving down the price. And because IRA is subsidizing renewable power investment, we see the possibility in certain markets that power prices could could could fall, could could fall very low to very low levels or even negative levels. Why could you have a negative wholesale price well. Why would somebody continue to produce at a negative wholesale price because they can still collect the production tax credit so if you're, if you're a solar producer or wind producer collecting the production tax credit, you still have an incentive to produce even though the wholesale price of electricity is negative. To respect the wholesale price of electricity to actually go negative. I think what what this will, what these subsidies will do is that they will lead to changes in electricity demand that will, that will try to take advantage of these low prices and that's an important thing to that's an important supply side effect that we need to think through. So, the upshot of all this investment in power generation is that the take up from these tax credits is expected to be larger than what was that what was originally scored. So, the left hand side bar shows the CBO JCT score of the climate provisions in the inflation reduction act. They score those provisions is costing roughly $400 billion over over the next decade. The fiscal estimate has a cost closer to $900 billion. It's important to emphasize here that the abatement cost so the, the, the reduction in CO2 emissions for for for a unit of fiscal cost is still below the social cost of carbon so in some sense that's the relevant metric for assessing, assessing whether it's worthwhile, but there is a bigger fiscal imprint that's coming from from from this legislation. There is uncertainty around this so we show two scenarios where in the lower scenario, which comes closer to the CBO JCT score that lower scenario involves macroeconomic conditions that are relatively unfavorable to clean investment and, and a reduction in the rate of, of, and relatively pessimistic assumptions about future reductions in cost. The high cost scenario shows greater take up of the various bonuses that are available in the inflation and more incentives to more uptake of, of carbon capture and, and green hydrogen incentives then in the baseline case. So, turning to the macroeconomic side what are the macroeconomic implications of IRA. We make the case in, in our work that on the longer in the long run IRA is delivering supply side benefits to the economy it's a supply side policy that increases output wages and, and productivity primarily through a lower price of electricity and an important input into into production, and so that that's the supply side benefit over over the long run. But in the short run IRA is going to stimulate considerable amount of investment demand and that investment demand will likely lead to a higher path of interest rates than might otherwise be expected. So relatively, you know we agree with the last presentation that these the boost in investment demand is relatively modest relative to the size of the economy. So we're not talking about large, large effects, but I think it's important to emphasize that the macroeconomic investment has a strong effect on on IR on on the on the incentives to invest in clean power. So, one, one, one point to make is that higher interest rates negatively disproportionately impact clean energy investment because they're relatively capital intensive. And what that panel is showing here is the sensitivity of the levelized cost of electricity with respect to interest rates. What you can see is that the flattest line is natural gas so natural gas where it has relatively more. The fuel cost is relatively more important to the capital cost. It's relatively less sensitive to interest rates, but other renewable technologies because they have the capital cost up front are more sensitive to interest rates. So if we think that the macroeconomic environment is shifting to one of higher interest rates that could negatively impact energy generation labor cost is also another factor where higher labor costs could could disproportionately impact clean energy investment. So, overall, the, the, we do see a big boost in energy investment but those impacts in terms of on the demand side are relatively modest. The reason they're modest is because electric power structures and transmission are a small part of the economy relative to GDP so even if you have a doubling of the amount of investment in in the power sector, that's still relatively relative to the economy. So when you stick it in a model like Ferbis which is the Federal Reserve US model, you have small increases in output and employment and core inflation initially but those are but those are but those are quite small. I think there's important limitations to this model, this modeling that may understate the macroeconomic effects. It does not include upstream investment effects, and it doesn't include the combined effects of the infrastructure act, and the chips of the infrastructure act which are all acting on sort of similar parts of of the economy that are all acting on manufacturing and on on construction. So, I'm out of time, I'm just going to leave up our takeaways, but look forward to talking more about this. Thanks. Thank you very much. We now have some time for questions and we're going to extend the Q&A until 1045 since we started a little bit late. Okay, let me, let me start with Jim stock. Okay, thanks. Thanks very much. This is just great. This is a really interesting panel. Thank you so much for your work. I have a question for john and Neil. The theme of this conference or this workshop is is climate issues in macro and of course one of those issues is exactly what you're grappling with which is take a large policy so the IRA is at least, you know, in principle potentially a large policy that might have macro implications and you need to investigate that you need to see in both of you conclude that maybe that's not that big of macro but you wouldn't know that a priority actually have to do the work for something of this magnitude. There are quite different approaches so one of them is running at using the using the macro module inside nems, which has well known, very well known limitations, I mean it's there but it has, it has limitations. Another is this sequential approach, I guess Neil that you took where you run this quite sophisticated every model and then you run a pretty quite sophisticated macro model you, which is furthest but these things are like just like separate blocks that are hanging out there. If you sort of said okay scale it back you know pretend that we're in 2019 and you had the secret whisperer telling you that you'd be asked to evaluate the IRA in 2023 what sort of toolkit would you like to have have seen in terms of the macro just focusing on the macro piece of it not on not on the energy system piece or anything but for focusing on the macro piece of it. John, would you like to go first or you go ahead Neil and then I'll go. Well, I mean, I've thought a little bit about this I think that you're exactly right in describing how we do it. And the sequential approach has has has definite limitations. I think that the spill overs that we are that we're missing through this approach is through commodity prices. And so the region model is clearly having big effects seeing big effects on electricity prices, and potentially it's taking as given fossil fuel prices natural gas and oil prices, but all of these things would be, you know, if you if you had the ideal that that that had both of these components together I think you'd have substantial spill overs there. And so when I'm when I'm sort of thinking you know backing up for models, if I'm thinking about how best to compare the expected macro economic impact of IRA, I think the relevant comparison is something like the shale revolution, and the shale revolution. Arguably like it had a really big effect on mining investment and and and mining employment. That didn't really show up in in the macro data in the sense that that was this was a time of slow recovery coming out of the coming out of the great recession, but it did show up in commodity prices, and those commodity prices arguably had, you know, some some important implications for for investment and for for GDP growth. So, I think that that if I had a model that could capture what the shale revolution was doing fairly well. And I don't think Ferbys did that but if you had a model that was doing was capturing that pretty well, then it's probably doing well on on thinking about the the near term impacts of IRA. Yeah, and just to add to that I mean first of all, I completely agree with Neil about the interactive effects with commodities. Now it's one that's actually one area where Nancy's actually fairly capable oil is an input but everything else all other fossil prices shake out in equilibrium with energy demand and supply costs. And we actually see gas prices go down, because the IRA, for example, because you now are displacing a large amount of gas demand and that's putting downward pressure on prices. But I think on the macro side, having a model with better granularity of the macroeconomic space generally like Ferbys would be one example, compared to what NEMS has, and better feedback between the energy system and that macroeconomic outlook is iterating over time again which is something NEMS does, but doesn't completely capture that macroeconomic feedback well at all. I'll be one of the first people to say would be quite quite useful because, you know, even if these effects are relatively small as we've talked about they they are still matter, and they may have spillovers on commodities and maybe even over the long run they may have device prices, you know, EV prices other things that could be quite important to the long term trajectory and we're not. No one's capturing that right now. And one more quick thing is that, you know, the Ferbys model doesn't do a great job of capturing supply constraints bottlenecks. We've seen, you know, over the last couple of years, how relevant those are to macroeconomic outcomes, especially inflation, and, and so ideally you would want them. IRA is taking place in a macroeconomic environment that's quite different than the shale revolution was taking place in a decade ago, and that's where I, my inclination would be that it's going to have more of a it's going to come up against some of those supply constraints more and, and that could have more important impacts. Thanks I probably should start over this but let me just, I think there's one clarification of something that Neil said you mentioned that, comparing the costs of the IRA provisions to the extent that they've been calculated on a per ton basis, compared favorably to the carbon I think you might have mentioned per fiscal cost, fiscal cost per ton but I think it's really system cost that you're that you that's what you calculate in the paper in your papers or just want to clarify that. Thank you, George Kovortov. Very interesting presentation this question is for this question is for the second two presenters. And it's sort of about the energy system mod models assumptions and that it's my understanding that there's basically kind of a relatively stable final energy structure, implementing some policies like carbon tax or IRA the subsidy. And so it's sort of implying a relatively smooth substitutability between the between basically electricity and the fossil fuels. I wonder how robust that is to some assumptions so for example, I guess, I guess the one hot topic is, is battery minerals, I think, in the energy systems models they have kind of a very detailed modeling of, like coal resources or oil resources, but I think I could be wrong but it seems like the battery minerals are a bit like more of a decent area. The second friction I was thinking of was the kind of human capital reallocation between brown industries and green industries like the skill reallocation. That probably wouldn't be an energy systems model questions more of a macro labor type thing but I wonder if that's maybe implicitly taken into account by like the cost of, I don't know, you know making a new power plant or factory or something. Yeah, I can speak to the first card, maybe just quickly. You had it mostly right with its service demand that's roughly constant different types of energy services so for example, mobility is a relative constant and the demand for travel which is different than the energy required to travel. And to point that out as an important distinction is over time consumers are making choices about what car to travel in the IRA influences those choices, the energy demand and types of energy use change, but the VMT the vehicle's travel largely doesn't right like so that that that's that's one way that that nuance gets captured that you are getting at. And, and there's like you can make the same thing about like he pumps and houses and other stuff too right so. So there's a stock turnover assumptions that drive that as well like each year a certain amount of consumers are buying new cars and then they make choices right and so that that's that part. And then the other thing I'll say is when on on minerals or other constraints we just at the moment would say okay well assume assume higher tech costs, generally, as a reflection of that and then how does that change people's choices. So that's kind of you know it's it's a little bit blunt, but that's the most kind of straightforward way to account for a lot of potential constraints and deployment or or you know like both labor and supply so that that's kind of how we approach it but that's you know ideally would be a little more in depth on that. On the, on the labor reallocation question I think it's a it's a very, a very interesting one. Again, in in aggregate terms the amount of labor used in power generation investment and transmission is not is not super large. It's relative to the construction industry which is small relative to the aggregate labor market. And so, so when you think about sort of scaling these things that you know you can, you think that it, it's, it's, it's fairly hard to get more solutions or get more more specialized construction workers, but you know during this during the shale revolution they were able to do that fairly fairly quickly. And, and so I sort of come out that that we have a tight labor market today so that's probably going to make it a little bit harder to reallocate labor, but, but we, you know, we also have for example, labor force participation among sort of prime age males has, you know is historically, you know it's been declining for for for a long time, perhaps in the communities where this investment takes place, there will be sort of labor supply at the at the margins to come in and quickly be trained up to to do this. Some of this work is specialized but a lot of it is not is not super super specialized. Heather Boucher. Thank you. And this is great I feel like I've been reading two years for this particular session so I think this is, it's very exciting and we know my colleagues have been so excited about this panel but certainly these conversations. And so I have a couple of both comments and in a quest, a couple of questions so on the one of the things just about electricians is that they often travel, which we, which I think people who think about it like macro may not know that they often travel around the country to do their work and so it's one of the things that in the back of our mind we should be aware of. And the other is that we are seeing this real big ramp up in con in construction but that's going to be temporary, and that is going to be a Mac like, you know, we're ramping this up and then it's going to probably fall down over time and we might want to be thinking about it. But actually wanted to come back to where this conversation started with the. I don't have all the names in front of me with the manuals first point about how the common reason for not doing or reversing these policies is concerns about unemployment and stranded assets. And yet, and these conversations, if the public could see them that oh there's no macro you know the macro economic effects seem somewhat muted or in a relatively small. One interpretation of that could be well the public shouldn't be worried about that, but underneath all of this is the challenges with supply site shocks critical minerals that labor reallocation which is a fancy word for geographic differences and I think one of the things that I'm just struggling with in the, in the connection between the model and the modeling work and the policy is how we open up that space for policymakers to see that to see some of those impacts. Through the macro through the through the macro conversations and tying it back to what you said Neil that the, what we do on the macro side may have big implications for how effective these policies are, if interest rates have figure more prominently in some of these big capital investment sectors then, you know, is the other monetary authorities working at cross purposes. So yes my question to all three of you is, as you're looking for to go to Jim's question, what do we want, like, what kinds of fantasies models can kind of imagine that we want out of these tools, we can help us connect better those dots between these, what are essentially micro spatial challenges, and the macro the macro way that we have these national conversations it feels like the models need to give us a little bit something and I'm struggling with how we, as the modeling community should be thinking about that, just again to connect back to Emmanuel's very first side and that very first question like why the public is so anxious about this transition. That was a lot but sorry. Emmanuel, do you want to maybe go first. Yeah, thank you for this. So let me just react quickly because I was also interested in joining Neil's presentations about this. So, it's true that several pieces of research including the ones that we've just seen seem to suggest that the macro economic implications of climate policies might be low or anyway not particularly dramatic. It struck me however how both modeling exercises showed that IRA is missing the target. Right. So, none of them are actually delivering what the administration announced that they would do. So, one question that I had for them was actually whether their model could also be used in order to derive let's say the optimal IRA let's say enhanced IRA that manages to actually get to the target and what would the macro economic implications of this enhanced IRA would be. So, it's very difficult to capture all of these implications, but the policies that we have might not be the ones that we actually need. I mean, so on the on Emmanuel's question about the Biden administration target and not make IRA not being sufficient on its own, the, the, the regulations on fossil fuel generation that were recently announced. Those together with IRA do get you I think substantially closer and perhaps, perhaps all the way. And so that was in our in our higher fiscal scenario, we were thinking about a scenario in which regulatory actions would devise more carbon capture and sequestration. And so those credits get used more but you get more more more emissions reduction. So I think the administration is thinking about, and Heather and others can talk about this is thinking about other levers to pull to get to 50%. And there is a range in all of our modeling about how much emissions reduction you get. On Heather's question about reallocation and whether we're really picking up sort of what's going underneath the hood and what might lead to more, what sort of maybe voters into it more about like the costs of transition. There's always, there's always loss aversion so people who are, who are working in a particular industry that is, that is, that is, that is losing out are going to feel it more acutely than those who may be joining a new industry. And so that's obviously a political issue that's obviously. But but does that translate on the macro side I don't, I'm, I'm not yet convinced of that. The, the question about does this look much more disruptive to the labor market, akin to the China shock like we were talking about, I think is a is a relevant is a relevant question. And our models don't do a good job of capturing those kind of sort of broad sectoral disruptions. So, so that's perhaps something where we should be focusing more of our attention on are we underestimating the, the potential scope of, of, of disruption. And just, just to add a couple things to answer Emmanuel's question we did, we wrote in group looked at kind of a pathway to the administration's target, which largely just layered on regulations as Neil described to alongside the IRA, and importantly not an optimal policy just like what you would use in the existing toolbox to that that's a key hobby out there. And, you know, what we saw is just like directly similar results just a bit more abatement in all the sectors and not not some major shift and wave and investment. So even, you know, I don't have the investment numbers off the top of my head but assuming even if it was double or triple what we found in the IRA that would still be a relatively small amount of the economy. And so that's, that's worth thinking about. And then Heather on your question, I mean two things one. I agree with Neil I don't think all the tools we have get at all the questions. And I think that is an important area for for more research and improvement, but also like it's it's, we say there's a lot in in house erode and it's like it's not just the tools is how you use them. And I think folks like ourselves and everybody in practitioners in this workshop should be thinking about well what are what are the right indicators to be looking at in these analyses with these tools to speak to these concerns which is maybe different than what we would think of as energy system experts, or like experts, because that's ultimately like when you're speaking to voters right or consumers or you know one one interesting interaction we have a lot is with the labor community, and they are much more interested in saving existing assets and retrofitting them for decarbonization instead of building something new because they know they have labor union members in those existing assets and they don't know what the new thing is going to look like. There are other relevant and material factors in the broader policy debate that are not well captured right in any kind of modeling framework and so I could go on there's like all sorts of other anecdotes but like but those are those are things I think everybody should be thinking about. We're almost out of time so let me take two more questions just in sequence and then we can be here one more time from everyone. So Tim Lenton and then Wendy Adelberg. My question in essence is to what degree other models are already considering reinforcing feedbacks between sectors or feedbacks in general between sectors but if they're reinforcing the potential for that accelerate transition further. So just to elaborate briefly, you know Neil remarked on the extraordinary cheap electricity at night as a UK consumer I can tell you I've got 100% renewable tariff with a peanuts price of electricity at night which is a delight. As I've just got an electric car which I can happily charge for almost nothing and if I could be bothered I'd buy a battery stick in the garage charge it up at night and run the house off the battery all day. And if I was a utility company fretting about not being able to drug the electricity in the night I'd team up with a green hydrogen manufacturer, make a little green hydrogen and then sell that at its market price. And my point is, there's reinforcement between the sectors right we know that ultimately, if you want to get 100% renewable electricity penetration, you need cheap forms of storage in the grid but we also know that the expanding battery and EV market is bringing the price down of batteries all the time and no doubt the same thing will happen in the green hydrogen economy they'll be economies of scale. So, how basically how much how much of feedbacks between sectors being considered especially the reinforcing ones. Wendy Adelberg. I want to offer some downside risk to Neil's excellent analysis. But it. So perhaps it's not surprising that the IRA is climate provisions have positive effects because they're all subsidies. But if, if the IRA had decided that it wanted to subsidize the fossil fuel industry to lower the cost of energy, just from using more fossil fuels that would also have positive economic effects. And so I worry about the negative economic effects, even if we were to finance the provisions in the IRA using lump sum taxes. I think it was totally waived away any of the negative incentive effects from those lump sum taxes. I think you still have to worry about whether or not we have a higher rate of return, or, you know, a bigger effect on GDP from all of the investment that is being created by the IRA relative to the investment that would have been done with those taxes. You know, less than one minute from from each of you on whichever question you want to answer. Manuel, do you want to go first. Well, I mean for the first one. I don't think maybe I'm on the right person to answer this but I certainly think that a promising avenue of research lies in this production networks approach that is able to capture sectoral interlinkages, both domestically and internationally and it's actually very difficult to make them work because you need a lot of granularity within sectors. Not often you have the data. Elasticities are not easily estimated, but if one could, then this could be answering some of your questions, Tim. I think. Yeah, John. So on on Wendy's question I completely take your point and we are very clear in our Brookings paper that optimal climate policy typically favors a carbon tax which has the opposite sort of effects then then the subsidies based approach. And, and, you know, the subsidies approach is is predicated on, you know, important learning by doing effects, and, and I think that that's, that's that that that's worth. You know, thinking about and interrogating and figuring out how how how how how large those effects are. I would say that on this question of capital being other capital being crowded out. That's true in the short run is not true in the long run in the long run, the, the model will also say that that capital is actually crowded in because you know if non energy capital is a compliment to to energy capital then then not even a compliment if it's just Cobb Douglas it'll you have the marginal product of non energy capital goes up. So, this I think comes back also to Tim's point about cheap energy, you know people will find productive ways, I think, of using cheap cheap cheap energy, and I as an electric vehicle owner also was surprised to learn that if I charge at night I would, you know, a price that's one fifth of the, the, the typical price and that's because they're trying to find ways to offload, you know, a renewable production at at night, and this may become, you know, a way in which, you know, utilities start to manage the, the, the intermittency that they're, they're actually going out there and supporting renewables and that's not something that that supporting electrification and EVs because, because they have this, this surplus of energy. And that's not something that our models are capturing. And quickly on the feedback point we do capture the feedback between clean and fossil but not a lot between clean and then perpetuating clean in large part because the learning by doing effects are hard to capture in in the in dodgingly in the modeling framework so, like I said earlier we see gas prices go down that actually has a negative headwind on additional clean energy investment, because you're the price you have to be is now lower. But, but overall the subsidies are sufficient enough to ship things. And I wish we could have real time rates represented in our energy modeling framework I think it would be an interesting outcome, until it's like a common thing in the US energy system I think it will be unlikely that it's built into the model. I, I, as an EV owner I wish I had the tariffs that other people in the conversation had and then the other thing just very quickly on investment. I mean we see more of a shift in investment not a crowding out by you know it's so it's, we're building less gas plants in the model and building more wind and solar farms right and, and same thing with electric vehicles it's like we're selling pretty much same number elect of cars you're just selling different cars. And I think there's an important just like thing to keep in mind there around how you know the net crowding out is relatively small compared to the gross right and, and so that's just another thing to think about as far as all this transition goes. Thank you, Bridget, let me turn it up turn it over to you. Great, thank you everyone for the interesting panel. We were going to break for the next, what is it about 10 minutes before moving into the next panel but we are opening the slido again. So for those that want to add more of their thoughts insights ideas, you can use the QR code on the screen, or we'll share the link in the chat in a second. But I think I saw at least one question in there for a speaker so I do encourage speakers to go on and you can reply to the questions that are in there. Otherwise, yeah we will reconvene at 11am Eastern.