Social Cost of Carbon and IAMs

Manage per ton of carbon

Resources

Slides 16 - Social Cost of Carbon and IAMs

Content (Day 1)

Introduction and Opening Remarks

The lecture begins with a welcome to students and an introduction of a special guest speaker, Christina Lundgren from the Institute on the Environment. The instructor notes that while some people are still arriving, it is time to get started with the day’s presentation. Christina is invited to speak about happenings at the Institute on the Environment and will take the first few minutes of the class.

Guest Speaker: Christina Lundgren and the Institute on the Environment

Introduction and Personal Background

Christina Lundgren begins by introducing herself and sharing that she uses she and her pronouns. She works at the Institute on the Environment, referred to as INE, and mentions that she was able to travel there via the Gopher Way despite the cold weather outside. She notes that the Institute is just a short walk up the stairs and over from the classroom. She asks the students if any of them have visited INE before, and upon receiving no clear affirmation, she decides to describe the space.

The Physical Space of INE

Christina describes INE as having a beautiful atrium space that offers great natural lighting and is an excellent place to study. She explains that this atrium space used to be an old facility for showing off livestock but is now an indoor part of the building. This historical transition from agricultural exhibition space to modern educational environment serves as a backdrop for understanding the institute’s mission.

The Mission of the Institute on the Environment

Christina advances to her next slide and explains that the Institute on the Environment focuses on understanding how people relate to and are included in the environment. She notes that the students in this economics class are already engaging with this concept by studying the economics aspect of environmental issues. She then explains that the main reason for her visit is that the course instructor sent out information about the Sustainability Studies minor, and the current class counts as an elective toward that minor.

The Sustainability Studies Minor Program

Christina emphasizes that the Sustainability Studies minor is super flexible and accessible. Even if students are taking another course that is environmentally related but is not technically listed as an elective, it is really easy to petition for it to count toward the minor. She highlights that the minor is particularly great for students interested in the people aspect of the environment. Taking the minor helps students meet people from across the university in different disciplines and hear how other people view sustainability from their unique perspectives.

Organizational Structure and Career Opportunities at INE

Christina explains how the Institute on the Environment conducts its work through several different teams and positions. The Institute has an education team, which is the team that Christina works on, focused on activating leaders. They also have a research team, and Christina notes that if anyone in the room is interested in research, there are often student positions that are posted on a regular basis. The Institute also works on communications and organizing, and these are student staff positions as well. Finally, there is a dedicated communications team.

Christina acknowledges that all the students are probably already interested in the environment, but if they are interested in additional classes, the Institute compiles a list of these courses. She notes that the slide deck she is using will be shared with the class later and is open to students of all degrees across the university. The idea behind this list is that if students are interested in sustainability and want to get a different lens on the topic, these courses will offer that alternative perspective.

Core Courses in the Sustainability Studies Minor

Christina asks if any students have taken SUS 3003, the introductory course, and notes that she sees at least one student who has taken it. She describes this as a really awesome course that is taught every fall and spring semester. It serves as a really great introductory course to thinking about how culture and sustainability fit together.

She then moves on to describe the capstone course, which is another core requirement. The first course she shared (SUS 3003) and this capstone course are the two core courses that students need to take for the minor. The remaining electives, like the current economics class, could be substituted in to count towards the minor. Students need to take three total additional electives to round out the minor, in addition to the two core courses of 3003 and 4004.

The Capstone Course and Community Engagement

Christina explains that the capstone course allows students to work with community partners on a project. She emphasizes that this kind of experience is great to add to a resume. In general, all the courses in the minor have some sort of component where students are doing community-engaged work. This focus on real-world application and community partnership is central to the minor’s philosophy.

Internship Opportunities Through the Minor

Christina describes a course that is great for anyone interested in doing an internship in something related to sustainability. The course instructor, Mary Hahnemann, works with many different community partners and compiles a list of host sites every year for internships. If students want to do an internship in the fall or next spring, the course is offered in both semesters. Some of the internships are paid from the beginning, while others have a scholarship that students can apply for. The process is pretty straightforward. As long as students have demonstrated financial need, they are pretty much guaranteed to get the scholarship to help fund that internship experience.

Upcoming Sustainability Events

Christina notes that one of the slides says to present at the Sustainability Symposium, but it should now say to attend. She explains that earlier in the week, on Monday, the presentation submissions for the symposium closed. However, if students want to meet students from all five campuses of the university system—Crookston, Morris, Duluth, Twin Cities, and Rochester—they will all be gathering at the Institute on the Environment, which she mentioned earlier in her talk. She describes it as a really cool event. There is free lunch, and students get to network with local professionals. Christina emphasizes that it is really a fun experience to be uplifted and see what cool things their peers are working on. She notes that sustainability can be kind of a downer sometimes when you are looking at what is happening in the world, so the symposium is a fun place to be inspired by others’ work.

Environmental Justice Event

Christina asks if anyone has heard of environmental justice and notes that it is kind of a jargony topic, but she sees head nods indicating that some students are familiar with it. For those who have not heard of it, she explains that environmental justice is the intersection of social justice and the environmental movement. The Institute has a really great event focused on this topic, and it will be their fourth or fifth year doing it. The event will be held over on West Bank at the Humphrey School of Public Affairs. At the event, there will be networking, job and internship tables where students can find out about opportunities. Students can pack some seeds if they are planning on having a spring garden, though this is optional. And once again, there will be lunch and really great community partners sharing about environmental justice in their local community.

Other Programs and Leadership Opportunities

Christina mentions that the Institute has so many other programs, and notes that the slides showing them do not even include all of them. Many of these are paid leadership programs, except for the Germany study abroad course. She asks if anyone has participated in Undergraduate Leaders, and receives no affirmative responses. She then asks about the Germany study abroad course and again receives no confirmations. Then she asks if anyone has participated in Clean Energy Leaders and notes that one or two students raise their hands. She comments that it is great that the Institute has a resident INE person who has done a lot of their programming, and notes that when someone does something with INE, they tend to come back for more programming because there are so many great continued opportunities for development.

Details About Leadership Programs

Christina provides details about the Undergraduate Leaders program, explaining that it has an equity focus, is year-long, and includes a one thousand six hundred dollar stipend. The Germany course, she notes, is super fun. It is a two-week study abroad program over winter break. So if students cannot take a whole semester to study abroad or want to squeeze in another study abroad experience, this one is really great. She explains that Germany is really far ahead in renewable energy, so it is fun to see what they are doing there and come back and make that transition in Minnesota, because Minnesota has been copying a lot of what has been happening in Germany’s renewable energy process.

Christina explains that Clean Energy Leaders is a year-round program that includes a summer internship with a focus on clean energy. She emphasizes that everything she has shared, even though it says clean energy or something like that, is open to students of all degrees. She notes that, as the students surely know, we need everybody when it comes to climate change. We need the artists telling the stories, we need the researchers doing the research, and we need every other discipline that one could imagine.

Partnership and Host Sites

Christina shows a final slide that displays a variety of host sites and partners that the Institute works with for internships and community engagement. Students can see a mix of nonprofits, government, and corporate partners. She notes that if students are curious about any of those avenues, the Institute’s programming touches these different areas. She highly recommends these opportunities if students are looking for resume experience and professional development.

Closing of Guest Presentation

Christina concludes by saying that this is all she has to share and asks for questions. When no questions are asked, she notes that she guesses she would just add that she loves INE. She came from working there for five years, and it is really an awesome thing to have just down the hallway. She thanks the instructor and class for having her. She mentions that she has a slide at the end for the newsletter, and if students decide to sign up, it basically shares everything she has shared with them, but when it is relevant for applications. She offers to send the slides around to the class, and the instructor thanks her for her time.

Main Lecture: Climate Change Economics

Transitioning to the Day’s Content

The instructor takes over and thanks Christina for her presentation. The instructor then announces that they are going to get started picking up where they left off yesterday. They note that they have updated the slides and will be using the same slides as they were before. They acknowledge that while the slides are the same, they have added a new slide for the Day Two Agenda.

The Challenge of Aligning Lectures with Topics

The instructor comments that it is always hard having the lectures line up perfectly with the topics, as topics are sometimes longer or shorter than a single lecture period. The approach of using the same slides for multiple days is explained as a workaround to this challenge.

Review of Previous Day’s Content

The instructor reviews what was covered the previous day. They explain that the class talked about the basics of climate change and the sort of irrefutable need to think about this issue. The class also set up the fact that there are different economists who think about climate change differently, and there is really interesting debate among them. The instructor notes that they then talked about the first and seminal piece of work in the area of climate economics: the Nordhaus DICE model. The instructor explains that they are going to finish that discussion up today, hence they are on the same slides.

Today’s Agenda Overview

The instructor outlines what will be covered in today’s lecture. After finishing the DICE model discussion, they will contrast it with a different economist’s report: Sir Nicholas Stern’s Stern Report, which approaches climate change from the other end of the spectrum. The instructor explains that Nordhaus is essentially saying there is no really big need to worry about climate change and that a little bit of investment is all that is needed, with anything beyond that being too much. Whereas Stern uses a lower discount rate and finds that the situation is actually very dire, though the analysis goes much deeper than that simple characterization.

The instructor then explains that they will talk about how these contentions about what is the optimal amount of climate mitigation depend on a cool set of models that will be referred to as integrated assessment models. They note that the class started to see the basics of those models when they talked about Nordhaus and the idea of maximizing utility subject to the choice of how much to abate. The instructor explains that since Nordhaus came out with that basic approach, it has blossomed into a huge area of research with many different integrated assessment models. The instructor notes that Stern uses one of these models too, but it is part of a trajectory where climate and economic science is going.

Conclusion of the Day’s Topics

The instructor notes that they will conclude the lecture by showing how integrated assessment models connect to a really important policy-relevant number called the social cost of carbon. They explain that they will talk about the literature on the social cost of carbon and how it links to policy.

Administrative Notes and Questions

Before diving into the content, the instructor asks if there are any questions regarding timing or the two different assignments that they have outstanding. They remind students that there are the weekly questions due Friday and an assignment due Monday. They ask if there are any other questions, and when none are forthcoming, they indicate they are ready to proceed.

The DICE Model Results and Logic

The Punchline of the DICE Model

The instructor begins by noting that where they left off, they concluded with the punchline of the DICE model, but they did not dive into the results in more detail. The punchline is that in one of the more recent versions of the model with a three percent discount rate, it would be optimal to abate carbon emissions or other greenhouse gases if it was cheaper than fifty-one dollars per ton of carbon dioxide emitted. The instructor explains that this is the social cost of carbon, and they will formalize this concept more broadly in modern integrated assessment models. However, before they do that, they want to talk through the basic argumentation that Nordhaus used in developing his model.

The Basic Approach and Logic

The instructor explains that essentially, Nordhaus’s approach was to optimize and find the optimal amount of climate mitigation to maximize human happiness. The instructor notes that this is just like standard economic growth, where economists try to maximize growth. Here, Nordhaus is going to do the exact same thing, but with the added factor that there will be potential damages to the economy. The instructor explains that Nordhaus has the option to spend money on abating emissions, and maybe that will be beneficial in terms of having a positive return on investment in those emissions reduction programs.

The Central Tension

The instructor identifies the basic tension in Nordhaus’s approach: the economy can take steps to slow emissions, but that is going to be expensive. All that expense spent on abating emissions will come at the cost of less consumption today. The instructor notes that this fundamentally relates to the concept of intergenerational equity in the discount rate. Present generations would be asked to reduce their consumption in return for lower damages and thereby higher consumption in the future.

The DICE Model Framework

The instructor explains that basically, all the DICE model does is identify the optimal path for consumption, just like the Ramsey model they discussed earlier, but with these extra bits of information about climate and damages. With that context established, the instructor is ready to discuss what the actual results of the DICE model are.

The DICE Results

The instructor presents Nordhaus’s key findings from the DICE model. The DICE model says that the optimal path will have a two point eight percent loss in GDP. The instructor emphasizes that this is not saying what would happen if the world just did not do anything about climate change. This is with optimal mitigation according to Nordhaus’s framework. Nordhaus is saying that the world will lose some GDP, but this is the least amount of GDP that could be lost given the constraints of the problem. The corresponding temperature increase with this optimal mitigation is three point five degrees centigrade warming. Both of these figures are essentially by the year twenty one hundred, though Nordhaus centers on twenty ninety-five in this particular publication.

The Historical Significance of the Research

The instructor notes that this is a really old journal article, but it is the work that led Nordhaus to win the Nobel Prize. The instructor then shows the basic results in a table from the publication, which displays the impact of alternative policies and how they affect the discounted consumption, measured as the net present value of consumption going way out into the future.

Nordhaus’s Policy Comparison

The instructor explains that Nordhaus looked at a business-as-usual plan where society does not do anything at all with respect to climate change: no controls, no emissions abatement. Under this scenario, the discounted value of utility is measured in trillions of dollars. If society does not do anything at all on climate policy, the value comes to seven hundred thirty-one point sixty-nine trillion dollars. That is a lot of value.

The instructor notes that the critical value to examine is how that number compares with different policy options that Nordhaus considers. The first policy he presents is the optimal policy. When this policy is implemented, the discounted value of utility comes to seven hundred thirty-nine trillion dollars when discounted to the present. The instructor notes that this is a little bit better than the business-as-usual scenario, representing about nine trillion dollars more in value. This is measured as two hundred seventy-one billion dollars of extra value. The instructor notes that this is good, but it is remarkably small. The instructor explains that this is the basic finding in a lot of Nordhaus’s results: even the very best policy just is not going to do very much to improve welfare compared to doing nothing.

Analysis of Delayed Action

Nordhaus also examines the cost of delaying climate action by ten years. The instructor explains that according to Nordhaus’s model, delaying action does not cost that much. The discounted value of utility goes down just a little bit with a delay, but that might be useful because the world can spend more time gathering science. According to this logic, there is not much cost to that approach of delaying action.

The Real Punchline: Drastic Measures

The instructor notes that the real punchline of Nordhaus’s analysis came when he looked at what if society implemented drastic measures that were being debated at the time to reduce emissions. At the time, the idea was to reduce emissions by twenty percent. The instructor notes that this might sound antiquated now because current climate policy discussions are focused on getting to net zero, literally zero emissions. So Nordhaus was considering a very weak policy by today’s standards. But a twenty percent reduction is what he plugged into his optimization model.

The instructor explains that this scenario of reducing emissions by twenty percent has a massive impact in Nordhaus’s model. This lowers the value of GDP going forward by ten trillion dollars. In other words, optimizing gives a little bit of benefit, but doing the existing climate policy that was being debated at the time would be massively costly. Nordhaus’s conclusion was that implementing such a policy would be a very dumb idea from an economic standpoint.

The Controversial Nature of These Findings

The instructor notes that this was such a controversial finding because it was hated from both directions. People who did not like to admit that climate change was happening did not like Nordhaus’s admission that it was happening. But people who did care about climate change did not like his findings either, because he was arguing that nothing should be done, or very little should be done. The instructor emphasizes that this was a very controversial first foray into systematically combining economics and climate science.

Nordhaus’s Recommendation for Carbon Pricing

The instructor explains that despite all of this analysis, Nordhaus showed that greenhouse gases should be cut a little bit. He was not a climate denier. He was saying that climate change is happening and society should do something about it, just not very much. His main point was that this should be done with a very small carbon tax. The optimal carbon tax in Nordhaus’s model slowly ramps up to between twelve and fifty dollars per ton. The instructor notes that this is really different than the tax that would be required to get a twenty percent reduction in emissions. That would require massive taxes measured in dollars per ton of carbon. Nordhaus’s point is that these high carbon taxes would hurt the economy. High taxes mean that society pays the government for each ton of carbon emitted, and this money goes toward preventing emissions instead of being used for consumption. And from an economic standpoint focused on maximizing consumption, that is framed as a bad thing.

Evolution of Nordhaus’s Work

The instructor notes that Nordhaus’s work has evolved extensively. The work that has been discussed so far is the old work, which is why they wanted to show the seminal work. The latest release was the two thousand twenty-three version of the DICE model, which shows different results. The instructor shows a graph with the green line in the middle representing the optimal temperature policy pathway according to this latest model.

If one takes into account Nordhaus’s framing of the DICE model and tries to make all of humanity as happy as possible, society should be reducing emissions. But what do the results show? The equilibrium value in twenty one hundred is about three point one degrees centigrade warmer than today. It is not much different than Nordhaus’s very initial estimates from decades earlier. He has changed a lot in how the model is computed. One thing is that he has admitted that the future could be much worse. The difference between the optimal policy and doing nothing at all is bigger now because some of the extreme risks of climate change have become more evident.

But the instructor notes that even with these updates, Nordhaus’s recommended warming is still not very much compared to what many policymakers are thinking about. This is illustrated by the reference to the two-degree target. If society were to implement something that kept warming below two degrees centigrade instead of Nordhaus’s optimal policy, it would look quite different on the graph. The instructor notes that Nordhaus’s preferred option still has much more warming than what the two-degree target would allow.

Criticisms of the DICE Model

Introduction to Criticism

The instructor says that the DICE model is controversial, and here is a good way to think about the main criticism. If there is a catastrophe coming—oh no, this is going to hurt the economy—well, as Keynes famously said, in the long run everyone is dead. If there is a catastrophe, it does not matter much if society loses out on a little bit of luxury consumption today if the whole system collapses. The instructor notes that of course, if one has a really high discount rate and does not think that catastrophe will happen until after one is dead, maybe it does make sense economically to let that catastrophe happen and worry about the economy instead. The instructor compares this to worrying about an incoming asteroid.

A Humorous Illustration

The instructor makes a side note about modern AI capabilities. They mention that there is a copyrighted comic they were basing their criticism on that they could not use directly. So instead, they typed into artificial intelligence software: make an image about dinosaurs talking about how the economy would be wrecked from a meteor. They got back an image that, while not quite as good as the original comic in terms of the humor and artistic quality, serves the same purpose and is not copyrighted. The basic idea of the comic and the AI image is the same: even if it is true that an asteroid would hurt the economy, that is a pretty small concern compared to the asteroid hitting and destroying all life.

The Discount Rate as Primary Criticism

The instructor turns to discuss the actual criticisms of the DICE model more seriously. They note that in upcoming problem sets, exams, and quizzes, the class is going to shift more toward arguing intelligently about issues like climate policy. Rather than just doing supply and demand and solving basic algebra, students will be expected to construct arguments about complex policy choices. The instructor notes that they are foreshadowing that if they were to write down some criticisms of the DICE model, it would not be off-target for them to ask a question like: what are the two most persuasive arguments against the DICE model in your perspective? The instructor advises students to keep this sort of analysis in mind as they proceed.

The instructor identifies the first major criticism: the discount rate. This has really been the primary criticism of the DICE model. The instructor explains the criticism by returning to the asteroid analogy. We should not care about the asteroid if we are going to die before it hits us. In other words, with a very high discount rate, we do not value future generations enough. Nordhaus had such a high discount rate that potential future damages mattered almost nothing in the model. So the criticism is that the discount rate is too high.

The Moral and Philosophical Dimension

The instructor then asks a provocative question: who is Nordhaus to choose how much society should care about future generations? That is ultimately his opinion. The instructor notes that there is much more behind the logic of how Nordhaus gets to his discount rate choice, but it is fundamentally a value judgment. Some people care a lot about their kids. There is literature out there trying to estimate behaviorally how much people actually care about future generations’ well-being, and one of the key events in a person’s life that increases how much they care about future generations is having children. And then the next one is having your children have children. These life events drastically reduce the discount rate that people apply because now people are getting utility from knowing that their grandkids are going to be alive rather than scavenging for meat in an apocalyptic landscape. The instructor emphasizes that this is a moral issue around sensitivity to the discount rate.

The Second Criticism: Damage Functions

The instructor presents the second major criticism of the DICE model: damage functions. The instructor recalls that the damage function that Nordhaus used was essentially temperature plus temperature squared. He had some coefficients to say how much temperature and temperature squared mattered in the damage function, but this was really simplistic. There are dozens, hundreds, thousands of ways that the economy will be affected by climate change. Many of them cannot even be expressed just as temperature. For example, sea level rise is a function of temperature, but it matters drastically whether society implements seawalls, has mangroves protecting the coast, and so on. The instructor notes that there is a whole body of work showing that the damage functions in the DICE model were artificially low, incorrectly low, or just wrong.

Tail Risks and Catastrophic Scenarios

One key issue that the instructor identifies is that the damage functions underestimated what are called tail risks. Tail risk is something that is not in the center of a distribution but in the tails of the distribution. With a normal distribution, extreme events are really improbable. But if the distribution had fatter tails, suddenly there would be non-zero probabilities that continue out quite far in the tails. Many people worry that climate change has fat tails because the world has only been experienced at current temperature conditions. We do not know what it is going to look like at much higher temperatures. If the distribution of climate outcomes does have fatter tails, then everything in Nordhaus’s damage function could be wrong.

Tipping Points and Feedback Mechanisms

Related to the tail risk issue is the absence of tipping points in the DICE model. The damage function just stays smooth—there is no feedback mechanism included. If it gets warmer, the permafrost might melt. If the permafrost melts, all the methane trapped there might start releasing, accelerating damages and warming further. This kind of feedback is terrifying because humanity has never seen it happen. We have never been at temperatures where permafrost melts, so scientists simply do not have direct observations or comprehensive science about what happens in those scenarios. But there is enough science to know that this kind of tipping point might matter a lot.

Technical Criticisms: The Global Agent Problem

The instructor notes that there are a few other technical criticisms of the DICE model. It is a super simple model, and the biggest challenge is that it is a single global agent, sort of like the Ramsey model from earlier in the course. There is no trade between regions, no regional inequity, no distributional effects. The instructor poses a hypothetical: what if the optimal policy is that ninety percent of countries should stop all agriculture and starve so that one country representing ten percent of the population becomes really happy? That would be unfair, but the Nordhaus model does not say anything about that kind of distribution of harms and benefits.

Technology Assumptions

Finally, the instructor identifies a criticism related to technology assumptions. Technology matters massively in determining the costs of addressing climate change. In the DICE model, there is no endogenous way that the model captures how we learn to mitigate carbon better or even get better air conditioners. Critically, the model does not represent how technology might evolve to make it cheaper to reduce emissions. This has been a critical oversight. When Nordhaus wrote the original DICE model, coal-fired power plants were producing at roughly six cents per kilowatt-hour, and solar was something like twenty-eight cents per kilowatt-hour. Given this cost structure, the idea that society would have to harm the economy to emit less matched the technology of the time.

The Technology Cost Revolution

The instructor notes that, fast forward to the present day, coal is still about six cents per kilowatt-hour. But what is solar now? The instructor asks if any students know, and then answers that it is way less than coal. Solar is now about three cents per kilowatt-hour, and it drastically depends on location. There are other challenges like distribution and getting the power grid up to specification to handle large amounts of renewable energy, so it is not trivial. But the fundamental equation has changed. The instructor notes that when the news is saying that the government is going to deregulate coal-fired power plants, they do not care that much because even the most selfish person would be pretty dumb to invest in coal right now. Even if someone had a guarantee that there would never be another negative law about coal, it would be like investing in whale oil for lights. Whale oil can be burned, and it was used extensively in the past for illuminating cities, but it is probably not a good investment, even if someone hates the environment and loves fossil fuels.

The Implications for the Model

The instructor explains that solar is just cheaper now, and wind is similar. There are going to be a lot of changes in the energy system. The way this affects the DICE model is that Nordhaus assumed really expensive mitigation because mitigation was expensive at the time. But what happens if mitigation is cheaper than not mitigating? It is kind of a paradox. It should happen automatically through profit incentives, right? Nonetheless, this has been a huge criticism of the DICE model. The model’s assumptions about technology costs have become substantially outdated as renewable energy has become much cheaper.

The Spectrum of Climate Economics Perspectives

Positioning the DICE Model on the Political Spectrum

The instructor turns now to contextualize the DICE model within the broader spectrum of economic thought on climate change. On the slide, the instructor shows a line representing the spectrum of political debate. On one end, all the way to the right, is where the instructor places Nordhaus’s position. The instructor then says that they are now going to talk about the other end of the spectrum: the Stern Review. The instructor notes that they would say that Nordhaus is farther to the right than the Stern Review is to the left. The Stern Review is definitely more to the left of consensus economics, but it definitely has much more emphasis on intergenerational equity and considers many of the real criticisms of the DICE model that have just been discussed.

The Stern Review

The Commission and the Author

The instructor begins the discussion of the Stern Review by explaining what this report is. The Stern Review was commissioned by Her Majesty’s Treasury in the United Kingdom under Chancellor Gordon Brown. The person who led the review, Nicholas Stern, is actually Sir Nicholas Stern because he is knighted. The instructor notes that that is how much the British care about him. This is a government-sponsored report on the economics of climate change.

Related Government Reviews

The instructor makes a side note that they are going to be talking about the Das Gupta Review later in the course. The instructor mentioned that a couple of times already. A similar story applies to the Das Gupta Review—it was commissioned by Her Majesty’s Treasury to do another report on a related topic. The instructor notes that their group was actually asked to contribute to the Das Gupta Review process, and they got flown out to Her Majesty’s Treasury, which turned out to be one of the coolest trips they have ever done. They went into this old castle-like building that was formerly the horse grounds where different sovereigns and kings and queens of the British Empire lived. It is pretty cool, and it turned out to be where a lot of movies had been filmed. However, the instructor notes that they want to get back to discussing the Stern Review specifically.

The Core Recommendation

The instructor explains that Stern’s basic recommendation was to spend one percent of global GDP per year on mitigation of climate change. The instructor notes that this was huge—hugely more than the Nordhaus approach suggested. This is a central and striking difference between the two models’ recommendations.

The Key Findings of the Stern Review

The instructor moves next to the key findings of the Stern Review. The biggest key finding of the Stern Review is that inaction is really bad. As economists, damages are measured based on loss of GDP. According to the Stern Review, inaction would result in loss of five to twenty percent of global GDP by the year twenty one hundred. This is in stark contrast to the two point eight percent loss in the Nordhaus work. This is a massive difference in the estimated costs of climate change between the two models.

The instructor notes that the other big part of the Stern Review is that although one percent of global GDP is a lot, and Stern is saying that a higher amount should be invested in climate mitigation than Nordhaus suggested, Stern also found that this level of investment is much more effective. Basically, for the Nordhaus assumptions about technology, to get the same level of emissions reduction that Stern recommends, it would have been more expensive. Here, Stern is saying invest one percent of global GDP per year, and that will have a huge impact on reducing climate change damages.

The instructor notes that other points in the Stern Review are that addressing climate change is urgent and very necessary to consider equity, but the key findings are that there is much more at stake from not acting than Nordhaus suggests, and the efficacy of investing in mitigation is really quite high.

Comparing DICE and the Stern Review

The Central Question

The instructor notes that they have been building to this point throughout the discussion. Why do Nordhaus and Stern have such different opinions on the optimal amount of GDP to spend on climate mitigation? The answer comes down to one key variable: the discount rate. That is why the instructor spent so much time on the discount rate earlier: it is fundamentally important to understanding this comparison.

The Discount Rate Comparison

The instructor presents a basic comparison of the two reports side by side. The Nordhaus DICE model has a pure rate of time preference of one point five percent. The Stern Review, by contrast, has a pure rate of time preference of zero point one percent, basically saying that the future matters a lot to society. This much lower discount rate leads to substantially higher estimates of the present value of future damages that Stern thinks would be optimal to abate. Stern in twenty oh six was saying that there is eighty-five dollars of damage per ton of carbon dioxide, while Nordhaus said that there is only thirty dollars of damage. Because all those damages are going to happen in the future, and when they are discounted back to the present using a higher discount rate, they do not matter as much. The key difference comes down to that discount rate and what that implies about how much society should care about its grandchildren.

Beyond the Discount Rate

The instructor notes that it is not just the discount rate, though. The Stern Review also includes other things that broaden the comparison. The instructor shows a visual with contributions from different factors. Five percent is Stern’s most conservative estimate of damages. But he also, and the instructor agrees with this assessment, identifies other types of damages like non-market damages—things that are not just saying that industry produces less output. Non-market damages include the fact that people die or suffer from consequences that do not filter through production functions. He also considers equity and the idea of fat tails, including some sort of risk of catastrophe. These all contribute to the much higher levels of losses that Stern estimates.

The Impact of the Stern Review

The Influence on Policy

The instructor notes that the influence of the Stern Review was essentially transforming the policy landscape into really caring about climate change. This review was behind the Kyoto Protocol, which was the first real substantial policy investment in addressing climate change, and then finally the Paris Accord, which is what is most binding on international policy today.

Summary of the Stern Report

To summarize, the instructor explains that Stern argued that the benefits of strong, early action on climate change outweigh the costs of not acting. The fundamental framework is the same as Nordhaus—utility maximization, cost-benefit analysis—but done differently, with different assumptions that the instructor argues are more realistic and defensible. This is a challenging debate, but a very, very important one for society to have.

Transition to Integrated Assessment Models

Recognizing the Range of Opinion

The instructor now moves toward the next section of the lecture. They note that the discussion has covered the space of different opinions and different ways to use models to analyze climate change. But the instructor asks: what should society actually do? Both Nordhaus and Stern are saying what their optimal choice is based on their models, but there is a range of optimal choices. On one hand, some economists argue that society should not do anything about climate change. On the other hand, others argue that society should do a whole lot right now. How is society going to make an informed decision? How can policymakers come up with the best estimate for the best policy?

Truman’s One-Handed Economist

The instructor references a famous quote from President Truman. Truman once said he really wants a one-handed economist because they always have this problem. They give you two hands. This is true with climate economics. There are two hands here: one saying do a little, one saying do a lot. How is society going to make an informed decision?

The Solution: Integrated Assessment Models and the Social Cost of Carbon

The instructor explains that the answer will come through using integrated assessment models to establish the correct social cost of carbon. This will be the framework through which society can come up with a more definitive estimate for the best policy.

Shifting Slide Decks

The instructor notes that they want to dive into more detail about how economists come up with these numbers. To do that, they are going to switch to the next slide deck. The class has been on Chapter fifteen on climate change. Now they are switching to the next one, which talks about the social cost of carbon and how it is calculated in different integrated assessment models.

The Social Cost of Carbon

Defining the Social Cost of Carbon

The instructor begins by defining what the social cost of carbon is. The instructor notes that they have mentioned it a few times, but now that the debate about climate policy is laid out, everyone is ready to define it precisely. The social cost of carbon, or SCC, is the monetary damages of one additional ton of carbon. It sounds simple, but there is a lot of complexity behind calculating this. The instructor notes that it is important to understand that the social cost of carbon is measured in the present. This is going to depend on calculating the net present value of all future damages from one additional ton of carbon emitted today.

The instructor emphasizes that the choice of how much carbon society should emit is being made right now, or society can think about the same calculation in future time steps, but the decision is fundamentally happening in the present. The benefits of reducing emissions give society a flow of value over time. This is not so different from any capital investment decision. A factory owner considering upgrading their machinery faces a big upfront cost that they have to consider in the present. But the point is that the new machinery gives a flow of value out into the future. So factory owners always calculate the net present value. That is basic accounting. And that is what society needs to do with climate change.

The Complexity of Calculating the Social Cost of Carbon

The instructor explains that the social cost of carbon is complex because it involves linking many different steps and many different models. With the production example, a firm understands what they are producing and their machine, so it is kind of simple. But here, society is integrating many different parts from climate science, the economy, carbon cycling, damages, and these all have different models. Society is going to need to integrate these different models into an integrated assessment model.

Integrated Assessment Models

Defining Integrated Assessment Models

The instructor now asks: so first, what is an integrated assessment model? It is going to combine models on each of five steps, and it is important to understand each of these steps. Society is trying to figure out the correct social cost of carbon, so it needs to consider each one of these steps carefully.

Step One: Baseline Emissions

The first step in an integrated assessment model is determining what is going to happen with baseline emissions. Given assumptions like population growth or technology, how much emissions will make it into the atmosphere? The instructor notes that this is not easy to figure out because if everyone switches to solar power, there will be a lot less emissions into the atmosphere. Climate scientists and economists make reasonable projections of things like economic growth, population, and technology, but also harder things like the preferences of consumers for meat versus vegetables. This matters because it determines the equilibrium level of meat production and the methane that cattle produce. The instructor notes that there are different emissions trajectories based on different assumptions of what type of economic activity is going to happen in the future.

Step Two: Emissions to Concentrations

The second step in an integrated assessment model is the emissions-to-concentrations step. Emitting is measuring it at the smokestack, essentially—how much carbon dioxide and other greenhouse gases get emitted into the atmosphere. But concentrations are actually more complicated because they involve biophysical detail. The carbon model shows what happens if humanity emits a bunch of carbon. Higher temperatures cause more plants to grow, like more algae in the ocean. This algae grows off carbon, so some of the emitted carbon gets photosynthesized by plants and removed from the atmosphere as a problem. Some of the carbon gets dissolved into the oceans, and of that, some sinks to the ocean floor and becomes mineralized—trapped forever. But some of the carbon gets emitted back out into the atmosphere. Going from the relatively easy-to-observe thing—emissions—to concentrations requires detailed modeling of these biophysical processes.

Representative Concentration Pathways

The instructor notes that the punchline here is that the scenarios community has defined something called Representative Concentration Pathways, or RCPs. The class has seen this before, and they are going to use it again for the rest of the course. Essentially, RCPs combine the assumptions that go into emissions with a model of what happens to those emissions to get the concentration of greenhouse gases that humanity is going to have in the atmosphere.

Step Three: Temperature from Concentration

The third step is temperature. From concentration, it is not obvious what happens to temperature. This step is driven by a very controversial number called climate sensitivity. Climate sensitivity measures what happens when carbon dioxide concentrations double. It measures what percentage increase in temperature results from that doubling of CO two concentrations. This is one of the very controversial numbers in climate science. If someone has a climate sensitivity parameter that is low, it means that society can emit lots of carbon dioxide and it will not really matter much because Earth’s system is not very sensitive to emissions and the planet will not get really hot. If climate sensitivity is high, each ton of carbon dioxide has a large effect on temperature and future warming. How much will temperature go up from an increase in carbon dioxide concentrations? There is a whole set of models trying to estimate climate sensitivity.

Step Four: Temperature to Damages

The fourth step is damages. How does temperature change damage the economy? The instructor notes that this is typically expressed as a fraction of GDP lost. As mentioned earlier, Nordhaus used a damage function of temperature plus temperature squared, but that is hilariously simple. Damages are probably going to depend on many different damage pathways. But the damage calculations account for many different things that matter beyond just temperature. Things like labor productivity matter. It is really hard to work in an agricultural field when it is super hot. The instructor shares a personal anecdote about doing corn detasseling as a youth. They were paid minimum wage to go pull the tassels off corn in a field. It required a ton of temporary staff, and the instructor’s high school class all did this. It felt like a lot of money to a sixteen-year-old, but it was awful because you are just standing out in the heat in the sun all day. Well, the instructor notes that experience gets much more painful if it is hotter out. So damages are really important, and there will be more discussion of that later.

Step Five: Discounting Future Damages

The final step is what the instructor has been emphasizing throughout: intergenerational consideration. How does society care about future value? How does society discount that stream of benefits back to the present?

Moving Forward with Analysis

The instructor notes that from here, society is going to take this basic pathway through the five steps and look at the data. They will rely on a very new article. The instructor is going to assign this as a reading for the next class. The article is Renert et al., and students should read it carefully. The instructor notes that the class is going to walk through the logic of computing the social cost of carbon and what the best evidence is on this thing. The instructor will send out an email to remind everyone of the reading, so students should not worry about that.

The Goal: Correct Estimate of Total Damages

The instructor explains that the class will step through each of these five steps, and they will conclude with what is the correct estimate of the total value of damages from one extra ton of carbon. This is important because it is essentially the benefit and cost ratio that society cares about. Society is trying to figure out the right amount of climate action. It is not zero because that is super expensive and it is not going to happen anyway, but it is probably also not this extreme of having three point five degrees centigrade warming that Nordhaus’s original model suggested.

Conclusion and Administrative Notes

Wrapping Up the Lecture

The instructor notes that they are going to leave it there for today. They will email out information about the Renert et al. reading that students need to do before the next class. The instructor reminds students to do their assignment and their weekly questions. The instructor notes that they should have said this in the assignment instructions, but Canvas defaulted to the questions being due by the end of the day on Friday. The instructor clarifies that they will not change this because that would be unfair to students who may have already completed the work. It is still technically due by the end of the day Friday, but if students can do it before class on Friday, that might inform the class discussions better. So the instructor informally advises students to aim for completing the weekly questions before Friday’s class.

Final Questions and Dismissal

The instructor asks if there are any questions about the material covered. When no questions are forthcoming, the instructor indicates they are good and thanks everybody for their attention during the lecture.

Content (Day 2)

Course Overview and Administrative Updates

Midterm Exams and Grading

The class began with the distribution of midterm exams, which the instructor noted had been intentionally designed to be difficult, with scores then curved upward to provide more realistic grades. The instructor explained that this approach serves an important pedagogical purpose: if exams are too easy and everyone receives perfect scores, it becomes impossible to distinguish between students who truly understand the material and those who do not. Students were given time to review their exams and the answer key immediately following the distribution. The instructor directed any students with questions about grading to contact Ryan McWay, the teaching assistant responsible for grading, whose email address was available on the course website.

Course Schedule and Trajectory

The instructor provided a detailed overview of the remaining course schedule, noting that while preliminary, the basic structure would remain consistent throughout the semester. The schedule was presented to give students a sense of how the course was pivoting toward more applied questions and practical tool usage. The course had just completed spring break and was currently positioned at the unit on the social cost of carbon and climate integrated assessment models, which represented Part B of this topic.

Upcoming Course Content and Practical Components

Scheduling and Learning Objectives

Looking ahead, the following week would introduce future scenarios, building upon the representative concentration pathways that had already been introduced. The concept of shared socioeconomic pathway scenarios would be explored in greater detail, with particular emphasis on how these pathways define changes in land use, deforestation, and cropland expansion. On Wednesday of the following week, the instructor would introduce Geographic Information Systems, or GIS, and requested that students bring laptops to class for hands-on practice. The instructor acknowledged that two students had indicated they had tablets instead of laptops and asked whether they could arrange to bring a laptop, as GIS work on tablets, while possible, is considerably more difficult. For students without laptop access, the instructor offered to work out alternative arrangements and requested that they contact him early.

Natural Capital and Ecosystem Services

Following the introduction to GIS and its foundational concepts, the course would transition to discussing natural capital and ecosystem services. The instructor emphasized that these topics are spatially explicit in nature and that most of them depend directly on land use patterns, which explained the sequence of the course structure. After introducing these concepts in general terms, the course would then address the original application of these frameworks: placing a price on nature through valuation methods. The instructor explained that economists use various methodologies to estimate the monetary value of ecosystem services, such as quantifying the worth of pollination or the value of wetlands.

Ecosystem Service Models and Applications

With the valuation toolkit in hand, the course would then spend considerable time examining specific ecosystem service models in depth. The focus would begin with carbon storage, then move to sediment retention, and would include pollination as a major area of study. The instructor noted that mid-week on April 8th would provide time for discussion and application exercises specifically focused on how these concepts would apply to the country reports that students would be developing throughout the course. This practical connection would be reinforced throughout the remainder of the semester. The course would further explore ecosystem services and uncertainty, then return to scenario analysis, examining not only the shared socioeconomic pathways but also emerging positive vision scenarios. The instructor noted that even the most optimistic shared socioeconomic pathway scenario available tends to be quite depressing, and that these positive vision scenarios would offer more encouraging possibilities for the future.

Policy and Guest Lectures

The course would transition to discussing both market-based and real-world policies designed to help society progress toward these positive visions. The instructor had arranged two guest lectures as part of the course. The first would feature Colleen Miller, the senior ecologist at NetCap Teams, who would address biodiversity and its foundational role in enabling ecosystem value. While biodiversity is not itself an ecosystem service, the instructor explained, it is the essential basis upon which ecosystem services depend and which allows ecosystems to thrive. The second guest lecturer would be Dr. C. Ford Rungi, who had previously taught the climate class at the university and would discuss land as an input to production from an economist’s perspective.

Course Conclusion and Student Presentations

The final portion of the course would focus on connecting all of these various topics into a combined framework for Earth economy modeling. On the second-to-last day of class, students would present their country reports in what the instructor called “lightning talks.” These rapid presentations would be an important part of the grading for the final project. All student slides would be combined into a single slide deck, and each student would have just three to five minutes to present the key findings about their assigned country. The instructor explained that keeping all presentations to one day would require them to be brief but impactful. A concluding lecture would follow on the last day of classes, and then students would take the final exam.

Hands-On Tools and Applied Economics

The instructor emphasized that many of the upcoming course components would be decidedly hands-on. For the GIS module, students would actually learn to use the software rather than simply study it conceptually. For the ecosystem service components, the course would use a tool called InVEST, which actually computes ecosystem service values for a given landscape. The instructor explained that these hands-on exercises would form the basis of the country reports and represented the applied nature of being within an applied economics department.

Integrated Assessment Models and the Social Cost of Carbon

The Five-Step Process Simplified

The instructor began the substantive content of the lecture by presenting a highly simplified version of the five-step integrated assessment model process. The core concept was that emissions lead to concentrations, which lead to temperatures, which lead to damages, which lead to discounting. The ultimate goal of this entire process is to arrive at a single, policy-relevant value: the social cost of carbon, expressed in dollar terms per additional ton of CO2 emitted. More precisely, the social cost of carbon represents the net present value of the additional damages expected to result from the emission of one ton of CO₂.

Current State of the Art

The instructor explained that rather than walking through this process conceptually, he would emphasize what the current state of the art reveals about each step. He noted that the discussion would be based on a recently published paper in Science that presented ranges and uncertainty estimates around each parameter according to state-of-the-art models. This approach would provide students with the most up-to-date understanding of where current economic science stands on these critical questions.

Step One: Emissions

Measuring Current Emissions

The first step in the integrated assessment process addresses how we determine emissions levels. While it might seem straightforward to simply report current emissions by measuring what is being emitted today, the process is actually considerably more complex than this suggests. Comprehensive tracking requires monitoring all different plants and industrial sources and accumulating emissions data across all of them, which represents a significant research and administrative task. However, modern technology has made this increasingly feasible. Satellites can now be used to keep track of where carbon is being emitted from various sources, and can similarly monitor other chemicals such as methane. The instructor noted that a recent article, published just a day or two before the lecture, had demonstrated that satellites can now pinpoint specific firms that are major methane emitters. These appear as tiny dots on satellite maps marking locations where enormous quantities of methane are being released, and observers can actually see the plume of methane blowing downwind from these particular plants.

Satellite Monitoring and Enforcement

The instructor found this technological development particularly compelling because satellites offer novel mechanisms for enforcing methane controls. It is much more difficult for polluters to get away with violating methane emission standards when such violations are detectable from space and can be constantly monitored. However, the instructor added an important side note regarding the artificial intelligence boom and its environmental implications. Data centers are being built in massive numbers and, according to the instructor, are driving increases in methane emissions. Many of these data centers are installing off-grid power plants, mostly powered by natural gas, and locating them right next to the data center to avoid the complexities of connecting to the grid. However, this approach has a significant downside: it resembles the Wild West, with many different entities rapidly building generators without proper coordination or standards. The instructor observed that this uncontrolled installation boom is unlikely to result in state-of-the-art methane emission control systems, and so these emissions can often be detected from space, which ideally provides a mechanism to identify and potentially control these emissions.

Projecting Future Emissions

Measuring current emissions is one challenge, but projecting future emissions is substantially more difficult. Future emissions cannot be simply observed with satellites; instead, they require making projections about numerous interconnected variables. The instructor noted that the paper highlighted three particularly important categories of projections.

The first projection category concerns population dynamics. The paper presented a central estimate, represented by a dark blue line, which comes from various sources but is fundamentally based on demographic models. These models track and incorporate factors such as fertility rates and death rates across different countries. However, enormous uncertainty surrounds population projections. The darker shaded area around the central estimate represents the range within which roughly two-thirds of all population estimates would fall, while the lighter outer boundary represents the range capturing about 95 percent of estimates. The central projection suggests that global population will peak at slightly above ten billion people around the year 2100, representing an important variable in determining future emissions because each additional person represents an additional consumer who drives a car and purchases goods and services.

The second critical set of projections involves the average per capita GDP growth rate in different countries. This matters substantially because countries that experience rapid economic growth will see increases in consumption levels, which directly drives emissions upward. The instructor made a somewhat provocative observation: being poor is environmentally beneficial from a climate perspective because poor people consume less. This observation connects to frequently cited statistics showing that people in the United States produce approximately eight times more emissions per capita than people living in India. The instructor explained that this disparity essentially comes down to the fact that Americans consume eight times more in economic value of goods and services. While the composition of consumption matters, the instructor emphasized that the dominant factor determining emissions levels is fundamentally the quantity of consumption, which is modeled through the GDP per capita growth rate.

The third and final major category involves translating all of these socioeconomic projections into an actual emissions pathway. More consumers and greater purchasing of energy-intensive goods and services inevitably leads to higher emissions. While many additional projections feed into emissions models beyond these three, the instructor explained that these three categories represent the fundamental basis for arriving at a given emissions projection path. The central estimate from the models examined in the paper shows emissions peaking approximately at the present time, which must occur because if the peak were delayed and emissions continued rising, the consequences would become extreme. However, the instructor noted that this projection aligns with what is currently being observed in the real world. China, for the first time in its modern history, is actively reducing its total emissions rather than merely slowing the rate of growth in emissions. This reduction is occurring because solar power is rapidly displacing coal-fired power generation, with coal plants operating at substantially reduced capacity. The instructor characterized this moment as the climax of the environmental story: if the story began at the start of the Industrial Revolution, humanity is now at the critical juncture where success or failure will be determined.

Uncertainty in Emissions Projections

The instructor introduced an important conceptual point about representing uncertainty in projections, noting that the paper demonstrated distributions around the central emissions estimates but that he would explain two fundamentally different ways of thinking about uncertainty, which would be contrasted in the subsequent lecture.

The first approach to representing uncertainty is through distributions. In this method, a central estimate is presented, with uncertainty represented through an envelope around that central line. This envelope shows what is considered most likely, though an even broader envelope representing less probable but still possible outcomes typically extends beyond. The instructor explained that in principle, if a third dimension were added to the graph, there would be a bell-shaped distribution centered on the central estimate. As one looks further into the future, these distributions become progressively wider. The reason for this widening is fundamental: the further one projects into the future from the present, the less accurate predictions become. Because this three-dimensional representation would create an unattractive and difficult-to-read visualization, one typically represents uncertainty in two dimensions using the envelope approach. However, the instructor emphasized that understanding the underlying three-dimensional distribution is critical: uncertainty starts relatively narrow based on what is currently known and expands progressively with time as our predictions become less certain.

The second approach to representing uncertainty involves scenarios. Rather than presenting a central estimate with a distribution of uncertainty around it, this method develops several distinct, explicit projections of what might occur under different circumstances. Instead of one large envelope representing all possible emissions trajectories, specific driver assumptions are selected, and the resulting single emissions pathway is traced under those particular assumptions. This approach is useful because it makes transparently clear what combination of factors produces what outcome. The instructor illustrated this with five representative concentration pathway scenarios. RCP 8.5 represents the very concerning scenario of fossil fuel-based development trajectories continuing forward. This can be contrasted with alternative scenarios such as RCP 7.0 or more optimistic scenarios such as RCP 1.9. The instructor noted an interesting feature: some scenarios actually show negative emissions values, which reflects the possibility that humanity might successfully remove carbon from the atmosphere through technological means and sequester it. The fundamental point the instructor wished to make was that uncertainty can be expressed quite differently through scenarios than through distributions. Rather than saying there is a wide range of possible emissions, one instead says there are multiple scenarios, and society does not yet know which scenario it will end up following. This scenario approach can provide a more useful framework for understanding how different drivers produce different outcomes, because one can ask directly: what is different about the RCP 8.5 scenario that produces such high emissions?

The instructor concluded this discussion by noting that throughout the Nature paper being examined, the discussion mixed and matched these two types of uncertainty representation, sometimes using distributions and sometimes using scenarios.

Step Two: Concentrations

Atmospheric CO₂ Concentrations and Carbon Pools

The second step in the integrated assessment model pipeline involves translating emissions into atmospheric concentrations. The instructor presented the same plots shown earlier but added another plot illustrating atmospheric CO₂ concentrations. This new plot takes into account the different carbon pools in the Earth system and models how emitted carbon is partitioned among them. A portion enters the atmosphere directly, while other portions are absorbed into the ocean or converted into trees through photosynthesis. Once a model of these processes is developed, scientists can attempt to project how a given emissions trajectory will translate into a concentrations trajectory. As with the emissions projections, this step introduces additional uncertainty into the model chain.

Recent Updates to Carbon Cycle Science

The instructor presented important recent findings that had changed the understanding of carbons cycles compared to assessments from approximately five years earlier. New evidence suggests that the ocean is not performing its carbon absorption function as effectively as previously thought. Specifically, as global temperatures increase, the ocean appears to absorb carbon less efficiently, which means that for any given level of human emissions, the resulting atmospheric CO₂ concentration will be higher than previous models had projected. The instructor characterized this as an example of science updating itself as new evidence emerges. The implication is that uncertainty bands in this particular step should perhaps be shifted upward as scientific understanding improves.

Climate Science and Uncertainty

The instructor took a moment to explain how this kind of updating represents how science is supposed to progress. Scientists slowly learn what the better parameter values are and gradually narrow the bounds around their uncertainty. However, the instructor noted that climate skeptics sometimes misuse this fact to claim that climate scientists were wrong in their previous assessments. The instructor clarified that this reasoning misunderstands the nature of scientific progress. The point is not that previous scientists were wrong, but rather that scientists did not know what was going to happen, and they are continuously improving their understanding. This is precisely how scientific knowledge advances: through gradual refinement and improved parameter estimates based on accumulating evidence.

Step Three: Temperature

Climate Sensitivity and Temperature Trajectories

Step three of the integrated assessment model uses the climate sensitivity parameter, which addresses a fundamental question: how much does global temperature increase in response to a doubling of atmospheric CO₂? This is a critical variable in the model. Using this climate sensitivity parameter, researchers project the different trajectories that global temperatures would follow under different emissions scenarios. The plots showing these temperature trajectories replace the concentration plots from the previous step, and now visualize the actual temperature increase expected to occur.

Temperature Projections Under Different Scenarios

Under the fossil fuel-intensive RCP 8.5 scenario, the models project that by the year 2100, global temperatures would be approaching five degrees Celsius above pre-industrial levels. The instructor noted that this temperature increase becomes even more alarming when expressed in Fahrenheit. In contrast, more optimistic scenarios such as RCP 1.9 and RCP 2.6 actually show slight decreases in temperature, with peak temperatures occurring sometime in the second half of the century and then declining somewhat thereafter.

The Business-as-Usual Scenario

One particular temperature scenario, SSP2-4.5, has become widely used and cited in climate policy discussions. This scenario has become important because it represents what would happen if humanity pursued a business-as-usual policy approach—that is, what if no additional climate policies beyond those already implemented were put into place. This scenario assumes that all climate policy gains achieved to date would be maintained but that no additional policies would be adopted. The instructor explained that this scenario is valuable because it is less pessimistic than the uncontrolled warming scenarios, given that significant climate policies have already been implemented, yet it is more realistic than assuming continued rapid emissions growth. The instructor observed that policy discussions typically focus on comparing future scenarios to this business-as-usual baseline rather than comparing different mitigation scenarios to the absolute worst-case scenario. This business-as-usual framing is where much of the actual policy action and debate occurs.

Nordhaus and Current Climate Modeling

The Nature paper being discussed presents the full range of temperature possibilities that might occur, with the central estimate being 3.4 degrees Celsius of warming. The instructor noted that this central estimate is essentially aligned with what the Nordhaus DICE model projects. This alignment suggests that the Nordhaus model may be reasonably accurate in projecting what would happen if humanity did not implement additional climate policies beyond those already committed to.

Step Four: Damages

From Biophysical to Economic Models

Up to this point, the integrated assessment model has linked together a series of biophysical models: emissions lead to concentrations, concentrations lead to temperatures. Now the model introduces an entirely new category of modeling: the damages model, which translates physical climate changes into economic losses. The instructor noted that in Nordhaus’s original DICE model, the damages function was remarkably simple. It was represented simply as a percentage loss in GDP as a function of temperature, specifically using the functional form of temperature plus temperature squared times some coefficients.

However, the instructor explained, substantial research since Nordhaus’s original work has produced far more detailed damage models. Rather than a single simple relationship between temperature and economic loss, contemporary models recognize that there are many different pathways through which climate change creates economic damages. The most recent graph in the Nature paper addresses this question: for any given increase in global temperature, what is the corresponding change in global mean sea level compared to 1900, and what are many other possible damage pathways?

Multiple Damage Pathways

The critical point emerging from this expanded approach to damages, compared to the Nordhaus DICE model, is the incorporation of many different damage pathways that collectively capture the full linkage between environmental changes and economic consequences. One important example the instructor provided is that temperature alone may be an insufficient proxy for climate damages. For instance, climate change is expected to produce not only increases in average temperature but also increases in precipitation and, importantly, decreases in climate stability. As the climate becomes less stable, extreme and unusual weather events become more likely. These extreme events—intense rainstorms, severe droughts, unprecedented heat waves—create sudden, localized losses that would not show up in simple average temperature data. Understanding which specific geographic locations are most vulnerable to increases in flooding or other extreme events becomes critical to understanding the true economic impacts on the economy.

Insurance and Risk Assessment

The instructor used the insurance industry as a concrete illustration of how climate damages are manifested in real economic markets. He noted that he owned a home with insurance and observed that obtaining homeowner’s insurance has become increasingly difficult in recent years. Insurance companies are withdrawing from whole regions of the United States where insurance markets are collapsing. While insurance companies are not primarily motivated by climate change concerns, they are very good at accurately assessing the risk that the assets they insure will suffer losses. This is essential to the functioning of the insurance business model: insurance companies must make accurate projections about how many claims they will pay and at what cost in order to set premiums appropriately and identify which customers are higher risk and which are lower risk.

The instructor noted that substantial economic opportunities exist for those who understand how climate change affects the insurance industry’s profitability. Insurance companies are developing highly detailed damage functions that examine factors such as where critical infrastructure and valuable real estate are located in regions where damaging storms historically occurred once every thousand years but are now projected to occur every hundred years or even every ten years. This understanding directly affects insurance company profitability by allowing them to calculate expected costs and set premiums appropriately. The instructor identified this as an emerging contentious political issue, noting that many insurance companies are simply withdrawing from certain high-risk areas entirely. Obtaining homeowner’s insurance in hurricane-prone regions has become quite difficult, with relatively few insurance companies willing to offer coverage independently.

Spatial Distribution of Climate Damages

The Nature paper analyzed by the instructor illustrated the translation of climate changes into specific damage pathways very clearly. The paper presents two main types of plots. The first shows temperature trajectories over an extended time horizon extending to the year 2200, with simulations that wiggle somewhat around the central estimate. The second plot, for each different RCP emissions scenario, displays a distribution of the expected costs to the economy. For the RCP 8.5 high-warming scenario, the frightening climate scenario, the estimate is that a risk premium of thirty-two trillion dollars would be added as a cost to the economy. The instructor clarified that this represents not simply direct damages but the additional costs that insurance companies would need to charge as a risk premium to cover expected damages. This represents a real cost that would be borne by society.

The nature of the damage distribution differs under each different climate scenario, but the figure powerfully illustrates why climate mitigation is economically valuable. As society reduces emissions and follows lower-temperature pathways, the distribution of expected damages shifts progressively to the left, toward lower dollar figures. For instance, under more optimistic scenarios, total damages might be three trillion dollars. While three trillion dollars still represents an enormous amount of money, those damages would be substantially lower than under the high-warming scenario, and the crucial point is that even after discounting, sufficient investments in climate mitigation could be economically justified if the social cost of carbon is sufficiently high. From an economist’s perspective, the key insight is that reducing emissions shifts the damage distribution leftward, reducing expected future economic losses.

Breakdown of Damage by Economic Sector

The Nature paper examined by the instructor included detailed analysis of the central estimates and distributions of which sectors of the economy would experience the most damage from climate change. A common public assumption is that the energy sector would experience the largest damages, given the need to transition away from coal-fired power plants and fossil fuel-based energy systems. The paper does account for these energy sector effects, but the damages to energy represent a relatively modest contribution to the total social cost of carbon of only nine dollars per ton of CO₂.

The two largest contributors to the projected social cost of carbon are agriculture and mortality, each representing substantial but highly uncertain damage estimates. Agriculture is estimated to contribute eighty-four dollars per ton of CO₂ to the social cost of carbon, but this estimate comes with a huge distribution of uncertainty. Some models in the plausible range even suggest that agriculture might actually benefit from climate change in certain scenarios.

Geographic and Geopolitical Considerations

The instructor interjected a side note about agriculture and climate winners. He asked whether students could think of any countries that possess large tracts of currently unproductive land. He suggested Canada and Russia, noting that Russia’s relatively low engagement with climate change policy may partly be explained by the fact that Russia stands to benefit substantially from warming. Much of Siberia has been thought to potentially become productive agricultural land if temperatures increase. However, the instructor added important recent qualifications to this assessment. The most recent studies suggest that the potential agricultural benefits to Russia have been greatly overestimated. Siberia currently lacks the soil necessary for high productivity. The region has very thin topsoil layers, which has developed because Siberia has been forested for an extremely long time, unlike the American Midwest, which was prairie for a prolonged period. Prairie land creates rich, deep topsoil through different ecological processes than forest land. Therefore, Siberia might become somewhat more productive initially, but this productivity gain would likely be much less substantial and more limited in duration than initially hoped.

Mortality and Climate Damages

Mortality represents the other major component of projected climate damages. This mortality comes from many different sources. Direct mortality from severe storms is one component, but the health impacts extend well beyond storm-related deaths. Heat waves and other extreme weather events contribute substantially to excess mortality. The relationship between elevated temperatures and mortality is particularly important when considering how climate damages are distributed across different populations.

Inequitable Distribution of Climate Damages

Climate change is not only damaging; it is also unjust. This injustice is particularly evident in examining the distribution of climate-related mortality burdens. The Nature paper presented an analysis of the distribution of increased deaths per hundred thousand people across different income deciles at the global level. This analysis revealed that while climate damages are borne by all income groups, they are not borne equally. The distribution shows a notable spike in the mortality burden falling on the second global income decile—that is, the portion of the global population that is poorer than eighty percent of the global population. This group (and the very poorest global decile) bears a disproportionately large share of climate-related mortality.

The instructor questioned why this pattern might emerge. One possible explanation is that many wealthy people also live on coastal areas and thus face exposure to sea-level rise and storm surge. However, even accounting for this factor, there remains a clear inequitable component to how climate damages are distributed. The poorest segments of the global population bear a larger burden of climate-related mortality than their proportional share of global population.

Adaptation Costs and Inequality

Another important dimension of climate inequality emerges when considering adaptation costs. The figure presented in the Nature paper shows how much different income deciles would need to spend to adapt to climate change and avoid the projected damages. This represents a different but related aspect of climate justice. Wealthy people possess the financial resources to purchase protection from climate damages. They can afford to install air conditioning systems, purchase homes with climate protection features, build seawalls, or invest in climate-resistant construction. The data show that wealthier deciles spend more on adaptation, which is advantageous for them but raises profound ethical questions. People in lower-income brackets would presumably wish to have access to the same adaptation options, but they simply lack the financial resources to secure equivalent protection. This difference in adaptive capacity based on wealth represents another important dimension of climate injustice.

Spatial Variation in Climate Damages

Agricultural Damage Functions and Regional Impacts

The instructor presented an interesting recent paper examining the spatial distribution of climate damages across geographic regions. The upper left graph from this paper showed agricultural yield impacts for the United States as a representative example, with the same principles evident in many other regions. The analysis revealed that agricultural yields would decline substantially, particularly in the southern and central regions of the American corn and soy belt. The instructor explained that this pattern occurs because corn and soy crops are particularly exposed to the negative effects of climate change, whereas other crops have different sensitivities.

The instructor pointed out that wheat is grown more extensively in the northern regions of the agricultural belt. This crop distribution pattern matters because wheat is one of the crops that may actually benefit from climate change in many scenarios. Wheat is limited not by excessive heat or water stress but rather by the insufficient number of warm growing days. In a warmer climate with an extended growing season, wheat productivity could increase. Therefore, the crops grown in the north might experience benefits, while those grown in the south face damages. However, the instructor emphasized that the vast majority of current agricultural economic value in the United States is concentrated in the southern and central regions, which are precisely where projected damages are largest. This geographic mismatch between where production is currently concentrated and where benefits would accrue represents a serious economic challenge.

Multiple Damage Functions and Economic Impact

The instructor explained that all of these various graphs represent different components of the overall damage function. Collectively, they describe how climate change translates into economic damages for each degree of temperature increase. The damage functions encompass multiple pathways. Crop damage is one major pathway, with regional patterns as described above. Mortality damages from heat and other climate-related causes represent another major pathway. Increased energy expenditure is yet another, with interesting geographical patterns.

The instructor noted that energy effects vary by location. In northern Minnesota, for example, people would actually reduce their use of energy for heating—specifically natural gas for radiators and furnaces—even as air conditioning use increases slightly. The net effect for such regions might actually be decreased energy expenditure and thus a regional benefit. However, other regions would experience net increases in energy costs.

Specialized Damage Functions

Beyond these major categories, numerous other damage functions contribute to the overall assessment. Labor damages fall disproportionately on high-risk occupational groups, particularly those working in agriculture or construction, who face direct exposure to heat stress and other climate hazards. Coastal damages represent another substantial category, reflecting how storms and flooding damage buildings and infrastructure in vulnerable regions. These coastal damages can be measured in concrete terms as a percentage of county GDP.

The instructor noted that the paper identified additional damage relationships that, while perhaps less obvious, are empirically significant. One such relationship is the strong correlation between increased temperatures and property crime and violent crime. To illustrate this relationship, the instructor noted that Minnesota has relatively low crime rates overall, but crime rates are particularly low when temperatures are very cold. The instructor suggested several possible explanations for this relationship. One possibility is that criminals themselves dislike cold weather just as ordinary people do, and when it is very cold, the costs of committing crimes—for instance, breaking into garages—increase substantially. Another possibility draws on the observation that heat affects people’s moods, making them more irritable and potentially more prone to aggression and violence. The instructor joked that he himself becomes cranky when it is hot out. While this particular damage relationship might be less obviously speculative than some of the others, the instructor emphasized that it has been measured empirically and included in comprehensive damage assessments.

Aggregate Damage Distribution Across Regions

The final analysis in the paper presents a spatial aggregation of all these various direct damages into a single map showing regional damage distributions. The south and southwest of the United States are projected to experience particularly severe damages across all categories combined. In contrast, northern regions face much lower damages, or in some cases, even experience benefits. The instructor observed that the audience lived in a fortunate geographic location. Minnesota’s location means that it would experience less severe climate damages than many other regions, and would possibly even experience net benefits in some impact categories. Additionally, Minnesota possesses substantial fresh water resources that provide protection against water scarcity—a challenge many other regions face. The instructor noted that he selects Minnesota’s climatic resilience to climate change as an illustrative example when making the case for why Minnesota is an exceptional place to live.

Step Five: Discounting and the Social Cost of Carbon

From Damages to Present Values

The instructor explained that the extensive discussion of damages was warranted because damages represent the foundation for policy-relevant calculations. However, the fifth step in the integrated assessment model process—discounting—had already been discussed at considerable length in previous lectures and therefore could be addressed more briefly at this point. Discounting involves applying the principle that damages occurring in the future are worth less to consider today than damages of equal magnitude occurring in the present. Using the appropriate discount rate, the distribution of damages across time can be aggregated into a present value representing the total expected climate damages.

Discount Rates and Social Cost of Carbon Estimates

The instructor explained the critical relationship between discount rate assumptions and estimates of the social cost of carbon through a concrete example. Using a high discount rate of three percent, the models produced an estimate of the social cost of carbon of eighty dollars per ton of CO₂. This estimate represents the idea that the social cost of carbon is relatively modest because damages occur in the future and are discounted heavily to present value. In contrast, if one uses a lower discount rate of just 1.5 percent, reflecting the view that one should care substantially more about the well-being of future individuals and should therefore discount future damages less, the social cost of carbon increases substantially to three hundred and eight dollars per ton of CO₂. At this higher social cost of carbon value, many investment projects and policies would pass a cost-benefit analysis test that would not pass the test at the lower social cost of carbon value. This dramatic sensitivity of the social cost of carbon to the discount rate selection underscores the profound importance of the discount rate assumption to policy conclusions.

The Web Application for Model Exploration

The instructor mentioned that he had created a web application intended to demonstrate the DICE model in an interactive format accessible through a web browser. He had hoped this application would allow users to explore how changes in various assumptions affected model outputs. However, the instructor acknowledged that the web application had not yet passed his scientific validity tests, so he was not yet prepared to share it publicly. Nevertheless, the instructor believed it would be valuable to walk through what the web application would look like and intended to continue refining the application until the numbers aligned properly with well-established results.

Policy Relevance of Integrated Assessment Models

The instructor emphasized that models like those created by economists such as Nordhaus and Stern possess significant value because they allow exploration of different policy scenarios and assessment of which policies perform best economically. Rather than discussing the five-step integrated assessment pathway only in abstract conceptual terms, the instructor proposed that the group actually work through computational examples to see concretely what can be said on the basis of these models. While skipping some steps, the fundamental calculation involves determining GDP damages with and without climate change.

Computational Exploration of Integrated Assessment Models

GDP and Climate Damages

The instructor explained the first computation step: calculating GDP damages under two scenarios: one highly optimistic but false world in which climate damages do not occur, and a realistic world in which actual climate damages are accounted for. Under the scenario without damages, GDP grows at a particular rate. When accounting for actual climate damages, this growth rate is reduced. The instructor noted that this figure obviously depends on numerous quantitative relationships included in the full model. Once fully developed, the web application would allow users to adjust discount rates or the projected cost of solar energy and observe how these changes affected the GDP comparison. The fundamental insight is that the difference between these two GDP trajectories—the world without climate damages and the world with climate damages—represents an enormous amount of wealth, measured in trillions of dollars, that could theoretically be preserved or recovered through better climate policy.

Why These Differences Occur

The source of these large GDP differences becomes evident when examining how damages change over time as temperatures increase. As global temperature rises, damages grow as a percentage of total GDP. The instructor presented a figure showing this relationship, with damages projected to cap out at something in excess of fifteen to seventeen percent of global GDP by the year 2200. This implies that in the absence of climate mitigation, economic output would shrink by fifteen to seventeen percent or more in the long run due to climate damages.

The Social Cost of Carbon and Cost-Benefit Analysis

The social cost of carbon figure can be computed from this damage analysis. The social cost of carbon represents the net present value of damages from one additional ton of emissions. It functions as a benchmark for policy analysis: any emission reduction policy that costs less than the social cost of carbon value to implement would be economically justified, because it would reduce damages at a cost less than the damages themselves. This principle provides the foundation for cost-benefit analysis of climate policies. Policymakers should pursue any policies with costs lower than the social cost of carbon, because such policies represent economically efficient investments.

Dynamic Policy Trajectories

The critical output from a full integrated assessment model is a time path of the social cost of carbon across multiple decades or centuries. According to the model presented, the current social cost of carbon is approximately ninety-three dollars per ton of CO₂ emitted. This value increases over time, primarily because the global economy grows larger and is therefore more valuable, meaning that damages to a larger economy represent larger dollar figures. Eventually, the social cost of carbon trajectory falls to zero. The reason for this decline is that the discount rate takes effect over very long time horizons, causing damage values far in the future to approach zero in present value terms.

Optimal Carbon Policy

The practical application of the integrated assessment model to policy is straightforward in principle, the instructor explained, following standard government methodology that has been mandated since the Reagan administration: conduct a cost-benefit analysis of environmental policy. The question addressed is whether there exists a policy intervention to mitigate climate damages, and if so, whether the benefits of reduced damages exceed the costs of the policy. The instructor illustrated this with a graph comparing GDP trajectories. The blue line represents GDP without any climate damages, while the red line represents GDP with climate damages. In a scenario where optimal policy is implemented, a gap would emerge between these two lines representing the damages that still occur despite mitigation efforts. However, the optimal policy would minimize the total area between these two curves—in other words, minimize the total present value of unmitigated damages and policy costs combined.

Short-Term Costs and Long-Term Gains

The particular optimal carbon policy path emerging from integrated assessment models reveals an important pattern: in the short run, the policy actually makes the economy perform slightly worse because the policy imposes costs (for instance, through carbon taxes) that reduce the resources available for consumption. These short-term losses manifest as a reduction in economic output in the near term. However, by constraining emissions in the near term, the policy prevents the accumulation of CO₂ in the atmosphere and the subsequent severe damages that would otherwise occur in the long run. Therefore, the near-term costs produce substantial long-term benefits by avoiding catastrophic damages. The instructor exaggerated slightly by suggesting that without mitigation, “society would collapse,” but noted that he might not be entirely off base in this characterization.

Optimal Climate Policy and Carbon Pricing

The fundamental implication of this cost-benefit analysis is straightforward: the optimal climate policy would involve setting a carbon tax equal to the social cost of carbon. This straightforward policy would incentivize emission reductions wherever they can be achieved at costs less than the social cost of carbon, which precisely captures what society should want: emission reductions that are cost-effective. However, the instructor noted candidly that economists rarely get to implement such elegant policies. Instead, more conventional cost-benefit analysis is applied project-by-project, examining which specific investments and policies pass a benefit-cost test using the social cost of carbon as the relevant damage valuation. This more applied approach can potentially achieve similar efficiency results to the elegant carbon tax approach.

The Positive Outlook: Falling Energy Costs

Renewable Energy Cost Trajectories

The instructor concluded the technical discussion with a more optimistic perspective than might normally accompany discussion of climate damages. He presented data on the profitability of different energy sources, measured using the levelized cost of energy, which he identified as the correct metric for such analysis. This metric shows that solar and wind energy have become dramatically less expensive over time. The instructor emphasized that he had made this point previously but believed it warranted deeper examination and explanation.

Economics of Renewable vs. Conventional Energy

The source of this cost advantage relates to the different economic structures of renewable and conventional energy sources. Renewable energy sources, particularly solar and wind, require very large upfront capital investment, but once built, require almost no fuel costs. In contrast, conventional energy sources, particularly coal, have relatively lower upfront costs but require continuous fuel purchases. A coal plant must constantly supply itself with coal through mining and transportation operations, generating ongoing costs that accumulate over the system’s lifetime. The true lifecycle cost to society of different energy systems must account for all these expenses. When levelized cost accounting captures all these factors, solar and wind have become increasingly cost-competitive and now offer lower-cost energy than conventional sources.

Climate mitigation as a cost-effective investment

This energy cost development has profound implications for the economics of climate change. The traditional narrative surrounding climate policy has been that society must make a sacrifice today—giving up consumption now—in order to save for the future through reduced climate damages. This narrative is logically sound but depends on the premise that renewable energy is expensive. The cost of renewables is what makes climate mitigation expensive and therefore makes the narrative of short-term sacrifice for long-term gain valid.

However, as renewable energy costs have fallen, the economic calculus has shifted fundamentally. With cheap renewable energy now available, solving climate change has actually become cheaper than continuing on the current pathway. When fossil fuels become more expensive than clean alternatives per unit of energy produced, then the economically optimal policy is to switch to clean energy not because of environmental idealism but because it makes financial sense. The instructor suggested this transformation represents a critical turning point in climate policy: rather than asking what sacrifices are necessary to address climate change, society can now ask how to profit from the transition.

Implications for Long-Term Projections

This cost trajectory has important implications for long-term climate policy projections. While policy changes like EPA regulatory rollbacks are negative developments, their timing may matter substantially. Had these policy changes occurred five years ago, they would have severely impeded renewable energy development, slowing the pace at which solar costs declined and reducing the pace of deployment. Five years later, when renewable energy has already become cost-competitive with fossil fuels, the same policy changes will still increase emissions on the margin relative to what would happen with proactive climate policy, but the increase will be modest compared to the additional emissions that would have resulted if the same policy changes had occurred earlier in the renewable energy cost trajectory.

Technological Progress in Transportation

The instructor illustrated this broader point about technological progress by discussing the electrification of the vehicle fleet. Long-standing predictions about when electric vehicles would begin to significantly penetrate the American vehicle market predicted a slow transition occurring decades into the future. In practice, the adoption of electric vehicles has dramatically outpaced these predictions. The energy and engineering communities have made substantial technological progress in battery technology and power electronics, driving down electric vehicle costs and improving performance while increasing range. The market is responding to these improvements, with rapid acceleration of EV adoption compared to historical predictions.

The instructor concluded this section with a personal note, observing that electric vehicles are also simply more fun to drive. The technological transitions now underway are driven not only by climate and economic imperatives but also by technology that produces tangible consumer benefits.

Remaining Course Content and Conclusion

Final Module on Country Reports

The instructor noted that while he had intended to discuss country reports in this lecture, time had not permitted such discussion. He explained that in the following class, he would conduct a matching exercise to assign each student to a specific country. These country reports would become increasingly important as the course progressed, incorporating the various concepts, tools, and frameworks being introduced throughout the remainder of the semester.

Closing Remarks

The instructor concluded by wishing the students a good weekend and noting that the remaining course content would address scenarios and then transition into other environmental aspects before shifting toward discussing natural capital and ecosystem services. With this overview of the social cost of carbon and the current state of integrated assessment modeling complete, the course was ready to move forward into these new topics. The instructor thanked the students and dismissed the class.

Transcript (Day 1)

Still missing a few people, but why don’t we get started? That sounds great! Alright, well, welcome everybody, let’s get started. We have a special guest first. I’d like to introduce Christina Lundgren. She’s going to take the first few minutes talking about some happenings at the Institute on the Environment.

So I can jump to the next slide. All right, so my name is Christina, as mentioned. I use she, her pronouns, and I work over at INE. I was able to take the Gopher Way to get here, which is very lovely with the cold weather outside. We’re just a little bit up the stairs and over that way. Has anyone been to INE before?

Maybe not? Okay. So it’s got a beautiful atrium space. If you ever want to study in there, it’s great with natural lighting. It used to be like an old space for showing off livestock that’s now an inside part of the building.

Okay, let’s see. Advancing the slide. All right, so IANE—Institute on the Environment—a lot of our focus is on how we include people in the environment. In your class, you’re already thinking about that with the economics aspect of it.

Part of the reason I’m here today is your instructor was kind enough. When we sent out an email about the Sustainability Studies minor, the class that you are currently taking counts as an elective toward the minor. It’s super flexible and super accessible, so even if you’re taking another course that’s environmentally related and it’s not technically listed as an elective, it’s really easy to petition it to count.

It’s really great if you’re interested in the people aspect of the environment. It’s a really helpful minor to meet students from across the university, all different disciplines, and hear how other people view sustainability.

So how does the Institute on the Environment do our work? We have an education team, that’s the team I’m on, activating leaders. We have a research team, so if anyone in this room is ever interested in research, there are often student positions that will be posted. We also work on communications and organizing, and those are student staff positions as well. And then we have a communications team.

I’m guessing all of you are already interested in the environment, but if you’re interested in additional classes, we compile these courses. This slide deck will be shared with you later across the university, open to all degrees. The idea is, if you’re interested in sustainability and wanting to get a different lens on it, these courses will offer that.

So the first course in the minor: has anyone here taken SUS 3003? No? Okay, I see yes. I know you took it. This is a really awesome course. It’s taught every fall and spring semester, and it’s a really great intro course to thinking about how culture and sustainability fit together.

This is the capstone course. So the first one I just shared and this slide—these are the two core courses that you need to take for the minor. The remaining electives, like this course, could be substituted in to count towards the minor. You need to take three total to round out the minor, in addition to 3003 and 4004.

This is the capstone course. You work with community partners in this course on a project. It’s great to add to your resume. In general, all the courses in the minor have some sort of component where you’re doing community-engaged work.

This course is great if anyone is interested in doing an internship in something related to sustainability. The course instructor, Mary Hahnemann, works with many different community partners and compiles a list of host sites every year for internships. So if you want to do an internship this fall or next spring, it’s offered both semesters. Some of the internships are paid to begin with. Others have a scholarship you can apply for. It’s pretty straightforward. As long as you have demonstrated financial need, you’re pretty much guaranteed to get the scholarship to help fund that internship experience.

Questions on this one? Nope? Okay, cool.

All right, so this says present at the Sustainability Symposium, but it should now say attend. This week, on Monday, the presentation submissions closed. But if you want to meet students from all five campuses—Crookston, Morris, Duluth, Twin Cities, and Rochester—we’ll all be gathering at the Institute on the Environment, which I mentioned earlier. It’s really cool. There’s free lunch, and you get to network with local professionals. It’s a really fun experience to be uplifted and see what cool things your peers are working on. Sustainability can be kind of a downer sometimes when you’re looking at what’s happening in the world. So it’s a fun place to be inspired.

Another great event: has anyone heard of environmental justice?

It’s kind of a jargony topic, but awesome—I see head nods. So if you haven’t heard of environmental justice, it’s the intersection of social justice and the environmental movement. We have a really great event. This will be our fourth or fifth year doing it. It’ll be over on West Bank at the Humphrey School of Public Affairs. There will be networking, job and internship tables where you can find out opportunities. You can pack some seeds if you’re planning on having a spring garden. And again, there’s lunch and really great community partners sharing about environmental justice in our local community.

And then we have so many other programs—this isn’t even all of them. These are paid leadership programs, except for the Germany study abroad course. Has anyone participated in Undergraduate Leaders? No? How about the Germany study abroad course? No? Okay. Has anyone participated in Clean Energy Leaders?

Nice, nice. We have a resident INE person who’s done a lot of our programming, which is awesome. When someone does something with INE, they tend to come back for more programming because there are so many great continued opportunities for development.

Undergraduate Leaders has an equity focus, it’s year-long, and it has a $1,600 stipend. The Germany course is super fun—it’s a two-week study abroad program over winter break. So if you can’t take a whole semester to study abroad or want to squeeze in another study abroad experience, this one’s really great. Germany’s really far ahead in renewable energy, so it’s fun to see what they’re doing there and come back and make that transition in Minnesota, because we’ve been copying a lot of what’s been happening in Germany’s process.

Clean Energy Leaders is a year-round program and includes a summer internship with a focus on clean energy.

Everything I’ve shared, even though it says clean energy or something like that, is open to students of all degrees. As I’m sure you all know, we need everybody when it comes to climate change. We need the artists telling the stories, we need the researchers doing the research, and we need every other discipline you could imagine.

And these are a variety of host sites and partners we work with for our internships and community engagement. You can see a mix of nonprofits, government, and corporate. So if you’re curious about any of those avenues, our programming touches these different areas. Highly, highly recommend if you’re looking for resume experience and professional development.

Okay, that’s all I got. Any questions? No questions. Okay, cool. I guess I would just add that I love INE. I came from there and worked there for five years, and it’s a really awesome thing to have just down the hallway.

Well, thanks so much for having me. I have a slide here at the end for the newsletter. If you decide to sign up, it basically shares everything I’ve shared with you, but when it’s relevant for applications. And I have these slides, so I’ll send them around to the class.

Great. Well, thank you so much. I really appreciate your time.

Thank you. All right, so let’s get started, picking up where we left off yesterday. I have updated the slides. We’re going to be on the same slides as we were before.

Okay, I’ve added a slide for Day 2 Agenda. It’s always hard having the lectures line up with the topics. Topics are sometimes longer or shorter than a lecture, and this is my workaround. But what did we do? We talked about the basics of climate change and the sort of irrefutable need to think about this. We also set up the fact that there are different economists who think about this differently, and there is really interesting debate among them.

We then talked about the sort of first and seminal piece of work in this area: the Nordhaus DICE model. We’re going to finish that up today, hence we’re on the same slides. Then we’ll contrast it with a different economist’s report: Sir Nicholas Stern’s Stern Report, which approaches it from the other end of the spectrum. Nordhaus is saying there’s no really big need to worry—a little bit of investment is all that’s needed, and anything beyond that is too much. Whereas Stern uses a lower discount rate and finds that it’s actually very dire, though it goes much deeper than that.

Then we’re going to talk about how these contentions about what’s the optimal amount of climate mitigation depend on a cool set of models that we’ll call integrated assessment models. We started to see the basics of those when we talked about Nordhaus and the idea of maximizing utility subject to the choice of how much to abate.

Since he came out with that basic approach, it’s blossomed into a huge area of research with many different integrated assessment models. Stern uses one too, but it’s actually the trajectory where climate and economic science is going.

And we’ll conclude by showing how IAMs connect to a really important policy-relevant number called the social cost of carbon. We’ll talk about the literature on that and how it links to policy.

Before I dive into content, any questions on timing or the two different assignments we have outstanding? There’s the weekly questions due Friday and the assignment due Monday. Any other questions? We’re good?

All right, let’s dive in.

So where we left off, I concluded with the punchline of DICE, but we didn’t dive into the results in more detail. The punchline is that in one of the more recent versions of the model with a 3% discount rate, it would be optimal to abate carbon emissions or other greenhouse gases if it was cheaper than $51 per ton of CO2 emitted. This is the social cost of carbon, and we’ll formalize it more broadly in modern integrated assessment models. But before we do that, I want to talk through the basic argumentation that Nordhaus used.

Essentially, his approach was to optimize and find the optimal amount of climate mitigation to maximize our happiness. It’s just like standard economic growth. We try to maximize our growth. Here, we’re going to do the exact same thing, but with the factor that there will be potential damages to the economy. We have the option to spend money on abating emissions, and maybe that will be beneficial in terms of having a positive return on investment in those emissions reduction programs.

But the basic tension is that the economy can take steps to slow emissions, but that’s going to be expensive. All that expense spent on abating emissions will come at the cost of less consumption today.

This fundamentally relates to the concept of intergenerational equity in the discount rate. Present generations would be asked to reduce their consumption in return for lower damages and thereby higher consumption in the future.

Basically, all the DICE model does is identify the optimal path for consumption, just like we did with Ramsey, but with these extra bits of information.

So with that context, what are the actual results?

DICE says that the optimal path will have a 2.8% loss in GDP. This is not saying what would happen if we just didn’t do anything. This is with optimal mitigation. Nordhaus is saying that we will lose GDP, but this is the least amount of GDP we could lose. The corresponding temperature is 3.5 degrees centigrade warming. Both of these are by essentially 2100, though he centers on 2095 in this particular publication.

This is a really old journal article, but it’s the stuff that led him to the Nobel Prize. We see his basic results in this table showing the impact of alternative policies and how they affect the discounted consumption—the net present value of consumption going way out into the future.

He’s looking at a business-as-usual plan where we don’t do anything at all: no controls, no emissions abatement. The discounted value of utility is measured in trillions of dollars. If you don’t do anything at all on climate policy, you get $731.69 trillion of value. That’s a lot.

But the critical value is how that number compares with our different policy options.

The first policy he presents is the optimal policy. When we implement that, we get $739 trillion of value when discounted to the present. It’s a little bit better—$9 trillion more. This is measured as $271 billion of extra value. That is good, but it’s remarkably small. This is the basic finding in a lot of his results: even the very best policy just isn’t going to do very much.

He looks at a few other ones, like the cost of delaying ten years. It doesn’t cost that much. It goes down just a little bit, but that might be useful because we can spend more time gathering science. There’s not much cost to that approach.

But the real punchline is, what if we implemented drastic measures that were being debated at the time to reduce emissions? At the time, the idea was to reduce emissions by 20%. That might sound antiquated now because we’re talking about getting to net zero, literally zero. So he was considering a very weak policy by today’s standards. But a 20% reduction is what he plugged into the maximization.

This has a massive impact. This lowers the value of GDP going forward by $10 trillion. In other words, optimizing gives a little bit of benefit, but doing the existing climate policy that was being debated would be massively costly. His conclusion was that this would be a very dumb idea.

This was such a controversial finding because it was hated from both directions. People who didn’t like to admit climate change happened didn’t like that he admitted it happened. But people who did care about climate change didn’t like that he admitted it happened because he was arguing that nothing should be done, or very little should be done.

This was a very controversial first foray into systematically combining economics and climate.

His results change so much over time. Regarding reducing emissions by 20%, the short answer is integrated assessment models. That’s where we’ll go for the rest of the lecture. To give you a sense of where we’re going, you plug different things into these models that say how much mitigation will reduce climate change and reduce damages, but also how much that costs. If it costs a lot, that means we’re not spending that money on luxury goods. And that’s why the valuation is so much lower—we’re not consuming as much.

He shows that greenhouse gases should be cut a little. He wasn’t a climate denier. He was saying climate change is happening and we should do something about it, just not very much.

His main point was that this should be done with a very small carbon tax. The optimal carbon tax slowly ramps up to between $12 and $50 per ton. This is really different than the tax that would get a 20% reduction in emissions. That has massive costs, measured in dollars per ton of carbon. His point is that these high carbon taxes would hurt the economy. High taxes mean we pay the government for each ton of carbon we emit, and this money goes toward preventing emissions instead of being used for consumption. And that’s a bad thing economically.

His work has evolved extensively. This was the old work, which is why I wanted to show the seminal work. The latest release was the 2023 version of the DICE model. This shows the key results. The green line in the middle represents the optimal temperature policy pathway.

If you take into account his framing of the DICE model and try to make all of humanity as happy as possible, we should be reducing emissions. But what do we see? The equilibrium value in 2100 is about 3.1 degrees centigrade warmer than today. It’s not much different than his very initial estimates. He’s changed a lot in how it’s computed. One thing is he’s admitted it could be much worse. The difference between the optimal policy and doing nothing at all is bigger now because some of the extreme risks of climate change have become more evident.

But it’s still not very much compared to what many policymakers are thinking. This is illustrated by the 2-degree target. If we were to implement something that kept us below 2 degrees centigrade instead of his optimal policy, it would look like this. You can see his preferred option still has much more warming.

So what’s the criticism of this? I say this is controversial, and here’s a good way to think about it. Catastrophe is coming! Oh no, this is going to hurt the economy! Well, as Keynes famously said, in the long run we are all dead. If we have a catastrophe, it doesn’t matter much if we lose out on a little bit of luxury consumption today if the whole system collapses.

Of course, if you have a really high discount rate and you don’t think this will happen until after you’re dead, maybe it does make sense to let this catastrophe happen and worry about the economy instead. It’s like worrying about an incoming asteroid.

As a side note, I love this modern AI because there’s a copyrighted comic I was basing this on that I couldn’t use. So what did I do? I typed into ChatGPT: make an image about dinosaurs talking about how the economy would be wrecked from a meteor. So I got this. The basic idea isn’t quite as good as the original, but it’s not copyrighted.

That’s the joking criticism, but a lot of debate has been had on this. Let’s talk about the actual criticisms of DICE.

In upcoming problem sets, exams, and quizzes, we’re going to shift more toward arguing intelligently about issues like these rather than just doing supply and demand and solving basic algebra. So I’m foreshadowing that if I were to write down some criticisms of DICE, it wouldn’t be off-target for me to ask a question like: what are the two most persuasive arguments against DICE in your perspective? Do keep these in mind.

The first one is the discount rate. This has really been the primary criticism. We shouldn’t care about the asteroid if we’re going to die before it hits us. In other words, we don’t value future generations enough. He had such a high discount rate that it mattered almost nothing. So the discount rate is too high.

And who is he to choose how much we should care about future generations? That’s his opinion. There’s much more behind the logic of how he gets to it, but it’s fundamentally a value choice. I care a lot about my kids. There’s literature out there trying to estimate behaviorally how much people care about future generations’ well-being, and one of the key events in a person’s life that increases how much they care about future generations is having future generations. And then the next one is their future generation having a future generation. These drastically reduce the discount rate because now you’re getting utility from knowing your grandkids are going to be alive rather than scavenging for meat in an apocalyptic landscape.

So this is a moral issue around sensitivity to the discount rate.

The second criticism is damage functions. You might remember that we had a damage function that was essentially T plus T squared. He had some coefficients to say how much temperature and temperature squared mattered, but this was really simplistic. There are dozens, hundreds, thousands of ways the economy will be affected by climate change. And many of them can’t even be expressed just as temperature. Sea level rise is a function of temperature, but it matters drastically whether you implement seawalls, have mangroves protecting the coast, and so on. There’s a whole area showing that these damage functions were artificially low, incorrectly low, or just wrong.

One key issue is that they underestimated what are called tail risks. Tail risk is something that is not in the center of a distribution—the tails of the distribution. With a normal distribution, extreme events are really improbable. But if the distribution had fatter tails, suddenly there are non-zero values that continue for quite a ways. Many people worry that climate change has fat tails because we’ve only seen the world around current conditions. We don’t know what it looks like out there. If it is like that, everything in his damage function could be wrong.

Related to this is the absence of tipping points. The damage function just stays smooth—there’s no feedback. If it gets warmer, the permafrost might melt. If the permafrost melts, all the methane trapped there might start releasing, accelerating damages and warming. This is terrifying because we’ve never seen it happen. We’ve never been at temperatures where permafrost melts, so we simply don’t have observations or science about that, but we have enough science to know it might matter.

Then there are a few other technical criticisms. It’s a super simple model. The biggest challenge is it’s a single global agent, sort of like the Ramsey model. There’s no trade, no regional inequity, no distributional effects. What if the optimal policy is that 90% of countries should stop all agriculture and starve so that one country representing 10% of the population becomes really happy? That would be unfair, but the Nordhaus model doesn’t say anything about that.

Finally, there’s technology assumptions. Technology matters massively. In his model, there’s no endogenous way we learn how to mitigate carbon better or even get better air conditioners. Critically, how might technology evolve in making it cheaper to reduce emissions? This has been critical because when he wrote this, coal-fired power plants were producing at roughly 6 cents per kilowatt-hour, and solar was something like 28 cents per kilowatt-hour. The idea that we would have to harm the economy to emit less matched the technology of the time.

But you know what? Coal is still about 6 cents per kilowatt-hour. What’s solar now? Anybody know? Way less than coal. It’s like 3 cents, and it drastically depends on location. There are other challenges like distribution and getting our power grid up to spec to handle this, so it’s not trivial. But when the news is saying we’re going to deregulate coal-fired power plants, I don’t care that much because even the most selfish person would be pretty dumb to invest in coal right now. Even if you had a guarantee that there would never be another negative law about coal, it would be like investing in whale oil for lights. You can burn it. It was used extensively in the past for illuminating cities, but it’s probably not a good investment, even if you hate the environment.

Solar is just cheaper now, and wind is similar. There are going to be a lot of changes. The way this affects the DICE model is he assumed really expensive mitigation because it was expensive at the time. But what happens if mitigation is cheaper than not mitigating? It’s kind of a paradox. It should happen automatically, right? Nonetheless, this has been a huge criticism.

Okay, so that’s the Nordhaus review. On my right-hand side, but your left-hand side, this is where Nordhaus is on the spectrum of political debate, all the way over here.

We’re now going to talk about the other end: the Stern Review. And I would say that Nordhaus is farther to the right than Stern is to the left. The Stern Review is definitely more to the left of consensus, but it definitely has much more emphasis on intergenerational equity and considers many real criticisms of Nordhaus.

What is this? The Stern Review was commissioned by Her Majesty’s Treasury in the UK under Chancellor Gordon Brown. This Stern guy is actually Sir Nicholas Stern because he is knighted. That’s how much the British care about him. This is a government-sponsored report on the economics of climate change.

As a side note, we’re going to be talking about the Das Gupta Review. I’ve mentioned that a couple of times. Similar story—it was commissioned by Her Majesty’s Treasury to do a report. We actually got asked to contribute to it, and I got flown out to Her Majesty’s Treasury, one of the coolest trips I’ve ever done. We went into this old castle-like building, formerly the horse grounds where different sovereigns and kings and queens of the British Empire lived. It’s pretty cool, and it turned out to be where a lot of movies had been filmed. Anyway, same deal—they flew me out there, I wrote an article.

But Stern’s basic recommendation was to spend 1% of global GDP per year on mitigation. That was huge—hugely more than the Nordhaus approach suggested.

Let me go to the key findings. The biggest key finding of the Stern Review is that inaction is really bad. As economists, we measure that based on loss of GDP. Inaction would be 5 to 20% of GDP lost by 2100. This is in contrast to the 2.8% loss in the Nordhaus work.

The other big part is that although 1% of global GDP is a lot, and he’s saying a higher amount should be invested, it was also found that this is much more effective. Basically, for the Nordhaus assumptions, to get this level of emissions reduction, it would have been more expensive. Here, Stern is saying invest 1%, and that will have a huge impact.

Other points are that this is urgent and very necessary to consider equity, but the key findings are that there’s much more at stake and the efficacy of investing in it is really quite high.

Okay, I’ve been building to this point. Why do they have such different opinions on the optimal amount of GDP to spend?

It comes down to the discount rate. That’s why we spent so much time on it: here’s a basic comparison of these reports. The Nordhaus DICE model has a pure rate of time preference of 1.5%, and the Stern Review has 0.1%, basically saying the future matters a lot to us.

This led to much higher levels of damages that Stern thinks would be optimal to abate. Stern in 2006 was saying there’s $85 of damage, while Nordhaus said that there’s only $30 of damage. Because all those damages are going to happen in the future, and when we discount them, they don’t matter as much.

The key difference comes down to that discount rate and what that implies about how much we care about our grandchildren.

It’s not just the discount rate, though. It also includes these other things. You can see the contributions. 5% is the most conservative estimate of damages. But he also, correctly I think, identifies other types of damages like non-market damages—things that aren’t just saying that industry produces less, but also that we die or we suffer from consequences that don’t filter through production functions.

He also considers equity and this idea of fat tails—including some sort of risk of catastrophe. These all contribute to much higher levels of losses.

I kind of want to get to the integrated assessment model part, so I’m going to skip the rest of this. I won’t test on any of it.

His influence was essentially transforming the policy landscape into really caring about climate change. This was behind the Kyoto Protocol, which was the first real substantial policy investment, and then finally the Paris Accord, which is what is most binding today.

To summarize, he argued that the benefits of strong, early action on climate change outweigh the costs of not acting. The fundamental framework is the same—utility maximization, cost-benefit analysis—but done better. This is a challenging debate, but very, very important.

Okay, so that’s the space of different opinions and different ways to use models.

But I want to dive into more detail about how they come up with these numbers. For that, we’re going to switch to the next slide deck.

So we were on Chapter 15 on climate change. Now we’re switching to the next one, which talks about the social cost of carbon and how it’s calculated in different integrated assessment models.

So where are we at? We have this spectrum of economists arguing for different policies. But what should we do? They’re just saying what their optimal choice is, but we have a range of optimal choices. On one hand, we shouldn’t do anything. On the other hand, we should do a whole lot right now.

Truman once said he really wants a one-handed economist because they always have this problem. They give you two hands. This is true. We have two hands here. How are we going to make an informed decision? How can we come up with the best estimate for the best policy? This will be through using integrated assessment models to establish the correct social cost of carbon.

Okay, so what is the social cost of carbon? I’ve mentioned it a few times, but I wanted to get the debate out there first. But now we’re ready to define it.

The social cost of carbon, or SCC, is the monetary damages of one additional ton of carbon. It sounds simple, but there’s a lot of complexity behind calculating this. We’re going to see that it’s measured in the present. This is going to depend on calculating the net present value.

The choice of how much carbon we should have is being made right now, or we can think about the same calculation in future time steps, but it’s a decision right now. The benefits give us a flow of value over time. This is not so different from any capital investment. A factory owner considering upgrading their machinery faces a big upfront cost they have to consider in the present. But the point is it gives a flow of value out into the future. So they always calculate the net present value. That’s basic accounting. And that’s what we’re going to do.

The social cost of carbon is complex because it involves linking many different steps and many different models. With the production example, a firm understands what they’re producing and their machine, so it’s kind of simple. But here, we’re integrating many different parts from climate, the economy, carbon cycling, damages, and these all have different models. We’re going to need to integrate these different models into an integrated assessment model.

So first, what is an integrated assessment model?

It’s going to combine models on each of five steps, and it’s important to understand each of these. We’re trying to figure out the correct social cost of carbon, so we need to consider each one.

The first one is what’s going to happen with baseline emissions. Given assumptions like population growth or technology, how much emissions will make it into the atmosphere? This isn’t easy to figure out because if we all switch to solar, we’re going to be a lot less. Climate scientists and economists make reasonable projections of things like economic growth, population, and technology, but also harder things like the preferences of consumers for meat versus vegetables. That matters because it determines the equilibrium level of meat production and the methane cows produce. We have different emissions trajectories based on different assumptions of what type of economic activity is going to happen.

The second step in an integrated assessment model is the emissions-to-concentrations step. Emitting is measuring it at the smokestack, essentially—how much CO2 and other greenhouse gases get emitted. But concentrations are actually more complicated because of biophysical detail. The carbon model shows what happens if you emit a bunch of carbon. Higher temperature causes more plants to grow, like more algae in the ocean. This algae grows off carbon, so some of the emitted carbon gets photosynthesized by plants and removed as a problem. Some gets dissolved into the ocean, and of that, some sinks to the floor and becomes mineralized—trapped forever. But some gets emitted back out. Going from the relatively easy-to-observe thing—emissions—to concentrations requires detailed modeling.

The punchline here is that the scenarios community has defined the Representative Concentration Pathways, or RCPs. We’ve seen this before, and we’re going to use it again for the rest of the course. Essentially, RCPs combine the assumptions that go into emissions with a model of what happens to those emissions to get the concentration we’re going to have.

The third step is temperature. From concentration, it’s not obvious what happens to temperature. This is driven by a very controversial number called climate sensitivity. Climate sensitivity measures what happens when CO2 concentrations double—it leads to what percentage increase in temperature. This is one of the very controversial numbers. If you have a climate sensitivity parameter that’s low, it means we can do lots of emissions and it won’t really matter because Earth’s system isn’t very sensitive to emissions and we won’t get really hot. If it’s high, each ton of CO2 has a large effect—maybe a 2 to 1 ratio. Basically, how much will temperature go up from an increase in CO2 concentrations?

There’s a whole set of models there.

The fourth step is damages. How does temperature change damage the economy? This is typically expressed as a fraction of GDP lost. Nordhaus used T plus T squared, but that’s hilariously simple. Damages are probably going to depend on many different damage pathways. But it accounts for the many different things that matter beyond just temperature. Things like labor productivity matter. It’s really hard to work in an agricultural field when it’s super hot. I did corn detasseling as a youth. They paid minimum wage to go pull the tassels off corn in a field. You need a ton of temporary staff, and my high school class all did this. It felt like a lot of money to a sixteen-year-old, but it was awful because you’re just standing out in the heat. Well, that gets much more painful if it’s hotter out.

So damages are important, and we’ll talk more about that.

The final step is what I’ve been emphasizing: intergenerational consideration—how we care about future value. How do we discount that stream of benefits back to the present?

From here, we’re going to take this basic pathway and look at the data. We’re going to rely on a very new article. I’m going to assign this as a reading for next class. You should read Renert et al. We’re going to walk through the logic of computing this and what the best evidence is on this thing. I’ll send out an email to remind you, so don’t worry about that.

We’re going to step through these, and we’ll conclude with what is the correct estimate of the total value of damages from one extra ton of carbon. This is important because it’s essentially the benefit and cost ratio that we care about. We’re trying to figure out the right amount of climate change. It’s not zero because that’s super expensive and it’s not going to happen anyway, but it’s probably also not this extreme of having 3.5 degrees centigrade warming.

Okay, so we’ll leave it there today. I’ll email out about that reading. Remember to do your assignment and your weekly questions. I should have said this in the assignment, but Canvas defaulted to the questions being due by the end of the day on Friday. I won’t change this because that would be unfair. It’s still due then, but if you can do it before class on Friday, that might inform our discussions better. So aim for that informally.

Any questions? We good?

Thanks, everybody!

Transcript (Day 2)

All right, well, let’s get started, everybody. Welcome to Day 2 of our class, where we’ll be focusing on the social cost of carbon and climate integrated assessment models.

This will be the same set of slides that we’ve been working from that we just started on last class, but I’ve updated them, so do refresh them in your browsers if you’re following along.

But the first thing I wanted to do was hand back midterms. So, great job. I put an announcement out in Canvas. You should be able to see your scores. I was quite pleased with how everybody did. You know, as I said in some of the very first classes, I have these exams intentionally be difficult, but then I’ll curve them up so that everybody gets a more realistic grade. This is because if they’re too easy, everybody gets 100%, then I can’t tell who knows what and doesn’t know what. And so you’ll see both your raw score as well as the actual score, the higher curved score. So why don’t we just pass these out?

[Students called to collect exams and answer key]

Good job. It’s always a risk handing these out at the beginning of class, because what everybody’s doing right now is looking through them. So feel free to take a couple of minutes.

If you have any questions about grading, remember to go ahead and email Ryan McWay. His email is on our website. He’s the one that did the grading, and so he’ll be able to answer any specific questions about that.

Okay, and as for the content of today, we’re gonna review first the schedule out till the end of class. It’s a little bit preliminary, but we will fill in and possibly make changes as we go, but this basic approach will hold. I think this will be useful because it’ll give you a sense of how the class is pivoting towards more applied questions and using tools.

Then we’ll dive back into our content, picking up where we left off: the five-step process for integrated assessment models. But we’ll really dive into the state of the art of using it. You know, we set up this spectrum of economists who range from climate change should hardly be dealt with at all, to climate change as an existential crisis. What is the best estimate, given all the knowledge on what the social cost of carbon should be, and what are the corresponding policies. That’s what we’ll focus on today.

Okay, so, taking a look at the schedule. Here’s where we are: March. We just had spring break, of course. And we’re right here on the Social Cost of Carbon and Climate IAMs Part B.

What we’re gonna then switch over to next week is talking about future scenarios. We’ve started to see this already with representative concentration pathways, but there’s a lot more there. And it’s more than just concentration of greenhouse gases in the atmosphere, but also what sort of future do we want to have? What would we like to work towards? We’ll talk about that in terms of the shared socioeconomic pathway scenarios.

One of the things that those define is much more detail about how land use will change, how deforestation or cropland expansion or all these other things might change, and so we’ll talk about that explicitly.

But on Wednesday, we’ll also introduce Geographic Information Systems, GIS. So I’ll send out a reminder, but I’m hoping everybody can bring a laptop. I’m looking at the two people with tablets. Is there any chance you could bring a laptop? That works? Okay, cool. You can try to do GIS on a tablet, it’s a little harder. But we’ll actually talk through that, and if you don’t have a laptop, contact me early, because I can probably set up alternatives.

Once we’ve got the basics of GIS under our belts, we’re gonna transition to natural capital and ecosystem services. And these are all things that are indeed spatially explicit, and most of them actually depend on land use. So that’s why we’ll do it in this order.

Once we’ve introduced them in general, we’ll then talk about the original way that they were used, really, which is: let’s just put a price on nature. We’ll talk about valuation methods, and how economists might say something interesting about how much pollination is worth, or how much a wetland is worth. We’ll talk about the different methods that they use to do that.

But with that tool in hand, we’ll then spend a lot of time diving into specific ecosystem service models. We’ll start with carbon storage, talk about sediment retention, pollination will be a big one we look at, and coming up middle of the week on April 8th, we’ll then have some time to do a discussion and application for how this applies to your country report, which I’ve been indicating all along, but we’ll start to get more details of that.

We’ll spend some time on ecosystem services and uncertainty. We’ll return then to the question of scenarios, and not just look at the shared socioeconomic pathways, but new emerging ones of positive visions. It turns out even the very most optimistic shared socioeconomic pathway scenario is pretty depressing. And so these will be more positive visions of what we might want to have happen.

Then we will transition to talking about the policies, market-based and real-world policies, that will get us, hopefully, towards those visions.

Then we’re gonna have two guest lectures. One from the senior ecologist at NetCap Teams, Colleen Miller. She’ll be talking about biodiversity and how it’s the basis of all of this value. It’s not an ecosystem service in itself, but it is what enables ecosystems to thrive.

And then we’ll have one from Dr. C. Ford Rungi, who taught the climate class. He’ll come in and talk about land as an input to production. He’s a great economist.

Then we’ll return to the very last part of class, which will be focused on connecting all of these things into a combined framework on Earth economy modeling.

On the second to last day of class, we’ll be having student presentations on your country reports. And this is part of the grading for the actual final project. You’ll actually present it. They’ll be very fast. We’re going to try to get through everybody in one day. These will be lightning talks, as we call them, where we’ll combine all your slides onto one slide deck, and you’ll come up here and give just a brief three to five minute blurb on your key points that you found out about your country.

Then I’ll have a concluding lecture on the last day of classes, and then there’s our final exam.

So that’s the arc. We’re within a vision of the end. We’re a little bit past halfway, so it’s still a long ways, but we can see through to the summer. Any questions or clarifications on any of this?

Okay, hopefully it’s pretty clear, but also hopefully you see that a lot of these things are going to be very hands-on. For the GIS one, I’ll actually be having you learn the software. And for the ecosystem service ones, we’ll be using a tool called InVEST, which actually computes those ecosystem service values. So this will give you hands-on exercise, and that will be what goes into your country reports. That’s the very applied nature of being in our applied economics department.

So, okay. Without further ado, let’s pick up where we left off last class. I’m going to really simplify the whole five-step process into a one-liner. We talked about how emissions lead to concentrations, which lead to temperatures, which lead to damages, which lead to discounting. The whole point of this is it gets us the value in dollar terms per extra ton. Specifically, the net present value of the additional damages expressed in terms of one ton of CO2 emitted.

That’s really quick, just to get it up there so we can remember all these different parts, but I want to walk through it now, not conceptually, but instead with an emphasis on what is the current state of the art.

This is all gonna be based on a paper that was recently published in Science, and it is putting ranges, including uncertainty, around each of these parameters if you calculated them through the sort of state-of-the-art model. And so this will give you the most up-to-date understanding of where the economy is at.

The first part, which actually has most of the graphs going into this, is how do we get to emissions? It’s really easy to say what the emissions are for today, and we just could look at it. It’s actually harder than you think. You need to keep track of all the different plants and monitor all of the emissions, and that’s a pretty big task, but at least it’s something you can do. Increasingly, we can do this from space. You can use satellites to keep track of where carbon is being emitted, and also other chemicals like methane. A cool article just came out a day or two ago showing that you can use satellites to pinpoint specific methane-emitting firms. There are little tiny dots on the map where tons and tons of methane is being emitted, and you can actually see the plume of methane blowing downwind from these particular plants.

That’s kind of cool, because with satellites, you can then have novel ways of enforcing methane controls. It’s a lot harder to get away with it by hoping nobody will notice.

Side note: you know we all know this AI boom that’s coming. They’re building tons of data centers. This is also driving methane. They’re installing off-grid power plants, mostly using natural gas, and they locate them right next to their data center so that they don’t even have to plug into the grid and deal with all the challenges that entails. But the downside is this is like the Wild West, a bunch of people building generators. That’s not gonna go well in terms of installing state-of-the-art methane emissions systems. And so we’re seeing those from space, which is hopefully a good way to catch those.

But yeah, so for the emissions part, we need to keep track of a bunch of things, and it’s much harder when you go out into the future. You can’t just observe them with satellites. You need to make projections of all sorts of things. Here the paper highlights three of those.

Number one is: what’s going to happen to population? They have a central estimate, which is the dark blue line. This comes from all sorts of sources, but essentially demographic models. We have things like keeping track of the fertility rate and the death rate of different countries. But of course, there’s huge uncertainty there. We don’t know where the population actually will be, and so that’s represented by these shaded colors around this line. The darker shaded area is roughly two-thirds of the estimates that are going to fall somewhere in this range. And then this goes out to 95% of the estimates we think will be in here. But there’s huge uncertainty. The central estimate is we’re gonna cap out at a little bit above 10 billion people around 2100. That’s an important variable, because each one of those people is a consumer. They drive a car, and they purchase things, and so this determines how much future consumption there will be.

The next set of projections that are really important to predicting climate change is keeping track of what is the average per capita GDP growth rate in different countries. If countries grow really fast, what’s the problem? They’ll consume a lot more. Here, in a very sort of sinister way, being poor is good for climate, right? You consume less. And so you’ve maybe seen those graphics that show how disproportionately responsible people in the United States are for climate change, showing that we produce on a per capita basis eight times more emissions than somebody living in India. Well, that’s essentially coming down to the fact that we consume eight times more dollars worth of stuff. The composition matters, but at the end of the day, the biggest thing is simply how much do you consume? That’s modeled by the GDP per capita growth rate.

And then the final one that goes into emissions is taking all those projections for the socioeconomic inputs and translating that into some pathway that actually represents emissions. More consumers and more purchasing of things that require energy-intensive production methods leads to more emissions. So A and B essentially finally get us to C. There are many more than just those projections, but that’s what represents the basis of what we need to get to this path.

And what we see is the central estimate is peaking right about now. And it has to, because if it delays and goes up in this area, things get very scary. But it is also true that we’re actually seeing this. China, for the first time ever, is reducing its emissions. I don’t mean to say slowing down the rate of the growth of their emissions, but literally reducing their emissions. This is essentially solar is taking over, and coal-fired power plants are not being online nearly as much as they were. And so we’re sort of at the climax of the environmental story. If this were a novel starting at the beginning of the Industrial Revolution, we’re right at the critical moment. Are the action heroes gonna succeed or not? So it’s terrifying, but hopefully it’s also very entertaining. We’re right at that crux of what’s gonna happen.

So that’s emissions, and I told you that there is this distribution around those, but I want to make this point a little bit more and set up for our next lecture. There are two ways of thinking about uncertainty. We’re gonna contrast those.

We’ve seen the first already: distributions. That’s gonna be something where you have the central estimate, and we’re gonna represent uncertainty by looking at some sort of envelope, which represents what we think is most likely. On the other one, we also had an even broader envelope, but basically what you can think of is if I could add a third dimension to this graph, there is a bell-shaped distribution. This line here represents the most likely outcome, and so you can sort of think of a series of distributions going out into the future. And they get wider. Why would they be wider in the future? Well, it’s because we don’t know what’s going to happen. The farther out we are from the present, the less accurate our predictions are going to be. So that’s an ugly graph, and you maybe don’t want to draw it, but it’s critical to understanding that we start out with a narrow distribution of what we think might happen, and the uncertainty gets wider and wider as we go.

The second way of looking at it is the one I’ve got on the slide here: scenarios. So now, instead of a central estimate, we’re going to have different explicit projections of what’s going to happen. Instead of uncertainty that shows this whole huge range of possible emissions, we’re going to take one specific set of drivers and say, under that set, we’re gonna have this emissions pathway. That’s useful, so here on the slide, we’re showing our five different representative concentration pathways. The 8.5 one, which is really scary, is the fossil fuel-based development trajectory going forward. But then we might contrast it with all of our different scenarios, like 7.0, or we might have a really optimistic scenario of 1.9. Side note: you can see they actually go negative, and that’s because we think we’ll be effective at actually sucking up carbon from the atmosphere and sequestering it.

But the point I’m trying to make is here the uncertainty is expressed in a different way. Rather than a distribution, we’re saying there are different scenarios, and we don’t know which one of these scenarios we’re going to be on. But it provides a potentially more useful way to think about how the different drivers behind emissions in this case are driving that, because you can say, well, in 8.5, what’s going on that gets us up to these really high levels?

Okay, so throughout, when we’re walking through this next paper, we sort of mix and match from the two different types of uncertainty.

Okay, so walking down our integrated assessment model pipeline, step two then is concentrations. Here we’ve got the same plots from before, but we’re adding another one here. This is the atmospheric CO2 concentrations, and this is what takes into account the different carbon pools. How much of it goes into the atmosphere versus how much of it gets absorbed into the ocean, how much of it gets turned into trees via photosynthesis. Once we have a model there, we can try to make a prediction of how those emissions will translate into concentrations. And we have yet more uncertainty around that.

Turns out the ocean actually is not doing as good of a job as we thought it would, so this is something that’s changed since about five years ago. More evidence is coming out in the state of the art that as we get to higher temperatures, the ocean doesn’t seem to be doing as good of a job of absorbing carbon, which means that there’ll be more carbon concentration in the atmosphere, even for the same level of emissions. That’s an example of science updating itself. So in other words, this bar maybe should be shifted up a little bit as the science becomes more certain.

Other side note is this is how science is supposed to progress. We slowly learn what the better parameter values are, and we narrow our uncertainty bounds. But sometimes climate skeptics will use this fact to say, oh, climate scientists were wrong. That’s not how it works. It’s that we don’t know what’s going on, but we’re improving towards a better understanding.

Okay, so that’s step two: concentrations.

Step 3 then is based on the climate sensitivity parameter, which, remember, we talked about: how does a doubling of CO2 translate into what percent increase in temperatures? That’s a critical variable. That’s the approach that’s then used to plot out the different trajectories of temperatures. So here, now, instead of concentrations, we’re actually seeing what the temperature rise would be. In this fossil fuel-driven RCP 8.5, by 2100 we’re seeing temperatures closing in on 5 degrees Celsius. And, you know, that’s even more scary in Fahrenheit. Whereas we see the optimistic scenarios, like 1.9 and 2.6, actually having a slight fall, peaking their temperatures somewhere in the beginning of the latter half of the century and falling a bit.

There’s one estimate that sort of is used by a lot of people: SSP2-4.5. It sort of stakes out a claim as being important because it represents what would happen if we pursued a business-as-usual policy. Like, what if we didn’t solve anything further? What if all the climate gains that we’ve had in policy so far stopped? Like, they’re included if they got in, but no additional policies came in. That’s the business-as-usual scenario.

This is nice because it’s less pessimistic than these ones where we had uncontrolled policy, but we are solving a fair amount of it already. And so we find ourselves more tracking with this one, but we sure would like to bend that curve down even further. This lingo about talking about business-as-usual and comparing all of our future mitigation scenarios to it is really where a lot of the action happens. We don’t so much think about comparing it to the worst case. We look at comparing it to how it performs better or worse than this central SSP2-4.5 scenario.

And so yeah, here it is back in the Nature article. This is the full range of possibilities, and so it centers at 3.4 degrees. This is essentially aligned with Nordhaus, and so maybe Nordhaus’s model is pretty good on the basis of we don’t do anything beyond what we’ve already committed to.

Okay, so far we’ve been essentially linking together a bunch of different biophysical models. But now we’re going to link in a new type of model: the damages model. We’ll talk more about this. In Nordhaus, the damages model was really easy to explain. That was the percentage loss in GDP from climate change, just temperature plus temperature squared, and some coefficients. But since Nordhaus’s DICE work, tons and tons of much more detailed analysis has gone in. On this last graph here, we’re asking the question: for any given increase in the previous graph, the temperature graph, what is the global mean sea level change relative to 1900, or one of many other different possible damage pathways?

So the key point here, expanding from Nordhaus and the DICE model, is to have many different pathways, which take into account what are the damage pathways needed to be included to understand the full linkage between the environment and the economy. Examples of where temperature alone might not be a good proxy is what happens if climate change also leads to extreme rain events. It’s very likely that temperature will go up and precipitation will go up, but also we see that there’s much less stability in the climate system, and we get much more freak storms. That’s a predicted result of climate change, but it wouldn’t show up in the average temperature. It would show up as sudden spikes of losses, as you know, you might have one really big flood in one location or another. Understanding which locations are most prone to an increase in flooding is really critical to understanding what would be the real impact back on the economy.

Anybody here own a house and have insurance on it? I do. It’s getting harder and harder to insure houses. There are whole regions of the United States right now where markets for insurance are collapsing. The reason for this is insurance companies, you know, they don’t really care about climate change per se, but they’re really good at estimating risk of assets they’re insuring having something bad happen. Because if they weren’t, you know, the whole insurance thing wouldn’t work. They wouldn’t be able to make good predictions about how much should people pay, and who is more at risk and less at risk.

But if you want to make a lot of money in the climate space, understanding how climate change affects the profitability of the insurance industry, there’s a lot of money to be made there. What we’re seeing is that they’re making very detailed damage function models, essentially, looking at things like where is their critical infrastructure or valuable real estate or whatever in areas that used to have a damaging storm every thousand years, now every 100 years, or every 10 years. Why do they care about this? Because they compute, well, how much will this cost them? And how much should they charge for insurance premiums? So this is going to be an increasingly contentious political debate, especially because what we’re really seeing is a lot of insurance companies are just pulling out of some areas. It’s really hard to insure property in hurricane zones at the moment. Not a lot of companies want to do it on their own.

So that’s just one example of how we would think about the specific translation of climate change into damaged pathways.

The Nature paper we’re looking at plots that very well, and I think it’s illustrated here. We’re gonna show two different plots. That’s the temperature plot over time, up to 2200. We see simulations that kind of wiggle around a bit. But then for each of these different RCP scenarios, it shows a distribution of the costs to the economy. So here what we’re saying is RCP 8.5, the scary climate scenario, it is going to have a risk premium that adds $32 trillion worth of cost to the economy. This isn’t damages, this is how much extra money would insurance companies have to pay to be able to cover those damages. So that’s a real cost that they would care about.

Yeah, so that’s going to be a different distribution underneath each of the different climate scenarios, but it also shows you the value. Why should we mitigate climate? It’s because the more we mitigate climate, and the more we get to these lowered temperature pathways, this distribution of damages shifts farther and farther to the left. And so here we’re just looking at three trillion in damages. That’s a lot of money. Even when you discount it, a lot of investment might make sense, if you have a sufficiently low discount rate, to shift the temperature curve down. But more relevant to us as economists, shift the damage distribution leftwards.

Okay. That’s the full pathway. Obviously one part left out, which is the discounting, which is saying that when these damages happen, it is going to determine how much we actually care about them. But first, I want to break down the details of what parts of the economy are going to get impacted.

From this paper, it looked at central estimates as well as distributions of what parts of the economy are going to be damaged most by climate change. What we see here is that it’s not energy. Everybody always thinks about energy as being the sector that will have the biggest losses, because we’ll have to close down a bunch of coal-fired power plants and such. They do account for that, and that comes up here as a relatively small US dollar per ton of CO2. So it’s saying $9 of damage to those energy industries happens.

But the two big ones are agriculture and mortality. Agriculture is $84 in terms of how it contributes to the social cost of carbon, but there’s a huge distribution around that. So there are even some estimates that are in the plausible range that show agriculture might actually benefit from climate change.

Side note: can you think of any country that has a lot of land that is currently unproductive? Canada and Russia. And Russia is one of the slowest actors on climate change, because frankly, they stand to benefit quite a bit from it. A lot of Siberia, they think, is now potentially posed to be better agricultural land. I would also add, though, that the most recent studies show that that’s been highly overestimated, because Siberia doesn’t have the soil necessary. They have very thin topsoil layers, and that’s just because they haven’t been prairie like the Midwest has been for a very long time. They’ve been just forests. That doesn’t generate the same rich topsoil that we have here. So it might be productive for a little bit, but will be less so going forward.

But yeah, so that’s the big distribution. Mortality is the other big one. This is mortality from many different sources. Storms is a small one. It’s also heat waves. Heat waves and other extreme events contribute to excess mortality.

Additionally, climate change is unfair. It’s not just bad, it’s unfair. This is focusing here on mortality, and this is the distribution of the burden of climate change, in this case measured as increased deaths per 100,000, and comparing that to the global income deciles.

So what this is saying is that, especially the second decile here, the portion of the global population that is the 20th percentile, so poorer than 80% of the population, they’re the ones that have this big spike here. There’s gonna be an unfairly large amount of death happening in that decile, with the next one being the very poorest, but then that falls off.

It’s interesting to question why might it rise here? Well, turns out a lot of rich people live right on the coast too, and so that exposes them more. But there’s still this unfair component.

The other one is: how much will the costs of adaptation be? So this is a bad thing you want to avoid. This is how much might the different deciles pay to avoid the damages. This is another aspect of the inequality: rich people, what do they have? A lot of money. What can money buy? Everything. It can buy air conditioners, it can buy protection from climate change, and that’s measured by adaptation costs here. They can spend money building seawalls or building buildings that are more resistant to hurricanes and such. This means they’ll spend more, which is good for them, but it’s unfair because people in lower-income groups probably would like to be able to do that. They simply can’t spend the money to be able to do that. That’s an interesting ethical question.

Okay, so in terms of spatial variation, here’s a fun paper that came out not too long ago, looking at the spatial distribution of these damages. Here is a graph, for instance, in the upper left, showing what do we expect will happen to agricultural yields? This is just the United States, but the same principle is evident in a lot of areas. What we see is there’s a big fall, especially in the southern and central part of the American corn and soy belt. That’s because corn and soy are quite a bit more exposed to climate change than other crops.

What’s grown up here? It’s actually wheat. It’s more grown up in the north. Turns out wheat is one of the crops that benefits more from climate change, simply because it’s limited by the number of growing days that are warm enough to grow, rather than being damaged by, for instance, overly wet or overly dry fields. And so we see what you might expect: you could make predictions that okay, well, if this is the case, these regions will stand to benefit more than these regions. Although I would point out that the vast majority of agricultural value comes from this area here, so that’s a real problem.

So that’s one type of damage function, and so basically what we’re saying is all those different graphs are different parts of this damage function linkage. They’re saying for a given increase in the temperature, what is the makeup of the damages? So we see this one through crop damage. Here’s mortality damages, increased energy expenditures. Look, up here, we spend a little bit less energy in northern Minnesota. Although we will spend a little bit more on our air conditioners, we don’t have to burn our natural gas and our radiators as much anymore, and so we actually benefit from that.

But then other ones would be damages to labor. High-risk labor groups, especially those working in agriculture or construction, are really exposed to climate damages.

Other ones would be coastal damages. This is essentially where do storms hit, and where will they wreck buildings. But this is real damage measured as percentage of county GDP. And then some other ones that are a little bit more interesting is there is a strong relationship between increased temperatures and property crime and violent crime.

The state of Minnesota is actually quite safe. You know there’s a lot less crime when it’s really cold out? And why is that? Well, criminals don’t like the cold either. If you’re gonna be breaking into garages or something like that, well, it’s just a whole lot less convenient. It lowers the relative return, essentially, because it’s quite costly. The other theory is that when it’s hot, it affects people’s moods, right? Who gets cranky when it’s hot out? I do. So yeah, that one’s maybe less speculative than some of the ones up here, but it’s listed here in this report.

But basically, this last one here then plots out the combination of all those direct damages, and we see the southeast of the United States as well as the southwest is really going to get walloped in terms of damages, and much less in the north, if not even positive gains. So, we all live in a lucky place, right? Also, we have a lot of lakes and fresh water that goes towards protecting us against water shortages. So I always use this when I’m trying to tell people how awesome Minnesota is. We’re quite resilient to climate change.

But yeah.

All right, so we spend a lot of time on damages, because that’s really important. But now we’ve spent so much time talking about discounting that we can sort of do this quickly. We’re saying given that set of damages, so essentially this plot of where the damages are, let’s aggregate that up to the whole region and then make an estimate of what is the total damages from climate change under different discount rates.

What we see is the expected result that at this point I’m hoping is intuitive: if you have a high discount rate, like 3%, the total damages are saying that the social cost of carbon is $80 per ton. But if you care a lot more about the future well-being of individuals, and when you compute the damages that we expect will actually happen, changing the discount rate to just 1.5%, this increases the social cost of carbon out to $308. So now all sorts of different projects would pass a cost-benefit analysis test with this much higher social cost of carbon.

Okay, I did make a web app for this, but I didn’t get it to pass my scientifically valid test, and so I’m not going to give you access to it. But I do want to show how this could look, and I think I’ll keep tweaking this until finally I get the numbers to work. I essentially set out to make the DICE model, but in a web app, just to show that it’s possible.

But let’s just walk through what this looks like in terms of computation. You know, these models that economists like Nordhaus and Stern use, they’re valuable because they let us try out different policies and see how good are those policies, and, you know, which one is the best, and which ones might we want to avoid.

So now, instead of talking about this five-step integrated assessment pathway in just general terms, here, let’s actually compute it and see what we can say. We’re gonna skip a number of steps, but the first thing that’s done is calculating the GDP damages with and without climate change.

This is the central estimate of what would happen if we had this false but very optimistic world where there are no climate damages. GDP would go up by this amount. But when we take into account the damages that are lost, we would actually have instead this growth rate. Obviously, this is going to be a function of all sorts of things in the model. Eventually, the web app will let you drag around the discount rates, as you’d expect, and also things like the cost of solar. But it’s this big difference expressed in terms of trillions of dollars of GDP that we could do a better job on.

Why is that? It’s because as the temperature goes up, we have this growing portion of damages, measured here as a percentage of GDP, capping out at you know, over 15% to 17% of GDP lost in 2200.

But what’s the social cost of carbon? From there, we can actually compute the optimal amount. This is a full-on integrated assessment model, and so if we know the damages and we also know what are the options for mitigating it, what would be the cost-benefit analysis? Well, the key point is that anything that is cheaper than the social cost of carbon here would be worth doing, because it lowers this gap, gets rid of some of those damages, and it does it in a cost-effective way. That’s all gonna depend on what the social cost of carbon is.

So the key thing relevant to policymakers, at least, that comes out of an integrated assessment model, is a time path like this. This is $93 per ton of carbon emitted today as the current social cost of carbon. It goes up, and this goes up primarily because the economy gets much bigger, and so there is more damages in terms of dollars. But then it eventually falls off to zero, and why is that? That’s because of the discount rate then taking over, where dollar values out in 2200 don’t matter too much to us.

Okay, and so I’ve been saying all along that the social cost of carbon is useful because it’s this policy-relevant tool. So how do we do it? Essentially, we’re gonna do what all good government bureaucrats do, and this has actually been mandated since Reagan: do a cost-benefit analysis on the environmental policy.

Basically there’ll exist some possibly expensive policy that could mitigate these damages. And do something like this: where instead of having this big gap in between the GDP trajectory without damages, the blue, and the one with damages, the red, there is, in this case, actually we computed what is the optimal policy. How much could we minimize the area lost in between those two curves? I’ve dragged around the numbers a little bit so it’s slightly different shapes, but what we see is the optimal carbon path that emerges from this integrated assessment model is going to be good, but only in the long run. And you can sort of see the problem here. It’s like, look, it’s actually making the economy worse off for a little bit, because we have high taxes. That means less money spent on consumption. But then eventually, it gets huge gains insofar as society doesn’t collapse. That’s, I guess I was a little bit hyperbolic, but maybe not.

But anyways, the point is there will be short-term losses for this long-term gain.

So in other words, the cost-benefit analysis is essentially taking that social cost of carbon and figuring out what policies should we do. The optimal policy is as simple as: set a carbon tax equal to the social cost of carbon. Oftentimes, we don’t get to do that, because nobody seems to listen to economists. Instead, we’ll just do the sort of standard cost-benefit analysis of which projects make sense to invest in, and that can also get us to a similar place.

Alright, so that’s in principle. But I want to leave you with a more positive thought: things have gotten better. They really have. Here’s the profitability, measured by the levelized cost of energy, which is the right way to look at this. It’s showing that solar and wind have just gotten cheaper. I’ve said this before, but it’s worth digging into that, and why is this?

Essentially, there is a big upfront cost for renewables, but they have no fuel, right? Here’s a coal plant constantly filling itself with coal, and if you take into account the costs of production, that’s what drives it up. It’s expensive to keep mining this, whereas here, it just keeps on going and going.

But when we look at these improvements that we see in energy production, it turns out that solving climate change is cheaper than doing nothing. So what I want to switch from is the narrative that says we need to give up now in order to save in the future. That’s a valid argument, but that’s dependent on renewables being expensive. The expense of renewables is the thing that causes this to be negative.

But if we have falling energy costs for renewables, it actually becomes the case that it’s cheaper to solve climate change. So that’s the positive note. That’s why many forces in society are arraying in such a way that hopefully the world is not quite as negative as it looked 10 years ago.

If you’re depressed by current news, like the EPA dismantling essentially the regulatory framework for climate change, that does matter, and it is certainly a negative thing. But the timing just might be such that we’ve pushed beyond a critical point where we have invested enough research and development in energy that it might not matter. It will matter. It will, on the margin, increase the amount of emissions that happened, but way less than if the same thing had happened five years ago.

So I think that’s the positive point.

I have a bunch more details on the slides. I won’t test on the details, other than to say: get an electric vehicle. Here’s a Nissan Leaf. That’s the other sort of advances. Technology has progressed, and the predictions for when electrification of our U.S. car fleet would slowly increase, we’ve actually accelerated and greatly outperformed. We’ve had much faster adoption of EVs. That’s my final point to leave you on. They’re also more fun to drive. I really like cars.

Okay, so have a good weekend. That’s it for climate change. We’re gonna pick up, like I said, with talking about scenarios and then plugging them into other aspects of the environment, and then start to switch to natural capital. I didn’t get a chance to talk about country reports. What we’ll actually do on Monday is talk about which countries you all are assigned to. I’m going to do a matching exercise this weekend.

All right. Thank you!