Host Kristin Hayes talks with Massimo Tavoni, the director of the RFF-CMCC European Institute on Economics and the Environment and an associate professor at the School of Management of Politecnico di Milano in Milan, Italy. They discuss integrated assessment models, what they are, how they're used in studying climate change, and why they matter for decisionmaking.
Kristin Hayes: Hello and welcome to Resources Radio, a weekly podcast from Resources for the Future. I'm your host, Kristin Hayes. This week, we talk with Massimo Tavoni, the director of the RFF CMCC European Institute on the Economy and the Environment. Max is also an associate professor at the School of Management of Politecnico di Milano in Milan, Italy. Max focuses his research on climate change policies, and he advises international organizations and institutions on the low carbon transition.
Today, Max and I talk about integrated assessment models—what they are, how they're used in studying climate change, and why they matter for decisionmaking. Stay with us.
Max, welcome to Resources Radio. It's great to have you here.
Massimo Tavoni: Thank you. Yeah, I'm excited.
Kristin Hayes: Max, you've ended up spending your professional life building and using models of the Earth's climate system. How did you get involved in this fairly technical field? What's your background that brought you to this point?
Massimo Tavoni: Yeah. This is indeed the outcome of something very personal but relates to my background. I started doing work as an applied mathematician, an engineer, and then worked in several jobs and was not very happy with those. And then at some point, I decided that climate was the thing I wanted to actually study more. So I got into that, and then did a masters along the way in economics and then I did a PhD afterwards. I was kind of blending things and the reason being, I wasn't sure exactly what I wanted to do in life, I guess. So it just took me a long time to actually figure out what I really wanted to do. That has downsides.
The upside was I learned a lot of different things along the way and picked up on so many different tools. And then I realized that the climate problem is such a big problem that you need to have these integrated views that range from engineering, math, climate, economics—just to understand it.
For me, the natural language for trying to address such a complicated issue (which otherwise sometimes seems just too big to be solvable) is just to use these kind of models that try to quantify and relate different components of the Earth, of the human systems, of our society, and write them down in a way that is actually formal but it's at the same time understandable. That's how we started and got going for several years.
Kristin Hayes: All right. Well, thank you for giving us a little bit more of an introduction. I know these models are a really critical part of decisionmaking, so it’ll be good to learn a little bit more about them from an expert.
My understanding is that the most common type of model used in understanding how the Earth's climate may change over time is called an integrated assessment model (or IAM). Can you break that down for us? What does each of those words mean in this context—integrated assessment model?
Massimo Tavoni: So, “integrated” stands for integrated different disciplines. If you want to understand what it takes to stabilize the climate and avoid climate disasters, we need to integrate what we know about the climate to begin with, obviously—that's pretty simple. But then we need to understand the sources of the problem. And that's everyday life. It's our economies, the way we use and consume and produce energy; it's the way we use land. And so you need to bring in these different elements. Essentially “integrated” because it integrates a series of things and disciplines and, in this specific case, a series of models related to climate, economics, social dimensions as well, and technological dimensions.
Assessment, that second letter, the "A" letter, stands for using these to assess policies. If we want to understand how we get to two degrees and avoid going to five or even more, we have to understand how to assess these transition policies and what it takes to get down to two degrees. Essentially, this is what the model is used for.
And then "M" stands for models, of course. This is may be the most difficult one. It is a model, after all. It gets a lot of confusion, of course. It's about essentially something which is written in the software type of environment. It's just [inaudible 00:04:07] computers, which is just too difficult to solve pen and paper. That's what it is.
Kristin Hayes: These are quite big computers, from what I understand. These models, given how elaborate and complex they are, take a tremendous amount of computing power to actually run that many pieces of information.
Massimo Tavoni: There's a lot of variety. I think that's also the beauty of it. Of course, the most complex models take a lot of time to run. Sometimes, however, you also have very simple models that actually take very little to run. They're very useful for teaching or for just understanding the basic things, and do a very good job at explaining some of the things but not all of them. This is not just some so complicated thing that cannot be understood.
In some cases, it can be understood. Indeed, the Nobel Prize in economics this year was awarded to a professor who developed such a simple model, which is publicly available and can be taught very easily at the university or even before, in high school. Then, of course, things can get more complicated in a more complex model, but that’s just to say there's a lot of variety here.
Kristin Hayes: Yeah. Okay. Well, that's great. So, specifically looking at integrated assessment models, these IAMs—what kind of questions are they designed to answer? And how does the information they produce get used in a decisionmaking context or a policymaking context?
Massimo Tavoni: I think we answer three main questions. The first one is: what's going to happen if you keep doing what we're doing right now? We know we're not doing it right—but how bad is it going to be? So this is kind of the scenario: how warmer will be the planet by the end of the century, or in 20 years from now, if we keep doing what we are doing? So, this is the first one. The second one (which is obviously the opposite side) is: what shall we do if we want to avoid still going up and emissions are rising? And we want emissions to go down and temperature to be within certain limits that we don't want to exceed. These are what most of these models are usually, most of the time. So it's telling you want kind of policies you have to deploy, what kind of technologies you need, what should us as individuals—as opposed to corporations, as opposed to organizations—do to stabilize the climate.
The third one, which is yet another thing, is understanding what are the economic costs and the economic benefits of reducing CO2 [carbon dioxide] emissions and making the planet a more healthy place—both in terms of the climate but also the other systems (better water, better land, better air quality) in general. So these are the main issues.
Kristin Hayes: Okay. So, just one follow-up question for you on that. Here at RFF, we have a number of models and oftentimes we use those to actually see what the impacts of a particular policy choice might be. So, what happens if you put a carbon tax into the US economy? What are the effects of that? It sounds to me like IAMs might be used slightly differently. Whereas we would put a policy into a model—you might actually start with a temperature target or an emissions target, and see what sort of policies might get you there. Is that right?
Massimo Tavoni: Yeah, you can do it both ways.
Kristin Hayes: You can do it both ways. Okay.
Massimo Tavoni: They can also be used the other way, as you said, just to evaluate the policy.
Kristin Hayes: Okay.
Massimo Tavoni: But they're also meant to inform policymakers when they're thinking about designing policies. That is a very normative approach. What should the world do—
Kristin Hayes: Yeah.
Massimo Tavoni: [...] in the best possible world to get ourselves to a safe path?
Kristin Hayes: Yeah.
Massimo Tavoni: In a way, this is a bigger, more complicated question—and sometimes it doesn't include political rallies.
Kristin Hayes: Sure.
Massimo Tavoni: Some of the many things that we know very well are making this transition so much more difficult than we think.
Kristin Hayes: Yeah, yeah. How do these models deal with technological change?
Massimo Tavoni: These are models which look at the very future, long future.
Kristin Hayes: Okay.
Massimo Tavoni: They generate scenarios for many decades, to the end of century—when most of us, I guess everyone, at least among us or least in my [crosstalk 00:08:11].
Kristin Hayes: I'm planning on living forever.
Massimo Tavoni: All right. So, here, to what extent technology will change is a really crucial parameter, but it's not just technology preferences or social habits, lifestyles. People’s preferences change as well; it's not just technology. Of course, technology changes very, very fast. This is one point where the models can do something and can also [...] should also be taken with care. Because we can project some of the changes that we see already happening, but we also cannot project many of those that will come. For that, we resort to large scenarios and samples, like sensitivity analysis. Plus, we use the models not as prediction tools—but as tools for thinking about what are the problems and their solutions. Also, you have to understand that a lot of the things related to climate and the energy sector take a lot of time to change. We have a lot of power plants in this country, in Europe and elsewhere, that are going to stay there for a long time.
So some of the things we know (actually, for a long time) will play out, some others we don't—and we have to be very humble and not trust the models too much, use our little salt grain.
Kristin Hayes: Yeah. Yeah, humility is always a good thing.
Massimo Tavoni: Yeah.
Kristin Hayes: Just one other context question: how many of these models, how many integrated assessment models, are there in the world? Are they common? Are they place-specific? Give us a little bit more.
Massimo Tavoni: There are several. Each country has several. Europe has a lot, maybe it's one of the leaders in this. The US also has several teams spread across the country in research institutes and universities that develop them. Then Japan has a history, and China. So that's actually the fun thing of working in this field. We get together. It's a community science. It's not just one team. It's a group of teams and each team being big, this means a lot of people. It's a community of international scientists. So we meet regularly around the word and discuss things and how to improve models, and we learn from each other, which is actually the very nice thing about it.
Kristin Hayes: I learned just this morning that your institute in Milan is actually the home or the secretariat for the integrated assessment model working group. Is that right? I think that’s the right name.
Massimo Tavoni: Yeah, it's a consortium. It's a kind of association—
Kristin Hayes: Okay.
Massimo Tavoni: —scientific association of all teams working on these kind of things that, as we were discussing before, have a lot of influence on the policy and the design of climate policies. Yeah, we are thrilled to have the secretariat just because it allows us to engage with so many different teams. This being so complicated, it really helps to do community science in this case and have sharing practices and learning from each other a lot.
Kristin Hayes: Yeah, that's great. It does sound like these models play a very important role in decisionmaking and, I think, sometimes can be perceived as having an outsized influence in decisionmaking. You've already talked about some of the challenges and some of the places where humility is needed in interpreting and understanding modeling results. Tell us a little bit more about other places where you've identified weaknesses in the model or ways to improve the modeling—ways where we should take that grain of salt.
Massimo Tavoni: One thing for example [...] I mean, there are many examples where we think we should take these kind of long-term scenarios with sufficient care and not just get too trapped into what the model says and trust the models too much. I think that's always useful. Human behavior is one example that always comes to mind. It's very easy (or, let's say it's relatively easier) to model things like technological transitions in some sectors. It's much more difficult to model human behavior. Humans—
Kristin Hayes: Humans are so complicated.
Massimo Tavoni: They're complicated. They behave differently and they are, by nature, different. Plus, this is a global problem—so, we have different cultures, people all over the place. So that is something where the models, for example, have not done a good job in the past. They're trying now to improve. They are using sciences coming from the behavioral world and psychological (social psychology studies) to get a better feeling of what actually people do when they're confronted with choices and also a lot of the mistakes that we do every day in every [...] in so many of our decisions. That part is very hard to capture. The other is the social part. This transformation is not just a technical fix. This is much more profound. This is a way also to influence the way our societies work and function. This also is very complicated to model very precisely. So we do a lot of scenarios and assumptions about the future which resemble different paths and, again, try to be humble as much as we can.
Kristin Hayes: Okay. Can you give an example? For example, your team in Milan runs an integrated assessment model. Could you tell us a little bit more about that one, and a way in which your team is directly using that or a way that you're working to make your particular model better?
Massimo Tavoni: Yeah, our team is a relatively small team, maybe, compared to others. We are about, maybe five, six people working on this kind of model. It's a model that [...] well, it's a model that actually combines two elements. On one hand it’s the global problem and the fact that countries do not act in the interest of the world. So we take this very seriously. This is reflected in international politics as we see today, in the US, in Brazil, in many other countries—that what is best for the world is often not necessarily best for single individual nations. So our model is this game theory approach, where we have these strategic incentives and strategic behavior of different countries that we compare and contrast toward this more paternalistic view of the world as a whole. The other element is technical change, which we discussed already. In this, the model brings together the two Nobel Prizes that were awarded this year. One was already mentioned, and the other went to Paul Romer, who developed and launched the technical change theory of economics.
So our model from the beginning thought that these two elements—the strategic game theoretic dimension and technological change dimension—would be important in solving the problem. Indeed, if we don't cooperate and if we don't develop technology sufficiently, we will never be able to solve it.
Kristin Hayes: Right, right. So can you tell me a little bit more about the types of data that go into these models? Given that it's a global model, how do you access global data sources of the kind that you might need to feed a tool like this?
Massimo Tavoni: Well, luckily (at least on that front), we’ve made progress in the past few years. There are now large databases—of socioeconomic indicators, climate indicators, energy indicators of all sorts—for most countries. If they are not precise [...] in some regions they are less precise than in others—but overall I think the quality has improved a lot. We are seeing this more and more with data sources that are new and were not there before. Now we have remote sensing and data coming from individuals, from regions. So we can get down at a level of disaggregation which is much better than we used to be.
Kristin Hayes: Okay.
Massimo Tavoni: Now, this is for today. This is the data that describe the world as it stands. We want to describe the world that changes though. That is much more complicated.
Kristin Hayes: Right.
Massimo Tavoni: Because we want the world which is nothing like the one that we've seen in the past—otherwise we would keep doing what we are doing. This is exactly not what we are looking for.
Kristin Hayes: Right, not the goal.
Massimo Tavoni: Exactly, with this model. So then the thing here, how you get to transition and to change and how you change these paradigms—that is much more complicated to answer and refers back to what we discussed before.
Kristin Hayes: One of the important ways in which integrated assessment models are used is in developing these large-scale global reports put out by the Intergovernmental Panel on Climate Change. They come out periodically. They give a high-level perspective on collective guidance on what we should be doing, what path we're on, what needs to change. Can you say a little bit more about how IAMs feed into those reports, which are pretty seminal to how the world thinks about climate change?
Massimo Tavoni: Yeah, thank you. That's an important question. They relate and speak very much to the scientific community. It happens in various ways. Many of the authors, lead authors of these reports, are typically people working in our field. So, we are directly part of these assessment report phases, which are pretty long and tedious. We have several meetings (week-long meetings) where, essentially, we discuss whatever is out there in terms of the published literature. Then we create scenarios for the IPCC to be as objective as possible. The IPCC reviews evidence; it doesn't produce any new evidence. So everything which we provide to the IPCC has been published and has been vetted by peer review. It's actually open and accessible. So all the repositories of all the papers and all the articles and all the scenarios are actually made publicly available. You can easily find them on our website, and these can be openly downloaded.
Kristin Hayes: But the IPCC, it sounds, is quite dependent on the kind of information that you are producing in order to generate its own conclusions.
Massimo Tavoni: It's certainly not the only thing that the IPCC cares about, but it's certainly a very integral part—not just on the climate side, but also on these kind of transitions toward low-carbon scenarios. The climate, we know already that it [...] more or less how it looks like, but we need to know, more and more, what we need to do to make it look different. That is where the scenarios are very useful and indeed the IPCC has been using them a lot, either to portray possible alternatives in the future or just also to provide more data on what we should actually do. So, for example, when should emissions start falling? In which year? At what rate? When should emissions finally reach zero? These are typical things that you read in the reports—peak [emissions] in 2025, steep decline in 2030, 40, reach neutrality by 2050, 60, 70. These are typical things that come out of the models that would be very hard to, let's say, have them coming out of something else.
Kristin Hayes: You mentioned that a lot of the work, or all of the work, that you do is in fact peer-reviewed. There's a commitment to illustrating what's behind the modeling results, but that's difficult. Models are, I think, inherently difficult for the layperson to understand. So how do you build trust in the community around the information that's coming out of these models? Are there avenues for transparency or other ways that you can convince people of the validity—at least, with all due humility—but the validity of what it is that you're producing or finding?
Massimo Tavoni: Yeah, you're touching upon a very important point. If we are not trusted, then no one will take whatever we say seriously.
Kristin Hayes: Yeah.
Massimo Tavoni: We have different ways of doing this and all should be perfected anyways. At a science level, as you said, we get published with peer review, so we follow the standard procedures. As I also mentioned already, we made our data publicly available. Plus, most of the models (I would say, now, the majority) are actually open source. People can download the code and run them. Of course, they are very complicated, so it's not an easy job to do.
Kristin Hayes: Sure.
Massimo Tavoni: The other thing is that we build visualization tools. That's where we're working a lot, to help everyone understand what's going on. There are a lot of ways now to visualize complex databases with the big data revolution. We are harnessing some of that power and pairing it up with experts who know how people understand images and can really understand visualization. [These are] very good experts in terms of designing them in a nice way, and creating more and more infographics and databases that can provide a glimpse or an overview of the key essential facts. Then, if you want to dig deeper, you can. You can download the data, go to the code. In this way, we are building several levels of transparency for different audiences, and that's what we are doing right now.
Kristin Hayes: That's great. That's great. Do you have a website you could recommend, where if somebody wanted to check out (someone being me) [...] if somebody wants to check out some of those data visualizations?
Massimo Tavoni: If you look for IAMC (which stands for Integrated Assessment Modeling Consortium) scenarios [...] just Google that, and you’re going to get things coming up.
Kristin Hayes: Okay. IAMC, got it. All right.
Massimo Tavoni: Yeah.
Kristin Hayes: Okay. So, Max, we usually close our [...] we always close our podcast with a feature that we call “Top of the Stack,” where we ask our guests to recommend something to our listeners—either something that you've read, something that you've watched, something that you've listened to—that you think might be of interest on this general energy, climate change, natural resource [...] in this area. What's on the top of your stack, Max? If it's in Italian, that might be a little challenging for some people, but that's fine. We'll take recommendations in all languages.
Massimo Tavoni: Yeah, thanks, thanks. I have two very quick recommendations. First one, if you want to know more about all this stuff, just a very good website: which is called Carbon Brief. I don't maintain it. It's a UK website where you can find a lot of information on all kind of scenarios ([e.g.,] two degrees, and how we get there) and what kind of science we need to back it up. It's very useful, very informative, and very simple. Now, with that said, I will also suggest to watch a documentary, which is called “Mountain.” It's just a very simple title— Mountain.
Kristin Hayes: Okay.
Massimo Tavoni: It's just these long images of mountains, with a narrator who tells you things about mountains and why we got fascinated by mountains. And why should I care? Well, obviously, this is personal. I care about mountains a lot.
Kristin Hayes: Yeah.
Massimo Tavoni: I like mountaineering a lot. It is also the mountains where we are seeing the biggest changes due to climate change.
Kristin Hayes: Yeah.
Massimo Tavoni: So, if you’ve had a chance to go and see the glaciers and see how fast they are retreating—and, in general, if you're fascinated by high altitude and big mountains—you're going to probably also understand why climate is changing and will be affecting ourselves and our future generations so much.
Kristin Hayes: Yeah.
Massimo Tavoni: It's beautiful movie.
Kristin Hayes: That sounds fantastic. Thanks. Those are great recommendations. Well, Max, thank you so much for joining us. It's been a pleasure.
Massimo Tavoni: Thank you. My pleasure.
Kristin Hayes: Thank you so much for joining us on Resources Radio. We'd love to hear what you think, so please rate us on iTunes or leave us a review—it helps us spread the word. Also, feel free to send us your suggestions for future episodes. Resources Radio is podcast from Resources for the Future. RFF is an independent, nonprofit research institution in Washington, DC. Our mission is to improve environmental, energy, and natural resource decisions through impartial economic research and policy engagement. Learn more about us at rff.org. The views expressed on this podcast are solely those of the participants. They do not necessarily represent the view of Resources for the Future, which does not take institutional positions on public policies. Resources Radio is produced by Kate Petersen, with music by Daniel Raimi. Join us next week for another episode.
Program Director, Energy and Climate
Director of the RFF-CMCC European Institute on Economics and the Environment