What do IAMs measure and how can they improve?
“Traditionally, modelers have been using IAMs to ask ‘What if’ questions. What if we shut down the power plants? What if we got rid of all the gasoline- based cars? The models can provide data on the answers to those questions,” said Peng. “To date, the IAMs are doing a fairly good job presenting the physical system such as energy technologies. But the political and human drivers aren’t as well represented in the modeling systems.”
Peng, a trained modeler, perceives climate change through quantitative data, but recognizes its shortcomings when looking for realistic solutions for decarbonization. “There’s a gap between the type of policies the model tells us we should do versus what we can actually implement given the sophisticated political and human environment we’re living in today," explained Peng.
Peng and her colleagues will incorporate political economy insights to examine two main issues in their models: domestic policy instrument choices, and global supply chain strategies within the United States. With this unique approach, they can upgrade IAMs to identify politically realistic and actionable solutions to take against climate change, according to Peng.
“Our mission is to ultimately tackle the climate problem through demonstration cases that are tied closely with the policy debates. We use these cases to push the frontier on how we can model the coupled human-natural system with improved representation of the human system,” said Peng.
Elizabeth Ransom, interim director of the School of International Affairs and associate professor of international affairs said, “Wei Peng’s work pushes the boundaries of what IAMs are currently capable of. This could change the way the IAM community approaches their research.”
The importance of process
With a variety of experts involved in this project, Peng noted that the team needs to develop an understanding of the other discipline to foster meaningful collaboration. She described how concrete examples are important to identify instances of disciplinary differences, as opposed to talking in an abstract, high-level way.
“Modelers and social scientists have different trainings, different sets of research questions, and fundamentally different methodology. Hopefully, through co-developing and co-revising the modeling strategy, we can gradually develop the shared language in order to do work together,” Peng said.
One of the first steps forward in uniting the modeler and political science communities started with a paper Peng led in June 2021, titled "Climate policy models need to get real about people—here’s how." Published in Nature, the other co-authors included leading modelers from around the world and several social scientists. “This is a consensus building piece," said Peng. "Many in the IAM community believe this is the direction we need to take."
For the new project, the team meets once a month to present concrete modeling demonstrations as a way to develop conversations between the members of the team and to make sure the modeling strategies represent the core insights from the political science field.
Bringing real change through policy
Peng and her team engaged with policymakers to pinpoint the two modeling issues the researchers will explore in the project. She is taking steps to keep them involved with their findings.
“We plan to have an annual workshop for researchers and policymakers in Washington, D.C. We’ll also work with them to develop a synthesis paper about what worked and what didn’t work in the project. It’s truly about knowledge co-production. We want to identify the good practices to ensure we produce actionable knowledge. Our modeling strategies need to be reflective of concrete policy needs. When we’re modeling realistic decisions, we’re going to have specific model users and decision makers in mind,” Peng said.
Peng’s research for this project started in September 2022 and is expected to end in August 2025.