New Faculty Q&A: Giacomo Lanzani on Learning, Misspecified beliefs, and Why Polarization Persists

Berkeley's economics department recently added microeconomic theorist Giacomo Lanzani to its faculty. In this Q&A with PhD student Serra Pelin, Lanzani discusses his path from reading Isaac Asimov's Foundation series in Italy to studying how people learn under Incomplete information. His research challenges the equilibrium assumption that underlies most economic analysis, showing how misspecified beliefs, selective memory, and network structures prevent societies from converging to correct beliefs, even when collectively they have enough information to do so.

Can you share a bit about your journey into the field of economics and what inspired you to pursue this career?
I did my undergraduate in Italy, where your "major" selection is much more consequential than in the US. In particular, I decided to pursue (mathematical) economics naively. I loved reading the Foundation cycle by Isaac Asimov, and I was looking for the discipline that was closest to using mathematical modeling to predict the future behavior of our societies.

Your research focuses on learning, decision theory, and networks in economic settings. How would you explain to a non-researcher why it’s important for economists to study how individuals learn under model misspecification or limited memory?
Most of the analysis in Economics is conducted under the equilibrium assumption, which puts a lot of discipline on our predictions. One of the main justifications for focusing on equilibrium is that agents who are repeatedly interacting will eventually converge to an equilibrium. However, such results assume agents who have a correctly specified view of the world (i.e., they correctly understand how to use what they observe to adjust their beliefs) have infinite and unselective memory and reach a feedback structure about what others do. My work shows how the equilibrium assumption is much less justified and nuances arise when agents are either misspecified, have selective memory, or can only learn from what they observe from their neighbors.

What advice would you provide to a graduate student interested in Microeconomic Theory?
My advice to students in Microeconomic Theory is dual. On the one hand, they should not give up on the mathematical rigor of the field, which has allowed it to be so helpful in the development of economics as a cohesive field and body of knowledge. On the other hand, they should keep the motivation for their work very grounded in the applied questions of other fields, ranging from Macroeconomics, to Industrial Organization, to the insights on human behavior that come from even outside Economics, for example, in Psychology. Then use our methods to help deal with those questions.

We all see people stuck in social media bubbles or disagreeing with no end in sight. Your research studies how people in a network influence each other's opinions. What does your work tell us about why it is so hard to change people’s minds or reach a consensus in a connected world?
My research provides a theoretical model where polarization and failure to agree on an underlying fact can happen even when there is a large number of agents who receive signals about this underlying state (think of the case of many people experiencing the outcomes of vaccination). Even if the joint information available out there would be enough to make us agree on this underlying state, I show that the network structure of cliques and attraction towards the most extreme opinion can prevent the convergence of the society to the correct belief.
What do you find most challenging about doing economic theory today?
There is a perception that many fundamental questions in economic theory have been settled once and for all, and that therefore, Economic Theory is doomed to become less and less relevant. I feel that this view is misguided from what in reality, is a strength of our field: it proceeds much more ‘vertically’ compared to applied fields, and it builds much more on our past discoveries.

Your research has looked at how a policymaker's own faulty beliefs can lead to real-world outcomes, like cycles in monetary policy. What is the single most important lesson from your work that a leader or politician needs to hear?
The key lesson is that politicians should use models, paradigms, and systems to make sense of what is happening in their countries and economy, while at the same time being very aware that these models and paradigms are useful but imperfect tools. My work highlights the significant gains that a central banker or policymaker can achieve by adjusting the trust in the model they use in the face of new data. 
 

Looking ahead, are there any specific topics you are eager to explore in the near future?
I am currently working on importing the tools I developed for analyzing learning and information transmission on networks into the theory of Production Networks. There are striking mathematical similarities that enable us to apply the tools we have developed to understand the problem of endogenous network formation in macroeconomic models, thereby shedding light on the fragility and resilience of different production networks.