Interviewer: Alycia Lee
Favorite book: This one is tough because I have spent my life surrounded by books. Three that have recently been on my mind are “How to write an autobiographical novel,” by Alexander Chee, “The Unbearable Lightness of Being,” by Milan Kundera, and “The Elegance of the Hedgehog,” by Muriel Barbery.
Favorite food: A bowl of tomato soup with a hearty slice of bread
Favorite place you have traveled to: Paris
Favorite economics theorem: The Revelation Principle—to the point that in my first year in Caltech, the only complaint I received in teaching evaluations was that I spent too much time on this result.
Fun fact that most people do not know about you: I never learned to ride a bike.
What motivated you to delve into the field of mechanism design and matching markets? What are their fundamental importance and practical applications? What do you foresee as the future direction of the field?
My entry way to economics was game theory: I have always been interested in understanding human behavior, and the possibility of using precise mathematical models to do so complemented my love for structure. Human behavior, however, rarely occurs in a vacuum, but within a wide array of economic, social, and computational systems. The outcomes we observe are thus a combination of human behavior together with the rules of these systems. The question in mechanism design is then how can one come up with rules that—coupled with human behavior in response to these rules—yield outcomes we desire. In a sense, I felt that mechanism design took everything that I loved about game theory and elevated it. It is the closest to the field of engineering within economics in that it provides us with tools to design institutions to achieve desirable outcomes.
Mechanism design is one of the most universal fields in economics, since it provides us with techniques not only to compare the performance of two institutions, but also with a systematic way of devising new institutions that are an improvement on the ones we have. Mechanism design started as a way to formalize post-war discussions about the benefits of a centralized economy, but since then, it has been applied to the design of optimal fiscal policy, social insurance, government auctions, and matching markets.
If you think about it, there are endless possibilities for how to design a particular economic system. What has made mechanism design so successful is that it has the ability to provide a set of tools that allows researchers to tractably understand the consequences of all possibilities without having to actually consider them all. These tools are the ones that we then use to advise governments on issues such as spectrum auctions, companies about ad auctions (Caltech played a key role in both), organ procurement agencies, and public school districts.
I think that the future of mechanism design lies in the intersection of economics and computer science (if you ever wondered how these two things can ever be related, I recommend a great article about the Center for Social Information Sciences at Caltech and a great podcast lecture by Tim Roughgarden at the London School of Economics). Many of the issues that arise in the world of big data, like privacy, virtual currencies, and the sharing economy, will require input from both fields. Incentive issues lie at the heart of these applications, but their scalability will require economists to think more like computer scientists when determining how to address them.
Who are your heroes and mentors in market design or—more broadly—in economics? Who influences you the most personally and academically in your professional career?
Andrea Rotnitzky, who was my undergraduate statistics professor, was an enormous influence in my academic life. She understood, probably even before I did, my academic calling and set me on the path to graduate school. Alejandro Manelli was also a huge influence. He introduced me to mechanism design and I never turned back.
And of course, there are my three wonderful graduate school advisors: Eddie Dekel, Jeff Ely, and Alessandro Pavan. They struck this perfect balance between detailed guidance and allowing me to develop questions on my own. I grew as a researcher thanks to them. They always emphasized the importance of the relevance of my work. Whenever I came to them with an idea, they pushed me to argue why we needed to study this problem and how solving it would be relevant to society. Economics is all about the allocation of scarce resources, so if I was going to devote time to a project, it needed to be useful.
My happiest memories from graduate school involve the three of them. Most importantly, they humanized graduate school for me and this was a great inspiration for my interest in applying to the Faculty in Residence Program at Caltech.
What ongoing projects or publications are you most excited about?
I am very excited about my ongoing work on school choice in Pasadena, with Federico Echenique and Adam Wierman, and on the allocation of deceased donor organs, with Federico Echenique, Matt Shum, and Yi Xin. The same question underlies these two seemingly different applications: how do you design algorithms to allocate children to schools or organs to patients, when you need to take into account the participants’ incentives? In this case, the main worry is that participants have an incentive to wait: for better school seats or private school admissions in school choice; for better quality organs in the allocation of deceased donor organs. This creates congestion in public school admissions, and increased waiting times in the allocation of deceased donor organs. We aim to devise algorithms that on the one hand, help participants find a good allocation and at the same time, balance everyone’s welfare in the system.
How does Caltech compare to other institutions you have worked in?
Caltech is my first job as an Assistant Professor. Previously, I was at Yale for a postdoc and at Northwestern as a graduate student. Something that has always struck me about Caltech is how inspiring it is; it is bustling with energy. Einstein’s quote “Play is the highest form of research” always comes to mind when I think about how to describe Caltech.
What are the biggest challenges in modeling real-world markets in order to solve for equilibria, devise optimal mechanisms, or achieve dynamic stability in the context of matching markets?
I believe that the two main ones are finding the right things to abstract from and the computational complexity of the solution you want to implement. An overly detailed model will give you an accurate picture, if you can solve it. Learning to isolate what matters or not for a particular problem is very important. On the one hand, tractability is key to be able to use the model to actually say something. On the other hand, you may want to use the solution to the model to design a mechanism or algorithm for a given market. Often, the model’s solution will not be easy to compute or implement at a large scale. But if your model isolates the key principles behind your problem, you will have a useful guide in terms of where you can simplify your implementation and where you cannot.
What advice would you give to young female students aspiring for a successful career in economics? In academia?
At the risk of saying something obvious, you should work on something you are passionate about. Figuring out what this is is not easy, and of course things like “not many women work in this field” certainly do not help. Research requires a bit of obsession: you will be thinking about how to solve the question you posed every waking hour of the day, so it should be something you like and not something someone told you you could do.
Another important piece of advice regards time management: the more you progress in your career, the less time you have for your own research. Learning to carve out time to develop your research program is as important as the rest of your academic responsibilities, but usually this is the time that goes out of the window first when our schedules become busier.