"In networks," says economic theorist Teddy Mekonnen, "how you connect one person to another really matters." When he came to Caltech as a Linde Postdoctoral Fellow, Mekonnen joined a network designed in a way he had never seen.
The Ronald and Maxine Linde Institute of Economic and Management Sciences and its Social and Information Sciences Laboratory (SISL), which are supported by gifts to Break Through: The Caltech Campaign, brought Mekonnen into daily contact with researchers in computation, the humanities, and the social sciences.
Mekonnen initially had a straightforward vision for his postdoctoral appointment. He aimed to work with leading economic theorists and continue promising studies he had already begun related to the design of incentives and mechanisms that help strategic individuals reach a mutually desirable outcome. He achieved those goals, and also expanded his research to explore theory related to search and matching scenarios. Mekonnen recently posited that when people newly enrolled in health insurance search for primary care physicians, they achieve better outcomes in the aggregate if they randomly sample the doctors available in their networks than if they screen doctors based on published rankings.
Mekonnen's vision has changed because Caltech's interdisciplinary environment gave him new views of his own field and others.
"Being at Caltech has opened my eyes," Mekonnen says. "We economic theorists come up with problems that are real and relevant, but some of our solutions are impractical, with implicit assumptions of infinite time, resources, or data. When you use computer science to approach economic questions, such as matching problems related to apps for ride-sharing or dating, those assumptions matter. The grand theories that we have … might fall a little flat if you can't implement them."
At Caltech, Mekonnen's friendship with a machine-learning expert across the hall has given him the confidence to ask rudimentary questions or pose preliminary ideas. He has enjoyed long days brainstorming with computational economists at a floor-to-ceiling whiteboard in a shared studio space near his office. "It's nice to bounce problems off of each other and see whether they're interesting, or even solvable," he says.
When Mekonnen first considered the link between computation and economics, he vaguely imagined data-science jobs at Amazon or Uber. Now, he interacts with SISL computer scientists who focus on topics in matching, network design, and game theory that he once saw as the domain of economists. He has become aware of deep theoretical questions where the two fields intersect.
Mekonnen routinely learns about scholarship in computation, engineering, and political science with relevance to his own studies at the many seminars and conferences that Caltech encourages him to attend. He and other SISL members also dine with seminar speakers and discuss the overlap in their interests.
"Interdisciplinarity doesn't necessarily mean you're completely immersed in two fields," Mekonnen says. "It could be realizing that there is a similar underlying thread in the kinds of research questions people ask and exploring tools from one field to answer questions from the other. That happens all the time here."
The Power of Lunch
SISL's cross-disciplinary matchmaking works because, as Mekonnen has discovered, it reinforces the curiosity and friendliness that are commonplaces at Caltech, an intimate community where faces quickly become familiar.
The first time Mekonnen walked onto campus, he encountered assistant professor Omer Tamuz, who remembered his postdoctoral interview and immediately asked him to lunch. Mekonnen also has gained perspective from gatherings where faculty, postdoctoral scholars, and students mingle. "At happy hours here on Fridays, I've talked with people who do philosophy and history, including heated debates about whether the assumption of completely rational players in economics is even practical and what that means."
At Caltech, such conversations can lead to research projects. That is how Mekonnen began a collaboration with Federico Echenique, the Allen and Lenabelle Davis Professor of Economics.
Mekonnen had asked SISL scholars to brainstorm with him on an employment-theory problem at one of the group's informal weekly lunches. That lunch discussion led Echenique to see Mekonnen as a potential partner in research about the Peter Principle, the maxim that organizations promote employees until they reach the level at which they are no longer competent. Because Mekonnen's philanthropically endowed postdoctoral fellowship enables him to conduct research that has no grant funding, he was free to accept Echenique's invitation.
Together, they are studying a perplexing trade-off: When organizations cannot easily fire or demote employees they have promoted one level too far, the organizations lose productivity. However, that loss may be worthwhile, because organizations that offer secure, well-paid work to internally promoted employees gain productivity from the high volume and quality of work that employees do to get those promotions.
The Explorations Ahead
Mekonnen feels confident that the breadth he gained at Caltech will
help him uncover central questions and solutions as he develops economic
theory in the future. He anticipates opportunities to carry forward
what he has learned from his Caltech network: the value of chance
conversations, not really by chance.
Written by Break Through: The Caltech Campaign