The Ins and Outs of Matching: A Conversation with Economist Laura Doval

Laura Doval, assistant professor of economics, is a new arrival on campus. She received her PhD in economics from Northwestern University in 2016, and came to Caltech from Yale University, where she spent a year as a postdoctoral associate at the Cowles Foundation.

Doval's research focuses on dynamic matching markets—those in which prices aren't the primary drivers and in which each side of a transaction has to be chosen by the other side. She studies how the characteristics of these markets affect the ability of its participants to find their match, and how those characteristics might be tailored to increase efficiency or fairness.

We sat down to talk with her about the dynamics at play when a person selects a kidney transplant from a deceased donor; how to make waiting lists work better; and her dog Pancho, who can sometimes be found sleeping on the floor in her office.

What are your research interests?

I am interested in how people behave and interact in dynamic environments. Dynamics can arise for different reasons. For instance, it could be that learning about the value of different alternatives takes time, like when we have to decide between colleges we have been admitted to, or that some alternatives may become available with time, like when a patient waits for an organ from a deceased donor. Sometimes we can choose whether we want to make a system static or dynamic, like when a school chooses whether to have early and late admissions.

My research explores questions of optimal consumer search, the feedback between rules and incentives to delay decisions in dynamic matching markets, and how we can use the timing of decisions to generate incentives.

What is a matching market?

Matching markets are a particular kind of market where prices don't do all the work and you cannot just choose what you want; you also have to be chosen. For instance, it is not enough that a student can pay tuition to be admitted to Caltech. Caltech has to offer them admission. Conversely, if Caltech wants to have a student in their ranks, the student has to choose Caltech from the schools available to them. This same logic applies to doctors applying for residencies in hospitals and children applying to a private high school. In all these instances, they also have to be chosen by the counterpart.

There are markets where there are no prices, such as allocating live transplants—we need to decide who gets what, but we cannot resort to a price system to make these decisions.

Those are examples of static matching markets since all participants are in the market at the same time: Each year Caltech has some slots that are offered to that year's desired applicants.

In contrast, in dynamic matching markets like child adoption, the allocation of deceased-donor organs, and public housing, participants and objects are entering the system over time, so your most preferred match may only become available in the future. Participants in these markets have to decide whether to match today, or wait until tomorrow to see if they can find a better match.

So, you're interested in what makes these markets work well?

These markets are operated by institutions that determine who gets offered what. Different institutions implement different outcomes, and these outcomes might have desirable properties, such as efficiency or fairness. My research focuses on understanding when the interplay between dynamic incentives and institutional rules results in desirable outcomes.

In dynamic markets, we face two new challenges. First, what does efficiency or fairness even mean when the population is evolving over time? Ideally, we would like a system that takes into account not only the welfare of the current participants, but also the welfare of future ones. For instance, if I decide to reject an offer to be matched to wait for a better one, I may crowd out the opportunities of future participants. The second challenge is how to design a system that actually implements the efficient allocations, assuming we settle on a definition of efficiency.

An example might be useful. Suppose we have a single patient on the waiting list today and one kidney in the system. This kidney, because it is a deceased-donor kidney, cannot be stored until tomorrow. Imagine the hospital knows that a better-quality kidney will become available tomorrow, but also that a new patient will join the waiting list tomorrow. If both patients are compatible with the kidney, intuitively, efficiency requires we match today's patient with today's kidney and tomorrow's patient with tomorrow's kidney. But if the rules are that if today's patient says no, he gets offered tomorrow's kidney, then we won't be able to achieve this outcome: Today's patient will want to wait and tomorrow's patient may end up with no kidney. So, the rules we write down affect our ability to implement desirable allocations, since they determine participants' decisions to wait or not for better assignments.

How did you become interested in economics?

I was brought up around science—my mom is a biochemist and my dad is a chemical engineer—so I was not inclined toward economics initially. In Argentina, debates about economics are tainted by political ideology, and what's discussed is often opinions rather than facts. So even though I enjoyed the social sciences, I did not see in them the harmony of the scientific method I had grown up with.

When I was still in high school, however, I was given a book on microeconomics written by the father of my boyfriend [now her husband]. This opened up a whole new world to me. Economics provides us with a language and a framework to analyze situations very much like the scientific method: To tackle a problem, we formulate and test precise hypotheses; we dissect problems trying to isolate the organizing principles, and we use the language of mathematics to structure our arguments.

What brought you to Caltech?

Caltech is well known for its research in mechanism design and market design; John Ledyard, Tom Palfrey, and Charlie Plott are pioneers in these areas. This tradition is still going strong. For instance, the Social and Information Sciences Laboratory brings together faculty from economics, computer science, engineering, and mathematics who study the theoretical and computational aspects of network markets and platforms. I can't think of a better place to be a part of.

So, you're excited to be here.

Absolutely. Caltech is an environment that's bustling with energy. Coming to work here reminds me of a quote by Einstein: "Play is the highest form of research." Research at Caltech is exactly like play: it is dynamic, spontaneous, collaborative, and it is fun. You come to work every day knowing you will learn something new.

Tell us about your dog.

I own a husky named Pancho. People often wonder why a husky given Pasadena's warm weather; I don't have a good answer except that he is a really good match.