Division of the Humanities and Social Sciences

Testing Models with Multiple Equilibria by Quantile Methods. This research is joint with Ivana Komunjer of UCSD. We explore a structure that provides testable implications of models with multiple equilibria.

Many economic models possess multiple equilibria, which poses a severe challenge for deriving testable implications. The standard practice is to impose additional assumptions, so that the model has a unique equilibrium. These additional assumptions are often very strong, and independent of the desired economic explanation.

We show that, when there are complementarities between the models' endogenous and exogenous variables, the model has testable implications. The implications are not on the conditional mean (regression) but rather on the extremes and quantiles of the endogenous variable. We present the limiting distribution of test statistics, and a likelihood-ratio test, under a fairly general structure on errors.

We show how our methods can be used for inference in macroeconomics (growth theory), industrial organization and other economically relevant applications.

Testing Models of two-sided markets. I explore the testable implications of the theory of two-sided markets.

There is an important literature on the testable implications of single-person decision theory, starting from Afriat's theorem. And of general-equilibrium model, starting from the Sonnenscheing-Mantel-Debreu Theorem. There are also recent efforts to study the testable implications of Nash equilibrium. It seems important to also study the same question in two-sided models, as they may offer a general model of markets that is nevertheless testable.

I am currently working on the refutability of the marriage market---two-sided contracting without transferable utility. I can prove that the theory does have testable implications, and I am optimistic that I will be able to characterize the sets of matchings that can arise as the core of a marriage model. I hope to be able to extend the results on the marriage market to markets with prices and transferable utility.