Research Interests

My research spans econometrics, judgement and decision, and analysis of social policy. Most of my early and some of my recent work concerns analysis of discrete choice, seeking to contribute by developing new econometric methods and by studying particular choice settings. I introduced the maximum score method for semiparametric analysis of choice data in my dissertation and I developed methods for inference from choice-based samples soon after. I have more recently developed survey research methods for collection of probabilistic expectations data that characterize how decision makers perceive their opportunities. My empirical work has devoted particular attention to analysis of schooling decisions. In the late 1980s I began a program of research on partial identification that has since become the main focus of my work. Concern with the credibility of traditional approaches to inference with missing outcome data led me to ask what partial observability reveals about outcome distributions if nothing is known about the missingness process or if assumptions weak enough to be widely credible are imposed. My findings on inference with missing outcome data provided the foundation for study of conditional prediction and analysis of treatment response under weak assumptions. My findings on these subjects then provided the basis for my most recent work on social planning under ambiguity. Here, I investigate the problem of making treatment choices with partial knowledge of treatment response.

Membership Type


Election Year


Primary Section

Section 54: Economic Sciences

Secondary Section

Section 32: Applied Mathematical Sciences