Establishing relationships of cause and effect is a complex task,
especially in samples from human populations where subjects may select themselves into treatment and control in ways that we as researchers do not fully observe. Ideally, we would like to explore such causal relationships in the controlled environment of a laboratory, and carefully assign subjects to well-specified interventions (the causes) and observe their outcomes (the effects). In reality, however, for most of the questions that interest decision-makers we cannot run such experiments because they are too expensive or because they are impractical or unethical. Examples are common in business research and in the medical and social sciences. For instance, what is the effect of ad campaigns in online media on sales? What is the effect of c-sections on mortality of very low birth-weight premature babies? Are prisons schools for crime or does incarceration reduce recidivism? A theory of causal inference is important for decision-making in varied research areas spanning from public policy to robot learning applications. In this course we will study mathematical assumptions and statistical methods for estimating causal effects from controlled, randomized experiments and from observational studies. This course is different to other econometric courses in that it emphasizes concepts and tools to approximate experimental design in observational studies. Topics for the course will include the potential outcomes framework for causal inference; randomization inference; matching, propensity score and regression methods for controlling for observed confounding; sensitivity analysis to unobserved bias; instrumental variables; and regression discontinuity designs (see the tentative timetable below). This course aims to be a platform for empirical research and help the student to develop a publishable research paper. By the end of the course the student should be able to critically assess published works that claim to estimate causal effects from observational data.
Assistant Professor of Business
Professor Zubizarreta teaches statistics to MBA and PhD students. From a methodological perspective his research focuses on new statistical methods for causal inference (impact evaluation) in randomized experiments and observational studies. From a substantive perspective, he is very interested in using these methods to address questions in health care and public policy. Besides the Division of Decision, Risk and Operations, Zubizarreta is also an...