Dr Valentino Larcinese CYBER

This course is available on the MRes Political Science and MSc in Political Science and Political Economy. This course is available with permission as an outside option to students on other programmes where regulations permit.

A good knowledge of statistics and of the generalized linear model at the level of MI452.

This course provides an introduction to the most commonly used methods for causal inference in the social sciences. After reviewing the rationale for simple OLS estimation, particularly focussing on the conditions for a causal interpretation of the coefficients, the course will cover matching estimators, instrumental variables, panel data, differences-in-differences, regression discontinuity, quantile regression. Students will be stimulated to think in experimental terms and the main theme will be how to replicate or get close to the experimental ideal of natural sciences by using non-experimental observational data. The lectures will present the techniques and illustrate their applications by making extensive use of the most up-to-date empirical literature in political science. The seminars (tutorials) will be conducted in a computer room and will provide an opportunity to learn how to use the methods with the statistical software Stata.

20 hours of lectures and 5 hours of seminars in the LT. 4 hours of lectures in the ST.

There will be two pieces of formative work to be submitted collectively to the lecturer in the form of short essays. Students will be divided into groups of 4-5 people (groups change each time). Each group will receive a dataset and will be asked to use some specific methods to answer an assigned question.

Angrist & Pischke: Mostly Harmless Econometrics, Princeton University Press 2009.
Stock & Watson: Introduction to Econometrics, Pearson 2007 (Second edition)

Exam (100%, duration: 2 hours) in the main exam period.