Skip to content

The Econometrics of Causality (ECO-AD-ECONCAU)

ECO-AD-ECONCAU


Department ECO
Course category ECO Advanced courses
Course type Course
Academic year 2024-2025
Term BLOCK 1
Credits 1 (EUI Economics Department)
Professors
Contact Simonsen, Sarah
Sessions

10/09/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

12/09/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

17/09/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

19/09/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

10/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

11/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

15/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

17/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

22/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

24/10/2024 8:45-10:45 @ Seminar Room B, Villa la Fonte

Purpose

This course will feature 10 2-hour lectures.
  1. Introduction to Causality and Causal Inference (1 lecture)
    • Setting up the problem: the potential outcome framework
    • The role of the assignment mechanism
    • Design and analysis of randomized experiments
    • Estimands and modes of inference: Fisher exact tests; Neyman frequentist per- spective; Regression analysis; Bayesian model-based imputation
    • Examples
  2. Introduction to observational studies under unconfoundedness (2 lectures)
    • The role of the propensity score
    • Designing observational studies: matching, weighting, trimming
    • Analysis of observational studies: stratification, weighting estimators, matching estimators, methods based on the outcome models and regression.
    • Methods combined: bias corrected estimators, doubly robust estimators
    • Sensitivity analysis
    • Enhancing causal inference with machine learning in high dimensional settings and heterogeneous effects
  3. Instrumental variables and Principal Stratification (3 lectures)
    • Homogeneous vs heterogenous causal effects
    • IV and RCT with noncompliance - revisited
    • Bayesian IV analysis: relaxing some of the assumptions
    • Point, partial, weak identification of causal effects
    • The role of covariates
    • Principal Stratification and application to censoring by death
    • IV and beyond: principal stratification
    • Principal Stratification and application to censoring by death and mechanisms
  4. Regression discontinuity designs (1 lecture)
    • The identification strategy: continuity vs local randomization
    • Sharp RDD
    • Fuzzy RDD and the local LATE interpretation of RDD
    • Graphical analysis, assessing identification assumptions
    • Bandwidth selection
    • Multiple thresholds, multiple forcing variables
    • Examples
 
  1. Difference in difference, synthetic controls and beyond (3 lectures)
    • DID and extensions (e.g., CIC, Synthetic DID)
    • Lagged dependent variables
    • Synthetic controls and permutation inference
    • Some recent developments (e.g., Matrix Completion, Time Series)
 

Exercise classes

Three practical sessions on Causal Inference methods using R.
 

Description

Teaching material

  • Imbens G. W., Rubin D. B. (2015) Causal Inference for Statistics, Social, and Biomed- ical Sciences, Cambridge University Press
  • Articles in journals.
  • Lecture notes by the instructor.
 

Final exam and Grading

There will be three take-home assignments (simulation and real data exercises).
Back to Overview
 

Register for this course

Page last updated on 05 September 2023