Causal inference reading group 2020

The reading group will last for the whole semester, and will cover the following topics, mostly related to causal inference, or endogeneity problems in an OLS setting:

  1. Robust and clustered standard errors to achieve robust inference in an OLS model
  2. Instrumental variable approach to attack the endogeneity problems wide spread in OLS models
  3. Panel data with fixed effects to solve for the unobservable heterogeneity problem, an important source of endogeneity issues
  4. GMM approach for dynamic panels, where fixed effects lead to inconsistent estimates using OLS
  5. Difference-in-difference approach for policy evaluation or shock identification, one of the most popular method to overcome the endogeneity issues in recent years
  6. Regression discontinuity design, an equally popular method for endogeneity problems in recent years
  7. Quantile regression, especially useful to analyze inequality issues and uncover heterogeneous effects of explanatory variables

Registration

Contact me directly via Email, and the registration deadline is 9/16. The reading group is mainly for junior and fourth year undergraduates, seeking to write a summer camp paper or dissertation.

Prerequisite:

  • Math: caculus, linear algebra, probability
  • Stats: statistics, elementary econometrics (familiar with multi variable OLS estimation)
  • Software: R, Stata

Each participant is required to contribute at least one presentation throughout the reading group

Reference books

  • 邱嘉平,《因果推断实用计量方法》,2020 (REQUIRED)
  • Angrist and Pischke, Mostly Harmless Econometrics, 2009
  • 赵西亮,《基本有用的计量经济学》,2017

Organizaiton and introduction

We meet every Thursday afternoon 2:00 – 4:30 pm in Big Data Institute (BDI). The first group meeting will take place on September 17th.

We will have 2 – 3 presentations each week. All participants need to present at least once to contribute to the reading group.

The following is a rough schedule:

  1. Presentations of own research by participants from the previous year, 2 – 3 weeks.
  2. Presentations by new participants (mostly first year graduates and third year undergraduates) on causal inference methodology, 4 – 5 weeks.
  3. Presentations by contribution, on research plans including the empirical questions, data, methodology, and most importantly, key references from the literature, 4 – 5 weeks.
Use the required PPT template to compose the presentation slides Download
Check the Introduction Slides first CIRG2020Intro Download
For guidelines on composing academic slides, see Link Download

Presentations

章节主题内容主讲人日期
第1章因果推断概念王健10/22
第2章线性回归基础第1-2节相耐汀10/22
第2章线性回归基础第3-4节蔡诺璇10/22
第3章线性回归运用第1-3节李非凡10/29
第3章线性回归运用第4-7节李思芃10/29
第4章标准误差第1-3节李浩芸10/29
第4章标准误差第4-6节曾卉琪11/5
第5章处置效应第1-4节江梦瑜11/5
第5章处置效应第5-8节陈远致11/5
第6章匹配方法第1-4节朱思颖11/19
第6章匹配方法第5-7节侯天宇11/19
第7章匹配方法与回归方法对比叶立欢11/19
第8章面板数据第1-4节陈露滢11/26
第8章面板数据第5-7节成思扬11/26
补充动态面板GMM郭倩美11/26
第9章双重差分第1-3节蒋涵琦12/3
第9章双重差分第4-5节王子萱12/3
第10章工具变量第1-3节马兆星12/3
第10章工具变量第4-7节李竞开12/10
第11章样本选择模型第1-3节杨宇彤12/10
第11章样本选择模型第4-6节瞿博洋12/17
第12章断点回归龙欣雨12/17