Description

The course introduces to the statistical methods used to interpret economic data. Specific topics include elements of probability theory, statistical tests, simple and multivariate regression, instrumental variables, and methods for qualitative variables. Examples complete the theoretical course and teach the practice of econometrics on real data using statistical software (Stata).  Problem sets (with solutions) and readings are proposed to the students that will be discussed in class and/or tutorials.

Syllabus

Prerequisites

This course requires some proficiency in basic calculus including the ability to calculate derivatives and integrals. Linear algebra is useful in econometrics but not for this course. 

Lecture notes

1. Random variables

   • Our broken economy in one chart - The New York Times

   • Problem set 1

2. Regression analysis and Ordinary Least Squares 

   • Problem set 2

3. Statistical inference

4. Asymptotics

5. Panel data and Difference-in-difference

   • Example: Breda and Ly (2015)

6. Causality, endogeneity and IV 

   • Example: returns to education

7. Maximum likelihood and qualitative dependent variables

Evaluation

Participation in the tutorials offered by a teaching assistant is greatly encouraged. The final grade will take this participation into account. 

The final exam takes the form of an empirical essay on a subject chosen by the students. The essay has to be submitted no later than two weeks after the last lecture. These papers should aim at testing or establishing causal effects, and use econometric techniques such as regression analysis, counterfactual analyses, instrumental variables, regression discontinuity, diff-in-diff, etc. A critical analysis concluding the essay is most welcome for literature reviews. 

Recommended textbooks

Probability and statistics:

Casella, G.C. and R.L. Berger (2002), Statistical Inference, Brooks/Cole, Edition: 2nd 

Econometrics:

James H. Stock and Mark W. Watson (2010), Introduction to Econometrics,  Edition: 3rd

Otherwise Wikipedia...

Statistical softwares

I recommend STATA. Students versions are available at a reasonable price. STATA is installed on the PCs at 28 rue des St Peres.

There are lots of STATA tutorials on the web. Lots of Stata resources can be found on there.

Data, Stata do files, and R programs

The folder contains a Stata tutorial, a data file drawn from the French Labour Force Survey 2002 (last issue before major sampling change making the survey less suitable for illustrating the course), and a set of Stata do-files and log-files illustrating different chapters of the course.