You can download the notes for the class:
Notes
by methods
Notes
by example
Class schedule
- Linear models for cross-section (running example will be labor
supply)
- Clarify concepts such as data, model, identification, estimation and
inference
- Study theats to identification (endogeneity, measurement error,
simultaneity)
- Finite sample gaussian inference & asymptotic inference (central
limit theorem)
- Lab
on labor
supply
- Heckman (77) paper
- Extras: clustering of standard errors, multiple testing, weak
IV
- Multinomial decisions (we will use transportation and monopolistic
competition)
- Introducing Maximum Likelihood Estimation
- Estimation of multinomial logit
- McFadden (74) paper
- Maximum score estimation, Manski (75) paper
- Lab
on discrete
choice
- Extras: Boostrap, Marriage market with Choo and Siow (03) paper
- Unobserved heterogeneity
- Revisit multinomial problem and IIA
- Random effect model and EM-algorithm
- Extras: introduction to Variational Auto Encoders VAE
- Topics in non-parametrics and high-dimensional econometrics
- model complexity and bias/variance trade-off
- many regressors (application to network of peers)
- non parametric regression
- Dynamic programing
- Introducing Bellman equations (through cake eating problem)
- Study the estimation of dynamic programing
- Lab
on dynamic
discrete choice
- Rust (74) paper
- Measuring sorting in the labor market (optional, if time allows)
- incidental parameter problems
- Fixed-effect approach, bias correction
- Group fixed effect approach
Homeworks
You can submit your homework either as solo or as a group of two. To
communicate about your homework, create a channel in our slack group
with a name starting with grp_ and then with the initials
of the members of your group. For instance the TA (Arjun) and myself
(Thibaut) would be grp_at. Then please add Arjun and myself
as members of that group. I recommend you make your channel private.
Submit the analytical parts directly to this slack group.
Computational homeworks should be submitted as a self contained R
markdown file. Please submit the Rmd file as well as the generated html
output.
End of class project
- The project can be done in groups of three. The goal is for you to
take an economic or policy question and to develop a DGP for that
question as well as a method adapted to answering that question given
data. The project should highlight the potential pitfalls of given
methods.
- 1 page proposal is due on week 7 or 8. This should include a
question and a DGP.
- final submission of the project will be the week after finals.
Midterm
- The midterm will in week 6
or 7.
Grading scheme
Overall grade will be 2/5 homeworks, 2/5 midterm, 1/5 project.
Books
These are some books that can be helpful to dig into more details of
the topics covered in class. These books are not
required for the class.