Welcome to 2W: Advanced Time Series/Panel!

I hope you are ready!

In the next two weeks we will set forth as comprehensive a treatment as we can in time series and panel data in ten sessions. With coffee in hand and the computer getting ready, let’s dig in!

A brief summary for the Instructor’s Presentation

Week 1: Time Series Day 1
- Stationarity and Unit Roots
- ARMA structures
- Dynamic Linear Models and their Interpretation
- VAR and Cointegration topics
- Equation balance and recent controversies
Week 2: Panel Data
- Fixed and random effects and Within/Between
- Estimating Models, xtmixed, xtreg, etc.
- STADL Up!
- Missing data
- GLM Extensions of Panel Models
- Dynamic Panel Data
- Causal Inference in Panel Data: TWFE/DID

Digital Resources

Depending upon your software of choice, you will want to have a look over the relevant manuals.

For Stata ts and xt.
For R there is timetk and fpp3 and plm along with panelr.
A github with the course data..
A github with this website..
Robert W. Walker’s website

A Useful Model Translator

The Social Science Computing Cooperative at the University of Wisconsin has a nice, albeit incomplete, rendition of model compatibilities with syntax in R and Stata.

Topics of Interest Cut by Time

We will read selections from a broader controversy in Political Analysis on Time Series and error correction models; the full list of papers can be found here as a large part of the Winter 2016 issue 24(1).

There is a digital special issue of Political Analysis on Regression Discontinuity and Time Series Cross-Section data. For day 6 of the class, the paper by Plumper and Troeger (2019) about fixed effects regression is definitely worth considering.

Factor augmented dynamic panel data models get an interesting treatment in Dynamic Panel Analysis under Cross-Sectional Dependence by Khusrav Gaibulloev, Todd Sandler and Donggyu Sul