Panel Data Analysis

The course consists of three parts

Part I involves a thorough discussion of the logic and assumptions underlying panel data methods. You’ll learn how the development of more advanced methods is driven by the need to address potential violations of these assumptions. 

Part II focuses on the various statistical approaches and 'tricks' available to social scientists to deal with such violations and problems hidden in their data, allowing them to obtain estimates of effects that are as close as possible to the true causal effects. 

Part III focuses on applying the wide range of panel data methods discussed in the previous parts to substantive research questions of interest. You will learn how these methods can be used to provide answers to your own research questions.

Monday

The course starts with a quick review of OLS regression, emphasising the key assumptions required for the OLS estimator to be the ‘best’ estimator. We then move on with simple panel data methods, namely two-period panel data analysis and first differencing.

Tuesday

We focus on slightly more advanced methods for estimating unobserved effects in the context of panel data analysis. We introduce fixed and random effect estimators, discuss their properties, and the assumptions needed for them to be valid. With these foundations, we then study a relatively new correlated random effects approach, a synthesis of fixed effects and random effects methods which has been shown to be very useful.

Wednesday

We begin with a lab session, during which we put into practice the methods introduced in the previous two days. We carry out these analyses in Stata and/or R and learn about the interpretation of the results. This is followed by a lecture on the instrumental variable (IV) method, which deals with violations to the strict exogeneity assumption.

Thursday

We move on to more advanced panel data methods that address further violations of the standard OLS assumptions, including clustered and robust standard error, panel-corrected standard error (PCSE) estimates and dynamic panel methods (Arellano-Bond and system GMM estimators).

Friday 

A second lab session and a seminar. If you want to earn extra credits, you can present your research or research proposal that uses panel data methods and receive feedback from the Instructor and fellow participants.

How the course will work online

Reading materials will be provided via CEU’s e-learning system Moodle, in case you do not have access to physical libraries. Readings will be supplemented by around eight hours of pre-recorded lectures covering the theoretical part of the materials. There will be around seven hours of live sessions during the week.

The live sessions are mainly devoted to:

1. Q&A with the Instructor
2. lab sessions during which the Instructor will demonstrate how to use Stata and R for panel data analysis
3. student presentations during the seminar on the last day.

The live components will also include getting to know each other on the first day and, potentially, an online social event. You are welcome to set up appointments with the Instructor or TA for one-to-one consultations, during office hours.

This course builds on OLS regression and extends it to data with a panel or TSCS structure. You should be familiar with basic theories of OLS, up to multiple regression.

To participate meaningfully in the lab sessions, you should also have basic knowledge of Stata and/or R. 

Some background in linear algebra would be helpful but is not required or assumed.