92816 - Analysis Of Panel Data: Methods And Applications

Learning outcomes

The lectures aim to provide an overview, both methodological and applied, of the econometric models for panel data. In a panel dataset we have at least two dimensions of the observations: classically we can have time-series cross-sectional data (CSTS); recently we talk about multilevel data (for example, companies within industries within regions within countries, observed over time). Models can be both static and dynamic. At the beginning of the course, to ease the comprehension and to introduce important topics, some introductory econometrics will be revised. At the end of the course, students are expected to be able to: (1) understand, estimate and interpret the main models for panel data; (2) use an econometric software and conduct their own research; (3) critically evaluate some applications of panel data presented in the empirical literature.

Course contents

The schedule of the program goes along the following points:

1. Introduction to panel data; longitudinal data, cross-section time-series data, multilevel data.
2. The fundamental role of heterogeneity: how to deal with it.
3. Static panel data models: two or more levels, the error components; fixed and random effects; robust Hausman test.
4. Mixed and hierarchical models.
5. Introduction to dynamic panel data model.

Readings/Bibliography

The material (articles, notes, programs and data-sets) will be distributed during the lectures and make available on the platform Virtuale.

Students are invited to have a look at the textbooks listed below (the order indicates an increasing level of difficulty):
Wooldridge J.M. 2020 Introductory Econometrics. A Modern Approach, Cengage, 7th Edition, Ch 13-14;
Verbeek M. 2017 A guide to Modern Econometrics, Wiley, 5th Edition, Ch. 10.
A preliminary look at chapters 1, 2, 3 and 4 of the Verbeek textbook could be a useful recap of the required basic econometrics.

Teaching methods

At each step of the course, the theoretical methodologies will be accompanied by hands-on empirical applications with an econometric software (Stata available using the CAMPUS license and students' university credentials).
At the end of the course, participants will be able to understand the main literature on basic panel data, and to model and estimate their own issue of interest, according to the problems at hand: static versus dynamic approaches, POLS, FE; RE; CRE, GMM.

Assessment methods

Some homework will be assigned during the course. These 'exercises' are intended to reinforce the concepts seen in class, to replicate the empirical analyses carried out together on the data and to familiarise students with the software. Students are encouraged to work together on the homework (groups of up to 4 participants are allowed). Each assignment handed in by the deadline will offer an additional evaluation. The final exam, which is compulsory, involves the presentation and class discussion of an empirical analysis on a research topic provided by the lecturer.

Not attending students will have a written examination on theoretical and applied issues, with open ended questions aimed at assessing their capacity in understanding models' specifications and estimated results, and in evaluating the strengths and weaknesses of alternative estimating methods.

Registration on Almaesami is mandatory for the final exam.

Evaluation of the home and class assignments:
3/28-30L: the write-ups reveal a proficient and exhaustive knowledge of the issue, and an excellent ability of comprehension and analytical execution.
2/24-27: from the write-ups it emerges an appreciable degree of knowledge of the issue, and a good ability of comprehension and analytical execution.
1/18-23: the write-ups are untidy, with theoretical and methodological inaccuracies.
0/<18: wrong or not delivered write-ups.

Teaching tools

Theoretical lectures are associated with working sessions (students can have their own laptop); during them students will receive the suggestions needed to run their own empirical analysis. The data-sets and the programming files to perfom applied analyses will be provided during the lectures. The distributed material (articles, notes, programs, and data-sets) will be make available on the Virtuale platform.

STATA software HERE

Office hours

See the website of Maria Elena Bontempi