B2163 - ECONOMETRICS OF PANEL DATA

Learning outcomes

At the end of the course, students know the most appropriate estimating techniques for dynamic panel data models, both microeconomic (large and with more than one cross-sectional dimension) and macroeconomic (over a long time span). Specifically, they can: - critically understand theoretical and applied aspects of the vast literature based on dynamic panel data models; - apply dynamic panel data models techniques to their own analyses by programming specific routines using the STATA software.

Course contents

Panel data, be it longitudinal data, cross-sectional time-series data (CSTS) and multilevel data, permeate empirical research in many fields of behavioural sciences, from economics to psychology. The aim of the course is to provide a methodological and applied overview of panel data econometrics.

Topics covered in this course:

1) strengths and weaknesses of various estimation methods: POLS, FE, FD, BE, RE;

2) double endogeneity and CRE;

3) dynamic models and the Generalised Method of Moments (GMM).

4) non-stationarity and unit roots; common correlated effects; slope heterogeneity (MG and PMG, meta analysis).

Clearly there are necessary prerequisities, specifically for Erasmus students:

1. At this page [https://corsi.unibo.it/2cycle/lmec/course-structure-diagram/piano/2024/5977/000/000/2024] take a look at the content of Econometric Methods, Microeconometrics, Macroeconometrics courses. Please assess your knowledge of Ordinary Least Squares, Generalized Least Squares, Instrumental Variables, Fixed Effects, Random Effects, Unit Root test, Moving Average, Error Correction Mechanism.

2. A knowledge, at least basic, of STATA software is also required

Readings/Bibliography

Discussion papers, commented notes & slides, stata programmes and datasets will be available on the VIRTUALE platform and explained during the lectures.

Useful textbooks:

Wooldridge J. M. 2010 Econometric Analysis of Cross-Section and Panel Data, Cambridge Mass. MIT Press, 2nd ed.;

Arellano, M. (2003) Panel Data Econometrics, Oxford University Press;

Baltagi B. H. (2021) Econometric Analysis of Panel Data, Springer, 6th ed.;

Hsiao, C. (2014) Analysis of Panel Data, Cambridge University Press, 3rd ed.

Teaching methods

During the lectures, the methodologies will be accompanied by empirical applications based on Stata econometric software (available using the CAMPUS licence and student university credentials).

You will be involved in implementing and commenting the empirical analyses, so have your laptop with you. Of course, attending and participating the lectures is highly recommended.

Assessment methods

During the course, you will be given an assignment to do at home (alone or in a group of up to three people) and to post on VIRTUALE at the deadline set by the lecturer. This is a research question relevant to the topics seen in class (the lecturer will provide datasets and questions), to be developed empirically, giving reasons. The homework contributes 30% to the overall assessment.
The final exam, compulsory and subject to registration on AlmaEsami, contributes the remaining 70%. It consists of an individual presentation in the classroom (maximum time 15 minutes) of an empirical analysis conducted on panel data. You can decide together with me which topic you would like to explore in more depth or agree that I will assign you a research question. The assessment focuses on your ability to analyse the topic in a theoretical and applied manner.
The overall grading scale is:
30 cum laude: impeccable work
28-30: excellent work, which demonstrates a thorough and comprehensive knowledge of the topic, as well as excellent comprehension and analytical skills.
24-27: good work that denotes an appreciable degree of knowledge of the subject.
18-23: there is disorder, with theoretical and methodological inaccuracies.
<18: failure of the test

Teaching tools

Theoretical lectures combined with empirical work sessions

Help in programming the software

Discussion of results and critical evaluation of the empirical literature

Capacity to select the best method for estimating different types of panel models.

Virtual room on TEAMS (just in case you cannot attend physically one of the lessons) for direct chat communication

Office hours

See the website of Maria Elena Bontempi