Syllabus for ECO 651: Econometric Analysis of Panel Data and Limited Dependent Variable Models

Spring, 2018

Course Instructor: Dr. Syed Mortuza Asif Ehsan (SME)

Assistant Professor, Department of Economics

Contact Information:
Office Room: NAC 831c
Phone: +880-2- 55668200, Ext: 1813
E-mail: syed.ehsan@northsouth.edu

Course Information:
Class Location: NAC 401
Class Time: Sunday 7.00 pm - 10:00 pm
Office Hours: ST:10:30 am – 12:30pm & 4.30pm – 5.30pm;
MW: 10.30am – 11.30am; or by appointment

Important Dates (tentative):

 Midterm Exam: March 04, 2018
 First presentation phase and project submission date: April 1, 2018
 Second presentation phase: April 8, 2018

NB: These dates are tentative and are subject to change if situations demand.

Course Description and Objectives

This class aims at learning and applying econometrics at the more advanced level than the undergraduate
econometrics class. The objective of this course is an in-depth examination of the modern econometric
approaches and techniques currently employed to study (i) dynamic relationships and to model the
differences (heterogeneity) among subjects, and (ii) the use of discrete choice models in different settings.
It is an advanced econometrics course, suitable for Masters students with an applied orientation and with
strong focus in constructing and testing models that employ panel data and cross-sectional data with
limited dependent variables. Having successfully completed this course, students will be able to

1. Evaluate the advantages and disadvantages of panel data as compared to other data structure
2. Evaluate the advantages and disadvantages of a collection of disparate panel data techniques as
compared to other methods
3. Formulate appropriate statistical inference to analyze panel data
 4. Extend their econometric knowledge through reading and replicating journal articles with good
Understand and critical perspective.
5. Understand different qualitative dependent variable models and their application in different
economics fields.

Prerequisite: At least one Econometrics course in prior semesters (undergraduate or Masters) is required
to better understand this class. In addition, a better understanding of mathematics, especially linear
algebra, is preferred. Basic knowledge about STATA or any standard statistical software (such as R,
Matlab) is required.

Textbook:

Part I: Panel Data

1. Baltagi, B.H. (2005): Econometric Analysis of Panel Data,John Wiley and Sons, Ltd.

2. Hsiao, C. (2003): The Analysis of Panel Data, Cambridge University Press.

3. Wooldridge, J.M. (2010). The Econometric Analysis of Cross-Section and Panel Data, 2nd Edition,
Massachusetts: MIT Press.

4. William H. Greene (2003): Econometric Analysis (5ed), Prentice Hall.

Part II: Limited Dependent Variable

1. Wooldridge, J.M. (2010). The Econometric Analysis of Cross-Section and Panel Data, 2nd Edition,
Massachusetts: MIT Press.

2. William H. Greene (2003): Econometric Analysis (5ed), Prentice Hall.

3. Maddala, G.S. (1983): Limited-Dependent and Qualitative Variables in Econometrics, Cambridge

University Press.

4. Manski, C.F. and McFadden, D. (ed.) (1981): Structural Analysis of Discrete Data with Econometric
Applications, The MIT Press.

Reference of any additional materials covered will be provided in class.

Student Evaluation:

 Homeworks 20%
 Project & Presentation 20%
 Midterm Exam 30%
 Final exam 30%

There will be two Homeworks due for this course. Homework-1 will be on the topics of Panel Data, and
Homework-2 will be on Limited dependent Variable part.
The Project is to replicate and extend a published paper. You can find an article in a journal (Journal of
Applied Econometrics is preferred) along with its dataset and replicate the published results. While
replicating, find a way to extend the existing idea of the paper. This is a group project of maximum 3
members. You can work individually as well. At the end of the semester (the last class), the project will
be presented in front of the whole class.

Midterm exam will be based on Panel Data Econometrics, and the Final exam will be based on
Limited Dependent Variable part of the course. Please note that Final exam may also contain some part of
Panel Data as Limited Dependent Variable models are also practiced in Panel Data set up.

NO late submission of Homework, and Project will be accepted.

Grading Policy:

North South University grading policy will be followed.

Tentative Topics:

First part: Panel Data models:

1. Introduction

(a) Advantages and Disadvantages of Panel Data

2. Models for longitudinal data

(a) Fixed effects model: Within estimator and test for fixed effects

(b) Random Effects model: GLS/FGLS estimator, Between estimator, computation of individual effects
and test random effects

(c) Random effects correlated with regressor

3. Models with instrumental variables and two-way models

(a) Consistent and efficient IV estimators

(b) Testing of absence of correlation between individual effects and regressors

(c) Two-way models

4. Dynamic Panel Data Models

(a) Inconsistency of LS estimators

(b) The Anderson-Hsiao Approach

(c) The Arellano-Bond approach

 (d) Exogenous regressors

(e) Autocorrelation and specification tests

(f) GMM estimation and parameter restrictions

Second part: Qualitative and Limited Dependent Variables:

1. Models for qualitative dependent variables: Binary choices

(a) Linear probability model

(b) Logit and Probit models: Economic and statistical underpinnings, ML estimation, interpretation
of coefficients and marginal effects

(c) Goodness of fit and prediction

(d) Logit and Probit models for panel data

2. Models for qualitative dependent variables: Multiple choices

(a) Multinomial and Conditional Logit Models

(b) IIA Assumptions

(c) Nested Logit models

(d) Models for ordered choices

3. Models for Quantitative, limited dependent variables

(a) Truncation and censoring

(b) Tobit models and ML estimation

(c) “Corrected LS”

(d) Models with stochastic thresholds

4. Special Topics

(a) Duration models

(b) Program Evaluation

Exams and Make up Exam Policy:

The final exam is semi-comprehensive, covering selected material from the entire course. Both the
midterms and the final exam have to be written. The exams will be comprised of the following: problems
solving questions.

You will receive a grade of zero for a missed exam. There are no make-up exams. If you miss an exam,
you need to obtain a letter signed by the parent/guardian with their contact numbers that excuses your
absence. You will have to provide documentation proving you had an emergency and could not attend the
exam. If you do not obtain such a letter, you will fail the exam. If you have to miss the first in-class
midterm exam, the second in-class midterm exam will be worth 50% of your grade. If you miss the
second in class exam, the final exam will be worth 50% of your grade.

Please note:

 You must come prepared for all your exams.

 You must come on time.
 Being late does not necessarily guarantee that you are going to get extra time for writing your
tests and exam.
 You must bring your own pencil, pen, eraser, calculator and any other permitted items that you
may need and you are allowed during the tests and exam.
 All cell phones must be switched off.
 Any deviation from the standard procedures will not be taken lightly.
 Any unfair means adopted in the tests and exam will be seriously dealt with.
 Academic misconduct or failure to comply with NSU Examination Code of Conduct may result
in F.

Attendance Policy
You must come to the class in due time, stay for the duration of the class, and submit in-class worksheets
to get attendance. You must not try to enter the classroom after 10 minutes of scheduled class time.

Communication Policy
All communications should take place using my official email. Announcements made in the classroom
will override any statement made here. It is your responsibility to be aware of any announcements made
in the classroom.

Accessing the Google Classroom:

Google Classroom will be used to post announcements, lecture slides, and other class related materials.
You are responsible for checking it on a regular basis.

In order to access Google Classroom go to classroom.google.com > sign-in with your nsu email > click
the plus sign at the upper right corner > Join this class with code jhb49lg.

Appropriate Use Policy

All members of the North South University community must use electronic communications in a
responsible manner. The University may restrict the use of its computers and network systems for
electronic communications subject to violations of university policies/codes or local laws or national
laws. Also, the university reserves the right to limit access to its networks through university-owned or
other computers, and to remove or limit access to material posted on university-owned computers.

Good Luck