ECON 330: Econometrics Fall Semester 2015-2016

Course Outline Muhammad Farooq Naseer Office: 244 (Extn: 8073) Email: farooqn@lums.edu.pk Office Hours: Mon-Wed 12:15-1:15 PM Description This is the second course in the statistics/econometrics sequence and looks

at the broad range of estimation problems that often arise in economic applications. In particular, we look at the criteria used to select a particular estimation method and the scenarios under which the OLS estimator becomes sub-optimal. The purpose of this course is to teach students the basics of econometric theory and also to give them hands-on experience with using a statistical package Stata, which will be helpful in later applications especially for those students who choose to do an empirical senior project.

Goals On successful completion students will:

1. be able to develop a suitable regression model for a variety of empirically interesting problems and validate the selected model via a battery of tests

2. be able to compare different estimators based on their finite sample and asymptotic properties

3. develop a basic understanding of time series econometrics and be able to handle and make use of panel data

4. be proficient in the use of Stata for econometric analysis Prerequisites

<Probability AND Statistics> OR <Statistics and Data Analysis>; Microeconomics 1 Or Principles of Microeconomics; Macroeconomics 1 Or Principles of Macroeconomics

Text Book

Wooldridge, Jeffrey M. 2006. Introductory Econometrics. 3rd edition. Thomson South-western.

Reference Texts

1. Kohler, Ulrich and Frauke Kreuter. 2012. Data Analysis using Stata. Stata

Press. 2. Banerjee, Abhijit V., and Esther Duflo. 2011. Poor Economics: A Radical

Rethinking of the Way to Fight Global Poverty. Public Affairs. 3. Hamilton, Lawrence C. 2006. Statistics with Stata. Thomson Brooks/Cole. 4. Levitt, Steven D., and Stephen J. Dubner. 2009. Freakonomics: A Rogue

Economist Explores the Hidden Side of Everything. Harper Perennial.

Lectures Two lectures of 100 minutes plus one 50-minute lab session (led by the TAs) per week. The course outline below refers to sections from your textbook. Relevant sections of the textbook are included in your course reading package and the reference texts may be obtained from the Library.

Online Resources To learn STATA you may use: http://www.ats.ucla.edu/stat/stata/ STATA illustrations for all our text book examples are at: http://fmwww.bc.edu/gstat/examples/wooldridge/wooldridge.html The power-point slides for the book are also available at: http://www.swlearning.com/economics/wooldridge/wooldridge2e/powerpoint.html Grading

Assignments (6) 20% Project 15% Quizzes (5) 30% Final 35%

Course Policies Quizzes: There will be four announced in-class quizzes, which will take place through the semester. There will be one announced in-lab quiz towards the end of the term. Lab Attendance: Attendance in the labs is highly recommended and we will be taking attendance during each lab session. Anyone who does not arrive within the first 15 minutes of the lab will be marked as absent from that lab. An individual who is absent in more than THREE labs will be given a grade of zero in one of his highest scoring lab assignments. Lab Submission: Students are encouraged to work on the assignments in groups of 2-3 students. However, the submission of assignment is to be done individually by each student in their own handwriting. There will be group grading of assignments (an individual’s assignment from within a group will be picked randomly for grading and the same grade will be assigned to the entire group for that lab). Please note that it is possible under this grading scheme for all group members to get zero even if one group member does not submit the assignment (or its correct solution). This is to improve learning by encouraging discussion within groups while also ensuring that everyone gets to do the assignment. Please note that sharing or discussing assignments with anyone outside your own group is NOT allowed and makes grounds for a disciplinary action. Group formation is voluntary but some groups may need to be adjusted. Project: The objective of the project is to provide you an opportunity to apply the skills you learn in class to a real world application. Several data sets will be made available to students for this purpose. The project would require you to pick a data set from this collection and write a short paper based on your analysis. The project grade will be determined on the basis of an intelligent use of this data to address the research question and an appropriate interpretation of results. Like the labs, the project would be group-

based and we may conduct vivas from any of the group members. Students are encouraged to discuss their project with me (during office hours) or with their TAs. Missed Quizzes/Assignments: As per the rules of the Student Handbook, students must contact the instructor with a petition form and valid supporting documents either before or within three days of missing an instrument. The decision on such petitions will be made on a case-by-case basis and may involve grade deduction before assigning the student’s quiz average. Under ordinary circumstances, there will be no make-up for missed assignments. Instrument Grading: All the course instruments are checked as thoroughly and fairly as possible and the process consumes a lot of your TAs’ and instructor’s time. Therefore, and to ensure uniformity in grading across all students, there will be no ad-hoc adjustment of marks ex-post. While we encourage student queries meant to improve learning, please note that your TAs are not authorized to change your marks once an instrument has been graded. Detailed Course Outline Sr. No.

Topic Readings Weeks

1 Introduction What is econometrics? Steps in empirical economic analysis The structure of economic data; random sampling Simple Regression Model Deriving the OLS estimates Algebraic properties Deriving statistical properties: mean and variance

Ch1. 1.1, 1.2, 1.4 Ch.2.1, 2.2, 2.4 Appendix B

1.5

2 Multiple Regression: Estimation [[Causality and Marginal effects]] Mechanics and Interpretation of OLS Classical Linear Model Assumptions The Gauss-Markov Theorem Properties of OLS – Mean and Variance Topics in OLS: Effects of Data Scaling: 6.1 Functional Form: 6.2 Goodness-of-Fit and Model Selection: 6.3 Functional form mis-specification: 9.1

Ch. 3 1.5 1

3 Multiple Regression: Inference Sampling Distribution of the OLS estimators The t-test – testing a single restriction Confidence Intervals Testing multiple restrictions Multiple Regression Analysis: OLS Asymptotics Law of Large Numbers and Central Limit Theorem

Ch. 4 Ch. 5; Appendix C

2

Consistency Asymptotic Normality and Large Sample Inference

4 Functional Form and Dummy Variables Dummy independent variables Using dummy variables for multiple categories Interactions using dummy variables Dummy dependent variable

Ch. 7 1

5 More Topics in OLS Prediction and Residual Analysis: 6.4 Missing Data, Outliers: 9.4

Ch. 6.4, 9.4 0.5

6 Heteroskedasticity Consequences of Heteroskedasticity Robust inference Testing for heteroskedasticity Weighted Least Squares

Ch. 8 1

7 Instrumental Variable Estimation and 2SLS Correlation between X and error; Omitted variable bias (3.3); OLS under measurement error (9.3); Using Proxy Variables for Unobserved Explanatory Variables (9.2); IV estimation and the 2SLS; Testing for endogeniety and over-identifying restrictions;

Ch 3.3, 9.2, 9.3, 15.1-15.5;

2

8 Simultaneous Equation Models The nature of simultaneous equation models; simultaneity bias in OLS; Identifying and estimating a structural equation (vs. reduced form); systems with more than two equations

Ch 16.1-16.3 0.5

9 Regression with Time Series Data Nature of time series data; Examples of TS models; Finite sample properties of OLS under Gauss-Markov assumptions; Functional form, dummy variables, index numbers; Trends and seasonality;

Ch. 10 1

10 Panel Data Models Pooling independent cross-sections across time; two-period panel data; differencing with more than two time periods: fixed-effects estimation; random-effects models; grouped data; policy analysis (difference-in-difference and panel estimation)

Ch 13.1-13.5; 14.1-14.3

1.5

11 Limited Dependent Variable Models and Sample Selection Logit and Probit models for binary response; [Depending on time: the Tobit model for corner-solution responses; Censored and Truncated regression;] Sample selection corrections

Ch 17.1-17.5 (excl. 17.3)

1

* The readings are all from the Wooldridge textbook unless otherwise indicated