Universiteit van Tilburg
Empirical Research In Econometrics
Prof. Jonathan B. Hill (with Prof. Meltem Daysal)
Prof. Jonathan B. Hill
Email: jbhill@email.unc.edu
Office :
Jump to: Announcements ** STATA Material ** Assignments/Data/Slides
Introduction The course presents the standard techniques of empirical investigation used in economics with applications in a wide range of fields such as labour economics, savings and consumption, endogenous growth, international finance, empirical macro-economics, etc. Students will learn to do empirical research themselves and to evaluate empirical research that others have done.
Course Topics The
following broad topics are typically covered. I will cover A and B and Prof.
Daysal will cover C and D. Note that changes may occur during the semester.
A. Standard linear regression model (model formulation and interpretation, estimation and hypothesis testing, goodness of fit, predictions, dummy variables, choice of functional form, omitted variable problems, applications).
B. Generalized Linear Model (heteroskedasticity, autocorrelation).
C. Models with endogenous regressors (instrumental variables estimation, two stage least squares).
D. Maximum likelihood estimation. Binary, ordered, and multinomial choice models.
Grading In the first half of the semester there will be 3 assignments worth 5% with blends of econometric theory and practice, involving both problem solving and STATA applications. Additionally, there will be one exam worth 35%.
Reading
There is one required text, and two useful hence recommended texts:
|
Required: |
Wooldridge, J. (2006).
Introduction Econometrics 4e, Online Datasets, Compressed
Datasets (master zip file with all Stata files) |
|
Useful: |
Verbeek, M. (2008). A
Guide to Modern Econometrics 3e, Online Datasets |
|
Useful: |
Baum, C. F. (2006). An
Introduction to Econometrics Using STATA |
|
(depending on you version of
STATA, not all commands in this textbook may be valid) |
First Half of Course The following is a refined list of topics by lecture, with Wooldridge chapters, for the first 6 weeks of class, covering the first three labs. Together, this material represents what the midterm exam will cover. Note these topics may change as the semester progresses.
|
Go to: Lecture 1, Lecture 2, Lecture 3, Lecture 4, Lecture 5, Lecture 6 |
||
|
Multivariate Regression: ordinary least squares Wooldridge Chapt: 2-3 (recommended: Verbeek 2) |
||
|
Multivariate Regression continued: dummy variables, properties of the OLS estimator, goodness-of-fit, confidence intervals and tests Wooldridge Chapt. 4, 7 (recommended: Verbeek 3) |
||
|
Lab #1 (Wed. Sept. 23) |
STATA structure, multivariate regression, testing |
|
|
Multivariate Regression continued: large sample properties of the OLS estimator, multicollinearity, prediction, interpreting the linear model, mis-specifying the regressor set Wooldridge Chapt. 5 (recommended: Verbeek 2) |
||
|
Multivariate Regression continued: selecting regressors, information criteria, comparing nested and non-nested models, testing the functional form (RESET), testing for a structural break (Chow), heteroskedasticity Wooldridge Chapt. 6, 9.1-9.2 (recommended: Verbeek 2) |
||
|
Lab #2 (Wed. Oct. 7) |
functional form, structural breaks |
|
|
Heteroscedasticity: consequences for OLS, robust standard errors (White), Generalized Least Squares, Feasible Generalized Least Squares, proportional heteroscedasticity, group-wise, Heteroscedasticity, testing for heteroscedasticity (White, Breusch-Pagan) Wooldridge Chapt. 8 (recommended: Verbeek 3) |
||
|
Lab #3 (Wed. Oct. 14) |
Heteroscedasticity |
|
|
Midterm
(Tues. Oct. 20) |
Autocorrelation: consequences for OLS, testing for autocorrelation (Q-test), First Order Autocorrelation (GLS, FGLS, Durbin-Watson test) Wooldridge Chapt. 10.1-10.4, 12.1—12.3 (recommended: Verbeek 4) The midterm exam will be given during the
second half of this last lecture. |
|
Online Data Sources
The following are excellent sources of free on-line data.
Board of Governors of the Federal Reserve
System
Federal
Reserve of Saint Louis: FRED Data Bank
Extensive array of time
series data; seasonally adjusted or unadjusted; real or nominal.
National
Center for Health Statistics
Maintained by the Center for
Disease Control, provides substantial data on
health related topics.
The BLS, a branch of the Dept. of Labor, provides data and
research on a variety of labor topics.
Current
Population Survey [CPS]
Multiple panels of a wide
variety of labor/health statistics.
Econ & financial data links :
Business & Economic DataLinks
International Monetary Fund
[IMF]: Data bases.
World
Bank: Data archives.
Organization
of Economic Cooperation and Developement [OECD].
Penn World Table: A
large archive of country specific data provided by the
National
Bureau of Economic Research [NBER] General Data
Eco5.com:
Data
archive, working papers, forecasts, job markets: omnibus website for
economists.
Economagic
Announcements
Oct. 20: Feel free to email me (jbhill@email.unc.edu)
on or after Oct. 23 for your midterms score.
STATA Material
|
STATA and Econometrics
(helpful guide for this class) |
Assignments, Answer Keys
|
Slides |
Assignments |
Datasets |
Data Descriptions |