Economics 312

Theory and Practice of Econometrics

Spring 2010

Jeffrey Parker


Course Outline and Reading List

Basic text materials

0. Review of basic statistics
1. Introduction to econometrics
2. The bivariate regression model
3. Inference in the bivariate regression model
4. Basics of multiple regression
5. Inference in the multiple regression model
6. Functional form and nonlinearities
7. Model assessment

Midterm exam

8. Models for pooled and panel data
9. Models with limited dependent variables
10. Endogenous regressors and instrumental variables
11. Basics of regression with time-series data
12. Vector autoregression, integration, and cointegration
13. Advanced topics in econometrics

Basic text materials

Most of the readings and assignments for Econ 312 will be taken from the following list of texts, which will be on reserve in the library. Notation varies some across texts, so be careful when switching among them. Class presentations will conform to the notation and sequencing of the main text, by Stock and Watson. On the list, ** indicates texts at about the same level of mathematical complexity as the Stock and Watson text. Texts marked with *** are more difficult; those marked * are more basic.

  • **Stock, James H., and Mark W. Watson, Introduction to Econometrics, 2nd ed., Boston: Pearson Addison Wesley, 2007. (The main text for the course. Most reading and many assignments will be drawn from here.)
  • **Berndt, Ernst, The Practice of Econometrics: Classic and Contemporary, Reading, Mass.: Addison Wesley, 1990. (Not a traditional econometrics text. Contains topical chapters on applications of econometrics with data sets and exercises. Some weekly econometrics projects will be drawn from here.)
  • **Griffiths, William E., R. Carter Hill, and George G. Judge, Learning and Practicing Econometrics, New York: John Wiley & Sons, 1993. (A former text that is perfect in level and detail, but very out of date.)
  • **Wooldridge, Jeffrey, Introductory Econometrics: A Modern Approach, 3rd ed., Mason, Ohio: Thomson South-Western, 2006. (A common textbook for this course. A little less mathematically complete than Stock and Watson.)
  • ***Davidson, Russell, and James G. MacKinnon, Econometric Theory and Methods, Oxford: Oxford University Press, 2004.
  • ***Greene, William, Econometric Analysis, 6th ed., Englewood Cliffs, N.J.: Prentice-Hall, 2008. (Note: The library has just ordered a copy of the 6th edition and it probably won't be available for a few weeks. Use the 4th edition until the 6th is available.)
  • ***Hamilton, James, Time Series Analysis, Princeton: Princeton University Press, 1994. (A specialized time-series book that is very difficult but authoritative.)
  • ***Enders, Walter, Applied Econometric Time Series, New York: John Wiley & Sons, 1995. (Another time-series text that we may use for special topics toward the end of the course.)
  • *Studenmund, A. H., Using Econometrics: A Practical Guide, 5th ed., Boston: Pearson Addison Wesley, 2006. (A somewhat easier text used for Econ 311 at Reed.)
  • *Murray, Michael, Econometrics: A Modern Introduction, Boston: Pearson Addison Wesley, 2006. (Another simpler text.)

Note that the material on the syllabus below will be adjusted as the semester proceeds. Dates are approximate.

Back to top

0. Review of basic statistics

January 20-22, 10-12 in Vollum 228

Required readings:

  • Stock and Watson, Chapters 2 and 3.
Notes

Back to top

1. Introduction to econometrics

January 25

Required readings:

  • Stock and Watson, Chapter 1

Additional sources:

  • Wooldridge, Chapter 1
Notes

Back to top

2. The bivariate regression model

January 27, 28, 29, February 1 and 3

Topics:

  • What regression does
  • Strategies for obtaining regression estimators: method of least-squares, method of moments, method of maximum likelihood
  • Least-squares regression model in matrix notation
  • Assumptions of OLS regression
  • Sampling distribution of OLS estimators
  • How good is the OLS estimator?
  • Measuring goodness of fit

Required readings:

  • Stock and Watson, Chapter 4, Sections 5.4 and 5.5, and Appendices 5.1 and 5.2.
  • Griffiths, Hill, and Judge, Section 5.4, augmented by their Appendix 3B and/or Stock & Watson's Appendix 18.1 as necessary.
Additional sources:
  • Griffiths, Hill, and Judge, rest of Chapter 5.
  • Wooldridge, Chapter 2.
  • Davidson and MacKinnon, Section 1.3 on specification of regressions, Section 1.5 on method of moments.
Notes
Back to top

3. Inference in the bivariate regression model

February 4, 5, 8, and 10

Topics:

  • Hypothesis tests about single regression coefficients
  • Confidence intervals for individual coefficients
  • Dummy independent variables
  • Asymptotic properties of OLS bivariate estimator
  • Weighted least squares

Required readings:

  • February 4: Session introducing first econometric project:
  • Stock and Watson, rest of Chapter 5 and Chapter 17.

Additional sources:

  • Griffiths, Hill, and Judge, Chapters 6 through 8.
  • Wooldridge, Chapter 2 in general, Chapter 8 on weighted least squares, Appendix C.3 on asymptotic distributions.
  • Dummy variables are covered in the multiple-regression context in Griffiths, Hill, and Judge, Chapter 12 and Wooldridge, Chapter 7.
  • Asympototic properties are covered in the multiple-regression context in Griffiths, Hill, and Judge, Chapter 14.
Notes
Back to top

4. Basics of multiple regression

February 11 and 12

Topics:

  • Omitted-variable bias
  • The multiple-regression model
  • Goodness of fit
  • OLS assumptions in multiple regression
  • Distribution of OLS multiple-regression estimators

Required readings:

  • Stock and Watson, Chapter 6 and Sections 18.1, 18.2, and 18.5.

Additional sources:

  • Griffiths, Hill, and Judge, Chapter 9.
  • Wooldridge, Chapter 3.
Notes

Back to top

5. Inference in the multiple regression model

February 15, 17, and 18

Topics:

  • Kinds of tests in a multiple regression
  • Hypothesis tests on a single coefficient
  • Testing joint hypotheses
  • Single hypotheses involving multiple coefficients
  • Multivariate confidence sets
  • Some specification issues
  • Applications of multiple regression

Required readings:

  • Stock and Watson, Chapter 7 and Sections 18.3 and 18.4.

Additional sources:

  • Griffiths, Hill, and Judge, Chapters 10 and 11.
  • Wooldridge, Chapter 4 through 6 and 8.
Notes

Back to top

6. Functional form and nonlinearities

February 19, 22, and 24

Topics:

  • Nonlinearity in variables vs. nonlinearity in parameters
  • Quadratic and higher-order polynomial models
  • Log-based models
  • Interaction effects
  • Nonlinear least squares

Required readings:

  • Stock and Watson, Chapter 8.

Additional sources:

  • Woodridge, Chapter 9.
  • Griffiths, Hill, and Judge, Chapter 8.
  • Greene, Chapters 6 and 7 (more advanced).
Notes
Back to top

7. Model assessment

February 25 and 26

Topics:

  • Internal vs. external validity
  • Assessing external validity
  • Assessing internal validity
  • Validity in forecasting/prediction

Required readings:

  • Stock and Watson, Chapter 9.

Additional sources:

  • Woodridge, Chapter 9.
Notes
Back to top

Midterm Exam

The midterm exam is scheduled as follows:

  • In-class part: March 1
  • Take-home part: Handed out March 1; due March 5
Back to top

8. Models for Pooled and Panel Data

March 3, 4, 5, and 8

Topics:

  • Pooled vs. panel data
  • Regression with pooled cross-sections
  • Structure of panel data
  • Before and after estimators with panel data
  • Fixed-effects regression models
  • Random-effects regression models

Required readings:

  • Stock and Watson, Chapter 10.
  • Wooldridge, Chapters 13 and 14 (Wooldridge devotes a lot more detail to these models than Stock and Watson).

Additional sources:

  • Greene, Chapter 9 (for a more advanced presentation).
  • Wooldridge, Econometric Analysis of Cross Section and Panel Data, Chapter 10 (a more advanced version of the material in the Wooldridge textbook).
Notes
Back to top

9. Models with limited dependent variables

March 10, 11, 12, 22, and 24

Topics:

  • Binary dependent variables: linear probability model, probit, and logit
  • Discrete-choice dependent variables: multinomial logit
  • Ordered dependent variables: ordered probit and logit
  • Censored and truncated regression models: tobit and heckit
  • Count variables: Poisson and negative-binomial regression

Required readings:

  • Stock and Watson, Chapter 11.
  • Wooldridge, Chapter 17 (Section 17.1 repeats Stock and Watson's Chapter 11; the remaining sections cover the material in Stock and Watson's Appendix 11.3 in more detail).

Additional sources:

  • Griffiths, Hill, and Judge, Chapter 23.
  • Davidson and MacKinnon, Chapter 11 (more advanced).
  • Greene, Chapters 23 through 25 (more advanced).
Notes
Back to top

10. Endogenous regressors and instrumental variables

March 25, 26, 29, and 31

Topics:

  • Approaches to endogeneity problems
  • Single-equation instrumental-variables estimation
  • Simultaneous equations and the identification problem
  • Estimation of systems of equations
  • Three-stage least squares and other full-information estimators
  • Experimental and quasi-experimental data

Required readings:

  • Stock and Watson, Chapters 12 and 13, Sections 18.6 and 18.7.
  • Griffiths, Hill, and Judge, Chapters 17 through 19.

Additional sources:

  • Wooldridge, Chapters 15 and 16.
  • Davidson and MacKinnon, Chapter 8 (more advanced).
  • Greene, Chapters 12 and 13 (more advanced).
Notes
Back to top

11. Basics of regression with time-series data

April 1, 2, 5, 7, 8, and 9

Topics:

  • Lags, differences, and autocorrelations
  • Autoregressive and moving-average models
  • Estimation with serially correlated errors
  • Distributed-lag models
  • Granger causality
  • Basics of nonstationary variables: trends, unit roots, and breaks

Required readings:

  • Stock and Watson, Chapters 14 and 15.

Additional sources:

  • Wooldridge, Chapters 10 through 12.
  • Davidson and MacKinnon, time-series sections of Chapter 7 (more advanced).
  • Greene, Chapters 19 and 20 (more advanced).
Notes
Back to top

12. Vector autoregression, integration, and cointegration

April 12, 14, 15, and 16

Topics:

  • Vector autoregressions
  • Unit roots and integrated processes
  • Cointegration and error-correction models
  • Autoregressive conditional heteroskedasticity

Required readings:

  • Stock and Watson, Chapter 16.
  • Recommended reading on VAR model: Enders, pages 294-316.

Additional sources:

  • Davidson and MacKinnon, Chapters 13 and 14 (more advanced).
  • Greene, Chapters 20 through 22 (more advanced).
Notes
Back to top

13. Advanced topics in econometrics

April 19 ...

Topics chosen among:

  • Specification search and data mining
  • Publication bias
  • Monte Carlo and bootstrap methods
  • Imputation methods for missing data
  • Varying-parameter models
  • Duration/hazard models
  • Quantile regression
  • Bayesian methods in econometrics

Readings:

  • Leamer, Edward E., "Let's Take the Con Out of Econometrics," American Economic Review, 73(1), March 1983, 31-43.
  • Lovell, Michael C., "Data Mining," Review of Economics and Statistics, 65(1), February 1983, 1-12.
  • De Long, J. Bradford, and Kevin Lang, "Are All Economic Hypotheses False?" Journal of Political Economy 100(6), December 1992, 1257-72.
  • Davidson and MacKinnon, Section 4.6: Simulation-Based Tests.
  • Greene, Sections 17.1, 17.2, 17.4: Simulation-Based Inference.
  • Little, Roderick J. A., "Regression with Missing X's: A Review," Journal of the American Statistical Association 87(420), December 1992, 1227-37.
  • Maddala, G. S., Econometrics, New York: McGraw-Hill, 1977, Chapter 17: Varying-Parameter Models.
  • Greene, Section 25.6: Models for Duration Data.
  • Koenker, Roger, and Kevin F. Hallock, "Quantile Regression," Journal of Economic Perspectives 15(4), Autumn 2001, 143-156.

Notes (complete)

Back to top