(Printer Friendly Version)

AREC 624

Applied Econometrics II

Spring 2008

Professors Alberini, Just and Kirwan

Course requirements 

1. One midterm (covering Professor Alberini’s material, part I, Professor Kirwan, part II, and Professor Just’s material)
2. One empirical project covering Professor Just’s topics,
3. One empirical project for Professor Alberini (to be described in a 10-page paper due on the day of the last class, May 12, 2008),*
4. One final exam covering the material in parts IV and V of the syllabus*

Professor Just’s part of the course accounts for 33% of the final grade, Professors Alberini and Kirwan’s the remaining 67%. Alberini and Kirwan’s part of the course can be satisfied in one of two ways: (i) by submitting an empirical paper as described in point 3 above and talking an abbreviated version of the final exam, or (ii) by taking a longer version of the final exam. You may choose whether you wish to pursue (i) or (ii), but you are required to notify Alberini by mid-April.

Teaching Assistant

The Teaching Assistant for this course is Michelle Michalek, Room 3109, mmichalek@arec.umd.edu.  She holds her office hours on Wednesdays from 1:30 – 3:30 p.m. Discussions are held on Fridays from 10 to 12:30 in Rm 0209.

Important Dates:

Jan 28-Feb 13: Professor Alberini’s lectures, part I of the course.

Feb 18-Mar 20: Two lectures on panel data models by Professor Kirwan. This is part II of the course.

Feb 25-Mar 14: Professor Just’s portion of the course (part III of the course).

Mar 17-Mar 21: Spring Break.

Mar 24: Professor Just’s empirical project due.

Mar 24: Midterm exam.

Mar 26-April 9: Professor Kirwan’s lectures (part IV of the course).

April 14-May 12: Professor Alberini’s lectures (part V of the course).

April 16: A one-page description of your empirical project for Professor Alberini is due.

May 12: Last lecture. 10-page paper describing your empirical project for Professor  Alberini due.

Part I--Alberini

(Jan 28-Feb 13, 2008)

Office hours:   Monday 3:00-5:00pm, Room 2210.

Textbook:        William Greene, Econometric Analysis, 5^th edition, Prentice Hall, 2003,

Suggested Topics

1. Heteroskedasticity (Greene, Ch. 11)

1. Nature of the problem and its effect on OLS estimates (review of material covered last semester)
2. GLS estimation and feasible GLS estimation
3. Tests: Goldfeld-Quandt, White’s information matrix test, and the Breusch-Pagan test. White’s heteroskedasticity-robust covariance matrix.
4. Alternative Ways to Compute Standard Errors: The Bootstrap and the Jackknife: Hogg --5.8-5.9.

2.   Serial correlation (Greene, Ch.12)

1. Nature of the problem and its effect on the OLS estimates
2. Autoregressive error terms of the first order, moving average error terms of the first order, and efficient estimation in the presence of serially correlated error terms. The Newey-West robust covariance matrix (Greene, Ch. 12, page 267).
3. Autoregressive conditional heteroskedasticity (ARCH) (Greene, Ch. 11)

3.   Optimization methods in econometrics (Greene, App. E.6).

Part II-- Kirwan

(Feb 18 and 20, 2008)

Office hours:               Tuesday & Thursday 2-3pm, Room 2120

Recommended Texts: Wooldridge, Jeffrey.  Econometric Analysis of Cross Section and  Panel Data.  MIT Press, Cambridge, MA, 2002.

                                    Cameron, A.C. and P.K. Trivedi. Microeconometrics.  Cambridge UP. New York, NY. 2005.

Additional Readings:   TBA

 I.    Regression and the Conditional Expectation Function

II.  Panel Data Models

      A.  Random Effects Models

      B.  Fixed Effects Models

      C.  Hausman Test

Part III--Just

(Feb 25-Mar 14, 2008)

Office hours:   To be Announced, Room 2213

Textbook:        William Greene, Econometric Analysis, 5^th edition, 2003,

Chapters 14 and 15.

Supplementary Readings:     

Section 11.2, Chapters 14 and 15 of Judge, Hill, Griffiths, Lutkepohl, and Lee, Introduction to the Theory and Practice of Econometrics, Second Edition, 1988.

Chapters 14 and 15 of Judge, Griffiths, Hill, Lutkepohl, and Lee, The Theory and Practice of Econometrics, Second Edition, 1985.

 1.   Review of distribution theory and application to systems models

 2.   Multivariate regression

 3.   Seemingly unrelated regressions

 4.   Simultaneous equations estimation – Estimation issues

A.   Assumptions

B.   Identification

C. Modeling issues

 5.   Simultaneous equations estimation – Single equation methods

A.   Indirect least squares

B.   Two-stage least squares

C. Limited information maximum likelihood

 6.   Simultaneous equations estimation – Multiple equation methods

A.   Three-stage least squares

B.   Full-information maximum likelihood

 7.   Simultaneous equations estimation – System structure and large systems

A.  Recursive system estimation

B.   Block recursive systems

 8.   Nonlinear simultaneous equations estimation

 9.   Generalized Method of Moments (GMM) estimation

Estimation exercise: 

During this section of the course, a major estimation exercise will be required to give each student experience with each major structural estimator.  Students can benefit from discussing together how to apply the various estimators.  However, each student is expected to develop and estimate his/her own systems of equations.  That is, each student should do his/her own specification and estimation independently.  Estimation must be done using MATLAB because of specific assignments about calculation and verification of estimators.  Since a short course was offered to familiarize students with MATLAB, no class time will be spent on learning MATLAB software.

Part IV-- Kirwan

(Mar 26 -Apr 9, 2008)

Recommended Texts: Wooldridge, Jeffrey.  Econometric Analysis of Cross Section and  Panel Data.  MIT Press, Cambridge, MA, 2002.

                                    Cameron, A.C. and P.K. Trivedi. Microeconometrics.  Cambridge UP. New York, NY. 2005.

Additional Readings:   TBA

I.    Special Topic: Difference-in-Differences

II.  Instrumental Variables

      A.   Wald Estimator

      B.  Constant Effects

      C.  Heterogeneous Effects – Local Average Treatment Effects

III.   Problems with IV

      A.   2SLS Bias

B. Clustering

IV. Special Topic: Regression Discontinuity

Part V--Alberini

(April 14-May 12, 2008)

Suggested Topics

 1.   Censored and Truncated Models, including the Tobit Model (Greene, Section 22.3)

 2.   Binary Data Models:  Probit and Logit Models (Greene, Ch. 21).

• The probit (logit) equations
• Derivation of the likelihood function, first-order conditions and covariance matrix of the estimates (Greene 21.1—21.4)
• Asymptotic properties of the ML probit (logit) estimates (Greene 21.1—21.4)
• Interpretation of coefficients, elasticities and marginal effects, and goodness of fit measures
• Applications: (a) models of participation in programs, derivation of the estimation curve and of its confidence intervals using the delta method; (b) dichotomous choice contingent valuation: how to recover mean/median WTP and compute its standard error using the delta method (Cameron and James, 1987)
• The fixed effects logit model and the random effects probit model (Greene, 21.5)
• The bivariate probit model.  

3.   Discrete/continuous models:

• Heckman two-step estimation and maximum likelihod estimation (lecture notes and Wooldridge, Ch. 17, page 560-571). Use of this model to address self-selection into a sample, an endogenous dummy in the RHS of a regression model.
• Testing for endogeneity of regressors in a discrete/continuous model (Rivers and Vuong, 1988)
• Switching regression models with endogenous switching
• Using the technique of propensity score matching in quasi experiments (Wooldridge, Ch. 18, page 614-642)

 4.   Models for Count Data (Greene, Ch. 21.9):

• the Poisson model: ML estimation and properties of the ML estimates
• the problem of overdispersion, and how to address it using (a) an explicit overdispersion parameter, or (b) the negative binomial model.

6.   Discrete Choice Models (Greene, section 21.7):

• The multinomial logit model and the conditional logit model: Assumptions, interpretation of coefficients, elasticities and marginal effects, goodness of fit measures, and a Hausman test for IIA

• The nested logit model
• The mixed logit model: comparison of estimated coefficients, predicted probabilities and elasticities from mixed logit and conditional logit (Train, 1998, 1999)
• Hierarchical Bayes Models (Train, 2003)
• An alternate approach to heterogeneity: Latent Class Models (Scarpa and Thiene, 2005)

References

Cameron, Trudy A. and Michelle D. James (1987), “Efficient Estimation Methods for ‘Closed Ended’ Contingent Valuation Surveys,” Review of Economics and Statistics, 69(2) , 269-286.

Rivers, D. and Vuong (1988), “Limited Information Estimators and Exogeneity Testing for Simultaneous Probit Models,” Journal of Econometrics, 39(3), 347-366.

Scarpa, Riccardo and Mara Thiene (2005), “Destination Choice Models for Rock Climbing in the Northeastern Alps: A Latent-Class Approach Based on Intensity of Preferences,” Land Economics, 81(3), 426-444.

Train, Kenneth E. (1998), “Recreation Demand Models with Taste Difference Over People,” Land Economics, 74(2), 230-239.

Train, Kenneth E. (1999), “Mixed Logit Models for Recreation Demand,” Chapter 4 in Joseph A. Herriges and Catherine L. Kling (eds.), Valuing Recreation and the Environment. Revealed Preference Methods in Theory and Practice, Cheltenham, UK: Edward Elgar Publishing.

Train, Kenneth E. (2003), Discrete Choice Methods with Simulation, Cambridge, UK: Cambridge University Press.