Econometrics II

Econ 620
Instructor: Nicholas M. Kiefer  (nmk1)

Cornell University
Department of Economics
Spring 2008

Introduction	This course deals with estimation and inference using economic models and economic data. We will cover the linear regression model and variations that arise when "ideal" conditions are not met. We will also cover, at the introductory level, maximum likelihood estimation, nonlinear models and simultaneous equation models. We rely heavily on lecture notes. These are available from the course web page : http://instruct1.cit.cornell.edu/courses/econ620   and lecture notes should be read in advance of the lectures.
Text Books	A good econometrics text is a valuable supplement: consider Judge et al Theory and Practice of Econometrics, Johnston's Econometric Methods, Goldberger A Course in Econometrics, or Ruud An Introduction to Classical Econometric Theory. If you are uncomfortable with matrix algebra you should review (perhaps in a group) as soon as possible - it will come fast in lecture. A guide to readings in Judge is included in the lecture notes.
Paper	A short (10 pages) paper reporting an application of econometrics is required. You may use any economic data sets. CISER is a potential source of data. Also check the business library. A project proposal of 1~2 pages (ideally 1 page) describing the question you will be looking at, the data you will use, and the relevant references to the literature must be submitted for approval by 2/14; it is to your advantage to do this as soon as possible. Preliminary results will be presented to the class with the aid of handouts or transparencies beginning after the break. You will have 20 minutes for presentation - this is about the time allocated at the ES/AEA winter meetings and the ASA August meetings. You should plan to present at these meetings when you are on the job market and you might as well start practicing now. Have fun with these projects; this is what you are preparing to do for the rest of your careers. The final version of the paper is due at the end of classes (last lecture, that is, 5/1).
Grades	Grades will be based on the paper (35%), a midterm (20%), a cumulative final (35%), and on promptness in turning problem sets, the quality of your presentation, and participation in class discussion during the student presentations (10%). Failure to meet deadlines in turning in the proposal or the papers will be penalized heavily.
Midterm	The midterm will cover the first 14 lectures; the material required is thorough knowledge of the linear regression model, including GLS, and asymptotics at the level covered in class and notes. The midterm will occur on 3/13.
	I encourage you to work together on problems and on the programming required for the projects.