1R Simultaneous Equation Models
Bob Luskin, University of Texas, Austin
9 - 20 July (two week course / 35 hrs)
Detailed Course Outline [PDF]
Course Content
This course is about statistical models of more than one equation, accounting for more than one dependent variable, still sometimes called "causal models." We shall concentrate on linear models of a standard linear econometric sort, although we shall also consider more briefly nonlinear models, models with discrete dependent variables, and models with measurement error, as time permits.
Exercises and laboratory sessions will provide practice at specifying models and generating and interpreting concrete results, but the lectures and readings will focus on more general questions of modeling, estimation, and inference: What sorts of models imply—and should reflect hypotheses of—what sorts of effects? What variables—and equations—must we include? What assumptions must we make, and what do they mean? How likely are the assumptions to be violated, and with what consequences? When is a model identified (roughly, estimable), and what can be done when it's not? What quantities should we be focusing on estimating? What estimators provide statistically desirable estimates? Where several different estimators might serve, what are their advantages and disadvantages? What do the estimates tell us, and how certainly?
The lectures and readings will treat these questions both abstractly and practically, referring both to x's and y's and to substantive examples. There will be mathematical notation and mathematically phrased argument but little proof or derivation. There will also be some algebraic/numerical and some practical/empirical exercises, the latter asking students to specify, estimate, and analyze models of their devising.
Course Objectives
The goal is to convey a good understanding of the how’s and why’s of constructing, estimating, and interpreting the estimates of these models—to understand the numbers a statistical package provides, to decide what to ask it to do in the first place, and to be able to appreciate critically what others analyzing these models are doing.
Course Prerequisites
Students taking this course should have already taken a course covering the single-equation ("regression") model, be fluent in ordinary algebra, have a good knowledge of basic mathematical statistics (including familiarity with probability distributions, expected values, and their properties), and be reasonably familiar with at least one suitable statistical software package, like SAS, S-plus, R, or STATA.
Reading:
The readings are mostly selected portions of econometrics texts, in particular:
Jeffrey M. Wooldridge. 2010. Econometric Analysis of Cross Section and Panel Data (2nd ed.). Cambridge, MA: MIT Press.
G.S. Maddala and Kajal Lahiri. 2009. Introduction to Econometrics (4thed.). New York: Wiley.
We shall also read fewer and shorter passages from some other texts and articles (see below), including:
Jan Kmenta. 1997. Elements of Econometrics (2nd ed.). Ann Arbor, MI: University of Michigan Press
and the whole of one monograph, namely
J. Scott Long. 1983. Covariance Structure Models: An Introduction to LISREL. Beverly Hills, CA: Sage, if we get to its topic (at the end of the course).