Econ 570 - Spring 2009

An A Paper

I thought you might be interested in reading a paper that I would
give an A in Econ 570. This paper has all the sections I like, the
introduction is catchy without sounding like the work of a college
student, and the tables are at the end. (OK, it does not have a
table of variable definitions, but it's close to perfect.)

I found the paper referenced here just tonight:

http://econlog.econlib.org/archives/2009/04/why_do_married.html

The paper is here:

http://www.policystudies.ca/documents/AreAlltheGoodMenMarried.pdf

As you read this paper, you should come to understand that, if you
could write papers like this, you could be a professor at a research
university. You could then go to college for the rest of your life.

Midterm Answers

Lags in State and Generating Normally Distributed Random Draws

Discussion of Income Projects and Time Series Econometrics

Midterm

Q & A

Elasticity

Getting Ready for a Midterm

Serially Correlated Errors

Time series economic data is very often serially correlated. Period t+1 is closely related to period t. The errors in regressions can be thought of as left-out variables and, therefore, we expect to find serially correlated errors.

Two-Stage Least Squares

Term Paper

We discussed term paper strategies in terms of a hypothetical average cost curve analysis.

]]>

Simultaneity

What we have here is a lack of identification!

Excluded exogenous variables give us identification.

Heteroskedasticity

There are some missing pictures. We went through a model with the error variance proportional to a variable Z. A plot of the residuals vs. Z is a good way to test for heteroskedasticity. The Goldfeld-Quant Test is given in the textbook, but has practical problems. The Breusch-Pagan Test is more convenient, but is not in the textbook. (These pictures are generally available for other semesters.)

Term Paper Specifications

Organization
Title Page
Sections
Topic Sentences
Subject-Verb
No Stream of Consciousness
8.5 x 11 paper, single-sided, no cover

Tobit, Sample Selectivity, Etc.