Syl5213 08
Syl5213 08
Syl5213 08
1 Purpose
Econometrics exists because much can go wrong in statistical analysis when data
are the outcomes of repeatable experiments. As part of your study you will learn
a few techniques about identifying when data are not being generated in a way
that will allow basic statistical models to be used. Some techniques that may
yield valid inference when things go wrong will be discussed. Unfortunately,
there is never any guarantee that your analysis is valid or that the assumptions
that permit valid inference are met by your data. In the end, though, obtaining
useful results from statistical analysis is as much art as science. Whether you
are able to become a competent artist or not is a function of your knowledge,
creativity, and humility. I can help you with the first part, as for the rest you
are on your own.
1
Fall 2008 2
2 Textbooks
Required
James H. Stock and Mark W. Watson, Introduction to Econometrics,
2nd edition, Addison-Wesley, 2007.
Adkins and Hill, Using Stata for Principles of Econometrics, 3rd
edition, John Wiley and Sons, 2008. Data sets used in this book are
found at http://www.principlesofeconometrics.com/
Other Sources
There is an excellent website that is provided by the authors and
publisher of your book. It can be found at
http://wps.aw.com/aw_stock_ie_2
If you purchase a new textbook, the material on this site is free. It
includes sample quizzes, data sets, tutorials, and other stuff.
Christorpher Baum’s Stata book should also be available at the
bookstore. Obviously, it is useful for learning about Stata, but it
is also useful for examples.
Adkins, Using gretl for Principles of Econometrics, 3rd edition, 2007.
It is freely available from my website,
http://www.LearnEconometrics.com/gretl/ebook.pdf.
Kennedy, Peter. A Guide to Econometrics, 6th edition. John Wiley,
2008. This is a particularly valuable book for anyone whose plan
of study includes more econometrics. Chapter 22 on what can go
wrong and what to do about it is worth $35 that Amazon charges
for this book.
3 Prerequisites
This course requires you to work with basic probability, statistics, algebra, and
to use Stata or Gretl. I may use a little matrix algebra and I will use a very
small amount of calculus. You will not be asked to derive estimators using either
of these tools. They are used in order for you to see where the estimators come
from (as opposed to believing that they come from the ether). As prerequisites I
recommend 2 undergraduate courses in statistics in addition to a good command
of algebra. You should have some notion about what random variables are, what
a probability distribution is, what a statistic is, and what a hypothesis test is.
Fall 2008 3
These are things that we will cover, but we move through them quickly. If you
haven’t learned about these before you’ll never be able to keep up. It is not
necessary that you have any previous experience with linear regression, though
this would be very helpful.
4 Course Outline
5 Software
There are two basic pieces of software that you can use to complete assignments
in class. Given a choice, I’d prefer you to use Stata. It is professional strength
software and learning how to use it will serve you well in the future. If you
commute and are unable to get regular access to Stata, gretl is offered as a free
Fall 2008 6
substitute. In addition, I have written a book that will show you how to use
gretl that is also free.
Please don’t use SPSS. It’s design encourages a particulary large number of bad
statistical practices and in poorly trained hands (95% of users) can produce
quite misleading results. Just because there is a button that does X, does not
mean you should do it! If you learn nothing else in this course, I want you
to develop a very sceptical attitude toward applied work in statistics. I would
venture to guess that fewer than 1% of applications are statistically valid. This
is due to user error and is not the fault of the software or techniques themselves.
Stata
The first is Stata. Stata is currently available in the CBA labs and on the CBA
Trading Floor.
For those interested in what Stata can do, here is a link to a Stata brochure:
http://www.stata.com/products
http://www.stata.com/capabilities
UCLA Comparison
Gretl
Gretl comes with many sample data files and a database of US macroeconomic
time series. From the Gretl web site, you have access to more sample data sets
Fall 2008 7
So, why use Gretl? Well, its free, its fast, it will work on any platform, and it
will do everything we are going to do in this class. If you want to use Gretl
instead of Stata, then feel free to do so.
Why use Stata? Stata is a professional piece of software that has many more
capabilities than Gretl. In the long-run, knowing how to use Stata could be
beneficial. On the other hand, by the time you get around to using Stata, you
may have forgotten it all and have to start from scratch anyway. As it turns
out, knowing one package well (any package) is a pretty good introduction to
other packages.
6 Data sets
Here is a list and brief description of the data sets used in the course. This gives
one an idea about the kinds of applications discussed. They are available for
download from Stock and Watson’s website as Stata data files, Microsoft Excel
spreadsheets, and (at least for some) EViews data format.
CPS data This consists of 11130 observations on 3 variables from the current
population survey.
California Test Score The California Standardized Testing and Reporting
(STAR) data set contains data on test performance, school characteristics
and student demographic backgrounds. The data used here are from all
420 K-6 and K-8 districts in California with data available for 1998 and
1999.
Economics Journal Subscriptions This data set contains data on 180 eco-
nomics journals for the year 2000.
Massachusetts Test Scores The Massachusetts data are district-wide aver-
ages for public elementary school districts in 1998. The test score is taken
from the Massachusetts Comprehensive Assessment System (MCAS) test,
administered to all fourth graders in Massachusetts public schools in the
spring of 1998.
Fall 2008 8
Auto Fatalities The data are for the ”lower 48” U.S. states (excluding Alaska
and Hawaii), annually for 1982 through 1988.
Boston Mortgage Applications The Boston HMDA data set was collected
by researchers at the Federal Reserve Bank of Boston. The data set com-
bines information from mortgage applications and a follow-up survey of
the banks and other lending institutions that received these mortgage ap-
plications. The data pertain to mortgage applications made in 1990 in
the greater Boston metropolitan area. The full data set has 2925 observa-
tions, consisting of all mortgage applications by blacks and Hispanics plus
a random sample of mortgage applications by whites.
Cigarette The data set consists of annual data for the 48 continental U.S.
states from 1985 - 1995. The series include annual per capita cigarette
sales in packs per fiscal year, the average retail cigarette price per pack,
per capita income, the average cents per pack tax (which includes the sales
tax), the tax applied to cigarettes only, and the cpi.
Tennessee STAR experiments The Project STAR public access data set
contains data on test scores, treatment groups, and student and teacher
characteristics for the four years of the experiment, from academic year
1985-86 to academic year 1988-89.
7 Exams
There will be 3 exams in the course. All exams must be taken at the designated
time. No make up exams will be given. If you miss an exam you will receive a
grade of zero.
8 Grades
Your grade in this class will be based on your performance on 3 exams, and
homework assignments.
Grades will carry the following weights and be measured according to the ac-
companying scale.
Grade Weights
Exam 1 28%
Exam 2 28%
Exam 3 29%
Homework 15%
Fall 2008 9
Grades
92%–100% A
79%–92% B
68%–79% C
57%–68% D
< 57% F
Note: One or more of the exams may contain a take-home portion that will
require you to do some work using Stata or Gretl. Although I encourage you to
collaborate with fellow students on homework assignments, I expect you to
work alone on any take-home tests.
9 Homework
There will be some homework in the course. The best way to learn econometrics
is to do econometrics. A large portion of your homework will require you to use
a computer. The computer software we are using is Stata or Gretl. Stata is a
Windows program that operates under the Windows XP operating system on
the microcomputers in the CBA lab. Gretl is free and has Mac, Windows, and
Linux versions available for download.
I will not accept late homework under any circumstance. I expect homework
to be legible and well organized. I encourage you to work with others in the
class while doing homework, and you may turn in assignments in groups of 2.
The homework receives style points, so identical answers may receive different
grades. I am predisposed to to look favorably upon work that is well organized
and legible.
Unless you are specifically told otherwise by me, all homework must be turned
in at the beginning of the class period on the date that it is due. Homework
will not be accepted if late.
10 Attendance
Regular attendance is expected. You are responsible for any material you miss
because of absence. In general, I do not permit students to copy my notes. If
you miss class and need a copy of the notes, please obtain them from one of
your classmates. And remember, if you miss an exam, you’ll earn a zero.
Fall 2008 10
11 Cheating Policy
There are two sections of this course being offered this semester. You are under
no circumstances permitted to share any information about the contents of an
exam with students in the other section. This is an extreme form of cheating. I
will seek to remove you from the course and you will receive the F! if it is within
my power to award it. I have successfully awarded this grade in this course and
will not hesitate to do it again if circumstances warrant.
Do not share information about exams with students in the other section, either
directly or indirectly. You have been warned.
Econometrics is Fun!
References
[1] L. C. Adkins, Using gretl for principles of econometrics, 3rd edition,
http://www.learneconometrics.com/gretl/ebook.pdf, 2008.
[2] L. C. Adkins and R. C. Hill, Using stata for principles of econometrics, 3rd
edition, John Wiley and Sons, 2008.
[3] Russell Davidson and James G. MacKinnon, Econometric theory and meth-
ods, Oxford University Press, 2004.
[4] Christopher Dougherty, Introduction to econometrics, 2nd ed., Oxford Uni-
versity Press, 2002.
[5] Peter Kennedy, A guide to econometrics, sixth ed., Wiley-Blackwell, 2008.
[6] SAS Institute, SAS/ETS, version 9, SAS Institute, Cary, NC, 2008,
ETS Factsheet.
[7] StataCorp LP, Stata statistical software: Release 10, 4905 Lakeway Drive,
College Station, Texas 77845, 2008, http://www.stata.com.
Fall 2008 11