ECONOMETRICS 2
Semester 2
Credits 6
Contact: Dr J.R. Hudson: J.R.Hudson@bath.ac.uk
Level: 3
Assessment: Exam 100%
AIMS & LEARNING OBJECTIVES
The aim of the course, which continues from Econometrics I taught in the first
semester, is to present a rigorous account of econometrics. By the end of the
course the students will be more confident in their use of matrix algebra, the
language of econometrics, and will have come into contact with further
empirical work using STATA. The course is a difficult and demanding one. But
econometrics is one of the foundations of the modern economists
toolkit and you cannot be an economist today or do postgraduate work without an
ability to use econometrics. At
CONTENT
The unit follows Johnson's classic text to a considerable degree, although with more emphasis on recent development in time series analysis and limited dependent variable analysis. You will cover a greater range of issues than in most other Universities. The topics covered include nonlinear least squares, ARIMA modelling, stationarity and cointegration and limited dependent variable analysis. A copy of the lecture notes can be accessed from this website. In addition, the following books will be of use:
Key Texts
1. Jack Johnson - Econometrics, McGraw Hill, paperback available. The classic text, excellent for matrix algebra and very much the text on which this and many other courses are based.
2. Pindyck and Rubinfeld, Econometric Models and Economic Forecasts. Excellent McGraw Hill text, uses matrix algebra, but not to the extent of Johnson. The best reference for ARIMA modelling, also used in the macroeconomic modelling course, paperback available.
3. Cuthbertson, Hall and Taylor, the only textbook to deal with the `new' approaches to econometrics, which in this course means stationarity and cointegration, paperback available, will shortly be in library.
4. Enders, W. Applied Econometric Time Series, Wiley and Sons. Also good for time series analysis. Heavily linked to RATS.
5. William Greene’s textbook on Econometrics – this is an excellent text
Also:
Download papers from: http://emlab.berkeley.edu/users/card/papers/geo_var_schooling.pdf
Levitt, S. D. 1996. The effect of prison
population size on crime
rates: evidence from prison overcrowding legislation. Quarterly Journal of
Economics 111:
319?351.
http://web.ebscohost.com.ezp1.bath.ac.uk/ehost/pdf?vid=4&hid=16&sid=fee321f3-f395-4956-88a1-8e8de3fc3165%40SRCSM1
The above two papers are necessary – especially the first which you will
be doing in the first class – for the classes
ORGANISATION
The course consists of a series of 12 two hour lectures + classes in the computer lab.
Topic Plan
Week 1: Non Linear Least Squares: http://staff.bath.ac.uk/hssjrh/lecture-notes1.pdf
Note in the above on page 5
On page 5 of the nonlinear least square lecture notes,
'The length of the step is chosen as to MINIMIZE (not maximise) the new value
of S. Also on the bottom of page 4 it should be:s di=gi=
-ds/dBi.
Week 2: Johansen's method of
estimating co-integrating vectors http://staff.bath.ac.uk/hssjrh/lecture-notes2.pdf
ALSO: http://staff.bath.ac.uk/hssjrh/VECM
STATA.pdf
Week 3: Probit and Logit - Multinomial http://staff.bath.ac.uk/hssjrh/MLOGIT1.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT2.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT3.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT1.pdf
Week 4: Probit and Logit
– Ordered http://staff.bath.ac.uk/hssjrh/MLOGIT4.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT5.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT6.pdf
http://staff.bath.ac.uk/hssjrh/MLOGIT7.pdf
Also see Probit measures of fit: http://staff.bath.ac.uk/hssjrh/Probit measures of fit.doc
Also see interview with Heckman and McFadden, Nobel Prize winners in Econometrics who worked with Limited dependent variables and also sample selection.
http://nobelprize.org/nobel_prizes/economics/laureates/2000/heckman-interview.html
Week 5: Heckman’s sample selection model. http://staff.bath.ac.uk/hssjrh/Heckman_LECTURE.doc
Week 6: Three stage least squares http://staff.bath.ac.uk/hssjrh/pg1.pdf
Week 7: Full Information Maximum
Likelihood http://staff.bath.ac.uk/hssjrh/pg1.pdf
Note in this lecture we introduce the concept of the trace
of a square matrix tr(A). This is simply the sum of the elements down the leading
diagonal of A. .We do no more than introduce the
concept. If you are interested see:
http://www.ee.ic.ac.uk/hp/staff/dmb/matrix/property.html
Week 8: Panel Data Techniques: http://staff.bath.ac.uk/hssjrh/PANEL.DOC
http://staff.bath.ac.uk/hssjrh/panelrat.doc
http://staff.bath.ac.uk/hssjrh/paneleg.doc
Week 9: Panel Data Techniques: Also look at this: http://video.google.com/videoplay?docid=5123879845363168993
Week 10: Panel Data Techniques
Week 11: Splines & restricted least squares http://staff.bath.ac.uk/hssjrh/splines.doc
Week 12: Revision
TYPED LECTURE NOTES:
http://staff.bath.ac.uk/hssjrh/TYPED Lecture 1 Nonlineaar least squares.pdf [
http://staff.bath.ac.uk/hssjrh/TYPED
Lecture 2 Johansen.pdf
http://staff.bath.ac.uk/hssjrh/TYPED
Lecture 3 Multin.pdf [Multinomial Logit and Probit]
http://staff.bath.ac.uk/hssjrh/TYPED
Lecture 4 Ordered.pdf
Ordered Probit]
http://staff.bath.ac.uk/hssjrh/TYPED
Lecture 7 3SLS.pdf
http://staff.bath.ac.uk/hssjrh/TYPED Lecture 8 FIML.pdf
http://staff.bath.ac.uk/hssjrh/TYPED Lecture 10 Splines.pdf
COURSEWORK: STATA None as such
But doing econometrics
without running programs is like learning to play the piano without actually
playing a piano. Hence learn this
now, or later, even after you graduate.
For notes on STATA see:
http://staff.bath.ac.uk/hssjrh/stataguide.doc
http://www.ats.ucla.edu/stat/stata/notes3/default.htm
http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm