Temple University

Department of Econoimcs

Economics 8009 and 8119

Econometrics I and II

Purpose of the Courses: Where, why and how does it begin? Econ 8009 will presume a working knowledge of the rudiments of probability and hypothesis testing. Students who have gone only as far as, say Stat 22, will have to run very hard to keep up. Students who have had an undergraduate econometrics course will have a much easier time of it. There is also the presumption that the students have a working knowledge of calculus and linear algebra. To refresh everyone's memory and get us all in synch I review statistics at the level one might find somewhere between Stat 22 and the Stat 501/502 sequence. We also review the rudiments of linear algebra and some special operations the students will need in later weeks of the course.

What do I and the students do as the course unfolds? Econ 8009 and 8119 can be construed as the process of generalizing the material covered in Stat 21 and 22. The generalization begins with the notion that the location parameter of a distribution can be modeled as a conditional mean. One must then be able to test hypotheses about the parameters in that conditional mean. As the course unfolds we relax the assumptions made in the original specification of the distribution of the random variable being modeled or its conditional mean and explore the ways in which the earlier tests of hypothesis must be altered.

What do I lecture about? Principally my lectures are derivations, statements of theorems and proofs. The students get their hands dirty with data in their homework assignments and in their papers.

What are the key assignments and/or student evaluations? The students must do homework exercises, write an empirical paper, and write two or more exams in each course.

How does the course end and why? Econ 8009 is organized along the lines of establishing the classical assumptions of statistical inference, then exploring the consequences of violating them in the classical regression model. One could not follow the material in 8119 without a good understanding of the pathologies considered in 8009. Econ 8119 is organized topically and is more closely related to solving problems encountered in applied economic research. Because of the uses to which economists put statistics, neither course could be said to end with a crescendo built from the earlier themes.

What do I want the students to believe and why? Because of the level of preparation of the students and the way economics is practiced in North America, my courses are not ideological. Nor do they introduce the students to the ideological divide which can be found in some statistics departments. The students should believe that there is received method for structuring empirical investigations and that there are ways to overcome technical problems encountered in real world problems. They should also recognize that the structure of empirical work and the treatments offered for pathological problems are based on the foundations of classical statistical theory.

Are there other ways to organize the course? My conclusion from
reading the Eno River materials is that econometrics is taught in a very standard format
throughout the country. There are some exceptions which reflect the idiosyncracies of some
of the most highly regarded econometricians. Unlike instructors at other institutions, I
do not assign original scholarly articles. The texts I use have done a superlative job of
integrating and explicating the literature.

In what ways does it teach students how scholars work in the field? The students are not prepared to do work in developing theoretical econometrics. My goal is to make scholarly tracts in econometrics accessible to the students so that they are able to apply those methods in their chosen subdiscipline of economics. By the end of the sequence the students should be able to question the empirical findings of others within the context of classical statistics. The students should also be able to conceive, design and conduct empirical projects by the end of the sequence.

How does it lead to the cutting edge? The sequence of econometrics courses, in combination with the theory sequences, should lead the students to the cutting edge of their chosen subdiscipline in economics. The students who come to Temple do not have the background to do econometric research. By the time I am finished with them they should be able to read the research relevant to their own applied research needs.

How does the course connect with other courses in the department? The econometrics sequence should be part of the foundation on which the field courses and dissertation are built.

Where will the students encounter their greatest difficulties? The class sessions are very difficult because of the need to integrate many past mathematical skills with new concepts in statistics. Necessarily, the classroom material is very abstract. The students are further challenged when they must then bring their newly acquired theoretical knowledge to bear on a numerical problem, often outside of the classroom.

How do I address these difficulties? Most often the problem is overcome by doing more of it. I am unrelenting in the assignment of homework. I grade the homework and return it quickly. Often times I will review problems which were particularly troublesome for many of the students.

Are there metaphors for your course? The class sessions are like an obstacle course. A paradigm is explained. An assumption in the paradigm is weakened and we look for a way around the problem. The homework and paper are like a football game. The students have learned the rules and trained for the game during class. In order to complete a paper or a homework assignment they must conceive and execute a game plan.