Teaching Undergraduate Econometrics (DEE 2007)

Presenter: David F. Hendry (University of Oxford)

This 38-minute video is also downloadable in high-resolution Windows Media format (269 MB).

Abstract

It is the perfect time to review the scope and practice of teaching undergraduate econometrics. Just published are two books: the first is the textbook, Hendry, D. F. and Nielsen, B. Econometric Modeling: A Likelihood Approach (Princeton University Press); the second is a major new upgrade for computer teaching of econometrics, Doornik, J. A. and Hendry, D. F. Empirical Econometric Modelling using PcGive, (Timberlake).

Although the Philosophy, Politics and Economics (PPE) degree at Oxford does not attract most maths-oriented students, David has raised option enrolment from 6 to 48 over 8 years - and yet the course goes in a year from no elementary statistical theory knowledge to (automatic) model selection in non-stationary data. The course has five central themes: likelihood; relevant empirical applications; mastery of software to implement; emphasis on graphics; rigorous evaluation of 'findings'. Theory and empirical work are taught simultaneously, with derivations introduced only from the need to understand properties. This common approach allows rapid progress - from binary IID, reinterpreted as regression on an intercept, leading to regression in general; applied to fish market data as if cross-section, reinterpreted as time series, then as a system, leading to identification, simultaneity, and cointegration; this then allows model selection to be tackled. Monte Carlo illustrates and evaluates methods and models.

Link: Slides in PDF format