Dienstag, 22. Oktober 2013

Lawrence Klein

Lawrence Klein ist am 20. Oktober gestorben.

Der Nobelpreisträger von 1980 war eine der großen Figuren im Bereich der makroökonometrischen Modellierung.

In einem Interview, das er 1987 mit der Zeitschrift "Econometric Theory" geführt hat, fand ich folgende Passage über den Fortschritt in der Makroökonometrie interessant:
"This is one idea that I thought a lot  about over the years, whether there is ever a breakthrough. Going back to the 1940s, we thought that simultaneous equations  methods  were going  to  be a breakthrough and that they would have enormous power and accuracy. Next we said that when we moved from annual to quarterly data we would have a new breakthrough, and then we said if  we can draw upon  cross-section data we would  have an additional breakthrough. Later we said if we had anticipations data we would have a breakthrough. The next big step was to have been through the use of control theory.  Now  people are saying if one models with consistent expectations, rational  expectations,  or  vector  autoregressions,  then  we  have promising new tools.  When we went to  nonlinearities and easy handling of  them and ARIMA methods through computers, we thought that we would have breakthroughs.  I regard none  of  these  as a complete  breakthrough in terms of making very significant  gains in accuracy of  economic-econometric  judgments.  But a lot  of  things build on  one  another,  with very tiny improvements."
Klein beharrte bis zuletzt auf dem Standpunkt, dass empirische Modelle nicht in erster Linie im Hinblick auf ihre rigorose Herleitung bzw. Mikrofundierung sondern hinsichtlich ihrer Fähigkeit, aggregierte Daten zu prognostizieren, bewertet werden sollten.

Keine Kommentare:

Kommentar veröffentlichen