Research indexBook chapter
Comment
Quah
15 Dec 1994
NobleIDark:/48914/w1/20260522/04CF9431DOI10.1093/oso/9780198288596.003.0017
Comment
Quah · 1994 · Subscription
NobleIDark:/48914/w1/20260522/04CF9431DOI10.1093/oso/9780198288596.003.0017
The full text is at the publisher
You can read the abstract here, or open it at the publisher.
Abstract
Abstract In the minds of many economic theorists and many traditional econometricians, vector autoregression (VAR) analysis is one of the worst things that could have happened to the scientific research programme that is business-cycle study. Users of VARs commit every crime imaginable: they estimate reduced forms—and hardly ever calculate formal specification tests—instead, preferring to work with a fiction known as a just-identified model. These practitioners refer to things as structural that no econometrician—working in either standard simultaneous equations, standard time series, or standard rational expectations paradigms—would recognize to be structural. (Within those three groups, of course, none acknowledges any of the others as structural either.) No undergraduate econometrics course teaches the tokens of VAR research dialogue: impulse response functions and forecast error variance decompositions.