A Critique of Structural VARs Using Real Business Cycle Theory

摘要: The main substantive finding of the recent structural vector autoregression literature with a differenced specification hours (DSVAR) is that technology shocks lead to fall in hours. Researchers have used these results argue business cycle models which rise should be discarded. We evaluate DSVAR approach by asking, derived from this misspecified when data are generated very model trying discard? find it misspecified. Moreover, misspecification so great leads mistaken inferences quantitatively large. show other popular uses level (LSVAR) also alternative state space approaches, including accounting approach, more fruitful techniques for guiding development theory.> > Replaced Staff Report No. 364(This abstract was borrowed another version item.)