This paper estimates the effects of technology shocks in VAR models of the U.S., identified by imposing restrictions on the sign of impulse responses. These restrictions are consistent with the implications of a popular class of DSGE models, with both real and nominal frictions, and with sufficiently wide ranges for their parameterers. This identification strategy thus substitutes theoretically-motivated restrictions for the atheoretical assumptions on the time-series properties of the data that are key to long-run restrictions. Stochastic technology improvements persistently increase real wages, consumption, investment and output in the data; hours worked are very likely to increase, displaying a hump-shaped pattern. Contrary to most of the related VAR evidence, results are not sensitive to a number of specification assumptions, including those on the stationarity properties of variables.
Bayesian VAR methods DSGE Models Identification Technology shocks