Working Paper: CEPR ID: DP4537
Authors: Luca Dedola; Stefano Neri
Abstract: This Paper estimates the effects of technology shocks in VAR models of the United States, Japan and Germany, identified imposing restrictions on the sign of impulse responses. These restrictions are motivated with priors on the parameters of a class of DSGE models with both real and nominal frictions. Estimated technology shocks lead to substantial and persistent increases in labour productivity, real wages, consumption, investment and output. In contrast with most results in the VAR literature, hours worked are much more likely to increase, displaying a hump-shaped pattern. These results are shown to stem primarily from the identification strategy proposed in the Paper, which substitutes theoretical restrictions for the atheoretical assumptions on the time series properties of the data, that are the hallmark of long-run restrictions.
Keywords: Bayesian VAR methods; DSGE models; Impulse responses; Technology shocks
JEL Codes: C30; E30
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Positive technology shock (O49) | Increase in labor productivity (O49) |
| Positive technology shock (O49) | Increase in real wages (J39) |
| Positive technology shock (O49) | Increase in output (E23) |
| Positive technology shock (O49) | Increase in consumption (E21) |
| Positive technology shock (O49) | Increase in investment (E22) |
| Positive technology shock (O49) | Increase in hours worked (J29) |
| Technology shocks (O33) | Variability in labor productivity (J24) |
| Technology shocks (O33) | Variability in output (C29) |
| Technology shocks (O33) | Variability in hours worked (J22) |