In this paper, we identify the technology shock at business cycle frequencies to improve the performance of structural vector autoregression models in small samples. To this end, we propose a new identification method based on the spectral decomposition of the variance, which targets the contributions of the shock in theoretical models. Results from a Monte-Carlo assessment show that the proposed method can deliver a precise estimate of the response of hours in small samples. We illustrate the application of our methodology using US data and a standard Real Business Cycle model. We find a positive response of hours in the short run following a non-significant, near-zero impact. This result is robust to a large set of credible parameterizations of the theoretical model.

The empirical importance of news shocks—anticipated future shocks—in business cycle fluctuations has been explored by using only actual data when estimating models augmented with news shocks. This paper additionally exploits forecast data to identify news shocks in a canonical dynamic stochastic general equilibrium model. The estimated model shows new empirical evidence that technology news shocks are a major source of fluctuations in US output growth. Exploiting the forecast data not only generates more precise estimates of news shocks and other parameters in the model, but also increases the contribution of technology news shocks to the fluctuations.

A recent finding of the SVAR literature is that the response of hours worked to a (positive) technology shock depends on the assumed order of integration of the hours. In this work we relax this assumption, allowing fractional integration in hours and productivity. We find that the sign and magnitude of the estimated responses depend crucially on the identification assumptions employed. Although the responses of hours recovered with short-run (SR) restrictions are positive in all data sets, long-run (LR) identification results in negative, although sometimes not significant responses. We check the validity of these assumptions with the Sims procedure, concluding that both LR and SR are appropriate to recover responses in a fractionally integrated VAR. However, the application of the LR scheme always results in an increase in sampling uncertainty. Results also show that even the negative responses found in the data could still be compatible with real business cycle models.