The Response of Hours to a Technology Shock: A Two‐Step Structural VAR Approach
提出一种两步结构向量自回归方法，用于一致估计工时对技术冲击的响应。模拟和实际数据表明，该方法优于传统SVAR，且美国数据呈现短期下降后驼峰状正响应。
The response of hours to a technology shock is a controversial issue in macroeconomics. Part of the difficulty lies in that the estimated response is sensitive to the specification of hours in structural vector autoregressions (SVARs). This paper uses a simple two‐step approach to consistently estimate the response of hours. The first step considers a SVAR model with a relevant stationary variable, but excluding hours. Given a consistent estimate of technology shocks in the first step, the response of hours to this shock is estimated in a second step. Simulation experiments from an estimated dynamic stochastic general equilibrium (DSGE) model show that this approach outperforms standard SVARs. When applied to U.S. data, the two‐step approach predicts a short‐run decrease followed by a hump‐shaped positive response. This result is robust to other specifications and data.