This paper investigates whether technology improvements are contractionary. Using a measure of aggregate technology change that controls for varying utilization of capital and labor, non-constant returns, and imperfect competition, the authors find that when technology improves, input use and non-residential investment fall sharply, while output changes little. With a lag of several years, inputs and investment return to normal, and output rises strongly. The results are consistent with sticky-price models, which predict that technology improvements reduce short-run investment demand and may lower output. However, they are inconsistent with standard one-sector real-business-cycle models, which generally predict that technology improvements are expansionary. The authors construct a measure of aggregate technology by "purifying" sectoral Solow residuals and aggregating across sectors. They find that purified technology varies about half as much as total factor productivity (TFP), and that technology fluctuations are countercyclical, being significantly negatively correlated with inputs but uncorrelated with output. The results suggest that technology improvements are contractionary on impact, with input use and investment falling in the short run and output rising in the long run. The paper discusses alternative interpretations of the results, including the possibility that demand shocks explain the procyclical nature of observed TFP. The authors also consider the implications of their findings for business-cycle modeling, noting that they are inconsistent with standard parameterizations of frictionless RBC models but consistent with dynamic general-equilibrium models with sticky prices. The results suggest that technology improvements are contractionary in the short run, but may be expansionary in the long run. The paper concludes that the findings are robust and suggest that technology improvements are contractionary on impact.This paper investigates whether technology improvements are contractionary. Using a measure of aggregate technology change that controls for varying utilization of capital and labor, non-constant returns, and imperfect competition, the authors find that when technology improves, input use and non-residential investment fall sharply, while output changes little. With a lag of several years, inputs and investment return to normal, and output rises strongly. The results are consistent with sticky-price models, which predict that technology improvements reduce short-run investment demand and may lower output. However, they are inconsistent with standard one-sector real-business-cycle models, which generally predict that technology improvements are expansionary. The authors construct a measure of aggregate technology by "purifying" sectoral Solow residuals and aggregating across sectors. They find that purified technology varies about half as much as total factor productivity (TFP), and that technology fluctuations are countercyclical, being significantly negatively correlated with inputs but uncorrelated with output. The results suggest that technology improvements are contractionary on impact, with input use and investment falling in the short run and output rising in the long run. The paper discusses alternative interpretations of the results, including the possibility that demand shocks explain the procyclical nature of observed TFP. The authors also consider the implications of their findings for business-cycle modeling, noting that they are inconsistent with standard parameterizations of frictionless RBC models but consistent with dynamic general-equilibrium models with sticky prices. The results suggest that technology improvements are contractionary in the short run, but may be expansionary in the long run. The paper concludes that the findings are robust and suggest that technology improvements are contractionary on impact.