Output Dynamics in Real-Business-Cycle Models

This 1995 American Economic Review paper by Timothy Cogley and James Nason tests whether standard real-business-cycle (RBC) models can reproduce two well-documented stylized facts about U.S. GNP growth: short-run positive autocorrelation (a “business-cycle” periodicity, with spectral power peaking around 3.2 years per cycle) and an important trend-reverting (transitory) component with a hump-shaped impulse response. Simulating several baseline RBC models (including King-Plosser-Rebelo-type and Christiano-Eichenbaum models) 1,000 times each and comparing simulated to actual autocorrelation and impulse-response statistics, the authors find that baseline models are rejected at the 1% level or better on the autocorrelation dimension, because their propagation mechanisms are weak: since productivity shocks are modeled as a random walk and capital accumulation does little to transform that impulse, model-implied output growth is “close to being white noise,” with output dynamics essentially inheriting the dynamics of the underlying technology and government-spending shock processes rather than generating genuine internal propagation. Adding gestation lags or quadratic capital-adjustment costs does not fix this – these mechanisms still fail to generate serial correlation or a plausible transitory component, because the resulting change in the capital flow is small relative to the capital stock. Models with labor adjustment costs or labor hoarding (following Burnside, Eichenbaum, and Rebelo) do partially succeed: they endogenously generate the correct sign and rough shape of output growth’s autocorrelation. However, even these labor-adjustment-cost models cannot match the magnitude of the actual transitory impulse response without implausibly large transitory shocks, so the authors conclude that current RBC models must rely heavily on exogenous, add-on sources of dynamics (e.g., serially correlated shock processes) rather than genuine internal propagation mechanisms, and that RBC theorists should devote more attention to developing stronger internal propagation.

James M. Nason | Macro Paper Warehouse

Output Dynamics in Real-Business-Cycle Models