SEARCHING FOR ENDOGENOUS BUSINESS CYCLES IN THE U.S. POSTWAR ECONOMY

This paper explores the possibility that booms and busts are endogenous with an econometric examination of the U.S. economy. Although existing empirical results suggest the contrary, a theoretical literature discounts the stability verdict on the grounds that it is based on linear econometric evidence which is unreliable when the underlying phenomenon is non-linear. We study a simple Keynesian cycle model which is theoretically general enough to include convergence to equilibrium or endogenous cycles as special cases. We find one particular nonlinear econometric specification which exhibits a limit cycle when estimated on postwar data, while the estimate of its global linear approximation converges to a fixed equilibrium. Our results are an empirical investigation of the boundary between stability and instability.