Business cycle dynamics in a small open economy.(ANTHOLOGY)

The small open economy (SOE) real business cycle model of Mendoza (AER, 1991, 797-818) uses Greenwood, Hereowitz, and Hoffman (AER, 1988, 402-17) preferences in order to yield a countercyclical trade balance present in small open economies such as Canada, wherein the correlation equals -0.29 (based on quarterly data). Subsequent attempts by Schmitt-Grohe and Uribe (JIE, 2001, 163-85) and Correia, Neves, and Rebelo (EER, 1995, 1089-113) to incorporate standard preferences of King, Plosser, and Rebelo (JME, 1988, 195-232) to the model have been unsuccessful in that they lead to a positive correlation between output and the trade balance. The fact that SOE models with GHH preferences can obtain a negative correlation (-0.89 for Canada in Letendre, 2004, CJE), while a standard preference model yields a procyclical trade balance, is indicative that the wealth effect, equal to zero in the GHH preferences, plays a crucial role in determining business cycle dynamics in a SOE model and should be further examined.

One way to examine the wealth effect in the SOE model is to consider preferences that allow for varying degrees of the effect. The preferences of Jaimovich and Rebelo (2007, working paper) are given by

U = [E.sub.0][[infinity].summation over (t=0)] [[beta].sup.t] [(C - [eta][N.sup.[theta].sub.t] [X.sub.t]).sup.1- [sigma]] -1/1 - [sigma],

where [X.sub.t] equals [C.sup.[gamma].sub.t] [X.sup.1-[gamma].sub.t-1], C is consumption, N is hours worked, [theta] determines the elasticity of labor supply, and [gamma][epsilon][0,1] determining the strength of the wealth effect. Moreover, standard preferences of King, Plosser and Rebelo and GHH preferences are nested in the the Jaimovich and Rebelo preferences, the former emerging when [gamma] is set at 1 and the latter when [gamma] equals 0. Thus, the intensity of the wealth effect rises as [gamma] increases.

The absence of a wealth effect in the SOE model with GHH preferences implies that consumption smoothing is eliminated. Hence, consumption volatility will be higher in a model with GHH preferences as compared to a model with non-GHH preferences. Moreover, as [gamma] increases, the strength of the wealth effect increases, decreasing the volatility of consumption. Preliminary results for a calibrated SOE model for Canada indicate that for [gamma]=0.25, 0.5, and 0.75, the volatility of consumption is 0.8656, 0.7406, and 0.5701, respectively (based on quarterly data). This has a direct impact on the trade balance. As agents smooth out consumption, with the rise in [gamma], they do not use the trade balance for consumption smoothing as they do in the GHH model. Thus, the trade balance volatility falls and, moreover, the trade balance moves more closely with output to keep the resource constraint balanced. The findings thus far show that setting [gamma] equal to 0.25, 0.5, and 0.75 obtains a trade balance output correlation of -0.7922, -0.7356, and -0.2887, respectively. Consequently, the lack of an operating wealth effect alone does not imply a countercyclical trade balance, it is the strength of the wealth effect on labor supply that determines trade balance dynamics.

Published online: 13 January 2009

JEL F41 * E32

Z. Janko ([mail])

University of Calgary, Calgary, AB, Canada

e-mail: zjanko@ucalgary.ca

Print Get the Mag Weekly Updates