In this paper, we present the results of an empirical and theoretical investigation into the effects of government spending shocks on consumption, output, the trade balance, and the real exchange rate. Our empirical analysis uses data from a panel of four industrialized countries, the United States, the United Kingdom, Canada, and Australia, over the post-Bretton Woods period. We employ a structural vector autoregressive representation of the data. Following Fatás and Mihov (2001) and Blanchard and Perotti (2002), we identify government spending shocks by assuming that no variable other than government spending shocks themselves can affect government spending contemporaneously.
We find that a positive innovation in government spending causes an expansion in output, an expansion in consumption, a deterioration of the trade balance, and a depreciation of the real exchange rate (that is, a decline in domestic prices relative to exchange-rate-adjusted foreign prices). The effects of government spending shocks on domestic aggregate activity and private absorption have been extensively studied in the related empirical literature. Our finding that government spending shocks raise output and consumption is consistent with previous studies that have used identification assumptions and estimation techniques similar to those we employ in the present paper (e.g., Rotemberg and Woodford, 1992; Blanchard and Perotti, 2002; Fatás and Mihov, 2001; Perotti, 2004, 2007; and Gal?, López-Salido, and Vallés, 2007).
The effects of government spending shocks on the external sector of the economy have received considerably less attention. Notable exceptions are Corsetti and Müller (2006) and Monacelli and Perotti (2006). Our results are most similar to those reported in the second of these two studies. The main difference between our empirical strategy and those adopted in the papers cited above is the pooling of data across countries. We justify a panel analysis by observing that the identified effects of government spending shocks, particularly on consumption and the real exchange rate, are to a large extent qualitatively similar across the individual countries considered. The purpose of our panel approach is to obtain an efficient estimate of a single benchmark against which to evaluate our proposed theoretical explanation of the transmission of government spending shocks.
A central contribution of our investigation is to advance and test a theoretical explanation for the observed effects of government spending shocks based on the deep habit mechanism developed by Ravn, Schmitt-Grohé, and Uribe (2006). To this end, we introduce deep habits into a two-country model. Under deep habits, an increase in aggregate demand provides an incentive for firms to lower markups. Thus, an increase in government spending in the domestic economy leads to a decline in domestic markups relative to foreign markups. In this way, the domestic economy becomes less expensive relative to the foreign economy, or, equivalently, the real exchange rate depreciates. At the same time, a decline in domestic markups shifts the labor demand curve outward, giving rise to an increase in domestic real wages. In turn, the rise in wages induces households to substitute consumption for leisure. This substitution effect may be strong enough to offset the negative wealth effect stemming from the increase in public absorption, resulting in an equilibrium increase inprivate consumption.
To stress the ability of deep habits to explain the empirical regularities associated with government spending shocks, we deliberately abstract from a number of alternative frictions that have been proposed as potential theoretical explanations in the related literature. Specifically, we abstract from sticky prices, rule-of-thumb consumers, and nonseparabilities between consumption and leisure in preferences.
