شکل گیری عادت و انتقال مالی در اقتصاد باز

In this paper we analyze the ability of an open economy version of the neoclassical model to account for the time-series evidence on fiscal policy transmission. Revisiting the evidence, we find that i) government spending raises output, while inducing a simultaneous decline of investment and the current account and ii) the responses of output and investment are more muted in more open economies while current account deficits tend to be larger. Turning to the model, we explore the role of habit formation for fiscal policy transmission. Specifically, we show that the model can account for the evidence if consumption behavior is characterized by habit formation and the terms of trade adjust endogenously.

Interest in specific aspects of the fiscal transmission mechanism has grown recently, both among policy makers and within the academic literature. A number of studies have applied different identification schemes to establish time-series evidence on the effects of fiscal shocks, see, e.g., Blanchard and Perotti, 2002, Mountford and Uhlig, 2009 and Ramey, 2011. Others have explored the ability of quantitative business cycle models, both of the neoclassical and of the New Keynesian variety, to account for the evidence, see Burnside et al., 2004 and Gali et al., 2007, respectively. Most of the analyses have been confined to closed economy models, however. In the present paper we take up the following question instead: to what extent can an open economy version of the neoclassical model account for the time-series evidence on fiscal policy transmission?1 To set the stage for our theoretical analysis, we briefly revisit the evidence on the fiscal transmission mechanism. We estimate a vector autoregression (VAR) model on quarterly time-series data for Australia, Canada, the UK and the US covering the period 1980-2007 and identify government spending shocks under the assumption that government spending is predetermined relative to the other variables included in the VAR model (see Blanchard and Perotti (2002)). Our results are largely in line with earlier findings. First, an exogenous increase in government spending has virtually no effect on private consumption, raises output, and lowers both investment and the current account, see, for instance, Monacelli and Perotti (2010).2 Second, we find that the decline of investment is less pronounced in those economies of our sample which are more open to trade. At the same time, the current account deficit tends to be larger and the effects on output more muted in more open economies.3 We put forward a variant of the neoclassical model and assess its ability to account for these findings. Specifically, we draw on earlier work by Baxter and King (1993) but consider a semi-small open economy. As in Buiter (1987), the economy is small in world capital markets so that the world interest rate is given, but large enough in the world good market to influence the relative price of the domestic good, that is, the terms of trade. Importantly, we also allow for the possibility that preferences are not time separable but characterized by habit forming behavior, as in, e. g., Campbell and Cochrane (1999) and Carroll et al. (2000). Both endogenous terms of trade and habit persistence in consumption play a crucial role for the fiscal transmission mechanism in open economies. To see this, consider an exogenous, but temporary increase of government spending. It induces households to reduce savings, as they try to avoid a large reduction in consumption and/or a large increase in labor supply. Reduced savings imply a decline of investment or the current account, or both. The decline of total savings, that is, the sum of investment and the current account, is larger, if preferences are characterized by habits, because in this case households' desire to smooth consumption is more pronounced. If the terms of trade are exogenous, the current account declines, but investment increases, because the increased labor supply raises the return to domestic capital. Investment and the current account may both decline, however, if increased government spending temporarily appreciates the terms of trade. In this case the return on foreign bonds increases because the terms of trade deteriorate in the long-run to clear the home good market. This, all else equal, makes borrowing from abroad relatively more expensive and induces a decline of domestic investment. We thus find that the model can account for a simultaneous decline of investment and the current account only if we assume habits and consider an endogenous adjustment of the terms of trade, that is, a semi-small open economy.4 Regarding our second set of findings, which concerns the role of trade openness, we also find the performance of the model increased, once we allow for habits. It is only in this case that we find current accounts deficits to increase in openness. We show, however, that the output effects of spending shocks decline in openness under time-separable preferences as well as under habits. This finding seems particularly noteworthy, because Perotti (2005) documents smaller output effects of spending shocks in a post-1980 sample relative to results for the pre-1980 period. Since the countries in his sample have become more open over time, this trend may be partly responsible for the reduced output effects of government spending. Two papers are particularly closely related to our analysis. Karayalcin (1999) simulates a small open economy model and finds that a temporary fiscal impulse stimulates investment and induces a current account deficit, as higher labor supply raises the marginal product of capital above the world interest rate. Karayalcin (2003) investigates the effects of a permanent public spending shock in a small open economy model with time non separable preferences. His framework differs in two major respects from ours: (i) the terms of trade are exogenous, and (ii) the stock of habits is introduced into the subjective time discount rate. The first assumption implies that the return on bonds remains fixed and investment decisions are determined by labor supply which alters the return on capital. In addition, as Karayaçin assumes that consumption and leisure are non-separable, he finds that labor supply, and hence output, falls on impact. The resulting decline in the return of capital triggers a fall in investment and a current account surplus. The second assumption implies that the marginal utility of wealth reacts more strongly to the fiscal shock in the short-run. Numerically, Karayalçin finds that labor overshoots on impact which in turn amplifies the current account surplus.5 The remainder of the paper is organized as follows. In Section 2, we discuss VAR evidence on the macroeconomic effects of government spending shocks. In Section 3, we develop an open economy version of the neoclassical model with habit formation. Section 4 provides an analytical exploration on the role of habits for the transmission of fiscal shocks. In Section 5, we report results from numerical simulations and discuss the role of trade openness. We also conduct a sensitivity analysis with respect to the shock duration and the financing scheme. Section 6 summarizes our main results and concludes.

In this paper we have analyzed the fiscal transmission mechanism in a semi-small open-economy version of the neoclassical model and how it is affected by habit formation in consumption. To guide our analysis, we have revisited VAR evidence on the effects of government spending shocks in Australia, Canada, the UK and the US. In line with several earlier studies, we find that government spending simultaneously reduces private investment and the current account. The model economy is able to account for this finding, in case we allow for habit formation. Following a temporary rise in government spending, consumption falls, labor increases and terms of trade improve on impact, whether agents care about their past consumption or not. Yet, in case of habits, the increase in labor supply is more moderate. This, ceteris paribus, lowers the return on domestic capital relative to the return on foreign bonds and induces a decumulation of domestic capital relative to foreign bonds. While overall savings decline in response to temporarily increased government spending, only in case of habits is the drop in savings large enough for both investment and the current account to decline. A second striking observation regarding the time-series evidence concerns how the degree of openness affects fiscal policy transmission. We therefore analyze the effect of trade openness on fiscal policy transmission within the neoclassical model. In line with the evidence, we find that higher trade openness moderates substantially the crowding-out of investment and amplifies the current account deficit (in the case of habits) following a temporary increase in government spending. In addition, we find that the output effects of government spending decline in openness, as suggested by the VAR evidence. This result may also provide a rationale for evidence put forward by Perotti (2005) who documents declining output effects of government spending when comparing time-series evidence for the pre-1980s with the post-1980s period; trade openness certainly increased considerably between the two sample periods. In concluding we stress a number of caveats. First, our time series evidence on openness can only be considered tentative, as it is based on observations for four countries. Reassuringly, however, it confirms earlier findings for a larger sample by Beetsma et al. (2008). Second, regarding the model's performance, we note that the model fails to match the evidence quantitatively along a number of dimensions. Notably, it fails to predict the size of current account deficits. Finally, while a number of studies provide evidence that government spending tends to depreciate exchange rates, at least for the US (see, e.g., Enders et al. (2011)), the neoclassical model predicts that government spending appreciates the terms of trade in the short-run. We leave a further analysis of these issues for future research.