اختیارات در مقابل چشم انداز بی انتها :مفاهیم نظریه و سیاست های پولی برای یک اقتصاد باز

Compared to the standard Phillips curve, an open-economy version that features a real exchange rate channel leads to a markedly different target rule in a New Keynesian optimizing framework. Under optimal policy from a timeless perspective (TP) the target rule involves additional history dependence in the form of lagged inflation. The target rule also depends on more parameters, notably the discount factor as well as two IS and two Phillips curve parameters. Stabilization policy in this open economy model is no longer isomorphic to policy in a closed economy. Because of the additional history dependence in an open economy target rule, price level targeting is no longer consistent with optimal policy. The gains from commitment are smaller in economies where the real exchange rate channel exerts a direct effect on inflation in the Phillips curve.

For some time now central banks and academics have been preoccupied with the way monetary policy ought to be conducted in an era of relative price stability. Woodford (1999a) proposes that the course of monetary policy in a forward-looking New Keynesian framework be set from a timeless perspective (commitment). This form of policy has a number of desirable features. To begin with, policy from a timeless perspective introduces history dependence into the conduct of monetary policy because it is based on an optimal policy rule that depends on the change in the output gap. The policy instrument responds to a cost-push shock in the current and subsequent periods until the target variables return to their original targets. The gradual adjustment process gives rise to persistence in the behavior of the output gap and the rate of inflation. Because the conduct of policy is history-dependent under policy from a timeless perspective, this strategy dominates pure discretion under which the response of the target variables to the cost-push shock is confined to the current period. Moreover, since policy from a timeless perspective is a time-consistent form of optimal policy under commitment it serves as a standard of comparison for forms of discretionary policy that also inject an element of history dependence into policymaking such as price level targeting, a speed limit policy, nominal income growth targeting, average inflation targeting or money growth targeting. Jensen, 2002, Nessén and Vestin, 2005, Soederstroem, 2005 and Vestin, 2006, and Walsh (2003) evaluate the aforementioned discretionary strategies in a closed economy setting and verify to what extent these policies achieve the optimal stabilization results under policy from a timeless perspective. This paper focuses on two key differences between policy in an open and a closed economy. First, it analyzes optimal monetary policy from a timeless perspective in a simple forward-looking framework to show that the open-economy target rule is far more complex than its closed-economy counterpart. Optimal stabilization policy in this open economy model is no longer isomorphic to policy in a closed economy. Second, the paper examines the connection between optimal policy and price level targeting and finds that the latter is not synonymous with the former in the proposed open economy model. Central to our discussion of policy in an open economy is a Phillips curve that features a real exchange rate channel. This exchange rate channel appears in the aggregate Phillips curve because domestic firms are concerned about their competitiveness at home and in world markets where their products compete with those produced by foreign firms. Briefly, an important objective of the typical cost-minimizing domestic firm is to avoid fluctuations in its firm-specific terms of trade. Hence an incipient rise in the foreign price of the competing foreign good or a rise in the nominal exchange rate leads the typical domestic firm to raise the price of its output. Thus external factors induce a firm to alter the price of output, which it sets in domestic currency. Earlier contributions that examine the implications of the existence of an exchange rate channel for the conduct of monetary policy are Ball, 1999, Walsh, 1999 and Svensson, 2000, and Guender (2006). Ball motivates the real exchange rate channel in the Phillips curve in a backward-looking framework by assuming that foreign producers are concerned only about receipts in their home currency. Any change in the nominal exchange rate is offset by adjusting the nominal price of the good in the foreign country. Positing a linear target rule, Ball finds that optimal policy requires a central bank to follow a monetary conditions index rather than a Taylor-type rule. Walsh (1999) derives an open-economy Phillips curve in a forward-looking model where the nominal wage demands are tied to the CPI. In a model that mixes elements of backward- and forward-looking behavior, Svensson (2000) introduces a Phillips curve where the expectation, formed in the past, of the change in the real exchange rate affects the current rate of inflation. He discusses a number of different policy strategies under discretion but does not derive the underlying endogenous target rules. Such an explicit target rule is derived in a forward-looking open economy model by Guender (2006) where firms are guided in their domestic pricing decisions by a benchmark price that is set in the world market. Such pricing behavior at the firm-level gives rise to an aggregate Phillips curve that depends on the real exchange rate and a target rule guiding optimal monetary policy that includes demand-side parameters. Guender considers only optimal policy under discretion.1 Other contributions downplay or dismiss the importance of a real exchange rate channel in the Phillips curve. Drawing on empirical evidence that shows only weak correlations between changes in the nominal exchange rates and inflation rates for a number of countries, McCallum and Nelson (2000, p. 89) are sceptical about the existence of a direct exchange rate channel and its relevance in policymaking. Moreover, in their theoretical set-up of an open economy the effect of the real exchange rate on the output gap is neutralized as it affects both the level of actual output and the level of potential output in the same way. As a consequence, their Phillips curve is the same as in a closed economy. Clarida et al., 2001 and Clarida et al., 2002 and Gali and Monacelli (2005) derive essentially a closed-economy Phillips curve too except that the coefficient on marginal cost is sensitive to the degree of openness of the economy. These papers assume that firms that operate in small open economies set domestic prices without reference to world market prices or consideration of the effects of terms of trade changes on their competitiveness. This paper takes a contrary view. It emphasizes that domestic firms value stability of their terms of trade in addition to stable prices. This concern forces cost-minimizing domestic firms to take account of expected changes in their firm-specific terms of trade when altering the domestic price of output. Changes in the nominal exchange rate and changes in the price charged by competing foreign firms are beyond the control of the typical domestic firm. Yet such changes affect its competitiveness. The only way that a domestic firm can counteract such pressure is to adjust its domestic price in such a way so that the overall cost to the firm is minimized. If such pricing behavior applies to the typical firm, aggregation over all firms leads to a Phillips curve where the expected change in the real exchange rate impacts on domestic inflation.2 The implications for optimal monetary policy are stark. Once the existence of a real exchange rate channel in the Phillips curve is acknowledged, the optimal target rule under policy from a timeless perspective becomes vastly different from the standard target rule.3 The new target rule depends on multiple parameters that appear in both the IS relation and the Phillips curve as well as on the discount factor. The proposed target rule is history-dependent too but differs from the standard target rule in a critical way: under policy from a timeless perspective the proposed target rule also features the lagged rate of inflation in addition to the lagged output gap. However, unlike the output gap, the rate of inflation does not enter the target rule in first-difference form. Optimal stabilization policy changes dramatically. In the absence of a real exchange rate channel in the Phillips curve it is optimal for the policymaker to fix the output gap as doing so ensures an optimal response to demand-side disturbances and leaves domestic inflation unaffected. If this channel does exist, however, the policymaker can no longer perfectly offset demand-side disturbances. Specifically, the policymaker’s relative aversion to inflation variability shapes his response to demand-side disturbances. Both inflation and the output gap deviate from target. Further analysis shows that a real exchange rate channel in the Phillips curve matters greatly in other policy-related contexts. Comparing policy from a timeless perspective to discretion, we find that the gains from commitment are smaller in an open economy where a real exchange rate channel is operative in the Phillips curve. Under policy from a timeless perspective, an adverse cost-push shock prompts the policymaker to “lean with the wind”, i.e. lower the nominal interest rate if there is no exchange rate channel in the Phillips curve. If this channel exists, then the policymaker raises the interest rate. Such an ambiguous response cannot occur under discretion irrespective of whether a real exchange channel exists or not. The final noteworthy finding concerns price level targeting in an open-economy framework. Woodford (1999a) and Vestin (2006) argue that in a simple closed economy forward-looking model price level targeting is consistent with optimal policy from a timeless perspective. This result does not carry over to the open economy framework proposed in this paper. There is a simple explanation for this result. Because of a real exchange rate channel in the Phillips curve the target rule under policy from a timeless perspective depends on the lagged rate of inflation. This has the effect of augmenting the history dependence of optimal policy and rules out expressing the target rule governing price level targeting in such a way so as to be consistent with the target rule underpinning optimal policy from a timeless perspective. With the targeting rules being incongruous, the delegation of a price level target to a central banker with the requisite aversion to price level variability does not conform to optimal policy from a timeless perspective in an open economy even if the shocks follow a white noise process. The organization of the remaining parts of the paper is as follows. Section 2 introduces the model. Section 3 analyzes the conduct of optimal monetary policy from a timeless perspective. Section 4 examines the case of pure discretion. Section 5 compares and contrasts the two forms of optimal policy. Section 6 takes up the discussion of the compatibility of price level targeting with optimal policy from a timeless perspective in an open economy framework. Section 7 offers a brief conclusion.

This paper has examined the design of optimal monetary policy in an open economy version of the forward-looking model. Its major finding is that the existence of a real exchange rate channel in the Phillips curve changes the design of optimal policy in no small measure. First and foremost, the target rule under policy from a timeless perspective becomes more complex. The lags of both target variables appear in it, thus making the conduct of policy more history-dependent. The discount factor as well as IS and Phillips curve parameters determine the weights attached to the change in the output gap and the lagged rate of inflation. With the target rule being significantly different if a real exchange rate channel is operative in the Phillips curve, it is not surprising that the character of stabilization policy changes. The policymaker can no longer perfectly stabilize both the domestic rate of inflation and the output gap in the wake of IS or UIP disturbances by mechanically adjusting the policy instrument. Demand-side disturbances thus cause temporary effects on both variables. Optimal stabilization policy in such an open economy is no longer isomorphic to policy in a closed economy. Whether or not a real exchange rate channel exists in the Phillips curve also matters for the delegation issue. Discretionary price level targeting is consistent with optimal policy from a timeless perspective in a closed economy because the target rules that govern both strategies are compatible in the sense that they can be expressed in terms of the same target variables. This result does not carry over to the open economy framework of this paper because the timeless perspective introduces additional history dependence through inflation into the conduct of monetary policy. This additional history dependence cannot be matched by price level targeting. The paper also assesses the welfare gains under commitment relative to discretion. The results suggest that the welfare gains under policy from a timeless perspective are somewhat lower in an open economy framework where a real exchange rate channel exists in the Phillips curve. Taken altogether, the findings reported in this paper warrant the conclusion that optimal policy in an open economy framework is substantially different from optimal policy in a closed economy. The mere existence of a real exchange rate channel in the Phillips curve suffices for the conduct of optimal monetary policy from a timeless perspective to be more complex and information-intensive. The realistic assumption that in a small open economy individual firms take account of fluctuations in the terms of trade when adjusting prices provides a conduit through which the real exchange rate enters the aggregate Phillips curve. Its presence there ensures a prominent role for the real exchange rate in the design of optimal monetary policy.