سیاست پولی در اقتصاد آزاد هریس دلاس

The recent literature on monetary policy in open economies has produced a strong presumption in favor of activistic policy and flexible exchange rates. We argue that this result may owe much to the combination of two commonly made assumptions: That nominal goods prices are rigid. And that the monetary authorities have a lot of information about the economy. When the source of nominal rigidity is found in wages and monetary policy is conducted according to less information demanding rules (such as a standard interest rate rule) policies that stabilize the money supply or the nominal exchange rate may perform better.

The properties of alternative international monetary arrangements have been studied extensively in the literature, first within the Mundell–Fleming and lately within the New Keynesian (NK) model. The former model has identified two key factors that make the fixing of the exchange rate costly: (a) Dissimilarities in economic structure, in particular regarding the degree of asymmetry in the shocks, and (b) a high degree of nominal rigidities. Asymmetric shocks generate a need for terms of trade adjustments. If the necessary adjustments cannot occur directly through wage and price changes, they may be accomplished indirectly via exchange rate changes (see Friedman's (1953) case for flexible exchange rates). The NK model uses diverse “formats” and has produced rather diverse findings. Nevertheless, its main conclusion echoes that of the traditional Mundell–Fleming model (at least under producer currency pricing) and for the same reasons. Namely, abstracting from non-fundamental fluctuations and speculative attacks, flexible exchange rate systems tend to fare better than regimes that severely restrict exchange rate fluctuations (Benigno and Benigno, 2003; Kollmann, 2002, Obstfeld and Rogoff, 2001, Pappa, 2004 and Stockman and Ohanian, 1993). Moreover, independent national monetary policy performs quite well, that is, there exist small gains from international policy coordination. While objections to the general validity of these results have been raised1 they have not undermined their wide acceptance. There are good reasons to believe that the alleged superiority of monetary policies that feature activism, absence of international coordination and a flexible exchange rate may not be as general as it appears. First, it is typically assumed that the monetary authorities have complete information about the structure of the economy and the shocks. Combining this with the assumption that monetary policy is conducted optimally, that is, that it aims at maximizing the utility of the representative agent, often allows these models to generate activist policy equilibria that replicate the efficient, flexible price (or wage) equilibrium. Consequently, when monetary policy is omniscient and omnipotent, it is not sensible to constrain it by making it target the exchange rate. This is especially true when beggar-thy-neighbor effects associated with independent policies are not strong (for instance, when domestic and foreign goods are poor substitutes, see Pappa (2004)). Second, the ability to manipulate the nominal exchange rate is more useful when there is no production interdependence across countries. If there is also trade in intermediate (capital) goods then an exchange rate depreciation may have adverse, direct effects on the cost of domestic production which go against those on relative demands and which make the exchange rate instrument less useful.2 Again it is typical in this literature to assume that trade involves only consumption goods.3 And third, most of the existing literature assumes nominal price rigidities. When prices are rigid, fixing the exchange rate incapacitates a mechanism that could bring about desired relative price changes (see Friedman's (1953) case for flexible rates). While this adjustment mechanism could remain in place even under nominal wage rigidities, its role in existing models with nominal wage rigidity tends to be limited due to the combination of the assumptions of imperfect competition and a single input (labor) which transfers nominal wage to nominal price rigidity (Obstfeld and Rogoff, 2000). The objective of this paper is to examine the role played by these factors for the properties of alternative exchange rate systems as well as the desirability of standard interest rate policies (such as the Henderson, McKibbin, Taylor, HMT, rule). This is important for a number of reasons. First, the assumption of perfect information may be a useful benchmark but is clearly unrealistic. Hence a policy recommendation that requires that monetary authorities possessed information that was not available might be of limited value. In recognition of this consideration most of the literature on monetary policy nowadays studies rules that are cast in terms of observable macroeconomic quantities rather than in terms of exogenous shocks (see Woodford, 2003). Second, there exists considerable uncertainty regarding the relative strength of alternative sources of nominal frictions (price vs. wage) in the real world (see Christiano et al., 2005 and Smets and Wouters, 2003). A policy prescription that would be valid under only one type of rigidity but not under another might be too risky to adopt. And third, international trade involves both consumption and intermediate goods. The model employed is highly stylized but contains all the points of contention raised above: nominal wage vs. price rigidities, international production and consumption interdependence and a variety of unobservable shocks. Issues of strategic interaction between the policymakers in the two countries are left out as this is a distinct issue with no direct relationship to the informational issues investigated here. The main results are the following: First, a standard interest rate policy rules and a flexible exchange rate regime fare well under nominal price rigidity as long as the degree of imperfect information is not too large. Second, production interdependence does not undermine the case for a flexible exchange rate regime. And third, nominal wage rigidity and imperfect information favor policies that target the money supply and/or the nominal exchange rate. These findings also have implications for international policy coordination. Obstfeld and Rogoff (2001) argue4 that in situations where global monetary policy can replicate the flexible wage equilibrium, “...lack of coordination in rule setting is a second-order problem compared to the gains from macroeconomic stabilization.” We find that under price rigidity independent national, uncoordinated, monetary policies can deliver superior results even when monetary policy does not have the ability to replicate the flexible price equilibrium. The rest of the paper is organized as follows: Section 1 presents the model. Section 2 reports the main findings.

The recent literature on the welfare properties of alternative exchange rate regimes typically comes out in favor of activistic monetary policy and flexible exchange rates. This paper examines the extent to which this result may hinge on three commonly made assumptions. First, that the monetary authorities know a lot about the state of the economy (that is, that they can perfectly observe and react to current shocks). Second, that prices (rather than wages) are the source of nominal rigidity. And third, that there is no direct international trade production interdependence. The first assumption induces a bias in favor of activistic policies. As the flexible exchange rate system is more suitable to the pursuit of activistic policies, it also induces a bias in favor of it. The second assumption also introduces a bias in favor of flexible exchange rates as it implies that the nominal exchange rate is the only means for bringing about needed terms of trade changes. It turns out that indeed the type of nominal rigidity and imperfect information play a key role in the ranking of alternative monetary policy arrangements. Under conditions of imperfect information and with a standard interest rate policy rule, price rigidity may still support a flexible exchange rate system. Under the same conditions, though, wage rigidity favors simpler targeting procedures, such as the targeting of the money supply or the exchange rate. These findings may have also something to say about the conditions under which international policy coordination could prove beneficial. Unsurprisingly, international policy coordination does not help when policy activism performs well. More interestingly, a simple form of international policy coordination (a global monetary targeting rule) seems desirable in situations in which national activist policies do not fare well.