اعتبار و تعهد از سیاست های پولی در اقتصاد باز

We study the delegation of monetary policy to independent central bankers in a two-country world with monetary spillovers. It is shown that, under the hypotheses of imperfect commitment and private information, the equilibrium degree of commitment depends on the correlation structure of the shocks hitting the economies. When the correlation is negative (as when the variance of output depends mainly on shocks to the terms of trade) there is strategic complementarity in the degree of commitment in the two countries. When the correlation is positive (common technological or demand shocks) there is strategic substitutability. In this latter case, the degree of commitment is shown to be increasing in the correlation among shocks. Common components in the international business cycle have been shown in several studies to be relatively more relevant in developed countries. Therefore, our results may contribute to explaining why the institutional solution to the inflationary bias has been adopted in the most advanced countries.

Since the end of the Bretton Woods agreement, the analysis of the role of the strategic interaction between national monetary authorities has been important in the academic and policy debate. The creation of the EMU and the establishment of the European Central Bank have further stimulated this line of research. Strategic interaction among monetary authorities in open economies is not, of course, the only major force shaping monetary policies in open economies. Since the seminal contributions of Kydland and Prescott (1977) and Barro and Gordon (1983), monetary policy has also been recognized to be affected by an inflationary bias. Economic, institutional and political aspects are at the core of the theoretical work in this field (see, for surveys, Eijffinger and de Haan, 1996 and Persson and Tabellini, 2000). Rogoff (1985a) explicitly proposed an institutional solution to the credibility problem based on the strategic precommitment of monetary policy to an independent central bank (cb henceforth) with suitable preferences. Political scientists also argue that “credible delegation of monetary policy making to an independent and conservative central bank can serve as a commitment device circumventing […] the inflationary bias” (Franzese, 1999). Are strategic interactions among monetary policy makers and openness relevant for the inflationary bias and for the design of monetary institutions? Intuition, theory and data suggest that they are. For example, Rogoff (1985b) and Romer (1993) argue that, in open economies, policy spillovers affect the inflationary bias and may affect institutional design and incentives to precommit monetary policy in non-trivial ways. Rogoff (1985b) shows that international cooperation may be counterproductive if the credibility loss associated to the cooperative regime among monetary policy makers countervails the gains from the internalization of the negative externalities associated with the terms of trade effect of surprise inflation. Romer (1993), in his study of the determinants of long-run inflation in open economies, analyzes the relationship between the inflationary bias and the degree of openness and argues that the institutional solution to the inflationary bias problem only seems to have worked for rich countries. If openness is relevant, how does it affect the institutional design of monetary authorities by sovereign governments? Giavazzi and Pagano (1988), by extending the model by Rogoff (1985a), show that governments of small open economies may have incentives to precommit monetary policy through a fixed exchange rate regime in order to gain credibility from abroad. Persson and Tabellini, 1995 and Persson and Tabellini, 1996 argue that cooperative stabilization policies may be implemented by suitable contracts between the government and central bankers. Extensions of the Obstfeld and Rogoff (1995) model by Corsetti et al. (2000), Corsetti and Pesenti, 2001a and Corsetti and Pesenti, 2001b and Benigno (2002) have revisited the normative analysis of welfare effects of strategic interdependence in monetary policy making and the foundations of policy coordination. In this framework, for example, Betts and Devereux (2000) reevaluate the issue of international monetary policy coordination and, implicitly, the interdependence of institutional design. Their model studies how alternative assumptions on pricing strategies by exporting firms may change the nature and the incentives for policy coordination and shows that the microfoundation of welfare analysis of international monetary policy regimes may not be inconsistent with negative spillovers across countries. The aim of this paper is to contribute to the positive analysis of the incentives of national governments to precommit monetary policies to independent central banks in open economies. We provide a simple extension of the basic credibility model showing that, if a commitment technology exists (see Rogoff, 1985a), although imperfect, its value to the government depends on several fundamentals of the economy, such as the degree of openness and the correlation of shocks across countries. In our model strategic interdependence at the monetary policy stage may entail strategic interdependence at the institutional design level. Openness and correlation of shocks may reinforce or countervail the traditional argument holding for the case of commitment in closed economies and they may contribute to explaining why some countries strategically commit monetary policy whereas others may rationally neglect such a solution. The study of how openness affects credibility and, therefore, the strategic design of monetary institutions is also motivated by the empirical analysis of the relationship between openness, commitment and inflation in Romer (1993), Campillo and Miron (1997) and Lane (1997). The idea that openness affects the inflationary bias and, therefore, inflation outcomes over the long run is analyzed in Romer (1993), who, using the extension of the Barro and Gordon (1983) model in Rogoff (1985b), considers an open economy in the presence of dynamic inconsistency of optimal monetary policy and shows that increased openness reduces equilibrium inflation.1 Although this prediction turns out to be robust across different subsamples of the data set, he finds no evidence that it holds for the most developed countries. Lane (1997) shows that openness has a strong effect on the subset of industrialized countries too, once the country size effect is taken into account. However, empirical evidence on the role of institutional design and its relation with trade openness in explaining inflation over the long run is less clear cut.2 In the regressions provided in Lane (1997), the index of central bank independence is highly significant in explaining inflation performance for the rich countries sample (see Table 3, column 4, p. 342) and for the OECD sample (see Table 4, column 4, p. 342) whereas it is not over the full sample (see Table 2, column 3, p. 341). These results provide support for Romer's interpretation of the role of commitment in inflation outcomes over the long run in more developed countries. Campillo and Miron (1997), on the other hand, provide evidence in support of the idea that economic fundamentals, such as openness and political stability, seem to be quite important for keeping inflation under control whereas “quick fixes […] do not make a big difference unless the underlying conditions for low inflation are present”, not denying, however, that institutions may play a role in monetary policy making. Overall, these studies seem to suggest that openness is a crucial determinant in explaining inflation over the long run whereas institutional solutions, along with openness, seem to matter only in rich countries. Why, therefore, have some countries effectively addressed the problem of inflationary bias through the institutional solutions whereas others have not? Is credibility a luxury good, so that only rich enough countries may afford or have incentives to buy it through suitable commitment? One way to explain cross country variation in the degree of independence of monetary institutions is to focus on the role of fundamentals such as the availability of financial instruments to hedge against inflation, the degree of efficiency of fiscal systems or the distribution of wealth regarded as determinants of societies' aversion to inflation (Posen, 1995). The distribution of benefits and costs of inflation within advanced western societies may indeed have evolved towards a larger weight given to inflation aversion and, therefore, can explain institutional reforms. This would drastically reduce the explanatory power of the strategic commitment hypothesis as a basis for understanding the institutional solution as the outcome of rational strategic delegation. Relying on the empirical results in Romer (1993) and Lane (1997), we believe that the strategic commitment hypothesis deserves further exploration. In this paper, we focus on the analysis of the forces that shape the strategic design of monetary policy institutions in open interdependent economies, taking as given the degree of societies' inflation aversion. To show the precise conditions under which these incentives arise, we develop our analysis in the framework of the precommitment approach to the solution of the inflation bias (see Rogoff, 1985a). The main features of our model are: (1) The existence of monetary spillovers across economies and strategic interdependence between monetary authorities willing to stabilize unpredictable shocks to real variables. (2) The existence of a commitment technology that can be adopted by governments as a way out to the credibility problem, albeit the commitment technology is assumed to be imperfect in a sense to be made precise in the following sections. (3) A degree of opaqueness in the broad political–economic environment surrounding the cbs, leading to private information by national policy makers about their assessment of the trade-offs involved in their policy actions. One of the results of our analysis is that incentives to precommitment are affected by the degree of synchronization and correlation among shocks hitting economies. The basic intuition for why monetary policy interdependence and openness may matter for the institutional design is easily described. Consider the case of positive correlation among shocks to the level of economic activity across countries. If monetary policies by central bankers are strategic complements, they will end up with closely matched policy stances. A less committed central banker in the foreign country (bearing the credibility costs of its government's choice) provides a larger expected level of stabilization at home (directly through the terms of trade effect and, indirectly, through strategic complementarity in the monetary policy game), in the states of the world when it is most needed and, for the same reasons, a larger expected variance of inflation imported from abroad. In this case, each government has incentives to increase its level of commitment to gain credibility, reducing the expected variance of inflation and free riding on the stabilization provided abroad. Openness, in this case, is shown to favor commitment in countries mainly affected by common shocks. The opposite incentives to commitment are provided when the correlation among shocks across countries is negative. Provided that most developed countries share a relatively larger component of common shocks compared to less developed countries, the predictions of our model may contribute to interpreting the empirical evidence reported above showing that more developed countries have relatively larger incentives to commit monetary policies for strategic reasons.3 The analysis of strategic issues within the institutional approach to delegation of monetary policy in the context of open economies is not completely new. Different contributors to this literature have recognized that, in the presence of strategic interdependence between monetary authorities, the design of the institution responsible for stabilization policy may have different characteristics than in the case of a closed economy (see Rogoff, 1985a, Persson and Tabellini, 1993 and Walsh, 1995). Persson and Tabellini (1995) extend to interdependent open economies the analysis of the optimal design of monetary institutions relying on the contractual approach. They analyze a two-country game showing that a cooperative outcome in the stabilization game can be implemented by a decentralized contracting scheme such that the central bankers' compensation in each country is made contingent on the outcome of the stabilization policies in both countries. The feasibility and enforceability of such “contracts” is largely discussed by the authors although not definitely established (see Persson and Tabellini, 1995, pp. 44–45). Dolado et al. (1994) analyzed similar issues by extending Rogoff, 1985a and Rogoff, 1985b analysis to interdependent economies in a framework very similar to that adopted in this paper. Our model, however, differs from their analysis in many respects: they model the decision of the governments at the institutional design stage as being taken after the observation of a (common) shock hitting the economy, whereas, following Rogoff (1985a), we model the government's decision so that the institutional arrangement has to be set ex ante, i.e. before the shocks are observed. They also assume perfect commitment and myopic private agents and, as a consequence, the credibility issue is disregarded in their analysis. Differently from Dolado et al. (1994), we set out our analysis using a model for open economies where the wage-price setting private agents have rational expectations, thus leading to a time consistency problem in policy making. The remainder of this paper is organized as follows: Section 2 briefly describes the economic model, the payoff functions of the agents and the timing of the game. In Section 3 we solve the commitment game under the hypothesis of common knowledge of the preference functions of the players. In Section 4 we solve the commitment game by introducing private information of the policy makers on their preferences. Section 5 concludes.

The last decades have been characterized by a strong interest in the institutional reforms of agencies in charge of monetary policy. There is a broad consensus in the literature that these reforms can be interpreted as the institutional solution to the credibility problem in monetary policy making. Economists' views about how effective institutional arrangements may be in controlling inflation have differed. After Rogoff, 1985a and Rogoff, 1985b influential contribution was apparently validated by empirical studies supporting the view that commitment was indeed important in explaining how societies keep inflation under control, recent studies (Campillo and Miron, 1997) conclude that politico-economic fundamentals are more important than institutions in the explanation of inflation. In Romer (1993) and Lane (1997), on the other hand, proxies of central bank independence are significantly negatively correlated with inflation over the long run in more developed countries. In this paper we have set out a model to analyze the main forces that shape institutional strategic design in open economies where the commitment technology may not be perfect and the political–economic environment is characterized by a degree of opaqueness leading to private information of national policy makers. The results of the model suggest that strategic interdependence at the institutional design level may be one of the reasons why more developed countries had incentives to commit monetary policy even in the presence of increasing openness. Our findings can be summarized as follows: in open economies the viability of the institutional solution to the inflationary bias problem depends on the nature of the correlation structure of economic shocks across countries. When the preferences in the two countries are similar enough (symmetric equilibrium) and the correlation between shocks to output is positive, a conservative cb result is obtained and the optimal degree of commitment turns out to be increasing in the degree of correlation among shocks to output. This result is driven by the fact that, for positive correlation, stabilization policy provided in each single country is a public good and, therefore, at the institutional design level, each country has an incentive to underprovide stabilization at the world level by saving on credibility costs. The model, on the other hand, predicts that when the shocks to the level of output are negatively correlated (as in the case when the bulk of the relative variance in the output process is due to shocks in the terms of trade) the incentive for strategic precommitment may be reduced because of interdependence. Therefore, it is possible that, in open economies, some countries may not find convenient to undertake monetary reforms to keep inflation under control. Empirical studies on the international business cycle show that, in developed countries, a larger than average share in the variability of output is accounted for by common components, whereas the idiosyncratic components account for more of the volatility in developing economies (see Kose et al., 2003). These results seem to support our model's prediction that incentives for strategic commitments are greater in developed economies. Our analysis shows that the conclusions of the study by Campillo and Miron (1997) on the irrelevance of institutional arrangements in controlling inflation are perfectly compatible with a simple extension of the Rogoff's (1985a) model of open economies. Our findings may contribute to explaining why only more developed countries have adopted institutional solutions as a strategic way out of the inflationary bias problem. Therefore, we should not be led to dismiss the institutional solution due to strategic commitment as ineffective.