اثر متلاشی شدن سیاست های پولی در مرکوسور

This paper shows that differences in the monetary policies in Argentina and Brazil have produced adverse effects on the structures of intra-regional trade and other target variables of economic integration in the MERCOSUR. The long-lasting coexistence of a strictly reserve-restrained monetary regime in Argentina with more flexible policies in Brazil has led to the development of a fundamental asymmetry in the adjustment mechanisms of both economies to common shocks. In several steps of modelling and econometric analysis the paper provides evidence for economic disintegration that has been caused by this lack of monetary integration.

The year 1991 marked the beginning of two policy experiments in South America that were to have far-reaching consequences for the economic development in the region. On 26 March 1991, Argentina, Brazil, Paraguay and Uruguay signed the Treaty of Asuncion by which the four countries created the institutional framework for MERCOSUR, the Common Market of the South. A few days later, on 1 April, Argentina's Congress approved a law to guarantee the full convertibility of the peso, the newly created currency, to the US dollar at the rate of one-to-one. Since the law confined the issue of pesos almost exclusively to the official dollar reserves, monetary policy in Argentina was subject to a (quasi-)currency-board rule. The formation of MERCOSUR was based on the idea that economic integration would foster economic growth in South America by intensifying trade and financial relations between Argentina and Brazil, the two dominant economies in the region.1 The Argentinian currency board was declared to be a precondition for macroeconomic stability and growth, because it was aimed at curbing the hyperinflation that had plagued the country throughout the 1980s. By the mid-1990s, both reforms were deemed to be successful. Argentinian inflation was under control and the growth rates of Argentinian GDP and intra-MERCOSUR trade rose faster than expected. At the same time Brazil had succeeded in getting rid of hyperinflation by pegging its new currency, the real, to the US dollar. Referring to Europe, where the 1991 Treaty of Maastricht had set monetary union on the agenda, politicians and economists began to speculate about combining economic integration in the MERCOSUR with monetary integration in the not too distant future. A few years later such speculations looked absurd. In January 2002 the Argentinian currency board collapsed in the middle of a deep and protracted crisis whose beginnings date back to the Asian crisis in 1997 and the Brazilian currency crisis in early 1999. With hindsight it seems almost trivial to say that the two experiments of the 1991 vintage—MERCOSUR and the Argentinian currency board—proved to be incompatible with each other. After the currency crisis Brazil had gone over to inflation targeting and floating exchange rates. Together with other factors the strong depreciation of the real against the dollar and the peso contributed to a catastrophical decline of industrial production in Argentina. This undermined the credibility of the currency board rule and created, at the same time, tensions within the MERCOSUR that threatened to halt the whole project of creating a common market in the region. In the end the “boarded peg” of the Argentinian peso to the dollar had to be abandoned. Yet it would be misleading to consider the period between 1999 and 2002 as an exceptional episode of inconsistent macroeconomic policies in an otherwise consistent story of economic integration. In our paper we show that fundamental differences in the monetary policies of Argentina and Brazil have existed even at the times when both countries had their currencies pegged to the dollar. We argue that the policy differences have led to asymmetries in the mechanisms by which the Argentinian and Brazilian economies adjust to common shocks. When Brazil was forced to abandon the dollar peg, these asymmetries generated adverse effects on real income convergence and trade integration in the MERCOSUR. The general idea behind this “joint hypothesis” can be described with reference to the recent discussion about the endogeneity of Optimum Currency Areas (OCA). In the tradition of Mundell (1961), there are two alternative OCA criteria for successful monetary integration through hard pegs or currency union. The first criterion is the degree of symmetry in the impact of shocks on the subregions of the currency area in question. Monetary integration is optimal, if shocks work symmetrically. In the case of asymmetric shocks, the second criterion would come into play: Monetary integration would be optimal only if adequate adjustment mechanisms (such as factor price flexibility, factor mobility and fiscal federalism) compensate for the loss of exchange rate flexibility. Frankel and Rose (1997) have pointed out that the symmetry of shocks may be an outcome rather than a precondition of monetary integration. The elimination of exchange-rate risks through hard pegs or currency union fosters trade integration, indicated by rising shares of intra-industry trade. Trade integration helps to synchronize the cyclical fluctuations of output and employment in the subregions. Shocks are thus made to work more symmetrically and the optimality of a currency area could be considered as endogenous. Conversely, we would argue that uncoordinated macroeconomic policies can produce adverse structural effects on trade by making business cycles less synchronous and shocks more asymmetric in their impact on the subregions. Our suggestion is that the lack of monetary integration between Argentina and Brazil has led to economic disintegration in the MERCOSUR. Following this introduction, the paper is organized in five further sections. In Section 2 we present results of cointegration analysis that provide evidence of fundamental differences in the monetary policies of Argentina and Brazil in the 1990s. In Section 3 we make use of a simple macroeconomic model to describe various scenarios in which the policy differences translate into asymmetric transmission of common shocks. In Section 4 we present structural VAR models that provide empirical support for these hypothetical conclusions, showing a lack of synchronization in the macroeconomic policies of Argentina and Brazil. In Section 5, we describe the behaviour of the Grubel–Lloyd index of intra-industry trade at high levels of sectoral disaggregation and present estimation results that suggest a strong influence of macroeconomic policies on trade structures in Argentina and Brazil. In Section 6 we sum up our arguments and draw some conclusions about the role that monetary policies play for economic integration in the MERCOSUR area.

Our analysis of the differences between the monetary policies in Brazil and Argentina and their consequences for macroeconomic developments suggests that monetary disintegration in the MERCOSUR has created obstacles to economic integration, and in particular to trade integration in the region. It may seem trivial to state that, with a view to economic integration, the exchange rate regimes of Argentina and Brazil were absolutely incompatible in the years between 1999 and 2002, when Brazil had switched to floating, whereas Argentina preferred to stick to its currency board arrangement. However, we have shown that the incompatibility problems were present even in the period from 1994 to 1999, when both countries had their exchange rates pegged to the US dollar. To put it simply, sterilization of foreign reserve flows to and from Brazil made the difference. As Argentina did not sterilize, various asymmetries developed in respect to the effectiveness of domestic macroeconomic policies and the impacts of external shocks on the two economies. The lack of comovement in macroeconomic variables that had existed prior to the foundation of MERCOSUR, as both economies were much less open and went through different cycles of stagflation in the 1980s, was thus perpetuated by a lack of policy coordination in the first decade of MERCOSUR. However, in the all-fixed exchange rates constellation of the mid-1990s (Scenario 1 in Section 3), trade integration was at least favoured by monetary expansion in Brazil or by fiscal expansion in Argentina. The difference in the monetary policy regimes may, moreover, have helped to buffer external shocks, as in the case of the Mexican crisis of 1994–1995. Yet, both types of expansionary policies did also undermine the constellation of (indirectly) fixed exchange rates, when the increasing force of external shocks in the wake of the Asian and Russian crises in 1997–1998 made the region's currencies look more and more overvalued, thus provoking demands for active stabilization policy. The underlying asymmetries of policy effectiveness and shock impacts were exacerbated, when Brazil switched to floating in 1999. Now even external shocks had clearly adverse effects on trade integration and the synchronization of economic development in the region. They worked their way through the two economies in a fashion that made Argentina being hit twice—first by the original shock and thereafter by its repercussions through the depreciation of the Brazilian currency. The filter analysis and the structural VAR model of Section 4 suggest that much of the observable divergence in the pattern of output fluctuation is explained by the above described asymmetries. The lack of coordination between the monetary regimes in the MERCOSUR has led to setbacks in the process of trade integration. The Grubel–Lloyd index of intraindustry trade between Argentina and Brazil shows (at three and four-digit levels of disaggregation) that the shares of intraindustry trade increased up to the year 1997 and declined thereafter. Our corresponding analysis of pseudo panel data suggests that the index is very sensitive to the bilateral exchange rate, interest rate differentials and output developments. This conforms with general observations that, between 1999 and early 2003, the depreciation of the Brazilian real and the severe contraction of aggregate demand in Argentina produced strong negative effects on the productive capacity of the Argentinean economy—an experience that hardly made a good base for economic integration. On the contrary, monetary disintegration spurred conflicts over remaining trade barriers in the region that might otherwise have been eliminated long ago. Argentina accused Brazil of pursuing a “beggar thy neighbour” policy, but eventually it had to accept that its own policy of “beggaring thy lender (of last resort)”, by sticking to the currency board, was unsustainable. In January 2002, the convertibility law was suspended and the peso quickly lost much of its value against the dollar. Since then the obstacles to macroeconomic policy coordination and monetary integration in the MERCOSUR may have become smaller. Optimistic visions of a monetary union or other forms of close cooperation have returned to the scene, both among policymakers and among economists (see, for instance, Giambiagi, 2003). Yet, until completion of this manuscript (in early 2004) there were no concrete plans to fix exchange rates between the two countries or to run coordinated sterilization policies. With floating exchange rates and enormous external debts in both countries, there remains much uncertainty, in particular for investors in transnational production and intraregional trade. The “monetary dilemmas” of Argentina in the MERCOSUR (Fanelli and Heymann, 2003) are far from solved. It seems that the monetary disintegration of the past has had so many detrimental effects that the prospects of economic and monetary integration in the MERCOSUR are bleaker than they used to be in its early years. This is not to say that monetary and trade integration in the MERCOSUR is impossible. It should be clear, however, that substantial progress in that direction requires more compatible monetary policies than those run in the recent past