Members of the European Union (EU)have long attempted to unify their monetary policies, whether by fixing exchange rates or moving toward a single currency.The success of these attempts has varied over time,and among EU member states.This article argues that the degree to which a country is integrated into EU trade and finance has a major impact on its willingness to make the sacrifices necessary to pursue monetary integration,especially stability against Europe's anchor currency,the Deutsche Mark(DM).Higher levels of intra-European trade and investment increase the desirability of stabilizing exchange rates between European countries. Statistical evidence indicates that greater integration of goods and capital markets is associated with greater success in fixing national exchange rates against the DM.This implies that pressures for monetary integration will continue but will vary among countries,along with the degree to which they are economically linked to European trade and investment. THE IMPACT OF GOODS AND CAPITAL MARKET INTEGRATION ON EUROPEAN MONETARY POLITICS JEFFRY A.FRIEDEN Harvard University Union Hv empeincneary ce ered history.They failed quite miserably from their inception in the early 1960s through the early 1980s.During the 1980s,however,the European AUTHOR'S NOTE:The author acknowledges support for this research from the Institute on Global Conflict and Cooperation.He also acknowledges research assistance from Roland Stephen and Michael Harrington and comments from Barry Eichengreen,John Goodman,Miles Kahler Peter Lange,Lisa Martin,Kathleen McNamara,George Tsebelis,and Guy Whitten.An earlier,and substantially different,version of this article was prepared for an SSRC project on Liberalization and Foreign Policy. 1.Throughout this article,I refer to the organization that has variously been known as the European Economic Community,the European Communities,and the European Union with this last (currently preferred)name.This may be somewhat misleading at times,especially in reference to historical developments,but it has the attraction of consistency. COMPARATIVE POLITICAL STUDIES,Vol.29 No.2,April 1996 193-222 1996 Sage Publications,Ine. 193
194 COMPARATIVE POLITICAL STUDIES April 1996 Monetary System(EMS)and its exchange rate mechanism(ERM)began to resemble a binding fixed-rate regime.Today,the future of the process is very much in doubt.On the one hand,several members of the EMS appear committed to continued monetary integration,leading eventually to a single European currency.On the other hand,especially in the wake of the 1992- 1993 currency crisis,there is little confidence in the ability of other EMS and EU members even to maintain stable exchange rates,let alone to move toward a single currency.A wide variety of alternative futures for European monetary integration have been discussed,from abandonment through a two-stage union to a single currency. Monetary integration,as used here,refers to a range of policies to achieve convergence among national monetary conditions.It includes measures to stabilize exchange rates,through formal fixed exchange-rate regimes (such as the Bretton Woods system or the EMS)to a single currency.At whatever level,monetary integration is difficult because stabilizing exchangerates,and at the limit adopting a single currency,can require national governments to implement politically unpopular economic policies in the interests of macro- economic convergence.It is not surprising that the willingness and ability of national governments to take the measures necessary to reduce exchange-rate fluctuations vary a great deal and that the policies of particular countries have varied over time. To understand the prospects for European monetary integration,it is important to understand its past trajectory.One way of doing so,particularly important for looking toward the future,is to examine the degree to which different European countries have been able and willing to take the steps necessary to hold their exchange rates constant against one another.This is of special importance because all plans for further monetary integration in Europe involve some degree of exchange rate stabilization,ranging from target zones through a single currency.Although the possible outcomes differ widely,they can all be seen as points on a continuum that measures the degree of permanence and rigidity of the fixed exchange-rate agreement. In this light,this article argues that a major determinant of national propensities to fix exchange rates is the degree to which the country in question is integrated into EU trade and finance.?The more important are a country's trade and investment ties with the EU,especially with the DM bloc around Germany,the more costly are fluctuations of its currency against the DM,and the more likely the government is to stabilize the exchange rate. 2.There are,of course,other potential explanations for the course of monetary integration. For a general survey,see Eichengreen and Frieden (1994a)
Frieden/IMPACT OF GOODS 195 Higher levels of capital mobility and intra-EU trade increase economic and political pressures for monetary integration.Closer links among EU financial markets heighten the trade-off between national monetary policy independence and exchange rate stability,forcing countries to choose be- tween them.More cross-border trade and investment within the EU expose more economic agents to currency risk and increase the demand to stabilize exchange rates.This effect is evident both over time and across country.As the EU has become more integrated,the demand for stable exchange rates has grown;and national support for monetary integration is correlated with national reliance on intra-EU trade and payments.The argument presented here grows out of,and is broadly consonant with,the economics literature on optimal currency areas and related macroeconomic policies. In section 1,I describe the history of attempts at European monetary integration.In section 2,I present various factors important in explaining European monetary politics,then develop my argument that the level of trade and capital market integration is a crucial determinant of currency policy.In section 3,I show how changes in the levels of intra-EU flows of goods,finance,and direct investment over time and across countries are correlated with Union-wide and national measures to reduce exchange rate variability.I conclude with some observations about the implications of this analysis for the future of monetary plans in the EU. 1.THE DIFFICULT COURSE OF EUROPEAN MONETARY INTEGRATION:FROM THE WERNER REPORT THROUGH THE EMS Fixing exchange rates between countries with different social,political,and economic conditions can be difficult.Although the determinants of exchange rates are still hotly debated among economists,there is little doubt that for the value of the currency of one country to be stable against that of another, the two countries'macroeconomic conditions cannot be too divergent.This is true,with varying degrees of stringency,whether the monetary integration in question involves simply stabilizing national currencies,or a more formal fixed exchange-rate system,or a single currency.In the case of the EU,in which the monetary anchor has long been Germany and the DM,monetary integration requires that other nations bring their macroeconomic conditions into line with those of Germany.This has typically meant reducing inflation, and-again,without entering the debates about this issuethe policies necessary to do so often have costs,typically involving recessionary mea- sures that increase unemployment and reduce real wages and consumption
196 COMPARATIVE POLITICAL STUDIES /April 1996 Although virtually everyone in the EU,then,gives lip service to the desirability of stable currency values,only in some countries and only at some times has rhetoric been backed up by the national sacrifices necessary to implement a stable exchange rate against the DM.The course of European monetary politics since the late 1960s illustrates this fact. Although stable exchange rates were a goal of the EU from its beginning, serious discussion of the matter began only once fissures appeared in the Bretton Woods system.Early talks led to the 1969 Werner Report,which recommended beginning a process of monetary union among EU members (Tsoukalis,1977,pp.51-111;van Ypersele,1985,pp.31-45).These recom- mendations were superseded by the end of the Bretton Woods regime.Over the course of the year that followed the August 1971 American decision to go off gold,EU member states established the snake,an arrangement to hold their currencies within a 2.25%band against each other.In addition to the original six,Great Britain,Ireland,and Denmark joined the snake on May 1, 1972,to prepare for their entry into the European Community 8 months later. Yet the goal of stabilizing EU currencies proved impossible to achieve.3 Britain and Ireland left the snake in June 1972,just weeks after joining;the Danes left shortly thereafter but rejoined in October.In February 1973,Italy withdrew from the arrangement.In addition,throughout 1973,only a series of parity changes allowed the system to hold together,and even then France chose to exit in January 1974.The French returned in July 1975,only to leave for good 8 months later. Within 3 years of its founding,then,the only EU members still in the snake were Germany,the Benelux countries,and Denmark.Even within this narrowed arrangement,realignments were frequent,typically to devalue the Danish krone and/or revalue the DM.Norway affiliated with the snake from May 1972 until December 1978,and Sweden from March 1973 to August 1977,even though the two were not EU members.Austria and Switzerland, also not EU members,had their currencies shadow the DM informally but did not join the snake. In the late 1970s,discussions of EU monetary integration began to gather momentum again,and in March 1979,the EMS and its exchange rate mech- anism went into effect.All EU members except the United Kingdom acceded to the ERM,which allowed a 2.25%band among currencies(6%for the lira). 3.Tsoukalis(197),p.112168:Ludlow(1982),pp.1-36:Coffey(1987),pp.616.A useful chronology of the snake is in Coffey,pp.123-125. 4.Ludlow (1982)is especially detailed on the negotiations and early operation of the EMS; see also van Ypersele (1985),pp.71-95;and Ungerer (1983).Excellent surveys of the EMS experience more generally are Giavazzi and Giovannini(1989);Fratianni and von Hagen(1991); Goodman(1992);and the articles in Eichengreen and Frieden (1994b)
Frieden IMPACT OF GOODS 197 The experience of the 1970s held out few hopes for success in the 1980s. The Union's two major high-inflation countries,France and Italy,had been unable to fix their exchange rates with other EU members.Resistance to austerity measures and pressures to maintain international competitiveness led to continual rounds of franc and lira depreciations against the DM. Therefore,for the first several years of its existence,almost no informed observer believed that the ERM would hold.Indeed,in the first 4 years of its operation,there were seven realignments of EMS currency values. However,after 1983,exchange-rate variability within the EMS declined substantially,whereas monetary policies converged on virtually every dimen- sion.Between April 1983 and January 1987,there were only four realign- ments,generally smaller than previous changes,and from January 1987 until September 1992,there were no realignments within the ERM.3 Meanwhile, Spain,the United Kingdom,and Portugal joined the mechanism and Finland, Sweden,and Norway linked their currencies to the European Currency Unit (ECU).In 1981,the Austrian government announced a unilateral peg to the DM,to which it held firmly. In the wake of the EMS's apparent success,European integration more generally picked up speed.In 1985 and 1986,EU members discussed and adopted the Single European Act,which called for the full mobility of goods, capital,and people within the Union by January 1,1993(Moravcsik,1991; Sandholtz Zysman,1989).Most capital controls were gone by 1991,and barriers to the movements of goods were reduced continually in the run-up to 1993.The 1991 Maastricht Treaty included plans for full monetary union. But as preparations for currency union gathered force,the 1992-1993 cur- rency crisis called even the less ambitious EMS into question(Eichengreen Wyplosz,1993).The German government's massive fiscal effort to finance unification led the Bundesbank to raise interest rates to counter potential inflation.EMS member governments faced the choice of either following tight German monetary policies,in an environment of already high unem- ployment,or leaving the ERM.Uncertainty about the future course of European integration was exacerbated by the June 1992 failure of the first Danish referendum on the Maastricht Treaty and by the closeness of the September 1992 French referendum.Faced with runs on their currencies, in September 1992,the British and Italian governments took sterling and the lira,respectively,out of the ERM and allowed them to float.In subsequent months,the currencies of Spain,Portugal,and Ireland were devalued,al- though they remained in the mechanism.In Summer 1993,with the system 5.Relative parity changes are reported in International Monetary Fund (1988),p.19; information on exchange rate variability is provided on pp.20-34
198 COMPARATIVE POLITICAL STUDIES April 1996 again under attack on currency markets,remaining ERM members agreed to widen fluctuation bands to 15%(the Dutch guilder and DM remained in a 2.25%band).Although the reduced ERM has been stable,questions persist about the future of monetary integration. This brief survey indicates how varied the fortunes of attempts to reduce exchange-rate variability within Europe have been.For more than 20 years, stable exchange rates have been a stated goal of the EU and its member governments,but success has varied greatly over time.And there has also been wide variation in the degree to which particular European countries have been able to sustain fixed exchange rates against the DM.I now turn to attempts to explain this variation,over time and across countries. 2.EXPLAINING THE COURSE OF EUROPEAN MONETARY INTEGRATION There is only a small body of thought on the political economy of national exchange-rate policy,of which European monetary integration is a subset.s However,two literatures are germane.One is specifically targeted at explain- ing aspects of European integration,including monetary integration.The other has to do with the national welfare effects of different exchange-rate regimes,including currency union.Both are useful for our purposes.I first summarize them,then go on to present my argument about the impact of goods and capital market integration on European exchange-rate politics The first body of knowledge relevant to explaining European monetary integration,not surprisingly,is that designed explicitly for this purpose.Here several factors that help explain the process have been widely commented upon.All played a role,and I do not argue against their importance.I do believe that they must be supplemented with an understanding of the differ- ential impact of goods and capital market integration,to which I turn after summarizing extant views. One factor that gave impetus to monetary integration was the Common Agricultural Policy (CAP).The EU's agricultural subsidies involve setting Union-wide food prices,and when a currency is devalued,the EU reference price would normally be raised in the devaluing country to counterbalance the devaluation.This "passing through"of the exchange-rate change to food prices would mitigate the devaluation's attempt to restore price competitive- 6.Among the exceptions,disparate in their coverage and concerns,are Destler and Henning (1989),Eichengreen (1992a),Henning(1994),and Gowa (1988).Edison and Melvin (1990)is a good survey
Frieden IMPACT OF GOODS 199 ness in the nonagricultural sectors.For this reason,the Union devised a series of compensatory arrangements and accounting exchange rates.For our purposes,what is important is that exchange-rate fluctuations complicate Union agricultural policy by changing compensatory farm payments in ways that disrupt EU farm policy.This was indeed one of the original reasons for monetary integration and remained important through the adoption of the EMS(McNamara,1993).However,the CAP is a constant,and although it explains the persistence of EU attempts at monetary integration,it cannot explain variation in their success.There is,in fact,no evidence that national support for monetary integration is related to reliance on the CAP. Another reason for variation in the success of monetary integration is institutional variation,both over time and across countries.It is certainly plausible that the EMS was more successful than the snake because of an increase in the amount of money available to EMS members for short-and long-term financing of payments deficits,as well as 5 billion ECU in concessionary development loans made available,essentially to Ireland and Italy,as a side payment to the two countries with the largest adjustment burden (van Ypersele,1985,pp.61-64;the subsidy component of the con- cessional loans was a billion ECU).All told,the resources committed to the EMS were about three times those committed to the snake,which made affiliation with the system that much more attractive to potential members. However,there is little evidence that the funds involved were particularly important to the process-Italy did not even use its concessionary finance, and the other funds were rarely central to EMS developments.And again, with the exception of the concessionary funds to Ireland,this factor cannot explain variation among EU members.It might be argued that the more institutionalized nature of the EMS helped cement its effects,but again this does not help much to explain differences in behavior among members of the system.Indeed,one of the more successful exchange rate pegs in Europe has been that of the Austrian schilling to the DM,engineered at a time when Austria was not even contemplating EU membership. A third,less tangible,but nonetheless crucial,explanation for monetary integration-and especially for the greater success of the EMS than of the snake-was the relationship between them and broader EU participation. There was never a sense that the snake was an essential component of the EU;neither national politicians nor Union leaders staked much political capital on the arrangement.The EMS was different.The French and German heads of state launched the attempt with great publicity,and the Commission regarded the EMS as of paramount importance
200 COMPARATIVE POLITICAL STUDIES/April 1996 This linkage of the EMS with broader EU participation was important.In the early 1980s,with the European economies beset by stagnation and unemployment,many segments of society began to look upon an intensifi- cation of European economic integration as the last best hope for the region (Katseli,1989).The threat of relegation to second-tier status within Europe led many to reconsider their position on monetary integration.Previously,it had been possible to oppose national policies to stabilize the exchange rate while evincing great enthusiasm for the EU generally.With the 1992 program tied to monetary integration,such a division of the question was less feasible. In many real and prospective EU members,participation in the EMS was seen as critical for full EU membership.This was important and has been dealt with in detail elsewhere (Frieden,1994b;Garrett,1994;Martin,1994). The second literature of relevance,which is very large,is that which analyzes the circumstances in which national welfare is improved by a fixed exchange rate,and at the limit,by a monetary union (see Genberg,1990,for a survey).Fixing the exchange rate imposes costs-the monetary authorities give up a policy instrument-and is only desirable from a national welfare standpoint if counterbalanced by greater benefits. One crucial observation is that in an economy financially integrated with the rest of the world (or the relevant region),the government is faced with a choice between monetary policy independence and exchange-rate stability. In a financially open economy,interest rates are constrained to world (or regional)levels.?Monetary policy largely involves the exchange rate:Mone- tary expansion drives the exchange rate down,makes locally produced goods cheaper in comparison to imports,and stimulates the demand for domesti- cally produced tradable goods.Fixing the exchange rate forgoes this instru- ment,removing the possibility of an independent monetary policy.Put differently,as countries become more financially integrated,the effectiveness of national monetary policy declines.In this sense,the social welfare gains to be had from a stable exchange rate tend to rise along with financial integration.8 7.To be precise,it is covered (exchange rate-adjusted)interest rates that are constrained to be equal.The insight is that of the famous Mundell-Fleming approach,which originated with Mundell (1962 and 1963).The basic model can be found in any good textbook discussion of open-economy macroeconomics;a useful survey is Corden (1986). 8.Or,what is the same thing,the costs of forgoing the exchange rate as a policy instrument tend to fall with financial integration.This is a bit oversimplified and assumes that exchange-rate stability is desirable in and of itself.However,there is no question that the trade-off between exchange-rate stability and monetary autonomy,absent or weakin a financially closed economy, grows in importance as the economy becomes more financially open
Frieden IMPACT OF GOODS 201 A related argument comes from the literature on optimal currency areas. Such areas,which are tantamount to monetary unions,can be regarded as particularly binding fixed exchange-rate regimes.In this framework,cur- rency union makes economic sense for regions among which factors are mobile and economic shocks are correlated (Mundell,1961,and McKinnon, 1963,are the classic statements).If two regions are so economically inte- grated that market conditions are closely linked between them,having a common monetary policy is efficient(and having separate monetary policies may be impossible). Here,too,the integration of goods and capital markets plays a crucial role The more integrated economies are among themselves,the less effective independent monetary policies will be(due to the ability of factors and goods to move in response to different policies),and the more desirable is monetary union.The economic analysis is therefore quite clear:The attractiveness of fixed exchange rates rises with the level of economic integration.This does not address the differences between full monetary union and other fixed-rate regimes.For the sake of this analysis,the two can be regarded as points relatively close on a continuum that runs from complete floating up to cur- rency union. These efficiency-based arguments for fixing exchange rates confront problems as explanations of European monetary integration in practice.There is,in fact,plenty of evidence that Europe has not met the criteria by which fixed exchange rates would improve social welfare;social welfare grounds do not support Economic and Monetary Union (EMU)as an economically efficient policy for EU members(Eichengreen,1992b,pp.4-25).Before the late 1980s,the EU was not very integrated financially-capital controls were common.0 Factors remain only imperfectly mobile among the members of the EU,and EU economies are not so integrated that they share common macroeconomic conditions.Neither fixed exchange rates nor currency union are clearly welfare-improving at this stage within the EU.Put differently, even if government policies were driven entirely by efficiency criteria,this could not explain the movement toward monetary integration in the EU,for neither fixed rates nor currency union are economically efficient policies for EU members.Certainly,as the EU has become more integrated over time, 9.Of course,economic integration and currency union can interact:Having one currency makes it easier for factors to move within a region.On such interactive effects in international monetary relations,see Frieden(1993). 10.Since the middle 1980s,capital controls have been removed and the EC has become more integrated financially.This,however,does not explain the course of the EMS before financial integration
202 COMPARATIVE POLITICAL STUDIES/April 1996 monetary integration is somewhat more economically defensible than it once was,but efficiency-based economic principles cannot be evinced to explain the EMS,EMU,or variations in national policy toward them. Although these factors are important,they are not sufficient to explain change over time or across countries in willingness to fix exchange rates within Europe.I present below,and evaluate empirically in what follows,an argument that variations in the level of trade and investment with potential currency-regime partners is a crucial contributor to explaining variations in exchange rate policy. Although current levels of goods and capital market integration do not make stabilizing exchangerates unambiguously welfare-improving for mem- bers of the EU as a whole,they do make it attractive to stabilize exchange rates for those economic agents heavily involved in intra-EU trade and payments.Currency arrangements have a differential effect on firms and individuals,which can be expected to translate into cross-cutting political pressures on national policy makers.The crucial political issue typically has to do with how important currency predictability is,relative to the ability of national monetary authorities to depreciate the exchange rate to stimulate the local economy or increase the competitiveness of national producers.Re- linquishing this option is not popular,other things being equal,even if it does lead to more stable currency values. However,higher levels of cross-border trade and investment increase the size and strength of domestic groups interested in predictable exchange rates. Firms with strong international ties support a reduction of currency fluctua- tions.These effects are especially important to banks and corporations with investments throughout the EU.In addition,tradable producers with EU-wide markets,and for whom price competition is relatively less important-those whose appeal is based primarily on quality or technological prowess-may be less concerned about ability to devalue than about currency stability.12 11.In this context,it is important to keep in mind that eliminating the ability to devalue for high-inflation countries typically leads to a transitional (inertial)real appreciation of the exchange rate.This is especially troublesome for producers of tradable goods that compete primarily on price,as fixing the exchange rate in conditions of inflation above the EU average exposes import competitors to substantial price pressure. 12.It is often objected that forward markets allow firms to protect themselves against potential currency fluctuations.Although this is true for short-and medium-term exchange-rate movements,it is not true over the longer time horizon typically of concern to investment planners.Indeed,the existing literature on corporate finance distinguishes clearly between transaction exposure,which can be effectively hedged,and operating exposure,which cannot. See,for a typical example,Shapiro(1992),chapter 10.I am indebted to Rich Lyons for bringing this to my attention