II Assessing the Long-Run Benefits of Euro Adoption

International Monetary Fund
Published Date:
April 2005
  • ShareShare
Show Summary Details

The first step in the CECs’ decisions about the timing of euro adoption entails reaching a view on long-term benefits and costs. Unlike for the United Kingdom and Denmark (Box 2.1), the outcome of this assessment does not change the ultimate requirement that countries adopt the euro. But an assessment of clear and sizable benefits relative to costs would argue for urgency in undertaking policy changes and preparations necessary for euro adoption. A balanced or negative assessment would suggest a slower approach—possibly even an indefinite delay—particularly if, for example, waiting until a greater degree of real convergence has been achieved would lower costs and raise benefits.

Benefits, which are the subject of this chapter, will come through a number of channels that differ widely in terms of likely strength and amenability to quantification. The most important will be through trade creation and attendant effects on income. An active area for current research, these gains are quantifiable even if controversy reigns over their exact size. The evidence on gains from other sources—in particular from the elimination of risk premia on interest rates, lower transaction costs, increased transparency in intercountry price comparisons, greater competition in goods and factor markets, and the discipline of the euro area’s macroeconomic policy framework—is more scanty. Indeed, it is not even clear to what extent these effects are already captured in estimates of the gains for income through trade creation. Gains from all these sources must be compared with the long-term costs of active participation in the euro area, primarily stemming from the loss of the monetary policy instrument. The latter will be examined in Section III, where conclusions on likely net effects are also drawn.

Gains from Trade

A spate of literature in the past few years has focused on measuring the trade creation effects of currency unions. Most of these studies take the form of gravity models, used to examine how trade between pairs of countries (the sum of exports of each to the other measured in dollars) is affected by a variety of variables. These models are typically estimated on large panel data sets drawn from countries in and outside currency unions and in most instances find significant and sizable effects of currency unions on trade. An important drawback of these models, however, is that they are based on reduced-form equations that do not have clear structural underpinnings. Therefore, they answer only the question “Do currency unions have a significant trade creation effect?” but not “Through what channels do currency unions affect the volume of trade between two countries?”

While most estimates of the effects of currency unions on trade are positive and significant, sizes vary widely. In a comprehensive survey, Rose (2002) examines 24 studies and finds a pooled estimate that currency unions double trade between union partners: even the lowest estimate among these studies shows a 15 percent increase in such trade stemming from the creation of a currency union.

Many of the studies feeding into this comprehensive survey were motivated by research suggesting startlingly large trade creation yielding substantial benefits for output growth. Rose (2000) and Frankel and Rose (2002) were particularly provocative. Using a two-stage approach, they estimate in the first stage a linear model of bilateral trade for a panel of cross-country data covering trade relationships among over 180 countries and dependencies.7 They model trade between pairs of countries as an increasing function of their sizes and a decreasing function of the distance between the two countries, with controls for colonial, political, and free trade agreements and a dummy variable for membership of the two countries in a currency union. The results suggest that membership in a currency union triples trade with other currency union members. The estimated coefficient on a dummy variable that is unity when one of the countries belongs to a currency union is taken as evidence that the stimulus to trade from a currency union does not come at the expense of diversion of trade with nonmembers; it does not, however, rule out intranational trade diversion. In a second stage, they estimate that every 1 percent increase in a country’s overall trade as a share of GDP raises per capita income by at least ⅓ of 1 percent. They conclude that EMU membership could boost GDP by 20 percent in Poland and 35 percent in Hungary over two decades.

Box 2.1.Staying Out of the Euro—Rationale and Perceived Benefits and Costs

Three EU members—Denmark, Sweden and the United Kingdom—remain outside the euro area. Denmark is a member of ERM2 operating within ±2¼ percent exchange rate bands agreed with the ECB; Sweden and the United Kingdom are inflation targeters. Denmark and the United Kingdom are members of EMU with permanent opt-out clauses; formally, Sweden will adopt the euro once it satisfies all the Maastricht entry criteria. Each country debated the pros and cons of euro adoption prior to the founding of EMU and has since reexamined its position. In Denmark and Sweden—where the authorities have endorsed euro adoption—referenda on this proposal were rejected (in Sweden, the referendum was on whether to join ERM2). The British government, in 1997, committed in principle to adopt the single currency, and the issue will be brought to a referendum once the government considers that euro adoption would be in the national interest. This box looks at why these countries have stayed out or, where the authorities have supported euro adoption, the basis for their decision.


The Danish government has recommended full participation in EMU in order to reap structural gains and reduce macroeconomic instability.1 Asymmetric shocks within the EU are seen as coming mostly from insufficient economic policy coordination or from domestic policy shocks (and joining EMU would reduce to some extent these types of shocks), rather than from external asymmetric sources. Danish fiscal policy, with its strong structural surplus, is considered well-suited for macroeconomic stabilization within EMU and has historically operated in a countercyclical manner, reflecting both large automatic stabilizers and discretionary policy measures. Notwithstanding the close monetary policy links with the EMU that already exist within the very narrow ERM2 band, the government sees significant advantages to (and implicitly, costs to remaining out of) full membership in EMU. Among the structural advantages they identify are: a lower interest rate differential; reduced currency transaction costs;2 and increased transparency and competitive pressures. In terms of economic stabilization, full participation in EMU is seen as eliminating fluctuations in interest rate differentials caused by international financial turbulence, and Danish economic conditions would have an influence on monetary policy decisions in the euro area. On the cost side, they note the one-off administrative expense associated with the currency changeover.


The Riksdag adopted in 1994 and again in 1997 a position in favor of joining the euro area from its inception. The government established a commission in 1995 to examine the consequences of EMU for Sweden and the effects of belonging to or remaining outside the monetary union.3 In its 1996 report, the commission concluded that while political considerations supported Sweden joining EMU in 1999, economic considerations—especially the need to put public finances in better shape for macroeconomic stabilization within the single currency—argued against. The four principal reasons underpinning the commission’s conclusion were as follows.

  • It would be risky to participate in the monetary union—thereby losing the exchange rate instrument to adjust to adverse economic shocks—when the unemployment rate is high.
  • An increase in unemployment would be damaging to public finances and macroeconomic stability.
  • To minimize the risk of a growing gulf between voters and political representatives, further time was needed for a broad, objective, and comprehensive debate on EMU.
  • Initially a number of EU countries would probably remain outside the EMU. This would reduce both the political cost of staying out and the gains for efficiency and transaction costs of going in.

The commission concluded that if unemployment were cut, public finances stabilized, and trade with the euro area increased, the case for entry into EMU would be strong. On the costs of waiting to join, the commission emphasized the positive spread on long-term nominal interest rates and risks that the Swedish crown would be hurt by doubts about prospects for low inflation.

Parliament decided in 1997 that Sweden would not join the euro area in 1999, but urged continued preparation so as to allow a smooth switch to the euro in the future. In 2000 and 2002, when compliance with the Maastricht criteria was assessed, Sweden was judged to have met the inflation, public finance, and interest rate criteria, but not the exchange rate criterion because it was not a member of ERM2. In September 2003, a proposal to adopt the euro in 2006 was rejected in a referendum.

United Kingdom

In June 2003, the U.K. government reexamined its position on euro adoption based on five criteria that were first applied in 1997 to assess whether the United Kingdom should be one of the founding members of EMU.4 The five economic tests were the following.

  • Are business cycles and economic structures compatible so that we and others could live comfortably with euro interest rates on a permanent basis?
  • If problems emerge is there sufficient flexibility to deal with them?
  • Would joining EMU create better conditions for firms making long-term decisions to invest in Britain?
  • What impact would entry into EMU have on the competitive position of the United Kingdom’s financial services industry, particularly the City’s wholesale markets?
  • In summary, will joining EMU promote higher growth, stability, and a lasting increase in jobs?

These tests form the basis for the government’s assessment of whether a clear and unambiguous economic case for euro membership exists. They are seen as the discipline to ensure that Britain’s experience in ERM—when it “joined at the wrong rate and at the wrong time without either convergence or flexibility”—is not repeated.5 The 2003 assessment examines the degree of sustainable convergence and the costs of delaying the benefits of joining EMU.

The 2003 assessment concludes that, since 1997, the United Kingdom has made progress toward meeting the five economic tests. It finds, however, that the tests of sustainable convergence and flexibility have not been met. It concludes that the lower is the degree of flexibility in the EU, the greater is the premium on U.K. flexibility. It estimates that if the convergence and flexibility preconditions had been satisfied and the United Kingdom were to join EMU, it “could enjoy a significant boost in trade with the euro area of up to 50 percent over 30 years and that U.K. national income could rise over a 30-year period by between 5 and 9 percent.” Failing to meet these preconditions might therefore be thought of as costing the U.K. economy up to ¼ of 1 percentage point in potential output growth per year in terms of higher trade alone. Despite this and other potential benefits of euro adoption—increased investment and a boost to financial services and jobs—the study finds that “a clear and unambiguous case for U.K. membership of EMU has not been made.”

The Treasury also reevaluated its 1997 decision not to join EMU, by comparing actual macroeconomic outcomes under inflation targeting with a stylized simulation of a “what if” scenario. This shows that had the United Kingdom entered EMU in 1999, economic instability in the subsequent five years would have been greater than was actually the case, with larger fluctuations in output growth, procyclical real interest rates, and a stronger real exchange rate. The better macroeconomic outcome under monetary policy independence is ascribed in part to policy reforms since 1997.

1 This paragraph is based on “The Danish Economy 2000: Medium-Term Economic Survey” (Denmark, Ministry of Finance, 2000).2 The Dansmarks Nationalbank notes, however, that many Danish shops accept payment in euro currency.3 This discussion draws on “The EMU Commission and the Decisions of the Riksdag,” The commission’s report is “EMU: A Swedish Perspective,” Calmfors and others (1997).4 “U.K. Membership of the Single Currency: An Assessment of the Five Economic Tests” (United Kingdom, HM Treasury, 2003c).5 Statement to Parliament by Chancellor Brown on June 9, 2003.

The conclusions from Frankel and Rose have spawned several investigations of possible upward biases in their results. These subsequent studies, included in Rose’s pooled estimate, generally find a smaller albeit still sizable effect. Persson (2001) examines selection bias in the Frankel and Rose estimates and finds that correcting for this lowers the estimates of the impact of currency union on trade to about 60 percent. Thom and Walsh (2002) examine Ireland’s break with sterling in 1979 and find no significant adverse effect on trade, which they suggest is consistent with such a selection bias arising from sampling of currency unions that involve overwhelmingly poor and/or small nations. Tenreyo and Persson (2001) find that existing trade patterns are a determining factor in the formation of currency unions, implying a simultaneity bias. Correcting for such a bias, they estimate trade gains of about 60 percent. Pakko and Wall (2001) and Clark and others (2003) suggest that the Frankel and Rose results may be affected by an estimation bias, whereby the benefits of a currency union are not the result of a common currency as such but of other omitted variables (for example, common history between country members in the currency union). Minford (2002) argues that large estimates of the currency union effect on trade capture not just the single currency effect but also other determinants of trade such as harmonized regulations and legal structures.

As EMU gains a history, early estimates of the actual gains for bilateral trade between member countries have begun to appear. These studies (De Nardis and Vicarelli, 2003; Micco, Stein, and Ordonez, 2003; and Faruqee, 2004) generally find that the euro has already had a positive and noticeable impact on trade, albeit modest in size compared with Rose’s pooled estimate derived from evidence on other currency unions. Any discrepancies may reflect that Rose is examining long-term impacts and the EMU results are of necessity short-term. An important aspect of this literature is that it tends to include all fixed factors associated with a given country pair that have affected bilateral trade flows historically, thus eliminating potential omitted variables bias. Specifically, using post-1999 data, Micco, Stein, and Ordonez (2003) find the EMU effect to be between 4 and 10 percent when compared with trade between all other pairs of countries, and between 8 and 16 percent when compared with trade among non-EMU countries. There is no evidence of trade diversion. Faruqee (2004) finds similar results, estimating that EMU has raised intra-area trade by some 10 percent relative to trade with other industrial countries. Moreover, he finds the trade impact of EMU has been largest and statistically significant in the last two years of the sample (2001–02), suggesting that trade gains may rise over time. Assuming the pace of intra-EMU trade gains in the last four years were to persist for the next 20 years, cumulative gains would reach some 60 percent. Partly reflecting findings for intra-EMU trade gains, the 2003 U.K. Treasury study concludes the potential increase in U.K. trade with the euro area resulting from U.K. membership of EMU would range from 5 to 50 percent: the lower bound reflects estimates of the increase in intra-euro area trade that has already occurred under EMU, and the upper bound is deemed closer to the likely outcome in the long term (United Kingdom, 2003e).

Notwithstanding the uncertainty that still surrounds the effect of currency union on trade, estimated ranges of such effects for the CECs after euro adoption can be large. Table 2.1, which draws on estimated parameters from a variety of studies, shows euro-area membership raising trade by between 6 and 60 percent over 20 years, except in Poland: there, owing to a smaller ratio of total trade to GDP than in the other CECs, estimates are lower. The upper end of the range (based on the pooled fixed-effect meta estimate in Rose, 2002) shows an increase in trade of some 85 percent with the euro area within two decades of euro adoption. The lower end shows a 10 percent increase in such trade, consistent with estimates of actual trade gains during the five years of EMU (Faruqee, 2004). As such, the latter is quite conservative for two reasons. First, these actual gains are assessed after only five years of EMU’s existence—estimates using non-EMU data focus on substantially longer gestation periods when deeper effects of restructuring are likely to produce larger gains. Second, the range of trade gains among EMU countries is large—from 6–15 percent—and with more flexible economies, CECs may expect to be toward the top end of the range even after only five years. Assuming that income will rise by one-third of the increase in trade/GDP, a reasonable range for the impact on income is between 2 and 20 percent, with lower estimates for Poland. Estimates in the Magyar Nemzeti Bank (MNB) (2002) show that the euro may raise GDP growth via the expansion of trade by 0.6–0.8 percentage points over 20 years fall into this range. Estimates in National Bank of Poland (NBP) (2004)—an increase in annual GDP growth of 0.2–0.4 percentage points—also fall in the range shown here.

Table 2.1.CECs: Potential Long-Run Increases in Trade and Per Capita Output Following Euro Adoption1(In percent)
High estimate2Low estimate4
Czech Republic602072
Slovak Republic592072
Sources: Frankel and Rose (2002); Rose (2002); IMF, World Economic Outlook; IMF, Direction of Trade Statistics; and IMF staff calculations.

Based on 2002 trade and GDP data.

Percent change in ratio of total trade to GDP. Assumes currency union increases trade (the sum of exports and imports) with current euro-area members and the other accession countries by 85 percent over 20 years. Based on Rose’s fixed-effect meta estimates (Rose, 2002). Assumes trade with other countries rises in line with GDP.

Assumes a 1 percent increase in total trade/GDP increases real GDP per capita by 0.33 percent. Estimate drawn from Frankel and Rose (2002).

Assumes currency union increases trade with current euro-area members and the other accession countries by 10 percent. Estimate drawn from analysis of average five-year gain from EMU experience.

Sources: Frankel and Rose (2002); Rose (2002); IMF, World Economic Outlook; IMF, Direction of Trade Statistics; and IMF staff calculations.

Based on 2002 trade and GDP data.

Percent change in ratio of total trade to GDP. Assumes currency union increases trade (the sum of exports and imports) with current euro-area members and the other accession countries by 85 percent over 20 years. Based on Rose’s fixed-effect meta estimates (Rose, 2002). Assumes trade with other countries rises in line with GDP.

Assumes a 1 percent increase in total trade/GDP increases real GDP per capita by 0.33 percent. Estimate drawn from Frankel and Rose (2002).

Assumes currency union increases trade with current euro-area members and the other accession countries by 10 percent. Estimate drawn from analysis of average five-year gain from EMU experience.

The actual magnitude of trade gains for the CECs will be determined by a confluence of factors, including policy influences on factor mobility and initial conditions related to openness and the pattern of trade. Faruqee (2004) finds evidence of considerable country-specific dispersion around the average EMU effect, reflecting a number of factors including the size of trade relative to GDP and the composition of trade. The importance of trade shares presumably reflects the outward orientation of a country’s producers. Also, countries with above-average shares of intraindustry trade tend to have experienced above-average trade gains. This finding may reflect a tendency for trade in similar products to be more sensitive to exchange rate changes than trade in goods that are less easily substitutable in production or consumption. Thus, for the CECs, where openness to trade and the share of intraindustry trade tend to be relatively high, actual trade gains from joining the euro area could be skewed toward the upper end of illustrative ranges.

The effect of currency unions on trade appears to stem from more than simply eliminating exchange rate volatility. Indeed, recent work has been rather ambiguous on the significance and size of the effect of exchange rate volatility on trade. Neither Bacchetta and van Wincoop (2000) nor Clark and others (2003) find evidence that exchange rate volatility systematically influences trade flows. Controlling for reverse causality, however, Broda and Romalis (2003) find that real exchange rate volatility reduces bilateral trade (in differentiated goods), but the size of the effect is small.8 Other methodologies find larger, significant effects. Skudelny (2003) observes that, in contrast to estimates using time-series data, studies using cross-section data find a significant and negative effect of exchange rate volatility on trade. He offers the explanation that in the time-series dimension, the exchange rate volatility term could be picking up volatility in the other variables. Thus, Dell’Ariccia (1998) finds that eliminating exchange rate volatility for the EU countries during 1975–94 would have produced trade between EU members some 12 percent higher than in fact occurred. Using a gravity model extended to capture common border effects, Taglioni (2002) pools three-digit industry-specific trade data on international and intranational transactions—the latter representing trade where currency volatility is nil by definition. He shows that exchange rate volatility indeed influences trade and accounts for an important part of the border effect in Europe.

The question remains, however, whether the effects of currency unions on trade go beyond the elimination of exchange rate volatility. Frankel and Rose (2002) attempt to shed some light on this issue by investigating whether there is a difference between the effect of a currency board and that of a currency union on trade. Estimates of equations including both variables find each to be positive and significant and similar in size, but they acknowledge that there is no compelling theoretical reason why this effect should be the same. The currency union effect on trade goes beyond simply the removal of exchange rate volatility—perhaps by eliminating all future uncertainty about exchange rate changes. Beyond direct effects on trade, they argue the certainty of long-term exchange rate stability can have indirect effects through improved incentives for FDI. This would be consistent with the finding that the gains from currency unions occur over periods of several decades.

Beyond Trade: Benefits on Income and Investment

Beyond trade creation, sharing a common currency can be expected to have other benefits that are even more difficult to quantify.

First, elimination of exchange rate risk should lower real interest rates. For the CECs, an important channel of risk reduction would be the lower probability of shocks to the financial sector caused by contagion. Using a general equilibrium model, MNB (2002) estimates that lower risk premia following the euro changeover would raise growth by about 0.1 percent per year over the long term via lower real interest rates. Estimates by the NBP (2004) suggest a similar effect, of about 0.2 percentage points per year.

Second, euro adoption will reduce transaction costs, but estimates of lower costs are generally small. In 1990, the European Commission estimated the potential savings on transaction costs through the elimination of national currencies at around 0.4 percent of EU GDP (EC, 1990). The EC suggests, however, that the larger countries, whose currencies are heavily traded, might experience savings in the region of 0.1 to 0.2 percent of GDP compared with gains of around 1 percent of GDP for the smaller EU economies. Borowski (2003) and MNB (2002) estimate that euro-area membership would result in savings on transaction costs equivalent to about 0.2 percent of GDP for Poland and Hungary.

Third, elimination of the potential for exchange rate swings should improve the attractiveness of the CECs as a platform for FDI, particularly in sectors open to trade. This effect would come on top of the effects of EU membership and the opportunities arising from privatization and deregulation (Bevan, Estrin, and Grabbe, 2001). It may, however, be captured implicitly in the estimates of trade creation. The U.K. Treasury study “EMU and Business Sectors” (2003a) acknowledges the potential for higher FDI but does not detect a clear effect so far in the EMU countries.

Fourth, joining the euro area would secure a clear framework for macroeconomic policy discipline. Currie (1997) singles out gains from enhanced credibility of monetary policy and lower inflation expectations as one of the most significant effects of joining the euro area. Frankel and Rose (2002), however, find that growth benefits from improved monetary credibility and stability of joining a currency union, per se, are not as clear-cut as benefits coming through trade.

In sum, euro adoption is expected to be associated with significant trade creation over the long term, as well as other benefits such as technology transfer through direct investment, lower transaction costs, lower emerging market financial risk, and a strong framework for policy discipline. With benefits from a currency union feeding into higher growth, convergence of the CECs to euro-area income levels should accelerate. While difficult to estimate, these effects should reduce the catch-up period for these countries.


This included all countries, dependencies, territories, overseas departments, and colonies for which the United Nations Statistical Office collects international trade data.


A doubling of real exchange rate volatility decreases trade in differentiated products by about 2 percent.

    Other Resources Citing This Publication