What event caused many EU countries to wait on adopting the euro as a common currency

Ten countries acceded to the European Union (EU) in 2004, including five in Central Europe and the three Baltic republics (along with Cyprus and Malta). Should they also join the European Monetary Union (EMU), as envisioned in the Maastricht Treaty? The advantages of belonging to such a currency area are fairly straightforward. They include most notably the promotion of trade and growth. The disadvantages are a bit more complicated but include most notably the loss of the ability to pursue an independent monetary policy.1 What would it take for the advantages to outweigh the disadvantages? The well-known theory of optimum currency areas weighs the advantages of fixed exchange rates against the advantages of floating.2 One standard textbook criterion is a particular kind of convergence of the candidate economy to the economy of the euro area: a synchronization of the business cycle or what is called in the jargon “symmetry of shocks.”3 If the candidate country experiences economic downturns when and only when the rest of the EMU experiences economic downturns, it will not be giving up much to allow its monetary policy to be set in Frankfurt. The interest rates that suit the rest of the EMU are likely to suit the candidate country as well. If cyclical correlation is low, on the other hand, having to accept the interest rate constraint is more likely to be a hardship.

To be sure, there are other possible definitions of real convergence. A second one is a similar structure of the economy’—for instance, similar shares of agriculture versus manufacturing. But this should not be a criterion for adopting the euro. True, differences in composition of the economy can be a source of asymmetric shocks, for example, in the event of a fall in world prices for agricultural commodities. But differences in composition are also a primary source of trade benefits.

A third definition of real convergence is similar levels of productivity, per capita income, and relative prices of nontraded goods. The relevance will be considered later in the paper, under “Convergence.” But the paper will focus primarily on the first definition of real convergence: synchronized or correlated business cycles.

I will particularly emphasize the endogeneity of cyclical correlations with respect to the decision to seek economic and monetary integration in the first place. The argument is that a link to the euro, like accession to the EU, promotes trade with Western Europe, which in turn raises the cyclical correlation, which in turn makes the country in question a better candidate for the EMU. That the creation of a common currency could alter patterns of international trade was one of the original motivations of the EMU’s architects. Nevertheless, it is only relatively recently that academic researchers have found convincing evidence that this is a major effect. This paper will explain what has been learned from recent research on (1) the effect of common currencies on trade among members,4 (2) the further implications for the benefits of a common currency, and (3) the further implications for the costs of a common currency as reflected in cyclical correlations—an important, but endogenous, criterion for membership in an optimum currency area. The paper concludes with thoughts about whether the countries of Central and Eastern Europe are ready to join the EMU.

The Effect of Common Currencies on Trade Among Members

Until relatively recently, economists were skeptical about whether a reduction in exchange rate variability gives a substantial boost to trade. The skepticism had both a theoretical and an empirical basis. Theoretically, the argument was that importers and exporters can hedge exchange rate uncertainty. Empirically, econometric studies found little evidence that exchange rate variability had an adverse effect on trade.

The problem with the theoretical argument, however, is that forward and futures markets (1) do not exist for most countries and for most longer-term horizons, (2) come with transactions costs when they do exist, and (3) come also with risk premiums, which can drive a wedge between the forward rate and the expected future spot rate. The problem with the empirical evidence is that it (1) was based mostly on time series, where it was difficult to sort out other influences on trade, and (2) was based mostly on large industrialized countries.5 When smaller countries were included in cross-section studies, some effects started to show up. This was particularly true for studies of bilateral trade. Data on trade among 100 countries offer 9,900 observations for each year (100 X 99). Having that volume of data allows the researcher to control for such other important determinants of trade as country size, bilateral distance, and common borders, as in the gravity model, thereby obtaining better estimates.6

The most important discovery was made by Andrew Rose (2000), who looked at a data set that included many very small countries and dependencies. To begin with, he found a statistically significant effect of bilateral exchange rate variability on bilateral trade. But, beyond that, he found a large effect of common currencies on bilateral trade. Enough small countries use some other country’s currency (most of them either the U.S. dollar, the French franc, the pound sterling, the Australian or New Zealand dollar, or the South African rand) that it was possible to isolate the effect. His remarkable estimate, which by now he has replicated in various forms many times, was that a common currency triples trade among members.

A threefold effect is very large, and the finding was, understandably, greeted with a lot of skepticism. There are four grounds for skepticism. First, one cannot infer from cross-section evidence what would be the effect in real time of countries adopting a common currency. Second, the statistical association between currency links and trade links might not be the result of causation running from currencies to trade but might arise instead because both sorts of links are caused by a third factor, such as colonial history, remaining political links, complementarity of endowments, or accidents of history. Third, the estimated effect on trade (and on income, to be discussed in the next section) seems too big to be believable. Fourth, Rose’s evidence came entirely from countries that were either small (e.g., Ireland, Panama, or African members of the CFA—Communauté Finaninancière de l’Afrique) or very small (e.g., Kiribati, Greenland, Mayotte), and so it was not clear that the estimates could be extended to larger countries.

While each of these four arguments has some validity, to each there is a better response than one might expect.

First, regarding the time dimension, subsequent research on time-series data finds that a substantial share of the tripling that Rose had estimated from the cross-section data (which is presumably the long-run effect) shows up within a few decades of a change. Using a 1948–97 sample that includes a number of countries that left currency unions during that period, Glick and Rose (2002) find that trade among the members was twice as high in the currency union period as afterward. This suggests that roughly two-thirds of the tripling effect may be reached within three decades of a change in regime. (This reasoning assumes symmetry with respect to entry into and exit from currency unions.)

Second, regarding the possible influence of third factors, Rose has done a thorough job of controlling for common languages, colonial history, and remaining political links. The large estimated effect of a common currency remains. It seems very possible that there remain other third factors (e.g., accidents of history) that influence both currency choices and trade links. Nevertheless, Rose’s various extensions of the original research—these robustness tests together with the time-series results—reduce the force of this critique.7

Third, regarding the surprisingly large magnitude of the estimates, it is important to consider something else that we have learned in recent years: there is a strong home-country bias in trade, which is especially surprising in light of all one hears about globalization. Many studies have found that people trade with their fellow citizens far more easily than with those living in other countries. This finding emerges whether one looks at the volume of trade flows between locations or at the ability of arbitrage to keep prices in line across locations. It holds even when one controls for the effects of distance; trade barriers; and linguistic, social, and historical differences. It holds even between the United States and Canada. The best-known finding is that Canadian provinces are 20 times more prone to trade with each other than with U.S. states (McCallum, 1995). This estimate was cut very roughly in half after the Canada-U.S. Free Trade Area (FTA) went into effect (Helliwell, 1998) and was cut further when a few more factors were controlled for (Wei, 1996). Nevertheless, a substantial bias—roughly threefold—remains even in this case. The home-country bias must certainly be higher for most other country pairs.8

Similarly, studies of the ability of arbitrage to narrow price differentials find that crossing the U.S.-Canada border discourages trade more than does traveling the entire length of Canada (Engel and Rogers, 1996) and that the barrier is even greater for other pairs of countries (Engel and Rogers, 1998; Parsley and Wei, 2001a, 2001b). What can explain these remarkable findings of home bias in both quantity data and price data? The difference in currencies is not an implausible explanation, given the paucity of alternative candidates.9

Regarding the applicability of the results to large countries, we will not know for sure until the EMU experiment is older. It would seem plausible that very small geographical units (the Kiribatis) are so dependent on international trade—due either to inadequate scale of the domestic market or to insufficiently diversified factors of production—that strategies such as currency unions or free trade areas would have a larger payoff for them than for larger, more self-sufficient, economies. But there are two counterarguments. First, Rose has tested whether there are any nonlinearities among his currency union sample (e.g., any difference between the effects among units that are merely small and those that are very small). He found no significant difference. Second, the home-country bias seems to be linear, regardless of the size of the country. That is, if two small units join together, thereby doubling the size of the economy, the ratio of trade to gross domestic product (GDP) falls—that is, home-country bias increases—as much (roughly by 0.2, in log form) as when two large units join together.10 To the extent that currencies explain this, the effect does not seem to be limited to small countries.

Finally, we now have four years of data since the EMU went into effect in January 1999. Econometri-cians have updated the gravity estimates to see what can be learned from the record so far. Micco, Stein, and Ordoñez (2003) find that for pairs of the 12 countries that joined the EMU, trade has increased significantly. The estimated effect is about 15 percent beyond what could be explained by growth and other factors—actually a range of 6 to 26 percent, depending on the use of country and year dummies, with a larger set of 22 industrialized countries. The Micco, Stein, and Ordoñez (2002) estimates of “differences in differences” reveal that between 1992 and 2001 the boost to intra-EMU trade was about 18 to 35 percent, depending on whether one uses country-pair dummies, or conditions on the standard gravity variables. These magnitudes are less than in the Rose studies. As the authors quite reasonably point out, however, the effects are not only statistically significant but also economically important, especially considering that the sample covers only the first four years of the EMU, a period in which the euro did not even circulate.

Other evidence confirms the finding. Bun and Klaassen (2002) also update gravity estimates and find that “the euro has significantly increased trade, with an effect of 4% in the first year” (p. 1) and a long-run effect projected to be about 40 percent. Takata (2002, p. 11) calculates that the United Kingdom–euro area intensity of trade rose gradually in the early 1990s and sharply in 1999–2000. (Trade intensities are more rudimentary estimates than full gravity models but are much easier to compute and usually give similar answers regarding changes over time.) Studies with price data have tended to confirm that the euro is facilitating arbitrage among the markets of member coun-tries.11 It seems that the trade effects of monetary union are not, after all, limited to small countries.

The Benefits of Monetary Union

Boosting trade is of interest primarily as a determinant of economic performance. (Noneconomic motivations for encouraging trade, such as binding countries together politically, are outside the scope of this study.) But there are several sorts of ways that an increase in trade among members of a group feeds into the question of the economic advisability of opting for a common currency. An increase in trade can influence the benefits of a common currency, or the costs.

One reason that trade patterns are relevant has to do with the advantage of a common currency for exporters and importers. The fact that the elimination of exchange rate uncertainty makes life easier for those engaged in trade will be more important, the higher is the share of trade in GDP, even if the level of trade does not change. Also, a fixed exchange rate will help stabilize the price level, the higher is the share of trade in GDP. For these reasons, McKinnon (1963) argued that a key factor determining the advisability of fixing the exchange rate is the ratio of tradable goods to GDP. Thus openness has long been on the standard textbook list of optimum currency area criteria. One implication is that if trade among the members of the EU is increasing over time, then they will satisfy the optimum currency area criteria more strongly in the future than in the past. A related implication is that even if a country does not satisfy the optimum currency area criteria ex ante, if it joins a currency area anyway and enough time passes to increase trade with other members substantially as a result of the common currency, it may satisfy the optimum currency area criteria in the future. In Frankel and Rose (1996, 1998), we called this the endogeneity of the optimum currency area criterion.

The second factor has to do with the long-run determination of growth. Trade is not just another sector. Theory and empirical evidence suggest that trade is good for the level or growth rate of income. Currency unions raise openness, and openness raises real income. Frankel and Rose (2002) combine estimates of the effects of a common currency on trade and the follow-up effects of higher trade on GDP, to derive estimates of the effects of common currencies on GDP. Table 1 shows that membership in a typical currency union raises the ratio of trade to GDP by an estimated 10 to 26 percent. But joining a currency union with particularly important trading partners (e.g., large and close neighbors) can have a larger impact. For example, if Poland were to join the EMU and thereby triple trade with euro countries, its ratio of total trade to GDP would eventually more than double, from 0.50 to 1.12. Once the increase in trade was realized, the estimated effect would be to raise real income by 20 percent over the subsequent 20 years—quite a substantial effect, if it is accurate. Similarly, if Hungary were to join and thereby triple trade with the euro area, its ratio of trade to GDP would also eventually more than double (from 0.76 to 1.83), and its real income would increase by 35 percent.12

Table 1.

The Effect of Currency Union Membership on Aggregate Trade/GDP

Source: Frankel and Rose (2000). Note: Regressand is log of trade/GDP, from Penn World Table. Robust standard errors are recorded in parentheses. GDP = gross domestic product; RMSE = Root Mean Squared Error; RoW = Rest of World.

1Data in this column are country-fixed effects.

2Data in this column are year-fixed effects.

Table 1.

The Effect of Currency Union Membership on Aggregate Trade/GDP

Source: Frankel and Rose (2000). Note: Regressand is log of trade/GDP, from Penn World Table. Robust standard errors are recorded in parentheses. GDP = gross domestic product; RMSE = Root Mean Squared Error; RoW = Rest of World.

1Data in this column are country-fixed effects.

2Data in this column are year-fixed effects.

The Implications of Trade Patterns for the Costs of a Common Currency: Asymmetric Shocks

The last factor concerns cyclical fluctuations. If joining a currency union has advantages, such as promoting trade or stabilizing the price level, what is the countervailing attraction of retaining an independent currency? Why don’t all countries fix their exchange rates or join currency unions? The most important advantage of flexible exchange rates is to retain the ability to respond to cyclical downturns by means of monetary policy—a reduction in real interest rates, a real depreciation of the currency, or both—and to respond to cyclical booms in the opposite direction. But this advantage is less important if the domestic economy is highly correlated with the other countries in a prospective currency area (i.e., if shocks are usually “symmetric”), because the changes in monetary policy that the other member countries choose will also be appropriate for the domestic economy. This is another key tenet of optimum currency area theory. But cyclical correlations are not timeless, unchanging parameters. If trade among members of a currency area increases, then the cyclical correlation is likely to change as well.

Artis and Zhang (1995) find that most European countries’ incomes were more highly correlated with that of the United States during 1961–79 but (with the exception of the United Kingdom) became more highly correlated with Germany after joining the European Exchange Rate Mechanism (ERM).13 Fatas (1997) also found increasing correlation within Western Europe. More recently, Darvas and Szapáry (2004) found that with the advent of the EMU, the synchronization of GDP (and components of GDP) among the members has continued to increase.

In Frankel and Rose (1998), we found on a broad cross-section of industrialized countries that an increase in bilateral trade raises the bilateral cyclical correlation. The geographic determinants of bilateral trade, via the gravity model, are used as instrumental variables to deal with what is otherwise likely to be the endogeneity of trade. (In other words, a pair of countries could show high trade links together with high cyclical correlation, but the connection could be spurious: both could be the result of currency links.) Clark and van Wincoop (2001) found that the historical lack of cyclical synchronization within Europe, as compared to within the United States, is explained by the lower level of internal trade (and to a lesser extent the higher degree of sectoral specialization).

Calderón, Chong, and Stein (2003) point out that the experience of developing countries might be different, in that the composition of their economies differs from that of high-income countries. They extend the finding—that trade links lead to cyclical correlation—to a larger sample of 147 countries, a majority of them developing countries, again using the gravity model to control for the endogeneity of trade. They find that the estimated effect on cyclical correlations is not quite as strong for developing countries as it is among pairs of industrialized countries. They attribute the difference to the dominance of interindustry trade for developing countries, versus intraindustry trade among high-income countries.14

Researchers have begun to look at the cyclical patterns in Central and Eastern European (CEE) countries. Boone and Maurel (1998, 1999) find high correlations of CEE countries with the German economy in particular. Fidrmuc (2004) examines all European countries, using output data through 2001; he confirms that correlations with the German economy increased for Hungary and Poland during the 1990s and that the correlation for Hungary is one of the highest in Europe but that the correlations are much lower in other CEE countries.15

Darvas and Szapáry (2004) recently examined eight CEE countries joining the EU in 2004. They found that three have achieved a high degree of synchronization with the EMU economy: Hungary, Poland, and Slovenia during 1998–2002, as compared with 1993–97. The implication is that these three now best satisfy the convergence criterion for joining the EMU optimum currency area. The lesser EMU correlation for the Czech and Slovak republics is attributed to financial crises in the late 1990s, and the lack of any EMU correlation for the Baltics is attributable to their greater relative exposure to Russia and Sweden. It is no coincidence that the EU makes up a higher share of the exports of Hungary, Poland, and Slovenia than of the exports of the other CEE countries (Figure 6 in Darvas and Szapáry, 2004). One could also note the importance of geographical proximity in determining trade links and cyclical correlations: except for the Czech Republic, the CEE countries that border the euro area are the ones with high EMU trade links and correlations.

The effect of trade on cyclical correlation holds as much when high bilateral trade originates in low bilateral exchange rate variability or adoption of a common currency as it does when the bilateral trade originates in proximity, common membership in free trade areas, or other determinants. This is another instance of the endogeneity of the optimum currency area criteria. A country is more likely to be suited to join a monetary union ex post than ex ante, because the cyclical correlation will have gone up in the meantime.

The internal history of the United States is consistent with these findings. Rockoff (2001) argues that it took 150 years before the United States met the criteria for an optimum currency area, asymmetric regional shocks having posed severe problems for much of the nation’s history.

All these findings contradict an earlier surmise of Eichengreen (1992, pp. 14–16), Bayoumi and Eichen-green (1994, pp. 4–5), and Paul Krugman (1993). These authors suggested that, because a higher trade level would lead to greater specialization, it would also lead to lower synchronization of shocks.16 Their view that specialization works against common currencies and that diversification of the economy works in favor of it, goes back to Kenen (1969).

In the growing literature on endogenous optimum currency area criteria, it is common to assume that the debate about whether trade raises cyclical correlations as Frankel and Rose (1998) claim,17 or lowers it as Krugman and Eichengreen claim, turns on whether the trade is primarily intraindustry or interindustry. Fidrmuc (2001, 2004) and Imbs (2003) extend the econometric estimation to take specific account of intraindustry trade as a determinant of cyclical correlation. The reasoning is that shocks in a world of interindustry trade take the form of shifts from one industry to another: one country’s loss is the other’s gain, yielding negative correlations. In a world of intra-industry trade, industry shifts are assumed to affect all the product varieties produced in different countries, thus yielding positive correlations. Tests by these authors seem to confirm the argument that intra-industry specialization is in fact the source of positive cyclical correlations, driving out total bilateral trade as an explanatory factor.

I believe that several things may be wrong with this argument. First, a large share of trade today is in inputs and intermediate products. Think of iron ore that is made into steel, which is in turn made into machinery parts, which are made into the finished machine tool that is used in the production of something else. It is a similar story with computers. A positive shock at one point in the chain of value-added in one country will tend to have positive spillover effects at the other points along the chain in other countries (e.g., Kose and Yi, 2001). Thus, trade in inputs and intermediate products gives rise to positive correlations and yet may be recorded as interindustry trade. Nevertheless, this is ultimately an empirical question, as Calderón, Chong, and Stein (2003) point out, and empirical studies such as Fidrmuc (2004) do indeed seem to find that intraindustry trade links are associated with cyclical correlation and interindustry trade links are not.

The second objection concerns supply versus demand shocks and may be harder to reject. It is worth stepping back for a moment to realize that we should be more interested in demand shocks than supply shocks. Recall that the point of the whole exercise is to see how much countries are giving up when they abandon independent/discretionary monetary policy. Discretionary monetary policy is not much good at addressing supply shocks anyway. Therefore, it does not much matter whether a country shares them with its neighbors.18 Discretionary monetary policy is more useful in addressing demand shocks. For these, bilateral transmission could come from either intraindustry trade or interindustry trade. A shortfall in demand, originating for example in a decrease in velocity or a fall in investment, will be transmitted to trading partners as a reduction in demand for imports of all sorts—varieties that are in the same industry as well as products in different industries. If the partners are unable to respond to shocks because they have given up their monetary independence, this will be less of a hardship to the extent that the common monetary policy is determined by a set of countries all experiencing the common loss in demand. But the distinction between intraindustry and interindustry trade may be less useful than often supposed. The more important question may be demand shocks versus supply shocks.19

“Convergence”: Are CEE Countries Ready to Join the EMU?

I believe that whether Europe experiences a large asymmetric shock within the next few decades will largely determine whether the EMU proves on net to benefit its members. The early 1990s saw a German spending boom associated with reunification, which implied a temporary real appreciation of the mark against the pound and other European currencies; fortunately, monetary integration had not proceeded so far as to make such a realignment impossible. Today it would not be possible.

One cannot predict the important shocks of the future. Their unpredictability is what makes them shocks. Possible shocks that would hit the eastern half of the continent more than the western half include repetitions of past instability in the former Soviet Union, which might particularly affect the Baltics, or of financial crises such as that experienced by the Czech Republic in the late 1990s. If such a shock occurs in one of the countries that are planning to join the EMU, even the other candidates lucky enough to escape the direct impact would likely be hit indirectly. The reason is the contagion phenomenon that has been so evident in past crises: Western Europe, 1992–93; Latin America in 1982, 1994–95, and 1998–99; East Asia. 1997–98; and other emerging markets subsequent to the Russian default of 1998.

Some of the CEE countries probably do not meet the optimum currency area criteria as well currently as do the existing members of the EMU. An argument for joining anyway is the endogeneity of the optimum currency area criteria. By adopting the euro, these countries will eventually promote trade with the rest of the euro area and increase the cyclical correlation. The increased trade will in turn further increase the advantages of a common currency, while the increased correlation will reduce the disadvantages of a common currency. Thus, the CEE countries may eventually qualify ex post even if they do not ex ante. On the other hand, the risks of asymmetric shocks in the meantime are substantial.

Trade links with the euro area have risen over the last decade anyway, as trade that had for half a century or more been distorted by enforced dependence on the Soviet Union reverts to more natural patterns. CEE countries now trade roughly as much with the euro area as the countries of the euro area trade with each other (Backe and Thimann, 2004, Figures 2.1 and 2.2; Boeri, 2005, Figure 1; Darvas and Szapñry, 2004). Trade with Western Europe can be expected to rise quite a bit more in reaction to the formal accession of the eight to the EU in 2004. Along with trade links, cyclical correlations can be expected to rise further. The shift in trade patterns and correlations will be drawn out over time because the effects of membership in an FTA or a common market develop with long lags.20 Therefore, the risk-averse strategy would be to wait five years or more for EU membership to take hold and integration to proceed further. By then, the convergence would have proceeded far enough that asymmetric shocks and contagion would pose less of a danger. Another factor working in favor of waiting is the opportunity to learn by watching the experiment unfold in the euro area (and—more unpredictably—among any additional joiners).

So far, this paper has interpreted “convergence” to refer to the synchronization of business cycles—that is, correlation of unanticipated shocks—that would reduce the need for individual countries to retain their own monetary policy. But another interpretation of convergence would look for narrowing in the gaps of productivity, or real income, between East and West. Levels of productivity and real income in the East are substantially below those in the West, but a reasonable expectation is that the former will increase more rapidly than the latter. It is not that low-income countries can always expect to catch up with rich countries (unconditional convergence). Rather, low-income countries that have the necessary preconditions or that become successfully integrated with the economies of richer neighbors tend to close the gap over time. In particular, economic integration with more advanced neighbors seems to accelerate productivity growth. For example, Coe and Helpman (1995) argue that intermediate goods embody technology and find that countries’ productivity levels are positively affected by measures of foreign partners’ research and development weighted by bilateral import shares.21 Indeed, one-seventh of the gap in per capita incomes vis-à-vis the EU15 has already been closed over the past 10 years. (Relative to EU income, incomes among the CEE eight had an unweighted average of 41 percent in 1993 but had risen to 48 percent in 2002.)22

The Balassa-Samuelson effect predicts that as a consequence of rapid productivity growth in Central and Eastern Europe, these countries will experience increases in their relative prices of nontraded goods and services, such as housing, and thus will experience real appreciation of their currencies. In this respect they will follow in the footsteps of Greece, which has experienced a trend of real appreciation over the past 10 years. Some estimates for CEE countries predict real appreciation of about 2 percent per year.23 Most of the eight have already experienced substantial real appreciation over the past 5 or 10 years, as Table 2 records. If they retain their own currencies and this trend in relative prices continues, it will probably show up partly as nominal appreciation of their currencies. But if they have joined the EMU before income convergence is achieved, the trend will show up entirely as a higher inflation rate in the new member countries than in the rest of the euro area. One issue is the parity at which they enter the EMU. But even assuming these countries are able to adjust their parities to the right competitive level before entering the EMU, or more precisely, before entering the two-year probationary period, the Balassa-Samuelson effect predicts that there will continue to be upward pressure on their price levels thereafter. Once the countries have joined the EMU, there will be no good way to address this upward pressure on their price levels.

Table 2.

Trends of Real Appreciation Among Countries Acceding to the EU over the Past Decade

Note: Calculated using annual data, unless otherwise noted: [1] Annual average of quarterly data; [2] Annual average of monthly data. CEE = Central and Eastern European; CHF = Swiss franc; CPI = Consumer Price Index; DEM = deutsche mark; EU = European Union; IMF = International Monetary Fund; NEER = Nominal effective exchange rate; PPI = Producer Price Index; SK = Slovak koruny; USD = United States dollar.

Table 2.

Trends of Real Appreciation Among Countries Acceding to the EU over the Past Decade

Note: Calculated using annual data, unless otherwise noted: [1] Annual average of quarterly data; [2] Annual average of monthly data. CEE = Central and Eastern European; CHF = Swiss franc; CPI = Consumer Price Index; DEM = deutsche mark; EU = European Union; IMF = International Monetary Fund; NEER = Nominal effective exchange rate; PPI = Producer Price Index; SK = Slovak koruny; USD = United States dollar.

Full convergence of real incomes can be expected to take a very long time. No country should be expected to wait this long to join the EMU, if joining is otherwise attractive. But the absence of income convergence suggests that the European Central Bank would have to take into account the Balassa-Samuelson effect. After the expansion of the EMU, it might be appropriate for the European Central Bank (ECB) to allow countries that experience the Balassa-Samuelson effect to run inflation rates above the unionwide target, to avoid imposing deflation on the original core members.

The bottom line is that waiting for substantial convergence of per capita incomes will take too long. But the CEE countries will meet the optimum currency area criteria for joining the euro area better in the future than they do today. Waiting five years for further trade integration and cyclical correlation between the CEE countries and the EMU would not necessarily be waiting too long.

According to the terms of the Maastricht Treaty, the new members must wait anyway, until certain conditions are met. The four “convergence criteria” that aspirants must legally satisfy before joining the EMU are price stability, budget discipline, exchange rate stability, and convergence of long-term interest rates. They are sometimes called “nominal convergence,” as opposed to the “real convergence” that has been the subject of this paper.

The four Maastricht conditions, particularly the fiscal criterion, are not very closely based on international monetary theory.24 The problem is not that different criteria might be relevant for transition countries than for their more advanced neighbors, or even just that different details of the specification are appropriate now that the EMU already exists. Rather, the four convergence criteria of the Maastricht Treaty never did correspond well to the standard criteria that textbooks lay out for optimum currency areas: trade integration, cyclical synchronization, labor mobility, and fiscal cushions. New research on currency boards and dollarization in the 1990s produced a third set of criteria that were supposed to qualify a country to be able to live without its own currency in an age of financial integration. Briefly, these criteria include popular support for integration with neighbors; high private use of the euro or the dollar; high pass-through; and, especially, a strong need to import monetary stability arising, for example, from a history of hyperinflation. But the Maastricht convergence criteria do not correspond to this 1990s list much better than they correspond to the traditional optimum currency area list.

Nevertheless, the delay that the Maastricht criteria would impose may not be too different from the sort of delay that is in any case desirable. It depends how strictly the criteria are interpreted.25

Heroes in Greek mythology were always being assigned seemingly irrelevant errands. King Aegeus placed his sword and sandals under a large stone and left behind instructions that only when his son Theseus was able to lift the stone and take the sword and sandals should he come to Athens to claim his kingdom. Theseus’ mother waited before even telling him to try to lift the stone, until he was old enough and strong enough to do it easily. Perhaps the Maastricht criteria will serve as an analogous sort of hurdle—a test that is less important in terms of its intrinsic relevance than as a means of delaying entry until the cyclical patterns have converged just a bit more.

What event caused many EU countries to wait on adopting the euro?

What has facilitated the adoption of the euro among EU countries?

What is the most likely reason some former communist nations refused to adopt the euro as a currency?

What event has been an important cause of the surge of immigration to European countries?