Journal of Empirical Economics Vol. 2, No. 4, 2014, 229-244

The Economic Impact Analysis of the Euro Currency on Twelve Member Countries of the European Union Oluwole Owoye 1, Olugbenga A. Onafowora 2 Abstract This paper examines the economic impact of the euro (€) on 12 member countries of the European Union (EU) who adopted the euro as their common currency on January 1, 2002. Prior to the period, the Maastricht Treaty (MT) of 1992 and the Stability and Growth Pact (SGP) of 1997 laid down the convergence criteria with respect to price stability, convergence of interest rates, exchange rate stability, and budgetary balance that must be met by member countries. We examine euro’s impact on key macroeconomic variables such as inflation, unemployment and long term interest rates, exchange rate, government budget deficits and debts as percent of GDP, which are core of the convergence criteria by testing the null hypotheses of no differences between the mean values of these variables before (1985-1998) and after (1999-2012) the adoption of the euro against the alternative hypotheses that differences exist between the means. The overwhelming conclusion arrived at based on our empirical results is that euro’s economic impact on these 12 Euro zone countries has been more positive than negative. Keywords: Economic impact, Euro zone, European Union, Euro, convergence criteria, currency union JEL Classification: E58, E62, F15, F31, F33, F42, H62. 1. Introduction Common currency union among a group of countries is a multifaceted arrangement that allows the affected countries to reduce, among other things, the uncertainty associated with the volatility of exchanges rates. Unquestionably, the ex-post benefits inherent in a currency are among the factors underlying the creation of the Economic and Monetary Union (EMU), which over the past couple of years led 17 countries to abandon their currencies to adopt a new currency – the euro (€) – to form the Euro zone or euro area. Six other countries who are not official members of the Euro zone are either using or are in the process of using the Euro. For example, Monaco, San Marino, and Vatican City have concluded formal agreements with the European Union (EU) to use the euro as their only official currency; Andorra negotiated similar agreement that allowed it to issue the euro, effective from July 1, 2013, while Kosovo and Montenegro have adopted the euro unilaterally. In a paper in the European Economic Review, Frankel and Rose (1997) asked: “Is EMU more justifiable ex-post than ex-ante?” They answered in the affirmative even though the euro’s launch date was not until January 1999 and its physical euro cash was not introduced until January 2002. In a recent paper in Centre Piece in which Silva and Tenreyro (2010) summarized their findings from another paper in Annual Review of Economics, they asked a related question: “Has the euro increased trade? Short answer: No.” They answered no and argued that there is little evidence that the creation of the euro has had an effect on trade between the original member states because the economies of the Euro zone countries were already deeply integrated before the euro currency was introduced. A decade-and-a-half of ex-post data now exist for

1 2

Professor of Economics Department of Social Sciences/Economics Western Connecticut State University Danbury Professor of Economics Department of Economics Susquehanna University Selinsgrove

© 2014 Research Academy of Social Sciences http://www.rassweb.com

229

O. Owoye & O. A. Onafowora. 12of the 17 Euro zone3economies –Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, Luxemburg, the Netherlands, Portugal, and Spain– that can be compared with ex-ante data to further provide clearer insight into these fundamentally important questions. To date, empirical studies of this issue have focused exclusively on common currency’s or euro’s trade effects in the past two or more decades [see for example, Frankel and Rose (1997)]. Indubitably, the stability of exchange rates among a group of countries implies more bilateral trade among the group, but the exclusive focus on euro’s trade effects neglects the importance of the other key convergence criteria, which policymakers in other EU countries must consider as they weigh the benefits and costs of abandoning their national currencies for the euro. For example, if policymakers know with a degree of certainty that inflation, unemployment and long-term interest rates will fall and that the standard of living or social welfare will increase after joining the EMU, they are more likely to favor the adoption of the euro. This paper is much broader in scope: its focus is not exclusively on euro and bilateral trade nexus, which many studies have examined using the gravity model.4While most studies conclude that a common currency such as the euro boosts trade, however, there exists areas of contention: the degree of currency union’s or euro’s trade effects, omitted variables bias, reverse causality or endogeneity, linearity or nonlinearity, and the choice of instrumental variables, which call into question the validity of these empirical findings. The magnitude of a currency union’s or euro’s trade effects has attracted the most criticisms due to econometric-specification errors in estimating the gravity equation. Baldwinet al. (2005) and Baldwin2006) dubbed these errors as ‘gold’, ‘silver’, and ‘bronze’ medal econometric errors. Baldwin (2006) further argued that the true gravity equation is nonlinear and that its linear estimations may be problematic. Other researchers have argued that it is difficult, if not impossible, to quantify common currency in any regression analysis. The plethora of econometric errors and the mixed results with respect to the magnitude of euro’s trade effects corroborate Artis (2006) assertion that “econometric analysis is not likely to be reliable in settling this matter.” Given Artis’ (2006) cautionary note, it is not the intention of this paper to question the magnitude of the Rose effect or to trivialize the various econometric pitfalls of those studies that examined the effects of a currency union on trade because each empirical study sheds additional light on the complexities of a currency union and the way forward in estimating its economic impact. As a contribution to the literature, this paper employs the less controversial economic impact analysis methodology5 to examine euro’s effects on each of the convergence criteria and other relevant macroeconomic performance indicators in each country. The decisive criterion for using the ex-ante and ex-post methodology, as opposed to the gravity model or any other methodology, is because it avoids some of the econometric pitfalls (such as omitted variables, model misspecification, and excessive use of dummy and instrumental variables) of previous studies – see Rose (2000) and Baldwin et al.(2008).More importantly, the conclusions arrived at based on our empirical analysis should prove useful to policymakers who must make the critical decision about whether or not their countries should join the euro. After all, the legislative policymaking bodies in these countries are not populated by high-powered econometricians and statisticians, but they can make better informed decisions if provided with logical and straightforward empirical evidence on the ex-ante and expost behavioral tendencies of any macroeconomic policy variable. 3

These were the 12 of the 15 EMU countries that fulfilled the convergence criteria by the time the Euro was launched on January 1, 2002. 4

According to the simple gravity equation, trade between a pair of countries is a positive function of their combined GDPs and a negative function of the distance between them. For a detailed theoretical foundation for the gravity model as to its applications to a wide variety of goods and factors moving over regional and national borders under differing circumstances, see Anderson (1979), and Anderson and Van Wincoop (2003). 5 The economic impact analysis allows one to measure or estimate the level of economic activity that occurred at a given time with the project or policy, and then calculate the difference from what would otherwise be expected if the project or policy did not occur. In other words, this analysis can be done either before or after the fact (i.e. ex-ante or ex-post) based on the actual data generated in both periods, thus one can avoid the use of proxy, dummy, and instrumental variables, which may lead to spurious estimated results and erroneous policy inferences. 230

Journal of Empirical Economics The 1992 Maastricht Treaty established the convergence criteria with respect to inflation rates, interest rates, government deficits and debts, and the stability of exchange rates that member states are required to fulfill in order to join the EMU. The exchange rates were irrevocably fixed for the 11 member states who adopted the euro on January 1, 1999. Therefore, one can argue that the exclusive focus on euro’s trade effects amount to a focus on just one of the five convergence criteria. In sharp contrast to the trade-specific crosscountry/panel data approach employed by previous studies, this paper uses a different statistical methodology to examine euro’s effects on each of the original 12 countries that adopted euro with emphasis on the core variables of the convergence criteria, and this method can be used to test for consistency and continued adherence to the convergence criteria after two or more decades of euroization. The overwhelming conclusion reached based on our empirical results is that euro’s economic impact on these 12 Euro zone countries has been more positive than negative. The rest of the paper is organized as follows. In Section II, we provide a brief historical timeline of EU’s economic and monetary union to help put into perspective the data used in the economic impact analysis. Section III provides a synopsis of the theoretical and empirical literature of currency unions. Section IV discusses the methodology and the data used in the empirical analysis. Section V presents the estimated results, while Section VI concludes with some policy implications. 2. A Brief Historical Timeline of the Economic and Monetary Union In this section, we provide a brief historical timeline of the Economic and Monetary Union (EMU) in order to put in proper perspective the euro’s economic impact on the original12 countries based on the data derived from the euroization of these economies since its launch. We argue that country specific analysis provides a much better insight because the benefits and costs of membership in a currency union differ across member countries.6 As research scholars who are familiar with the historical timeline of the European Union may recall, the Werner Report of 1970 laid out a three-stage approach to EMU, but its implementation was delayed because of the difficult macroeconomic conditions that prevailed during the early 1970s. The European Monetary System which consisted of an Exchange Rate Mechanism (ERM) and the European Currency Unit (ECU) was not launched until 1978; and it was not until 1989 that the Jacques Delors Report provided a clear road map of the three stages to the EMU. The first stage of EMU, which spanned between July 1990 and December 1993, called for closer cooperation between central banks and the coordination of macroeconomic policies, the liberalization of capital movements, and the narrowing of the exchange rate band (known as ERM). During this stage, the Maastricht Treaty (MT), which laid out the timetable of EMU and the five convergence criteria was signed in 1992. The second stage of EMU spanned between January 1994 and December 1998; and during this period, European Monetary Institute (EMI) of the European Monetary System was created. Among the functions of the EMI was the coordination of national monetary policies with the aim of ensuring price level stability and the promotion of efforts towards convergence. The 1995 Madrid EU summit named the new single currency for the member states as ‘the euro’ and set the scenario for the third stage of EMU with a three-year transition period between the introduction of the new currency and the launch of euro cash. The 1997 Amsterdam EU summit yielded the Stability and Growth Pact (SGP) to ensure that member states maintain budgetary discipline within the EMU and a revised exchange rate mechanism (known as ERM II) that will link the euro and the currencies of non-participating member states. In May 1998, the European Council and the EMI agreed to launch the third stage of EMU on January 1, 1999 and published their convergence reports which indicated that 11 of the 15 member states – Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Portugal, and Spain – met the criteria and could therefore adopt the single currency. In addition, the European Council established the 6

By comparing the ex-ante and ex-post data for these countries, we can answer the empirical questions posed by Frankel and Rose (1997), Micco et al. (2003), and Silva and Tenreyro (2010) with supportive statistical evidence. 231

O. Owoye & O. A. Onafowora. European Central Bank (ECB) to replace the European Monetary Institute, effective June 1, 1998; and by this effective date, the ECB started operation with the mandate to conduct and implement monetary policy for the euro area with the primary objective of price level stability. By December 31, 1998, the exchange rates between the euro and the currencies of the member states that adopted the euro were irrevocably fixed effective January 1, 1999. The third stage of the EMU started with the launch of the physical euro currency on January 1, 1999. As we discussed earlier, the Madrid EU summit agreement provided for a three-year transition period between the introduction of the new euro cash. In essence, during the transition period – between January 1, 1999 and January 1, 2002 – the euro existed as a virtual currency. At the beginning of 2001, Greece complied with the MT criteria and this brought the initial number of countries that adopted the euro to 12 at its launch in 2002. 3. Literature Review The theoretical analyses on the importance of national currency areas or a single currency area for Western Europe were articulated in the early studies by Meade (1955, 1957) and Scitovsky (1958). Meade (1957) argued that the conditions for a common currency in Western Europe did not exist because of the lack of labor mobility; therefore, a system of flexible exchange rates would be the more effective channel with which to promote balance-of-payments equilibrium and internal stability. In contrast, Scitovsky (1958) argued in favor of a single currency area in Western Europe because he believed that it would induce a greater degree of capital mobility if the countries in Western Europe took steps to make labor more mobile and facilitate supranational employments policies. Mundell’s (1960) study of the monetary dynamics of international adjustment under fixed and flexible exchange rates elaborated on the earlier studies of Friedman (1953), Lutz (1954), and Meade (1955), which highlighted the importance of flexible exchange rates. Furthermore, Mundell’s (1961) theoretical framework about the concept of optimum currency area (OCA) suggested the conditions (similarity of the economic shocks that member countries experience, wage and price flexibility, and the mobility of capital and labor) that can reduce the cost of relinquishing monetary independence for a currency union. Also, this was an attempt to reconcile the apparent contradiction between Meade’s (1957) and Scitovsky’s (1958) studies and to reduce it to an empirical rather than a theoretical question by arguing that “neither writer disputes that the optimum currency area is the region – defined in terms of internal factor mobility and external factor immobility – but there is an implicit difference in views on the precise degree of factor mobility required to delineate a region. The question thus reduces to whether or not Western Europe can be considered a single region, and this is essentially the empirical question.” In contribution to the debate on common currency union, Frankel (1997) and Frankel and Rose (1998) argued that business cycles may become more synchronized across countries and thus can been do genous because increased trade integration will lead to increased business cycle correlation; and that currency unions tend to lower the opportunity cost of national monetary policy. Along the same line of argument, Alesina and Barro (2000) pointed out that currency union may be an efficient institutional arrangement to handle the problem of credibility. The empirical investigation of the effect of currency unions on trade became an issue for empirical debate when Rose (2000) used the gravity model and found a large positive effect of a currency union on trade. Rose’s finding corroborated the earlier findings of McCallum (1995) and Helliwel (1998) that used the gravity model to show the “border effect” or “home bias” on trade between Canadian provinces with the same currency and US states with a different currency. Both studies suggested that trade between two Canadian provinces was 10 to 20 times larger than trade between Canadian provinces and the US states. In response to the criticisms of Rose’s (2000) findings, Rose and van Wincoop (2001) employed Anderson and van Wincoop’s (2001) gravity equation and found the trade-creating effects of currency union

232

Journal of Empirical Economics to be 58 percent7 among the then 11 Euroland members. Their results were consistent with those of Frankel and Rose (2000), Glitch and Rose (2001), and Rose (2004).Despite these findings, Rose’s (2000) initial claim that “these effects are statistically significant and imply that two countries that share the same currency trade three times as much as they would with different currencies” continued to be subjected to rigorous examination and criticisms by other scholars. Persson (2001) considered Rose’s (2000) estimates to be biased due to nonlinearity and non-random selection. To address these problems, he proposed the nonparametric matching technique that allowed for systematic selection into currency unions as well as nonlinear effects of trading costs on trade. Using the matching techniques in which he grouped countries into treatment and control groups, he produced a “propensity score” for each country pair and obtained considerably smaller treatment effects of a common currency than the 200 plus percent that Rose (2000) obtained. According to Persson (2001), “my preferred point estimates range from 13 to 65 percent, but with large enough standard errors that they are not significantly different from zero.” Kenen (2002) used a different matching technique and obtained very different regression estimates. Devereux and Lane (2003) considered currency union as an extreme form of exchange rate stabilization since nations tend to stabilize their bilateral exchange rates with other nations with whom they trade a lot. According to Tenreyro (2001), the problem of endogenous selection into a currency union and the omitted factors could strengthen trade relations and increase the propensity of joining currency unions, thus, this could lead to a positive bias in the OLS estimates. To overcome the endogeneity problem, Tenreyro (2001) estimated the trade equation with the joint decision to participate in a currency union and found that currency union affects trade by 50 percent; however, the effect was statistically insignificant. In other related studies, Estevadeordal et al. (2003) used the experience of countries who participated in the gold standard during the 1870-1939; and they found that common participation in the gold standard increased trade by 34 to 72 percent. López-Córdova and Meissner (2003) examined the same gold standard participation between 1870 and 1910; and they concluded that the gold standard increased trade by 60 percent. Furthermore, they found that currency unions double trade, which corroborates the earlier finding of Glick and Rose (2001). Pakko and Wall (2001) used Rose’s (2000) dataset; and they preserved the bidirectional trade flows, which allowed them to introduce two direction-specific dummies. In the absence of these pair dummies, they found estimates similar to the Rose effect; however, with both dummies included, they found negative coefficient estimate that was not significantly different from zero. De Souza (2002) used the basic gravity model and added a time trend for 15 EU countries and found no evidence of a significant Rose effect. Similarly, Anderton et al. (2002) applied the three-stage least squares methods to estimate the import demand functions and their findings was consistent with De Souza’s. In contrast, Blun and Klaasen (2002) updated the gravity model by using a dynamic fixed effect estimator and found that “the euro has significantly increased trade, with an effect of 4% in the first year,” and with the long-run effect projected to be 40 percent. DeNardis and Vicarelli (2003) took a different approach with the aim of controlling for reverse causality, and they found that trade increased by about 10 percent and 20 percent in the short-run and longrun, respectively. Flam and Nordström (2003) used the direction-specific bilateral trade flow as suggested by the gravity model and employed the bilateral export data rather than the average of bilateral exports and imports. This allowed the authors to explore the issue that concerns non-Eurozone nations: that the euro puts their exports at a disadvantage in Euroland. Using three non-Eurozone and eight other rich countries as the control group, they found the Rose effect to be about 15 percent more trade. As one of many studies that estimated the early effect of the EMU on trade, Micco,et al.(2003) used the gravity model based on panel data on bilateral trade for 22 countries from 1992 through 2002. To address some of the econometric problems attributed to Rose’s (2000) study, they added country pair fixed effect, a dummy for membership in a free trade agreements (FTA) to account for a country that once experienced 7

See Table 2, p. 389 of Rose and van Wincoop (2001). 233

O. Owoye & O. A. Onafowora. increase in trade as a member of an FTA but now in a currency union, an EU dummy as a way to recognize that the impact on trade on membership in the EU may be larger than that of other shallow FTA, and an EU Trend defined as the EU dummy multiplied by the year since the beginning of their sample. They found “that the effect of EMU on bilateral trade between member countries ranges between 5 and 10 percent, when compared to trade between all other pairs of countries, and between 9 and 20 percent when compared to trade among non-EMU countries…….that monetary union increases trade not just with EMU countries, but also with the rest of the world.” Berger and Nitsch (2005) questioned Micco, et al.’s (2003) estimates on several grounds including the fact that Micco, et al.’s estimates were based on four years of data and trade among EMU members appeared to jump in the 1998 – the last year of the stage three of the EMU– therefore, they argued that Micco, et al. may be unable to separate the effects of monetary integration and non-monetary integration among the Eurozone countries. They further argued that the size of the Rose effect could be quite sensitive to disaggregation by country. Berger and Nitsch added an additional year to Micco, et al.’s data and found an increase in Rose effect. Their finding was consistent with Nitsch (2001). Other studies by Flam and Nordström (2006) and Chintrakarn (2008) also found positive Rose effect of varying degrees. According Frankel (2008), Rose’s (2000) study brought the brigade of studies meant “to shrink the Rose effect – or to make it disappear altogether.” In defense that Rose’s estimates may be spurious, Frankel (2008) laid out the five grounds of skepticism of Rose’s study – time series dimension, omitted variables, causality problems, implausible magnitude of the estimate, and country size. Frankel (2008) assessed the critiques on each ground and concluded that the discrepancy in the magnitude of euro’s effect on trade “might stem from sample size.” As anyone can observe from these studies, their empirical findings are mixed for a host of factors. Studies covered different time periods, employed different data sizes, utilized different econometric methodologies, and used different variables and/or instrumental variables. Are there other less controversial methodologies that can be used to show euro’s economic impact on key performance indicators for the Euro zone countries? This question is similar to the one asked by Micco, et al.(2003): “Does the EMU effect show by looking at the trade data?” 4. Methodology and Data Some research scholars have attributed the econometric flaws in previous studies of the euro’s effect on trade to the lack of time series data of sufficient length for individual countries included in the sample, which may have forced many studies to use cross-country or panel data approach to examine the effects of currency unions and exchange rate volatility on trade. Even, Rose (2000) was cognizant of the potential problems of using cross-country approach to investigating currency unions when he pointed out that “One might imagine that trying to measure the effects of a common currency on trade is purely academic (i.e. trivial) exercise. The only countries that have adopted a common currency of late are the EMU-11, for whom there are necessarily few data. True enough. But there is no reason to rely on before and after differences to estimate the effect of currency unions on trade, just as one need not use time-series variation to discern the effects of exchange rate volatility on trade.” If there had been enough data size, these econometric problems could have been avoided and some studies could have opted for a methodology that relied on ex-ante and ex-post data to estimate currency unions’ effect on trade. It is now almost 15 years since the launch of the euro as single currency, therefore, we compare the exante and ex-post data to estimate euro’s effects on key macroeconomic performance indicators – inflation rates, long-term interest rates, government budget deficits and debts, net exports as well as current account balances, unemployment rates, and per capita GDP in each of the original 12 of the 17 EU member countries who adopted the euro as their common currency on January 1, 2002.In addition, we consider the examination of euro’s impact on the level of unemployment rates, per capita GDP, public social expenditure as a way to gauge the effects of euro on the social welfare in each country in the sample. 234

Journal of Empirical Economics In the attempt to simplify the econometric methodology and provide reliable evidence, some studies estimated the euro’s effect by employing the “difference-in-difference” (D-in-D) approach. This approach is based on the comparison of relevant economic variables for the periods before (ex-ante) and after (ex-post) the euro was launched for two groups of countries. Using the D-in-D approach, those countries that joined euro during the period under observation are regarded as the “treatment group,” and the comparable group of countries who did not join the euro is regarded as the “control group.” The effect of the euro is then estimated as the difference in the changes from the ex-ante to ex-post periods for the two groups of countries (see Tenreyro, 2003).This difference-in-difference method is commonly used in labor economics and health economics.8The advantage of applying this technique in examining euro’s effect is that it avoids the unnecessary use of assumptions and instruments than can lead to spurious estimates. We implement the differences-in-means approach (D-in-M) rather than use the difference-in-difference estimation method because the focus is not on comparing two – “treatment and control”– groups of countries, which many studies have done. Rather, our focus is on euro’s economic impact on each of the original group of 12 countries in our sample. In other words, we examine each country in the sample on two different occasions – before and after the implementation of the euro currency. This is a case in which we assumed that all the relevant macroeconomic variables in each country are measured repeatedly at two different points in time [period 1 (before) and period 2 (after)]. The main assumption here is that the mean values of these macroeconomic variables may be different due to euro’s impact. To estimate euro’s effects on each country in the sample, we test the following null and research hypotheses:

HO : X B  X A

(1)

HR : X B  X A

where X is a vector of all relevant macroeconomic performance indicators, X B and

XA

are their ex-

ante(B) and ex-post (A)mean values in each country in the sample. It is important to note that X A ' s are the past decade-and-a-half of euroized mean values of the relevant variables. According to Levin, Fox and Forde (2010), testing for differences in means for the sample or variable measured twice is just one of several applications of the t test for dependent samples/variables whose computed t-values can be expressed in general form as:

t 

 D2 N

XB  XA 

X

B

(2)  XA

2

N 1

where D is the difference (obtained by subtracting the ex-post from ex-ante euro raw-data), N is the observed number of years for both periods,

D N

8

2

  X B  X A  is the standard deviation for the distribution of before– 2

The difference-in-difference approach used in other studies is appropriate when comparing the “treatment” and “control” groups, that is, two independent samples. Since we look at the same variables in two different periods and not comparing them with other “control” groups, we consider the difference-in-means as the appropriate method for measuring euro’s economic impact. For detailed procedures in conducting both (D-in-D and D-in-M) tests, see any standard statistics textbook. 235

O. Owoye & O. A. Onafowora.

D after difference values or scores,

2

N



X

B

 XA

2

N 1

is the standard error of the difference between the

mean values, and N – 1 equals the degrees of freedom.9 We reject the null hypotheses of no differences in the mean values if the estimatedt-value given by equation (2) is greater than its table value. If one fails to reject the null hypotheses, then one can conclude that the euro had no statistically significant economic impact on those variables. If one rejects the null hypotheses in favor of the research hypotheses, then one can conclude that there are statistically significant differences between the ex-ante and ex-post mean values. The data used in this study are obtained from the World Bank Economic Indicators from various years and the OECD Annual National Accounts and the Annual Macroeconomic Database of the European Commission, with our range between 1985 and 2012 – see http://stats.oecd.org. 5. Data Evidence and Estimated Results Some studies have asked whether EMU’s (or euro’s) effect should be visible by looking at the trade data– see Micco et al. (2003). This assertion also applies to other macroeconomic data. Other studies have also asked if joining the EMU is more justifiable ex-post than ex-ante or whether the euro has increased trade [see Frankel and Rose (1997), Bun and Klaasen (2002), Silva and Tenreyro (2010)]. To answer these questions, we provide two visual diagrams (Figures 1 and 2) to show the effects of euro as these countries struggle to comply with the fiscal synchronization outlined by the Maastricht Treaty-Convergence (MTC) criteria. During stages one and two of EMU, only two countries (Finland and Luxembourg) had a positive budget position. As Figure 1 shows, those countries with budget deficits above the three percent benchmark made concerted efforts toward compliance. Similarly, for the other threshold: the 60 percent debt as percent of GDP, Figure 2 shows that about half of these countries had debts that were consistently below the 60 percent threshold; and the rest of the countries (except for Belgium, Greece, and Italy) had their debts hovering around the 60 percent target. The observed divergence from the three percent and the 60 percent targets occurred in10 of the 12 countries in response to the 2008-2009 financial perturbations that required these countries to use their fiscal space. This was a policy initiative that the World Bank and the International Monetary Fund recommended for countries worldwide in the wake of the 2008-2009 financial crises. In Table 1, we provide additional visual evidence by showing other channels or macroeconomic variables through which we examine euro’s economic impact on the original 12 euro adopters. We report the ex-ante and ex-post mean values of these variables as other ways with which to discern euro’s economic impact. As we can see from these mean values, inflation rates fell, ex-post, in nearly all countries except for Luxembourg and the Netherlands. Similarly, long-term interest rates fell in all countries since the adoption on the euro. As for trade, we also observed that net exports and current account balances improved in some countries and worsened in others. From socio-economic perspective, one can see that unemployment rates declined in all countries, post-euro period, except for Austria, Luxembourg, and Portugal. During the post-euro period, the growth in industrial production declined in all countries, except for Austria, Belgium, Germany, and the Netherlands. Similarly, GDP growth rates declined in nearly all countries, while public social expenditures increased in nearly in all countries, except for Finland, Ireland, and the Netherlands; but more importantly, per capita income increased in all countries with no exception. Given the observed increases in per capita income since 1999, one can safely conclude that the euroization of these economies contributed to the improvement in the standard of living in Euro zone countries.

9

This is a straightforward statistical methodology commonly used by researchers in the social sciences. While this simple statistical technique may put us in any of the medal categories for its oddities, its importance lies in the fact that it adds more to the knowledge of euro’s impact on the original 12 Euro zone countries. 236

Journal of Empirical Economics Figure 1: Government Budget Deficits as Percent of GDP 10 5 AUT BEL FIN FRA GER GRC IRE ITA LUX NED PRT

-5

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

0

-10 -15 -20 -25 -30 -35

Figure 2: Government Debts as Percent of GDP 180 160 AUT 140

BEL FIN

120

FRA GER

100

GRC IRE

80

ITA LUX

60

NED PRT

40

SPA MTC

20

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

0

237

O. Owoye & O. A. Onafowora. Table 1: Euro’s Economic Impact on the Original 12 Adopters

1

Ex-Ante and Austria Belgium Finland Ex-Post Mean Values IRB 2.45 2.31 3.19 IRA 2.01 2.19 1.87

France 2.57 1.76

Germany 2.24 1.57

Greece 13.89 3.43

Ireland Italy 2.86 2.75

Luxembourg

Netherlands Portugal

Spain

5.07 2.32

2.13 2.43

1.84 2.11

8.63 2.79

5.31 2.89

2

LIRB LIRA

7.30 4.09

7.95 4.19

------

8.14 4.02

6.77 3.77

------

8.92 5.09

------

------

6.85 3.96

------

10.93 4.57

3

GDBB GDBA

-3.29 -2.09

-6.56 -1.48

4.47 2.29

-3.59 -3.68

-3.26 -2.00

-9.47 -7.12

-3.03 -4.19

-9.13 -3.21

(2.66) (0.68)

(-3.30) (-4.80)

-6.67 -4.05

-4.48 -3.07

4

GDB GDA

49.31 66.54

-----97.55

103.78 37.95

68.54

20.72 68.45

85.88 116.45

-----50.85

98.86 110.57

2.24 10.36

56.30 56.00

54.81 72.91

53.81

5

NETXB NEXTA

-0.61 -0.36

0.41 1.21

0.35 0.51

0.04 -4.06

4.38 14.59

-1.05 -3.05

0.60 3.45

0.53 -0.62

-0.13 -0.37

0.83 3.52

-0.67 -1.82

-1.69 -6.02

6

CABB CABA

-0.88 1.81

3.49 2.09

-0.42 8.11

1.04 3.99

6.03 120.51

0.59 -3.71

2.01 -35.55

------

12.34 36.25

-1.51 -17.41

-1.13 -5.21

7

URB URA

3.54 4.32

8.75 7.71

9.40 8.93

6.83 11.21 10.49 9.23

7.25 8.38

14.11 7.14

9.83 8.45

2.10 4.38

6.25 3.79

6.16 7.42

20.01 13.99

8

YPCB YPCA

20,260 35,250

19,372 32,261

17,180 31,899

17,784 19,352 30,487 32,421

13,050 24,248

14,701 37,336

18,388 29,519

32,215 71,322

18,701 36,391

11,457 21,855

13,834 27,858

9

PSEB PSEA

25.33 27.42

25.33 27.42

27.31 26.31

27.16 30.30

24.52 26.66

16.67 21.42

18.42 17.97

21.03 25.25

20.06 22.08

24.32 21.55

13.74 22.39

20.01 22.35

10 INVB INVA

25.03 22.93

21.04 21.38

22.24 20.28

18.94 19.86

23.09 18.67

21.32 21.59

18.47 20.75

21.29 20.53

21.75 21.49

22.09 19.71

26.91 23.19

22.65 26.34

11 IPGB IPGA

3.40 3.59

1.96 2.87

3.50 1.84

1.69 -0.50

1.74 2.03

1.16 -1.32

9.74 4.67

2.10 -1.08

2.32 2.23

1.26 1.35

3.06 -1.26

2.34 -1.02

7.25 8.23

238

Journal of Empirical Economics 12 YGB YGA

2.67 1.87

2.30 1.63

2.12 2.31

2.18 1.45

2.39 1.37

1.75 1.09

5.69 3.69

2.10 0.49

5.29 3.31

3.03 1.59

3.79 0.73

3.08 2.11

Note: Authors’ calculations IR = inflation rates, LIR = long-term interest rates, GDB = government budget deficits (% of GDP), GD = government debts (% of GDP), NETX = net exports ($ billions), CAB = current account balances ($ billions), UR = unemployment rates, YPC = income per capita (constant prices, $, and PPP) , PSE = public social expenditures (% of GDP), INV = total investment (% of GDP), IPG = industrial production growth (% change over prior year), YG = GDP growth (%); B = before or ex-ante period covering 1985 through1998, and A = after or ex-post period covering1999 through 2012.

From the cursory observation in Figures 1 and 2 as well as Table 1 above, one can conclude that the euro has more positive economic impact in these countries, however, the question is whether or not the observed mean differences between the ex-ante and ex-post euro periods are statistically significant to convince policymakers in Britain, Denmark, Switzerland, and other countries in Central and Eastern Europe to abandon their currencies for the euro.10 To address this issue, we provide the estimated results for the null and research hypotheses in Table 2. The null and research hypotheses reported in the first six rows are germane to the convergence criteria, while those of the last six rows touch somehow on issues that can be regarded as euro’s socio-economic (social welfare) impact in each country. As Table 2 shows, the reductions in inflation rates (IR) show that the observed differences in the mean values in both periods are statistically significant in Belgium, France, Greece, Italy, Portugal, and Spain. Similarly, the decreases in the long-term interest rates (LIR) also show that the differences in means in both periods are statistically significant in all countries except Portugal. In addition, we also reject the null hypotheses of no differences in the mean values of government budget deficits (GBD) as percent of GDP in half of the countries in the sample. These findings corroborate the visual evidence provided in Figure 1. We reject the null hypotheses of equality between the ex-ante and ex-post mean values of government debts (GD) as percent of GDP in seven of the eight countries for which we have available data. The statistical significance of the hypotheses with respect to GBD and GD highlights the importance of benchmarks or criteria and corroborates the movements toward MTC in both Figures 1 and 2. Furthermore, we reject the null hypotheses of no differences in the means of net exports (NETX) in all countries, except for Finland and Italy. For current account balances (CAB), we fail to reject the null hypotheses of no differences in mean values for Belgium and France. Simply put, CAB positions improved in some countries and worsened in others, post-euro adoption. Also, we reject the null hypotheses of no differences in the pre- and post-euro mean values of unemployment rates in 10 of the 12 countries (except for Finland and Portugal). In other words, the observed differences in the means values of UR during both periods are statistically significant. Similarly, we reject the null hypotheses of no differences in the pre- and post-euro mean values of PSE in 10 of the 12 countries (except for Finland and Ireland).

10

Allam (2009) pointed out that countries in Central and Eastern Europe (CEE) are unsure and do not have consistent and target dates for accession into EU and EMU. Shifting the accession dates can be attributed to the inability to fulfill the initial ERM participation requirement, which is one of the convergence or Maastricht criteria. 239

O. Owoye & O. A. Onafowora. Table 2: Null and Research Hypotheses of Euro’s Economic Impact on the Original 12 Adopters

1

2 3

4

5

6

7

8

9

10

Null and Austria Research Hypotheses IRB = IRA IRB ≠ IRA 1.11 LIRB = LIRA LIRB ≠ LIRA GDBB = GDBA GDBB ≠ GDBA

Belgium

Finland

France

Germany

Greece

Ireland

0.28

1.67

2.08**

1.59

6.86*

0.26

Italy

Luxembourg

Netherlands

Portugal

Spain

4.89*

0.52

0.51

4.71*

4.24*

12.26*

9.22*

10.80*

10.38*

9.45*

-------

4.59*

3.86*

-------

8.56*

0.52

7.45*

2.74*

4.21*

2.42*

0.17

1.41

1.43

0.36

6.18*

(7.25*)

(0.86)

1.85**

1.04

GDB = GDA GDB ≠ GDA

11.14*

-------

53.05*

-------

28.84*

6.61*

-------

3.64*

6.07*

0.12

3.18*

-------

NETXB = NETXA NETXB ≠ NEXTA

2.04**

3.59*

0.74

3.44*

6.34*

7.80*

21.34*

1.55

8.11*

9.30*

8.13*

5.76*

CABB = CABA CABB ≠ CABA

3.99*

1.24

4.29*

1.43

2.03**

3.42*

2.78*

3.11*

-------

5.02*

9.19*

5.98*

URB = URA URB ≠ URA

3.28*

2.17**

0.42

4.27*

2.32**

2.36**

5.25*

3.06*

8.53*

4.95*

1.31

4.24*

YPCB = YPCA YPCB ≠ YPCA

31.49*

40.50*

17.55*

36.79*

24.18*

15.51*

24.61*

33.24*

20.71*

27.87*

37.75*

23.85 *

PSEB = PSEA PSEB ≠ PSEA

7.06*

3.14*

0.93

10.66*

3.13*

19.50*

0.31

9.93*

55.54*

3.78*

28.60*

3.88*

INVB = INVA INVB ≠ INVA

8.06*

0.55

1.39

1.38

7.94*

0.24

1.08

1.54

0.35

3.89*

3.39*

3.76* 240

Journal of Empirical Economics

11

12

IPGB = IPGA IPGB ≠ IPGA

0.11

0.47

0.71

1.45

0.13

1.55

1.71

1,47

1.25

0.07

2.77*

1.54

YGB = YGA YGB ≠ YGA

1.25

1.01

0.12

1.12

1.21

0.43

0.99

2.15**

1.88**

1.89**

3.01*

1.12

Note: The numbers in columns 3 to 14 are the estimated t-values, * = significant at the 1% level, and ** = significant at the 5% level. IR = inflation rates, LIR = long-term interest rates, GDB = government budget deficits (% of GDP), GD = government debts (% of GDP), NETX = net exports, CAB = current account balances, UR = unemployment rates, YPC = income per capita, PSE = public social expenditures (% of GDP), INV = total investment (% of GDP), IPG = industrial production growth (% change over prior year), YG = GDP growth (%); B = before or ex-ante period covering 1985 through1998, and A = after or ex-post period covering1999 through 2012.

6. Conclusions and Implications Until now, almost all studies focused on euro’s trade effects leaving many unexplored dimensions of euro’s economic impact. This paper employs a less controversial methodology – the differences-in-means approach to examine many unexplored dimensions– inflation rates, long-term interest rates, government budget deficits and debts as percent of GDP, net exports and current account balances, unemployment rates, per capita income, public social expenditures, total investment, and the growth in industrial production and output – of euro’s economic impact. Some of these dimensions are the core components of the convergence criteria while the others may be regarded as euro’s socio-economic issues. First, the mean values of inflation rates declined in all but one of the 12 countries during the post-euro period; and based on the difference-in-means methodology, we reject the null hypotheses of no difference in means in half of the countries in the sample. Second, we observed remarkable reductions in the mean values of long-term interest rates in nearly all countries, post-euro era, and the rejection of the null hypotheses of no difference in means in nearly all countries(except Portugal) confirmed the statistical significance of the observed differences in the mean values. Third, as for euro’s impact on net exports and current account balances, we reject the null hypotheses of no difference in the mean values in 10 of the 12 countries. Fourth, the average growth in industrial production and output decreased during the post-euro period. For the mean values of growth in industrial production, we fail to reject the null hypotheses in all countries except for Portugal, while for output growth; we reject the null hypotheses for Italy, Luxembourg, the Netherlands, and Portugal. Also, the mean values of unemployment rates decreased during the same period, and we reject the null hypotheses of no difference in the mean values in all countries except for Finland and Portugal.

241

O. Owoye & O. A. Onafowora. Fifth, we observed remarkable increases in the mean values of public social expenditures and per capita incomes in these 12 countries since the adoption of the euro. With respect to the mean values of public social expenditures, we fail to reject the null hypotheses only in Finland and Ireland; and for the mean values of per capita income, we reject the null hypotheses of no differences in all countries in our sample. Interpretively, the statistical significance of the increases in the mean values of per capita income lends credence to the conclusion that the standards of living improved in these 12 Euro zone countries. Finally, many research scholars, particularly Frankel and Rose (1997) asked: Is EMU more justifiable ex-post than ex- ante? If we rely, objectively, on the visual evidence in Figures 1 and 2,which showed the apparent compliance to benchmarks as well as the data evidence reported in Tables 1 and 2,which showed and statistically confirmed the substantial reductions in the mean values of inflation rates, long-term interest rates, unemployment rates, and the improvements in net exports and current account balances, and the sizeable increases in per capita incomes, one can conclude that the euro has positive economic impact in these 12 Euro zone countries. References Alesina, Alberto and Robert J. Barro, 2000. Currency unions. National Bureau of Economic Research Working Paper No. 7927, Cambridge, MA. Anderson, James E., 1979. A theoretical foundation for the gravity equation. American Economic Review69(1): 106-116. Anderson, James E, and Eric van Wincoop, 2003. Gravity with gravitas: A solution to the border puzzle. American Economic Review 93(1): 170-192. Allam, Miriam, 2009. The adoption of the euro in the new EU member states: Repercussions of the financial crisis, EIPASCOPE 2009(1): 27-34. Artis, Michael, 2006. What do we now know about currency unions?” CEPR Discussion Paper No. 5677;and Bank of England Quarterly Bulletin. Baldwin, Richard, Virginia DiNino, Lionel Fontagné, Roberto A. De Santis, and Daria Taglioni, 2008. Study on the impact of the euro on trade and foreign direct investment. Economic Paper 321/May 2008, European Commission, Directorate-General for Economic and Financial Affairs. Baldwin, Richard, 2006. Euro’s trade effects. Working Paper Series, No. 594/March 2006, European Central Bank. Baldwin, Richard, Frauke Skudelny and Daria Taglioni, 2005. Trade effects of the euro: Evidence from sectoral data. Working Paper Series, No. 446/February 2005, European Central Bank. Berger, Helge and Volker Nitsch, 2005. Zooming out: perspective. CESifo Working Paper No. 1435.

The trade effect of the euro in historical

Bitãns, Martinš and Egils Kaužĕns, 2004. Impact of the euro adoption on the economy of Latvia. Working Paper 2/2004 of Latvijas Banka. Bun, Maurice J.G. and Franc J. G. M. Klaasen, 2002. Has the euro increased trade?. Amsterdam, Netherland: University of Amsterdam and Tinbergen Institute Discussion Paper, TI (2002): 108-2. Chintrakarn, Pandej, 2008. Estimating the euro effects on trade with propensity score matching. Review of International Economics 16(1): 186-198. Costa-i-Font, Joan, 2010. Regional single currency effects on bilateral trade with the European union. LEQS Paper No. 26. 242

Journal of Empirical Economics De Nardis, Sergio and Claudio Vicarelli, 2003. Currency unions and trade: The special case of EMU. World Review of Economics 139(4): 625-649. De Souza, Lucio V., 2002. Trade effects of monetary integration in large, mature economies. A Primer on European Union, Kiel Working Paper No. 1137. Devereux, Michael B. and Philip R. Lane, 2003. Understanding bilateral exchange rate volatility. Journal of International Economics 60(1): 109-132. Eicher, Theo S., and Christian Henn, 2009.One money, one market – A revised benchmark. International Monetary Funds Working Paper WP/09/186. Estevadeordal, Antoni, Brian Frantz, and Alan M. Taylor, 2003. The Rise and fall of world trade, 18701939.Quarterly Journal of Economics 118(2): 359-407. Fernández, Francisco de Castro and Pablo Hernández de Cos, 2006. The economic effects of exogenous fiscal shocks in Spain: A SVAR approach. Working Paper Series, No. 647/June, European Central Bank. Flam, Harry and Hakan Nordström, 2003. Trade volume effects of the euro: Aggregate and sectoral estimates. Institute for International Economic Studies, Stockholm. Frankel, Jeffrey A. and Andrew K. Rose, 2002. An estimate of the effect of common currencies on trade and income. Quarterly Journal of Economics 117(2): 437-466. Frankel, Jeffrey A. and Andrew K. Rose, 2000. An estimate of the effect of currency union on trade and growth. National Bureau of Economic Research Working Paper No. 7857, Cambridge, MA. Frankel Jeffrey A. and Andrew K. Rose, 1998. The endogeneity of the optimum currency area criteria. The Economic Journal108 (1): 1009-1025 Frankel, Jeffrey A. and Andrew K. Rose, 1997. Is EMU more justifiable ex post than ex ante? European Economic Review41(3-5): 753-760. Friedman, Milton, 1953. The case for flexible exchange rates. Essays in Positive Economics. Chicago. Glick, Reuven and Andrew K. Rose, 2002. Does a currency union affect trade? The time series evidence. European Economic Review46(6): 1125-1151. Helliwell, John, 1998.How much do national borders matter? Washington, DC: Brookings Institution Press. Kenen P. B., 2002. Currency unions and trade: Variations on themes by Rose and Persson. Reserve Bank of New Zealand Discussion Paper Series DP2002/08, Wellington: Reserve Bank of New Zealand. Levin, Jack, James Alan Fox, and David R. Forde, 2010.Elementary Statistics in Social Research, New York: Allyn and Bacon/Pearson Publisher. Lutz, Friedrich A., 1954. The case for flexible exchange rates. Banca Nazionale del Lavoro Quarterly Review7(31): 175-185. López-Córdova, J. Ernesto and Christopher M. Meissner, (003. Exchange-rate regimes and international trade: Evidence from the classical gold era. American Economic Review93(1): 344-353. Meade, James E., 1957. The balance of payments of a free trade area. Economic Journal 67(4): 379-396. McCallum, John, 1995. National borders matter: Economic Review85(4): 615-623.

Canada-US regional trade patterns:

American

McKinnon, Ronald I., 1963. Optimum currency Areas. American Economic Review 53(4): 717-725.

243

O. Owoye & O. A. Onafowora. Mundell, Robert A., 1997. Currency areas, common currencies, and EMU. American Economic Review87(2): 214-216. Mundell, Robert A., 1961. A theory of optimum currency areas. American Economic Review51(4): 657665. Mundell, Robert A., 1960. The dynamics of international adjustment under fixed and flexible exchange rates. Quarterly Journal of Economics 74(2): 227-257. Micco, Alejandro, Ernesto Stein, and Guillermo Ordoñez, 2003. The currency union effect on trade: Early evidence from EMU. Economic Policy 18(37): 315-356. Nitsch, Volker, 2001. Honey, I shrunk the currency union effect on trade. World Economy 25(4): 437-474. Pakko, Michael R. and Howard J. Wall, 2001.Reconsidering the trade-creating effects of a currency union. Federal Reserve Board of St, Louis Review 83(5): 37-45 Persson, Torsten, 2001. Currency unions and trade: How large is the treatment effect? Economic Policy 16 (33): 433-448. Pop, Napoleon, Iulia Lupu, Camelia Milea, Floarea Iordache, Elena Ana Moşneanu, Silviu Ionaşcu, Adina Criste, and Alina Glod. 2010. Trends in the new EU member states macroeconomic policies to adopt the euro: Performance of the Romanian macroeconomic policies to adopt the euro. Financial Studies. 1(1): 124-140. Rose, Andrew K., 2008. Is EMU becoming an optimum currency area? The evidence on trade and business cycle synchronization. Unpublished manuscript. Rose, Andrew K., and D. Stanley, 2004. Meta-analysis of the effect of common currencies on international trade.Journal of Economic Surveys 19(2): 347-360. Rose, Andrew K., 2001. Currency unions and trade: The effect is large. Economic Policy 16(33):449-461. Rose, Andrew K., 2000. One money, one market: Estimating the effect of common currencies on trade. Economic Policy 15 (30): 7-46. Rose, Andrew K., and Eric van Wincoop, 2001. National money as a barrier to international trade: The real case for currency union. American Economic Review 91(2): 386-390. Scitovsky, Tibor, 1958. Economic Theory and Western European Integration, Stanford. Silva, João Santos and SilvanaTenreyro, 2010. Has the euro increased trade? Short answer: No. Centre Piece, Autumn 2010: 6-7. Silva, João Santos and SilvanaTenreyro, 2010. Currency unions in prospect and retrospect” Annual Review of Economics 2(1):51-74. Silva, João Santos and SilvanaTenreyro, 2006. The log of gravity. Review of Economics and Statistics 88(4): 641-658. Tenreyro, Silvana and Robert J. Barro, 2003. Economic effects of currency unions. National Bureau of Economic Research Working Paper No. 9435, Cambridge, Massachusetts. Tenreyro, Silvana, 2001. On the causes and consequences of currency union. Boston, Massachusetts: Harvard university press.

244