An optimum currency crisis.

• Author: Pasimeni, Paolo
• Position: Report

1. Introduction

   The decision to establish a common currency in Europe had long been debated (Meade, 1957; Scitovsky, 1957; Mundell, 1961; McKinnon, 1963), before it was actually taken and implemented. At the summit of The Hague in December 1969 it was decided that the Community should evolve by stages into an Economic and Monetary Union (EMU). The objective of a full EMU was explicitly stated in the Maastricht Treaty in 1992. In 1999 the euro was introduced as the official currency of 11 Member States (MS) replacing national currencies (2).

   This new institutional setting represented a peculiar monetary union, more integrated than past agreements of fixed exchange rates, but still far from complete economic and political unions (3). Monetary policies of the participating Member States (MS) became the responsibility of the European Central Bank (ECB). Fiscal policies remained responsibility of national authorities, even though subject to restrictive common rules on public finances known as the Stability and Growth Pact. This created an unprecedented divorce between the main monetary and fiscal authorities (Goodhart, 1998).

   The reasons for such an ambitious experiment were mainly political, but several economic advantages were also expected: a reduction in transaction costs, promoting trade (Rose, 2000), and the elimination of the exchange rate risks, favouring financial integration. At the same time, participating countries lost a great deal of flexibility, renouncing to national monetary policies and to a mechanism for adjustment to shocks (Krugman, 2012).

   The theory of Optimum Currency Areas (OCA), as developed by Mundell (1961), McKinnon (1963), Kenen (1969), Fleming (1971), proposes a set of necessary conditions for monetary unions and provides an analytical framework to assess risks and opportunities a region might be confronted with. This approach has been widely discussed in the literature (Robson, 1987; Bayoumi, 1994; Bayoumi and Eichengreen, 1997; Goodhart, 1998; Alesina et al, 2002; McKinnon, 2004; Tavlas, 2009; Krugman, 2012; O'Rourke and Taylor, 2013), and it has been used to assess ex-ante the feasibility of the EMU (Mundell, 1961; McKinnon, 1963; Kenen, 1969; Eichengreen, 1991; Eichengreen, 1993a; Tavlas, 1994; Obstfeld, 1997, Alesina et al, 2002).

   The standard theory of OCA was later on defined "exogenous", in contrast to an "endogenous" approach (Frankel and Rose, 1998), which admitted the possibility that OCA properties, even if not fulfilled ex-ante, could be gradually satisfied during the existence of the monetary union.

   Its applicability to the specific case of the EMU has been rather controversial. The report "One Market, One Money" (EC, 1990), for instance, discarded the theory, explaining that:

   there is no ready-to-use theory for assessing the costs and benefits of EMU. Despite its early insights, the 'theory of optimum currency areas' provides a too narrow and somewhat outdated framework of analysis. Recent developments in both micro- and macroeconomics have not yet led to a unified theory of monetary unions (p.31). Critics of the OCA theory highlighted that it lacks a formalized model allowing a measurement of the "OCA test" for potential currency unions (Robson, 1987), that it has little or no predictive capacity (Goodhart, 1998), and that analyses investigating OCA properties are by necessity backward-looking (Mongelli, 2008).

   In this work, however, we apply the theory of OCA to study ex-post the developments of the EMU, following the main conditions for a suitable monetary union prescribed by the theory:

   * factors mobility (capital and labour) across the area;

   * price and wage flexibility;

   * similarity of business cycles among participating countries;

   * common fiscal capacity as a mechanism of shock absorption and risk-sharing.

   The paper analyzes each one of these criteria, in order to assess to what extent they were satisfied before or during the EMU. The actual development of the EMU is then put in perspective, comparing the lessons drawn from the theory with the current status.

   Sections 2 to 5 review the main conditions for OCAs and the extent to which the Eurozone complies with each of them; section 6 describes the process of building up the EMU in light of the theory; section 7 explains how the "private insurance channel" worked during the first decade; section 8 shows the developments of the EMU during the crisis; section 9 the adjustment process currently taking place; section 10 proposes a simulation of a common budget; and section 11 concludes.

2. Factors mobility

   Factors mobility was proposed as a key criterion to define an OCA since the seminal work by Robert Mundell (1961), who defined an OCA "in terms of internal factor mobility and external factor immobility". We can analyze in detail the degree of mobility of capital and labour in the EMU.

   2.1. Free circulation of capital

   The free circulation of capital has been established in Europe in parallel with the development of the single market. The complete liberalisation of capital flows in the EU was not foreseen initially by the Treaties. The steps towards the EMU and the introduction of the single currency required a stricter coordination, which brought to the Council Directive 88/361/EEC fully liberalising capital movements between residents of the MS, by removing all remaining restrictions as of 1 July 1990. The Maastricht Treaty (1992) introduced the free movement of capital as a Treaty freedom.

   Today, Article 63 TFEU prohibits all restrictions on the movement of capital and payments between MS, as well as between MS and third countries. Even though some exceptions clauses were foreseen, since 1999 the safeguard clause to remedy crises in the balance of payments is only applicable to those member states which are not part of the Eurozone (4).

   The criterion of capital mobility can therefore be considered as fully satisfied by the EMU. Moreover, this growing financial integration is key to understand the development of the Eurozone (Obstfeld, 2013).

   2.2. Labour mobility

   Labour mobility in the Eurozone has not reached the same extent as capital mobility, due to cultural and language barriers, and regulatory constraints. The measurement of labour mobility has often been a complex topic addressed by many authors (Harris and Todaro, 1970; Diamond, 1981; Pissarides, 1990; Layard et al, 1991, Mortensen and Pissarides, 1994; Molloy et al, 2011). Many have also tried to study the specific case of intra-EMU labour mobility (Eichengreen, 1991; Vandenbrande et al, 2006; Bonin et al, 2008; Fenge and von Weizsaecker, 2009; OECD, 2012; Kahanec 2012 and 2013; EPC, 2013), which is of specific relevance for our analysis.

   A first, rough, estimate can be based on the number of foreign-born residents in each country. The Eurostat Labour Force Survey provides these data for each EU MS since 2009. The number of foreign-born residents in each Eurozone country is increasing in both absolute terms (from 35,650,225 in 2009 to 38,315,569 in 2012) and in relation to the total population of the Eurozone (from 10.98% in 2009 to 11.53% in 2012).

   The comparison with the US can give an idea of the degree of labour mobility in Europe and of its importance as a mechanism for adjustment: Blanchard et al. (1992) explain how in response to an adverse shock in demand, relative nominal wages might decline, but in general they do not decline enough to prevent increases in unemployment. In the US, what they trigger is mostly labour out-migration, rather than job in-migration or job creation.

   Similarly, Krugman (1993a) shows that in the US relative higher unemployment of some states is often reduced not so much by employment creation, but by reducing their labour force via outwards migration. The case of Massachusetts, experiencing a hard recession at the end of the 80s after a big boom, is explanatory: it lost a relevant part of its workforce, which never came back. This was the main driver of unemployment reduction.

   Another way of measuring the differences in labour mobility between the US and the Eurozone is by plotting the unemployment rates at two distinct points in time, in order to see whether differences among regions or states persist or not.

   We can observe that between 1999 and 2012 the differences in unemployment rates changed quite significantly among US states, the two series have in fact a low correlation (p=.382). If we analyze the same differences among Eurozone NUTS2 regions, between 1999 and 2012, we observe that they tend to be quite similar, definitely more than in the US (correlation p=.727).

   While in the US labour mobility plays a role in reducing and rebalancing unemployment rates across states, in the Eurozone differences in unemployment rates are more persistent. Labour mobility, then, seems a weaker mechanism of adjustment in the Eurozone, compared with the US. The figure suggests that unemployment in the Eurozone is a localized problem, with permanent differences among regions, and labour mobility has a limited impact in mitigating the problem.

   Molloy et al. (2011), however, have estimated the share of the population in 2005 who actually moved residence in the previous year in EU MS in comparison with the US: they found similar mobility rates between the US and some EU MS.

   Recent analyses (EC, 2013; EPC, 2013; Dao et al, 2014) show that geographical mobility of workers as an immediate response to shocks in the EU is significantly increasing over time. The willingness of Europeans to move to another MS for work doubled in the last three years; actual moves from deficit to surplus countries doubled and tripled in some Eurozone countries, during the last two years.

   Record-high unemployment levels as consequence of the crisis in the countries of origin exemplifies the growing importance of "push" factors versus "pull" factors, in explaining the increasing labour mobility within the Eurozone (EPC, 2013).

   To sum up, labour...