European Monetary Union
Managing the Transition

The politics of the treaty negotiations before the Maastricht summit were highly charged, and the debate over EMU's implications has diverted attention away from the important issues involved in Stage Two the transition from the current Exchange Rate Mechanism (ERM) of the EMS. The challenges policy-makers will face in this transition were the primary focus of a CEPR joint conference with the Paolo Baffi Centre for Monetary and Financial Economics of the Università Bocconi, held in Milan on 27/28 September, which brought together academics and central bankers from the major European countries. The conference was organized by Franco Bruni, Professor of Economics at the Paolo Baffi Centre, Francesco Giavazzi, Co-Director of CEPR's International Macroeconomics programme, Rafael Repullo, Director of Economic Research at the Banco de España, and Riccardo Rovelli, Professor of Economics at the Paolo Baffi Centre. It formed part of CEPR's research programme on Financial and Monetary Integration in Europe, which is supported by the Commission of the European Communities under its SPES programme and also by the Ford Foundation. A Report of its proceedings is now available from CEPR.
At the time of the conference there was already agreement that the European Central Bank (ECB) would be independent of political control, that it would not be established until Stage Three, that national governments would retain control over monetary policy until then, and that only those countries that met tough economic convergence requirements would be allowed to join. The conference focused on how much prior economic convergence is required for EMU to work, and it assessed the extent to which the risk of realignments, the abolition of exchange controls or continuing economic integration may destabilize the ERM during transition. It considered the functions of the European Monetary Institute (EMI) in Stage Two, the constitution of the European System of Central Banks (ESCB) in Stage Three, the new risks and complications involved in setting macroeconomic policy at the European level, and a possible supervisory and prudential role for the ECB.

The ECB's Macroeconomic Role
Opening the first session with his paper, `The Draft Statute of the European System of Central Banks: A Commentary', Charles Goodhart (LSE) maintained that low inflation and central bank independence may be less related than they appear. If an electorate's desire for price stability is weak, granting a central bank the independent power to pursue low inflation may make it politically unpopular if this leads to high output losses and unemployment: far from separating monetary policy from politics, independence may politicize the central bank. Further, the ECB's performance in delivering price stability is not governed by any measurable contract and it has no pecuniary or re-election incentive do so. Also, allowing the European Council to determine external exchange rate policy as currently proposed may undermine the ECB's ability to deliver low inflation.
Xavier Vives (Universitat Autònoma de Barcelona and CEPR) noted that exchange rates are determined by the interaction of fiscal and monetary policies. ECB independence is therefore wholly incompatible with assigning exchange rate policy to a committee of national finance ministers, which is a recipe for high inflation. Alfred Steinherr (European Investment Bank) argued that political support for price stability in Europe will initially be weak; the ECB's institutional independence could be bolstered by tighter constitutional requirements, such as a money growth rule to pin down price stability.
In his paper, `Delors and the Core: Cooperative Monetary Policy Games and the Transition to EMU', Martin Klein (Universität Bonn and CEPR) developed a two-country model of national governments' decisions to coordinate monetary policies in Stage Two or to implement EMU. The countries first decide whether to join the fixed exchange rate system; they then bargain over the common monetary policy. Their preferences for the cooperative arrangement increase with the extent of their economic integration and the destabilizing effects of nominal exchange rate swings on real exchange rates. Countries with lower incentives to join the ERM clearly have greater bargaining power; so the larger European countries, whose shares of imports in total GDP are relatively small, will determine both the terms of the EMU agreement and monetary policy during the transition.
David Currie (London Business School and CEPR) argued that the most important inter-country asymmetry is not size or openness, but the current credibility and reputation of their national central banks. It is the relative attractiveness of the alternative to fixed exchange rates facing Germany that has strengthened the Bundesbank's hand in negotiating over the EMI's role.
Michele Fratianni (Indiana University) then presented a paper written with Jürgen von Hagen and Christopher Waller, `From EMS to EMU', which considered routes to EMU that differed in the length of Stage Two and in whether or not realignments are permitted. Governments reneging on the commitment to peg to the Deutschmark face higher future interest rates; but the loss of monetary policy autonomy also entails costs if wages and prices adjust slowly to economic shocks. A country will gain from moving sooner rather than later if the real relative shocks it faces are rare and the interest rate premium is high. Fratianni maintained that differential real shocks remain sufficiently likely to prevent either a rapid move to EMU or the fixing of exchange rates in Stage Two. He proposed an alternative policy of regular realignments, which need not impair their anti-inflationary credibility if they operate within 6% bands in the ERM, with the option to narrow these bands gradually as real shocks disappear.
Patrick Minford (University of Liverpool and CEPR) welcomed the proposal for frequent exchange rate changes in a lengthy transition, since the move to a single currency should only occur once the singlemarket programme is complete and differential shocks are no longer possible. Charles Wyplosz (INSEAD, DELTA and CEPR) disputed their assumptions that asymmetric shocks will disappear over time and that wage and price flexibility are unaffected by the exchange rate regime's credibility.
In a paper on `Policy Outcomes in Stage Two and in the EMS Versus Those in a Monetary Union A Comparison', Alex Cukierman (Tel Aviv University) used a variant of the Barro-Gordon model to examine whether the shift to collective decision-making in Stage Three will itself reduce the possible inflationary bias implied by the possibility of a devaluation in Stage Two. If monetary policy decisions are then made by a majority vote of individual national members who face shocks that are not perfectly correlated, a shock is less likely to hit a majority of the countries than one country alone, so individual countries are more likely to renege during Stage Two than the ECB is in Stage Three. Such realignments could undermine the credibility of the transition and of the whole EMU enterprise, so Stage Two should be as short as possible.
Alessandra Casella (University of California at Berkeley and CEPR) disputed Cukierman's identical treatment of all EMU member countries: the high-inflation members joined the ERM precisely because they wanted to peg to the others' more credible anti inflationary policies.

Financial Stability and Regulatory Responsibilities
While the macroeconomic responsibilities of the ECB are insufficiently spelled out in its draft statutes, its microeconomic responsibilities as guardian of the European banking system are hardly specified at all. These responsibilities are vaguely defined and clearly intended to be secondary to the maintenance of price stability, and this ambiguity masks profound disagreement among EC member states about the prudential and supervisory role of the ECB. The Bundesbank argues that a lender-of-last-resort function raises moral hazard problems and creates the potential for inflation bias; others maintain that giving the ECB responsibility for banking supervision would unduly centralize this function.
In his paper, `The European Central Bank: Financial Regulation Versus Price Stability', written with Pierre-André Chiappori, Colin Mayer and Damien Neven, Xavier Vives (Universitat Autònoma de Barcelona and CEPR) argued that Europe-wide prudential regulation is needed because national regulators do not take account of all the spillover effects arising from their own regulatory failure; but it is much easier to harmonize countries' prudential rules than their regulatory authorities' discretion to act in particular cases. Vives proposed that a regulatory agency separate from the ECB administer insolvencies, supervise bank closure and provide minimum deposit insurance; the ECB should act as lender of last resort and assume responsibility for bank authorization. Securing liquidity need not conflict with maintaining price stability, if liquidity operations are sterilized when monetary conditions permit.
Jeffrey Shafer (OECD) said that the lender-of-lastresort duty was an integral part of the ECB's responsibilities for monetary policy rather than a special function or task; but it need not take on the other supervisory and regulatory roles required in a integrated financial system. Manfred Neumann (Universität Bonn) stressed that the ECB should not lend to failing institutions; it need only keep open a discount window where solvent institutions can borrow if they need to do so.
David Folkerts-Landau (IMF) presented his paper, `The European Central Bank: A Bank or a Monetary Policy Rule?', with Peter Garber, which argued that the recent growth in non-bank intermediated finance across Europe has increased the need for active central bank supervision. The growth of bank holdings of short-term, non-bank corporate obligations has increased the risk that liquidity crises within the payments system will spill over and destabilize the banking system; and banks' knowledge that the central bank will bail them out if necessary may reduce their incentive to act prudently. These risks may be reduced by the ECB's careful supervision, monitoring and prudential regu lation to minimize expected losses on bank assets acquired during liquidity support operations. The lender-of-last-resort function cannot be decentralized to the national central banks, since their support operations could then conflict with the ECB's actions to maintain price stability.
José Perez (Banco de España) agreed that separating the lender-of-last-resort and price stability functions could have destabilizing effects; but these could be overcome by establishing clear `rules of conduct' that could be decentralized to the national central banks. Riccardo Rovelli (Università Bocconi, Milano) maintained that defining the ECB's role too narrowly would make Europe's integrated financial markets more risky and force more regulatory and support functions on to the national central banks.
The question of convergence criteria was reconsidered at the panel discussion which ended the first day. Anthony Loehnis (S G Warburg and CEPR) stressed that a treaty without convergence conditions during Stage Two would provide insufficient political commitment to low inflation in Stage Three. Reimut Jochimsen (Landeszentralbank Westfalen) said that the final EMU treaty should preclude all national vetoes, arbitrary exclusion of any country and coercion to join. Christian de Boissieu (Université de Paris I) stressed the need for flexibility: for example, the direction of change of national debt ratios may be more important than their level. Niels Thygesen (Universitet Kobenhavn) agreed that any specific convergence criteria should be a starting-point only, since the final membership of any EMU is ultimately a political rather than an economic decision. Tommaso Padoa-Schioppa (Banca d'Italia) closed the session by noting that only Italy is currently divergent but eager to join EMU. Italy must make substantial progress in stabilizing its debt ratio if it is to exploit the political opportunity to get its fiscal system in order, whether or not this is necessary from a strictly economic point of view.

Exchange Rate Adjustment and Currency Substitution
In his paper, `Destabilizing Effects of Exchange-Rate Escape Clauses', Maurice Obstfeld (University of California at Berkeley and CEPR) extended the Barro Gordon model to an open economy to examine a `pegged-but-adjustable' exchange rate system. The government can renege on its commitment to the peg by invoking an escape clause and expanding the domestic money supply, but it will only be worth while to do so for a real output shock above a certain threshold, which depends on the cost of reneging. The escape clause may entail costs even in periods with no shocks by undermining the government's credibility. In practice it cannot realign, but agents' expectations that it will do so may nevertheless lead to high unemployment, making the country worse off than if it had completely fixed the exchange rate in Stage Two.
Vittorio Grilli (Birkbeck College, London, and CEPR) doubted whether any society could make the costs of reneging high enough to sustain fixed rates in the face of differential real shocks; even EMU may not remove these potentially destabilizing pressures. Giorgio Basevi (Università degli Studi di Bologna and CEPR) suggested extending the model to include the impact of foreign inflation or output shocks on the decision whether to renege.
Luigi Spaventa (Università degli Studi di Roma, `La Sapienza', and CEPR) presented his joint paper with David Begg, Francesco Giavazzi and Charles Wyplosz, `The European Central Bank: Price Stability and Exchange Rate Policy', which unequivocally rejected the view that regionally differentiated shocks are an obstacle to EMU and advocated moving to such a union as rapidly as possible. There is no need for convergence criteria before EMU, since convergence may best be achieved by irrevocably fixing the current parities. Even in its current `hardened' form, with no general realignments since 1987, the ERM has failed to remove inflation differentials; convergence can only come with a single currency. Fiscal convergence requirements are also arbitrary and unnecessary and would straitjacket national fiscal policies; indeed, with monetary policy no longer available to offset regional shocks, fiscal policy should assume a greater stabilizing role. Three policies should ensure that borrowing governments bear their share of default risk: requiring high-debt countries to issue indexed debt; forbidding individual banks to hold too much of any country's debt; and including a `no bail-out' clause vis-à-vis national governments in the ECB's constitution.
Eduard Hochreiter (Österreichische Nationalbank) agreed that there should be no final realignment before the EMU treaty is signed. The early locking in of exchange rates should boost credibility and speed the wage adjustment process. José Viñals (Committee of Governors of the EC Central Banks, Basle, and CEPR) argued that inflation convergence would inevitably be required in practice; otherwise the low-inflation countries would block progress to EMU. Also the ECB may be unable to avoid bailing out commercial banks to prevent financial instability; even its refusal to do so could trigger a general flight from public debt.
Potential instability may also arise during Stage Two from the growth of currency substitution as investors switch among currencies in search of the highest return. In his paper, `Two Concepts of Currency Substitution and their Implications for Exchange Rate Volatility', with Behzad Diba and Alberto Giovannini, Matthew Canzoneri (Georgetown University) developed two simple two-country models that distinguish between currencies' uses for transactions purposes and as stores of value. In the first model, the demand for foreign goods determines the demand for foreign currency, and the substitutability of the transactions services the two currencies supply increases with the substitutability of the two countries' goods. An increase in the output of foreign goods raises demand for foreign currency and the domestic currency depreciates; but their higher relative supply leads to a fall in their price so the domestic currency appreciates. As substitution between the two goods increases, the offsetting appreciation will become smaller, and exchange rate volatility will increase. In the second model, domestic and foreign goods are already perfect substitutes, and the currencies' substitutability is inversely related to their actual substitution, since relative shifts in demand can be offset by smaller changes in interest rates. So long as countries are expected to maintain fixed exchange rates, increased currency substitution need not lead to greater spot rate volatility.
William Branson (Princeton University and CEPR) felt that the second model was more relevant to the current European debate, but he suggested using a factor other than monetary policy to destabilize the exchange rate. Jürgen von Hagen (Indiana University) questioned the relevance of the transactions-based model, since in practice different currencies are only used for transactions purposes if it is hard to anticipate changes in relative prices because of high and variable inflation rates which are not to be part of Europe's future.
Currency substitution and the possibility that exchange rates may have to be fixed early in Stage Two both suggest the need for closer coordination of monetary policy during the transition. In his paper, `Monetary Policy in Stage Two of EMU: What Can We Learn from the 1980s?', Michael Artis (University of Manchester and CEPR) noted that national monetary aggregates should become less important as intra-European currency substitution increases. He therefore investigated whether the EMI and ECB could use a Europe-wide monetary aggregate to measure the tightness of the common monetary policy. Applying time-series techniques to data for the major European countries and the US for the 1980s, he first confirmed that the relationship between monetary aggregates and both output and inflation was generally weakening. That between interest rates and prices or output was generally much stronger. The demand-for-money function at the European level was very stable indeed, so the prospects for monetary targeting at the European level are quite favourable.
Lucas Papademos (Bank of Greece) attributed the instability of the national monetary aggregates to shifts between different types of money within countries, rather than currency substitution, which he argued was inconsistent with the stable European aggregate.

European Monetary Targets
Rather than establish a European monetary aggregate, the EMI or national governments could adjust the national aggregates to include `cross-border deposits'. Excluding these from national aggregates understates national money in circulation, which might introduce an inflationary bias into European monetary policy. In his paper, `Operating Targets of Monetary Policy Within the EMS', with Thomas O'Connell, Michael Moore (Queen's University, Belfast) argued that even these extended aggregates could be misleading, since they focus on ex post money demand, whereas ex ante indicators are also important under fixed exchange rates. Domestic money creation leads to downward pressure on the exchange rate and a loss of foreign exchange as the central bank defends the parity. Its standard indicator the domestic assets on the central bank's balance sheet omits sales of government debt to foreign investors. Gradual and modest increases in some ERM member countries' central banks' assets have been accompanied by high rates of domestic high-powered money creation due to foreign debt sales, and central banks should take account of their Treasuries' direct and indirect foreign borrowing during Stage Two.
Georg Winckler (Universität Wien) cautioned that even this proposed consolidation of central bank and Treasury balance sheets may not fully capture domestic money creation; open or unbalanced foreign exchange positions of private non-banks may also put upward pressure on the domestic currency.
Ignazio Angeloni (Banca d'Italia), Carlo Cottarelli (IMF) and Aviram Levy (Banca d'Italia) presented their paper on `Cross-Border Deposits and Monetary Aggregates in the Transition to EMU', which examined the links between nominal income and three extended monetary aggregates that classified cross-border deposits by depositor's nationality, currency of denomination, and the bank's location. The traditional aggregates were unstable and dominated by the extended aggregates; the informational contents of those based on depositors' nationality and currency of denomination were similar, while that based on bank location was insignificant.
Participants expressed unease about the authors' chosen estimation techniques. Christian Pfister (Banque de France) expected the increasing credibility of the ERM parities to enhance the relative importance of the aggregate based on the holders' nationality. Alessandro Penati (Università Bocconi, Milano) argued that the EMI may not be interested solely in the monetary aggregates: differences in national interest rates and the term structures of national debts may better signal changes in inflationary bias or the nature of economic shocks.
The EMI may use a combination of these techniques to monitor policy-making in Stage Two; but some conference participants felt that it should have independent powers. In a paper on `Reserve Requirements and EMU in Stage Two', Daniel Gros (Centre for European Policy Studies, Brussels) advocated imposing a further reserve requirement on all deposits held by commercial banks in the Community, to be held in ecus with the EMI. This would enable the EMI to control overall monetary growth and also give it experience of operating in a Community-wide money market. Although the European money supply would then technically be over-determined, the upper ecu bound would generally not bind in practice. The division of power between the national central banks and the EMI would be determined by the degree to which they stuck to their monetary targets.
Joseph Bisignano (Bank for International Settlements) noted that there is little relationship between inflation performance and reserve requirements, which are in effect a tax on bank intermediation. Franco Bruni (Università Bocconi, Milano) argued that the demand for ecus for reserve purposes could only be sufficient to allow the development of a working ecu market if a mature market for ecus already existed, in which case the reserve requirements Gros proposed would be unnecessary.
In the panel discussion that ended the conference, Reimut Jochimsen said that focusing on the role of the EMI during Stage Two distracted attention from the unfinished business of Stage One: achieving free capital mobility and completing the single-market programme. José Perez noted that de facto monetary policy cooperation may increase with national central banks' independence, but otherwise there is little difference between Stages One and Two; the real transition concerns fiscal policy. Christian de Boissieu proposed empowering the EMI to intervene in the foreign exchange markets, for which it would need to control official foreign exchange reserves. Charles Goodhart said that the conference should stop worrying about a policy role for the EMI and focus instead on the ECB's operations in Stage Three. EMU would require a considerable transition period: the UK took more than three years simply to introduce a decimal currency.
There was general agreement at the conference that much more thought must be given to the powers and responsibilities of the ECB to safeguard the stability of the European financial system. It pointed to gaps in proposed arrangements for Stage Three and highlighted outstanding issues in Stage Two. The delegates agreed that if these questions are not addressed and Stage Two destabilizes the ERM as a result, EMU may remain a dream of Europe's politicians rather than an economic reality for its citizens.