Comments on “The Maastricht Criteria on Price and Exchange Rate Stability and the ERM II,” by György Szapàry†
- Susan Schadler
- Published Date:
- April 2005
The paper gives an excellent overview of the issues involved in the participation in the Exchange Rate Mechanism II (ERM II). In my remarks, I would like to discuss the following four questions: (1) Is there a conflict between exchange rate stability and price stability within the ERM II for accession countries, considering that these countries are catching-up economies? (2) What are the main benefits and risks of ERM II participation? (3) What considerations should be taken into account in deciding the appropriate timing for entry into the ERM II? (4) What is the optimal length of stay in the ERM II?
I will begin by addressing the first question, which concerns exchange rate stability and price stability—both Maastricht criteria. In principle, there need not be a conflict between these two objectives, since a fixed peg to the currency of a low-inflation economy, in this case the euro area, is not incompatible with price stability if domestic policies, notably fiscal, monetary, and incomes policies, are supportive. However, a conflict may arise between exchange rate and price stability in the ERM II for countries where the Balassa-Samuelson inflation is larger than what can be accommodated within the 1.5 percentage point cushion provided by the Maastricht inflation criterion. Balassa-Samuelson inflation stems from productivity growth differentials and typically affects nontraded goods prices. Estimates of the Balassa-Samuelson effect in the new European Union (EU) members reported in the literature vary a great deal.15 My primary point is that the Balassa-Samuelson effect is an equilibrium phenomenon that can be accommodated within the framework of the ERM II if there is enough flexibility in the interpretation of what is regarded as exchange rate stability.
Prior to entry into the ERM II, a country has to agree with the euro area member countries, the European Central Bank (ECB), and other member states participating in the ERM on a central rate and the width of the exchange rate band. The normal width is ±15 percent, which is also the maximum width. A narrower band can also be agreed upon, as in the case of Denmark. The entry central rate should be as close as possible to the “equilibrium rate,” which can be broadly defined as the rate that helps ensure both the maintenance of competitiveness and the achievement of price stability. This latter condition is relevant for small, open economies where the exchange rate plays an important role in the determination of domestic prices. It is well known that the equilibrium exchange rate is difficult to calculate and the selection of the central rate will necessarily involve some judgment. But the central rate should be credible in the eyes of the markets.
Equally important is the recognition that the inflation and exchange rate Maastricht criteria are nominal concepts, while the equilibrium exchange rate is a real concept that may change over time. This is typically the case for a catching-up economy where productivity grows faster, resulting in an equilibrium appreciation of the real exchange rate. In such circumstances, the equilibrium exchange real rate can appreciate by higher Balassa-Samuelson-induced inflation, by a nominal appreciation of the exchange rate, or by a combination of the two. Thus, a nominal appreciation that offsets the Balassa-Samuelson-induced inflation is an equilibrium phenomenon that should be allowed to be accommodated within the concept of the exchange rate stability criterion in the ERM II. Such appreciation could help the achievement of price stability without undermining competitiveness. This is in fact explicitly recognized by the ECB when it states that, in judging whether a country has satisfied the exchange rate stability requirement, account will be taken of the factors that may have led to an appreciation.
Why would there be an appreciation of the nominal exchange rate when inflation is higher in an ERM II participant than in the euro area? This is because a participant country that is credibly on a path toward adopting the euro in the foreseeable future and where interest rates are still above euro interest rates will attract capital inflows in the expectation of the convergence of the domestic interest rate to the euro area level and the possibility of an extra gain in the event that the final conversion exchange rate is appreciated prior to the adoption of the euro, as was the case in Greece and Ireland. There is even considerable risk of the capital inflows becoming so intense that the authorities have to resort to sterilized intervention to prevent an excessive appreciation of the currency, as was done in Greece. Sterilization is necessary to avoid jeopardizing achievement of the inflation target, but it can be costly for the government budget.
Inflation differentials in the new EU members can be driven also by factors other than the Balassa-Samuelson effect, such as the liberalization of administered prices, reforms in the health system, higher demand for services (which typically accompanies the catching-up process), and, more generally, loose demand management policies. If such an “extra” inflation were to be offset by an appreciation of the nominal exchange rate, the resulting real appreciation would negatively affect competitiveness. It is therefore advisable to postpone ERM II participation until the non-Balassa-Samuelson inflation pressure has been reduced to modest levels.
I now turn to the second question, which concerns the risks and rewards of ERM II participation. The main benefits of ERM II participation is that it can anchor expectations, so that the latter can work for stability, alleviating the burden on policies. “Regime change” and “new deal” have been the terms used to describe the conditions in which policies could be framed more effectively in the run-up to the EMU in Portugal and Greece. For that to happen, there needs to be broad political and social consensus to support entry into the euro area and the policies required. In the absence of such support, the benefits of ERM II participation are doubtful. The maxim that there is only one opportunity to make a good first impression seems to apply here. If participation in the ERM II is dotted with depreciations and reversals in disinflation, the opportunity to make a good impression on market participants is lost and will be hard to regain. The ERM II can be a discipline multiplier, but only if there is enough political will to get disciplined.
The third question concerns the timing of ERM II entry. Certain conclusions regarding the appropriate timing for entry into the ERM II follow from the above discussion. First, a country should enter the ERM II only if demand management policies are on a firm path toward price stability and a significant degree of nominal convergence—in inflation, interest rates, and the fiscal deficit—has already occurred. Second, it is advisable to affect the remaining major administrative price adjustments prior to the ERM II so that they do not interfere with the disinflation process by raising the price level and inflation expectations. Finally, incomes policy based on wage agreements can play a key supportive role in the run-up to the euro adoption, as has been demonstrated by several current euro area members.
Regarding the optimal length of stay in the ERM II—the subject of the fourth question—the Maastricht criteria set a minimum of two years, though exceptions are possible and have been granted for Italy and Finland. The strongest argument in favor of not exceeding the minimum required length is that it minimizes the potential for disruptive capital flows: excessive inflows of convergence capital when policies are credibly on a path toward euro adoption and sudden reversals if policies are not seen as being conducive to convergence. To be sure, a country is exposed to such dangers outside of the ERM II, but the risks seem to be greater inside, particularly the risk of excessive upward pressure on the exchange rate due to intense convergence capital inflows.
It is often argued that a longer stay in the ERM II allows more time for the exchange rate to find its “equilibrium” level. However, the new EU members are small, open economies where the exchange rate is heavily influenced by speculative capital flows driven by the markets’ expectation of convergence or the lack thereof. Under such circumstances, the prospect is not very promising that the exchange rate will find its “equilibrium” level better if the country stays longer in the ERM II.
To conclude, I believe that a country should enter the ERM II only when policies are firmly and credibly set toward meeting the Maastricht criteria and the exit from the ERM II—that is, the adoption of the euro—is targeted for the not-too-distant future. Whether this means staying in the ERM for two years or more depends on whether the authorities see ERM II participation as a helpful disciplinary force. Irrespective of whether a shorter or a longer stay is preferred, broad political support for the required stability-oriented policies is a sine qua non. The European Monetary Union was a long-nurtured dream that enjoyed broad popular support, making it easier for people in the current euro area member countries to accept the needed short-term sacrifices in the interest of becoming a member of the union. Now that the euro is a reality, many tend to blame the monetary union and the discipline imposed by the Stability and Growth Pact for the problems besetting their economies, even if the euro has little to do with these problems. The authorities of the new EU members might now find it more difficult to muster support for the euro, since it is easier to sell a dream than it is to sell reality.
European Commission, DG Economic and Financial Affairs, Brussels. The views expressed in this paper are those of the authors and do not necessarily reflect the views of the European Commission. Research assistance by Moises Orellana, Jiri Plecity, Stéphanie Riso, and Tony Tallon is gratefully acknowledged. Sean Berrigan and Max Watson commented on an earlier draft.
This is the length of transition from the launch of the first stage, which did not require changes in the treaty, on July 1, 1990, to the actual start of the third stage on January 1, 1999.
Tsoukalis (2000) recalls that the Maastricht Treaty was the result of a debate between the “economist” and “monetarist” views on the plans set out in the Delors Committee Report for the achievement of economic and monetary union. The “economist” view was that the monetary union would be sustainable only if the joining member states first achieved a low level of inflation and resolved fiscal imbalances. This was a popular view among monetary authorities (e.g., the Bundesbank, which labeled it the “coronation approach”), where participation in the monetary union was seen as the last step of successful efforts to eliminate inflationary behavior. The “monetarist” view—popular with academic economists (e.g., Begg and others, 1991)—was that the creation of a new currency with its own central bank would radically change economic incentives and behavior. Accordingly, pre-EMU behavior of public and private economic actors would be a poor predictor of behavior in EMU. Instead of convergence criteria, what was needed, in the latter view, were solid institutions, and in particular for the EMU project an independent central bank: convergence criteria “create pain with no assured gain” (Wyplosz, 1997, p.11).
Interestingly, Boltho (1998) argues that one neglected aspect of the optimal currency area (OCA) theory, which is the standard economic framework for analyzing the decision to join a monetary union, is the degree of similarity in policy preferences. Most of the OCA literature has focused on two other aspects of the theory: the degree of openness and integration of the different economies, and the degree of similarity in their economic structures. For Boltho, a clear lesson from OCA theory is that the requirement for convergence in policy preferences can be equally important for a successful monetary union. The Maastricht criteria can then be interpreted, using this neglected aspect of OCA theory, as ensuring that countries that are about to pool monetary policy share some basic policy attitudes: the goals of eradicating inflation and reducing fiscal imbalances prior to participation in the monetary union.
The benefits that accrue to the high-inflation country are related to a stable currency and price stability in the monetary union. As outlined by Begg and others (1991), associated benefits are lower interest rates, since financial markets would no longer require risk premiums to compensate for the prospect of recurrent devaluations. To the extent that lower interest rates reduce the cost of servicing public debt, this may also lead to lower taxes. On the other hand, countries that already have low inflation and a stable currency or that have invested considerably to reach a low-inflation environment are not likely to want to compromise the results of such investments by forming a currency union with high-inflation countries, unless given a credible guarantee as to the price-stability orientation of the currency union.
Bini-Smaghi and Del Giovane (1996) show in a simple two-country model that a monetary union improves overall welfare only if the commitment “technology” underlying monetary policy—that is, the degree of central bank independence—is set at the highest level (i.e., that of the most independent of the two countries). Interestingly, these authors also show that if the reputation of the central bank were damaged by a lack of convergence at the start of the monetary union, this would increase the welfare loss from joining for the low-inflation country and reduce the gains for the high-inflation country. Since lack of inflation convergence may damage a country’s reputation, it is in the interest of both countries to achieve convergence before moving to monetary union. In their model, inflation convergence is best achieved by exchange rate pegging, with the high-inflation country giving up its monetary sovereignty to the low-inflation country and moving to monetary union only once convergence has been reached. They indicate that most countries in the European Monetary System followed this strategy even after the 1992–93 crisis.
This is what Begg and others (1991) call an “incentive compatible design”—that is, criteria that limit the risks incurred by the countries already pursuing prudent policies, while not imposing conditions that are too strict to remove all incentives for other countries to join the union.
It may be noted that because most of the member states have adopted the single currency, the argument of competitive devaluations may now have less weight for countries that have not yet joined the euro area. Monetary integration in the EU may have rendered their monetary and exchange rate policy more effective and thus may have changed the cost calculation pertaining to abandoning independent monetary and exchange rate policies.
This argument draws on earlier work by Froot and Rogoff (1992), who looked at the role of government consumption in explaining real exchange rate movements in the European Monetary System between 1986 and 1991.
Note that the impact of the Balassa-Samuelson phenomenon on the inflation differential with the euro area is, in principle, not dependent on the existence of a difference in price levels between the two regions. Even among regions with similar price structures, inflation differentials due to the Balassa-Samuelson phenomenon can appear if productivity growth differentials between the tradable and nontradable sectors significantly differ between the two regions.
The exchange rate mechanism underlying the EMS was based on bilateral parities among the participating countries, with bilateral “normal” fluctuation margins of 2.25 percent around the central bilateral rate for a number of currencies and “wide” margins of 6 percent for other currencies. Following the turbulence in exchange markets in 1993, the bilateral fluctuation margins were temporarily widened to ±15 percent.
“A country running a current account surplus is receiving more payments from abroad than it disburses. If the private financial account is in balance more domestic currency must be flowing in than is leaving the country. This is possible only if someone makes up the difference. In fact, the imbalance between inflows and outflows translates into excess demand for the domestic currency on exchange rate markets world-wide. This demand tends to appreciate the domestic currency. If the monetary authorities want to avoid such an appreciation, they must relieve the pressure by selling the ‘missing’ domestic currency against foreign currencies” (Burda and Wyplosz, 2003, p. 451).
See, for instance, Buiter and Grafe (2001).
The views expressed are the author’s and do not necessarily reflect the official view of the Magyar Nemzeti Bank.