Monetary Policy In EMU

Many believed monetary policy would be tight in Europe once the euro was adopted. But this economist believes it might be just the opposite.
When the European Central Bank (ECB) began operations in January 1999, there was a widespread expectation that its monetary policy would be firm and that its currency, the euro, would be strong. That expectation is now in some doubt; interest rates have been cut faster than expected and the euro depreciated by about 10 percent relative to the dollar in its first three months of its existence. Although there are good arguments for monetary expansion in Europe under present circumstances, there are also reasons to believe that monetary policy in Europe may remain more expansionary than has been expected, at least in the short to medium term. This tendency toward expansion grows out of a problem that is inherent in the operation of the European Monetary Union (EMU): How is the European Central Bank to set a single monetary policy for eleven separate national economies?
Is Europe an Optimum Currency Area?
A single monetary policy can be expected to work well for the members of the European Monetary Union only as long as these countries have similar rates of inflation and unemployment and therefore need approximately the same monetary conditions. If, however, business-cycle conditions differ widely among the members, a one-size-fits-all monetary policy will not succeed or at least will be quite painful for some countries.
The academic debate over the desirability of a currency union in Europe began with Robert Mundell's 1961 introduction of the concept of an optimal currency area.(1) He argued that a monetary union should be no larger than the area over which similar business cycles typically prevailed or over which labor is easily mobile. A currency union larger than that will allow localized recessions and high levels of unemployment, about which monetary policy can do nothing. Business-cycle conditions often differ among EMU members, as indicated by the contrast between the current expansionary boom in the Netherlands and Ireland and the soft economic conditions in France and Germany, and differences in language and culture remain major barriers to labor mobility. Highly educated French and German citizens, who have foreign language skills, can easily move to other European countries to find employment, but ordinary workers, who only speak their national language, would find this difficult or impossible. Europe would certainly appear to be far too large to meet Mundell's criterion for a successful union, and at times even the United States appears to be larger than ideal for a single monetary policy.(2)
During 1986-87 the United States as a whole was in a period of considerable prosperity and appeared to need a restrictive monetary policy. The Dallas Federal Reserve District, however, was in a recession, caused by a sharp decline in the prices of oil and natural gas, and would have benefited from relatively easy money. The policy of the Federal Reserve System, of course, was determined by the apparent needs of eleven districts rather than one, meaning that Texas and neighboring states had to survive a monetary policy that was inappropriate for their needs. Within the United States, however, the federal government absorbs about 35 percent of the costs of a localized recession through reduced tax receipts and increased transfer payments. If output and earned incomes in Dallas decline by $10 billion, disposable income will decline only about $6.5 billion because of declines in personal and corporate income tax payments to Washington, and increases in unemployment insurance payments and other federal transfers.
There are no such fiscal arrangements in Europe, which means that a localized recession in Finland will not be partially absorbed by the other members of the European Union (EU) through reduced tax payments and increased transfer payments. The costs of a Finnish recession will be carried entirely by the Finns without help from Brussels, making localized recessions within the European Monetary Union (EMU) considerably more painful than those experienced by regions within the United States.

Advantages of the EMU

If a single monetary policy will be as difficult for some countries as has been suggested here, one might ask why the members of the EMU decided to proceed with this enterprise. Despite its likely problems, the EMU retains some major advantages:

1. Price stability in southern Europe. Countries such as Italy and Spain have had a history of excessive inflation, and may well be better off not running their own monetary policies. The European Central Bank begins operations with a level of credibility that is not readily available for the national central banks of Spain, Italy, or Portugal.
2. Reduced transaction costs. A single currency ends the need to exchange one currency for another within the EMU and thereby reduces transactions costs. Bid/asked spreads in European exchange markets may have been narrow for large transactions, but now they will be zero. Savings in small retail transactions will be far larger; anyone who has lost 6 or 8 percent of their money exchanging currencies at an airport or railroad station "bank" will appreciate the savings that Europeans will experience as tourists in other EMU member countries.
3. Transparency in pricing. It is often argued that European firms can get away with charging quite different prices for the same products in various European Union countries because the prices are in different currencies. A single currency is supposed to make this more difficult and make uniform pricing almost universal.
4. Rivalry with the dollar and New York. The euro will be the money for an enormous economy and can be expected to emerge rapidly as a widely used international currency, which may allow it to supplant the dollar. This would help make Frankfurt and Amsterdam successful competitors with New York as world financial centers. This argument would be more persuasive if the United Kingdom had joined the EMU at the outset; London is a more believable rival for New York than is Amsterdam or Frankfurt.
5. Ease of management for other E U institutions. The Common Agricultural Policy, for example, would be more manageable with a single currency than with many. In the past, if the Italian lira fell in value against the deutsche mark, either German food prices had to be reduced or prices had to be raised in Italy, complicating the agricultural support program's finances. Many EU institutions will be easier to manage with monetary union, but would be even easier if the four nonmember countries (Denmark, Greece, Sweden, and the United Kingdom) were to join.
. Elimination of competitive devaluations. It is no longer possible for an EMU member country to devalue its currency in order to gain a competitive price advantage over other member countries in export markets.
7. Advancement of European unity. The most important argument for monetary union has little to do with economics and far more to do with politics and history. Centuries of war in Europe have led a generation of its leaders to conclude that only a federation of western Europe can preclude a return to such chaos and pain. The EMU is a huge step on the road to that federation, and if it causes some economic pain, it is argued, the political gains make it more than worth it.

How Monetary Policy Is Determined in the EMU

Returning to the earlier issue of monetary policy goals among EMU members, setting a single monetary policy for Europe is likely to create major conflicts within the ECB management whenever member countries face different macroeconomic conditions. This leads to the question of how monetary policy changes are decided within the monetary union and therefore how these disagreements might be settled.
The ECB management structure bears a striking resemblance to that of the Federal Reserve System. Short-term management decisions are made by a six-member Executive Board, which corresponds to the Board of Governors in the United States. Longer-term monetary policy is set by the Governing Council, which consists of the six members of the Executive Board plus the eleven governors of the member central banks. The Governing Council is very similar to the Federal Open Market Committee of the Federal Reserve System, with the governors of the member central banks playing the role of the presidents of the district federal reserve banks. In the case of the Governing Council of the European Central Bank, however, all eleven governors vote, creating a total of seventeen votes on the council. At present there is no public record of the meetings of these boards, making it impossible to determine which members voted for or against a particular policy
The Executive Board now has members from Germany, the Netherlands, France, Italy, Spain, and Finland, giving each of these six countries two votes on the Governing Council, while Belgium, Ireland, Luxembourg, Portugal, and Austria each have one.
In trying to predict monetary policy decision-making within the ECB, many hope that nationalism is dead and that the members of the Executive Board and the Governing Council will vote solely on the basis of a broad European interest. Such idealism seems very unlikely Alternatively, it might be assumed that nationalism is alive and well in Europe and that the members of these boards will vote on the basis of circumstances prevailing in each of their home economies. This paper assumes the latter, more nationalistic, approach to decision-making within the ECB management committees.

Monetary Policy Decision Rules

Central bankers are expected to care about, and therefore respond to, both inflation and unemployment. Monetary-policy decision rules normally include both goals and predict that monetary policy is tightened if the rate of inflation is high or if the unemployment rate is low, and that monetary policy is eased in the opposite circumstances. Including both inflation and unemployment in such a decision rule has the additional advantage of distinguishing between aggregate demand shocks, which monetary policy is expected to address, and aggregate supply shocks, about which monetary policy can do very little. A positive aggregate demand shock, for example, would cause inflation to accelerate and unemployment to decline, causing the monetary-policy decision rule to indicate a clear need to tighten, and vice versa. A negative aggregate supply shock, such as the impact of the 1974 decision by the Organization of Petroleum Exporting Countries to increase oil prices to the United States, would instead cause both inflation and unemployment to increase, producing a small change in the monetary-policy decision rule of uncertain direction. Such a supply shock, of course, cannot be dealt with through monetary-policy changes, so a decision rule that suggests little or no policy adjustment in such a circumstance would be appropriate.
A decision rule that includes only inflation (tighten when it is rapid, loosen when it slows) has the major disadvantage that it does not distinguish between aggregate demand and supply shocks. Either a positive aggregate demand shock or a negative aggregate supply shock would cause prices to increase, signaling a tightening of monetary policy. Such a response would be appropriate in the case of the demand shock, but not for the supply shock, where it would further reduce output and increase unemployment.
The purpose of this paper is to use monetary-policy decision rules to suggest which members of the ECB management boards might be more likely to favor tight or expansionary monetary policy as of early 1999.(3) The best-known monetary-policy decision rule is the Taylor rule, which has been used to predict changes in the U.S. federal funds rate.(4) It presumes that the real fed funds rate responds equally to a one-percentage-point change in the rate of inflation or a one-percentage-point change in the unemployment rate, and that the nominal fed funds rate responds to these changes in inflation and unemployment by a ratio of 3:1. This rule will be applied to recent data for members of the EMU to suggest which countries would now prefer a tighter or more expansionary monetary policy. Then a simpler variant of the Taylor rule, which is also based on both inflation and unemployment, will be applied to the same data. This rule will also be used to suggest which members of the ECB management boards would be likely to prefer a tighter or more expansionary policy as of early 1999. Finally, a rule will be used that normalizes the inflation and unemployment rates for the recent experience of each country. If inflation in a country exceeds its average rate for the previous three years, or if the unemployment rate is less than its average for the previous three years, tightening is indicated; when the opposite situations exist, loosening is called for.
Returning to the original Taylor rule, which was derived from U.S. data for recent decades, the desired nominal federal funds rate is:
Inflation + 2.0 + 0.5(Inflation - 2.0) - 0.5(GDP gap)
For the purposes of this exercise, full employment is defined as a measured unemployment rate of 5 percent, so the gross domestic product gap becomes the unemployment rate minus 5.0. The following numbers are not intended to suggest what interest rates the central banks of various countries would prefer, but instead are merely indicators of each country's likely desire for a tight or expansionary monetary policy. A low (negative) number indicates a preference for an expansionary policy, and a high number for a tight policy.

Data and Results for the Taylor Rule

The inflation rate in these calculations is that prevailing in each country during the six months to October 1998, which is the latest data available from the Organization for Economic Cooperation and Development's (OECD) Main Economic Indicators Internet site at the time of this writing. The range among the eleven countries is rather narrow, from a high of 2.4 percent in Portugal to a low of -0.4 percent in France, the average being 0.93 percent. The unemployment rate used is from the same source and is the last monthly rate available, which was October 1998 for most countries, September for Luxembourg and the Netherlands, and July for Italy. The range of unemployment rates among the eleven members is much wider, from a high of 18.2 percent in Spain to a low of 2.2 percent in Luxembourg, with an average of 8.6 percent. Since there is little difference in rates of inflation among the countries, but the differences in rates of unemployment are large, most of the difference in the monetary-policy decision rules that follow, with the implied differences in desired monetary policy, will result from unemployment rates rather than inflation. This, however, need not be true in the future. Despite free trade and a single currency, rates of inflation can continue to differ among EMU members, just as they frequently do among regions of the United States.(5) While the continued operation of the EMU might lead to long-run convergence in business-cycle conditions among its members, this is far from certain, particularly with regard to unemployment rates where language and culture will remain as barriers to labor mobility.

The Taylor rule statistic for the eleven countries is shown in Table 1, with the countries with seats on the Executive Board, and therefore having two votes on the Governing Council, being listed on the left, and those with only one vote on the right.

The range for this statistic is from a high of 4.9 for Portugal (preference for tight money) to a low of -3.2 for France (prefers expansionary policy). Although it is not possible to predict a precise vote count within the Governing Council of the European Central Bank, these numbers do suggest who will lead the two sides of the debate. France and Spain should have a clear preference for an expansionary monetary policy, while Portugal, Luxembourg, and the Netherlands have the strongest reasons to prefer a restrictive policy. These numbers may indicate why the French government was so strongly opposed to a Dutch president of the ECB. If each member of the Governing Council voted the interests of his or her country, there would probably be four votes on each side with nine undetermined. Among the remaining members, the Finnish representatives would be the most likely to vote for expansion, followed by Germany, Italy, and Belgium. Austria would be the most likely to vote for a tight policy followed by Ireland. There might be an indication of a majority for an expansionary policy in these numbers, particularly since very recent data for Germany, as shown in the back pages of the Economist for January and February 1999, have shown an increase in the unemployment rate relative to the October 1998 number used in this study. The main conclusion of these data, however, is not to suggest what policy would necessarily get a majority vote but, instead, to indicate the probability of conflict over the direction of policy and the likely leaders of the two sides of the discussion.

A Modified Taylor Rule for Europe

It might be argued that this use of Taylor's formula for monetary policy, which was derived from U.S. data, is unrealistic and that a somewhat simpler version of the rule that would be more appropriate for Europe might be devised. If, as the Maastricht treaty establishing the European Union strongly implies, price stability is a primary goal, with a desire for zero inflation, and if unemployment is equally important but 6 percent is viewed as a historic norm, the new monetary policy decision rule statistic becomes:
Inflation Rate - (Unemployment Rate - 6.0)
If this number is positive, a country would be expected to prefer a tight monetary policy; and if it is negative, an expansionary one. Calculating this statistic for the members of the EMU, using the same OECD data as in the previous exercise, produces Table 2, again with those countries having two votes on the Governing Council at the left, and those with one vote at the right.
It is interesting that five of the six countries with two votes on the Governing Council appear to favor an expansionary policy, while a majority of those with only one vote seem to favor a tighter policy. However, one cannot predict exact voting preferences with this table, since the assumptions of a 6 percent norm for unemployment and of a zero tolerance for inflation could be viewed as arbitrary; but the nature of likely conflicts within the Governing Council can be seen. Spain, France, Italy, and Finland, all of which have two votes, have a likelihood of favoring expansion, while Luxembourg, the Netherlands, and Portugal should favor tight policies, with only one of these countries having two votes. The current membership of the Governing Council, in terms of which countries have two votes, appears to create a tendency toward an expansionary policy under this decision rule.

A Taylor Rule That Incorporates Past Experience

Thus far the same decision rules have been applied to all eleven countries, meaning that differences in past national experiences are not taken into account. It might be expected, however, that each country evaluates current rates of inflation and unemployment relative to the average experience during recent years in deciding what monetary policy it prefers. A restrictive policy would be desired if inflation is higher than the recent norm or if unemployment is lower than normal, and vice versa. A Taylor-like monetary-policy decision rule that allows for these differences in recent national experiences might be the following:
Inflation - Three-Year Average Inflation Minus Unemployment - Three-Year Average Unemployment

This statistic is really each country's current rate of inflation minus its current rate of unemployment, with each number normalized for the experience of the previous three years. Using 1995-97 as the base three years and the same current data that were used in the previous two exercises, the numbers in Table 3 result for the eleven countries, with a positive number indicating a desire for a restrictive monetary policy; and vice versa.
This statistical exercise produces quite different results from the two preceding it. France and Italy remain in favor of an expansionary policy, but Spain now favors more restraint, as do Finland and Ireland. Portugal and the Netherlands remain supporters of tighter policies, but Germany and Austria now support an expansionary approach. Luxembourg and Belgium might be expected to prefer no change from the current policy. Allowing for the countries that currently have two votes on the Governing Council, there should be eight votes for a restrictive policy, seven for expansion, and two for no change. If national preferences are accurately described by this rule, there could easily be a stalemate, which suggests little or no change in monetary policy.

Conclusion

The management boards of the European Central Bank can expect to function without serious internal conflicts only if macro-economic conditions are very similar in the EMU member countries or if the members of the boards forgo the national interests of their countries in favor of a broad European viewpoint. Neither seems likely. Macroeconomic conditions varied considerably among the member countries as of late 1998, and there is little reason to expect national interests to be forgotten by the members of the Governing Council of the ECB.
If it is assumed that members vote the interests of their home nations, it is possible to use variations of the Taylor rule and recent inflation and unemployment data for EMU member countries to suggest where conflicts might occur in determining European monetary policy.
The original Taylor rule and two variations of it have been used to suggest which countries would be most likely to favor tighter or more expansionary policies as of early 1999. Under all three rules, France would strongly support monetary expansion, and the Netherlands and Portugal would firmly oppose it. Germany would favor expansion under two of the rules and be slightly inclined toward expansion under the third. Italy is in a similar circumstance, while Luxembourg would support a tighter policy under two of the rules.
Allowing for which countries have two votes on the Governing Council, two of the three rules indicate the possibility that a majority of the members would favor an expansionary policy; while the third rule suggests a stalemate and therefore a likelihood of no major policy change. The fact that a number of the two-vote countries (Germany, France, Italy, and Spain) now appear to have chronically high unemployment rates suggests a possible continuing tendency toward easy money in the Governing Council.
Readers who believe that one of the Taylor rule variations in this paper might be expected to reflect how members of the Governing Council of the ECB would vote can easily insert newly available national data on inflation and unemployment into that rule to indicate how monetary policy in Europe might evolve. They can also change the Taylor rules by adjusting the target rates of inflation and unemployment, or by varying the inflation/unemployment trade-off to match that which they believe members of the ECB Governing Council may be using.

Notes

1. Robert Mundell, "The Theory of Optimum Currency Areas," American Economic Review (September 1961): 657-66. For an opposing view, which supports large currency unions on the ground that they encourage price stability, see Ronald McKinnon, "Optimum Currency Areas," American Economic Review (September 1963): 717-25.
2. For other studies of the potential problem of EMU member countries encountering different business cycles while being required to accept the same monetary policy, see Tamin Bayoumi and Barry Eichengreen, "One Money or Many: Analyzing the Prospects for Monetary Unification in Various Parts of the World," Princeton Studies in International Finance 76 (1994). See also Paul de Grauwe, "Monetary Union and Convergence," European Economic Review (April 1996), and Barry Eichengreen, "Should the Maastricht Treaty Be Saved?" Princeton Study in International Finance 74 (1992). See also Guillermo de la Dehasa and Paul Krugman, EMU and the Regions, Occasional Paper no. 39 (Washington: Group of Thirty, 1992).
3. For a study of which members of EMU have tended to have similar business-cycle patterns in the 1970-95 period, and which therefore might be expected to agree on monetary policy decisions within the Governing Council of the ECB in the future, see Robert Dunn, "EMU Without Britain: Reasons for Scepticism," Economic Affairs (september 1998): 45-51
4. John Taylor, "The Inflation/Output Variability Tradeoff Revisited," in Goals, Guidelines, and Constraints on Monetary Policy, ed. J. Fuhrer (Boston: Boston Federal Reserve Bank, 1994).
5. See Economic Trends (October 1997): 7, for data on differing rates of inflation among major U.S. cities, which ranged from 1 percent in Baltimore to 3.6 percent in San Francisco and Miami in 1996-97. Economic Trends is a publication of the Cleveland Federal Reserve Bank.
ROBERT DUNN, JR., is professor of economics at George Washington University