Printer Friendly

European currency union and the EMS.


The Delors report on monetary union in Europe has enlivened the debate on European monetary integration. We would argue that the debate has avoided one of the major issues it should be addressing. There has been very little discussion of the performance of a monetary union when there are external shocks. It is not obvious that a monetary union would cope better with external shocks than would the individual components of the union if they were free to respond. Both exchange-rate and monetary policies can be seen as shock absorbers. Their use can reflect the structural differences between economies, allowing both for differences in comparative advantage that arise from natural endowments such as oil reserves and for differences in wage and price behaviour which, although not immutable, may take some time to remove.

This note addresses two issues. We ask whether, given past experience, it would be possible to create a monetary union. We then discuss the potential effects of shocks on a union, and compare them to the effects of the same shocks on the individual countries operating a flexible exchange-rate mechanism. The note uses the Institute's world model, GEM, to undertake various simulations to illustrate our arguments.

We have chosen to illustrate our argument by comparing a system of floating rates to a monetary union because this allows us to make a comparison of the benefits from using the exchange rate as a shock absorber compared to the costs of completely abandoning this role. We do not think that there are no workable exchange-rate mechanisms between these two extremes. The issues involved in discussing monetary unions are, however, clearer if we make a comparison of this sort.

The economies of Europe are very diverse and have behaved very differently in the past. However, the formation of a monetary union may cause the economies of Europe to structurally converge more rapidly than they have done in the past. if this is the case then the simulations presented below, which are based on a model estimated over the past, can be seen as a worst case scenario. However, the empirical work associated with the construction of the model suggests that the process of structural convergence in wage and price behaviour has been slow. We would argue that the convergence of inflation rates in Europe in the 1980s is associated more with the convergence of economic policy than of economic structure.

Monetary unions

There has been much discussion in the theoretical literature on international monetary economics of the situations in which it is desirable to engineer a currency union. The early literature (1) presumed that there was some policy exploitable trade-off between inflation and unemployment in the long run. In the late 1960s, and in the very early 1970s there was much discussion of the concept of currency block areas, with the US being the dominant player in one block, and the European economies banded together in another block. Structural similarities between economies were seen as essential in the formation of the union and a considerable degree of trade linkage was judged to be necessary. it was assumed that if the degree of bilateral trade between two countries is low, and if they do not compete with each other in third markets, then major and substantial real exchange-rate misalignments are not important in that they will not disrupt either economy. It was claimed that the gains from currency stability were sufficiently great that economies with similar output-inflation trade-offs would be advised to form a currency union where exchange rates were irrevocably fixed and a central monetary authority would issue the common currency. The gains to be had from exchange-rate stability were not always spelt out, but it was widely assumed that they are positive and significant.

The nature of the argument over currency unions changed as the 1970s brought a realisation that there might be no long run output and employment gains to be had from a higher inflation rate. If there are no such gains then the arguments against currency unions must rest on significant differences between countries in the short run output inflation trade-off, or the benefits to be gained from a country being able in the long run to determine its own inflation rate and exchange rate, or on the possibility that there could be country specific shocks. in this latter case policy independence could bring advantages.

The costs of a stable nominal rate might also be associated with differential responses to common inflationary shocks. If wages and prices react quickly and symmetrically to inflation-inducing shocks then there are few costs involved in accepting the shock and then later undertaking the policies to reduce inflation. The inflation inducing shock could come from commodity prices, from the expansionary effects of policies elsewhere, such as the US, or from internal causes such as unanticipated changes in structure or conscious policy design. If the shock is from outside then it may be beneficial for a closely linked group of trading partners to float their exchange rates up together. Changes in the nominal exchange rate against a close trading partner may be seen as undesirable and if the bloc chooses, or is forced, to accept some of the inflationary shock then the costs to each country will vary depending upon differences in the speed of response. Some countries may find it relatively costless to allow the inflation rate to rise and then to fall, whilst others may take some time to adjust to higher and then lower rates of inflation. Our analysis below suggests that there are currently large differences in the speed of response to shocks amongst the members of the

European Community.

The Delors report and the internal market The Delors report on Economic and Monetary Union and the subsequent report on 'Economic and Monetary Union' (1990) probably propose the most radical change in the conduct of macroeconomic policy in Europe since the war. A Monetary Union would mean the 'irrevocable locking' of exchange rates between the currencies of the member states, whilst the Delors Report claims that an economic union also requires the coordination of macroeconomic and budgetary policies. Monetary Union requires a European central banking system (or a central bank) which would be given the main responsibility for keeping down European inflation. Although this appears a necessary consequence of union, the Delors Report also requires considerable concentration of fiscal policy powers. The premise of the Report is that European central banks, when faced with inflationary pressures, would raise interest rates and there is clearly a worry that their task would be made more difficult if the governments of member states were pursuing independent and expansionary fiscal policy. Although this may well be a problem, a European Central Bank could be designed to ensure that fiscal disharmony was not a problem. A number of federal states, such as the US, Canada and Australia, allow considerable fiscal autonomy to their component members.

A monetary union is, at its extreme, characterised by rigidly fixed exchange rates and by a central foreign exchange reserve manager. The existence of rigidly fixed exchange rates would lead to an immediate movement of interest rates toward a common level. If interest rates on equivalent risk assets differed between countries in the union then all funds would flow to the higher paying source until rates equalised. interest rates differ in part because inflation expectations differ between countries and to the extent that the monetary union brought an immediate adjustment in expectations then the transition toward interest rate equivalence would be easy. Inflation rates may not of course converge so rapidly, but ultimately they have to be similar in all members of the union. However, interest rates do not differ just because of inflation differentials. Real interest rates may differ between economies because they reflect differences in capital productivity and profitability. More profitable economies would generally find their real exchange rates depreciating. A currency union will change the relative real rates of return available to investors in different countries but it is still possible for real exchange rates to change inside the union, and real rates of return may hence continue to differ. interest rates may also differ between regions for reasons of portfolio preference. A region that runs continuing deficits with the rest of the world must on balance contain individuals who are increasing their indebtedness to the rest of the world. As outsiders' portfolios become saturated with such assets the rate of return needed to keep them holding these assets will rise relative to that on other assets. Hence the interest rate actually charged to a region will for instance rise as a sequence of deficits builds up a net debtor position.(2)

The Delors Report must be seen in the light of the move toward a unified market in Europe. The nature of the reforms involved in the market may, indeed, ultimately remove profitability differentials and cause a major move toward structural homogeneity in the Community. Increased competition on a Communitywide scale may well cause the large-scale relocation of industry. The unified market involves the removal of all barriers to trade that currently exist. This initially involves the removal of discriminatory tax regimes, and ultimately the removal of barriers to the movement of factors of production. This latter process is likely to be lengthy in that it requires a unified capital market and free movement of labour, which in turn requires the recognition of education and training qualifications obtained in other countries.

However, the unified market alone does not imply that domestic policy makers lose control of monetary policy. It is true that with free capital markets it is not possible to have an exchange-rate policy that is independent of interest rate and monetary policy. Policy makers still have one degree of freedom available to them; they can choose the inflation rate that they see as desirable, and they have to accept the exchangerate consequences. If the UK chooses to inflate faster than the rest of Europe then we would expect UK interest rates to be higher than elsewhere and exchange rates would move in line with the inflation differential.

Monetary union: some simulation results using GEM (3)

We used our model to construct a synthetic monetary union in Europe between Germany, France, italy and the UK (along with the Netherlands and Belgium). We used three criteria by which to judge the successful creation of a union. Firstly, and obviously, that exchange-rate changes against any single outside currency would be the same everywhere. Secondly, that trade and competition along with fiscal policy would have to ensure the approximate convergence of inflation rates within eight years. Thirdly, although current balances no longer need to be financed' or reported, they are not irrelevant in the union. in the longer term the current balances on our model reflect the savings and investment behaviour of individuals in the economy. In the longer term countries with persistent surpluses such as Germany are displaying a preference for accumulating a stock of overseas assets, whilst persistent deficit countries display a desire to borrow. in equilibrium those flows will depend on demographic factors and on comparative advantage, and we do not expect a monetary union to drastically change these. Hence we require that inter-regional financial flows under a monetary union are broadly similar to our base current account flows.

In our analysis we have assumed that exchange rates were irrevocably fixed from 1991Q1, and interest rates were assumed to converge toward German rates. We presume that large risk premia would exist in the early stages of a union, but that interest rates would gradually converge as the credibility of maintaining a successful union rose. (4)

The reduction of nominal interest rates toward German levels would raise the level of demand, especially in the UK and Italy, even given the concomitant reduction in inflation that would come from having a stronger exchange rate. We presume that fiscal policy would be used to maintain competitiveness by reducing inflation to German levels. This would also constrain increases in consumers' expenditure in France, italy and the UK. We have assumed that equilibrium savings and investment flows would be produced if (i) the Italian budget deficit were progressively reduced from around its current 10 per cent of GDP to around 5 per cent by 1999, (ii) if French fiscal policy were tightened compared to our base assumption by around 0.5 per cent of GDP and (iii) if the UK tightened fiscal policy through the 1990s by up to 2 per cent of GDP. (5)

On our base exchange rates change in line with risk adjusted interest rate differentials, with the exchange-rate path for the future linking on to the past at current exchange rates. Each economy on our model has an interest rate and exchange-rate equation. We assume that the authorities aim to keep real interest rates constant in the long run, and our exchange-rate equation is essentially forward looking, in that a 1 per cent increase in interest rates sustained for five years will cause the exchange rate initially to rise by about 5 per cent, and then to depreciate from this peak by around 1 per cent per annum. The jump and the subsequent depreciation are ameliorated by variable risk premia associated with current account flows (details of the equations may be found in the latest issue of the Institute's GEM model manual).

Our base assumes a successful EMS, with the Franc depreciating against the D-mark by only 1 per cent a year and the Lire only by an average of 1.7 per cent a year. These depreciations have to be removed when we create the union, and hence interest rates have to be cut by just enough everywhere to prevent any further realignments after 1991/Q1. Fiscal policy has to be tightened, as described above in order that the financial flows associated with the Union remain feasible.

Each of the European countries on the model has a long run NAIRU built into it, and hence in the long run changes in the rate of inflation do not change the real wage rate and hence do not affect the level of output. However, the speed of response to inflationary shocks differs between economies. Our wage and price equations are estimated over the longest possible structurally stable period, and although the long run properties are similar in different economies the estimated equations do differ. in particular Italy and the UK react more rapidly to inflationary shocks than do France and Germany. They also have higher inflation rates on our base, and in order to remove them we have assumed that there is a progressive tightening of fiscal policy; hence in the UK and italy the union has progressively greater effects on output and prices. Eventually, however, the inflation rates on our model will converge and output will no longer have to be held below base. The time horizon involved does appear to be rather long.

Our synthetic monetary union appears to be easy to create. We assume that the Bundesbank's approach to policy is adopted by the New European Central Bank, and inflation is seen as the major policy target. Table 1 describes our union in terms of the differences it produces from our base forecast over the period 1991Q1 to 1999Q4. It appears relatively costless to unite France and Germany. The effects on demand and prices of lower French interest rates are offset by the effects of higher than base exchange-rate path. The slightly tighter fiscal stance in France lowers the level of output and in the transition to inflation convergence there is a loss of competitiveness especially against Germany. The process of adapting is however relatively slow. The process of adaptation for the UK is more costly in terms of lost output. The lower interest rate path and associated higher exchange rate produce a lower rate of inflation, and the fiscal policy tightening keeps the financial flows to foreigners approximately on the base path. (6)

Our model of the Italian economy suggests that they may find it considerably more difficult to join a monetary union than either the UK or France. This may of course reflect our inability to capture the recent structural changes in the Italian economy, and hence our message must be seen as cautionary, and any actual outturn in a union may be less painful. Our estimated equations suggest that although the union will help reduce inflation through the effects of lower trade prices it is not enough. Fiscal policy will have to be tightened both to reduce inflation in the short run and hence in the long run keep within bounds the financial flows required from foreigners. in order to achieve these aims it appears that italian growth would have to be held .75 of a per cent below capacity for some years. This might well be considered a very high price to pay in order to get inflation down to German levels, rather than keeping it 1 1/2 to 2 per cent higher. Eventually of course, the structure embedded in our model allows the italian economy to return to its capacity growth path, but the costs on the way to recovery could be large as the response lags, based on historical evidence, are liable to be slow.

The evaluation of a monetary union

The main objective of this note is not to evaluate the costs of forming a monetary union, but rather to analyse the results of abandoning the use of the exchange rate as a shock absorber. We present the effects of two shocks, each with and without a monetary union. The differing reactions to the shocks reflect the estimated structure of the model rather than the policy reactions of the authorities. We assume that for each country and for the union as a whole, the authorities target the real interest rate, and that it converges on its base level rapidly. The exchange-rate reactions are also similar across countries and in the union. The effects of interest rate changes on exchange rates are approximately the same everywhere, whilst the estimated effects of current balances on exchange rates are generally small and essentially similar across countries.

In order to bring out the role of exchange rates we have made a further assumption about fiscal policy. We assume that it is not used as a policy instrument. Fiscal policy responses would optimally differ between economies and in the union, and we would assume that their use would produce a better outcome in all circumstances. The first shock we consider is a 50 per cent rise in oil prices sustained for six years. Table 2 reports the effect on the members of the union whilst table 3 reports the effects that might be observed if their exchange rates were free to float. As table 2 shows, inflation rises everywhere in response to this shock, but a tighter monetary policy keeps the increase in bounds. Output is above base in Germany as both increased demand for exports from oil producers, and improved competitiveness relative to France and the UK raise the contribution of the external sector. This contribution rises over time. The oil exporting countries on our model respond only slowly to their increased revenue, and their import volumes (and hence world trade) keep rising for at least 12 years after such on oil price shock. It is this slow response to the shock that keeps inflation above base for some years after the shock.

If there were no union then the experience of the UK would be very different from that reported in table 2. As an oil producer, the UK's balance of payments would improve, and hence its dollar rate would rise (7), as we can see from table 3. As a consequence the UK would suffer less inflation outside the union than it would inside it. The reverse is true for France and Italy. Both respond rather rapidly to the price shock on our model, both inside and outside the union. The deterioration of their balance of payments would, if they were outside the union, lead to an exchangerate depreciation, and the consequent rise in inflation would cause interest rates to rise. The union exchange rate would not initially depreciate so strongly, and this would reduce the inflationary shock to at least Italy, but the rather slower response of prices to exchange rates that is embedded in our model for France mean that the union does not help them reduce inflation as much as the Italians. The consequent change in relative prices reduces the level of income in France.

Tables 4 and 5 report the effects of an expenditure shock in the US sustained again for six years. The level of government spending rises by 2 per cent of GNP, and as a consequence the interest rate in the US rises by 1.0 per cent throughout. As a consequence exchange rates in Europe initially jump downwards with and without the union, and interest rates rise. The demand shock raises inflation everywhere initially, both with the union and without it. Without the union the policy reactions in both France and Italy produce smaller downward movements in the exchange rate than they would experience under a union. This leads to both a smaller rise in output and a smaller inflationary response that they might experience under a monetary union run along the policy lines of the Bundesbank. As a consequence under the union France and italy slowly lose competitiveness vis-a-vis Germany, and hence German output rises more with the union than without it.


Our two sets of simulations bring out two main points, firstly that structurally dissimilar economies react differently to shocks, and secondly that policy reactions can be designed to allow economies to react differently. There are two major types of structural difference. Those that derive from comparative advantage, such as the existence of oil reserves, are immutable. Those that reflect differences in the behaviour of markets for, say, labour can of course be ameliorated. However, there do appear to have been large differences in structure between European economies, especially in relation to the speed of response of prices to shocks and in the degree of influence overseas prices have on domestic prices. If these structural differences remain large and persistent then the costs of membership of a monetary union may also be large.

Not all differences in structure need lead to differences in outcome if the countries are not members of a union. As our second set of simulations showed, it is possible to design simple policy rules in the face of external shocks that, through the diversity of exchange-rate responses, lead to similar outcomes in different economies. Abandoning the exchange rate as a shock absorber in a monetary union could lead to worse outcomes in response to shocks than we would see without a union. Only a serious attempt to understand and then reduce structural difference would allow the economies of Europe to reduce the costs of removing independent shock absorbers.


(1) There are a number of theoretical papers in the early, essentially Keynesian, traditional analysis using the simple Phillips curve concept, as in J. Marcus Fleming (1971). in many cases this is entirely derivative from the debate between R. Mundell, (1961) and R. McKinnon (1963).

(2) Another way to make this point is to look at the exchange-rate equations used in National institute models. We assume that the change in the exchange rate is determined by three factors, firstly the interest differential, secondly by a portfolio preference term such as the ratio of net overseas assets to GNP, and thirdly by a pure risk term. The equation may be written as:

[delta] log (RX) = r [sup.d] + a(NAR) + e where RX is the exchange rate, r [sup.d] is the interest differential and NAR is the net assets ratio. if RX is irrevocably fixed then this equation may be inverted to become a portfolio based equation for the interest rate: - r [] = r [sup. world] - a(NAR) - e The higher the country's net indebtedness the higher its interest rate.

(3) The institute has developed a 640 equation model of the world economy, and uses it for forecasting and research. The latest forecast is reported in Chapter 2 of this Review Each major European economy is modelled by around 60 equations, with sectors for demand, trade and the balance of payments, money and interest rates, and it also has a supply side set of wage and price equations with endogenous employment.

(4) it is difficult to calculate just what risk premia would apply under a monetary union, as the risks associated with a lack of credibility of the policy stance would be reduced. Over the period 1978-88 the ex-post interest differentials, after allowing for exchange-rate changes, between France and Germany and italy and Germany were 0.5 and 4.0 per cent respectively. The French differential is probably a reasonable estimate of the risk premium on assets located in France, but that for Italy reflects both the greater than anticipated success of policy, and also the influence of strong and strengthened capital controls on capital outflows. Both issues are discussed in R. Barrell and A. Gurney (1989). We have judged that risk premia will still exist inside our union, and they will be 0.5 per cent per annum for France, 1.0 per cent for the UK and 2.0 per cent for Italy.

(5) The base forecast was produced in February 1990 and is discussed in Chapter 2 of the February issue of the Review. it assumes that the EMS will progressively get harder, and already assumes a continual tightening of fiscal policy in both italy and the UK. The assumptions on fiscal policy for the UK are similar to those produced analytically in Simon Wren-Lewis (et. al) (1990).

(6) The oscillatory pattern displayed by the UK current balance reflects our assumptions that UK fiscal policy is tightened every other year and that the tightening is unanticipated.

(7) The UK exchange rate on GEM is defined as pounds sterling per dollar and hence a negative sign indicates an appreciation.


Barrell, R. and Gurney, A. (1989) `The World Economy', Chapter 2 of National Institute Economic Review, No. 130. November 1989.

Delors, J. (1989) Report on economic and monetary union in the European Community', Committee for the Study of Economic and Monetary Union.

European Commission (1 990) 'Economic and monetary union: the economic rationale and design of the system', Bank for International Settlements Review, No. 62, March.

Fleming, J.M. (1971) 'On exchange rate unification', Economic Journal, No. 18, September, pp. 467-88.

McKinnon, R. (1963) 'Optimal currency areas', American Economic Review Vol. 53, pp. 717-24.

Mundell, R. (1961) `A theory of optimal currency areas', American Economic Review, Vol. 51 pp 657-65.

Wren-Lewis, S., Westaway, P., Soteri, S., and Barrell R., (1990) `Choosing the rate: an analysis of the optimum level of entry of sterling into the ERM', National Institute Discussion Paper No. 171.
COPYRIGHT 1990 National Institute of Economic and Social Research
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:European monetary system
Author:Barrell, Ray
Publication:National Institute Economic Review
Date:May 1, 1990
Previous Article:The evolution of special employment measures.
Next Article:International labour costs in manufacturing, 1960-88.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters