Most economists agree that the EU as a whole does not form an optimal currency area. This means that countries will benefit differently from joining the EMU. It is therefore highly important to find out what factors are likely to determine whether a given country will benefit or lose.

There are basically two questions: 1) what factors make

This paper will firstly look at the factors determining the costs of joining. Then it moves on to analyse the factors that determine the benefits.

Distinction long/short run.

I What is likely to determine the cost of giving up the exchange rate ?

In this first part we have a look at the cost side of the cost-benefit analysis. In so doing, we mainly rely on the 'Optimum Currency Area' Theory. This Theory has been pioneered by Mundell (1961), McKinnon (1963) and Kenen (1969).

It precisely concentrates on the cost side. Its main insight is to say that countries with very different economic structures (labour market, economic preferences, growth rates, and fiscal systems) should not form a monetary union. In this line, we will try here to determine what factors increase the cost of EMU.

A - The basic cost of EMU : managing asymmetric shocks without monetary instruments

1 - Monetary Union implies to lose control over monetary policy

The costs of a monetary union derive from the fact that when a country joins, it loses the ability to conduct a national monetary policy, i.e. :

- to chose the price of its currency (devaluation/revaluation, use of exchange rate)

- to chose its quantity (money supply, determined by the interest rate)

2 - Monetary policies are useful against asymmetric demand shocks (Mundell 1962)

Mundell´s theory of Optimal Currency Area assumes that nominal exchange rates have real effects as otherwise there would be no costs in a nation giving up its own currency. This is typically a Keynesian view of the world.

Let us consider the case of a demand shift developed by Mundell in his celebrated article on OCA. For some reason EU consumers shift their preferences away from French-made to German-made products.

France Germany

The result is that the output declines in France and that it increases in Germany. In France the value of output has declined as a result of the shift in the aggregate demand. If spending by French Residents does not decline by the same amount, France will have a current account deficit. This is the most likely outcome, since the social security system automatically pays unemployment benefits.

Current account = domestic output - domestic spending.

In Germany, the situation will be the reverse. It will show a surplus on its current account. Both countries will have an adjustment problem :

- France is plagued with unemployment and a current account deficit

- Germany experiences upwards pressures in its price level and it accumulates current account surplus

Germany will now face a dilemma situation. If the authorities try to reduce inflation, then the restrictive monetary and fiscal policy would increase its account surplus. If they try to get ride of the account surplus, they would have to accept higher inflation.

The dilemma can only be solved by revaluing the mark against the Franc. The effect of the revaluation is that the demand curve shifts back to the left. In France the opposite occurs. The devaluation of the Franc, increases the competitiveness of the French products. This shifts the aggregate demand curve upwards.

France Germany

Thus France solves its unemployment problem, and Germany avoids accepting inflationary pressures. At the same time the current account deficit of France and surplus of Germany tend to disappear.

3 - Monetary policies are useful against asymmetric supply shock

Similarly, in the case of an asymmetric supply shock the exchange rate can correct again unbalances. Let's imagine that a sudden increase in the price of production inputs affects all EU countries. In Germany, the centralised labour union accepts a wage decrease meanwhile French labour unions do not accept a similar decrease. This ends up in an asymmetric supply shock, in favour of Germany, and therefore leads to the same kind of macroeconomic problem as before, with the same solution : devaluation of the French Franc.

4 - re-wording of the question

The question of what factors increase the costs of a monetary union therefore amounts to the question of what factors increase the utility of monetary policy, which is frozen under monetary union.

As we have seen, we are mainly concerned with the management of asymmetric shocks. The theory of OCA precisely sets out the conditions under which joining a monetary union is not costly. The cost directly depends upon the capacity of the economies to deal with asymmetric shocks without using monetary policy.

We shall now systematically set out the factor influencing the capability of a country to deal with asymmetric shock without monetary instruments.

B - The importance of the labour market for asymmetric shocks

OCA Theory says that in principle, adjustment problems disappear automatically if wages are flexible and/or if the mobility of labour is sufficiently high. Moreover, institutional features of the labour market, as social security and trade unions are also very important

1 - wage flexibility

Wage flexibility is the first mechanism, which can bring back equilibrium in the two countries affected by asymmetric shocks.

a) case of an asymmetric demand shock

In the context of the aforementioned example, French workers who are unemployed will reduce their wage claims. In Germany, the excess demand for labour will push up the wage rate. The effect of this adjustment mechanism is shown in Fig 1.2 :

In France, the price of output decline, making French product more competitive, improving the supply capacity, and stimulating demand. The opposite occurs in Germany. Therefore, the French current account improves and the German surplus is reduced, decreasing the inflationary pressures.

Conclusion : the greater the wage flexibility, the lower the cost

b) case asymmetric productivity growth rates

Let us assume that there is a differential in productivity growth between countries. In that case, nominal wage growth between countries should be different and should reflect the differences in productivity growth between these countries. This notably implies that central wage barraging would be harmful.

2 - Labour Mobility

A second mechanism that will lead to a new equilibrium following an asymmetric shock involves labour mobility. The French unemployed workers move to Germany where there is excess demand for labour. This movement of labour eliminates the need to let wages decline in France and increase in Germany. Both the French unemployment and the German inflationary pressures vanish.

The current account desequilibria will also decrease. The reason is hat the French unemployeds were spending on goods and services before, without producing anything.

Conclusion : the greater the labour mobility, the lower the cost

3 - Institutional features of Trade Unions : centralised or decentralised ?

a) the case for centralised wage bargaining Þ institutional differences can transform a symmetric into an asymmetric shock.

There are important institutional differences in the labour markets of European countries. Some labour markets are dominated by highly centralised labour unions (e.g. Germany). In other countries labour unions are decentralised (e.g. the UK). These differences may introduce significant costs for a monetary union.

The main reason is that these institutional differences can lead to divergent wage and price developments, even if countries face the same disturbances (symmetric shock).

For example when two countries are subject to the same oil price increase the effect in the domestic wages and prices very much depend on how labour union react.

b) the case for centralised wage bargaining Þ better to respond to a supply shock (Bruno & Sachs)

BRUNO & SACHS (1985).

Recent macroeconomic theories over the importance of labour market institutions. The idea can be formulated as follows. Supply shocks have very different macroeconomic effects depending on the degree of centralisation of wage bargaining.

When wage bargaining is centralised ('corporatist' countries), labour unions take into account the inflationary effect of wage increase. They have no incentive to make excessive wage claims.

In countries with less centralised wage bargaining, individual unions know that the effect of the nominal wage increase they claim, only have a very small impact on the aggregate price level. There is a free-riding problem. Everyone has an incentive to claim more. In equilibrium, this non co-operative game will produce a higher nominal wage level than the co-operative (centralised) game.

c) the case for decentralised wage bargaining Þ very decentralised wage bargaining are also able to manage a supply shock (Calmfors & Driffill)

CALMFORS & DRIFFILL (1988)

This co-operation story has been extended by noticing that in a very decentralised system (e.g. wage bargaining at the firm level) the labour unions know that their wage claims will have a direct effect on the competitiveness if the firm, and therefore on the employment prospect of individual union member.

Thus, unions in such a decentralised system may exhibit a considerable degree of wage restraint.

Conclusion : The more extreme the centralisation/decentralisation, the lower the cost.

· countries with either strong centralisation or strong decentralisation of wage bargaining are better equipped to face supply shocks (e.g. 1979-1980) than countries with an intermediate degree of centralisation. Empirical evidences support this idea.

d) the case for decentralised wage bargaining Þ in the presence of asymmetric shocks (e.g. productivity differential) wages should be flexible

e) the case for decentralised wage bargaining Þ better to manage regional asymmetries

Although in a unified Europe asymmetric shocks between countries may become less prevalent, specialisation would still lead to large regional divergences. These regional divergences may even increase in a unified Europe.

The characteristic feature of this specialisation is that it would cross borders. In such an environment a centralised wage bargaining system would be harmful for many European regions.

f) the case for decentralised wage bargaining Þ better to manage asymmetric shocks between different sectors

Technological changes tend to lead to uneven changes in output and employment changes between sectors. A cross-sectoral wage bargaining system would therefore be harmful for sectors tat experiences less favourable developments.

g) the case for decentralised wage bargaining Þ better to manage asymmetric shock within a given sector

There is even a lot of empirical evidence indicating that many of the asymmetric shocks occur within the same sectors. Again, a centralised wage bargaining system would be very detrimental to output and employment for firms that lag behind other ones in the same sectors.

Conclusion : on the whole, a centralised wage bargaining at the European level is not desirable.

4 - Social protection

In the case of asymmetric unemployment between two countries, a current account desequilibria emerges. As we have seen, this is partly because of the unemployed people can still spend money without producing, thanks to the social security system which compensate for losses due to unemployment (current account = national output - national spending i.e. absorption). This means that a way to adjust without monetary instrument would be to cut of social protection.

High standards of social protection would maintain the level of imports in case of a negative and asymmetric demand shock increasing unemployment and reducing the level of output. Therefore, a current account deficit would emerge.

Conclusion : the lower the social protection, the more a country is able to adjust without monetary policies, the lower the cost. (This may have tremendous implications in the future)

C - The importance of a centralised budget and of political union in dealing with asymmetric shock

1 - direct fiscal transfers as a solution to asymmetric shocks

Let us consider the previous example whereby France experiences a negative demand shock and Germany a positive demand shock. The consequences in France are unemployment, and a current account deficit. In Germany, the wages and therefore the prices go up (inflationary pressure) and a current account surplus appear.

In the absence of any adjustment of the exchange rate, of wage flexibility and of labour mobility, an adjustment can still be produced through a fiscal transfer. The German authorities could increase the taxes in Germany so as to reduce the aggregate demand. These tax revenues could thereafter be transferred to France (inducing an upward shift of the aggregate demand).

2 - This solution is already applied within nation states, between regions

This solution to the problem is still difficult to contemplate between sovereign nation states, especially since it would have to be repeated every year if the demand shift is a permanent one.

This solution however is frequently applied between regions of the same nation. Many countries have implicit or explicit regional redistribution schemes through the federal budget.

· Implicit regional redistribution within a nation occurs because of the fact that a large part of the government budget is centralised. As a result, when output decline in a region :

- the tax revenue of the federal government from this region decline

- at the same time the social security system (which is quite often centralised) will increase transfers to this region (e.g. unemployment benefits).

Þ the net result of all this is that the central budget automatically redistributes in favour of regions whose income declines.

· In some stated there also exist explicit regional redistribution schemes.

Probably the best known is the German system of 'Finanzausgleich'. This system came into existence after the second world war. Its basic philosophy is that Landërs whose tax revenues fall below some predetermined range should receive compensation from Landërs whose tax revenues exceed that range. As a result, there has been a remarkable reduction of regional income inequalities in Germany.

3 - However fiscal transfers are only suitable for temporary (not permanent) demand shocks

Fiscal transfers between regions do not solve the adjustment problem following an asymmetric shock. They just make life easier in the country region experiencing a negative demand shock.

When the demand shock is a permanent one, price and wage adjustments and factor mobility will be necessary to deal with the problem.

Thus, fiscal transfers are only suitable to deal with temporary shocks in aggregate demand. Actually if fiscal transfer take on a permanent character, they may even make the adjustment to permanent demand shocks more difficult (as they become a substitute for wage and price changes and for mobility of labour).

4 - The more the political unity, the lower the costs

As we have seen, having a centralised budgetary process helps a lot. It enables monetary transfers, in order to handle asymmetric shocks.

Therefore, there is an economic rationale for a greater political unity. Indeed, only such a political unity enables the monetary union to organise transfers without costly political bickering.

As Paul de GRAUWE puts it : " the continued existence of nation states is an important source of asymmetric shocks ".

Conclusion : the greater the political union, the lower the costs and risks

D - The importance of growth rates differential within EMU

Some countries grow faster than others do. E.G. during the 1980s some southern European countries (and Ireland) experienced growth rates of their GDP which were higher than in the northern part of Europe.

1 - Trade balance problems for fast growing countries of a monetary union

Such difference in growth rates could lead to a problem when countries form a monetary union. Suppose that Country A's GDP is growing at 5% per year. Suppose that the income elasticity of A's import from B is one, and similarly that B's income elasticity of import from A is equal to one.

Then country A's imports from B will grow at 5% per year, whereas B's imports from A will grow at only 3% per year. This will lead to a trade balance problem of the fast-growing country A, whose imports tend to grow faster than its exports.

2 - The necessity to alter the terms of trade (devaluation or deflationary policy)

In order to avoid chronic deficits of trade account, country A will have to reduce the price of its exports to country B, so that the latter country increases its purchases of goods from country A. In other words, country A's terms of trade must decline so as to make its products more competitive. Country can do this in two ways :

- a depreciation of the currency

- or lower rate of domestic price increases than in the country B

If it joins a monetary union with country B, only the second option is available. This will require country A to follow deflationary policies, which will in turn constrain the growth process.

Conclusion : the monetary union had a cost for the fast-growing country. The very fast growing economies (Ireland ?) are likely to be penalised.

II What is likely to determine the cost of giving up control over inflation rate ?

There are different reasons why different countries may have different optimal inflation rates. One is that they may have different fiscal systems and therefore different seigniorage needs. Another it that may simply have different preferences over the combination of inflation and unemployment they want (in a Keynesian view, where the Phillips curve holds)

A - Different fiscal system, different optimal inflation rate, and the seigniorage problem

1 - Fiscal systems and optima inflation rates

Countries differ also because they have different fiscal systems. Therefore they use different combination of debt and monetary financing of the government budget deficit. When these countries join a monetary union they will be constrained in the way they finance their budget deficit, especially as regard their use of seigniorage. Seigniorage is a way to finance a deficit by raising revenue by inflation (i.e. issuing money).

The theory of optimal public finance tells us that rational government will use the different sources of revenue so that the marginal cost of raising revenue through these different means is equalised :

G - T + rB = dB/dt + dM/dt

Where G is the level of government spending, T is the tax revenue, r is the interest rate, B the government debt. The right hand-side is the financing side. The budget deficit can be financed by issuing debt (dB/dt) or by issuing high-powered money (dM/dt).

All this means that different countries will have different optimal inflation rates. In general, countries with an underdeveloped tax system will find it more advantageous to raise revenue by inflation (seigniorage).

2 - The cost of EMU : limits the use of seigniorage

By joining a low inflation northern monetary zone, southern EC countries will have to increase taxes. For these countries the cost of a monetary union is that they will have to rely too much on a costly way of raising revenue.

Empirical evidence : since the 1980s, seigniorage revenue in these countries has declined (reduction of inflation and improvement of tax system)

Conclusion : the more efficient the fiscal system, the lower the cost (C.F. extra costs for southern countries)

B - Preferences over inflation and unemployment

1 - Different preferences over inflation / unemployment

Countries differ also because they have different preferences over inflation and unemployment. Let us assume that the Phillips curve is valid in the long run, and that each country has to truly chose what combination of unemployment and inflation it wants (this is a Keynesian, non monetarist view of the world).

2 - The cost of EMU : one inflation rate for all

If two countries decide to join a monetary union, the exchange rate is fixed. Therefore, the rates of inflation should be equal. If it is not the case (e.g. German inflation is lower than the Italian inflation), then Italy will increasingly lose competitiveness.

The cost of a monetary union for two countries now consists in the fact that the inflation rate will be the same (for competitiveness problems)

However, countries have 1) different Phillips curves 2) different preferences. In order to equalise their inflation rates, they will have to choose another (less preferred) point on their Phillips curve.

III A RECENT REASSEMENT OF THE COSTS

In this section we propose a criticism of the theory of OCA, notably developed by Gros and Thygesen (1991). This criticism has been formulated at three different levels (A, B1, B2).

A - Do differences between countries really matter ?

There is no doubt that countries are different. To what extent can these differences affect the economic functioning of monetary union ?

1 - Is a demand shock concentrated in one country a likely event ?

There are two ways to answer such a question.

a) over time, national economies will converge towards similar economic structure

There is still a vivid debate on this point. Roughly one can oppose two view.

· The European Commission's view

According to the European Commission, differential shocks in demand will occur less frequently in a monetary union. This is because trade between the industrial nations is to a large extent intra-industry trade. The trade is based on the existence of economies of scales and imperfect competition (product differentiation). It leads to a structure of trade in which countries buy and sell to each other the same categories of products. Thus France, sells cars to and buy cars from Germany. And so does Germany.

The removal of barriers with the completion of the single market will reinforce these tendencies. As a result, most demand shocks will tend to have similar effect, that is to be symmetric (and not asymmetric).

· Krugman's view

The opposite view has been defended by Paul Krugman (1991). In his view, one cannot discard Mundell's analysis, for there is another feature of the dynamics of trade with economies of scale : trade integration (with economies of scale) leads to regional concentration of industrial activities.

When impediment to trade decline, this has two basic effects on the localisation of industries:

- it makes it possible to produce closer to the market

- it makes it possible to concentrate production so as to profit from economies of scale.

Þ trade integration may lead to more concentration of regional activities than less.( e.g. in the USA, the production of automobiles is far more concentrated than in Europe. The feature is found in many other industrial sector).

· Empirical Assessment

What is the right view of the world ? There is a reasonable presumption in favour of the European Commission view.

Theoretically speaking, economists cannot settle the question. However, some empirical studies have shown that their is a firm relationship between strong trade linkage and tightly correlated economic activity (see FRANKEL & ROSE, 1996). Similarly, ARTIS & ZHANG (1995) found that as the European countries have become more integrated during the 1980s and 1990s, the business cycles of these countries have become more correlated.

B) the national unit is less and less pertinent : asymmetric shocks occur more and more at the regional, sectoral or even intra-sectoral level. Monetary policy cannot deal with this type of shocks.

· Indeed, asymmetric shocks occur less and less at the national level. However, they occur :

- at the regional level, regardless national borders

- at the sectoral level, affecting a certain type of industry throughout Europe (e.g. change in technology)

- at the intra-sectoral level (certain firms lag behind others of the same sector)

Manipulating the national currency is not a suitable way to handle such non-national asymmetric shocks.

· The implication of Krugman's view

What Krugman seems to forget is that it becomes more and more likely that the relevant regions in which some activities is centralised will transgress one or more borders.

Here we do not deny that integration may entail more regional specialisation. The argument is rather that national borders will increasingly be less relevant in influencing the shape of these regions. An asymmetric shock hitting a cross-border economic region, cannot be dealt with by a national monetary policy.

Trade integration and utility of the exchange rate

2 - Do differences in growth rates really matter ?

Fast growing countries experience fast growing imports. In order to allow their exports to increase at the same rate, these countries will have to make their exports competitive either by depreciation of their currency (option non-available in a monetary union) or by a deflationary policy (in order to have a lower rate of inflation). As a result these countries will be constrained in their growth.

This popular view has very little empirical support. An IMF study over the period 1976-1995 shows a lack of clear relation between economic growth and real depreciation.

An elegant interpretation has been given by Paul Krugman. Fast growing countries are those which develop and exports new products. The result is that the income elasticities of the exports of fast-growing countries are typically higher than the income elasticity of their imports. As a result these countries can grow faster without incurring trade balance problem.

Þ differences in the growth rates cannot be considered as an obstacle to monetary integration.

3 - The factors that really produce asymmetric shocks

a) The continuing existence of nation states

An important source of asymmetric shock will remain, namely the continuing existence of nation states which exercise considerable sovereignty in a number of economic areas. The two most important fields of sovereignty are :

· The budgetary field

In the monetary union, most of the spending and taxing powers will continue to be vested in the hands of national authorities. Today, in most EU countries spending and taxation by the national authorities amount to close to 50 percents of GDP. The spending and tax power of the European authorities represent less than 1,5 percent of GDP.

By changing taxes and spending the authorities of an individual countries can create large asymmetric shock. By their very nature these shocks are well contained within national borders. This raises the issue of how budgetary policies should be conducted in a monetary union.

· The wage bargaining process

The wage bargaining process differs widely between countries and can create asymmetric shocks.

· Conclusion

For some economists, a monetary union can only function satisfactorily if further steps towards political unification are taken. In this view, the absence of political union will create great risks of difficult adjustment of political disturbances.

It is equally possible that the existence of a monetary union will exert sufficient pressure on the MS to accelerate their efforts towards a political union.

b) The institutional differences in labour market

The institutional differences in the national labour markets will continue to exists for quite some time after the introduction of a common currency. This may lead too divergent wage and employment tendencies, and sever adjustment problems when the exchange rate instrument has disappeared.

B - The long run cost of losing the instrument of monetary policy : a Monetarist view (zero cost in the long run)

1 - The long-run ineffectiveness of exchange rate changes in handling asymmetric shocks

Nominal exchange rate changes have only temporary effects on the competitiveness of countries. Over time, the nominal devaluation leads to domestic cost and price increase which tend to restore the initial level of competitiveness. The short-term gain in price competitiveness of exports is soon compensated by the increase of import prices and therefore of production costs.

In other words, nominal devaluations only lead to temporary real devaluation. In the long run, nominal exchange rate does not affect the real exchange rate of a country. Devaluation does not have a permanent effect on competitiveness and output.

2 - Losing the inflation rate : a Monetarist view (zero cost)

What is the true cost of having a single inflation rate for very different countries. We have seen that different countries may have different preferences over their optimal combination of inflation and unemployment.

This idea, of course, relies on the assumption that each country has to arbitrate between inflation and unemployment. This Keynesian idea is embodied in the Phillips curve.

The core of the monetarist critique is that countries which choose too high an inflation rate will find that their Phillips curve shift upwards. In the monetarist view of the world, the Phillips curve is a vertical line in the long run. Therefore, in the long run, the authorities cannot choose an optimal combination of inflation and employment.

Þ therefore, there is nothing to lose in having a unique inflation rate for all Europe. This analysis is now generally accepted.

C - However, monetary policies can be efficient in the short run.

1 - In the short run, devaluation can be of great use

Even though in the long run, money seems to be a veil and is ineffective in solving a long-term asymmetric shock , it remains that the instrument of exchange rate is useful in the short run.

The alternative policy to devaluation in case of a negative demand shock considered by Mundell, is a policy of expenditure reduction. Indeed, such a policy reduce the current account deficit by reducing the level of absorption (current account = value of national output - value of national spending (absorption)).

The major difference between the two policies is to be found in their short run dynamics. The cost of devaluation is that there will be inflation. The cost of the second policy (expenditure reduction) is that output decline during the transition period.

For an example of successful devaluation, see the devaluation of the Belgian Franc in1982.

2 - Therefore, there is a real loss in giving up monetary policy

The greater cost in the increase difficulty to manage large scale and unexpected asymmetric shocks. The adjustment processes will be necessarily more painful, even though the long-term equilibrium is likely to be unchanged.

D - A synthetic criteria : openness of a country and costs of EMU

1 - The relationship between costs of a monetary union and openness of a country

· First there is a relationship between the degree of openness and the occurrence of asymmetric shocks. We have presented two view (Commission v Krugman).

· The second idea which matters has to do with the degree of effectiveness of the exchange rate in dealing with asymmetric shocks.

Let us consider two countries, one relatively open, the other relatively closed. As far as the demand-side effects are concerned the same devaluation has a stronger effect in the relatively open economy. As regard the supply-side effects, the upwards shift of the supply curve following the devaluation is more pronounced for the relatively open economy (more inputs are imported at a higher cost).

Þ we cannot say a priori in which country the devaluation is most effective in stimulating the output.

Þ however, it is clear that a systematic use of the exchange rate instrument will lead to more price variability in the more pen economy.

Þ that price variability involves costs

Þ for the same effect in output, the devaluation is likely to be less costly for the relatively open economy than for the relatively close one

Þ the cost of a monetary union lost likely declines with the degree of openness of a country.

The cost of a monetary union and the openness of a country

p.49

What will determine the extent to which a country will suffer from this costs ?

IV WHAT IS LIKELY TO DETERMINE THE BENEFITS ?

A - A set of microeconomic benefits

Whereas the costs of a common currency have much to do with the macroeconomic management of the economy, the benefits are mostly situated at the microeconomic level. Moving to a common currency can be expected to lead to gains in economic efficiency.

1 - Elimination of transaction costs

a) The most visible gain

Eliminating the costs of exchanging one currency into another is the most visible gain, and most easily quantifiable) gain from a monetary union.

We should note however, that it is only when national moneys are replaced by a common currency, that transaction costs will be eliminated. As long as national currencies remain in existence (even if the exchange rate is 'irrevocable') currencies will not be perfect substitutes. The need to convert will remain.

b) Estimation of the gains

The European Commission has estimated these gains, and arrives at a number between 13 and 20 billions per year. This represents between 0.25 and 1 percent of the Community GPD. This may seem peanuts, but still constitutes a real gain.

c) Transition problem for the banks

The banks will have a problem of transition, since approximately 5 percents of their revenues are the commission paid in exchange of national currencies. When this has been done, society will have gained (employees previously engaged in exchanging money will be free to perform more useful tasks to society).

2 - Price transparency : less price discrimination

a) Price transparency v price discrimination

· The elimination of transaction costs will also have an indirect gains : it will reduce the scope for price discrimination between national markets.

· There is la lot of empirical evidences that price discrimination is still practised widely in Europe. (e.g. in 1995, an automobile was on average 28 percents more expansive in Germany than in Italy).

b) ROGERS's analysis shows the importance of currencies in price discrimination

· Richard ROGERS (1995) analysed the factors that influence the price differential of the same good in different locations, in the US and in Canada. He found that 1) firstly distance matter 2) borders powerfully segment markets, largely due to the difference in currencies.

Þ having the same currency truly unify the market, leaving less room for price discrimination.

3 - Less uncertainty & risk : a better price system (better allocation of resources)

a) a better price system, a better allocation of resources

· Exchange rate uncertainty introduces uncertainty about the future prices of goods and services. Economic agents base their decisions concerning production, investment and consumption on the information that the price system provides for them. If the prices become more certain, the quality of these decisions will improve.

· In other words, the price system becomes a more reliable mechanism to allocate resources, a better guide to make the right economic decisions. These efficiency gains are difficult to quantify, even though they are probably very important. Let us consider two situations.

b) The importance of an efficient price system

· A way to show the importance of price certainty is to consider the investment decision of a firm in a foreign country. The exchange rate on which the decision is made may be wrong ; this forecast error may make the investment unprofitable.

· Another way to show the important of price certainty, is to consider the extreme situation of hyperinflation. We observe that in countries experiencing hyperinflation the wrong production and investment decisions are made in a massive scale. Quite often we observe strong increases in output and investment, but very frequently in the e wrong sectors or product lines. After a while these production have to be abandoned. Massive amount of resources is wasted in the process.

4 - Less uncertainty & risk : a lower real interest rate

a) lower interest rate: decrease in the risk premium

· An increase in risk, due to price uncertainty, will in general increase the real interest rate. Indeed, risk-averse investors require a higher risk premium to compensate them for the increased riskiness if the project. The lower the risk, the lower the interest rates, the better for entrepreneurs.

b) less risky investment projects are selected

· Moreover, we should not that higher interest rates lead to less efficient ways in selecting investment projects. Indeed, due to problems of moral hazard and adverse selection higher interest rates led to the selection of more risky investment projects. Thus the systemic risk increases.

· Conclusion : eliminating the risk by moving towards a common currency reduces the amount of risky projects that are selected by the market.

5 - Less uncertainty & risk : an ambiguous effect on growth rate

a) Monetary union is largely believed to boost growth rates

Many economists have argued that the elimination of the exchange risk, by permanently lowering the interest rate, will lead to an increase in economic growth. This analysis features prominently in the EC Report 'One Market, One Money' (1990) which in turn was very much influenced by Baldwin (1989).

This can be shown by the neo-classical growth model.

b) The neo-classical growth model predicts a temporary increase in the rate of growth

The neo-classical growth model is represented here. The horizontal axis shows the capital stock per worker, the vertical axis the output per worker. The line f(k) is the production function which has the usual convex shape (implying diminishing marginal productivities).

The equilibrium in this model is obtained where the marginal productivity of capital is equal to he interest rate consumers use to discount their future consumption. The equilibrium is represented by the point An where the line rr (whose slope is equal to the discount rate) is tangent to the production function.

Now assume that the elimination of the exchange risk reduces the systemic risk so that the real interest rate declines. We represent this effect in the next graph. The reduction of the interest rate makes the rr-line flatter. As a result the equilibrium moves from A to B.

There will be an accumulation of capital and an increase in the growth rate while the economy moves from A to B. In the new equilibrium, output per worker and the capital stock per worker will have increased.

However, the growth rate of output then returns to its initial level, which is exogenous (determined by the rate of technological change and the rate of growth of the population). Thus, in this neo-classical growth model, the effect of a monetary union is a temporary increase in the growth rate, du to a reduction in interest rate.

c) The new neo-classical growth model (with endogenous growth) predicts a permanent higher growth path

The previous model has been extended by introducing dynamic economies of scale (ROMER, 1986). Suppose the productivity of capital increases when the capital stock increases (learning effect, accumulation of knowledge). In this new growth model the growth path is endogenous, and is sensitive to the initial conditions.

A lowering of the interest rate can put the economy on a permanent higher growth path. The economy first moves from A to B, and then to C (after the increase in capital productivity embodied by the shift of the f(k) line)

d) However, this prediction is probably too optimistic

The previous analysis ignores that a reduction of the exchange rate variability has a double effect :

1) it reduces the real interest rate (which in the previous analysis generated the growth effect

2) it also reduces the expected return of investment

Þ the reduction of risk has an ambiguous effect on investment activity, and thus also on the growth of output.

Paul de Grauwe : this point is not generally recognised by the public. Quite often it is stated that a reduction of the risk will boost investment activity. Economic theory does not allow us to draw this conclusion.

e) the empirical evidences are even more pessimistic

Most of the many econometric studies that have been performed recently tend to confirm that the degree of exchange rate variability has a very weak impact on the growth rate of investment, trade and output.

One has only to consider the growth rates of EMS countries during the 80s as compared to those of non-EMS countries. It is clear that the greater exchange stability that the EMS countries have experienced does not seem to have provided a great boost to the growth rates of output and investment.

f) a decrease in the exchange rate uncertainty does not entail a decrease in the systemic risk (W. POOLE, 1970)

The reduction in exchange rate uncertainty may not necessarily reduce the systemic risk. Less exchange rate uncertainty may be compensate by greater uncertainty elsewhere (output, employment). A greater monetary zone may therefore not be a less risky environment.

There is a whole theoretical literature, starting with William POOLE (1970).

Conclusion (Paul DE GRAUWE) : the benefits of a monetary union are to be found elsewhere than in its alleged growth-stimulating effects.

B - What will determine the extent to which a country benefits from them ?

1 - Relationship between benefits and openness of a country

As in the analysis of the costs of EMU, we can derive a relationship between the benefits of a monetary union and the openness of a country.

The welfare gains of a monetary union that we have identified are likely to increase with the degree of openness of an economy :

- the elimination of transaction costs will weight more heavily in a country that buy and sell a larger fraction of goods and services in foreign countries.

- Price transparency will have a deeper impact on a economy which is already used to deal with external markets.

- Reduction in exchange rate uncertainty entails a larger welfare gain in small and open economies (rather than in large and closed economies)

2 - Graphic representation

We can represent this relationship between the benefits of a monetary union and the openness of the countries graphically. On the horizontal axis we show the openness (share of external trade in the GDP). On the vertical axis, we represent the benefits (as percentage of GDP). The relationship is positive.

V WHAT IS LIKELY TO DETERMINE THE BALANCE ?

A - Determination of the critical threshold of openness

1 - A unified model of cost/benefit analysis

It is useful to combine the figures relating benefits and costs to the openness of a country. The intersection point of the benefit and the cost line, determines the critical level of openness that makes it worthwhile for a country to join a monetary union with its trading partners.

To the left of that point, the country is better off keeping its national currency. To the right, it is better off when it relinquishes its national money.

Costs and benefits of a monetary union according to openness (trade integration)

2 - The shapes of the curves (brief reminder)

· The benefit-line

· The cost-line

3 - The slope of the cost-curve : the role of the theoretical framework

The slope of the cost curve depends largely to a large extent on one's view about the effectiveness of the exchange rate instrument in correcting fir the effects if asymmetric shocks (in demand and supply - i.e. costs).

The monetarist view of the cost-line The Keynesian view of the cost-line

a) the monetarist view

At one extreme there is the monetarist view of the economy, which claims that exchange rate changes are ineffective as instrument to correct for different demand and cost development in different countries.

If one is sufficiently monetarist, one could argue that for countries with low degree of openness, the benefits could still outweigh the costs.

b) the Keynesian view

At the other extreme, there is the Keynesian view that the world is full of rigidities (wages and prices are rigid, labour is immobile), so that the exchange rate is a powerful instrument in eliminating desiquilibria. This view is well represented by the original Mundell model previously discussed.

In this view the cost curve is far away from the origin. Therefore, relatively few countries should find it in their interest to join a monetary union.

c) domination of monetarism since the early 80s in Europe

Since the early 1980s, the monetarist view has gained adherents, and has changed the view of many economists about the desirability of a monetary union. Today the consensus seems to have evolved favouring monetary unification, certainly in Europe.

4 - The position of the cost-curve: the role of the labour market

The cost-benefit calculus of a monetary union is also very much influenced by the degree of wage and price rigidities. A decline in wage and price rigidities has the effect of shifting the cost-line downwards. As a result, the critical point at which it becomes advantageous for a country to join is lowered. More countries can become candidates for a monetary union.

Impact of an increase in wage flexibility or labour mobility:

5 - The position of the cost-curve: the role of asymmetric shocks

Not only the degree of labour market flexibility (wage and labour mobility) matter for determining the position of the cost-curve, but also the size and frequency of asymmetric shocks.

Countries which experience very different demand and supply shocks will find it costly to form a monetary union. In the framework of our cost/benefit graph, this means that the cost-line shifts to the right.

B - Determination of the suitable combinations of divergence and flexibility

1 - A new representation

We are now in position to analyse the relation between labour market flexibility (wage and mobility) and asymmetric shocks, in a monetary union. This is done in the following graph :

- on the vertical axis, we set out the degree of 'real divergence' between countries. (With real divergence is meant the degree to which growth rates of output and employment tend to diverge as a result of asymmetric shocks)

- on the horizontal axis we have the degree of flexibility of the labour market in these countries. (relates to wage flexibility and labour mobility)

The central insight of the theory of the Optimal Currency Areas is that countries that experience a high divergence in output and employment growth need a lot of flexibility in their labour markets if they want to benefit from joining a monetary union. The larger the degree of 'real divergence', the greater the need for flexibility.

This relationship between real divergence and flexibility is represented by the upward sloping line AA. Countries or regions located below the AA line can benefit form a monetary union.

2 - What set of countries is EMU good for according to this criteria ?

a) The EU - 15 is not an OCA

There is a broad consensus among economists that the EU is not an OCA. See EICHENGREEN (1990). DE GRAUWE and HEENS (1993). Thus, according to most empirical studies the EU as a whole is located above the AA line. From an economic point of view, a monetary union involving all EU member countries is a bad idea.

b) The EU - 5 is an OCA

Whereas there is a strong consensus that the EU as a whole should not form a monetary union, there is an equally strong consensus that there is a subset of countries which form an OCA. The minimum set of countries that could form a monetary union is generally believed to include Germany, the Benelux countries and France (EU-5). This conclusion is buttressed by the same empirical studies.

c) Recent empirical analysis tend to enlarge the group of countries that would benefit

Recent empirical analysis, however, has tended to enlarge the group of EU countries that would benefit from monetary union. The study of ARTIS and ZHANG (1995) shows that the business cycles of EU countries have become much more correlated since the early 1980s. There are now fewer asymmetric shocks among a relatively large number of EU countries than 15 years ago. (This confirms the Commission's view of divergence & integration, at the expense of the Krugman's view).

On the top of that, many recent studies tend to find that in most EU countries monetary policies are powerless to affect real variables. Finally, some other recent studies show that asymmetric shock in the EU countries occur at the sectoral level and not so much at the national level. These shock cannot be dealt with by exchange rate changes.

Thus the most recent empirical studies lead to more optimism. The number of EU countries that can form an OCA is much larger than economists thought before. The EU-10 (EU-5 + Ireland, Spain, Portugal, Italy, Austria) is probably an OCA.

C - In the long run, economic divergence and trade integration matter

1 - The model

On the vertical axis, we set out the degree of economic divergence. On the horizontal axis, we have the degree of trade integration between the countries.

The upward sloping line (called OCA) represent the combination of divergence and trade integration that makes a monetary union a break-even operation (benefits = costs). All the points to the right of the OCA line are points for which the benefits of a monetary union exceed the costs. We call it the OCA zone.

2 - The optimistic view (Commission)

The downwards sloping line (TT) says that as trade integration increases the degree of economic divergence between countries involved declines Þ the countries become more alike and face less asymmetric shocks (the Commission view).

As trade integration within the EU proceeds, this point will move downwards along the TT line. In this view, monetary union will over time be beneficial for all countries in the European Union. In this sense, monetary union is inevitable.

3 - The pessimistic view (Krugman)

In the Krugman analysis, economic integration leads to more economic divergence between countries, due to the process of regional specialisation. This is represented by the upwards sloping TT and TT' lines. We have to consider two possibilities.

a) The TT line slope is flatter than the slope of the OCA line (TT)

In this case, although today EU - 15 may not be an OCA, it will move into the OCA zone over time. In this case more integration leads to more specialisation and thus more asymmetric shocks. However, the benefits of a Monetary Union. Also increase steeply with the degree of integration.

Þ as a result, despite the increase in asymmetric shocks, more integration will lead us into the OCA zone.

b) The TT line slope is stepper than the slope of the OCA line (TT')

The second case is represented by the steep TT' line. Here integration brings us increasingly farther away from the OCA zone. This is so because the gains from a monetary union do not increase fast enough with the degree of integration.

4 - Conclusion

From the discussion of the Krugman model we conclude that even if integration leads to more asymmetric shocks, this may still lead to increasing net gains of a monetary union for the EU-15. We cannot exclude however the possibility that the process of integration will make monetary union more and more unattractive.

Conclusion

On the benefit side, what is truly decisive is the :

- degree of openness of the country.

On the cost side, three factors are crucial :

- the degree of openness of the country

- the characteristic of the labour market

- the degree of political union (for a centralised budget)

Finally any answer rely on two theoretical framework :

- the utility of the exchange rate (Monetarist v Keynesian)

- the relationship between integration and divergence (Krugman v Commission)