The Maastricht treaty gives the central bank a very high level of autonomy. There is certainly a strong economic rationale for it. However this very feature raises problems of democratic accountability which are not addressed in the Maastricht Treaty.

There is therefore a balance to be found between independence and accountability. We argue in this paper that procedures can be designed that accommodate both objectives.

I Political independence is necessary to ensure price stability

A - The BARRO-GORDON model : inefficiency of the political decision process

The intellectual case for political independence of the ECB has been much influenced by the BARRO - GORDON model (path-breaking article in 1983).

1 - Hypothesis of the model

· Rational expectation assumption (economic agents use all relevant information, ant they cannot be systematically wrong in making forecasts : on average, their forecasts are good).

· Monetarist view of the world : in the 'long term' Phillips curve is vertical. Therefore there is a natural rate of unemployment (NAIRU = Non-Accelerating-Inflation Rate of Unemployment)

· Monetary authorities are supposed to care about both unemployment and inflation :

- 'hard nosed' monetary authorities care more about inflation

- 'wet' monetary authorities care more about unemployment

· The model stars from the standard Phillips curve, which takes into account the role of inflationary expectation. We specify the curve as follow : U = Un + a( Pe - P).

U is the unemployment rate, Un is the natural unemployment rate, P is the observed rate of inflation and Pe is the expected rate of inflation.

Þ this equation expresses the idea that only unexpected inflation affects the unemployment rate.

· Politicians are short sighted and give a low weigh to future welfare loss.

2 -The time inconsistency of the optimal short term monetary policy: the lack of credibility

From these hypothesis it logically results that a government who cars about the short-term rate of unemployment has a strong incentive to create unexpected inflation.

In order to did so, the government has an interest in publicly announcing a zero inflation rate (in order to influence private actors expectation) and then to betray their trust by a loose monetary policy.

It is thus rational for the government to cheat. The government has an incentive to renege on its promise to maintain a zero inflation rate. This feature leads to the so-called 'time inconsistency' of the optimal monetary policy.

3 - Implications : the Phillips curve shifts upwards

Economic agents are likely to react by increasing their expectations of inflation. Thus during the next period the Phillips curve is likely to shift upwards.

The government should therefore evaluate the short-term gain from cheating against the future losses that result from the upward sifting of the Phillips curve.

4 - The problem with politicians...

Suppose now that it consists of short sighted politicians who gives a low weight to future losses and that decide to cheat again and again. The Phillips curve will go upwards until the equilibrium reaches a point situated on the 'long term' Phillips curve (which is vertical).

At this point, the authorities have no incentive any more to surprise economic agents. A move upward along the Phillips curve going through E would lead to an indifferent curve located higher and therefore to a loss of welfare.

5 - Outcome : governments are stuck in a sub-optimal equilibrium with too high an inflation rate

Therefore the political decision process leads the economy to an unattractive equilibrium whereby expected and actual inflation rates are both very high, with no gain in terms of unemployment which is at its natural level anyway.

6 - Conclusion : in the long run governments have an incentive to acquire a reputation of low inflation. Importance of credibility.

Politicians are supposed to give a strong weight to the short-term gains of inflationary policies and a low weight to future welfare looses (du to too high an inflation rate).

A government which is known to care about unemployment will not credibly announce a zero inflation rate. The price of non-credibility is for the economy to be stuck in a sub-optimal equilibrium.

Þ the idea is therefore to delegate the monetary power to an organ independent from the government, which will have enough credibility to achieve low both expected ad actual inflation rates.

B - Empirical evidences confirm its conclusion

1 - Empirically, independence goes with lower inflation rates

The idea that political independence will lead to less inflation has been subjected to much empirical analysis, recently. E.g. studies SCHALING (1995). These studies show that central banks that are politically independent tend to produce less inflation than central banks that have to take orders from the government.

Basically, countries where the central banks have a great deal of political independence enjoy a lower rate of inflation.

2 - No empirical relationship between independence and growth or unemployment rates

An important aspect of these studies is that they also reveal that on average and in the long run, political independence does not lead to more unemployment or to a lower growth rate of the economy.

Þ all this seems to suggest that the basic BARRO - GORDON paradigm is correct.

II The ECB's statutes : a strong unbalance between independence and accountability

A - The design of the ECB

1 - The influencing forces

· The economic case for independence, which enables for less inflation without damaging the long run level of output (c.f. BARRO - GORDON model)

· The concern for inflation was particularly important to Germany.

Þ Indeed there is a strong asymmetry in incentive for high and low inflation countries to enter EMU. This fact is very important in the design of the ECB.

The low inflation countries have very little to gain from joining a monetary union constituted by higher inflation countries. Therefore low inflation countries may refuse to make the final step to EMU.

The only way able to improve the incentive for Germany to join EMU is to ensure that the individuals who will rune the ECB are even more 'hard-nosed' about inflation than Germany itself.

This is largely why, strikingly enough, the statutes of the ECB give even more emphasis to price stability than do the Bundesbank statutes. The Treaty implements two important principles.

2 - The first principle : the maintenance of price stability

The first is that the primary objective of the ECB should be the maintenance of price stability (Article 105).

3 - The second principle : political independence

The second principle is political independence. It is in fact seen as the necessary condition to ensure that the national budget deficits will not be financed by printing money. It is also intellectually influenced by the conclusions of the BARRO - GORDON model.

The language used by the drafters of the statutes of the ECB is tougher on inflation and political independence that those of the Bundesbank.

B - The statutes only provide a weak accountability

Whereas the Treaty is quite explicit in the formulating the principle of political independence, it has little to say on the issue of accountability. The Treaty has created an institution with weak accountability. The problem of accountability, as we shall see, arise at two levels.

1 - Article 109b of the Treaty

· The ECB must address an annual report on its activities to the EP, the Council and the Commission.

· The president of the ECB shall present this report to the council and the EP which may hold a general debate.

· The President of the Council and a member of the Commission mar participate without the right to vote, in meetings of the Governing Council of the ECB.

2 - Accountability in defining the objectives of the ECB

The problem of accountability arises at different levels. The first one the society must chose the objectives for the central bank (and not the other way round). The Treaty is so vague that is quite likely that the ECB will in fact fix them.

This may lead to tension when the ECB pursues objectives that are not shared by the rest of society.

Þ in particular, the ECB may choose a target level of inflation that is too low for the welfare of society.

Þ the ECB may accord too little importance to short-term output stabilisation.

This tensions may lead to the feeling that ECB lacks legitimacy, which may endanger the long-term survival of the monetary union.

3 - Accountability in evaluating and possibly sanctioning the ECB

A second problem has to do with accountability in the narrow sense. Once the objectives have been determined, there should be a procedure that allows a society to evaluate how well these objectives have been achieved.

If there is a perception of a systematic failure, a sanctioning mechanism should exist that allows society to redress the situation.

4 - Accountability in reforming the statutes

It is interesting to note that the Bundesbank, which served as the role model for the ECB can be said more accountable than the ECB. The reason is that the law which describes the responsibilities and duties of the Bundesbank can be changed by a simple majority in the German Parliament. This is a very direct way in which society can pressure the Bundesbank to pursue more short terms society's interests.

This is not the case with the ECB. Their statutes are part of the Maastricht Treaty and therefore can only be changed by unanimity. There is very little chance that society can change ECB's behaviour.

5 - The only accountable area of ECB's behaviour : the exchange rate

The Treaty stipulates that, although the ECB will independently decide the conduct of monetary policies, it will not independently decide the Euro exchange rate with non-EU currencies.

In Article 109, the Treaty states that the Council of Ministers (i.e. politicians) shall decide whether the Euro can enter into formal exchange rate arrangements with third currencies. Only the Council can decide to devalue the Euro.

In absence of formal arrangements, the Council may formulate general orientations for exchange rate policy. The Treaty adds however that these general orientations 'shall be without prejudice to the primary objective if the ECB to maintain price stability'.

Some observers have argued that the power of the Council to determine the exchange rate policies in relation to third currencies undermines the political independence of the ECB. It is indeed clear that is the Council decided to conclude a formal exchange rate arrangement with, say, the dollar, this would severally constrain the ECB's independence in pursuing its monetary policy.

However, the chances are good that the exchange rate regime will be flexible (as it has been for most of the last 25 years). In such an exchange rate regime, there is very little the Council can do.

III Moreover, the ECB is likely to run policies which are sub-optimal to society and therefore undermine its legitimacy

The social acceptability of the ECB will be very much influenced by the inflation target. Thus, two issues arise :

- What level of inflation should the ECB aim at ?

- What level of variability should the ECB allow around the target ?

A - The likely inflation rate target of the ECB is sub-optimal to society

1 - The optimal inflation rate

In the 1950s Milton Friedman formulated the view that the optimal inflation rate is zero. The basic reason for this conclusion is that a zero inflation rate maximises the total utility of holding money.

Two factors, which have been recently much researched, casts doubts on this conclusion.

Þ Measurement bias GORDON (1996) First, there is evidence that the conventional measure of inflation (the rate of the consumer price index) tend to overestimate the true inflation rate by 1 or 2 percents. One of the key reason is that the conventional measure of inflation do not take into account quality improvements (example of personal computers). We conclude that if we observe an inflation rate of 1 to 2 percents, the true underlying inflation rate is probably zero.

Þ Necessary real wage effect. There are arguments to be made for a rate of inflation a little higher than 0 percent. The main one is that sectoral or micro-economic shocks require adjustments in real wages. Given the fact that the resistance of nominal wage reduction is high, some inflation is necessary to adjust. AKERLOF (1996) comes to the conclusion that this effect may require the monetary authorities to target an inflation rate of 1 or 2 percents per year.

Conclusion : The previous two analysis lead to the conclusion that the optimal inflation rate may be of the order of 2 to 4 percents per year.

2 - Probable stance of the ECB : lower than 2 percents

There is a great likelihood that the ECB will want to keep the annual inflation rate below 2 percents per year. This is because the central banks that are hailed for their anti-inflationary success have lowered inflation below 2 percents.

Therefore, the ECB will probably run a target inflation rate below the optimal one.

B - The likely ECB target of inflation rate variability is sub-optimal to society

1 - A legal obligation to support the national economic policies

A major interrogation over the ECB has to do with the degree of compromise this institution will accept for short-term fluctuation in the inflation rate around the targeted one. In this connection the Treaty stipulates the following :

Article 105 " Without prejudice to the objective of price stability the ECB shall support the general economic policies in the Community with a view to contributing to the achievement of the Community objectives as laid down in Article 2".

It worth noting that Article 2 includes as an objective 'a high level of unemployment'. Thus the Treaty recognises the need for the ECB to pursue other objectives. Most societies except the central bank not to abandon completely the ambition of stabilising the economy. With each recession, social pressure will push the ECB to relax its monetary policy stance.

2 - Yet the optimal level of short run fluctuation in inflation is to the ECB's gift

The problem here is that the ECB will determine in a sovereign way 1) whether it is willing to allow for short-term deviations of inflation rate from its targeted level 2) and whether it is willing to accommodate its monetary policies to a recession.

IV Can we eliminate the inflation bias while maintaining society's desire for stabilisation (accommodate independence and accountability) ? Yes

A -The need for a balance, and for a sanctioning system

1 - The need for an institutional device ensuring a balance between independence and accountability

It is clear that a balance should exist between political independence and accountability. In this line one could argue that a sanctioning mechanism should be set up. This could take different forms. The important thing is that it should exist.

A control mechanism should be instituted to ensure ECB legitimacy while preserving its independence. What could be its form ?

2 - Two forms of accountability

There are actually two issue at stake :

1) installing a control procedure to ensure that the ECB comply with its Treaty duties

2) finding a way to make sure that the ECB is not totally impervious to the short-term economic needs of society.

B - Controlling the level of the inflation rate

1 - Who should control the inflation rate ?

This raises the question of who should determine the inflation target. The Treaty has not really settled this issue. It only stipulates that the ECB should pursue price stability , without giving any precise definition of this concept.

It is likely that the ECB will be quite autonomous in deciding what this means. Several proposal have been made to introduce some accountability of the ECB on this issue, and to ensure that the ECB fulfils its obligation.

2 - Proposal 1 : Incentive schemes

Incentive schemes have been proposed whereby the salaries of the ECB policy-makers would be significantly reduces if the inflation rate exceed some level.

3 - Proposal 2 : procedure for removal of the board of directors

Others have proposed procedures whereby the board of directors could be removed, should it fail to maintain price stability.

C - Controlling the level of variability of the inflation rate

Several proposals have been made to make the ECB take into account the short term needs of national economies.

1 - The problem (ROGOFF, 1985)

We know from the BARRO -GORDON model that the appointment of an independent and 'conservative' central banker will result in a lower expected inflation rate and therefore in lower inflation rate. By conservative we mean a central banker whose preferences over the inflation rate are lower than those of the society (politicians).

There is a gain from appointing this conservative central banker : a lower long-term inflation rate. However, there is a cost as well. This conservative central, if there is a recession, will use less stimulus so that unemployment increases more than in the case where the central bank pursues the same objectives as society.

This situation may lead to problem of legitimacy of the ECB. How can this problem be solved ?

2 - Proposal 1 : a performance contract (WALSH, 1995)

Recently WALSH has shown that this problem can be solved by designing a 'performance contract' with the central banker in which the latter pays a penalty if inflation is too high.

This penalty scheme as the same effect as reducing the inflation target.

PAUL DE GRAUWE :

'The nice thing about the solution is that the central banker pursues exactly the same stabilisation effort as society desire. The central banker operates along an expansion path which has the same slope as society'.

'Thus, the long term legitimacy of this central banker is likely to be stronger than of the conservative banker'

'This solution combines independence with accountability. The central banker has a contract with society to achieve certain objectives and is evaluated ex post on his performance. Apart from that, he is independent'.

3 - Proposal 2 : inflation targeting (SVENSSON, 1995)

A 'performance contract' proposed by Walsh is not the only way to eliminate the inflation bias while maintaining society's desire for stabilisation. Recently SEVENSSON (1995) has shown that inflation targeting can be made equivalent to WALSH's 'performance contract'.

Society now makes a contract with the central bank in which it sets the inflation target at the level P1. The central bank pursues this target and will on average achieve P2, which is the socially desired rate of inflation.

In this inflation target contract, the central bank attaches the same weight to stabilisation as society. In this contract we do not need to impose a penalty on non-performance.

However, there is a problem : the lack of credibility of this contract. Indeed, the central bank announces a target P1 which it misses systematically.

Conclusion

The major problem in the design of the ECB concerns the balance between independence and accountability which are conflicting and both desirable. While the Maastricht Treaty is very explicit in guaranteeing political independence, it is much less so on the need for an accountable central bank. This is unfortunate. For it is essential that the central bank should be held accountable for failures in its monetary policies.

We argue that a control mechanism should be instituted to avoid the ECB losing its legitimacy. We also argued that this can be done without compromising on the independence of the ECB (c.f. WALSH & SVENSSON).

BACK TO PUBLIC POLICY ANALYSIS