ARENA Working Papers
WP 97/11

EMU and the French Generals
Some Notes on the Swedish EMU Report (SOU 1996:158)

Ton Notermans
Department for Social Science, University of Tromsø and ARENA

Introduction

Even though the European Commission argues that Sweden will have no choice but to join EMU if it satisfies the Maastricht convergence criteria the Swedish government has a decidedly different opinion. As minister of finance Erik Åsbrink again pointed out during the parliamentary debates about EMU on November 6, 1996 it is the Swedish parliament which will take the final decision. Moreover, the government feels that the decision should be preceded by a broad public debate about the pros and cons of EMU membership. In order to provide the basis for such a discussion in October 1995 the government appointed a committee of economists and political scientists with the task of producing a policy recommendation. In its report, which was published on November 5, 1996 the commission recommended that Sweden should not join EMU from the beginning but should do so at an, unspecified, later date. (See SOU 1996:158). The committee takes care to point out that it is not possible to reach a definitive conclusion based on scientific reasoning (p. 401). The recommendation it has produced is based, in part, on subjective judgements and hence has to be seen as a basis for discussion.

According to the report the balance of the political pros and cons is in favour of EMU participation. On the economic side, however, the costs outweigh the benefits in the short run. Moreover, the short run economic costs are seen to outweigh the political benefits. In the somewhat longer run, however, the political costs of not joining can be expected to increase whereas the economic benefits of staying outside are likely to decrease. Accordingly, in the longer run EMU membership is advisable.

Joining EMU will give foreign policy access to greater power resources and hence will allow Sweden to more effectively defend its interests in an ever more internationalised world. A decision not to join, especially if Sweden does satisfy the convergence criteria, may be interpreted by other EMU countries as an attempt to free ride and might hence isolate Sweden politically, just as Britain and Denmark are said to be isolated. Moreover, as more and more countries can be expected to join the initial core group, political isolation is likely to become more pronounced over time. Although the political arguments overall favour EMU membership, an important political argument against joining EMU from the start, is that such a decision is likely to increase the rift between politicians and the population. Given the widespread scepticism in Sweden concerning EMU and the absence of a broad political discussion, a decision in favour of joining from the beginning is likely to make EMU appear as a project for the political and economic elite. In the short run, the report hence deems it imperative that a public debate take place in Sweden (p. 434). Moreover, if the Swedes are given a chance to observe that EMU functions well the risk that it will be considered an elitist project is reduced (p. 435).

On the economic side the report agrees with one of the European Commission's main arguments, namely that elimination of separate currencies will lead to gains in the form of reduced transaction costs. Yet these cost are judged to be small. Moreover, this effect will even be smaller the fewer countries join from the start. The report also agrees that joining the EMU is likely to improve the credibility of a low inflation policy and hence will most likely reduce nominal interest rates in Sweden. Yet, this same effect can also be achieved by giving the Swedish central bank a larger degree of independence without joining EMU. The decisive argument against joining EMU in the short run is that the loss of monetary policy autonomy entailed by EMU membership carries unacceptable risks given the presently high unemployment rate. EMU membership entails that Swedish monetary policy can no longer be employed to counteract so-called asymmetric shocks, i.e. economic shocks which do not affect all member countries equally. If, for example, Sweden, as the only country in EMU were to experience a reduction in domestic demand it is unlikely that the ECB will become more expansionary because its policies probably will be determined by developments in the bigger countries. (p. 52). The result would be higher unemployment in Sweden.

In theory labour market flexibility and fiscal policies could compensate for the loss of monetary policy autonomy. In practice neither of these are deemed feasible alternatives (p. 412-413). Because of language and culture barriers the high degree of labour mobility which characterises the US market cannot be achieved in Europe. Given that labour is relatively immobile, flexibility of nominal wages is an alternative mechanism by which to iron out asymmetric shocks. If, for example, in response to a negative demand shock in Sweden wages would fall, thereby improving relative cost competitiveness with respect to the rest of the EU, no increase in unemployment would have to result. Given the presently very low inflation rate a fall in relative wages would imply an absolute reduction in nominal wages. Historical experience, according to the report, shows that such wage cuts are very difficult to achieve and likely to promote social conflict. A radical reduction of unemployment benefits might increase the pressures for wage reduction in times of unemployment. But the report argues that such a solution is neither likely nor desirable (p. 413) Moreover, it is pointed out that even the US labour market with its low level of unemployment benefits is characterised by a fairly high degree of nominal wage rigidity. Finally, discretionary fiscal management in theory can compensate for the loss of monetary policy autonomy. In case of a negative demand shock Swedish fiscal policy could improve the domestic cost level by reducing tax burdens on employers and increasing other taxes. Yet the political bargaining process concerning changes in taxation is such that discretionary fiscal management is an unwieldy instrument.

In short, the loss of monetary policy autonomy implies a higher risk of increased unemployment, but given the already very high level of unemployment in Sweden the report finds such a risk unacceptable. In the short run therefore Sweden should reform its labour market institutions so as to improve real wage flexibility and bring down unemployment. With a considerably lower unemployment rate and more flexible labour markets, however, the economic costs of participation are small and are outweighed by the political benefits.

In theoretical terms the position of the committee implies a step away from the norm-based / Neo-liberal discourse which has dominated Swedish economic policy debates since the early nineties, towards a more Keynesian interpretation. The concept of norm-based policies was closely affiliated to monetarist concepts as it assumed that monetary stability was a necessary and sufficient condition for stability of the real sector of the economy. In contrast to monetarist views, where stabilisation of the price level was to take place by means of stabilisation of the money supply, the Swedish concepts, influenced by the rather unhappy experiences with monetary targeting in other European countries, formulated stabilisation of the price level, either in the form of a fixed exchange rate or an inflation norm, by means of more or less discretionary policy management as the overriding priority of economic policies. Economic instability in this view was primarily the result of destabilising economic policies. The emphasis the committee places on (asymmetric) economic shocks, however, means a return to a pre-monetarist view where it is the task of economic policies to stabilise a potentially unstable economy. Accordingly the EMU report also displays a greater confidence in the ability of the Swedish political system to pursue policies which benefit the economy. As will be remembered the decision to tie the Krona to the ECU in early 1991 was primarily a result of the desire to create an external constraint in the face of the political system's perceived inability to pursue stabilising economic policies. As Johan A. Lybeck (1991, p. 556) put it: "The link to the ECU shall not, in my view, be interpreted aggressively or optimistically but as the final proof that a 19-year period in which Swedish policies deviated from those in the surrounding world has come to an end. We have given up! We have come to the conclusion that Swedish economic policy is not competent to manage the Swedish economy. We leave the economic policy decisions to the surrounding world.” ^[1]

The argument that Sweden should not join the EMU in the short run in order not to loose it ability to pursue stabilisation policies hence can be seen as a rather fundamental turnabout. Moreover, in traditional Keynesian fashion economic policies are interpreted to have short-run effects only. The underlying model is one of an economy which stabilises itself in the long run by means of price flexibility. Yet due to short-run price rigidities the possibility exists for economic policies to stabilise the economy. From this follows the view that EMU as such, will have no significant long term effects on the real development trends of the Swedish economy. Accordingly no good economic arguments exist against EMU participation in the longer run, although admittedly EMU membership may imply a greater variability around the exogenously given long-term trend.

In addition the argumentation again displays somewhat more of a historical-institutionalist flavour typical of many Keynesian arguments. For example, the monetarist argument that restrictive monetary policies would have negative effects on employment only for at best two years was mainly based on a rationality argument. An unexpected reduction in inflation meant that wage earners were forced off their labour supply curve as they now were supplying less labour at a higher price. The rationality postulate accordingly dictated that wage earners, at the next available occasion, i.e. the next wage bargaining round, would adjust wages so as to return to their labour supply curve. And from this type of reasoning followed the accusation that Keynesianism was logically faulty because its view that macroeconomic policies were required to move the economy out of an unemployment equilibrium was incompatible with rational behaviour by labour market actors. The EMU report, however, is less willing to ignore nominal rigidities because they may not be compatible with a certain view of rationality but rather tends to point to the historical record which indicates that nominal rigidity is a pervasive problem. In political terms this means that decision-makers again are advised to take real-world institutions into account rather than formulating economic policies an the basis of an abstract theoretical model.

To be sure, the position taken in the EMU report does not mean a full-fledged return to Keynesianism. Although it is argued that expansionary macroeconomic policies are needed in order to reduce unemployment, such policies are seen as supporting real wage adjustments. In other words, unemployment is still seen as a problem mainly related to the price of labour rather than to aggregate demand. Moreover, whereas a need for more discretionary policies is recognised, the long-term orientation of monetary policies should remain strictly focused on the price level. Whereas in Keynesian analyses the responsibility for low inflation and high employment is considered to be shared by the central bank and the labour market partners, with the bulk of the responsibility for low inflation falling on wage bargaining and most of the responsibility for employment assigned to demand management, the EMU report maintains the monetarist view of an undivided policy responsibility of the central bank for low inflation. Accordingly the report is not willing to accept that the central bank should be subjected to the fiscal authority but instead advocates further strengthening the political independence of the Riksbank.

In terms of the history of economic policy concept the position advocated in the EMU report hence might perhaps best be characterised as lying in between Neoclassical and Keynesian positions with the arrow of motion pointing to the Keynesian direction. Historically this development is not too surprising. Although one might like to think that changes in economic policy views are inspired by advances in economic science, it has been pointed out repeatedly ^[2] that the discussion between what is presently known as Keynesianism and monetarism goes a long way back and that who seems to be winning the debate is closely related to the reigning economic constellation. Whereas the developments of the 1930s may have led the Keynesians of the 1950s and 1960s (i.e. almost all economists) to argue that the neoclassical quantity theory now was finally dead and buried, such turned out not to be the case as the rise of monetarism in the seventies showed. Yet also the monetarist view that Keynes was finally dead because a fatal logical flaw had been discovered in his theories did not prove correct. Looking at the economic policy discussion in Western Europe at present it might seem that Keynes is not dead in the long run after all.

Historically the monetarist view, or to be more precise, the neutrality of money doctrine, has become the dominant economic policy discourse in response to high inflation. ^[3] In times of high inflation policy makers come to embrace the view that monetary instability is a political phenomenon, i.e. due to misguided monetary policy, and that the restoration of monetary stability at the same time is a necessary and sufficient condition in order to restore stability also to the real economy. The persistent unemployment which generally results from severe monetary disinflations (or deflations) sooner or later, however, tends to give rise to a re-evaluation of policy. The sluggish adjustment of markets now comes to be blamed for stagnation in the real sector and economic policy accordingly again comes to appear as a stabilising rather than a destabilising element.

Indeed this stylised version confirms rather well to the Swedish economic policy history of the last years. The breakthrough for a policy which gave absolute priority to low inflation came when Alan Larsson replaced Kjell-Olof Feldt as minister of finance in 1990 and it had been preceded by rapidly rising inflation and an inability to moderate wage and price escalation by the traditional Swedish way of bipartite centralised wage bargaining. The policies started by the Social Democrat Larsson where pursued with even more vigour by his conservative successor Anne Wibble. And although such policies did rather effectively put an end to inflation they also lead to the most severe recession experienced in Sweden since the early twenties, when a social democratic government, succeeded by a liberal government, embarked on a policy of price deflation in order to prepare the Krona for its return to the Gold Standard at the pre-war parity. Finance Minister Wibble eventually had to abandon the defence of the Krona parity when, in the exchange rate crisis of the second half of 1992 the very high interest rates charged by the Riksbank threatened the complete collapse of the Swedish financial system. ^[4] Yet keeping inflation down remained the overriding aim of monetary policy as the fixed exchange rate target now was replaced with a inflation norm (2%) for the Riksbank. The experience that a policy aimed at defending the exchange rate came close to precipitating a complete collapse of the economy, ^[5] obviously could only serve to undermine the believe that a norm-based policy is a precondition for real sector stability. Moreover, quite in line with the common historical pattern following macroeconomic disinflations, Sweden experienced that the unemployment created in the early nineties tended to stay at a high level. Accordingly, the view that an anti-inflation policy, if only sufficiently credible would lead to rapid adjustments of wages and prices thereby minimising the costs to the real economy became less and less tenable.

From the Swedish policy experience of the last few years it is hence quite understandable that the policy discussion should move back in a Keynesian direction. Yet, if the longer term historic record of economic policy discussion in Europe is any guide, it is quite likely that the position advocated in the EMU report is not the endpoint of the move away from policies which stress macroeconomic disinflation. Although the report recommends maintaining monetary policy autonomy in the short run while joining the EMU in the long run, the long run results which generally have been predicted to follow from macroeconomic disinflation have historically frequently failed to materialise. Indeed the historical record suggests that the promise of long run employment neutrality has mainly served as an argument to facilitate the political acceptability of short term policies aimed at breaking inflation. The long run in which full employment was supposedly restored never materialised but instead history came to be a succession of short runs. In the longer run, the effects of monetary disinflations generally have not been a restoration of full-employment but a policy reinterpretation which again assigned monetary policies the task of promoting growth.

The obvious historical example is the Great Depression. The Great Depression in fact started in the early twenties when Western European countries started to pursue severely restrictive monetary policies in order to break inflation and prepare their currencies for a return to the Gold Standard. Unemployment increased in the short run, as expected, but the long run held no relief. Instead the long runs brought the catastrophe of the Great Depression as virtually all Western countries pursued strongly restrictive macroeconomic policies in order to maintain exchange rate stability. ^[6] The policy outcome of the Great Depression, was not so much a turn to countercyclical fiscal management, as Keynesian historiography would have us believe, but a different assignment in monetary policies which assigned promoting growth and investment by means of cheap money policy to the central bank whereas the concern for price stability became, in part, delegated to a host of microeconomic arrangement of which incomes policies (with or without state involvement) was probably the most important. ^[7]

From the historical experiences with macroeconomic disinflations it would seem not unlikely that EMU will give rise to similar effects as the interwar gold standard. With a completely independent ECB which is institutionally required to give absolute priority to low inflation and which politically probably will feel the need to demonstrate that it is at least as good, if not better, as the Bundesbank, a restrictive monetary regime will most likely prevent any significant upturn in private sector investment and growth. ^[8] Economic stagnation in turn is likely to intensify efforts of individual EMU participants to cut domestic costs, which leads to the worst of imaginable worlds namely restrictive macroeconomic policies combined with competitive internal devaluations. Politically and socially the result can only be more conflicts.

In short, the EMU project seems to be an outstanding example of the type of economic policy which Hamilton and Rolander (1993) have described with the metaphor of the French generals. As is well known, at the eve of World War II the French generals, with their Maginot line, were preparing to fight World War I all over again. The results of this policy inertia were disastrous for France and Europe. Likewise the EMU is basically a device designed to meet a constellation, which most West-European governments had to confront at a point, in which containing inflation by the traditional means of negotiated price and wage moderation seemed impossible leaving restrictive macroeconomic policies which created unemployment as the only alternative. And although it would obviously be quite ridiculous to want to compare EMU to the Nazi dictatorship, the effects of institutionalising a policy regime which was designed for a constellation which lays in the past can only perpetuate Europe's economic and social crisis.

Rather than preparing for EMU participation in the longer run, Swedish economic policies hence might best serve their country be embarking on two different priorities. (1) Trying to prevent monetary integration from taking place in the form as planned now. Since a monetary policy regime aimed at promoting growth is incompatible with the set-up chosen in the Maastricht treaty, the EMU plans are best abandoned. ^[9] Even if Sweden is not to join EMU in the longer run, it can not benefit from an arrangement which perpetuates restrictive policies in the EU. (2) Concentrating on a durable reform of domestic labour relations which will make it possible to durably combine low nominal wage increases with low unemployment. (3) Abstaining from any plans to make the Riksbank more independent, but instead making it clear to the bank that it has a shared responsibility for inflation and employment in the short as well as long run. ^[10]

The structure of the paper is as follows: Section two argues that the report suffers from an unresolved tension between a theoretical argument which predicts that restrictive monetary regimes have no real long-term effects and the empirical recognition that the mass-unemployment which has resulted from macroeconomic disinflation has proven to be quite permanent indeed. As a result of this tensions the report advocates, on the one hand, institutionalising a policy regime which gives priority to macroeconomic disinflation by giving the Riksbank more independence and by joining the EMU in the longer run, although it lacks a theoretical argument why inflation is bad at all. On the other hand it advocates that the concern for employment should again be given more weight in short run macroeconomic policies, thereby in fact pointing to a need to actually reduce the political independence of the central bank. Section three argues that the institutionalisation of a disinflationary regime has long term negative effects on unemployment, not because of wage rigidities, but because it discourages growth. Section four concludes that a the EMU in its present form is the main obstacle to a recovery of the European economy. Instead of spending so much political energy on pushing through EMU, European countries should concentrate on finding an alternative form of monetary co-operation which allows for growth oriented policies, and on labour market reforms which will prevent inflationary pressures once unemployment has been brought down again.

Inflation and the Fight Against Unemployment

As mentioned, the weight the report attaches to reducing unemployment means a change in discourse compared to the previous years when the goal of low inflation was allowed to prevail over all other objectives. A change in professed policy priorities, however, does not necessarily have to result in a change in policies simply because the relationship between means and ends is a matter of economic theory and hence a great many different policies can be interpreted as being informed by the priority of reducing unemployment. The conflict between giving overriding priority to inflation or employment, e.g. could simply be solved by postulating low inflation as a necessary and sufficient condition for employment growth. On that view one might even interpret the policies of the Bundesbank as really being driven by a desire to optimise employment.

Yet the report does advocate a change in policy making, as it recognises a need for discretionary monetary policies to assist in the fight against unemployment. As will be remembered, the norm-based approach implied that macroeconomic policies had no role to play with respect to unemployment. Employment in this view was determined by the real wage which in turn resulted from the wage earners' trade-off between work and leisure. High inflation resulted because wage earners expected the government to pursue inflationary policies thereby prompting them to increase nominal wage demands in order to arrive at their preferred real wage. A credible anti-inflation policy would simply eliminate this effect. The unemployment that maintained under a credible anti-inflation regime was seen not to be the responsibility of the government because it had no business trying to force wage earners to accept a different work-leisure trade-off than the one they preferred.

This view, however, has become increasingly untenable as every single European country which embarked on macroeconomic disinflation came to experience radically higher and durable unemployment in the aftermath (p. 197). To be sure, the EMU report does not make a full U-turn back to Keynesianism in the sense that it analyses unemployment again as purely a problem of aggregate demand. It does strongly emphasise the need to reduce real wages and proposes several structural changes to that extent (p. 220-223, 369-373). Amongst those are a reduction in unemployment benefits, more restricted access to labour market programmes and regulations which will place more of the costs for unemployment on the employed. Yet, it is argued, these supply side measures will not suffice. They will have to be complemented by macroeconomic policies which increase demand and since fiscal expansion (at present) is not a feasible alternative this means more expansionary monetary policies (p. 374).

The recognition of the need for short term discretionary policies in order to combat unemployment, however, sits uneasy with the acceptance of low inflation as the sole policy goal of the Riksbank, and the demand for increased political independence of the bank. Like all views which are based on the assumption of long run neutrality of money, the report has no good answer why inflation simply should not be ignored. Long term money neutrality does mean that it is not possible to "buy permanently lower unemployment at the price of higher inflation" (p. 121) But this of course is no argument against inflation at all. If inflation is stimulative in the short run but neutral in the long run, this should imply that low inflation is a rather irrelevant policy goal.

At the same time disinflation has high costs, which not only occur in the short run. Most important, the unemployment which the report now wants to reduce seems to be the cause first and foremost of disinflation. As is argued on p. 366 e.g.: "It is clear, however, that the triggering factor behind the dramatic increase in unemployment during 1991-93 was the strong reduction in demand.” ^[11] Following the report's labour market analysis the main mechanism was that the severe disinflation which was started in 1991 had the effect of increasing real wages due to nominal wage rigidity. ^[12] Since, as the report points out, the increase in unemployment which resulted from a reduction in demand has tended to become permanent, one must come to the conclusion that what has happened here is simply that macroeconomic policies have created durable mass-unemployment in an attempt to reduce inflation. ^[13]

It should also follow from the report's analysis that the most effective way to get rid of the present unemployment is to reduce real wages by means of inflation. Since the report argues that it is hard to bring about a reduction in real wages if inflation is very low or even negative, then a reasonable conclusion would be that Sweden needs more inflation. It may be true, that a substantial part of the Swedish unemployment despite its relatively short duration has become 'structural' e.g. because the unemployed no longer have the required skills and therefore in practice are unemployable. (p. 366-7) But inflation should at least be able to reduce that part of unemployment which is due to high real wages. Moreover, the distinction between "structural' and cyclical factors is hard to draw. Many long-term unemployed at present may be almost unemployable. Yet as the overall demand for labour increases firms become more willing to invest in retraining and what may seem to be structurally unemployable people in a lax labour market will turn out not to be so during an upswing.

In addition it would seem that the discretionary policies the report advocates are in fact made more difficult by the target of 2% inflation currently in force. Unless the Riksbank's commitment to the inflation target is beyond any reasonable doubt, it is argued, a reduction in interest rates with the aim of stimulating the domestic economy may easily lead to a depreciation of the Krona and higher inflation which will require a turnaround in monetary policies (p. 123-4). It is concluded from this that the political independence of the Riksbank should be increased so as to strengthen the credibility of the commitment to low inflation. A more obvious conclusion would be to abandon the low inflation target thereby abandoning the need for a policy turnaround, and instead allowing exchange rate movements to reinforce the effects of expansionary monetary policies. Given that nominal wages are said to be sticky the depreciation of the Krona which may follow lower interest rates would provide an additional stimulus to the economy as it lowers real wages and increases competitiveness.

Some cost of inflation are recognised (p. 46-49), yet those costs cannot outweigh what are analysed as the costs of disinflation. First the famous "shoe leather costs" which simply means that because inflation undermines the value of money the public will at any time hold less cash and hence will have to make more frequent trips to the banks. Secondly inflation means that producers have to change their listed prices more frequently. Thirdly inflation may lead to lower growth. Fourth, inflation leads to more uncertainty. Fifth, inflation may mean that it becomes more difficult for market participants to discern the correct relative prices thereby leading to misallocation. Finally inflation changes the income distribution because it implies a tax increase in systems based on nominal values and because it benefits debtors at the costs of creditors.

As the report notes the first two effects cannot be considered significant (p. 47, 85) and both the theoretical and empirical basis for the third argument is rather weak. But, in my view, the remaining three effects can hardly be decisive either. Why inflation should lead to more uncertainty is not clear at all. If e.g. the government can convince the public that it will pursue a durable high inflation policy, uncertainty would not have increased. Moreover, it is not clear why uncertainty in a market economy is a bad thing. After all it is one of the cornerstone of competition that one does not know exactly what one's competitors will do next. The fifth argument, though widespread is theoretically inadmissible. One might think it a bit inelegant to first arrive at the conclusion that money is neutral in the long run by introducing the absence of money illusion, i.e. the notion that wage earners can very well distinguish absolute from relative prices, only to reintroduce some form of money illusion in the next round in order to arrive at costs of inflation. More important, the claim that allocation based on correct perceptions of relative prices is optimal is in fact an unsolved issue in general equilibrium theory. It is a rather weak, ad hoc and inconsistent argument to assume that in a real-world economy characterised by the absence of future markets, and hence by the pervasiveness of imperfect information, the additional distortion in information brought about by inflation should affect allocation only in negative ways except as far as real wages and unemployment is concerned where no distortion in the perception of relative prices occurs at all. Finally, although it is true that inflation changes the income distribution, this is at heart a political question which cannot be decided by economic theory. That inflation benefits the productive sector, which in the aggregate has a net debtor position, and the common tax payer with little financial assets at the costs of those who hold considerable wealth in financial assets might just as well be considered an argument in favour of inflation. Moreover, if considered undesirable, the tax system could in principle be adjusted (p. 86) thereby eliminating another argument against inflation.

The peculiar tension between, on the one hand accepting the goal of low inflation, but on the other hand employing an analysis which is unable to justify this goal, is also reflected in the recommendation the report makes for monetary policies. It is not likely that the recommended route of increasing the Riksbank's political independence will increase the credibility of low inflation policy and allow for a more discretionary type of monetary policy which can pay greater attention to the unemployment problem.

The report argues that the credibility of low inflation policies in Sweden needs to be strengthened. This could be done be means of EMU participation, but the same effect can achieved by implementing the proposals made by another public commission in 1993 (SOU 1993:20). In that report it was proposed, amongst other things, that the Bank should be given the explicit goal of maintaining price stability, that the mandate period of the banks board members should be extended so as to no longer coincide with parliamentary elections, and that members of parliament, members of government and persons connected to the central organisations of the political parties should be non-eligible for membership.

The increased credibility for low inflation policies should in turn make it possible for the Riksbank to pursue a more flexible policy aimed at helping the fight against unemployment. In principle the report's argument that it should be possible to give employment considerations more weight in monetary policy by committing the central bank stronger to the goal of low inflation, is not necessarily contradictory. It is possible that a greater credibility of the policy regime, i.e. the long term orientation towards low inflation, will allow for more flexibility in short term policies without casting doubt about the main priorities of the bank. Yet, the question is whether the bank will have an incentive to do so. Already at present the Riksbank manages its low inflation goal very rigidly. Although the official target is 2% inflation (plus or minus one percent) the experience of the last two years rather would suggest that it is aiming for an inflation rate of close to zero. Giving the bank more independence, and an explicit mandate to keep inflation low carries the risk of making monetary policies even more inflexible. As Stanley Fischer (1994, p. 293) argued: "Shielded as they are from public opinion, cocooned within an anti-inflationary temple, central bankers can all too easily deny - and perhaps even convince themselves - that there is a short-run trade -off between inflation and unemployment, and that cyclical unemployment can be reduced by easing monetary policy." An independent Riksbank which has been given a clear single policy goal namely two percent inflation has no incentive whatsoever to risk not achieving the only indicator by which it is judged in order to stimulate employment. Strict monetary policies which perpetuate stagnation and mass unemployment are the safest way to keep inflation down. Expansionary monetary policies always carry the risk, however slight, that the additional monetary space will be used by trade unions and producers for wage and price increases. Assuming a rational Riksbank which aims to fulfil its task such an expansionary policy hence must seem as taking unnecessary risks.

Moreover, the recommendation to join the EMU at a later date is likely to cement that type of monetary policies which the report holds to be too rigid. In a national context a government which has a majority in parliament can always override a legally independent central bank by changing the law. This is not the case for the European Central Bank (ECB) because the Treaty on European Union (TEU) accords the ECB a degree of political independence hitherto unparalleled in history. For the ECB it will not be the case that a simple parliamentary majority can change its statutes. Instead a change in the statutes of the ECB will require a renegotiation of the treaty and hence gives each single country a veto right. In principle this means that, let's say, the Luxembourg government, representing roughly 300,000 people could block any change. In addition the lack of a fiscal policy authority and a proper parliament at EU level will contribute to the Bank's greater independence. Even independent Banks like the Bundesbank have historically been quite sensitive to the general political climate, if only because of the need to maintain their legitimacy as a policy institution. The ECB in this sense will not only be politically independent but also politically isolated. ^[14] Finally, the ECB board will most likely feel an urgent need to demonstrate that it is at least as inflation averse as the Bundesbank. During its first years the Bank may hence be expected to display a pronounced tendency to try and be more catholic than the pope. And because a reputation is only built up over quite a number of years, this first period may last a considerable time. ^[15]

Put differently, since monetary policies apparently do have a significant effect on unemployment in what reasonably must be called the long run, it would seem to follow that the policy priorities set for the bank by the political authorities must also include employment alongside inflation. The decisive argument in favour of central bank independence during the seventies and eighties was that no trade-off existed between inflation and unemployment. In that case the only thing the bank can do is to lower inflation, and that is a politically legitimate task. If, however, it turns out that a serious trade-off does exist then the bank is in fact making a political choice which only a democratically elected government should have the right to make. To quote Stanley Fischer (1994, p. 294) again: "...central banks cannot merely be given the task of keeping inflation low: they have also to be made accountable for their performance, especially their counter-cyclical performance, to be asked whether they are making the right judgement about the speed at which to reduce inflation, or to return to full employment. They cannot take refuge in the claim that there is no long-run trade-off."

The report, to some extent recognises the danger of an independent Riksbank becoming obsessed with low inflation, as it argues that there might be reasons to include a so-called emergency clause in the central bank statutes (as well as in the TEU) (p. 137, 145, 288). An emergency clause is a clause which gives the government the right to override the central bank in exceptional circumstances. Or, in other words, the government should have the right to issue directives to the Bank. At this point it becomes somewhat difficult to follow the argument of the report, at least for the author of the present notes. The government's right to issue directives to the bank used to be the central characteristic of a dependent central bank. As the authors of the report note themselves (p. 132): "A problem, however, is that emergency clauses can be abused to generally reduced the independence of the central bank. In that case the delegation of monetary policy to the central bank in normal times becomes meaningless and any gains in the form of lower inflation never occur." ^[16] To argue that such a right should only exist in exceptional circumstances is of no help because it must obviously be the government which will decide what is an exceptional circumstance. Accordingly such an emergency clause simply means that the central bank can be forced to adhere to the governments views on how the trade off between inflation and unemployment should be handled. In sum, the proposals of the report seem to come down to trying to increase flexibility in short-term monetary policy by reducing the influence the government can exert over the Riksbank, just to return this influence to the government in the second round.

Finally, one should consider the possibility that convincing the markets of the sincerity of the commitment to low inflation will actually not strengthen the confidence in economic policies. It is assumed in the report that the positive interest rate differential which presently exists between Swedish and international interest rates is a result of lingering doubts about the commitment to low inflation. It is just as well, possible however, that the interest rate differential reflects doubts about the political viability of the current strategy. It may be hard for many to believe that in a country with a long and entrenched tradition of full-employment the government will not sooner or later try expansionary macroeconomic policies to improve the employment situation. To the extent that a more independent Riksbank is seen to make a solution to the unemployment problem less likely and therefore domestic political conflicts more likely, it might actually increase interest rate differentials. It is hard to say to what extent these expectations really do exist. But at least the experience of the devaluation of 1982 which actually restored the confidence in the currency and reduced international interest rate differentials is an example that such mechanisms do occur. ^[17]

Policy Regimes and the Long Term Effects of Economic Policies

In fact the whole discussion about central bank independence would seem somewhat of a confusion. What has happened during the last two decades in Europe is not that governments have given their central banks the freedom to pursue a policy which is more inflation averse than governments themselves would have desired. Rather governments themselves have decided to embark upon a policy regime which gave priority to breaking inflation at the cost of increased unemployment. Obviously there have been many changes in West European central bank laws during the last two decades which have served to reduce the direct influence of governments over monetary policy. And these changes certainly have not been meaningless. Granting more legal independence does grant the central bank more practical independence because it often leaves a change in central bank statues as the only way for governments to change central bank policies. Since this is an instrument which carries more political costs than direct government influence over specific central bank measure, the threshold for government intervention has increased.

Yet the point remains that the statutes of the central banks have been proposed by governments and passed by parliaments and could have been changed at any time if governments had disagreed with the fundamental policy orientation of their central banks. In several countries, like the Netherlands, Britain and Sweden central banks received no legal independence in the sense that the government maintained the right to issue directives to the bank. Yet all of these countries at some point during the last two decades have embarked on a similar monetary policy regime as those with more 'independent' central banks. Moreover, in all countries, with the exception of Sweden, the responsibility for exchange rate policy remained with the government implying an ability to determine a substantial part of monetary policy. Governments may at times put the blame for impopular policies on the central bank, ^[18] but they remain responsible for the overall policies pursued.

That, what was in fact a government decision to install an anti-inflation regime would come to appear analytically as a shift in monetary policy authority from the government to the central bank is due to the peculiarities of the neutrality of money doctrine which is unable to derive the possibility of market driven inflation and hence must treat inflation as a purely political phenomenon. The neutrality of money doctrine arrives at this conclusion by the peculiar method of first assuming an economy in which no money exists and then assuming that such a pure market economy would produce stable and pareto optimal outcomes. In a next step money is introduced as a pure government creation. Since it has been assumed already that real outcomes are determined money can obviously only causes nominal changes, and since money is a government creation the government must be responsible for any instabilities of the price level. In other words, inflation is a political and not an economic phenomenon.

Postulating inflation to be a political phenomenon, however, gives rise to several peculiar problems:

(1) If the government's inflationary bias, which is seen to be endemic, is not a policy which benefits the majority of the electorate but instead creates unnecessary economic instability, then it is not clear why in a democratic polity such policies could be electorally successful. Accordingly the normal democratic process should remove governments from office which pursue such policies and independent central banks would not be needed. That an inflationary bias can benefit the majority of the electorate, however, must be excluded by the neutrality of money assumption. The so called problem of dynamic inconsistency, which is the most common explanation for the inflationary bias, ^[19] is not a convincing argument because it is incompatible with a vote maximising strategy on the part of the government. Dynamic inconsistency basically means that governments are trying to fool wage earners (i.e. the majority of the electorate) by engineering surprise inflation. If market participants are led to conclude contracts on the basis of low inflation expectations then, it is argued, the government has an incentive to pursue a more expansionary monetary policy which will lead to more inflation and hence lower real wages and higher output. That governments can electorally benefit from creating higher output would seem a rather reasonable assumption. Yet, the assumption of the absence of money illusion, which is necessary in order for the neutrality of money to hold, contradicts this. If wage earners do not suffer from money illusion, then they realise immediately that they are forced off their preferred real wage and work/leisure trade-off by the state. In other words they are tricked by the state into accepting economic outcomes they do not prefer. It is hard to understand how an economic policy which runs counter to the interests of the electorate could be electorally beneficent for the incumbent government. Accordingly, it must remain unclear why there should be an inflationary bias at all, or why one cannot rely on the normal electoral process to remove governments from office which do exhibit such a bias. Hence there would seem to be no need for an independent central bank. ^[20]

(2) If, for some reason or other, the government does have an incentive to pursue inflationary policies and the democratic process cannot correct this bias, then governments should have no incentive to grant independence to their central banks. In sum, either independent central banks are not needed or they will not be legislated into existence.

(3) The changes in monetary policies which have taken place in Western Europe during the last two decades did not involve stabilisation of escalating inflation rates, but disinflation. Since disinflation, according to the neutrality of money view, means depressing the economy temporarily just to return to the original position in real terms, there simply should be no good political motive for a government to do so.

The neutrality of money doctrine hence leads to a fundamental ambiguity concerning the relations between government and central banks, because it seems one needs to rely on the government to institute a policy regime which the government in fact does not want. More important, it also leads to fundamental difficulties in understanding the history of economic policy-making. Why have governments repeatedly been willing to incur great economic costs, as e.g. in the interwar period and the last two decades, to break inflation? In addition, this type of "money illusion" is not confined to governments of a certain political hue. At present as well as during the interwar period disinflationary regimes have been installed by left, centre as well as rightist governments.

What is needed to solve the puzzle is an argument which can explain why inflation will eventually pose so severe problems to governments of any political hue that they are willing to get rid of it even at the costs of a durable recession. In order to do, so, however it is necessary to abandon the framework of long term neutrality of money.

Inflation, instead of being a political phenomenon, is better understood as a market driven process which results from the combination of price and wage inflation. Escalating inflation primarily is a signal that at the present level of resource utilisation employers and trade unions are unable to contain nominal prices and wages. Under very tight labour markets not even centralised unions will be able to contain nominal wages, and very high levels of demand allow firms a considerable freedom in price setting. But inflation is obviously not confined to situations of (over) full employment but can occur at substantial levels of unemployment. Wage setting is to a large extent determined by institutional and political dynamics within and between the trade unions and employers' associations, and only at extreme levels of employment and unemployment is mainly market determined. A reduced ability of, e.g., trade unions to co-ordinate wage bargaining hence may lead to higher wage inflation even without a tightening of the labour market.

The increasing rates of inflation in Sweden in the eighties have been a result of both mechanisms operating simultaneously. During the second part of the eighties labour markets were so tight that probably any trade union would have lost control. Moreover the centralised model of national level LO-SAF negotiations, which worked so well in the fifties and sixties, was increasingly confronted with problems of internal co-ordination independent of the labour market situation. As a result the traditional means of containing nominal wage rises broke down in the eighties.

Although some advocates of the neutrality of money doctrine recognise that inflationary pressures can build up in labour and product markets independent of monetary expansion, they maintain that the government has to carry the sole blame because inflation cannot materialise if inflationary pressures are not accommodated monetarily. This argument is not necessarily correct as it assumes a fixed velocity. ^[21] Yet inflation itself increases monetary velocity simply because it punishes money-holding. Recent Swedish history has provided an impressive example of that as the speculative inflation economy which developed during the late eighties was fuelled to a not insignificant extent by a substantial increase in velocity.

Nevertheless it remains true that inflation cannot go on indefinitely without being accommodated monetarily. Yet since inflation is a signal that resources utilisation is too high given the present state of labour market institutions, a reduction in inflation cannot be had without an increase in unemployment, as the last two decades have shown impressively. It is indeed true that inflation expectations come to be included in wage bargaining thereby giving inflation a cumulative character. Yet instituting a credible anti-inflation regime cannot allow for a costless reduction of inflation rates because even if it eliminates expectations it does not remove the root cause. Notwithstanding the large theoretical literature on credibility effects on sacrifice ratios, the decision to give priority to low disinflation have necessarily also been decision to increase unemployment. ^[22] Accordingly it is simply not within the government's powers to guarantee nominal and real stability by pursuing a stable monetary policy. It also follows that the hope that in the longer run disinflationary policies will not lead to durable unemployment is illusory as it is the essence of a macroeconomic disinflation regime to keep unemployment up.

Yet governments do have a good reason to combat inflation which is that inflation, when tolerated too long, tends to become cumulative and disrupt the financial system and with it the financing of productive investment. It is a fundamental trait in much of the economic policy debate that market economies in principle are assumed to be self equilibrating. There may be some need for short term political intervention given the presence of nominal rigidities, yet in the long run it would seem difficult to discern another task for economic policies than preventing that the free operation of market forces is obstructed. Surprisingly such an assumption has no firm basis in economic theory.

The question of how a pure market economy would behave is the topic of general equilibrium theory. General equilibrium theory, however, has never been able to document that market adjustment in an idealtypical pure market economy will result in stability and pareto optimality. ^[23] The basic reason for this is the reflexive nature of social (including economic) interaction. What most economists seem to have in the back of their mind when discussing a market economy is a model derived from Newtonian physics, like for example the model of gravity. In a world with gravity its inhabitants will over the longer run discover the optimal strategy to cope with it. Those that do not will most likely be eliminated. Social phenomenon however have a quite different nature because, to stay with the metaphor, in a social world the gravity which each individual faces is the result of the sum of the actions of other individuals. Accordingly individual, uncoordinated, adjustment processes pose no guarantee to arrive at stability because the attempts of all individuals to adjust to the constraints they face changes the constraints. This problem of reflexivity has lead general equilibrium theory to the conclusion that a stable pareto optimal equilibrium solution in a pure market economy can only be arrived at via non-market processes. This mechanisms is known in the literature under the name of the auctioneer, or alternatively, as the assumption of the absence of time (i.e. the presence of perfect future markets). The auctioneer collects all relevant information from market actors and then communicates to each market participant the price vector which is pareto optimal. Similarly the assumption of the absence of time, and hence money, assures that market adjustment process do not change the constraints other market participants face.

In the real world there is no auctioneer plus there is time, and the reflexive nature of social interaction accordingly implies that a basic co-ordination problem exists which economic policies will necessarily have to address. In an economy where a neo-classical auctioneer is not present co-ordination of economic activity de facto takes place through the medium of money. Or, As Martin Shubik (1975, p. 563) put it: "In the running of the economic system in disequilibrium the guiding mechanism is provided primarily by the monetary mechanisms and trade in ownership and claims. Money is not a veil; the monetary and financial system is the neural network guiding the real system." Money, however can only perform this role if prices can be maintained relatively stable. Rapid de-or inflation simply destroys the conditions under which a market economy can operate. ^[24]

Continuous deflation reduces the incentive to lend and at the same time increases the attractiveness of holding money. Deflation reduces the credit-worthiness of borrowers as it increases their real indebtedness. At the same time deflation means that money holdings yield a return. The resulting contraction in lending to the economy provokes a (further) downturn which increases the risk of default and hence further curtails lending. The value of the collateral of debtors declines with respect to their money obligations. Firms and real estate owners may be driven into bankruptcy, taking the financial institutions which hold their debts with them, further propagating deflationary impulses through the economy. Declining demand and investment produce unemployment. If, however, the labour market behaves the way most neo-classical economists think it should, namely by reducing wages in response to unemployment the fall in prices becomes cumulative and the financial and industrial system will eventually completely collapse.

Wage flexibility in the common understanding denotes real wage flexibility. But since wage earners do not control the price level they cannot set real wages. Rather the only available way to affect real wages is through changes in money wages. Money wages are variable costs and an overall cut in nominal wages is rather likely to provoke an overall fall in the price level thereby exacerbating the deflationary crisis.

Inflation reduces real indebtedness and penalises money holdings thereby stimulating lending. The resulting upswing puts additional pressures on the price level and further stimulates lending and investment. Again the process becomes cumulative if tight labour markets lead to an escalation of nominal wages. Although inflation initially has a stimulating effect, this is not of duration. Escalating inflation implies a continuous erosion of the value of debt contracts and money which eventually will provoke a flight out of money and into real assets. The result is a cessation of lending and the proliferation of speculation.

That inflation undermines the financial and productive system becomes most obvious during hyperinflations. The main problem during e.g. the last phases of the German hyperinflation of the early twenties was not that people had to make very frequent trips to the bank. Instead during the last phases of the German hyperinflation the Reichsmark was only used for tax payments and nothing else. Inflation led not only to a pronounced unwillingness to hold cash but also to a refusal to lend. Without lending, however, the financial system and industry must break down. Accordingly during the last months of hyperinflation the German real economy collapsed. Under hyperinflation, unemployment rose from 3.5% in July 1923 to 19.1% in October of the same year. ^[25]

In sum, even though disinflation by means of monetary restriction inevitably involves recession, this price must be paid if wage and incomes policies are no longer available because the alternative is an eventual hyperinflation and a complete collapse of the economy. At times the false promise of the neutrality of money doctrine of a costless elimination of inflation can even serve a useful purpose to the extent that it allows governments to overcome political resistance to disinflation. Yet at the same time the acceptance of this view guarantees the perpetuation of unemployment as it denies that restrictive money is the main cause of recession.

Now obviously most West European nations were not close to hyperinflation during the last decades. And in some countries like Germany and the Netherlands, which initiated restrictive policies during the early seventies, the behaviour of their trade unions would suggest that there was an alternative to the ferocious monetary disinflation actually pursued. The restrictive monetary regimes of those two countries might have stayed a peculiar exception in Western Europe. Yet as the seventies an eighties progressed those countries in which the goal of full-employment had a stronger foothold in the political system (See Therborn 1985) had to realise that at the point where negotiated wage and incomes policies can no longer be relied on to contain inflation, creating unemployment became unavoidable because if inflation is tolerated for so long that it starts to enter all economic calculations the economy will spin out of control.

In Sweden, were full-employment traditionally had priority and the government accordingly was willing to accommodate inflation longer than most other west European countries, this stage was being reached by the late 1980s. It is a bit peculiar that a report which purports to analyse policy options in a country which has recently produced what was probably the most spectacular financial crisis in the Europe since the collapse of the Austrian Creditanstalt in 1931 would pay so little attention to financial sector dynamics. The speculative finance economy which developed in Sweden during the second part of the eighties is best interpreted as the long run outcome of inflation accommodation. Although much of the blame for the explosion of speculative finance is placed on credit market deregulation, much of the increase in velocity prompted by expectations of inflation was well underway before as witnessed by ever extensive circumvention of the regulatory system'. Inflationary expectations hence forced deregulation but without a sharply restrictive monetary policy this could only further fuel the development of a pure speculation economy. ^[26] It is hard to see what alternative the Swedish government had, especially after Kjell-Olof Feldt's attempt to introduce a wage and price stop failed, other than to create a recession to prevent the economy from getting completely out of hand.

Yet that does not mean that it is wise to stay with a macroeconomic disinflation strategy after the circumstances which made such a policy inevitable have disappeared. Since unemployment is primarily the result of macroeconomic disinflation it will require a macroeconomic solution. No doubt the employment crisis in Sweden offers a historical opportunity which should not be missed to dispose of some of the perverse incentives which have arisen in the social security system over the last decades. Yet to expect a significant reduction in unemployment from reducing real wage flexibility is incorrect. No doubt unemployment can be mitigated by improving export competitiveness. Yet such a solution cannot work in the aggregate. Since EU countries carry out most of their trade amongst themselves, no substantial solution to the unemployment problem can be expected from increasing real wage flexibility. Indeed as many countries of the EU who have had time to introduce a lot of the labour market reforms recommended by the report have experienced, the long run in which unemployment returns to its pre-crisis level fails to materialise. ^[27]

What is needed in Europe to reduce unemployment is first and foremost private sector growth and not beggar-your-neighbour real wage, tax and social security cuts. Yet also the short term countercyclical policies of a basically Keynesian nature which the report recommends are not likely to contribute significantly to growth. The main reason for this is that as long as the macroeconomic disinflation regime remains in place, no decisive upturn in private sector investment can be expected. To pursue sharply restrictive macroeconomic policies certainly is a very effective way to eliminate inflation expectations as the Swedish experience shows. Unfortunately it is also a very effective way of eliminating expectations of growth. Hence, the fundamental long term effects of the policy regime instituted since the early nineties do not derive from real wage rigidity but from the fact that such a regime discourages investment. Investment decision also display a reflexive, or self-fulfilling nature in the sense that if a large number of investors expect growth and decide to invest, growth will indeed be forthcoming. The macroeconomic disinflation regime however is founded on the logic of killing growth as soon as it tends to lead, or is feared to lead, to a substantial reduction in unemployment which might again contribute to wage inflation. Accordingly there is no good reason for market actors in the aggregate to expect growth in the present regime. ^[28] As Tobin (1980, p. 19) argued: "As Keynes also knew, protracted underproduction and under-utilization severely damages the marginal efficiency of capital. In mild and short-lived recessions investment is buoyed by the belief that high employment and prosperity are the long term norm. Once this confidence is destroyed, as contemporary events again demonstrate, it is terribly difficult to revive it." Instead a revitalisation of private sector activity will require that a mechanism be found which can contain inflationary pressures without relying on recession and unemployment, thereby making a credible turn to a growth regime possible.

EMU: Towards the Second Great Depression of the 20th Century?

Applied to monetary integration, the inability to realise that monetary policy regimes can very well have serious long-term effects leads to the misunderstanding that EMU basically is a political project. After all, if monetary policies have no long term effects whereas the abdication from monetary policy autonomy is apt to lead to larger swings in economic activity in the short run, there would seem to be no good economic argument for EMU.

EMU could hence be seen as a political project in a double sense. First as a project which originated as a economic means to the political end of integration (p. 25). The unification of Germany is often seen as the catalyst for EMU. Given that Germany has historically had the best inflation performance in Europe since the seventies there would seem to be no good economic reason why it would be one of the driving forces behind a development which may risk to undermine its performance. Unification, however, made stronger political union a pressing issue if only because Germany can only gain legitimacy for its foreign policy as part of Europe (p. 238). Unification in turn confronted France, and other countries with the danger of a politically dominant Germany making closer integration seem advisable in order to tie Germany. Certainly France may also have had a economic interest in EMU, namely to gain a say over European monetary policy which it considered to be unilaterally determined in Frankfurt am Main. Yet since Germany did not seem to have an economic interest in EMU ^[29] and since its participation was essential it could turn EMU into a quid pro quo for closer political integration. Secondly, as the report emphasizes strongly, the main benefit Sweden can hope to gain from joining EMU is political, namely an increased ability to promote its interests (p. 289).

Yet, EMU is not primarily a economic means to a political end. If it were, then those who call for a revisions or replacement of the treaty because the political and economic tensions it creates has come to pose a threat for political integration might find more open ears. Rather EMU is a device which is primarily created with the aim of institutionalising an economic policy which originally was designed for the inflationary problems Europe confronted in the seventies and eighties. And because of that the increase of power resources which Sweden may receive from EMU membership must be balanced against the long-term economic costs. ^[30]

The basis for the Maastricht treaty was the switch to macroeconomic disinflation which all West-European countries - from Germany and the Netherlands in 1973 to Sweden in 1991 - have individually undertaken during the last two decades. Given the pivotal role of France in EU negotiations the failure of Mitterand's reflation was of crucial importance in this respect. As long as France held on to some form of Keynesianism, agreement on monetary union with Germany was impossible.

The emergence of macroeconomic disinflation regimes is a necessary but not a sufficient condition for EMU. After all Britain under Thatcher was no great fan of monetary co-operation. What put the EMS and later EMU plans on the agenda was that for the traditional weak currency countries, like Italy, Spain and France, where disinflation remained politically precarious, EMU served a political role in cementing the new disinflationary regime. Economically, tying the currency to Germany benefited the credibility of the new policies. More important probably was that European monetary integration strengthened the political acceptability of disinflation as it would allow strict monetary policies to be interpreted as policies promoting integration. Moreover a threat to the exchange rate, much more so than the failure to meet monetary targets, provided a clear signal of danger around which political support could be mustered. ^[31] Monetary integration also allowed politicians to detract attention from what was a domestic economic policy defeat by presenting a new foreign economic policy programme.

This latter mechanism was most obvious in France. Since the mid eighties France has claimed that EMU is desirable in order to end the unilateral dominance of Germany in monetary policies. Economically this argument does not make much sense, however. If France indeed wanted to increase its monetary policy autonomy, EMU would seem a rather awkward way. First, because the French representative on the ECB board will obviously be in a minority and will be independent of the French government. Secondly, because the Maastricht treaty institutionalises the German type of monetary policy which France allegedly wanted to change by means of EMU. Thirdly it should be pointed out that the official French argument is the exact opposite of the position taken in the Swedish EMU report. If France wanted to safeguard its monetary policy autonomy, then a floating exchange rate, rather than a monetary union with an independent central bank would be the obvious choice. Yet a floating exchange rate was something which French politicians did not wish simply because they lacked the domestic institutions with which to contain inflation in the absence of a clear external constraint. Given that for domestic reasons France needed the discipline of a link to the D-Mark, the argument that France was going to dethrone the Bundesbank however made very much political sense as it enlisted the traditionally strong feeling of grandeur in the service of disinflation.

In Sweden, by the way, something similar has happened. It is no surprise that shortly after the spectacular collapse of the SAP's "Third Way" policy, social-democrats started to argue that it was time to export the Swedish model to Europe by joining the EU.

For hard currency countries like Germany and the Benelux, who did not need the reference to an external constraint to make disinflation politically acceptable, the interest in monetary integration was primarily economic. Ever since the early eighties Belgium and the Netherlands have pursued aggressive domestic cost cutting strategies in order to promote domestic exports. Combined with restrictive monetary policies such a strategy however is vulnerable to revaluation of the currency. Accordingly both countries displayed a strong interest in a fixed exchange rate arrangement on the basis of disinflationary policies in order to be able to pursue a beggar your neighbour strategy. Germany has been less vulnerable to the effects of revaluation as its exports are to a large degree concentrated in less price elastic niche. Yet also German industry is not insensitive to revaluation. Chancellor Schmidt's successful effort to create an EMS was largely inspired by the wish to combine export competitiveness with restrictive monetary policies. Given the Bundesbank's restrictive policies a freely floating D-Mark would, no doubt, have appreciated substantially. Not only would weaker currency countries, in that case, have been able to continue with a more expansionary macroeconomic policy but in addition they would have been able to improve their export competitiveness relative to Germany. Moreover, given that the export sector is the economic engine of Germany, a severe real appreciation of the D-Mark would certainly have raised serious political opposition to the policies of the Bundesbank.

Chancellor Kohl, during EMU negotiations, admittedly stressed that progress towards political union was a quid pro quo for EMU. Yet the Maastricht treaty mainly delegated such issue to a subsequent Intergovernmental Conference (IGC). This IGC, which is currently under way, is in a complete impasse as no single country seems to be willing to actually make binding steps towards greater political union. If, Germany could only agree to EMU if it implied greater political union, then Kohl at present should be signalling that a successful start of EMU will require some breakthrough on political union. Indeed this would also seem likely to increase his support in the German electorate whose majority is unable to identify a good economic reason for EMU. Yet, instead Kohl has made a rather drastic turn in his arguments as he now holds that too much emphasis on political union would threaten EMU end hence presently is undesirable. It seems likely therefore, especially after the experience of period of D-Mark revaluation during the last four years and the disastrous employment situation in Germany, that Kohl has discovered similar benefits in EMU as Chancellor Schmidt in EMS.

Certainly it is not correct to claim that EMU was a necessary outcome of the economic policy constellation in Europe. The desire to promote integration for it's own sake does play a role in many European countries. Yet, this desire could only materialise around EMU in its present form because of the reasons given above. And the reluctance to review the Maastricht Treaty is not primarily informed by the fear to setback integration but by the fear to weaken the commitment to disinflation and hence return to the inflationary situation of the seventies and eighties, In other words, at present the French Generals are in the EMU driving seat. ^[32]

The closer one looks at EMU the more it must seem like an attempt to re-enact the interwar gold standard. Like in the interwar period, restrictive macro regimes were installed in response to escalating inflation. Like in the interwar period, politicians deemed an institutionalisation of such policies in the form of a quite rigid fixed exchange rate regime and independent central banks necessary in order to prevent a repetition of the inflationary sins of the past. And like the interwar period, durable mass-unemployment and social tensions resulted. At present, e.g., the virulently xenophobic French Front National has managed to become the most popular political party amongst blue collar workers on the basis of slogans like: "Globalisation threatens your job. The Front National fights globalisation." And although Jean-Marie Le Pen should not be compared to Hitler the parallels with the NSDAP's programme of economic autarchy cannot be missed.

It seems that Europe can handle a significant reduction in unemployment without risk of setting off a wage price spiral. Unit labour costs in several countries have actually been falling during the last year or so. Under the pressure of durable mass-unemployment even giant unions like the German IG Metall are loosing their ability to prevent local undercutting of wage agreements. Sweden has even experienced a slight deflation as the consumer price index fell by 0.1% between October 1995 and October 1996. A more expansionary monetary policy would hence be rather appropriate in Europe rather than the tightening which can be expected if EMU is realised.

Yet, in the longer run unsolved problems remain. As the failure to co-ordinate wage bargaining in the LO area during the 1995 bargaining round showed, one of the central problems which contributed to Swedish inflation may still be present. One might argue that in the longer run this problem will solve itself because mass-unemployment eventually disintegrates the strongest unions. Yet one might not want to wait that long with a change in policy. Moreover it is not clear whether an atomised labour movement will be desirable in times of higher employment. If wage bargaining definitively shifts to the local level one might well find out that the complete lack of co-ordination might actually increase inflation tendencies under tighter labour markets. The decentralisation of wage bargaining which the report advocates in order to adjust real wages downward (p. 222) may hence become counterproductive from the viewpoint of keeping inflation down.

At present the best option to contain possible nominal wage escalation in countries like Sweden might be a greater government involvement in wage bargaining in order to prevent particularly aggressive individual unions or firms from setting a spiral in motion. It is true that Swedish unions have traditionally resisted government involvement in wage bargaining. Yet, given the dramatic situation in the labour market it is not so clear whether this opposition is still so pronounced, especially if larger co-ordination and more relaxed monetary policies can be achieved that way. The Norwegian model in which the state can force unions to accept mediation might be worth trying. Although in order to make this instrument politically acceptable it's activation should not be at the discretion of the government. If governments can force labour market parties to accept mediation the risk exists that unions will interpret this as a purely political measure when used by conservative governments, and employers might have the same view under a social democratic government. Activation of such an instrument hence may be best left to a qualified majority of a board composed of union and employer representatives. Moreover the same board should appoint mediators. Where the state is needed nevertheless is to make the mediation outcomes legally binding on both parties to the conflict. Moreover, the government should have the right to ask the board to consider using compulsory mediation.

In the longer run, a larger degree of profit-sharing may be helpful. A substantial degree of profit sharing would help reduce wage demands even in tight labour markets. Moreover, it would also allow for a more differentiated wage setting, helping to undo some of the compression which has taken place over the last decades and which has caused so much trouble within the trade unions.

The biggest potential problem at present may, however be relations in the public sector. Whereas mass-unemployment does weaken private sector unions over the long run this may not be the case in the public sector. Rather unemployment and job cuts may lead to further politicisation here. It is important to notice that those who protested most virulently against the French budget cuts in connection with the EMU convergence criteria were public sector workers. The report's suggestion to increase competition in public services hence is useful. Moreover, one might consider maintaining the fiscal criteria as the only element of the present EMU treaty as it may help preventing that a relaxation of macroeconomic policies will contribute to an expansion of the in many cases already overgrown public employment instead of promoting private employment.

References

de Beus, Jos (1996), "De Veerkracht van de Sociaal-Democratie." Voor Socialisme en Democratie, No. 53, 1996.

Calmfors, Lars (1996)"Nationalekonomernas roll under det senaste decenniet - vilka är lärdomarna?" In: Lars Jonung, ed., Ekonomerna I Debatten - Gör de någon nytta? Stockholm: Ekerlids Förlag.

Collignon, Stefan, et al (1994), Europe's Monetary Future. London: Pinter Publishers.

Dow, Alexander C. & Scheila C. Dow (1989), "Endogenous Money Creation and Idle Balances" In: John Pheby, ed., New Directions in Post Keynesian Economics. Aldershot: Edward Elgar.

Fischer, Andreas M. (1996), "Central Bank Independence and Sacrifice Ratios." Open Economies Review, 7:5-18.

Fischer, Stanley (1994), "Modern Central Banking." In: Forrest Capie et al, The Future of Central Banking Cambridge: Cambridge University Press.

Forsyth, Douglas J. & Ton Notermans (1996), "The Political Consequences of Price Flexibility" ARENA Working Paper, No. 6, July 1996. Oslo: ARENA. Forthcoming in Italian translation in Stato E Mercato, 4/1996.

Goodhart, Charles, Forrest Capie & Norbert Schnadt (1994), "The Development of Central Banking." In: Forrest Capie et al, The Future of Central Banking. Cambridge: Cambridge University Press.

Hahn, Frank (1984), Equilibrium and Macroeconomics. Oxford: Basil Blackwell.

Hamilton, Carl & Dag Rolander (1993), Att leda Sverige in i krisen. Stockholm: Norstedts.

Kindleberger, Charles P. (1985), “Keynesianism vs. Monetarism in Eighteenth- and Nineteenth-Century France.” In: Charles P. Kindleberger, Keynesianism Versus Monetarism. London: George Allen & Unwin.

Kitson, Michael & Jonathan Michie (1993), Coordinated Deflation: The Tale of Two Recessions. Watford: Full Employment Forum.

Kratz, Anita (1996) Ingvar Carlsson. Erlanders siste pojke. Stockholm: Bonnier Alba.

Larsson, Mats & Hans Sjögren (1995), Vägen till och från bankkrisen. Stockholm: Carlssons.

Laursen, Karsten & Jørgen Pedersen (1964), The German Inflation 1918-1923. Amsterdam: North-Holland Publishing Company.

Lybeck, Johan A. (1991), "Sveriges väg mot EMS." Ekonomisk Debatt, 6:554-561.

Reinius, Ulla (1996) Stålbadet. Finanskrisen, Penserkraschen och Nordbankens Rekonstruktion. Stockholm: Ekerlids Förlag.

Riese, Hajo (1986), Theorie der Inflation. Tübingen: J.C.B. Mohr (Paul Siebeck).

Shubik, Martin (1975), "The General Equilibrium Model Is Incomplete And Not Adequate For The Reconciliation of Micro and Macroeconomic Theory." Kyklos, Vol. 28, Fasc. 3:545-573.

SOU (1993:20), Riksbanken och Prisstabiliteten. Stockholm: Finansdepartementet.

SOU (1996:158), Sverige och EMU. Stockholm: Finansdepartementet.

Stenberg, Ewa & Gunnar Örn (1996A), "Kronan hade inte skuggan av en chans." Dagens Nyheter, November 1, 1996: A6-A7.

Stenberg, Ewa & Gunnar Örn (1996B), "De spelade bort sexton miljarder." Dagens Nyheter, November 3, 1996: A6-A7.

Temin, Peter (1989), Lessons from the Great Depression, Cambridge: The MIT Press.

Therborn, Göran (1985), Arbeitslosigkeit. Hamburg: VSA:

Tobin, James (1980), Asset Accumulation and Economic Activity. Chicago: The University of Chicago Press.

Footnotes

^[1] My Translation.

^[2]See Kindleberger 1985 and Goodhart, Capie & Schnadt 1994.

^[3]"As always in a period of inflation, the monetarists appear to be winning the intellectual debate today, in periods of recession and unemployment, the tables are normally turned." Kindleberger (1985, p. 41)

^[4]See e.g. Larsson & Sjögren 1995 and Reinius 1996.

^[5]For an account of those dramatic days in November 1992 see Sweberg & Örn 1996A,B.

^[6]On this interpretation of the Great Depression see Temin 1989 and Kitson & Michie 1993.

^[7]See Forsyth & Notermans 1996.

^[8]Indeed the President designate of the ECB, Wim Duisenberg has built up a reputation for ultra-orthodox monetary policies during his tenure as president of the Dutch central bank.

^[9]See also de Beus (1996, p. 19) who argues that European social democrats should take the lead in developing an alternative to the TEU.

^[10]The US Federal Reserve might serve as a possible example here. The Humphrey-Hawkins act of 1978 specifies the following goal for the bank: "to maintain long-run growth of the monetary and credit aggregates commensurate with the economy's long run potential to increase production, so as to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates." Quoted in Fischer 1994, p. 265.

^[11] My translation.

^[12]See e.g. p. 125 were it is argued that a restrictive demand management policy in a situation of high inflation expectations will create unemployment because it raises real wages.

^[13]It is argued that the increase in unemployment during the early nineties was partly due to reduced export competitiveness resulting from high wage increases in Sweden (p. 125). However, this real wage increase is also an effect of a (mild form of ) disinflationary policy as it results from an attempt to cling to a nominal anchor (fixed nominal exchange rate) during a period when domestic inflation rates exceed those in most trading partners.

^[14]The German government can also directly influence the Bundesbank because it has the right to postpone a decision by the Zentralbankrat by a period of up to two weeks. Such an arrangement has not been included in the TEU.

^[15]Indeed, in a reservation to some of the report's conclusions, one of its authors, Nils Gottfries, argues that Sweden should not even join in the long run because the economic costs of loosing monetary policy autonomy are not outweighed by the political benefits.

^[16] My translation.

^[17]Ignoring political reasons for interest rate differentials seems to become a bit of a tradition in the Swedish debate. As Lars Calmfors (1996, p. 237) has pointed out "During September-November 1992 it was almost impossible to watch television news programmes without coming across some economist who claimed that the credibility for the fixed exchange rate could only be maintained if the budget deficit was quickly reduced. (..) However, almost no attention was paid to the fact that devaluation expectations might be related to the quickly rising unemployment which might become politically unacceptable and hence - as in previous cases - force an exchange rate adjustment.” My translation.

^[18]Private interest organisations at times also have used the central bank as a scapegoat. One might argue that, at least from the mid seventies to the mid eighties, the German DGB was actually satisfied with a monetary policy which created a base level of unemployment. The alternative would namely have been an nominal incomes policy which might easily have exacerbated internal tensions with the DGB.

^[19]See Fischer 1994, p. 286-288.

^[20]One could surmise, although it is not specifically stated in the report, that the main aim in proposing more independence for the Riksbank is to convince the public that inflation should be avoided. But in that case one would expect the report to have spent more energy on demonstrating why inflation is actually bad.

^[21]That the velocity of circulation can be highly variable is now generally accepted in Europe. In fact it was the reason for the shift from targeting monetary aggregates to targeting prices (i.e. the inflation rate and/or the exchange rate).

^[22]Fischer (1996) even finds a positive correlation between central bank independence and sacrifice ratios, i.e. in complete contrast to the credibility argument, disinflation in countries with independent central banks leads to more loss of output.

^[23]See Hahn 1984.

^[24]See e.g. Laursen & Pedersen 1964, Riese 1986, Tobin 1980, Dow & Dow 1993.

^[25]Riese 1986, p.217-8.

^[26]See also Kratz 1996, p. 169: "The mistake with the credit expansion was, according to Kjell-Olof Feldt, that Sweden had high inflation for many years which created expectations of continued inflation. (...) At present, fall 1995, we would be able to deregulate the credit market under the most idyllic of circumstances. Now we have low inflation expectations and an economy which is definitely not overheated.” My translation.

^[27] In a report commissioned by the European parliament Stefan Colignon (1994), e.g. has recently come to the conclusion that in order to depart from the low growth / high unemployment trap more expansionary monetary policies will be necessary which in turn require a nominal incomes policy.

^[28]For the same reason Keynesian fiscal policies cannot be expected to bring any serious relief. Fiscal expansion which does not rekindle growth is doomed to falter on escalating budget deficits. It is a historical misunderstanding to attribute the full employment of the first three decades after WWII to Keynesian interventionism. The nature of a regime which could rely on other than macroeconomic means for containing inflation provided the engine for growth. Whereas Keynesian policies at best allowed for some manipulation at the margins. As most West European countries had to experience during the seventies and early eighties, once the monetary regime has switched to a disinflationary stance, Keynesian countercyclical spending becomes rather ineffective, just at a time when it seems to be needed most.

^[29]Unification is said to have increased the importance for Germany of being able to pursue independent monetary policies (p. 242).

^[30]Apart from that, it is by no means clear that a Swedish refusal to participate in EMU will isolate it in all other policy areas.

^[31]As e.g. the period 1990-92 in Sweden showed.

^[32]For other examples of politicians fighting the last war see e.g. Temin 1989 who argues that the Great Depression resulted because politicians, in a deflationary economic situation still were combating the inflationary tendencies of the late teens and early twenties. For a more recent argument along these lines see also Hamilton & Rolander 1993 according to whom the deep recession in Sweden during the first part of the nineties was due to the Bildt governments desperate attempt to fight non-existent inflationary pressures.

[Date of publication in the ARENA Working Paper series: 15.4.1997]