The Economic and Monetary Union. The Interdependence of Monetary and Fiscal Policy in the Eurozone

Table of Contents

1 Introduction

2 Literature Review

3 Theoretical Discussion
3.1 Interaction of Monetary and Fiscal Policy
3.1.1 The Influence of Interest Rates on Government Deficits
3.1.2 The Impact of Fiscal on Monetary Policy
3.1.3 Monetisation of Debt
3.2 Game-Theoretical Approaches
3.2.1 Fiscal Consolidation as a Prisoners’ Dilemma
3.2.2 Fiscal Coordination as a Prisoners’ Dilemma
3.2.3 Fiscal and Monetary Policy as a Game of Chicken
3.2.4 Games of Monetary-Fiscal Interaction

4 The Political Economy of Eurozone Budgets
4.1 Budgetary Rules and the SGP
4.1.1 Arguments Against Budgetary Limitations
4.1.2 The Status Quo and Arguments in Favour
4.1.3 Evaluation of the Stability and Growth Pact
4.2 The Way Forward
4.2.1 Proposals for Changes and Supplements
4.2.2 Transaction-cost Limits to Improvement?
4.2.3 An Evolving Informal Structure

5 Conclusion

6 References

Acknowledgements

The author would like to thank Professor Eijffinger for valuable help, support, and guidance through his supervision. He is also very grateful to Didier Reynders, Belgian Minister of Finance, for letting him know his views about the future development of fiscal coordination within the Eurogroup, as well as to Professor Peter Praet, Director of the Belgian National Bank, and several officials in the European Central Bank and the Belgian Ministry of Finance for openly discussing the ideas developed in this thesis. Finally, my love goes to Frances for providing the right mixture of distraction and encouragement throughout my time in Bruges.

Abstract

This paper addresses the interdependence of monetary and fiscal policy in a monetary union and the ensuing consequences for the economic constitution of the eurozone. Monetary-fiscal interactions are approached from economic theory and game-theoretical perspectives, which provides the basis for a discussion of the political economy of the present institutional framework.

It draws extensively on the existing literature covering the linkages between monetary and fiscal policy as well as game-theoretical approaches to the interaction of the two major branches of macroeconomic management. The recently proliferating work on the political economy of Economic and Monetary Union is condensed into its main arguments and critically reviewed. This includes contributions to the understanding of the politics of economic policymaking as well as normative statements about the design of institutions in a monetary union. Most of the sources are papers that focus on specific issues or start from different assumptions, which implies that their conclusions are often diverging or not even compatible. While the thesis tries to provide a balanced representation of the scientific discussion in this area, it does at times stand for a clear-cut choice on controversial normative issues.

The paper consists of a theoretical and an applied part. After briefly reviewing the relevant literature in this field, the analytical section develops the necessity of fiscal coordination in a monetary union, based on the interdependence of monetary and fiscal policy as well as the strategic interaction of the institutional players involved. This insight is then applied to the present framework of macroeconomic policy in the eurozone – characterised by the fact that monetary policy is supranationally concentrated, whereas fiscal policy remains decentralised and national. The Stability and Growth Pact is discussed as an effective means of safeguarding the independence of the central bank, while falling short of the degree of fiscal coordination that would be desirable on the grounds of both stability and growth in Europe. This evaluation leads to a discussion of suggested improvements and a careful judgement of what kind of evolution would be feasible and advantageous for Economic and Monetary Union in the future.

The principal findings of this work lie in the field of the political economy of Economic and Monetary Union. It is demonstrated that the relation between monetary and fiscal policy necessitates not only a rule-based mechanism that avoids excessive deficits but also a more far-reaching coordination of budgetary decisions in order to establish monetary leadership and to safeguard the values of monetary commitment as a precondition for sustainable growth. The discussion shows that the present situation only insufficiently delivers these results and needs institutional transformation to improve the currently suboptimal state of affairs. This matter of fact, combined with the need for budgetary reform caused by the demographic development in our societies, is projected to lead to an initially informal and shallow modus vivendi that is based on the Eurogroup of Finance and Economics Ministers and will mark an incremental process towards more comprehensive concertation of fiscal policy.

Annex I: A Taylor Rule for Europe?

Annex II: The Impact of Interest Payments on German Deficits

Annex III: A Model with Politically Induced Deficit Biases

Annex IV: Stackelberg Leadership

List of Tables

Table 1: A Fiscal-Consolidation Prisoners’ Dilemma

Table 2: A Fiscal-Coordination Prisoners’ Dilemma

Table 3: A Fiscal Game of Chicken

List of Abbreviations

AWM Area-Wide Model

ECB European Central Bank

ECOFIN Council of Economic and Finance Ministers

EDP Excessive Deficit Procedure

EMS European Monetary System

EMU Economic and Monetary Union

EU European Union

GDP Gross Domestic Product

SGP Stability and Growth Pact

1 Introduction

The acronym EMU is often wrongly seen to stand for “European Monetary Union”. This corresponds to a belief widely held outside of our discipline: With the transfer of control over monetary policy to the European Central Bank (ECB) and the issue of euro notes and coins, a well-defined process of monetary integration in Europe is completed. – Far from that, EMU stands for “Economic and Monetary Union” and we are talking about the end only of the beginning.

It is an experiment unprecedented in history for both its political as well as economic dimensions. One of its unique features is that it combines a centralised and supranational monetary union with a set of independent, decentralised, and national fiscal policies. Therefore the conduct of monetary and fiscal policy within the EMU framework has become a subject of lively academic debate and is one of the most important questions of contemporary political economy. It requires us to leave the textbook-world of treating the monetary and fiscal fields of economic policymaking as separate and brings together the need for a better understanding of the linkages between the two as well as of the functioning (and the improvement!) of the corresponding institutional framework.

This Master’s paper aims to be a contribution to that effort. By discussing the macroeconomic relations between the two branches of economic policy and by reviewing their existing institutional environment, the author hopes to support the argument that the success of the project requires a profound improvement of the latter towards supranational coordination of fiscal policies. The first part summarises the academic literature on the interaction of monetary and fiscal policy, leading up to a game-theoretic interpretation of the structural setting. In the second part, we will discuss the political economy of the present institutional solution and show its shortcomings, pointing towards areas of potential improvement.

2 Literature Review

This section briefly summarises the most relevant contributions to the existing literature on the interaction of monetary and fiscal policy.

At the conceptual base of the problem, Kydland and Prescott 1977 discussed the issues of time inconsistency and credibility in a seminal article. They introduced the problem of time-inconsistent macroeconomic policy in the form of ex-post optimality as an additional constraint on ex-ante policy. Time inconsistency has turned out to characterise the incentive problems of many policy areas, in particular monetary and fiscal policy, as exemplified in the case of the inflation bias (see also Calvo 1978). It highlights the relevance of credibility issues (private agents anticipating the deviation from ex-ante optima and, with rational expectations and backward induction, completely anticipating the ex-post equilibrium), second-best environments (which can be turned into third-best outcomes due to rational anticipation), and institutional characteristics of the policy-making framework as possible solutions to the problem (commitment technology, rules over discretion). Its basic message is that lack of credibility is generic to many problems of monetary and fiscal policy and that it creates the need for efficient mechanisms that go at least some way in alleviating them. Persson, Persson and Svensson 1988 show that an optimal fiscal and monetary policy under commitment can be made time-consistent under full government discretion if each government leaves a particular maturity structure of the public debt to its successor. Calvo and Guidotti 1993 stress the relevance of an optimal taxation framework for the advantages of flexible monetary policy, based on the determination of monetary policy by public finance considerations and their impact on inflation. In their approach, monetary policy flexibility takes advantage of unanticipated inflation in a public finance context, e.g. by smoothing conventional taxes. Calvo 1988 writes about the problem of multiple equilibria of high interest rates and high repudiation of debt or low interest rates and low repudiation, triggered by the existence of government bonds. He sees the nominal interest rate as one of the main determinants of inflation due to public expectations about debt repudiation – hence a credible anti-inflationary policy should avoid unduly high nominal interest rates (e.g. by price indexation of public debt or by keeping the issuance of bonds to an interest-rate determined limit).

Among the research that has been done on the issue of central bank independence, Alesina and Grilli 1992 discuss the provisions for the ECB and Eijffinger and de Haan 1996 provide a critical review of the debate on central bank independence, including a discussion of the theoretical aspects as well as clear definitions, measurement approaches and a summary of empirical studies. The origin of discussions of government finance is usually associated with the Modigliani-Miller theorem as stated, for example, in Sargent 1987. However, analogous to corporate finance theory, the irrelevance proposition of the Ricardian-equivalence hypothesis breaks down when incentive incompatibilities along the lines of the principal-agent problem are introduced, as developed, among others, by Lucas’ and Stokey’s 1983 work on government debt structure. Bohn 1988 focuses on the role of nominal government debt in the context of discretionary and distortionary fiscal and monetary policy. Based on an overlapping-generations model with stochastic real and monetary shocks, he shows that nominal debt can act as a hedge in the face of real shocks by driving up inflation, which allows to government to raise finance via the inflation tax. However, the incentive to create surprise inflation is greater the higher the level of existing debt. Important empirical studies of the interaction of monetary and fiscal policy in EMU do not yet exist, given the short time horizon, as do simulations of future developments. The lack of the latter is also due to an absence of good macro models of the eurozone economy, of which the ECB’s recently published Area-Wide Model (AWM, Fagan et al. 2001) is unfortunately no exception.¹

Advances in game theory allow to discuss varying policy designs, such as time-consistency enabling mechanisms, and their efficiency performance across space and time. The initial attempt addressed the role of reputation and co-operation in a repeated game between policymaker and agents (Barro and Gordon 1983a, they show the possibility of low-inflation equilibria for a monetary policy game), an approach that is less insightful when applied to fiscal policy. Subsequently, game theory contributed to reputational models, among others, by Rogoff 1989, Chari and Kehoe 1993, Kotlikoff, Persson and Svensson 1988. It became evident that reputation cannot substitute institutional solutions to the time inconsistency problem but helps to explain their success. Institutional aspects were first discussed with respect to delegation, which was particularly interesting in the context of the credibility problems in monetary policy. Delegation to an independent central bank came to be seen as a possibility to relax binding incentive constraints in games of strategic interaction. The seminal article on the “conservative central banker” was Rogoff 1985. The contemporary game-theoretical debate on monetary and fiscal policy encompasses the contingencies of monetary union and its institutional framework, including for fiscal policy. There have been several papers on monetary-fiscal interaction in a monetary union, for example Sibert 1992, Levine and Brociner 1994, or Beetsma and Bovenberg 1998. Other studies have treated the desirability of fiscal constraints, among which Chari and Kehoe 1998 and Dornbush 1997 argue against fiscal limitations, whereas Beetsma and Bovenberg 1995 and Beetsma and Uhlig 1999 support them. The most important recent contributions are principally associated with Dixit, relevant articles include Dixit and Lambertini 2000a, b, and c. These issues will be taken up and addressed extensively in this paper after having discussed the linkages between monetary and fiscal policy from a perspective of economic theory.

3 Theoretical Discussion

The following section treats the linkages between monetary and fiscal policy, specifically looking at two major channels of transmission represented by the impact of monetary on fiscal policy and vice versa as well as at the game-theoretical implications of that interdependence. The fiscal issues related to EMU are in several respects more complex than the monetary implications, which is the major reason why the questions they pose have – unlike the monetary side – not yet received an adequate answer in institutional terms. The most important areas of scientific investigation and political debate include the issues of fiscal policy coordination, constraints on national budgets and debt levels, the stabilisation role of fiscal policy in the context of asymmetric shocks and its implications for a potential European budget as well as aspects of fiscal consolidation, tax competition and state aid. Given the limited scope of this paper, we will disregard tax and subsidy issues of more microeconomic orientation and limit ourselves to the question of the desirability of coordination among national budgetary policies and the underlying rationale derived from the interaction of decentralised budgets with a common monetary policy.

3.1 Interaction of Monetary and Fiscal Policy

The standard textbook approach to monetary and fiscal policy is to cast both as separable activities, often assigned to different bodies. According to this view, monetary policy has authority over short-term nominal interest rates and – more or less precisely – over the growth of money supply, whereas fiscal policy deals with tax rates as well as public spending and borrowing. In fact, almost the entire debate between Monetarism and Keynesianism can be framed within this dichotomy as a way of disputing which family of policies is better suited to manage the economy. However, as Bhattacharya and Haslag (1999, p. 26) state, “(a)lthough it may be appropriate to think of monetary and fiscal policy actions as separate ventures, it is important to understand that the two interact.”

In this section, we want to propose three arguments that sketch some of the ways in which this interaction takes place. Let us assume that the goal of monetary policy is price stability defined as low inflation π and its main instrument the interest rate i. Let us further assume that, besides distributional, stabilisational and other concerns, the goal of fiscal policy is growth (i.e. the rate of change of output) °y and its main instruments taxation T and government expenditure G. The essence of all following arguments is that, ceteris paribus, each instrument and goal of one policy-maker is a function of what is controlled by the other:

Abbildung in dieser Leseprobe nicht enthalten

It is impossible to cover and model the entire range of potential channels of transmission between monetary and fiscal policy, nor is this necessary in order to construct a sound argument in favour of fiscal policy coordination. We will therefore in the next chapters concentrate on the two most obvious relationships; that between interest rates and fiscal policy measures and that between budgetary policy and the determinants of monetary policy.

3.1.1 The Influence of Interest Rates on Government Deficits

The first channel of transmission can be treated in a very textbook-like manner before we embark on more sophisticated approaches. Suppose that, from an initially balanced budget, the government cuts taxes and leaves expenditure unchanged, thereby creating a deficit, which we define for the year t as

Abbildung in dieser Leseprobe nicht enthalten

Bt-1 is the resulting public debt at the end of year t-1, r being the real interest rate, thus rBt-1 representing the volume of real interest payments on the existing debt. The other component, G-T, is called the primary deficit. The government budget constraint says that the change in public debt is equal to the deficit in any given year:

Abbildung in dieser Leseprobe nicht enthalten

This shows that the deficit is linked to the initial level of debt (via interest payments) and to current spending minus taxes. Simple as this algebra might be, it shows how monetary policy impacts on fiscal policy: For a given level of inflation, the real interest rate is under the influence of the central bank, which sets the nominal interest rate independently of the government:

Abbildung in dieser Leseprobe nicht enthalten

Hence the central bank has a relatively direct influence on the size of the government’s interest payments and thereby on its deficit as well as the level of public debt. A good example of this relationship is the experience of Germany during the 1990s, where an elevated actual deficit occurred throughout the period despite the fact that the government earned more in taxes than it spent in expenditure (primary surplus) except for 1991 and 1993. The reason are the vast interest payments incurred on the existing level of debt, which due to the very high interest rates set by the then pronouncedly conservative Bundesbank, turned the deficit strongly positive.²

3.1.2 The Impact of Fiscal on Monetary Policy

It is more interesting to look at the relationship from the other end, i.e. to assess how fiscal policy impinges on monetary policy. We start with a discussion of de Grauwe’s argument (de Grauwe 1996, summarised in Eijffinger and de Haan 2000, p.81-82) that the debt ratio affects the equilibrium inflation rate and extend it to show that this drives up the interest rate before summing it up within the context of debt sustainability, which leads over to the Sargent-Wallace argument on debt monetisation.

Starting with the government budget constraint,

Abbildung in dieser Leseprobe nicht enthalten

we obtain the natural rate of taxation

Abbildung in dieser Leseprobe nicht enthalten

for stable debt, i.e. Bt = Bt-1. We now introduce a loss function for the government:

Abbildung in dieser Leseprobe nicht enthalten

which, if minimised, yields

Abbildung in dieser Leseprobe nicht enthalten

as the equilibrium inflation rate (a being the weight attached by the government to inflation in its loss function). Important about this result is that the level of public debt has a positive influence on the inflation rate. Combining this insight with a standard central bank reaction function as proposed by Taylor such as

Abbildung in dieser Leseprobe nicht enthalten

we see that the higher inflation rate increases the inflation gap as perceived by the central bank, which should force it to increase the nominal interest rate. In other words, “achieving convergence to a low debt to GDP ratio reduces the risk tat the monetary union will have an inflationary bias.” (Eijffinger and de Haan 2000, p.82) In fact, the author was confirmed by a Director of the Belgian National Bank that the recently more procyclical stance of fiscal policies in Europe is one of the reasons why the ECB is currently very reluctant to lower interest rates.³ A crystallisation of the preceding discussion and of potential dangers of fiscal and monetary problems reinforcing each other is provided by the issue of debt sustainability.

This argument is based on a trivial manipulation of the government budget constraint. Recalling that

Abbildung in dieser Leseprobe nicht enthalten

and requiring the debt ratio B/PY (PY being nominal GDP) to be constant at level k, we can state that

Abbildung in dieser Leseprobe nicht enthalten

where g is the real rate of GDP growth, and

Abbildung in dieser Leseprobe nicht enthalten

What this formula shows is that the required primary surplus to ensure a constant debt ratio k depends positively on this ratio and on the difference between the real interest rate and the real growth rate. As long as interest rates are lower than growth, rising or explosive debt tracks are easy to avoid. If however, as in the 1980s and 1990s, interest rates become strongly positive (partly due to the impact of fiscal policy, which induced the markets to require higher interest rates as risk premia!), problems of debt sustainability can become extremely serious.

Already now it is evident that the linkages between monetary and fiscal policy can manifest themselves in the form of a vicious circle where a higher deficit leads to higher interest rates, which again increases the deficit. This issue is closely related to the question of debt monetisation developed by the Sargent-Wallace argument, to which we turn next.

3.1.3 Monetisation of Debt

Sargent and Wallace’s (1981) article “Some Unpleasant Monetarist Arithmetic” was the decisive blow against the conceptual separation of monetary and fiscal policy and showed that neither can be seen to operate in a vacuum independently of the other. They base their argumentation on the assumptions of real interest rates being above the growth rate – i.e. the previously mentioned scenario of rising and potentially explosive debt tracks – and a given path of fiscal policy independent of monetary policy, hence a fixed set of values for the real primary deficit in a framework of fiscal leadership. The first assumption is generated by an overlapping generations model which ensures the necessary form of public demand for interest-bearing government debt. Both monetary and fiscal measures interact within the framework of the government budget constraint and a setting where fiscal policy dominates over monetary policy. Suppose that (a portion of) public spending is financed by selling government bonds and that current money growth is initially unaffected. As discussed above, the interest payments of the government increase, possibly amounting to more than the revenue from the additional sale of bonds. This leaves the “printing of money” as the only way to pay off the deficit, normally taken to be the open market acquisition of government bonds by the subservient central bank, which of course increases the money supply and, indirectly, inflation. Sargent and Wallace (1981, p.2) state that “(i)f the fiscal authority’s deficits cannot be financed solely by new bond sales, then the monetary authority is forced to create money and tolerate additional inflation.” They also show that the higher interest payments force the government to print money at a faster rate than what would have been necessary without the initial issue of bonds. As a consequence, debt finance via seignorage could ultimately be even more inflationary than direct inflation finance. One possible remedy would be to install the monetary authority as the first mover with primacy over the fiscal sequence, which is thereby disciplined and has to choose its deficits in accordance with the announced monetary policy. Alternative mechanisms are represented by a fixed exchange rate regime or a commodity standard (e.g. gold).

Sargent and Wallace’s approach has been criticised not for its results and the major insight of monetary and fiscal interaction but for the assumption of an interest rate exceeding the growth rate (see for example Darby 1984), which seems not to be confirmed by the data. However, Battacharya and Haslag (1999) and other recent contributions reframe the argument in a way that holds despite bond prices below growth rates if there is any asset with a real return exceeding the growth rate, such as equities – a plausible situation for almost all countries. These models still yield the result that the treasury’s financing needs could have to be met by seignorage despite rather low real interest on public debt.

So, at the end, we are still left with Sargent and Wallace’s insight into the influence of fiscal on monetary policy and vice versa, as they remark (1981, p.7) that “(s)ince the monetary authority affects the extent to which seignorage is exploited as a revenue source, monetary and fiscal policies simply have to be coordinated.” Among the empirical confirmations of this statement is, for example, a study by Grilli et al. (1991) who find evidence for a link between budget deficits and seignorage across countries, which obviously disappears for countries with independent central banks. The general conclusion of the pervasive interaction of the two wings of economic policy is central to our argument in the context of EMU. The second conjectural step we are going to take from here ventures into game theory and some of its applications for our purposes.

3.2 Game-Theoretical Approaches

Both fiscal consolidation as well as coordination can be discussed in the simple framework of a prisoners’ dilemma. This leads over to a representation of the interaction of fiscal authorities with the central bank in a game of chicken and finally an advanced model of monetary-fiscal interaction.

3.2.1 Fiscal Consolidation as a Prisoners’ Dilemma

One important game-theoretical issue is associated with the fact that EMU has been launched in the context of high levels of public debt and the need for fiscal consolidation. Based on fiscal linkages between member states, Allsopp, McKibbin and Vines 1999 argue in favour of a coordinated approach to fiscal contraction. They model fiscal consolidation in a monetary union as a prisoners’ dilemma where, if both countries cut expenditures, output losses are small and consolidation is achieved, whereas large losses without consolidation occur when one country cuts expenditures and the others deviate. This reasoning provides strong support in favour of the SGP and the Maastricht criteria as the solution to a collective action problem, helping to achieve EMU if seen as an exercise in co-ordinated fiscal retrenchment.

Allsopp et al. identify two major channels of influence, based on short-run nominal rigidities, supporting their view that the output-loss of coordinated retrenchment is lower than that of individual consolidation. Firstly, the constraints of a common monetary policy imply short-run Keynesian effects of falling output, employment and consumption while prices and wages are sticky. The assumption is that monetary policy could be more lenient if a coordinated reduction of the deficits occurred. Secondly, collective consolidation allows the internalisation of a positive externality in terms of lower interest rates and the corresponding beneficial developments for investment and growth, which any individual country could not entirely appropriate (due to the fact that the interest-rate is set union-wide) and therefore would underprovide consolidation. The externality-effect of consolidation on interest rates is precisely the source of free-riding incentives and the reason why it makes sense to look at this issue as a prisoners’ dilemma. The model itself is represented by supposing that each country in a monetary union aims to achieve fiscal consolidation by reducing government expenditure. It faces the policy alternatives ‘Cut’ and ‘No cut’. The payoffs are structured such that small losses for everybody are incurred if all countries play ‘Cut’, but no losses have to be borne by a country that deviates by playing ‘No cut’, whilst the others do cut and experience higher losses, which makes consolidation no longer worthwhile. Developing the approach of Allsopp et al. 1999, this is the game in normal form for a two-country framework:

Table 1: A Fiscal-Consolidation Prisoners’ Dilemma

Abbildung in dieser Leseprobe nicht enthalten

source: adapted from Allsopp et al. 1999

Under the premise to consolidate, ‘Cut / Cut’ would be the least costless solution, but ‘No cut’ is the dominant strategy for either country in a one-shot simultaneous run of the game, the outcomes being symbolised by a slightly positive or neutral payoff. It is most costly to cut expenditure while the other player deviates, leaving the consolidating country with a large loss. Allsop et al. simulate the game with an empirical model which confirms that, in EMU, collective fiscal consolidation is associated with a much lower level of output loss than unilateral retrenchment. They suggest “the Maastricht criteria, and the Stability Pact, can be seen as a response to this coordination difficulty by enforcing the cooperative solution between the fiscal authorities.” (1999, p.307) But they maybe overstretch their interpretation of the SGP in seeing it as a “process of game-theoretic reasoning by Europe’s leaders” (ibid.). Nevertheless, it seems evident that EMU allows to achieve fiscal consolidation collectively in an easier manner than would otherwise be possible unilaterally.

3.2.2 Fiscal Coordination as a Prisoners’ Dilemma

The prisoners’ dilemma of fiscal coordination also entails externality effects, which arise in three ways within the framework of a common currency: Firstly, we know from standard Mundell-Fleming analysis that a regime of fixed exchange rates (such as EMU or its institutionally less cemented precursor, the European Monetary System / EMS) enhances the role of fiscal policy in expanding domestic output, whereas the effect on foreign output is ambiguous. Higher export demand for the foreign country can be more than offset by higher interest rates (Eichengreen 1997), which makes the net spillover of fiscal expansion potentially negative – such as was precisely the case for most members to the EMS after German reunification. The size of this kind of effect might not be very significant but, with increased economic integration, fiscal linkages between countries will grow and thereby provide a case for policy coordination (Masson 1996).

Secondly, a European-wide integrated capital market will experience fiscal spillovers across national boundaries in the form of universally crowding-out private investment to some degree due a higher common interest rate induced by one or several national debt levels being too elevated or rising rapidly. A member state can then cause a negative externality by running a fiscal deficit and, more seriously, even a current account deficit (partially due to its fiscal policy stance) – i.e. if the sum of the deficit and private investment exceeds national saving (Pisani-Ferry 1996). This scenario is one justification for expressing the SGP provisions in terms of the actual deficit because it reflects more directly than structural or primary figures the burden on the financial markets (Thygesen 1996). Additionally, Artis and Winkler (1999, p.162) point out that “the conduct of fiscal policy, both deficits and debt, may affect both the short-run and the long-run credibility of the common monetary policy and thereby nominal interest rates, inflation expectations and actual inflation.”

Thirdly, fiscal policy in a member state has effects on the external value of the euro, which can be expected to increase in response to fiscal expansion and higher interest rates, thereby depressing the exports of the entire union. Taken together with the worst-case scenario of a de-facto bail-out of a member state in crisis and the corresponding damage to the common capital market, these three channels of externalities in the context of financial integration and monetary union provide the reasoning for a game along the lines of a Prisoners’ Dilemma with the following payoff matrix (adapted from Artis and Winkler 1999):

Table 2: A Fiscal-Coordination Prisoners’ Dilemma

Abbildung in dieser Leseprobe nicht enthalten

source: adapted from Artis and Winkler 1999

In analogy to the original game, a one-shot non-cooperative run produces ‘lax policy’ as the dominant strategy for both players, yielding the suboptimal result 0 / 0. Fiscal discipline would be the preferable outcome but is costly to implement, given the free-riding incentives for the other player. Conversely, the negative effects of fiscal expansion (higher interest rates, crowding-out of private investment and exports, risk of financial instability) are socialised throughout the union, whereas the deviating country can appropriate the benefits in terms of higher output and increased employment. Therefore it is rational for both sides to implement lax policy, which means that both countries are stuck with the Pareto-inferior Nash equilibrium at the bottom right.

The same reasoning holds for the domestic deficit bias, where different fiscal players within a country find it optimal to increase spending or to avoid budget cuts for their clientele and leave the socially optimal but costly consolidation to the others. Budgetary policy can also be modelled as strategic interaction in a dynamic game, which evolves as a ‘war of attrition’, where each side holds out as long as possible, waiting for the other to give in first and contribute to the public good of deficit reduction (Bliss and Nalebuff 1984, Alesina and Drazen 1991).

Interpreting debt reduction and fiscal restraint as a prisoners’ dilemma allows us to understand both the Maastricht criteria and the SGP as contracts that are designed to change the payoff-matrix of the players in a way that makes the co-operative outcome achievable. In order for them to work, they have to modify the results of lax policy such that ‘tight policy’ becomes the dominant strategy or ‘tight / lax’ and ‘lax / tight’ at least equivalent to ‘lax / lax’ in terms of the payoff-structure. The effectiveness of the SGP provisions depends essentially on the degree to which they achieve such a modification of the game in reality.

3.2.3 Fiscal and Monetary Policy as a Game of Chicken

The relation between fiscal and monetary policy is even more relevant for our topic and can also be studied in game-theoretical terms. The principal motive of the Maastricht criteria and the SGP is to facilitate the ECB’s primary task of monetary stability, i.e. stable and low levels of inflation. Based on the preceding discussion, we have seen that government finances in several direct and indirect ways impinge on monetary policy. The raison d’être of the SGP in this respect then is to safeguard the long-run credibility and independence of the central bank by preventing the build-up of dangerous levels of public debt and, in the short run, to contribute to fiscal discipline by reducing the risks of imbalances in the macroeconomic policy-mix. In the absence of a union-wide Stabilitätskultur and a stock of reputational credibility, the SGP seems to be a surrogate mechanism to ensure discipline and coordination needed to back up the “empty shell” of central bank independence (Artis and Winkler 1999). Its functioning can be modelled as a game of chicken between the institutional players involved:

Table 3: A Fiscal Game of Chicken

Abbildung in dieser Leseprobe nicht enthalten

source: Artis and Winkler 1999

Suppose that, in analogy to two cars racing at each other, policymakers are on a collision course in the sense of fiscal policy being lax and monetary policy tight: The outcome is an unbalanced policy-mix with high interest rates, currency appreciation and output losses, represented by a negative payoff (-1,-1). This is not a Nash equilibrium: both sides would be better off joining the line of the other, but the benefits are unequally distributed for the outcomes of symmetrical strategies (as one of the players is the ‘loser’). Hence both institutions prefer different Nash equilibria as the outcome; the central bank wants to end up top left, national governments would prefer bottom right. Social welfare is best served by making the ECB win, as the sum of the payoffs is greater in a ‘tight / tight’ outcome compared to the ‘lax / lax’ scenario – although the difference for the two players in terms of ‘winning the game’ is the same.

The purpose of the SGP then is to change the payoffs in such a way that ‘losing’ becomes attractive to the fiscal authorities. This is necessary as, at least in the early years of EMU, it is not at all certain that the ECB on its own will prevail in the game of chicken, backed up by credibility and popular support as the Bundesbank would have been, and equally important, that the fiscal authorities anticipate to lose anyway. In the terminology of Canzoneri and Diba (1996), the SGP essentially is about establishing ‘monetary dominance’ through a precommitment device binding the fiscal authorities, which – if it works – transforms the simultaneous game into one of sequential Stackelberg leadership where the central bank chooses the outcome it prefers on the reaction function of the fiscal authorities (see Appendix IV). This, at least in theory, should enable the socially preferable equilibrium to prevail and at the same time avoid costly leadership battles with mutually suboptimal outcomes.

In the long-run, where monetary and fiscal policy is linked via the previously discussed intertemporal government budget constraint, ‘fiscal dominance’ would imply the Sargent-Wallace scenario of a stream of precommitted net deficits and the ultimate necessity to impose an inflation tax via monetary policy. ‘Monetary dominance’, on the other hand, would mean that the central bank is credibly committed not to inflate and thereby not to bail out the debt-issuing government(s) who in turn are forced to adjust their financial paths accordingly. Artis and Winkler argue that “a deficit rule, like the one endorsed by Maastricht and the Stability Pact, can be shown to (…) establish monetary dominance and the ‘functional’ as opposed to ‘legal’ independence of the ECB.” (1999, p.173) If they work, that is.

In the short-run, the stock of debt can affect the credibility of monetary policy through the volatility of interest payments that it gives rise to and which impinge on the policy-mix of macroeconomic management. The 1980s in the US and the 1990s in Germany both demonstrate the high costs that an inappropriate policy-mix of tight money and lax public finances imposes on the economy. It can be interpreted as a collision of the monetary and fiscal authorities along the lines of a game of chicken, the central banks ultimately but costly prevailing. In terms of EMU, the SGP appears to be a crude instrument designed to limit the extent to which the ECB will either be induced to accommodate fiscal expansion and succumb to fiscal dominance or be forced to raise interest rates aggressively in order to re-establish fiscal dominance. It is a second-best alternative to the as yet unrealistic best of all possible worlds of full-scale fiscal coordination under monetary leadersip, a union-wide Stabilitätskultur and the corresponding institutional framework.

3.2.4 Games of Monetary-Fiscal Interaction

Among several voting models describing strategic interaction in the Governing Council of the ECB and the influence of member states on the central bank’s inflation rate, Avinash K. Dixit and Louisa Lambertini have produced a very insightful game-theoretical approach to the interaction of monetary and fiscal policy in EMU⁴. Based on a reduced-form Barro-Gordon type model (Barro and Gordon 1983b) extended to many countries, they describe how each country’s fiscal policy inflicts positive and negative externalities in the form of output expansion as well as higher inflation and increased interest rates on the other economies in the union, complemented by a common monetary policy that suffers from time inconsistency if it is not independent. In Dixit and Lambertini (2000c, p.2) they write that “(t)he ability of each country to choose its own fiscal policy leads to an equilibrium that is suboptimal for them all because of interactions of externalities.” If the net sum of externalities is negative, then the noncooperative Nash equilibrium leads to deficits and spending levels that are too high. Their model delineates what they call a ‘symbiosis’ of uncoordinated fiscal policies and a monetary policy that mitigates fiscal externalities as the first-best solution and, short of that, fiscal coordination under monetary leadership as the second-best outcome. We will briefly summarise their model as follows:

In a monetary union of member countries i = 1, 2, … n, a common central bank chooses its policy variable p 0. The fiscal policy variable of each country is denoted by xi, the GDP levels in the separate countries by yi and the common inflation level by p. p e are the private sector’s rational expectations of p. The individual GDP levels are attained as

Abbildung in dieser Leseprobe nicht enthalten

The parameter ŷi can be seen as the natural private output in the country. aii shows the GDP effect of the country’s own fiscal policy, whereas aij for j ¹ i are the fiscal spillovers from other countries, which Dixit and Lambertini debatably assume to be positive.⁵ The last term of (1) denotes an output effect of surprise inflation. In vector notation, the GDP levels of all countries can be aggregated as

Abbildung in dieser Leseprobe nicht enthalten

where A is the matrix of fiscal policy own and cross effects and b the vector of the supply effects of surprise inflation, both (as well as ŷ) being stochastic shocks. The common inflation level is given by

Abbildung in dieser Leseprobe nicht enthalten

where e’ represents transposes of the n -dimensional unit vector. Note that the inflation level is only partially controlled by the central bank and entails a weighted component of the sum of the fiscal policies. A strict interpretation of Europe’s monetary constitution would require the stochastic variable c to be zero, but Dixit and Lambertini arrive at the first-best solution even with a positive c, i.e. monetary accommodation. Countries are minimising their respective loss functions

Abbildung in dieser Leseprobe nicht enthalten

where ỹ i is potential output, which is assumed to be ỹ i > yi, and q i denotes the weight for a country’s preference for higher output (i.e. a lower output gap) relative to its preference for lower inflation. One could criticise equation (4) for the fact that inflation might give rise to structural costs indirectly related to a fiscal expansion, which does not feature in the model and therefore overstates the benefits of fiscal leniency. Dixit and Lambertini propose the following sequence of events in the game:

Firstly, the central bank chooses a function for its policy variable p 0 = p 0 ( ŷ , A, b, c) and thereby determines its future response to the stochastic shocks in the bracket. Then the private sector forms its (rational) expectations p e. Thirdly, the shocks occur, followed by the central bank setting p 0. If it sticks completely to its commitment, p 0 will be realised mechanically according to step one (as in a Taylor rule). Now the separate governments determine their fiscal policies xi. They can either cooperate in a framework of fiscal coordination or choose independently, which results in a non-cooperative Nash equilibrium. Interestingly, the first-best outcome of this sequential game does not depend on the order of the monetary and fiscal moves. In the case of fiscal coordination, the authors suppose that the union as a whole minimises a weighted average of the member states’ loss functions:

Abbildung in dieser Leseprobe nicht enthalten

where the weights w i could represent GDP shares, are positive and sum to 1. In vector notation, (5) becomes

Abbildung in dieser Leseprobe nicht enthalten

where the prime again denotes transposes and Q and W are diagonal matrices with the entries q i and w i. If the central bank differs from this common objective and is more conservative than the union, as advocated by Rogoff (1985), then it minimises

Abbildung in dieser Leseprobe nicht enthalten

A strict reading of the Maastricht Treaty would imply the ECB to be ultra-conservative, that is δ = 0, but the results hold for any arbitrary value of δ. Important is the relative weight attached by the two authorities to the output and the inflation terms, i.e. the case with δ = 0 can also be seen as one where the fiscal authorities do not care about inflation at all. The central feature of (6) and (7) is that both loss functions share the same ideal outcome, which thereby becomes the first-best solution: y = ỹ and p = 0 at any δ – in other words, output at capacity with no inflation. The rest of the model is about picking the right institutional structure to achieve this outcome. Dixit and Lambertini consider four alternatives: Fiscal coordination with monetary discretion, fiscal coordination with monetary commitment, fiscal non-coordination with monetary discretion, and fiscal non-coordination with monetary commitment. The first scenario turns out to yield the first-best outcome under the condition that both sides agree about the optimal solution.

When the fiscal authorities act cooperatively and coordinate their actions with those of the central bank, we get the following FOC’s:

Abbildung in dieser Leseprobe nicht enthalten

In vector form: Abbildung in dieser Leseprobe nicht enthalten

Abbildung in dieser Leseprobe nicht enthalten

In vector form: Abbildung in dieser Leseprobe nicht enthalten

Equation (8) is the FOC for fiscal policy xj in country j, whereas (9) represents the condition for an optimal common monetary policy. Multiplying (9) by ce’ and substituting into (8) yields

Abbildung in dieser Leseprobe nicht enthalten

which has ỹ – y = 0 as the only solution, hence actual output equals potential output, which on its turn delivers p = 0 when substituted into one of the FOC’s. When rational expectations hold, the private sector will attain p e = 0. Putting these values into (2) and (3), we get the solutions of the policy variables that achieve the first-best outcome of ideal GDP levels in all countries and zero inflation for all realisations of the stochastic shocks:

Abbildung in dieser Leseprobe nicht enthalten

and

Abbildung in dieser Leseprobe nicht enthalten

The nature of the first-best result, ‘fiscal-monetary symbiosis’ in Dixit-Lambertini terminology, is not that there are no inflationary effects from fiscal policy, but they can be countered by appropriately setting p 0, and there is no need to have recourse to surprise inflation in order to attain potential output.

The first-best outcome can also be achieved in the second scenario, where fiscal policy is coordinated within a framework of leadership by a committed monetary policy. It requires that monetary policy be conducted at an earlier stage, leaving the fiscal authorities with a subgame of the sequential structure. Our interpretation of this scenario is that the central bank can pick the subgame-perfect outcome on the fiscal reaction curve very much like a Stackelberg leader in sequential quantity-setting games.⁶ It is necessary to stress that the existence of the symbiosis-outcome depends on the strong assumption of agreement between monetary and fiscal authorities on the optimal levels of inflation and output. The best of all worlds disappears when this assumption is relaxed. Dixit (2000) takes a slightly different approach but arrives at the same result and concludes for the absence of symbiosis: “(F)iscal freedom with discretion entirely negates the value of monetary commitment.” (Dixit 2000, p.25) The models used in this discussion are not fully conclusive as to whether monetary discretion is preferable to monetary commitment, but the fundamental result is that “(t)o the extent that monetary commitment is desirable, fiscal constraints need to be designed optimally to retain that value of commitment.” (ibid. p.26) Our interpretation of these findings is that, short of the rather hypothetical situation of monetary-fiscal symbiosis, institutional answers must be found to the question of how best to establish and protect the benefits of monetary stability against the sub-optimal scenario of non-cooperative fiscal Nash equilibria. The question is not whether we should coordinate fiscal policies in EMU but whether the SGP does that to a sufficient extent. This is the topic of the following section, which addresses the issue from a perspective of political economy.

4 The Political Economy of Eurozone Budgets

A purely technical economic perspective on the institutional solutions to the issues evoked by EMU is invariably bound to be incomplete and even misleading. Willet, among others, argues that “a broader public choice or political economy perspective is required.” (Willet 1999, p.38) Under political economy considerations, one discovers strong pragmatic arguments in favour of the SGP stipulations and also in support of more far-reaching arrangements of coordination. This is essentially because, again Willet, “they should be seen not so much as external limitations being imposed on unified national actors, but rather as processes which strengthen the hands of some actors – i.e. pro-stability monetary and fiscal officials vis-à-vis others in the domestic political economy process.” (ibid., p.39) One significant consequence of a political economy approach is that simplicity and transparency become the most important requirements for successful rules, which often makes the institutional settings second-best from an economic perspective.

4.1 Budgetary Rules and the SGP

This sub-section juxtaposes the arguments against and in favour of budgetary norms in general and evaluates the SGP in particular. This will provide some ground for analysing suggestions for improvement as well as the contemporary evolution of the policymaking framework.

4.1.1 Arguments Against Budgetary Limitations

The standard counterargument to budgetary rules in general and the SGP in particular is that financial markets should be sufficient in exerting control over public spending.⁷ However, applying this reasoning to a monetary union requires an implicit premise to be satisfied, which is that the markets distinguish the different public sector debtors in the eurozone and price credit risk accordingly. Yet, based on experience in federal states, it seems that “the premia likely to develop on sovereign debt issued by EMU governments may be unable in themselves to trigger restraint in government behaviour.” (Thygesen 1999, pp. 20/21) The argument against budget rules is further weakened by the fact that financial markets have in fact upgraded the creditworthiness of weaker treasuries in the run-up to EMU⁸, which means that market participants anticipate the bailing-out of a member state in financial distress. Such a scenario distorts government incentives and encourages free-riding in the form of running excessive deficits. Thirdly, financial markets, being far from perfect, react usually with a lag, have a tendency to overshoot the equilibrium rates and can cause contagion effects as well as spillovers to other countries. So, in conclusion to this argument, “it remains doubtful that market discipline is effective, let alone efficient, in offsetting a domestic political deficit bias.” (Artis and Winkler 1999, p.161)

An argument with Keynesian flavour against fiscal rules in general and specifically the SGP states that it sets the stage for a counterproductive interaction of the ECB and the fiscal authorities, resulting in a suboptimal or even damaging policy-mix. As Hallet and McAdam (1999, p.216) write about the interdependence of fiscal and monetary policy, “the costs of non-cooperation behaviour between those two instruments can be rather high” – higher than the effort needed to develop a comprehensive coordinative framework for EMU. It seems plausible that a more encompassing and coordinative institutional structure could not only help to solve problems that the existing SGP is unable to address but also alleviate some of the concerns that authors have about the present setting on its own, such as Hallet and McAdam’s (1999) warning with respect to the long-term implications of the SGP on potential growth in Europe. However, it is debatable whether the full-scale concertation of macroeconomic policy which they propose is really necessary or whether fiscal coordination within a framework of monetary leadership is the better solution. The argument of critics like Hallet and McAdam⁹ is that the SGP implies a large-scale fiscal retrenchment across the union, leading not to financial solidity but to a loss of investment and output capacity in the long run. In their view, the benefits of economic stabilisation and consolidation have to be discounted against the costs of arguably inevitable conflicts between the ECB and the treasuries. Nordhaus (1994) provides a similar argument in predicting non-cooperative deficit reductions leading to higher unemployment in order to pressurise the ECB into monetary leniency, the suboptimal policy-mix then yielding output and consumption below trend for a decade or more, worsened by substantial capital outflows. The baseline of both arguments is that “if we constrain fiscal or monetary policies, we lose the ability to adjust the policy mix to get the outcomes we want.” (Hallet and McAdam 1999, p.217) The authors have in mind to achieve a lower deficit ratio, if at all, only while at the same time keeping output constant through monetary expansion in order to avoid real prices falling below real interest rates, which would result in the opposite of what the SGP was intended to achieve, namely rising levels of deficits and debt. Through an econometric simulation, they claim to have detected a pro-cyclical deflationary bias in the SGP, turning it into an “‘instability pact’ which imposes fiscal discipline in the short run, but destroys it in the long run.” (Hallet and McAdam 1999, p.218) The panacea they propose is “to maintain growth rates through a careful coordination of fiscal and monetary policies” (ibid.)

However, macroeconomic coordination is not the same as fiscal coordination. What Hallet and McAdam as well as some other authors of a Keynesian persuasion argue for, at least implicitly, is the subjection of both monetary and fiscal discretion to a common guidance, such as the gouvernement économique supported by France, based on a tradition of bureaucratic dirigisme. On the other hand, we would argue that fiscal coordination is sufficient within a system that clearly establishes monetary leadership. The SGP allows enough freedom of manoeuvre to avoid Hallet and McAdam’s worst-case scenario and is not necessarily procyclical, thanks to generous exemptions from the excessive deficit procedure (EDP) in the face of significant adverse shocks. In fact, they modify the SGP provisions by introducing a 1% deficit limit instead of the 3% allowance in order to get their undesirable results, based on a simulation exercise with debatable assumptions. The one remaining shortcoming the SGP could be reproached for is probably that it might, instead of going too far, not go far enough in clarifying the rules of the game – that is a game of Stackelberg leadership for the central bank.¹⁰ If that were fully established, the system would lose the costs as well as the risks of non-cooperative behaviour between the fiscal and monetary authorities that lies at the base of Hallet and McAdam’s concerns. Having discussed arguments against fiscal constraints, we now look at some that support these arrangements.

4.1.2 The Status Quo and Arguments in Favour

The present situation is defined by the SGP, essentially a rule-based mechanism for national fiscal policies in the eurozone with the aim to prevent deficits rising to and above 3% of GDP. The three basic arguments advanced in favour of binding guidelines and upper limits to national deficits are: (1) maintaining a proper policy mix (as high public debt can incur unnecessarily high interest rates), (2) avoiding union-wide interest rate effects from misbehaviour of individual members, and (3) offsetting additional incentives for a lack of fiscal discipline.¹¹

Budgetary discipline shall be secured by the threat of the EDP, which foresees a system of lagged sanctions and fines in the event of a non-recessionary overshooting of the 3% criterion by a member state. The rationale for finding ways to ensure fiscal discipline alongside a stability-oriented monetary policy is based on the insights that the gravest challenges to monetary solidity under an independent central bank stems not from monetary but from fiscal policy in general and, as argued, among others, by Giovannini and Spaventa (1991) and Grilli, Masicandaro, and Tabellini (1991), from public debt in particular. Fiscal rules guaranteeing the long-run solvency of member states are useful because of their credibility-enhancing effect on monetary union. We postpone arguments (1) and (2) and now address in more detail point (3), which justifies the SGP on the reasoning that monetary union increases already existing incentives of a government to run deficits. This point provides the most persuasive cause in favour of budgetary rules among the three mentioned above.

The argument is based on the rationale that countries generally and independently of a project of monetary unification have a bias towards running deficits. A public choice argument for this proposition is, for example, that beneficiaries of tax cuts and expenditure increases are politically more active than the ordinary taxpayer and thereby exert pressure to underfinance and overspend beyond a socially optimal level. Such a tendency is strengthened by entry into a monetary union with other countries due to a number of additional free-riding incentives. These arise as the probability of higher inflation caused by excessive deficits is reduced thanks to a union with more disciplined members and because, if it occurs, the inflation-cost of one country’s deficit now falls onto all member states and therefore represents only a fraction of the original cost to the initial culprit. A third important incentive is the fact that EMU should deliver lower interest rates on average, thereby lower costs of borrowing in terms of lower interest payments, which makes public debt seem more attractive.

A technical approach of this line has been has been developed by Roel Beetsma (1999, for an adaptation see Appendix III). Here the deficit bias is generated by ‘political myopia’ based on a non-zero probability that the incumbent political party will not be re-elected and therefore has not to take responsibility for the future costs (in terms of higher taxation) of present deficits. This bias is exacerbated by membership to a monetary union, since a fiscal expansion is less likely to push up inflation – a mechanism that used to constrain fiscal policy before monetary union and therefore provides a good reason for rule-based mechanisms that enforce fiscal discipline within the union. Beetsma argues that excessive deficits in one or several member states, notably the big ones, will create pressure on the ECB to water down its anti-inflation stance below a socially optimal level and monetarily accommodate the lack of fiscal prudence. It is important to note that explicit external pressure on the ECB by a particular government can already be preceded by a shift of opinion among the ECB’s governing council, when representatives of countries with high debt become more prone to favour monetary laxity. Gerlach (1998, p.110) argues that “the fact that a tightening of (monetary) policy is likely to worsen the debt situation and increase the risk of financial instability may lead the average (ECB) council member to be marginally less willing to tighten or marginally more willing to relax monetary policy. Large public debts could therefore impart an inflation bias to the ECB’s monetary policy.”

The argument of institutionally limiting governments’ discretion has therefore recently been seen in analogy to the rationale for a conservative central banker (notably Agell, Calmfors and Jonsson 1996). The desirability of such limitations is enhanced by the insight that “one cannot safely assume full effectiveness of the ‘no-bail-out’ and central bank independence provisions of the Maastricht Treaty” (Willet 1999, p.47) plus the aspect that the ECB, despite its institutional independence (see Eijffinger and de Haan 1996), is a young central bank without a stock of credibility and is already now under heavy fire from several governments. In fact, according to Beetsma, a fine increasing with the size of public debt can exactly do the job of correcting the deficit bias exacerbated by monetary union. The design of the SGP seems to be in line with these proposals (although it focuses on deficits rather than on the debt level) and represents an important step in the right direction, but several authors and already quite some experience in practise point to a number of shortcomings.¹²

4.1.3 Evaluation of the Stability and Growth Pact

Arguments (1) and (2) mentioned in favour of budgetary rules, i.e. the need to improve the policy mix and to avoid interest-rate effects from fiscal laxity, seem to demand a solution that goes beyond the SGP as it stands now and “give support to better coordination of non-monetary stabilization policies in the member states”. (Thygesen 1999, p.21) We will take a number of issues in turn and thereby assess the SGP as well as its deficiencies.

Anticipating the conclusion, one could quote Artis and Winkler (1999, p.176) who argue: “The Stability Pact attempts to pre-empt any potential leadership battles between fiscal and monetary policy in favour of the ECB and to prevent an unbalanced policy mix of a lax fiscal stance and tight money. (…) The Maastricht limits are an imperfect substitute for full coordination.” But Willet (1999) provides an interesting political-economy issue pointing to an important deficiency of the SGP: According to his analysis, fiscal limitations are needed to protect the economic targets of the union but also to offset domestic political pressures. However, a rule-based system like the SGP is likely to increase political pressures against it and at the same time to reduce the room for fiscal manoeuvre. If policymakers diverge from the attitude in which the SGP was framed, there is little hope that the institutional setting will prove effective in ensuring discipline, which in a context of low growth, high unemployment and political priorities might turn out to be wishful thinking. However, Willet can be criticised for being overtly pessimistic since the SGP was drafted and the Maastricht deficit criterion more or less achieved in the less then favourable circumstances of the last decade. Nevertheless, the basic concern of a serious political weakness of the pact is confirmed by a power-political point advanced by von Hagen (1998, p.18): “In large countries (…) the role of external political constraints (…) is simply too weak to coerce internal politics.” But what holds generally for the institutions of a polity is unsurprisingly also true for its monetary and fiscal constitution: Things can be and are subject to change when enough political momentum has built up. We cannot expect the SGP, of all social norms, to be an exemption from the general rule that their power of influencing the forces of change is reciprocal and possibly weaker than the developments it tries to contain and to predetermine.

A more pragmatic dilemma of the SGP is the fact that the fines it envisages as a last resort precisely worsen the problem of high deficits and thereby potentially aggravate the financial situation of the country concerned. This argument is voiced, among others, by Andersen and Dogonowski (1999, p.88): “(W)e find that there are realistic recessions which will bring countries into problems with the 3 per cent budget norm, even if they have a structurally balanced budget, and in which the escape clause defined in the Stability Pact will not apply. Moreover, in a probable long transition period, most countries will have a structural deficit and thus be very vulnerable to even mild recessions.” On the other hand, it seems laudable that the EDP is designed in a way that is bound to spark a public debate across the union, which would probably alert the financial markets and thereby lead to higher risk premia on the non-adhering country’s debt. This is to say that the effects of peer pressure and political and financial stigmatisation might become effective before the fines would have to be brought into play. But can we simply assume that the yet hypothetical public debate across the EU in such a case would actually result in a clear and strong support of fiscal discipline and monetary stability? This would doubtlessly be the case in Germany, whose monetary institutions have come to shape the European ones, but we should not underestimate the importance of the historically founded tradition of Stabilitätskultur. Sure enough, the Netherlands prove that a strong social preference for monetary stability does not need the traumatising experience of hyperinflation, but it is far from certain that these attitudes will spread swiftly and extensively across the eurozone.

The argument can be taken even further in displaying the SGP itself as a potential danger to monetary independence: Suppose higher deficits or purely monetary conditions spur the need to raise interest rates – countries with high levels of public debt will immediately blame the ECB that this increases their interest payments to an intolerable extent and thereby pushes them closer to or beyond the 3% limit. In this way the pact provides additional ammunition for high-debt countries to attack the ECB. This scenario is not at all unlikely at least for economies with a short-term maturity structure of their debt, which immediately feeds through the effects of higher interest rates to higher deficits. Artis and Winkler (1999, p.176) formulate that “(t)ying the fiscal authorities’ hands may well turn out to increase (sic.) rather than decrease the burden on monetary policy with respect to stabilisation policy.” With regard to the loss of fiscal discretion for a country in such a situation, de Grauwe points to the similar danger of a perverse effect of the SGP (1997, p.207): “As countries will be hindered in their desire to use the automatic stabilisation in their budgets during recessions, they will increase their pressure on the ECB to relax monetary policies. Thus, paradoxically, the stability pact whose aim it was to protect the ECB from political pressure may in fact have increased the risk of such pressure.” On the other hand, these concerns are limited to the extreme and unlikely case of a country coming in conflict with the SGP out of its own accord and not due to a recession.

Other commentators criticise the SGP for not taking into account debt ratios but focussing instead exclusively on budget deficits. A weakness of the present system seems to be that it applies identical deficit rules to countries with fundamentally different positions as far as long-term sustainability is concerned. The long-run financial health of the eurozone is rather determined by the stock measure of the largely differing debt ratios across the union than the less indicative flow measure of the deficit, which fits the SGP with too much of a short-term perspective. This valid critique leads to a more general appraisal of the pervasive lack of definitions and the vagueness of at least some of the provisions of the SGP, which opens, deliberately or not, the way for further politicisation of an essentially economic issue. On the other hand, one could provide the counter-argument that the Bundesbank Law itself was even less precise and deliberately designed as a “lex imperfecta” that should cause politicisation and a public debate in case of conflict between the central bank and the government. But the counter-argument loses much of its bite when one considers that such a provision, in order to enhance central bank independence, depends on the strong presence of some Stabilitätskultur, which – as discussed above – is arguably lacking in the EU as a whole. So it seems justified to disapprove of the lack of detailed concreteness of the Stability Pact, but then again its simplicity also appears to be an asset worth balancing against thorough specification, if one bears in mind that macroeconomic management in democratic societies is also always political and thereby ultimately a public process that benefits from simple and transparent rules.

Many commentators blame the SGP for simply being too weak in its application, especially with regard to the fudging and fiddling that was wide-spread in order to meet the Maastricht criteria and which might have set a bad precedent for the future application of the EDP. The pact is hence seen as a “Paper Tiger” lacking teeth through cumbersome and politicised procedures. Eijffinger and de Haan (2000, p.89) write that “(i)f national fiscal policy is restricted, it is important that the restrictions are credible.” Of course, this ultimately depends on the political will of the stability-oriented countries in the euro area to exert their influence over those who will try to take the edge off the pact. At least some concern if not outright doubt whether that would effectively happen seems justified in this respect. On the other hand, the more countries have reached a comfortable debt ratio, the more stability-oriented can we expect the Council of Economic and Finance Ministers (ECOFIN) to become. The relative strength of the EDP is hence influenced by the general economic climate in the union and, especially in these early years, would benefit from a phase of sustained growth that could help a majority of countries reach debt levels below 60 percent of GDP.

Taking a micro-economic approach, Corsetti and Pesenti (1999) provide an analysis of the long-run welfare effects of the budget cuts implicitly stipulated by the SGP arrangement. They look at the spillovers and interactions among member states and conclude that also these considerations point strongly towards the need for stronger fiscal coordination. Eichengreen and Wyplosz (1998, p.106) address a different problem and point out that “(t)he problem with the Pact as presently framed is that it is all stick and no carrot; rewarding good behaviour in booms rather than, or in addition to, punishing bad behaviour in slumps would surely make better sense.” Their point is valid but its realisation, as it has sometimes been proposed e.g. in the form of tradable deficit permits (Casella 1999), would imply at least some indirect form of redistribution across member states, which – at least for the time being – seems out of reach.

The preceding discussion can be neatly summed up in the following statement by Willet (1999, p.60): “The basic form of the Stability Pact is not necessarily seriously deficient. What is clear, however, is that it is not sufficient by itself to meet its objectives.” In particular, the present arrangement faces three kinds of problems: Firstly, it is doubtful whether its enforceability will be sufficient, secondly, if it should be possible, a full enforcement of the SGP provisions in the absence of budgetary reforms in the member states could generate substantial costs, predominantly in terms of lost flexibility vis-à-vis asymmetric shocks, and thirdly, this situation could possibly lead to the perverse effect of increasing the pressure on monetary policy. Related to the second point is a concern voiced by Eichengreen and Wyplosz (1998, p.69) with respect to the difficulties of achieving economic reform: “The danger is (…) that the Stability Pact will divert efforts from the fundamental reforms needed.” It can already be observed that countries in the presently favourable economic climate think it sufficient to achieve their deficit targets based on the existing organisation of their economies instead of tackling budgetary reforms. What is needed instead is a concerted effort to implement the goal of article 99 of the Maastricht Treaty in the sense that the member states should aim to make their budget procedures conducive to fiscal discipline. Beyond that, there is the daunting challenge of preparing the national budgets for an ageing population and all the consequences this has for pension systems and health budgets. Countries would indeed be very ill advised to think their tasks achieved by not violating the SGP. Lately, the Commission has become increasingly active in reminding governments of the pensions problem, and a number of countries have started to tackle the issue seriously. It is to hope that, instead from deflecting from it, monetary unification provides the unique chance to piggyback badly needed reform to the EMU project.

Taking the previously derived game-theoretical shortcomings together with the political-economy treatment of the SGP, its second-best nature becomes clear and can be traced back to political constraints. This view is shared by Artis and Winkler (1999, p.159), who write that “(c)learly more sophisticated rules could be derived from public finance considerations (Buiter, Corsetti and Roubini 1993). Equally clearly it would be preferable to address the (political) (sic.) distortion of budget policy at source if possible, instead of imposing arbitrary numbers. However, the Maastricht criteria are best understood as a simple commitment device in a second-best world of incomplete contracts.” Both our analysis as well as events since the establishment of the SGP – only the most obvious example being provided by the ‘Irish case’ of cyclical fiscal policy in a time of budget surpluses – confirm that the present arrangement is not yet the absolute solution, or in the words of Thygesen (1999, p.32): “The simple rule-bound policies that have been given prominence in the Maastricht framework, useful as they are, can hardly be the full story of economic policy in EMU (…)” The desirability of coordination of fiscal policies in the eurozone seems now established. The remaining question is whether a more formal approach is needed or whether the present setting evolves in such a way that no further formalisation is needed.

4.2 The Way Forward

Having discussed the rationale of the SGP and its shortcomings from an economic, game-theoretical and political-economy perspective, based on the interaction of monetary and fiscal policy in the eurozone, it is now time to look at what change should take place based on this analysis. It is a matter of fact, as Artis and Winkler (1999, p.177) write that “(t)he process of institution-building in the eurozone is (…) as yet far from complete.” They go on to say (ibid., p.183:) “(C)onceding concerted fiscal discipline in order to safeguard the leadership and credibility of the ECB may be a first significant step towards further implicit or explicit coordination among economic policy makers.” The preceding sections of this thesis have had the aim to justify further coordination of fiscal policies in euroland, whereas we now turn to proposed and actual developments.

4.2.1 Proposals for Changes and Supplements

Both academics and practitioners have become aware of the insufficiency of the institutional organisation of macroeconomic management at the European level. Out of the increasingly voluminous literature has emerged quite a range of interesting proposals for changes and supplements, especially with relation to the SGP.

The most far-reaching among these is also the most commonly voiced one and envisages fiscal federalism or at least some sort of intra-regional insurance for the eurozone. This would supposedly address the lack of coordination as well as any deficits the present SGP might have in view of countering asymmetric shocks by a European system of stabilisation and transfers. A strong preference for fiscal federalism can be associated for example with Eichengreen 1992, but many other authors have taken that line. A concrete example of how such a scenario could look like in practice is provided by the European Commission (1993): The Commission system is based on cross-country transfers triggered by unemployment differentials and is designed to function without the need for shifting vast amounts of money (the envisaged increase of the budget lies at 0.2 percent of EU GDP). Also Kletzer (1999, p.119) argues that “monetary union raises the welfare benefits of introducing a system of fiscal insurance between member states.” His analysis is based on a general equilibrium model and entails a compensation mechanism for the asymmetric effects of an independent monetary policy, this system being arguably better at addressing asymmetric shocks than fiscal decentralisation. Williamson 1990, Eichengreen 1992 and Von Hagen and Hammond 1997 mention similar inter-member fiscal redistribution mechanisms. Jensen 1999 provides analysis of how the stabilisation of the effects of asymmetric shocks could occur through international risk-sharing, transfers being triggered by the magnitude of shocks and geared towards the specific goal of redressing real output divergence within the union by short-term stabilisation. However, he states in his conclusion (Jensen 1999, p.141) that the proposal is “quite unrealistic for political reasons, however desirable” and mentions that moral hazard problems would undermine the functioning of both fiscal federalism and more moderate forms of fiscal transfer systems. Therefore, as interesting as these suggestions might be for theoretical reasons, they share a very hypothetical nature as their greatest disadvantage and seem politically infeasible at least in the foreseeable future, simply because of the fact that the EU is not a mature federation. On top of that, their benefits are theoretically disputed, e.g. by Thygesen, who concludes his analysis (1999, p.34) as follows: “The case for a scheme of fiscal transfers as an essential complement to EMU is currently unproven.”

A rather sophisticated proposal to improve the SGP is to impose budgetary discipline by a mechanism linked to national risk ratings based on the debt ratio as a superior alternative to the deficit criterion. The theoretical exposition of this idea is to be found in Bishop 1990 – but it also remains of hypothetical value due to the lack of political will to seriously consider such a procedure. The problem of the questionable enforceability of SGP sanctions is addressed by Corsetti and Roubini 1993 and Gros 1996 who propose the loss of EU voting rights for countries with excessive deficits as self-enforcing sanctions. Unlike the not very credible imposition of fines, the loss of voting rights or, for that matter, the suspension of EU payments would not have to be externally enforced by the other member states, which are probably susceptible to generous leniency in the present set-up, depending on their own position. But also this idea seems politically out of reach.

Some slight modification of the SGP is proposed by defining the deficit norm as the structural and not in terms of the actual deficit in order to avoid hampering the automatic stabilisers (e.g. Buiter, Corsetti and Roubini 1993, Eichengreen 1996). On the other hand, the criticism related to the functioning of the stabilisers in Europe seems generally exaggerated, firstly with regard to the rather wide margins of manoeuvre granted by the 3 percent criterion, and secondly in account of the probably not very stringent – if at all – implementation of the EDP. Hence, “if countries stick to the medium-term objective of keeping their budget in balance, the Stability and Growth pact offers substantial room for automatic stabilizers to function.” (Eijffinger and de Haan 2000, p.86) Besides that, the proposal of reframing the deficit definition leaves fundamental problems of the SGP, such as regards definitions and implementation, unresolved. Similarly, Willet 1999 demands looser targets but stricter enforcement than the Maastricht criteria.

Several authors strike a different tone by pointing out the much higher priority that should be assigned to national reforms, to which the SGP and any supplementary structures are only secondary supranational complements with the character of a contingency procedure. Among these, Bayoumi, Eichengreen and von Hagen (1997, p.84/85) write in the context of EMU enlargement that “an alternative to the numerical guidelines and politicised procedures of the EDP would (…) be to encourage countries seeking to qualify for monetary union to reform their fiscal procedures and institutions (…)” This, they argue, could happen along the lines of creating “independent agencies at the national level to monitor the budget and prevent spending ministers and legislative coalitions from engaging in creative budgeting. Still more drastic reform would establish in each country a national Debt Board with the power to set a binding ceiling on the annual increase in public debt.” (ibid., p.84) Such a proposal amounts to nothing less but advancing an agency solution, a step which in terms of institutional reform is probably even more far-reaching than the implementation of the previously discussed plans for fiscal federalism. As for that, it is thereby also even more unrealistic as long as the present framework does not run into a serious crisis – which some nevertheless see building up through the pensions problem associated with ageing societies in Europe. Additionally, an agency solution, which could as well be conceived at the supranational level in the form of EU Debt Boards, runs into problems of its own in terms of accountability and democratic legitimacy. But a good part of this vision’s intellectual attractiveness seems to be the analogy it would create for fiscal policy in parallel to monetary policy, which is evidently run by the agency par excellence and does not seriously suffer from a lack of legitimacy. It would also offer the possibility of designing optimal contracts for the agents, compensation being linked to the adherence of fiscal policy to sound deficit targets. Nevertheless, for the time being it appears to be bound to remain just that: a vision.

The preceding paragraphs are not comprehensive but try to summarise the most important arguments of the debate about how to improve the institutional framework of fiscal and monetary policy in Europe. We have not yet touched upon the most likely development, which sees the Eurogroup as the locus of at least initially informal coordination of macroeconomic and particularly fiscal policies, embedded in a number of loose supervisory dialogues (the Luxembourg, Cologne and Cardiff processes) and the broad economic guidelines. Before we discuss this evolution and argue that, together with the SGP, it looks like the best feasible way among the options that realistically exist, it might be worth taking a brief look at why economic reform despite considerable expected improvements might be impossible due to the political process inherent in modern democratic societies.

4.2.2 Transaction-cost Limits to Improvement?

In discussing the proposals for improvement of Europe’s macroeconomic management, we often ran into the problem of “political infeasibility”. What does this exactly mean, and can something be done about it? A very insightful political-economy approach to answering that question has been pursued by Dixit 1996. A brief summary of his transaction-cost perspective on economic policy may make informal coordination and gradual improvement centred on the Eurogroup appear as the only realistic way of upgrading the present system.

Dixit suggests seeing markets and governments both as imperfect systems that mutually influence each other. The standard model of the political process from this perspective is a dynamic game of multiple principals and a common agency, the results being invariably of a second best nature, due to – above all – informational problems. Hence economic policy in particular is a dynamic game with changing and uncertain conditions, the rules being written as the game proceeds. It can hence be described as an ongoing, necessarily imperfect and incomplete process with powerful but slow dynamics. The process itself is structured by ‘constitutions’, i.e. contracts that are necessarily incomplete because of imperfect foresight, asymmetric information and the complexity of the issues they deal with. A constitution in the economic realm, such as the Maastricht Treaty and the SGP, needs to be particularly general in order to be able to adapt to future developments that cannot be foreseen when the contract is designed. The double function of such a constitution is to set rules in a continuum between specificity and generality and, secondly, to lay down dispute-settlement procedures that enables the polity come up with new rules and thereby to deal with unforeseen situations which are not covered by the already existing rules. Dixit stresses the importance of hysteresis and path dependency implied in the design of constitutions, especially when they involve the setting up of an agency.

The baseline of Dixit’s approach is the idea to apply the economic concept of transaction costs to politics. He uses Williamson’s definition (1989, p.142) of transaction costs as the “comparative costs of planning, adapting, and monitoring task completion under alternative governance structures.” Based on this conceptual framework, transaction cost politics aims to analyse and assess the evolution of governance configurations, especially institutions, agencies, contractual agreements etc., as a way of how societies cope with precisely these costs. Formal models of transaction cost analysis in politics are derived from time-inconsistency considerations, which applies in particular to the fields of fiscal and monetary policy. In his discussion with a Director of the Belgian National Bank¹³, the author found confirmation for exactly the idea that a more accommodating position of the ECB vis-à-vis budgetary imbalances is impeded by time-inconsistency in the sense that the fiscal authorities cannot credibly commitment to improve their positions as the ECB support will distort their incentives away from consolidation. Similarly, the temporary hardships of structural reform could in principle be alleviated by monetary leniency up to a point and over a limited period of time. However, as soon as the central bank gives its support, the incentive to carry the reform through will be corrupted, as it seems less costly to benefit from the brief upturn caused by monetary expansion and do nothing else.

North 1990 argues that politics is even more subject to transaction costs than the economy and hence yet less efficient. This can be traced back to the general absence of markets and competition from the political sphere, although one could argue that the phenomenon of globalisation increasingly introduces a global market for foreign direct investment into politics and thereby creates competition also between countries. Dixit 1996 provides some proposals of how to deal with transaction costs in politics, based on the rationale that economising on them is Pareto superior for all parties involved, despite potential (or probable) divergence of interests, informational asymmetries, different degrees of freedom etc. Due to the fact that transaction costs are more pervasive in politics and at the same time external enforcement mechanisms less available and less effective, the way of dealing with the problem in the political realm is different from an economic approach. Fundamental is the role of institutions and institutionalised processes for amelioration, since they are best suited to offer credible enforcement mechanisms. Commitment, which needs to be credible, can be achieved or at least aimed at by locking-in actions, delegation, and reputation.

Taking the three in turn, we find that the concept of locking-in actions goes back to Thomas Schelling’s seminal work (1960), where he advised “Make it true!” (p.24) as the best way to create a credible commitment – i.e. by taking an action in advance that makes the choice one wants to commit to ex-post optimal and thereby achieves time consistency. Such actions can involve a deliberate limitation of one’s own future freedom of action through contracts, or, as Persson, Persson and Svensson 1987 propose for the example of fiscal policy, a structure of the public debt that maintains zero net nominal balance with the private sector and thereby eliminates the ex post temptation to create surprise inflation. Secondly, delegation is a specific form of locking-in in the sense of reducing one’s future degrees of freedom by invariably assigning executive power to someone who does not suffer from the same temptation or opportunism because of a different preference set or a fixed mandate – the standard example of delegation in monetary policy being an independent central bank. Finally, the role of reputation and its attainment through repetition has been underlined by its significance for achieving cooperative outcomes in prisoners’-dilemma games. These cases need uncertainty about the last period of the game and are helped by a long time horizon, a low discount factor (vulgo ‘patience’), noise reduction and an increased cost of deviation. Signals, of course, are very important and can be crucial in establishing reputation – even or especially when they come at some cost. For example, the current wait-and-see attitude of the ECB, which at the time of writing persistently refuses to lower the interest rates, can be seen as a move to establish reputation for independence and hard-nosed stability preferences.

Commitment devices in the practice of economic policy are mainly to be found in the form of (regulatory) agencies, an independent central bank and rule-based mechanisms for fiscal policy. However, as many agencies face multiple tasks and multiple principals (‘common agency’) and operate in an environment void of competition and poor in transparency, there is plenty of scope for corrupting the incentives of the relevant agents. Rule-based mechanisms, for their part, are often weak and open to manipulation. This leaves us with a relatively accurate description of the present situation in the interacting fields of monetary and fiscal policy. It underlines the need to improve the commitment credibility of fiscal authorities, enhance the de facto independence of the central bank and establish effectiveness for fiscal rules. However, transaction-cost politics also teaches us that change will be slow and incremental and inevitably lead, if at all, to a second best.

Our assessment of the institutional future of monetary and fiscal policy in Europe is that, at least for some years, feeble but increasing coordination will be based on relatively informal proceedings centred on the Eurogroup of EMU Finance and Economics Ministers. Only incrementally will we see more structured concertation with concrete parameters and measurable results, maybe eventually leading to a more formalised institutional solution. The last part of this thesis is going to look at actual evidence in support of this anticipated scenario.

4.2.3 An Evolving Informal Structure

Thygesen, among many other commentators, is prepared to assign a crucial role to the ECOFIN council in the further evolution of coordinative structures of economic policy (1999, p.32): “Some visible role for ECOFIN (…) in the coordination of non-monetary policies may tend to protect rather than endanger the independence of the ECB.” Indeed, informal concertation leading to increasingly effective coordination at the level of ECOFIN and in particular among the ministers of the Eurogroup, seems to be the development that is, above all, more realistic than the ones we have discussed before and, secondly, actually observable on the political scene.

From the perspective of policy-makers¹⁴, great importance is attached to the Eurogroup as a collegial form of coordination and to ECOFIN as the base of several coordinating processes that aim to bring all relevant actors in line with the goals of stability and growth. The need for a more pro-active stance of the Eurogroup, especially on fields of structural reform, is clearly recognised, but decisionmakers anticipate another two to three years until this can be achieved and based on a more formalised structure. Among the more imminent improvements, as far as the coordinating processes are concerned, will be a standardisation of the stability programmes that member states have to present as part of the budget-review procedure. These have so far been lacking in significance due to diverging assumptions and incoherent timing. There seems to be a consensus that the general format and quality of information underlying the decisionmaking has to be improved in order for the decisions themselves to become more concrete. The Commission formulates in a recent communication: “Economic policy coordination within the euro area is based on consensus. (…) It comprises (…) a common assessment of the economic situation; agreement on appropriate economic policy responses; acceptance of peer pressure and, where necessary, adjustment of policies being pursued.” (European Commission, 2001, p.2) The Cologne Process (“Macroeconomic Dialogue”) is also expected to improve the matching of fiscal and monetary policy in the wider context of macroeconomic management by including instruments and actors of the labour markets as well as the ECB and the treasuries. Finally, the application of the Cardiff Process (a benchmarking procedure aimed at identifying best practices) to fiscal policy might result in strong support of reforms that are conducive to fiscal discipline, such as stronger prerogatives for the treasuries or the implementation of concrete planning horizons.

Of course, review processes and peer pressure are not among the most effective policy tools one could think of. The rebuke Ireland received for its mildly procyclical stance is at most annoying, and even less so would it be for more powerful countries. On the other hand, if the evolving informal structure leads to more preventive information and coordination, governments will at least try to avoid annoying situations like that. We cannot expect any direct effects from the present arrangements, but that is not to say that they might not have a sizeable impact in the medium run. This should indeed enhance the scope and quality of fiscal policy coordination in a framework of monetary leadership and thereby further the common goals of stability and growth – while stopping short of a gouvernement économique above both fiscal and monetary policy, which would destroy the benefits of central bank independence and multiply all previously described problems of time inconsistency and transaction costs.

5 Conclusion

This Master Thesis has tried to construct a logical argument in favour of more formal coordination of fiscal policies in the eurozone. The premises of the argument have been the interaction of monetary and fiscal policy and the suboptimal nature of the present institutional framework, largely defined by the SGP provisions. The validity of the first premise is beyond doubt and has been discussed from an economic and a game-theoretical perspective. The second one is somewhat more controversial, but the author hopes to have conclusively made his case from the angles of economic concepts, game theory, and political economy.

We have described, analysed and evaluated the workings of new institutions and new procedures, aware of the fact that they have to develop their functioning and right now find themselves in a difficult phase of reputation-building and learning. It goes without saying that the success of this endeavour and thereby of EMU as a whole would be greatly helped by a sustained period of robust economic growth to ease the inevitable difficulties of transition from one arrangement to the next. We have the chance of entering a virtuous circle of smooth transition and higher growth – but the danger of a resurfacing and exaggerated Keynesian paradigm for economic policy has to be seen with great concern in this respect. We have implicitly tried to argue against it. Claims for monetary expansion instead of structural reform make it that much harder for the ECB to implement stability in a non-conflictual environment and will be potentially very costly.

Which way forward for EMU? It is to hope that policymakers and academics alike are understanding the need for and the direction of ongoing reform and institutional improvement. The implications of monetary union make the supranational framework for fiscal policy appear suboptimal. Additionally, the ageing of our societies makes most present national arrangements of fiscal policy virtually redundant as they stand. We can either face that truth and address the looming twin-crisis with decision – or be overrun by it in the not too distant future. This seems to be the answer to one of the most important questions of economic policy of our times.

6 References

Abbildung in dieser Leseprobe nicht enthalten

ANNEX

Annex I

A Taylor Rule for Europe? Summary of an Empirical Estimation of a Taylor Rule based on the AWM Data Set

1. Introduction and Theoretical Remarks

This section shall address the empirical question of the relevance of Taylor rules for the eurozone over the last three decades. The setting is somewhat paradoxical as the eurozone was non-existent prior to 1999, but – given the importance of convergence in the project of Economic and Monetary Union – it will be interesting to see whether a Taylor rule holds at all or at least for a sub-period of that time frame for the area.

John Taylor in his seminal paper (Taylor 1993) proposed a simple reaction function for the interest-rate instrument of an independent central bank, based on the observed deviations from an output goal (given by potential output) and from an inflation goal defined by the bank. The Taylor rule can be stated as

Abbildung in dieser Leseprobe nicht enthalten

where D it denotes the change in the nominal interest rate at time t, a and b are weights for differentials in the output gap (between potential output ytpot and actual output ytact) and in the inflation gap (between desired inflation p t goal and actual inflation p t act).

The following estimations are going to look at the question whether a traditional Taylor rule as defined in (1) applies for the countries in the present euro-zone over the period of 1973 to 1998. It should be rather surprising if this were the case as the present policy-area of the ECB was then still fragmented into separate monetary jurisdictions over which the Bundesbank only gradually and partially acquired an indirect grasp. Therefore, estimations of a Taylor function for the entire euro-area over the whole period can be expected to suffer from misspecification problems, omitted variables or gradual central bank responses and hence should exhibit serial correlation and heteroscedasticity. Given the data set, we will be unable to reduce the group of countries assessed, however the time period of the EMS (1980 to 1998) and in particular the era of the “hard EMS” (1987 to 1992) might yield a better fit for a Taylor-like policy function, whereas the effects of the 1992/3 turbulances might have to be excluded by dummy variables.

2. The Econometric Approach

The discussion is based on an adaptation of the Taylor rule for empirical purposes. The model is

Abbildung in dieser Leseprobe nicht enthalten

where STN is the short-term interest rate, INF inflation (i.e. the change in the private consumption deflator) and OGAP the output gap. This is translated into error correction form as follows:

Abbildung in dieser Leseprobe nicht enthalten

where DSTN is the change in the short term interest rate, DINF is the change in inflation and DOGAP the change in the output gap. During the econometric procedures, INF and correspondingly DINF are going to be subdivided into DINF1 and DINF4, which corresponds to a one-year lag derived in the first case from 4*(logPCDt – logPCDt-1) and in the second case from (logPCDt – logPCDt-4) in order to arrive at annual lags based on the quarterly data. The method of estimation is going to be OLS.

3. The AWM Data Set

The database used for this project is identical to the one underlying the ECB’s area-wide model AWM¹⁵. It is quarterly with series from 1973q1 to 1998q2 and provides seasonally adjusted euro-zone figures as an aggregate of country data. Aggregation is achieved by an index of the eleven member countries with individual weighting factors based on constant GDP at market prices (PPP) for 1995.

4. Results

The variables are defined in Sub-Annex I.1. Modelling changes in the interest rate based on the change in inflation and the output gap with four or three lags gives results that are partially significant but suffer, as expected, from serial correlation and heteroscedasticity (Sub-Annex I.2 to I.5). This problem is not removed by introducing a dummy variable to neutralise the effects of financial turbulences in the years 1992 to 1993 with four or three lags (Sub-Annex I.6 to I.7).

Starting the observations in 1985 and still excluding 1992/3 covers the period of strong and increasing Bundesbank dominance. The results are no longer subject to heteroscedasticity but still serially correlated (Sub-Annex I.8). However, the period of the so-called “hard EMS” from 1987 to 1992 was characterised by virtually all member countries “shadowing” the Bundesbank, that is following the Bundesbank’s interest rate movements. Restricting the model to this time-frame removes the problems of serial correlation and heteroscedasticity (Sub-Annex I.9) but does not deliver significant parameters. But we now can go on to test the significance of each variable on its own and see whether removing lags improves the quality of the model.

Unfortunately, none of the variables but the constant is individually significant, and removing lags does not help but even reintroduces serial correlation (Sub-Annex I.10). Sub-Annex I.11 delivers a graphical representation of the estimated model with all lags and the real-world changes in the interest rate, which fits reasonably well until towards the end. The last periods of the observation (1990q4 to 1991q4) showed a relatively constant interest rate but estimates strong changes for the model, which probably already shows the break-down of the Taylor function, if it ever held for the eurozone has a whole. This view is confirmed when using the Taylor rule as a forecasting instrument: It would have been totally useless already in 1991 (Sub-Annex I.12).

5. Conclusion

We have tried to answer the question whether a Taylor rule could be a reasonably accurate rule-of-thumb to describe interest rate changes in what has now become the eurozone prior to the introduction of the common currency. The underlying rationale of this argument was that by following the Bundesbank, Europe already had something like a common monetary policy rule. However, the model suffers from serial correlation and heteroscedasticity unless one restricts the observations to the period of the hard EMS. And even then does it fail to produce significant variables. The answer to the intial question and hence the conclusion of this short paper is that a traditional Taylor rule does not apply to interest changes in the area that has recently been merged into a monetary union prior to 1998.

This result is interesting with regard to the fact that the authors of the ECB’s area-wide model rely on a Taylor rule as a substitute for the fiscal block of their model.

6. References

Clarida, Richard, J. Galí and M. Gertler. 1998. “Monetary policy rules in practice: Some international evidence”. European Economic Review, 42, pp. 1033-67

Fagan, Gabriel, J. Henry and R. Mestre. 2001. “An area-wide model (AWM) for the euro area”. ECB Working Paper, 42

Gerlach, Stefan and G. Schnabel. 1999. „The Taylor rule and interest rates in the EMU area: A note”. BIS Working Papers, 73

Hendry, David F. and J.A. Doornik. 1996. Empirical Modelling Using PcGive for Windows. London: International Thomson Business Press

Rudebusch, Glenn and L.E.O. Svensson. 1998. „Policy rules for inflation targeting“. Sveriges Riksbank Working Paper, 49

Taylor, John. 1993. “Discretion versus policy rules in practise”. Carnegie-Rochester Conference Series on Public Policy, 39, pp. 195-214

7. Sub-Annexes

Annex I.1: Definition of Variables

Abbildung in dieser Leseprobe nicht enthalten

Annex I.2: Modelling DSTN on DINF4 and DOGAP, four lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.3: Modelling DSTN on DINF4 and DOGAP, three lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.4: Modelling DSTN on DINF1 and DOGAP, four lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.5: Modelling DSTN on DINF1 and DOGAP, three lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.6: Introducing a dummy to remove 1992q1 to 1993q4, four lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.7: Introducing a dummy to remove 1992q1 to 1993q4, three lags

Abbildung in dieser Leseprobe nicht enthalten

Annex I.8: Time period 1985q1 to 1998q2, excluding 1992q1 to 1993q4

Abbildung in dieser Leseprobe nicht enthalten

Annex I.9: Time period 1987q1 to 1991q4 / Hard EMS

Abbildung in dieser Leseprobe nicht enthalten

Annex I.10: Significance Tests and Removing Lags

Analysis of lag structure

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each variable

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each lag

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of all lags up to 4

Abbildung in dieser Leseprobe nicht enthalten

Analysis of lag structure

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each variable

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each lag

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of all lags up to 3

Abbildung in dieser Leseprobe nicht enthalten

Analysis of lag structure

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each variable

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of each lag

Abbildung in dieser Leseprobe nicht enthalten

Tests on the significance of all lags up to 2

Abbildung in dieser Leseprobe nicht enthalten

Annex I.11: Graphical Representation

Abbildung in dieser Leseprobe nicht enthalten

Annex I.12: Forecast

Abbildung in dieser Leseprobe nicht enthalten

Analysis of 1-step forecasts

Abbildung in dieser Leseprobe nicht enthalten

Tests of parameter constancy over: 1990 (4) to 1991 (4)

Abbildung in dieser Leseprobe nicht enthalten

Annex II

The impact of interest payments on actual deficits - the case of Germany

Abbildung in dieser Leseprobe nicht enthalten

Annex III

A model with politically induced deficit biases (adapted from Beetsma 1999)

Consider country i (i = 1, …, n) taking part in a monetary union. Its expected utility is given by

Abbildung in dieser Leseprobe nicht enthalten

where f and g are spending on public goods F and G in periods (t = 1,2) and p is the inflation rate, determined in the second period. f represents the inverse of the degree of inflation aversion; E(.) denotes the expectations operator. The political process in society i is modelled along the lines of Alesina and Tabellini (1990), where two political parties F and G are associated exclusively with the corresponding one of the two public goods. They are selected to run the government by an election with random outcome and have the utilities

Abbildung in dieser Leseprobe nicht enthalten

Assume that in each country, party F is in power in period 1. It will be re-elected at the end of the period with an exogenous probability p, where 0 < p < 1. The exogenous endowment income for each government is one unit in each period. The initial debt is zero and all debt is paid off in the second period. The budget constraints for the two periods then are as follows:

Abbildung in dieser Leseprobe nicht enthalten

bi is the amount of debt issued in period 1; e i represents a shock to first-period resources, capturing all kinds of unexpected effects such as additional needs for finance etc. The fines of the stylised stability pact are given by y 1 (bi - ^bi) in period 1 and ψ 2 (bi - ^bi) in period 2, where ^bi represents the reference level of the debt (which, in a two period model, is equal to the deficit). Assume that the fines paid by country i are equally distributed among the remaining n – 1 countries, which is captured by the last term in (4) and (5). The fines payable can be indexed to the size of the shocks according to

Abbildung in dieser Leseprobe nicht enthalten

where δ > 0 and hence a negative shock increases the reference level and reduces the fine for country i.

The ECB conducts monetary policy and thereby selects the inflation rate in period 2. In the absence of political pressure, it would maximise – p 2 / 2 f. However, if the ECB is not completely independent, it will attach the relative weight λ (0 ≤ λ ≤ 1) to its own objective and simultaneously takes into account an equally weighted average of the objectives of the governments in period 2, multiplied by 1 – λ. The ECB’s actual objective function is then given by

Abbildung in dieser Leseprobe nicht enthalten

where α ≡ φ(1 – λ) and °b denotes the cross-country average °b ≡ (1/n)Σnj=1bj. If 0 < λ < 1 or 0 < α < φ, the ECB is only partially independent but still more conservative than the average union (along the lines of Rogoff 1985).

Signing the SGP involves setting the values for ^b, δ, and ψ. The sequence continues with shock εi hitting each country i. Then the governments issue debt and pay fines in period 1. The ECB determines the inflation rate, debts are repaid and period 2 fines occur at the end of the sequence. The model is solvable through backward induction by assuming that the ECB takes inflation expectations as given (since it cannot credibly commit to its announcements) and maximises equation (7) over π, yielding

Abbildung in dieser Leseprobe nicht enthalten

which means that inflation rises in proportion to average debt °b. The rationale behind this relationship is that higher debt increases the incentive to reduce its real value through higher inflation once inflation expectations are set. The equilibrium solution anticipates this through higher nominal interest rates.

All period 2 resources are spent on the public good associated with the party in power. Setting the debt determines the inflation expectations, so that πe = π still holds when the governments are choosing their debt levels. Hence, period 1 governments maximise

Abbildung in dieser Leseprobe nicht enthalten

where f1i and f2i are given by the right-hand sides of (4) and (5). The FOC for each government is obtained as

Abbildung in dieser Leseprobe nicht enthalten

As f1i is the same for all i, we can state

Abbildung in dieser Leseprobe nicht enthalten

The average union debt level is then determined by

Abbildung in dieser Leseprobe nicht enthalten

Individual countries’ debt levels are given by

Abbildung in dieser Leseprobe nicht enthalten

This result shows that the model yields higher debt levels for countries in a monetary union since autonomous countries have ψ1 = ψ2 = 0 and n=1, which leaves °γ = 1, whereas it is greater than one when the country joins.

Monetary unification, this is the important insight, can lead to higher debt accumulation because the cost in terms of future inflation of unilaterally increasing the deficit becomes smaller. However, sanctions can be structured so as to neutralise this increase in the deficit bias.

Annex IV

Stackelberg Leadership

In a sequential game with known reaction functions, player 1 as Stackelberg leader can pick his preferred outcome on player 2’s reaction function. This equilibrium S lies on a lower isprofit line for player 1 than the Cournot equilibrium C that would have been obtained in a simultaneous game.

Applied to monetary-fiscal interaction, we would say that the institutional framework has to be desinged in such a way that the central bank can act as a Stackelberg leader vis-à-vis the fiscal authorities

Abbildung in dieser Leseprobe nicht enthalten

[...]

---

¹ In fact, the interaction of fiscal and monetary policy in the ECB’s model is reduced to a Taylor rule. This is a very doubtful approach as the author found it virtually impossible to estimate a meaningful Taylor rule on the AWM data set. (see Appendix I for a summary of the estimation)

² See Annex II for a graphical representation of this development.

³ Discussion with Peter Praet, Director of the Belgian National Bank, Brussels, 03 April 2001

⁴ in Dixit and Lambertini (2000a, 2000b, 2000c) and Dixit (2000)

⁵ If fiscal expansions lead to higher interest rates, the spillover effects can very well be negative – as the EU experienced in the wake of German reunification.

⁶ The Stackelberg diagram can be found in Annex IV.

⁷ Lamfallussy 1989 provides a comprehensive overview of this argument.

⁸ The risk premium on Italian ten-year government bonds above German Bunds has fallen below 30 base points without any reduction of the debt ratio.

⁹ See also Giavazzi and Pagano (1990, 1995) or Vines et al. (1997)

¹⁰ see Appendix IV

¹¹ from Thygesen 1999

¹² The SGP still looks quite effective in comparison to US balanced budget rules (see Eijffinger and de Haan 2000, pp. 89/92).

¹³ Interview with Director Peter Praet, 03.04.2001, Brussels

¹⁴ Discussions with Didier Reynders, Belgian Minister of Finance, Brussels, 30 January 2001 and 24 April 2001, as well as with Koenraad van Loo, Advisor to the Ministry of Finance, Brussels, 24 April 2001

¹⁵ Fagan et al. (2001)