Macro Policy
European Interdependence

Increased economic interdependence, together with the unsatisfactory economic performance of the 1970s, has led to a growing debate on the prospects for international economic policy coordination. The ESRC and the Centre National de la Recherche Scientifique have jointly sponsored a CEPR research programme on Economic Interdependence and Macroeconomic Policy in Europe. A November 28-29 workshop organized by CEPR was the first UK meeting of the French and British economists involved in this research, and brought together not only academic researchers but also participants from the Bank of England and Banque de France.

The European Monetary System (EMS) has been seen by many observers as the precursor of a European monetary union with a common currency. In the first paper of the conference, entitled 'Pros and Cons of a Unique European Currency', Daniel Cohen (INSEE) presented a general framework in which the potential gains and losses involved in the movement to such a European currency union could be analyzed.

There were disadvantages to a monetary union, Cohen argued. Countries would lose the ability to pursue independent monetary policies and in addition would be unable to print legal tender, which created a risk of insolvency. Governments had little role to play in the conventional 'Arrow-Debreu' models studied by economists, where all markets were assumed to exist and to function competitively. But under more realistic assumptions it was desirable for governments to play a more substantial role, reacting to unforeseen events. Optimum fiscal policy required them to be able to print money and raise government debt through monetary action, and a monetary union would make this difficult.

Cohen argued that the benefits of a common currency were threefold. There was first a gain in efficiency due to the elimination of the exchange rate system itself (whether fixed or flexible) among the participating countries. Under fixed rates the inefficiencies arose because individual countries were obliged to adjust domestic interest rates to attract foreign currency; under flexible rates inefficiency resulted when a number of countries attempted simultaneously to appreciate or depreciate, in order to combat inflation or increase competitiveness. The second benefit to a common currency lay in the renewed effectiveness of fiscal policy, which would no longer be constrained by exchange rate considerations. Thirdly, the common currency would increase European bargaining power in international negotiations. There would be gains from seigniorage and, if the common currency established itself as a reserve currency, then Europe would have the option of printing its own currency if it needed to borrow.

Although Cohen's paper was seen as a useful framework for analysis, there was much discussion of the specific gains he had identified. Would it be possible for European bargaining power to increase as a result of the union, at the same time as each country made greater use of fiscal policy? In order for European bargaining power to increase, Europe would have to adopt a unified fiscal stance, but this would limit the independence of fiscal policy for individual countries. In addition, it was noted that currency union would produce a policy 'game' with two large players - the US and Europe. It was not clear that the outcome would be preferable to the 'Stackelberg' equilibrium that currently arose from the leadership of the one large player (the US) which dominated the smaller European players.

Subsidies to investment are widespread throughout Europe, as well as elsewhere; they are usually a response to local or regional unemployment. Philippe Michel (University of Paris), in a paper entitled 'Investment and Keynesian Unemployment', examined the effects of an investment subsidy in a model characterized by disequilibrium in the labour market. His model postulated a single representative consumer and a single producer. The government, unable to act directly on the labour market, attempted to restore it to equilibrium through investment subsidies, financed by income taxes. Michel's results showed that investment subsidies would only restore the economy to full employment if the labour market was initially quite close to its equilibrium position. Otherwise, the subsidies would have to be accompanied by an incomes policy to achieve full employment.

The third paper of the conference, presented by Paul Levine (LBS), was entitled 'International Coordination of Monetary and Fiscal Policy' and represented joint research with CEPR Programme Director David Currie (QMC). Currie and Levine studied policy coordination in a two-country world. There was a three-way interaction among the two governments and the private sector, which was in turn composed of groups of domestic wage-setters and an international player, agents participating in the capital market. Their analysis dealt only with government policies which were 'time-inconsistent'. This was a reasonable restriction, they argued, and could arise out of a 'noncooperative' game in which the governments acted as a (Stackelberg) leader with respect to the private sector. Governments could exercise this leadership because the private sector agents were small relative to the government. The decisions of a single private sector agent therefore have no perceptible effect on government policies. The government by contrast can perceive and act on the private sector's reaction to its policies and is therefore able to take strategic leadership.

The two countries' governments can choose to cooperate and jointly select optimal policies. In the absence of such cooperation, each country will act independently and optimize its own objective function; the result will be a non-cooperative equilibrium.

Currie and Levine examined two types of non-cooperative equilibria, 'open-loop Nash' and 'closed-loop Nash'. In the open-loop Nash equilibrium each country chooses its policy given its expectations of the paths of the other country's policy instruments. In the closed-loop Nash equilibrium it is the expected policy rule of the other country which is taken as given. Currie and Levine found that when governments pursue time-inconsistent optimal policies, the closed-loop Nash equilibrium cannot exist, and the open-loop Nash equilibrium is the only feasible non-cooperative outcome.
Even if the two governments can devise coordinated policies, it is also essential to determine whether such cooperation can be sustained. Currie and Levine argue that a useful framework for this problem is the concept of a supergame, in which the policy 'game' described above, whose outcome is either the open-loop Nash equilibrium or a cooperative agreement, is played repeatedly over time. A strategy for this supergame then involves a choice of policy for an individual game based on the strategies chosen by all players in previous games.

Currie and Levine found that two elements are necessary for a sustainable cooperative solution: that each government honour the cooperative agreement provided the other government has done so in the past; and that if either government deviates from the cooperative policy, the other government chooses the non- cooperative open-loop Nash policy. The 'threat' of non- cooperation will be an effective deterrent only if the non- cooperative equilibrium results in a higher welfare loss for both countries compared to the cooperative outcome, but there is no guarantee that this will be true. Even if global welfare is always lower in the non-cooperative equilibrium, one country may still benefit by reneging if there are asymmetries between the two countries' economic structures or between the shocks which affect them.

The reactions of the private sector must also be taken into account in assessing the sustainability of policy coordination: will a government's policies remain credible with the private sector, if that government has already reneged on an earlier time-inconsistent cooperative policy? The authors argued that such a switch in policy will be credible provided that the time- inconsistent cooperative policy is clearly seen to be conditional upon the international agreement continuing.

The afternoon session began with a round-table discussion of computing software. Garry MacDonald (Warwick) outlined the software available at Warwick University and suggested that it could be used as a central source of programmes to be made available over the UK academic computing network. It was decided that in the short term, programmes should be made available on magnetic tapes, but that efforts should be made soon to transfer this software to personal computers.

In the final paper of the first day, entitled 'Monetary Stabilization Policy in an Open Economy', Marcus Miller (Warwick and CEPR) presented a simplified analysis of how monetary policy could be designed to control output and inflation in an open economy under floating exchange rates. Miller assumed that inflation was influenced by its own past values, possibly through overlapping wage contracts. The openness of the economy meant that inflation was also affected by the exchange rate. The foreign exchange market was assumed to be forward-looking and to form its expectations rationally, hence current and expected monetary policy influenced the exchange rate and therefore inflation.

Miller noted that forward-looking expectations in the exchange market introduced the problem of time inconsistency into the model: there was a systematic temptation for policy makers to announce a sequence of current and future policies from which they would later be tempted to depart. He compared two policy rules: a time-consistent policy from which the government would (by design) experience no subsequent temptation to depart; and a policy which takes the exchange rate as exogenously fixed, in effect ignoring the effects of monetary policy on the exchange rate. Miller noted that in this model, when high real interest rates are used to combat inflation, the exchange rate tends to be overvalued when inflation is high. Paradoxically, however, the time-consistent policy recognises this and chooses to underrespond to inflation in its choice of interest rates. Why should this occur? Because the policy is required to be time- consistent, there must be no temptation to depart from it at some future date. But the perception that future policy makers will use high interest rates reduces the incentive to increase them now, since such a rise in the exchange rate reduces the inflation problem and the incentive for higher interest rates in the future!

Miller also found that the policy of turning a blind eye to exchange rate effects generates a more appropriate, activist response to inflation. Indeed for the initial set of parameters he considered, this 'Nelsonian' policy resulted in the optimal linear feedback policy, although it was not in general true that optimal policy could ignore the effects of policy on forward- looking variables.

There have been many calls for national governments to coordinate their economic policies but few estimates of the benefits which might arise from cooperation. Andrew Hughes Hallett (University of Newcastle and CEPR) opened the second day with his paper 'How Much Could the International Coordination of Economic Policies Achieve? An Example from US-EEC Policy Making'. This is now available as CEPR Discussion Paper No. 77 and is summarized elsewhere in this Bulletin.

In his analysis Hughes Hallett used an empirical multi-country model that was similar to those employed by policy makers, but it differed from previous work in two important respects. Firstly, it examined the interdependence of two policy blocs with asymmetric economic structures, so that the potential gains from exploiting 'comparative policy advantages' could be explored. Secondly, dynamics played a more important role. Cooperation made it possible to improve the timing of policy actions. The results illustrated the inefficiency of uncooperative strategies. Cooperation restored the effectiveness of domestic policy responses weakened by interdependence and cut intervention costs by speeding up these responses.

The second paper of the day was presented by Shaziye Gazioglu (Birkbeck College) and represented research done jointly with Michael Artis (Manchester University and CEPR). Their paper, entitled 'Currency Substitution: A Simulation Approach', described the simulation results of a two-country, two-asset model containing thirty equations. The long-run model exhibited steady inflation in the two countries, while nominal exchange rates adjusted with the inflation differential to maintain a constant real exchange rate. Currency substitution could affect this long-run rate of inflation, since a portfolio shift from domestic currency to foreign currency raised the domestic level of inflation by reducing the base for the domestic inflation tax (and would conversely lower foreign inflation by increasing the foreign inflation tax base). The authors simulated the estimated model while varying its parameters so as to produce different degrees of currency substitution. This process revealed interesting short-run dynamics reminiscent of the Dornbusch 'overshooting' model, and Gazioglu and Artis were able to identify channels through which currency substitution could lead to shocks in the real exchange rate.

In the first paper of the afternoon, entitled 'The Prospect of a Depreciating Dollar and Possible Tension Inside the EMS', Jacques Melitz (INSEE and CEPR) analyzed the potential for policy conflict between two EMS members. He modelled the EMS as containing only the Deutschmark and the Franc, assuming that the respective countries have different policy preferences: Germany is taken to have a lower tolerance for inflation than France. Melitz also assumed that, due to inertia or poor forecasting ability, the two countries cannot neutralize the dollar depreciation without altering the Franc/Deutschmark rate. This results in policy conflict between the two countries, and Melitz found that one of them could be better off by leaving the EMS and pursuing a non-cooperative strategy. The EMS would be able to achieve the same results as this non-cooperative outcome but there would be no advantage to membership. Political rigidities would be likely to lead to the collapse of the system.

Melitz's analysis led naturally into the discussion by Michael Artis of the November report by the Treasury and Civil Service Select Committee on the consequences of 'UK entry into the EMS'. Before the report had been released, opinion had moved in favour of UK entry into the exchange rate mechanism of the EMS, but the Select Committee recommended against entry. Artis outlined three reasons for the Committee's decision. The first was the bipolarity argument: the EMS would contain two strong international currencies, the Deutschmark and Sterling. They are both 'reserve' currencies in open capital markets, and large flows into and out of these currencies would make it difficult to maintain parities without exchange controls. Moreover, Sterling and the DM are very likely to react differently to oil-price shocks.

Secondly, Artis noted the sovereignty argument. The EMS is currently seen as essentially a 'Deutschmark zone'. This, combined with the problems of bi-polarity, could lead to conflict over policy formulation. The UK, it is argued, might be obliged to follow German monetary and fiscal policy in order to preserve EMS parities, and this could lead to tensions in Anglo-German relations.

Thirdly, Artis drew attention to Sterling's exchange rate upon entry into the EMS. Most commentators argued that the Pound was undervalued against the US dollar, but overvalued against the DM and other currencies in general. UK membership of EMS could create tensions if Sterling later moved significantly against other currencies, and it may be desirable for Sterling to remain outside the EMS until it has moved closer to its long-run equilibrium value. Marcus Miller attempted to explore this question further by showing how movements in Sterling, the Dollar and the Deutschmark affected the EMS under differing assumptions about the width of the parity bands maintained by Sterling after UK entry. Artis noted finally that a continuing fall in the Dollar could prompt a quick reassessment of the case for Sterling going into the exchange rate mechanism of the EMS.

The final session of the workshop was a discussion between the academics and two leading central bank economists who were present - John Flemming (Bank of England) and Robert Raymond (Banque de France). After a preliminary outline of the group's research interests Marcus Miller invited comments from Flemming and Raymond concerning the areas of research which would be most useful to policy makers.

John Flemming's response focussed on the radical changes that were taking place in domestic financial markets. He argued that research on how innovations in the domestic market are likely to affect public policy would be especially beneficial. Robert Raymond discussed research in the area of macroeconomic policy coordination. He stressed the distinction between cooperation within Europe and coordination between the major trading blocks, e.g., Europe, the US and Japan. He also pointed out that much of current research concentrated on the effect of monetary policy, and he asked what the effects of fiscal policy and the policy mix would be in cooperative models. Raymond concluded with three remarks concerning research methodology. Firstly, he wondered if a common measure of the stance and efficiency of monetary policy could be defined. Secondly, he stressed that although there exist many useful theoretical models for policy makers, they need to be applied to a period of time that is relevant to policy making. Finally, he asked if it would be possible to define general long-run policy prescriptions in these models, although he conceded that this would be difficult in the face of a lack of agreement on long-term targets.