The costs of inclusion of a country in EMU have been used as the main determinant of choice of participating countries. The implicit choice of excluding Italy, was based on the mostly German analysis that Italy inside would imply excessive high costs because of its public debt and budget deficit situation. Until recently, the costs of excluding a EU country from EMU have not been very much discussed. This is because the analysis of the relation between the ins and the outs seems to have been conducted as if the creation of a partial EMU would have no impact on the monetary policy of excluded countries.
The episode of the competitive devaluations of the Lira and the Peseta has demonstrated that countries can benefit from such a policy at a surprisingly low inflation cost. One can thus interpret the 1992 devaluations of the Lira and of the Peseta as a small preview of what EMU members have to expect Their impact may only be short term but it is a powerful one. Through EMU, its members will forbid themselves such destabilising monetary strategies. The excluded countries will neither have the means nor the interest of such stability.
The incentive for competitive devaluations for outside countries will in effect increase due to the creation of large zone characterised by a stable monetary policy. Hence, excluded countries will optimally exploit this free-rider position given that they are offered no alternative (1). Any other policy would anyway not be credible in the eyes of the markets which will perfectly understand these new incentives. From that perspective, there is not much difference in interpretation to do between a depreciation chosen as a deliberate monetary strategy and one imposed by the markets.
The economic impact for EMU members will be significant as the more active monetary and exchange rate policies of the excluded countries will tend to destabilise the output of EMU members. The political consequences of exclusion may be more dramatic as specific sectors hit by competitive devaluations will naturally ask for protection and ultimately for a restriction of Single Market rules to EMU members. Again, the reaction of French industrialists at the Lira devaluation was a good preview of the political pressures to expect. Flexible exchange rates may be economically feasible among countries of the Single Market. It is not sure it will be so from a political point of view. Also, the optimal monetary reaction to the policies of excluded countries will not be the same in Germany and in France for example. The exclusion of certain EU countries from EMU will thus not only generate conflicts between these countries and EMU members but also among EMU countries.
Different solutions to the difficult relations to expect between the ins and the outs of EMU have been discussed.
An EMS type of agreement between the Euro and the other EU currencies has been proposed. The advantage of such an agreement for the EMU countries is obvious: if it works, it enables them to avoid the problem of competitive devaluations but does not oblige them to share monetary policy with countries which carry a fiscal risk. Can it work? A simple replicate of the old EMS will not do. It was not viable even for currencies such as the French Franc. It will not stand the market pressure for currencies such as the Lira which in addition will be handicapped by the suspicion linked to the free rider problem that goes with the creation of a partial EMU.
An agreement for the transition period is necessary but it needs to be much stronger than the present one. In particular, it should involve commitments from both parts. One of the reasons of the failure of the EMS has been the discretionary nature of interventions of the Bundesbank. In a new EMS, these interventions should, as has been proposed by Charles Wyplosz and Jean Pisani-Ferry (2), be part of a contract between the ECB and the non EMU countries: the ECB should intervene to defend the exchange rate in case of speculative tensions; in exchange the non EMU countries should engage themselves in fiscal convergence.
No need to say that these commitments will be tested by the markets especially around the times when the contract will be reviewed by the partners. This seems however the only way to go in order to make the relations between the ins and the outs less bumpy. Some additional commitments may help. For example, the outside countries which want to enter should quite rapidly be obliged to convert their new emissions of public debt in euros. This would make the future entry to EMU of outside countries and their commitment to stable exchange rates more credible in the meantime as a competitive devaluation would automatically imply an increase of the value of the debt.
Another answer to the question of the ins and the outs is implicitly provided by the Maastricht treaty and it is that the problem should only be temporary at least for countries which, like Italy, have expressed their desire to enter as soon as possible. The problem is that the convergence criteria are such that they could transform a temporary exclusion into a permanent one may be even chosen by the excluded countries if the free-rider gains are large enough.
Because of the creation of a partial EMU, the differential in credibility, inflation and interest rates will increase between the ins and the outs. These differentials are at the origin and will perpetuate the decision to exclude the later. They could then be trapped in a vicious circle which would make it difficult, may be impossible, or even non desirable to enter EMU later on. The penalty in terms of interest rates that will be imposed on those countries left out will make it also more difficult for them to converge fiscally. Imposing the same convergence criteria for all countries as an entry condition to EMU is dangerous because it may transform a temporary exclusion into a permanent one.
There is a good argument to be made that once a partial EMU is created, the entry conditions for those left out should be altered and made less strict. Adding some flexibility into the Maastricht criteria will be politically difficult to sell especially in Germany but the case should be made that the creation of a partial EMU constitutes a radical change in regime that makes the existing criteria obsolete and dangerous.
Because the creation of a partial EMU (or its mere expectation) will automatically increase the differential of interest rates between the ins and the outs, the budget deficit computed for the convergence criteria should be calculated with a debt service that takes into account the interest rates among EMU and not those of the outside countries.
This should not be viewed as a way to artificially make the fiscal criteria more lenient for some countries. The aim should be to avoid penalising the outside countries. As they stand, the convergence criteria are in effect less strict for the countries that will come in first than for those countries left out. The objective to make the relations between the ins and the outs as harmonious as possible and the aim to avoid the scenario of a permanent exclusion are linked but the second one deserves special attention.
Both require a strengthening of cooperation on monetary policies and exchange rate policies between the ECB and those left out. The contract for conditional interventions between the ECB and the outs should not only specify the conditions of the convergence program but also be precise on the timetable of integration, consistent with the convergence program, so as to let no doubt, to markets as well as to governments, that the exclusion of countries left out is indeed only temporary.