In the negotiation of the Maastricht Treaty there was little question that there should ultimately be a single currency in the ensuing monetary union and that the single currency should be the ECU. There was discussion about the way such a single currency might come about - through successful competition with existing currencies as suggested by the UK or by administrative replacement - but no real discussion of the means of implementation of such a single currency. The benefits in terms of reduced transactions costs were clear and other benefits suggested in the Commission's assessment One Market, One Money, vol. 44 of European Economy contribute to the case for its introduction. There is no matching discussion of the difficulty of implementation nor of the associated costs. In practice both the difficulty and the costs are substantial - sufficiently substantial that it is unlikely that the ECU could be implemented as the single European currency before the end of the century even if the member states have agreed on the introduction of Stage 3 by then. (These issues are discussed in detail in a report for the Commission by Mark Burridge and David Mayes (Burridge and Mayes, 1992), summarised in `A Single European Currency by the Year 2000?', in De Pecunia; 1993.)
In this paper we discuss the nature of these difficulties
and how they might be overcome. The difficulties are of two sorts, those
which relate to the practical application of monetary policy and those
which relate to the adoption of the single currency by economic agents
in the economy, principally banks and other financial institutions, commercial
firms, house-holds and public sector organisations. In this article we
concentrate primarily on the second group of problems.
Problems for the monetary authorities
Implementing a single currency involves a long series of decisions and subsequent actions, starting with the simple need to establish what the form of the notes and coin are to be. Size and design need to be agreed at an early stage so that firms that have to make physical changes, such as those with cash dispensers, ticket machines, slot meters, etc. can have a long enough period to adjust. Even with major changes within single countries nothing on the scale proposed has been approached, let alone attempted. The US for example has the same notes and coins that it did in the early part of the century. Although electronic machines handling cash may be easier to adjust than their mechanical counterparts, it remains the case that all such machines will have to be adjusted. If all have to be done by the date of implementation of the ECU (we labelled this E-day in Burridge and Mayes (1992) - the introduction of decimalisation in the UK was labelled D-day at the time) this will be a monumental task. Although there are means of programming machines to make them all change on a particular day for most this is not practicable and a physical changeover is required. This would normally involve the machines being out of service for a while. If such a changeover period took even a week the chaos involved would be considerable.
To some extent the monetary authorities are shooting in the dark as the adoption of Stage 3 will be a political decision. While it now seems unlikely that the earlier of the two dates laid down in the Maastricht treaty will be chosen there are doubts about whether the second date will be altered or even if it is not, which countries will be involved in the initial EMU. This could include some or all of the existing member states and the current applicants, Austria, Finland, Norway and Sweden. Even given a fixed date for stage 3, the treaty does not prescribe an exact date for the introduction of the single currency except that it is to happen rapidly. Optimists might feel this could be achieved in six months but it could take much longer. Running stage 3 with multiple currencies but fixed exchange rates may prove very difficult in the face of speculative pressures. The rapid introduction of the ECU may therefore prove to be necessary just from a practical point of view. In that case the plans for its introduction have to be highly advanced by the time that Stage 3 starts, i.e. the nature of the currency must have been known for several years, the processes for change clear to companies, the public information programme ready to run and so on.
Much of this is not very likely unless the date of introduction of the ECU becomes far more of a certainty. It is not just that politicians have to agree a date, it is that that date has to be widely believed by the public at large. Changing over to the ECU is a costly procedure for many businesses. They will therefore tend to postpone action if they are uncertain, particularly because the costs are incurred first and the benefits flow only later (and not necessarily to the people who incurred the costs - a purely domestic enterprise like the Metro will have to change all its coin machines and ticket printing machines but it gains almost no benefit from the fact that the new currency circulates over a much wider area than the old, it has not normally had to trade in foreign currencies.)
Most public debate has been conducted as if the changeover to the ECU would occur on a single day. There is of course no reason why the ECU should not be introduced progressively, taking some of the smaller member states first so that those who follow can learn from their experience. This lead in period could take 18 months without reducing the pressure on the authorities. The progression could of course occur with the currency, certain coins being replaced first with ECU equivalents. This, however, has a number of drawbacks as the new coins would need to have very close equivalent values if hoarding is not to take place. Where currencies are close to a simple alignment, such as 2 DM to the ECU then the process is possible but in other cases the two systems would be very much in parallel and trading between them would be required.
In any case further thought needs to be expended on the order in which the coins are minted. If entirely new metal were needed it would represent a sizeable proportion of the world's annual nickel production. A more sensible approach is to retire the larger coins (in terms of metal content) first and produce more of the equivalent nominal value in ECU by using smaller coins. Indeed since the demand for 1 and 2 ECU coins will tend to be substantial the sensible approach may be for ECU notes to be used in the short run so that the large weight of existing coins can be recycled and ECU coins introduced after reminting. Although notes may last for much shorter periods of time they are much cheaper to make compared with their face value. (The typical banknote in New Zealand costs six cents to print and values nearer 2½ cents per note have been quoted for the longer print runs in the US.) Hence using them for a limited period to ease the changeover may make good economic sense even though it may at first glance appear retrograde because small value notes have been phased out in recent years.
Although the monetary authorities have responsibility
for issuing the currency, in one sense they face the least problems because
these actions form part of their normal job. They are not an external task
thrust upon them. Clearly they will have further problems in their own
accounting and in their transactions with the commercial banks, but these
highly sophisticated computer systems will be rather more straight forward
to change than smaller scale more mechanical arrangements. Clearing systems
will have some problems with converting existing balances but the change
in operation from the last domestic currency day to the first ECU day will
be largely a matter of multiplication as all the member states have decimal
currencies. They will also have problems with managing public debt, which
will almost all be incorrectly denominated (however, the UK and Italy for
example have had some ECU denominated issues for many years).
Problems for banks
The commercial banks and other financial institutions will be among the organisations most affected as they will have to undertake the issuing of the new currency to customers and the collection of the old. All their accounts will need converting as will the denomination of assets. Of course they face all the same costs of adjusting their own in house payment, invoicing and accounting systems as does any other company. The banks will have a particularly difficult time round E-day itself as they will have to make all the conversions in as short a period of closure as possible. A week has been quoted but something nearer a long weekend would be desirable, although there will be an inevitable string of further action to be taken after E-day. It will take some time convert everybody and the old currency will take some time to flow bank into the banks.
One of the real problems for banks would occur if the two systems have to work in parallel for a while. There was no great problem in the case of decimalisation as everyone could make the conversion relatively easily in their heads as it was only the last few pence where such computation was required. In the case of the ECU it will be much more complex as there will be few simple relations between ECU and the pre-existing currencies.
Setting up parallel systems in the banks would be computerised.
The real difficulty would fall on the ordinary person who will have problems
both in conversion and in establishing a new set of values. The changeover
in the banks will entail both long hours round E-day and considerable numbers
of extra staff to handle the physical aspects of the changeover in the
Problems for firms and public sector organisations
As we have indicated there are three main areas of effect on organisations.
The first concerns their accounting, billing and payment systems. These will all need conversion but since most are computer based and the problems common we can expect that a wide range of services will be offered. It is the larger companies with tailor-made systems and the smaller companies, especially those which do not have computer systems, where the greatest difficulties will occur and we can expect that special arrangements will be made to help small business.
The second area is pricing. For companies other than retailers this is a relatively simple exercise as it will be possible to put both domestic and ECU prices on all invoices for a while both before and after E-day to ensure that the transaction is fair and the currency change has not altered the cost. For retailers prices will need to be posted on all products in both currencies for a period either side of the changeover. Providing turnover is rapid this relabelling can probably be accomplished just by the marking of new stock, rather than relabelling existing items on the shelves or in warehouses.
Lastly and most tedious is the conversion of currency
handling machines, which range from cash dispensers to turnstiles, slot
machines and cash registers. The problem for most cash registers will be
trivial as one decimal system is being replaced by another and no physical
conversion is required. Of course if the two currencies are run in parallel
a problem arises, which will be solved by using to sets of registers or
conversion charts. The costs are potentially high for the owners of slot
machines. However, there are ways round this. Many machines already operate
with tokens rather than currency. There will be no shortage of the old
currency and since it is unlikely to trade at a discount no problem for
either operators or users getting access to it. The old coinage could be
purchased in the establishment where the machines are. (If as is expected
the new currency is issued in advance of it becoming legal tender through
souvenir sets and educational packs, it might also be possible to supply
it to owners of slot machines to encourage them to convert early and to
enable new machines handling ECU coins to be sold in the run up to E-day
- otherwise the market for the machines could dry up for a while.)
Problems for households
While decimalisation proved relatively easy to cope with, as have rebasings of other currencies, the change over to the ECU presents a difficulty as many of the conversions will be difficult to do in the head. One might question why the conversions have to be undertaken at all as it is likely to be a matter of a step shift from one currency to the other round E-day, unless substantial numbers of firms are physically unable to change quickly.
Publicity and information are the keys to helping households.
This can be aided by a clear pack of material including durable conversion
charts that can be carried round. Despite the high sounding costs of this
sort of activity it will be small compared with the currency costs and
costs to banks and firms. The individual will also have problems with accounts,
including tax returns, as the change is bound to straddle a tax year. (Previous
changes have usually been made at quiet periods, avoiding year ends, ends
of accounting periods, etc. A date in say February when most people are
likely to be around might be suitable for example.)
A timetable for change
It is difficult to say how long the process of change will take. There is a temptation to look at the introduction of the DM in the new Lander and to argue that quite rapid progress would be possible - even though that itself was not without its problems. However, it is a long step from implementing an existing currency in a small part of Europe to agreeing a complex system which will replace all the currencies in the European Union, with co-ordinating problems among a variety of authorities.
If we look at other previous examples of currency change such as decimalisation in the UK or the introduction of the dollar in New Zealand we can get an idea of the complexities involved in even these simple instances. Once the decision had been made in New Zealand it took from May 1964 to July 1967 to organise the change. (The NZ dollar replaced the NZ pound, both currencies independent of others with the same units, at the rate 2$=1£.) It was only necessary to replace the notes, and to replace the smallest coins: half, one and three pence. The half crown was also withdrawn but this was rather more because it seemed an unnecessary coin. It was `replaced' by a 50 cent piece only slightly larger and more complex in milling but twice the nominal value. Thus in currency terms the changeover was rather easier in that most of coinage remained in place and only needed to be changed slowly by coins of identical dimension but with the new value system stamped on them. When the UK decimalised it actually introduced the decimal 50 pence ahead of the change, further assisting the ease of the changeover.
In some respects the change from pounds, shillings and pence to dollars and cents is more complex than the change to the ECU in that the arithmetic system also had to change. However, having said that the value system did not change 1/- became 10c, 1/6 became 15c, 10/- a dollar, not a great step forward. The great difficulty came in converting cash registers, accounting and other machines including computer systems. The physical job took eighteen months to complete despite bringing in new machinery and using a factory in Australia as well.
As we have noted the amount of physical conversion as opposed to electronic reprogramming is now much smaller despite the vastly increased volume of transactions as a result of the demise of most mechanical and electro-mechanical machines. In any case switching from one decimal currency to another does not entail any reprogramming except perhaps in cases where the unit of measurement is stated. By contrast the number of computer systems to be converted has increased out of all proportion. The time for the changeover has to include the development of conversion software as well as the identification of the systems to be changed and the organisation of their changeover.
In Burridge and Mayes (1992) we suggested that a period of six years preparation might be needed for an orderly changeover. In other words it would be touch and go if one started today whether the process could be completed before the year 2000. Providing a lot of the ground work has been done beforehand then a shorter time period once the definite dates are known would be possible, perhaps just three years. The nature of the publicity, the new coins and notes, the steps to be taken in ensuring the conversion of machinery and the systems of banks and firms can be substantially developed even without a specific date. However, until the change has a fixed date and is clearly going to happen most firms will avoid giving it any serious attention.
When we conducted our survey of firms and financial institutions in 1991 to try to assess the costs that might be involved in the process of implementation we found that most had given it almost no thought and those which had studied the subject had only done so in a relatively sweeping manner. It is only now, three years after the event, that the UK commercial banking system is beginning to question the estimates (and come up with rather larger costs). It will be longer before commercial firms address the issue in any significant numbers. As was very clearly established in our extensive interviews for A Strategy for the ECU (Ernst and Young - NIESR, 1990) firms are very reluctant to devote management time to such issues until they have reasonably firm dates within the normal planning horizon. Even with the 1992 programme it took several years of strong persuasive effort before the bulk of firms began to face the issues and even then action was frequently put off. Given the popular uncertainty about EMU, it will be difficult to persuade them to address the problems of implementing the single currency for some time.
The main initial progress on needs and costs in the meantime
will have to be made by working groups of knowledgeable individuals drawn
from large firms and representative organisations. We were surprised in
our survey for Burridge and Mayes (1992) that even the main computer firms
and accountancy and other organisations had not really addressed the subject
although they might have expected to benefit commercially from the change.
Establishing a currency board
We have therefore recommended strongly that the EU should set up a currency board or similar organisation representing all the affected parties: central banks, finance ministries, commercial banks and other financial institutions, small and large firms and the ordinary householder and customer. I gather that a body is being set up by the European Commission at present to study the whole range of technical issues concerning implementation, however, unlike the currency boards it will not have any powers to act. Given the uncertainties which still exist this is probably the practical way forward in the short run, although clearly there needs to be thorough co-operation with the European Monetary Institute and the member states if the outputs from this group are to be regarded as definitive.
If one wants to get an idea of the complexity of what
is involved, the New Zealand currency board had to meet 67 times during
its life of approximately four years.
I have left what is many respects the most important question till last. How much is this changeover going to cost and who is going to pay? I have not seen this addressed in any of the official documentation. The costs are non-trivial. An issue of new banknotes is a relatively cheap exercise, first because the life expectancy of banknotes, especially the smaller denomination ones is short. Secondly, it has been the custom in several of the member states to change the notes at reasonably frequent intervals. In part this has been a reduction in size, to an extent reflecting the inroads of inflation, and partly an attempt to stay ahead of counterfeiters. The additional costs of the note issue would normally therefore be borne in effect by the tax payer through an increase in central bank costs. Coins are a different matter because they have a long expected lifetime and the vast majority would be retired early. Since they can be recovered and melted down the loss is not the full value of the new minting. However, since we are discussing the minting of billions of new coins, this is a non-trivial exercise.
Clearly there will be a publicity campaign but beyond general information to firms and households one would probably expect that extra training in new systems and familiarity will be provided through the private sector. The cost of this would then be borne either by the customer of the service or by the shareholders of the companies concerned through a reduction in profits.
The costs to firms will vary markedly depending upon the alterations required. Where it is merely accounting, invoicing, payment systems and pricing, expecting the firm to incur the costs might be reasonable but where substantial alterations are required to machinery then this could have a serious impact on the competitiveness of the business. When the decimalisation changes took place South Africa, New Zealand and Australia took a different view from the UK. The UK did not compensate. Now in New Zealand even when banknotes change a small compensation payment is made by the central bank to the commercial banks towards the costs of converted all the automatic cash dispensers. This amounted to $250,000 out of a total cost of $9,400,000 in issuing a completely new set of banknotes in 1992.
It is not quite clear why compensation should stop at the banks as all note recognition machines are affected. In 1967 with the changeover to the NZ dollar, the compensation depended on the age of the machine to be converted, the newest being converted at government expense, the owners of older ones receiving a grant towards the cost and the oldest machines (over 15 years) attracted nothing.
In our enquiries into the costs of changing the various computer systems in commercial and financial companies in 1991 we received a wide range of estimates, some of which, particularly from the banks, were substantial. Applying a somewhat heroic grossing up to the level of the economy as a whole we estimated that the cost would lie between a quarter and a half of 1% of EU GDP - a major sum, but nevertheless one considerably smaller than the expected benefits, especially if we take the dynamic effects discussed in One Market, One Money into account. However, the costs accrue first and there will be many instances where those who incur the costs will not be major beneficiaries. Clearly one of the important first steps for the new technical study group will be to develop a more accurate assessment of the likely costs and their incidence so that a clear view of their size and equity can be established.
Burridge M and Mayes D G (1992) The Implications for Firms and Industry of the Adoption of the ECU as the Single Currency of the EC, report to the European Commission.
Burridge M and Mayes D G (1993) A Single Currency for Europe by the Year 2000? De Pecunia.
Ernst and Young, National Institute of Economic and Social Research (1990) A Strategy for the ECU, Kogan Page, for the Association for the Monetary Union of Europe.
I am grateful to Brian Lang for providing information on costs.