IMF WORKING PAPER

c 1990 International Monetary Fund

This is a working paper and the author would welcome anycomments on the present text. Citations should refer to anunpublished manuscript, mentioning the author and thedate of issuance by the International Monetary Fund Theviews expressed are those of the author and do not neces-sarily represent those of the Fund.

WP/90/24 INTERNATIONAL MONETARY FUND

Research Department

Fiscal Policy Independence in a European Monetary Union

Prepared by Paul Masson and Jacques Melitz *

March 1990

Abstract

The issue of whether constraints should be placed on f i s c a l policieswhen moving to European monetary union is examined in the context of the useof f i s c a l policy for macroeconomic stabilization purposes. Examples ofshocks hitting French and German economies are considered: an appreciationof their joint exchange rate against other currencies, an inflation shock,and an o i l price increase. Except in the third case, flexible use of f i s c a lpolicies in the two countries is li k e l y to give better outcomes than asystem with constraints on their use. For the o i l price shock, there seemsto be a good case for policy coordination, not for ceilings on f i s c a ldeficits.

JEL Classification Numbers:3 2 1 , 431, 4 3 2

* Jacques Melitz is in the Research Department of INSEE (Paris),Professor at HEC and a CEPR fellow. The latter would like to thank the FordFoundation and the Alfred P. Sloan Foundation for support. This paper wasdone in large part while he was a consultant for the Fund. The authors aregrateful to Lans Bovenberg, B i l l Branson, Hugh Bredenkamp, Matt Canzoneri,Michael Deppler, Morris Goldstein, and Klaus Riechel for helpful discussion.

- ii -

Contents Page

I. Introduction 1

II . The Simulation Analysis 41. Long-run simulations: Joint current account balance

relative to the rest of the world 42. Medium-run simulation 1: French resistance to a current

account d e f i c i t resulting from an appreciation of the ecu 83. Medium-run simulation 2: German resistance to i n f l a t i o n

resulting from monetary policy 114. Medium-run simulation 3: Contractionary responses by both

France and Germany to an o i l shock 14

I I I . Some Theoretical Analysis of Current Accounts in the EMU 14

IV. Concluding Remarks 17

Text Tables

1. France and The Federal Republic of Germany in a EuropeanMonetary Union: Long-Run Scenarios 5

2. French Fiscal Targeting of Current Balance in Response toAppreciation 9-10

3. German Fiscal Targeting of Zero Inflation in Response toEuropean Money Growth Shocks 12-13

References 20

I. Introduction

Further progress toward monetary union in Europe is likely to involveloss of monetary independence on the part of member countries; this paperconsiders the issue of possible limits on fiscal policy independence aswell. There is a concern that fiscal policy independence might bringconflicts in the European Monetary Union (EMU), leading to proposals fordeficit ceilings, for policy coordination, and for a community-wide systemof taxes and transfers. Would individual governments borrow too much on theassumption that their debt would be monetized by a European central bank?Would member governments independently attempt to reduce the relative pricesof their goods through contractionary fiscal policy in order to gain compet-itiveness; or alternatively, would they use expansionary fiscal policy toimprove their terms of trade? The fear of looser fiscal policy in the EMUhas led the authors of the Delors Report (1989) to advocate a statutoryceiling on deficit spending. But deficit ceilings would not even addressthe problem of the competitive use of contractionary fisc a l policy. Thevarying dangers have induced some academic opinion to support fiscal policycoordination (see Portes and Vinals, eds., (1989) and van der Ploeg (1989)).In a recent paper, Sachs and Sala-i-Martin (1989) go further, and argue thata joint fiscal policy may be required for the survival of an EMU, since inthe absence of a major Community-wide mechanism of transfers a member hit byan adverse country-specific blow might be tempted to move out of a monetaryunion and devalue. Social transfers to afflicted regions through nationalfiscal systems exist in many countries on a scale that far exceeds thecurrent level of such transfers in the Community.

One way to treat the previous questions is to consider optimal macro-economic management in the context of monetary union. As a rule, fixedexchange rate regimes are successful in handling shocks hitting a l l of themembers similarly as they avoid opposing efforts to move the exchange rate.Asymmetric shocks, on the other hand, tend to require f l e x i b i l i t y of realexchange rates. Cohen and Wyplosz (1989) bring out another relevant aspect:the duration of the shocks. In the face of stickiness of wages and prices,fixed-rate and common-currency systems do not easily achieve adjustment totransitory shocks. Thus, a combination of temporary and asymmetric shocks,as opposed to permanent and general shocks, would be the most d i f f i c u l t foran EMU to handle without the active use of other policy instruments.

Our approach to these questions in this paper w i l l be to examine theconsequences of an independent use of government spending to influenceaggregate demand. We focus on this case not because demand management isthe only--or even primary--objective of fiscal policy, but because i t is inthis context that conflicts may arise when countries are faced with tempo-rary, asymmetric shocks. We w i l l not consider the possibility of debtmonetization, nor the question of policies with respect to taxes andtransfers. 1/ Our procedure w i l l be to do some simulation analysis ofalternative fiscal policy responses to various shocks. This w i l l enable us

1/ On these matters, see Padoa-Schioppa (1990).

- 2 -

to take the current situation as a point of departure, to study some casesof interest, and to base our results on standard models with plausibleempirical values of the fundamental parameters. But the drawback w i l l be ananalysis rooted in certain scenarios, therefore lacking generality. Weshall need to keep this limitation in mind in discussing the results.

Our focus w i l l be on France and the Federal Republic of Germany, whichw i l l thus represent for us the EMU. We shall simulate the French and Germaneconomies together using the IMF's multi-regional model, MULTIMOD, 1/while treating the rest of the world (ROW) as exogenous except for the ROWcurrent account, which w i l l perforce be the negative of the joint French andGerman one. Implicitly, we may be said to consider France and Germanycombined, therefore, as a small open economy. In a l l of the simulations,the growth rate of money w i l l be set for France and Germany jointly (i.e.,for the EMU), and the two countries w i l l be supposed to have a commonnominal interest rate and a common exchange rate relative to the ROW. 2/

Our f i r s t set of simulations assumes a sufficient appreciation of thecurrency of the EMU to drive down its joint current account balance, whichis now positive, to zero. This w i l l enable us to show the extent to whichthe capital flows between France and Germany (which can only be expected tobecome more elastic in a monetary union) w i l l be conditioned by the d i s t r i -bution of the adjustment of home absorption between the two countries. Thedistribution of the adjustment can in principle vary from a l l or nothing byeither country depending on the mutual settings of their f i s c a l policyinstruments. This f i r s t set of simulations w i l l a l l concern the long run.

Our remaining simulations w i l l focus on the medium-term adjustment fromyear to year over five years. In order to highlight the possibility of con-f l i c t , we w i l l assume different preference functions on the part of Frenchand German authorities. In particular, we shall suppose that the Germansplace more emphasis on price level stability, while the French emphasizecurrent account balance, in the short run. A few things should be said hereabout this essential assumption.

Of course, national preferences could evolve in the context of amonetary union, and indeed we w i l l examine in a later section (III) whethercurrent account targets s t i l l retain a role in this context. However, notonly is i t reasonable to suppose some persistence of different French andGerman preferences at the outset of the union, but there are really nogrounds to think that French and German objectives must eventually convergein the union. The fact that both countries must accept a common long-runrate of inflation does not remove the need for adjustments to the terms of

1/ See Paul Masson, Steven Symansky, Richard Haas, and Michael Dooley(1988).2/ We assume that the equations of the model would not be affected by the

move to monetary union, even though we recognize that there are good reasonsto expect a change in some relationships, for instance, the demand formoney, trade equations, and wage/price f l e x i b i l i t y .

- 3 -

trade between them, and this requires short-run deviations from price levelstability for one or the other. Therefore, differences in their ranking ofinflation relative to other objectives w i l l remain relevant.

Since 1987, the French rate of inflation has dropped to the Germanlevel, around 3 percent. This equivalent price performance is the fruit ofan intense effort in France to gain monetary discipline and a solid reputa-tion for i t . Nevertheless, i f we take the 1950-88 period as a whole, theaverage French rate of inflation, measured by the change in the GNP defla-tor, was 7.1 percent, as compared with 4.0 percent in Germany. If we lookat the experience of France and Germany regarding current account balance,we find an equally marked contrast, and this is true whether we focus on thelast few years or a longer period. In the years since 1986, the currentaccount was virtually balanced for France, but in large surplus for Germany,attaining levels of around 4 percent of GNP. Over the longer 1963-88period, the average imbalance was -0.4 percent of GNP for France +1.2 per-cent of GNP for Germany. As regards the absolute values of these imbalanceswithout regard to sign for 1963-88, the corresponding values are 0.8 percentfor France and 1.5 for Germany. Thus, the idea that France put relativelymore emphasis on current account balance and Germany on price level stabil-ity looks very plausible as a description of historical experience.

The f i r s t of our three medium-term scenarios, based on the previoushypothesis about French and German priorities, supposes a 10 percent appre-ciation of the EMU currency which gives rise to a French attempt to usecontractionary fi s c a l policy in order to maintain current account balance.In the second medium-term scenario, the joint monetary authority of the EMUallows a growth rate of money one percent higher than i t was previously (asa weighted average in France and Germany), thereby prompting the Germans toadopt tighter f i s c a l policy to bring inflation down to earlier levels. Ourthird and last scenario starts off with a doubling of the price of o i l :France fights the current-account implications while Germany opposes theinflationary ones.

In Section III, as already mentioned, we take a closer look at the ideathat France would continue to place some emphasis on current accountbalance. Our hypothesis that current-account objectives would be retainedby some countries in the EMU is a f a i r l y common one. It has been used byothers to show that f i s c a l policy independence may result in externalitiesthat c a l l for cooperation (see Cohen and Wyplosz (1989) and Krugman (1989)).Yet in a monetary union, the significance of current account imbalancesclearly alters. Such imbalances can no longer yield exchange rate changesor balance-of-payments crises. They do, nonetheless, continue to affect theinternational distribution of wealth and the geographical location of eco-nomic activity. We conclude, based on our general examination of the issue,that there are reasons for believing that differences In country preferenceswithin an EMU may persist, and that f l e x i b i l i t y of fi s c a l policy would helpin achieving those different objectives. A final section offers someremarks on a different issue: namely, the need for a system of taxes andtransfers at the community level to make an EMU viable.

- 4 -

II. The Simulation Analysis

In doing the simulations with MULTIMOD, we use as a reference scenario,or baseline, the projections of the IMF's World Economic Outlook, extrapo-lated, where necessary, further into the future (IMF (1989)). The startingyear, taken to be 1990, has a large current account surplus for the FederalRepublic of Germany (about 4 percent of GNP), and a small deficit for France(about 0.2 percent of GNP). If monetary union were to begin in that year,and at current nominal exchange rates between the French franc and thedeutsche mark (and between both of them and third currencies), then i t islikely that the current account imbalances would persist into the mediumterm, given stickiness of inflation rates, which are about equal in the twocountries. Indeed, monetary union would probably yield larger currentaccount deficits for France, at least i n i t i a l l y , by lowering French interestrates to the German level, since the elimination of the expectation or therisk of a devaluation associated with the franc would equalize nominalinterest rates in the two countries (presumably at a level close to theprevailing one in Germany). French short-term interest rates are currently(December 1989) about 10 percent, while German rates are a l i t t l e over8 percent. The adverse current-account effects of the reduction of Frenchinterest rates would come from the demand side, and should prevail at f i r s t ;but in the long run the beneficial supply-side effects of a higher Frenchcapital stock would be expected to prevail, or at least to do so on thetrade balance.

The presence of monetary union means that EMU monetary policy would seta rate of growth of the joint money stock of France and Germany. Theseparate demands for money in the two countries would simply determine thedistribution of the increase in the money stock between them, while the sumof these two demands would decide the countries' common nominal interestrate. It is assumed that these changes from present circumstances, togetherwith the perfect f i x i t y of the franc/mark exchange rate, would be viewed asirrevocable by the market. We do not necessarily suppose a new common unitof account, and therefore do not presume that the benefits of a commoncurrency would be fully reaped. In any event, these benefits, involvingincreased ease of calculation and lower transactions costs, cannot easily bereflected in the model. Nevertheless, where convenient, we shall refer tothe ecu as an alternative to saying "either francs or marks." The commonvariables in France and Germany are shown in bold type in the tablescontaining the simulations so as to set them out.

1. Long-run simulations: Joint current account balance relative tothe rest of the world

The current account surplus of the monetary union vis-a-vis the rest ofthe world would be roughly 2 percent of GNP on the basis of current projec-tions (Table 1). The current account surplus of the EC as a whole is reallymuch smaller, the German surplus being largely offset by a deficit in theUnited Kingdom, and to a lesser extent, in Spain and Italy. For i l l u s t r a -tive purposes, we look at only France and Germany, however. We suppose thata reduction of the U.S. current account deficit would involve larger U.S.

- 5 -

Table 1. France and The Federal Republic of Germany in aEuropean Monetary Union: Long-Run Scenarios

Baseline EMU Current Balance Constrained to Zero v i a ChangesValue in in Exchange Rates, and by Absorption Changes in:

1990 Germany only France only Shared equally

1/ As a percent of GNP.2/ France and Germany combined, as a percent of their combined GNP.3/ Deviations from baseline, in percent.4/ A positive number indicates an appreciation of both the franc and the deutsche mark

a g a i n s t t h e U.S. d o l l a r .5/ The reason that the French deficit is lower than the German surplus as a ratio to

GNP, yet the EMU current balance is zero, is that the value of nominal GNP in Franceincreases relative to that in Germany.

Current account balances 1/

FranceGermanyEMU 2/

Real domestic absorption 3/

FranceGermany

Output price 3/

FranceGermany

D o l l a r exchange rate 3/

EMU 4/

-0.23.92.1

-- 11.3

-4.69.4

9.9

-3.9 5/4.1 5/

12.3--

22.6-6.7

12.4

-2.11.8

6.16.1

8.51.6

11.7

- 6 -

surpluses in the future relative to the EMU, and thus would imply an equaldollar depreciation relative to both the franc and the deutsche mark.

Let us assume that the dollar exchange rate of the ecu adjusts toachieve a zero current account balance for the union (with the ecu movinguniformly relative to a l l third currencies, not only to the dollar) accom-panied by changes in absorption in France and Germany, as discussed below.The growth rate of the joint money stock in France and Germany follows thebaseline trend and is assumed to be unaffected. Therefore, the nominalappreciation also leads to real appreciation (but not necessarily one forone because of endogenous responses in German and French output prices).However, the current account balance between France and Germany need not bezero. Indeed, as a basic consequence of monetary union, capital flows wouldeasily finance the excess of spending over income in any region. This easeof finance is reflected in our simulations by the common interest rate inthe union, implying that the flow of capital from the surplus into the defi-cit area is not hampered by any rise in the interest rate in the deficitarea relative to the rest of the union. The responsiveness of capital flowsto shocks is correspondingly higher. 1/

Nonetheless, the extent of these capital flows w i l l depend on thedistribution of the rise in absorption that is needed to eliminate the2 percent joint surplus vis-a-vis the outside world and to bring about jointbalance in the French and Germany current account. This distribution, inturn, w i l l depend not only on the appreciation of the ecu (which is alwayssufficiently high in our simulations to balance the current account), butalso on the separate fiscal stances of France and Germany. By setting theirfiscal policies sufficiently far apart from one another, France and Germanycan always ensure that the rise in absorption takes place entirely on oneside of the Rhine or the other. For the moment we are not interested in thefiscal policy considerations themselves, but the size of the capital flowsinside the monetary union.

Table 1 shows us these flows in three separate cases: f u l l adjustmentof the required rise in absorption by Germany (with absorption unchanged inFrance), f u l l adjustment by France (with absorption unchanged in Germany),and equally shared adjustment. In a l l three cases, every coefficient in theFrench and Germany models is set at its cumulative value over time in orderto eliminate distributed lags and the whole adjustment process. This istherefore a long-run simulation, but only in a limited sense, since we donot impose the satisfaction of a l l long-run equilibrium conditions, whichwould include, notably, the stabilization of the net capital flows betweenthe two regions of the union based upon life-cycle considerations anddifferentials in growth of factors and productivity. Thus, the only long-run condition which is necessarily satisfied in these simulations is f u l ladjustment to current account balance between the two countries and the ROW.

1/ For an insightful theoretical and empirical discussion of capitalflows in a monetary union (partly inspired by James Ingram's earlier work onPuerto Rico (1962)), see Barry Eichengreen (1989).

- 7 -

The case of f u l l adjustment by Germany is the one leading toapproximate balance in the separate French and German current accountsindividually, since the German current account is the only one of the two tobe markedly out of balance i n i t i a l l y . This particular scenario requires an11 percent rise in absorption by Germany. It also requires (independentlyof baseline growth of the EMU money stock) an appreciation of the EMUcurrency against the dollar of about 10 percent. In this scenario, further,German prices rise by nearly 10 percent. But in France, where some fiscalcontraction is implicit (in order to explain the lack of any rise in thecurrent account deficit despite the appreciation of the franc), prices f a l lby nearly 5 percent.

The case of f u l l adjustment by France is the alternative that achievesoutside adjustment entirely at the expense of greater internal imbalance.This alternative (implicitly requiring a combination of fiscal ease inFrance and fiscal stringency in Germany) necessitates a 12 percent appre-ciation of the ecu. The German surplus stays the same (approximately) whileFrance acquires a deficit larger than 4 percent of GNP. The widening of theGerman advantage in competiveness relative to France is enormous: nearly30 percent (22.6 + 6.7). This case is clearly unrealistic, and has perverseconsequences.

The final alternative, which assumes that domestic absorption increasesin equal proportions in France and Germany, requires about a 6 percent risein absorption in each country. While the German current account surplusdrops in half from the baseline, this surplus remains substantial and nowcorresponds to a French current account deficit of about 2 percent of Frenchoutput. Prices rise both in Germany and France, but do so by about7 percent more in France than Germany.

Thus, the stakes are high in the distribution of the adjustment betweenthe two countries. If the predominant burden of the adjustment is onGermany, inflation there increases for a time. If the adjustment is f i f t y -f i f t y — a situation which we might associate more or less with the case of ajoint fiscal authority-- the Germans might be happy enough (as the smallinflation in their home output prices w i l l be offset by a large deflation intheir import prices), but the French would experience inflation in consumerprices together with a large current account deficit. The case of exclusiveadjustment on the French side seems least favorable for France. Viewing thef u l l range of possibilities, we can see that the presence of separate fiscalauthorities in the two countries, as opposed to a single one in both ofthem, could lead to a better outcome for both countries, since the two aredifferently situated i n i t i a l l y , and aside from this, disagree about what isbest. Of course, policy coordination in principle could respond to thesedifferences. However, a fully optimal coordinated policy is unlikely toemerge. Instead, a joint EMU fiscal policy might well lead to a convergenceto identical policy settings, even though such convergence is not what isrequired, given the different circumstances and the different preferences inthe two countries. In the next three simulations, we focus on these last

- 8 -

policy disagreements, and consider the dynamic sequence of adjustments overa five-year period.

2. Medium-run simulation 1: French resistance to a current accountdeficit resulting from an appreciation of the ecu

The f i r s t of our medium-term simulations begins with an appreciation ofthe ecu of 10 percent against the dollar (and a l l other currencies) as aresult of pressures in exchange markets that do not depend on France andGermany, but come, for example, from a concern about the U.S. currentaccount deficit. We model this as i f i t were a portfolio shift on the partof international investors. Monetary policy in the union and fiscal poli-cies in France and Germany are assumed not to respond, but to continue alongthe baseline path derived from the World Economic Outlook, The appreciationof the ecu yields a reduction of the German current account surplus and anincrease in the French deficit (raising the latter to more than 2 percent ofGNP by 1995) after the f i r s t year of the J-curve effect. No constraintbeing imposed on the joint current account balance of the two countries, asmall joint surplus continues. We then repeat the simulation of the shockwith an accompanying endogenous response of French government spending ongoods and services in order to resist a current account deficit. In thiscase, as opposed to the preceding one, we treat the ecu as endogenous.Quite specifically, following the i n i t i a l 10 percent appreciation, wesuppose the ecu to move in accord with open interest parity as well asrational exchange rate expectations.

Table 2 compares the two simulations. Hence, i t focuses on the changesresulting from the French unwillingness to accept the emergence of currentaccount deficits. The French fiscal policy feedback rule is based oni n i t i a l experiments which were designed to yield approximate balance in thecurrent account. 1/ As seen from the table, this feedback rule reducesFrance's current account deficit by 1.7 percent relative to GNP by 1995, ascompared with an alternative deficit of about 2.3 percent in that year (notshown).

The decrease in government spending lowers demand in France by morethan 4 percent by 1995 and output (real GNP) by almost one third as much.French output prices also go down by almost 5 percent by 1995 (or onepercent annually). The impact of the French policy actions on Germany is ofspecial interest. The French fiscal contraction tends to increase output inGermany by depreciating the ecu. Fiscal contraction in France leads to adecline in demand for French goods, requiring a real depreciation of thefranc; because of sticky prices, this occurs through a nominal depreciationof the ecu. Paradoxically, interest rates rise at f i r s t because deprecia-tion pushes up German output and prices, and on balance EMU demand for money

1/ The particular rule took the the formAG/GDP = 5 CAB/YGNP

where G and GDP are stated in real terms, CAB (current account balance) andYGNP in nominal terms.

- 9 -

Table 2. French Fiscal Targeting of Current Balance in Response to Appreciation

(deviations from the scenario with no f i s c a l response) 1/

France

National Income

Real GDP (percent)Real GDP (annual growth rate)Real GNP (percent)

Domestic demand (percent)Consumption expenditure (percent)Gross private investment (percent)Exports of goods and services (percent)Imports of goods and services (percent)

Government

Gen. gov't financial balance (FFb)Gen. gov't financial balance/GNP (percent)Government debt (FFb)Government debt/GNP

Interest Rates

Short-term interest rateLong-term interest rate

Prices and Supply

1990

-0.4-0.4-0.4

-0.8-0.8-3.80.2

-1.4

-31.3-0.532.70.6

0.60.6

Absorption deflator (percent) 0.0Absorption deflator (annual i n f l a t i o n rate) 0.0GNP deflator (percent) -0.1Export price deflator (percent) 0.5Import price deflator (percent) 1.0Capacity u t i l i z a t i o n rate -0.4Long-term real interest rate 1.3Real user cost of capital 1.3

International Accounts

Trade balance (FFb)Current account balance (FFb)Net foreign assets ($b)

As a percent of nominal GNPTrade balanceCurrent account balanceNet foreign assets

Nominal effective exchange rate (percent)Real effective exchange rate (percent)

15.515.52.6

0.30.20.2

-1.0-0.6

1991

-0.9-0.6-0.9

-1.9-1.8-4.50.4

-3.3

-7.9-0.143.31.0

0.70.5

-0.1-0.2-0.50.51.7

-0.81.41.5

40.541.79.9

0.60.60.8

-1.5-1.1

1992

-1.5-0.6-1.4

-3.0-2.5-4 .70.8

-5.0

33.10.5

14.60.7

0.70.4

-0.6-0.5-1.10.52.4

-1.31.41.5

66.871.322.7

1.01.01.7

-2.0-1.9

1993

-1.8-0.4-1.7

-3.8-2.9-4 .41.2

-6.3

79.11.1

-58.9-0.0

0.60.2

-1.3-0.8-2.20.13.1

-1.51.21.3

87.597.640.5

1.21.32.8

-2.6-2.8

1994

-1.9-0.1-1.7

-4.4-2.8-3.61.9

-7.2

126.91.7

-179.4-1.4

0.4-0.0

-2.2-1.0-3.4-0.43.8

-1.50.91.0

103.0120.362.7

1.41.54.1

-3.0-3.9

1995

-1.70.2

-1.4

-4.6-2.4-2.42.6

-7.8

172.62.2

-345.4-3.2

0.2-0.2

-3.3-1.1-4.8-1.24.3

-1.20.40.6

114.5140.788.9

1.51.75.5

-3.4-5.1

10 -

Table 2 (concluded). French Fiscal Targeting of Current Balance in Response to Appreciation

(deviations from the scenario with no fiscal response) 1/

1/ Deviations are for the level of the variable r e l a t i v e to i t s value in the correspondingyear for the scenario with no f i s c a l response.

Germany, F.R.

National Income

Real GDP (percent)Real GDP (annual growth rate)Real GNP (percent)

Domestic demand (percent)Consumption expenditure (percent)Gross private investment (percent)Exports of goods and services (percent)Imports of goods and services (percent)

Government

Gen. gov't financial balance (DMb)Gen. gov't financial balance/GNP (percent)Government debt (DMb)Government debt/GNP

Interest Rates

Short-term interest rateLong-term interest rate

Prices and Supply

Absorption deflator (percent)Absorption deflator (annual inflation rate)GNP deflator (percent)Export price deflator (percent)Import price deflator (percent)Capacity utilization rateLong-term real interest rateReal user cost of capital

International Accounts

Trade balance (DMb)Current account balance (DMb)Net foreign assets ($b)

As a percent of nominal GNPTrade balanceCurrent account balanceNet foreign assets

Nominal effective exchange rate (percent)Real effective exchange rate (percent)

EMU Nominal exchange rate (percent in $/ECU)

1990

0.40.40.4

0.20.11.10.1

-0.4

1.10.1

-1.7-0.2

0.60.6

0.50.50.20.41.20.40.1

-0.2

-1.7-1.3-2.0

-0.1-0. 1-0.0

-1.1-0.8

-1.3

1991

0.50.10.5

0.30.21.10.2

-0.4

-0.9

-0.0

-1 .5

-0.3

0.70.5

1.00.50.70.91.70.50.0

-0.2

-1.2-0.7-4.3

-0.1-0.1-0.1

-1.5-0.8

-1.8

1992

0.60.10.6

0.40.31.10.2

-0.4

-1.7-0.1-1.0-0.4

0.70.4

1.60.61.31.42.40.60.0

-0.2

-1.0-0.3-7.0

-0.1-0.1-0.3

-2.1-0.9

-2.5

1993

0.60.00.6

0.40.40.90.2

-0.4

-2.1

-0.1

-0.5

-0.5

0.60.2

2.20.62.02.13.10.6

-0.0-0.1

-0.8-0.1

-10.3

-0.1-0.1-0.5

-2.7-0.8

-3.2

1994

0.6-0.00.6

0.40.50.60.2

-0.4

-2.1-0.1

-0.2-0.6

0.4-0.0

2.80.62.62.63.60.6

-0.0-0.0

-0.60.1

-14.3

-0.1-0.1-0.6

-3.2-0.7

-3.8

1995

0.6-0.00.6

0.40.60.30.1

-0.4

-2.1-0.10.1

-0.6

0.2-0.2

3.20.43.13.14.00.5

-0.00.1

-0.40.2

-19.0

-0.2-0.1-0.8

-3.5-0.5

-4.2

- 11 -

rises. 1/ The German current account surplus drops, ever so slightly.Two points are manifest. The f i r s t is the "negative transmission" of fiscalpolicy that is characteristic of a currency union under flexible exchangerates relative to the rest of the world in a Mundell-Fleming model:contraction in one member country of the union leads to expansion in theother (and vice versa). The second is the importance of the rest of theworld in diminishing the interaction between the two members of the union.Virtually the entire improvement in France's current account balance comesat the expense of the ROW. Of course, this point would carry less weight i fwe included the other members of the EC in the EMU, but i t would s t i l lremain significant.

We see more clearly, in this light, the possible advantage of freedomto use government expenditure flexibly within the EMU. France is ablevirtually to neutralize a 2 percent current account deficit relative to GNPover a five-year horizon by taking action benefiting Germany slightly interms of real output and costing i t about half of one percent of extraannual inflation. Since the shock to which France responds lowers Germanoutput and prices independently, the impact on Germany is even better thani t might otherwise seem. Even i f the result were not entirely to Germanliking because of the net inflationary consequences of France's action, theoutcome could easily be to lighten the cost for France a lot without muchdisturbing Germany. If so, any German fiscal reaction to the i n i t i a limpulse and France's response to i t would only be small and not alter theresult substantially.

3. Medium-run simulation 2: German resistance to inflation resultingfrom monetary policy

The next simulation begins with the assumption that the EMU monetaryauthorities let the total money stock grow by one percent more annually thanpreviously (how this rise is divided up as between French and German port-folios is assumed to be endogenous). A l l other exogenous values followalong the baseline trend. We then repeat the simulation while supposingthat Germany responds by contracting its fiscal policy in order to cut backinflation. Table 3 compares the two results. By depreciating the ecu, thetight fiscal policy in Germany at f i r s t promotes inflation. But inflationhas already come down as of the second year (1991). By 1995, the Germanfiscal program has offset 40 percent of the extra inflation. 2/

The cumulative f a l l in aggregate demand is substantial in Germany(9.3 percent by 1995), though GNP growth drops only in the i n i t i a l year and

1/ If demands for money were formulated in terms of the output deflatorsrather than absorption deflators, a f a l l in interest rates would likelyoccur instead.2/ The feedback rule for Germany in this case is

AG/GDP = -2rwhere n is the rate of change of the absorption deflator and G and GDP areas defined above.

- 12 -

Table 3. German Fiscal Targeting of Zero Inflation in Response to European Money Growth Shocks

(deviations from the scenario with no f i s c a l response) 1/

France

National Income

Real GDP (percent)Real GDP (annual growth rate)Real GNP (percent)

1990

0.10.10.1

Domestic demand (percent) 0.2Consumption expenditure (percent) -0.0Gross private investment (percent) 1,4Exports of goods and services (percent) -0.6Imports of goods and services (percent) 0.1

Government

Gen. gov't financial balance (FFb) -6.8Gen. gov't financial balance/GNP (percent)-0.1Government debt (FFb) 5.8Government debt/GNP 0.0

Interest Rates

Short-term interest rateLong-term interest rate

Prices and Supply

1.00.6

Absorption deflator (percent) 0.8Absorption deflator (annual i n f l a t i o n rate) 0.8GNP deflator (percent) 0.3Export price deflator (percent) 2.0Import price deflator (percent) 3.7Capacity u t i l i z a t i o n rate 0.1Long-term real interest rate -0.2Real user cost of capital -0.5

International Accounts

Trade balance (FFb)Current account balance (FFb)Net foreign assets ($b)

As a percent of nominal GNPTrade balanceCurrent account balanceNet foreign assets

-37.9-37.8-6.1

-0.6-0.6-0.7

Nominal effective exchange rate (percent) -3.6Real effective exchange rate (percent) -1.8

1991

0.50.40.4

0.30.01.40.2

-0.5

-13.0-0.216.4-0.0

0.50.5

1.40.70.92.44.50.4

-0.2-0.4

-20.5-24.2-10.2

-0.3-0.3-1.0

-4.3-1.9

1992

0.60.20.6

0.20.00.90.5

-1.2

-14.9-0.227.2-0.1

0.40.3

2.20.81.82.84.70.6

-0.1-0.2

-0.4-6.6

-11.7

-0.0-0.1-1.1

-4.6-1.8

1993

0.6-0.10.5

-0.0-0.10.20.4

-1.7

-23.0-0.245.5-0.1

0.50.1

3.10.92.73.24.80.50.10.0

15.88.6

-10.9

0.20.1

-0.9

-4.8-1.5

1994

0.3-0.20.2

-0.3-0.4-0.50.1

-2.0

-32.4-0.373.50.1

0.6-0.2

3.90.83.63.75.00.20.10.3

24.317.1-8.9

0.20.2

-0.7

-5.1-1.2

1995

0.1-0.30.0

-0.5-0.6-1.0-0.1-2.1

-38.3-0.3

107.90.3

0.4-0.6

4.50.54.14.05.2

-0.00.10.4

25.719.2-6.6

0.20.2

-0.6

-5.4-1.0

13 -

Table 3 (concluded). German Fiscal Targeting of Zero In f l a t i o n in Response to EuropeanMoney Growth Shocks

(deviations from the scenario with no f i s c a l response) 1/

1/ Deviations are for the level of the variable relative to its value in the correspondingyear for the scenario with no fiscal response.

Germany, F.R.

National Income

Real GDP (percent)Real GDP (annual growth rate)Real GNP (percent)

1990

-2.2-2.3-2.2

Domestic demand (percent) -5.2Consumption expenditure (percent) -1.9Gross private investment (percent) -3.4Exports of goods and services (percent) 0.7Imports of goods and services (percent) -7.8

Government

Gen. gov't financial balance (DMb) 57.7Gen. gov't financial balance/GNP (percent) 2.5Government debt (DMb) -57.7Government debt/GNP -2.0

Interest Rates

Short-term interest rateLong-term interest rate

Prices and Supply

1.00.6

Absorption deflator (percent) 1.0Absorption deflator (annual i n f l a t i o n rate) 1.0GNP deflator (percent) -0.2Export price deflator (percent) 0.5Import price deflator (percent) 4.2Capacity u t i l i z a t i o n rate -2.2Long-term real interest rate 0.8Real user cost of capital 0.7

International Accounts

Trade balance (DMb)Current account balance (DMb)Net foreign assets ($b)

As a percent of nominal GNPTrade balanceCurrent account balanceNet foreign assets

37.439.216.5

1.71.83.0

Nominal effective exchange rate (percent) -3.7Real effective exchange rate (percent) -3.6

EMU Nominal exchange rate (percent in $/ECU) -4.5

1991

-1.90.4

-1.7

-5.6-1.4-3.62.0

-8.4

64.42.6

-120.9-4.4

0.50.5

0.9-0.1-0.80.15.0

-1.80.80.9

46.451.738.6

2.02.25.0

-4.4-4.8

-5.3

1992

-1.60.3

-1.2

-6.0-0.9-3.82.9

-9.1

72.52.8

-192.3-6.9

0.40.3

0.6-0.2-1.5-0.45.6

-1.50.81.0

55.764.867.2

2.32.67.2

-4.8-5.8

-5.7

1993

-1.7-0.1-1.1

-7.0-0.5-3.73.7

-10.5

91.53.4

-282.6-9.8

0.50.1

0.4-0.2-2.3-1.15.9

-1.50.81.0

70.284.2

105.9

2.83.2

10.0

-5.0-6.7

-6.0

1994

-1.9-0.3-1.2

-8.3-0.1-3.44.4

-12.3

122.04.3

-403.3-13.4

0.6-0.2

0.2-0.2-3.1-1.86.4

-1.70.60.9

89.3109.5157.9

3.44.0

13.5

-5.3-7.7

-6.4

1995

-1.90.0

-0.8

-9.30.6

-2.55.2

-13.9

152.85.1

-554.8-17.7

0.4-0.6

-0.2-0.4-4.2-2.76.8

-1.60.30.7

107.3135.8223.1

3.94.7

17.6

-5.7-9.0

-6.8

- 14 -

thereafter GNP grows faster than in the same situation without German fiscaltightening. As this implies, the current account balance swings sharplyinto surplus. Such is the rise in this surplus--both because of the f a l l inabsorption and the induced depreciation of the ecu--that there might bel i t t l e room for further tightening of fiscal policy to combat inflation. Onthe French side, real output is slightly higher, inflation goes up by lessthan one percent annually, and the current account worsens a l i t t l e . Frenchnet foreign assets are lower by about half of a percent of GNP by 1995. A l lin a l l , the story is much the same as previously. Mundell-Fleming isconfirmed. But the confirmation is attenuated by a tendency for theimprovement of the current account in the contracting country to come mostlyat the expense of the outside world. Once again, the possibility thatfiscal policy independence yields a reasonable outcome for both countries isnot remote.

4. Medium-run simulation 3: Contractionary responses by both Franceand Germany to an o i l shock

In a final simulation, we consider the possibility of fiscal responsesto an o i l shock. This shock has stagflationary effects on output and pricesin both countries. Moreover, i t tends to lower their current account sur-pluses (or, in the case of France, to increase the deficit). France wantsto adopt contractionary fiscal policy to combat the current account deficit.Germany wants to do the same to fight inflation.

We had no difficulty simulating the French response without a Germanreaction, or simulating the German response without a French one, but wecould not simulate the two simultaneously. The simulations would notconverge. This could obviously be a reflection of our French and Germanmodels. But i t could be also the result of genuine instability, since eachcountry's fiscal contraction reinforces the other's contractionary tendency.The French reaction improves the French current account, but does so partlyby depreciating the ecu, thereby worsening the inflationary problem forGermany. The German reaction attenuates inflation at home (after a lag),but yields a French current account deficit in the short run (because of aJ-curve effect of the depreciation of the ecu). Moreover, unlike the twoprevious simulations, both of the countries seem about as much affected bythe shock i n i t i a l l y and therefore would be as much impelled to respond.Even i f we dismiss the possibility of genuine instability, there is primafacie evidence of notable room for improvement through fiscal policycooperation. Such cooperation would remove the conflicting attempts ofFrance to reduce its output prices relative to German ones, thuscontributing to a depreciation of the ecu, and of Germany to resist a risein inflation, which becomes a l l the harder as the ecu fa l l s .

III. Some Theoretical Analysis of Current Accounts in the EMU

Up to now the discussion assumes persisting French concern with currentaccount balance. But, as mentioned above, current account deficits nolonger have the same implications in a monetary union: they can be easily

- 15 -

financed and do not portend reserve losses, possible balance-of-paymentscrises, and devaluations, as they did for France most of the time in recentdecades. Do current account targets s t i l l make sense in an EMU? In orderto get some insights into the question, let us return to the scenario inwhich the monetary union as a whole has a balanced current account, bur.France has a deficit and Germany a surplus. To keep the focus on theimportant issues, let us further assume that the French deficit (Germansurplus) is not warranted by structural differences regarding productivity,time preference, or demography. Therefore, the current account imbalancesare not sustainable and the policy question is the optimal path towardequilibrium.

There are two mechanisms that operate, even i f the nominal exchangerate is fixed, to bring about an adjustment in the current account balancesand therefore a rise in Germany's absorption relative to output and a f a l lin France's. First, continuing German surpluses lead to transfers of wealthfrom France to Germany. These wealth changes w i l l lower consumption inFrance and raise i t in Germany. Second, and equally important, insofar asFrench current account deficits correspond to higher investment in France--due to i n i t i a l l y lower capital intensity and higher marginal productivity ofinvestment--the deficits w i l l be associated with an outward shift of theFrench productivity frontier relative to the German one. This shift w i l lincrease the marginal product of French labor relative to German labor overtime, hence result in an increase in the competitiveness of France relativeto Germany.

The second mechanism allows a real exchange rate adjustment to occurwithout either a rise in German prices or a recession-induced f a l l in Frenchreal wages. Indeed, French real wages would rise. The basic reason forconsidering this second mechanism to be important is that the past decadehas seen insufficient investment in France. One cause has been restrictiveaggregate demand policies. But another cause--more significant from ourviewpoint-- is that French real interest rates have been high and have evenrisen since the mid-eighties relative to the German level. The specialsignificance of this last factor stems from its link to the past operationof the EMS and its relationship to the move toward monetary union.

In the early years of operation of the EMS, realignments compensatedfor differentials in inflation. As French inflation has declined to theGerman level, however, adjustments in exchange rates have become smallerthan inflation differentials, and French relative prices have risen. Frenchnominal interest rates have also failed to decline correspondingly, becausea premium has developed on the franc based on the expectation or the risk ofa downward realignment which would offset past increases in French relativeprices. The result has been a rise in real interest rates in France, whichhas led to lower capital accumulation and to a higher marginal product ofcapital in France.

Because of the need for faster capital accumulation, therefore, werethe EMU to be formed today, France would have less cause to resist current

- 16 -

account deficits than i t bad in the past. 1/ The country therefore hasseveral good reasons to resist these deficits less, since as we indicatedbefore, the deficits w i l l also become easier to finance and less troublesomein general in an EMU. Hence, i t is plausible to suppose that the Frenchemphasis on targeting current account balance w i l l decrease.

Let us suppose, next that the German fiscal authorities interfere withthe appreciation in the real exchange rate of German output in the EMU byrestraining the rise in German prices. This would sustain current accountdeficits in France. On the assumptions of the earlier sections, the con-f l i c t with French objectives would be marked. But on present assumptions,the conflict would be lessened since higher current account deficits wouldbe more acceptable to France. Moreover, the German policy would not unam-biguously interfere with the process of international adjustment; whilehindering this adjustment through shifts in relative demand, i t would assistthe adjustment through shifts in relative supply. More precisely, byencouraging the financing of investment in France, the German policy wouldpromote an equilibrating rise in capital intensity in France. 2/ Is theFrench current account deficit, therefore, no longer the source of anylegitimate concern?

Current account deficits correspond to accumulation of net indebted-ness. If the deficit finances productive investment and the rate of returnon investment is greater than the borrowing cost, then the indebtednessshould be self liquidating. If, however, there are distortions that makethe private return on investment higher than the social return, i t is not soclear that deficits are benign. Further, current account deficits maycorrespond to increased consumption; while i t may be optimal to smooth tem-porary consumption shocks, the persistence of high consumption in the faceof reduced income may be cause for concern. 3/ Therefore, the existenceof current account deficits may signal inappropriately stimulative policies,and accumulation of indebtedness that is not self liquidating. In thesecircumstances, i f accumulated indebtedness by France needs to be reversedlater, or at least stabilized as a ratio to GNP (and the economy's rate ofinterest is above the growth rate of output), then French prices w i l l haveto decline relative to foreign prices since the higher interest payments onthe debt w i l l need to be offset by a larger trade surplus. Indeed, i f thedebt is to be repaid, then the ratio of French to German output prices w i l lneed to undershoot its long-run equilibrium value at some stage in theadjustment process. 4/ The period of undershooting is clearly a

1/ This is also the conclusion of Bredenkamp and Deppler (1989).2/ France might also do the same, in particular, by lowering taxes on

capital income. Discussion of the effect of tax policies on the location ofinvestment and capital may be found in Dooley and Isard (1989), Isard(1989), and Tanzi and Bovenberg (1990).3/ See Jacob Frenkel, Morris Goldstein, and Paul Masson (1989).4/ William Branson may be cited as the source of a number of clear

expositions of the dynamics of the adjustment process. For a recent examplein a policy context, see Branson (1988).

- 17 -

disadvantage to French consumers since i t means higher relative prices ofimports. From this perspective, therefore, i t is not the case that per-fectly elastic capital flows in the EMU would make the current accountbalance a matter of indifference.

It is therefore s t i l l reasonable to suppose that German anti-inflationpolicy could lead to French current account deficits that are not desired byFrance. But is there any cause to modify our earlier emphasis on the bene-fits of fiscal policy independence? We think not. We see l i t t l e reason todeviate from our earlier assessment in the case of overpriced French goodsthat fiscal policy independence is a good bet to yield better results thanwould come about without i t . If France and Germany begin from differenti n i t i a l positions or i f their preferences differ, the ability of the Frenchand German authorities to wield a separate instrument adds a desirableelement of flex i b i l i t y . At the opposite extreme, a joint fiscal authoritymight well force convergence to a policy that would represent some mix ofFrench and German preferences. But i f Germans place greater emphasis onprice level stability than the French, while the French place more weight,say, on output smoothing, there is a real possibility that both countrieswould be better off with some difference in fiscal policies, made possibleby retaining fiscal policy flexibility. The answer to fiscal policy con-fl i c t s over aggregate demand policies, therefore, is not to adopt measuresinterfering with the ability of fiscal policies to differ within the EMU,like ceilings on fiscal deficits, but to add an element of fiscal policycooperation. Failing such cooperation, i t would be desirable to design theconstraints on national fiscal policies in a way that kept a degree offlex i b i l i t y .

IV. Concluding Remarks

In sum, our analysis serves to highlight the increased possibility offinancing current account deficits in a monetary union and the value offiscal policy independence in that context. 1/ Our simulations haveillustrated cases in which there would be clear value to retention of fiscalf l e x i b i l i t y , because of different i n i t i a l conditions (or asymmetric shocks)and different preferences. How much fiscal policy cooperation would bedesirable in order to cope with fiscal policy conflicts remains largely amoot point in our discussion, we admit. We presented one example--that ofan o i l shock--where the French and German fiscal authorities seem to belocked in a conflict implying notable room for improvement through coop-eration. Indeed cooperation could even be essential in this example. Butin light of our central concern with showing the value of fiscal policyindependence, the benefits of cooperation have been left mostly in the

1/ For a highly complementary analysis, see Tommaso Padoa-Schioppa(1990).

- 18 -

background. 1/ Also left in the background is the issue of possibleincentives governments may have to run excessive fiscal deficits in amonetary union, and the role of deficit ceilings in limiting thispossibility.

Obviously, our medium-term baseline projections might be radicallyaltered by developments in Eastern Europe, and in particular, by unificationof the Federal Republic of Germany with the German Democratic Republic.Though we do not discuss German unification here, i t could be analyzed as ashock that would likely affect France and Germany quite differently. Wewould argue that in these circumstances fiscal f l e x i b i l i t y would be a l l themore desirable, and deficit ceilings a l l the more unproductive.

Another basic question that we have not considered is the possible useof regional transfers to alleviate undesirable consequences of monetaryunion. No doubt such transfers might be an advantageous means of puttingbounds on the redistributions associated with fiscal policy independence.But as indicated in the introduction, strong views have been expressed aboutthe importance of engaging in such transfers, some going so far as tosuggest that a fully integrated fiscal system may even be essential for amonetary union to work. Having emphasized the value of retaining somefiscal policy independence in a European Monetary Union, i t behooves us toreturn to this strong form of the argument for fiscal integration. 2/Though we do not pretend to do justice to the whole issue, we would like toraise a few skeptical observations about the c r i t i c a l need for a system oftaxes and transfers in order to compensate for unfavorable country-specificshocks.

There is a considerable question, at the start, about how far thepreceding argument for an integrated fiscal system really concerns theviability of a monetary union rather than an independent question of dis-tributive justice. The most important fiscal mechanisms in the automatictransfer of resources to adversely hit areas--including the income taxsystem above all--essentially arrived late: only in the course of thiscentury, in many countries mostly since the Second World War. Yet theearlier meagerness of transfers to depressed areas rarely gave rise to callsfor monetary independence in particular regions. Politically speaking,depressed areas tend to call for aid, not for monetary independence. Thismight be different, of course, in the instance of newcomers in a monetary

1/ Bredenkamp and Deppler (1989) concluded that the problems fromnoncooperative fiscal policies might in fact be weaker in a monetary unionthan in a flexible exchange rate regime. Insofar as cooperative adjustmentto shocks nonetheless remains desirable, and insofar as monetary policy isnot independent of fiscal policy, there may be an argument for limitingfiscal policy f l e x i b i l i t y in order to gain credibility for monetary policy.See Matthew Canzoneri (1985).2/ The classic reference on this matter is Peter Kenen (1969), a work

discussing Robert Mundell's important early contribution to the subject ofoptimal currency areas, Mundell (1961) .

- 19 -

union. But how many prospective entrants in the EMU could look back to anearlier time when monetary independence was an effective means of respondingto adverse shocks?

There is another set of major questions. The sort of system oftransfers that advocates of fiscal integration have in mind would operateautomatically. What would therefore prevent the system from coming intooperation because of problems that a country brought upon i t s e l f and thataroused no sympathy elsewhere in the union? To what extent would residentsof one country be prepared to make transfer payments to that of anotherbecause of an enormous wage settlement in the latter? To what extent wouldthey be prepared to compensate others for high unemployment in some indus-t r i a l sectors that resulted from very tight fiscal policy? But is there anyway to set up an automatic system of transfers at the Community level thatwould draw distinctions of this sort? And i f not, is there scope for devis-ing a fis c a l system in a monetary union to correct regional problems thatwould otherwise have been solved through exchange rate changes? As thesequestions indicate, a basic tension exists between the notion of independentstates wishing to preserve their national identities and any major system oftaxes and transfer payments between them. Whereas such a system might verywell be acceptable as a way of funneling aid to poorer members, i t might notbe acceptable i f i t were a means of providing collective insurance foreveryone.

- 20 -

References

Branson, William, "International Adjustment and the Dollar: PolicyIllusions and Economic Constraints," in Wilfrid Guth, moderator,Economic Policy Coordination (IMF and HWWA-Hamburg, 1988).

Bredenkamp, Hugh, and Michael Deppler, "Fiscal Constraints of a FixedExchange Rate Regime," IMF Working Paper No. WP/89/107 (Washington:International Monetary Fund, December 1989).

Canzoneri, Matthew, "Monetary Policy Games and the Role of PrivateInformation," American Economic Review. Vol. 75 (December 1985),pp. 1056-70.

Cohen, Daniel, and Charles Wyplosz, "The European Monetary Union: AnAgnostic Evaluation," CEPR Discussion Paper No. 306 (April 1989).

Delors Report, "Report on Economic and Monetary Union in the EuropeanCummunity," European Community (April 1989).

Dooley, Michael and Peter Isard, "A Note on Fiscal Policy, LocationalDecisions, and Exchange Rates," IMF Research Department, September 8,1989

Eichengreen, Barry, "One Money for Europe? Conceptual Issues and Lessonsfrom the U.S. Currency Union," prepared for Economic Policy (September1989).

Frenkel, Jacob A., Morris Goldstein, and Paul Masson, "InternationalDimensions of Monetary Policy: Coordination Versus Autonomy," paperprepared for a symposium Monetary Policy Issues in the 1990's, JacksonHole, Wyoming, August 30-September 1, 1989 (Federal Reserve Bank ofKansas City, forthcoming.

International Monetary Fund, World Economic Outlook (Washington:International Monetary Fund, October 1989).

Ingram, James, Regional Payments Mechanisms: The Case of Puerto Rico(University of North Carolina Press, 1962).

Isard, Peter, "Corporate Tax Harmonization and European MonetaryIntegration," IMF Research Department, September 8, 1989.

Kenen, Peter, "The Theory of Optimum Currency Areas: An Eclectic View," inRobert Mundell and Alexander Swoboda, eds., Monetary Problems of theInternational Economy (University of Chicago Press, 1969).

Krugman, Paul, "Policy Problems of a Monetary Union," prepared for Bank ofGreece and CEPR conference on "The EMS in the 1990s," (August 1989).

- 21 -

Masson, Paul, Steven Symansky, Richard Haas, and Michael Dooley,"MULTIMOD--A Multi-Region Econometric Model," Staff Studies for theWorld Economic Outlook (Washington: International Monetary Fund, July1988) pp. 50-104.

Mundell, Robert, "A Theory of Optimum Currency Areas," American EconomicReview, Vol. 51 (September 1961), pp. 657-65.

van der Ploeg, Frederick, "Fiscal Aspects of Monetary Integration InEurope," CEPR Discussion Paper No. 340 (August 1989).

Padoa-Schioppa, Tommaso, "Fiscal Prerequisites of a European MonetaryUnion," prepared for Bank of Israel and David Horowitz Instituteconference on "Aspects of Central Bank Policy" (January 1990).

Portes, Richard and Jose Vinals, eds., The European Monetary System inTransition, a CEPR Report (September 1989).

Sachs, Jeffrey and Xavier Sala-i-Martin, "Federal Fiscal Policy and OptimumCurrency Areas," (mimeographed, Cambridge, Massachusetts: HarvardUniversity, 1989).

Tanzi, Vito and Lans Bovenberg, "Is There a Need for Harmonizing CapitalIncome Taxes within EC Countries?" in H. Seidel, ed., Reforming CapitalIncome Taxation (Kiel Institute of World Economics, forthcoming, 1990).