1. Monetary constraints

It has long been recognized that the two arms of macroeconomic policy—monetary and fiscal—are not entirely independent. Blinder and Solow (1973) demonstrated that purely bond-financed increases in the budget deficit are unsustainable and must eventually be reversed or monetized. The consequences of the government’s financing constraint were explored further by Sargent and Wallace (1981), who showed that—if there is an upper limit on the stock of debt (per capita) which the market will bear, and if the prevailing interest rate exceeds the growth rate of the economy—the authorities will eventually lose control of the monetary base (and hence inflation) unless fiscal policy is subordinated to monetary policy. ^2/ Since budget deficits must be financed either by bond sales or money creation, simultaneous constraints on both sources of finance must imply a constraint on deficits themselves.

A number of commentators (see, for example, Tanzi and Ter-Minassian (1987), Dornbusch (1988a,b), Gros (1989), Roubini and Sachs (1989)), have drawn attention to the potential implications this analysis has for the fiscal policy of EMS members. Countries with large budget deficits are denied, or have highly restricted, access to revenue from seigniorage once they join the EMS and are obliged to maintain a rate of monetary growth in line with their partners. Hence, it could be argued, if they are to avoid unsustainable increases in their debt-GDP ratios, their budget deficits must be tailored accordingly. Fiscal policy is no longer a free instrument. As is well known, this constraint tends to be significant for very high inflation countries. In these countries, monetary financing of fiscal deficits often tends to be large, and other things being equal, proportional to the size of deficits. ^3/ However, these are likely to be countries which—perhaps because their revenue base is narrowed by widespread tax evasion, or because of a history of financial and political instability—are up against the limit in terms of how much of their debt the market will bear; that is, they conform to the Sargent-Wallace type. If it were the case that prospective members of the hard currency regime would be faced with rising debt-GDP ratios as a result of loss of seigniorage, this would suggest that they, too, could soon reach the Sargent-Wallace constraint and be forced to tighten fiscal policy.

But how likely are they to hit this constraint in practice? One way to approach this question is to compare prevailing budget deficits with those that might be necessary to stabilize debt ratios in a hard currency regime. ^4/ Following this approach, consider two extreme cases: France, which has the lowest debt-GDP ratio in the EMS; ^5/ and Belgium, which has the second highest (after Ireland). Assuming that the income velocity of high-powered money is constant over time, it is possible to calculate the ratios of primary budget deficit or surplus (i.e. excluding interest payments) to GDP necessary to stabilize existing debt-GDP ratios for these countries, given plausible values for the main economic variables in a hard currency regime. This is done in Table 1, where we assume 3 percent real growth, 2 percent inflation and real interest rates of 5 percent. Under these assumptions, it would appear that France’s budgetary position, not surprisingly, is already substantially better (its primary surplus larger) than would be required to prevent any rise in the existing debt-GDP ratio. This, combined with its comparatively low initial level of debt, would suggest that a country in France’s position could retain a significant margin for maneuver on fiscal policy while adhering to the strict monetary constraints of a hard currency regime.

Table 1.

Debt Stabilization in a Hard Currency Regime

Definitions:

VH	=	Ratio of nominal GDP to base money (calculated as the average observed in 1984-88).
n	=	Growth rate of GDP.
ρ	=	Inflation rate.
r	=	Nominal interest rate.
y	=	Nominal GDP.
b	=	Net general government debt.
d	=	Primary general government deficit (negative sign indicates surplus).
(d/y)	=	$\mathrm{\left(}\mathrm{n}\mathrm{-}\mathrm{r}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\frac{\mathrm{b}}{\mathrm{y}}\mathrm{+}\mathrm{\left(}\mathrm{n}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\mathrm{/}{\mathrm{V}}_{\mathrm{H}}$ (See Argy (1989) for derivation.)

View Table

VH	=	Ratio of nominal GDP to base money (calculated as the average observed in 1984-88).
n	=	Growth rate of GDP.
ρ	=	Inflation rate.
r	=	Nominal interest rate.
y	=	Nominal GDP.
b	=	Net general government debt.
d	=	Primary general government deficit (negative sign indicates surplus).
(d/y)	=	$\mathrm{\left(}\mathrm{n}\mathrm{-}\mathrm{r}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\frac{\mathrm{b}}{\mathrm{y}}\mathrm{+}\mathrm{\left(}\mathrm{n}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\mathrm{/}{\mathrm{V}}_{\mathrm{H}}$ (See Argy (1989) for derivation.)

• Download Figure
• Download figure as PowerPoint slide

View Full Size

• Download Figure
• Download figure as PowerPoint slide

Table 1.

Debt Stabilization in a Hard Currency Regime

Parameter	France	Belgium
VH	15.30	12.30
n	.03	.03
ρ	.02	.02
r	.07	.07
b/y	.38	1.25
Required (d/y)	-0.4 %	-2.1 %
Actual (d/y) in 1988	-1.3 %	-1.9 %

Definitions:

VH	=	Ratio of nominal GDP to base money (calculated as the average observed in 1984-88).
n	=	Growth rate of GDP.
ρ	=	Inflation rate.
r	=	Nominal interest rate.
y	=	Nominal GDP.
b	=	Net general government debt.
d	=	Primary general government deficit (negative sign indicates surplus).
(d/y)	=	$\mathrm{\left(}\mathrm{n}\mathrm{-}\mathrm{r}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\frac{\mathrm{b}}{\mathrm{y}}\mathrm{+}\mathrm{\left(}\mathrm{n}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\mathrm{/}{\mathrm{V}}_{\mathrm{H}}$ (See Argy (1989) for derivation.)

• Download Figure
• Download figure as PowerPoint slide

View Table

Table 1.

Debt Stabilization in a Hard Currency Regime

Parameter	France	Belgium
VH	15.30	12.30
n	.03	.03
ρ	.02	.02
r	.07	.07
b/y	.38	1.25
Required (d/y)	-0.4 %	-2.1 %
Actual (d/y) in 1988	-1.3 %	-1.9 %

Definitions:

VH	=	Ratio of nominal GDP to base money (calculated as the average observed in 1984-88).
n	=	Growth rate of GDP.
ρ	=	Inflation rate.
r	=	Nominal interest rate.
y	=	Nominal GDP.
b	=	Net general government debt.
d	=	Primary general government deficit (negative sign indicates surplus).
(d/y)	=	$\mathrm{\left(}\mathrm{n}\mathrm{-}\mathrm{r}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\frac{\mathrm{b}}{\mathrm{y}}\mathrm{+}\mathrm{\left(}\mathrm{n}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\mathrm{/}{\mathrm{V}}_{\mathrm{H}}$ (See Argy (1989) for derivation.)

View Table

VH	=	Ratio of nominal GDP to base money (calculated as the average observed in 1984-88).
n	=	Growth rate of GDP.
ρ	=	Inflation rate.
r	=	Nominal interest rate.
y	=	Nominal GDP.
b	=	Net general government debt.
d	=	Primary general government deficit (negative sign indicates surplus).
(d/y)	=	$\mathrm{\left(}\mathrm{n}\mathrm{-}\mathrm{r}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\frac{\mathrm{b}}{\mathrm{y}}\mathrm{+}\mathrm{\left(}\mathrm{n}\mathrm{+}\mathrm{\rho }\mathrm{\right)}\mathrm{/}{\mathrm{V}}_{\mathrm{H}}$ (See Argy (1989) for derivation.)

• Download Figure
• Download figure as PowerPoint slide

View Full Size

• Download Figure
• Download figure as PowerPoint slide

More significantly, it appears that even a highly-indebted country such as Belgium would be able to stabilize its debt ratio at a level close to that currently prevailing with little or no reduction in its budget deficit. ^6/ In addition, it could be argued that the very fact of joining a hard currency club would enhance the degree of integration between domestic capital markets and those in partner countries (by reducing exchange rate uncertainty). This could make Belgian government debt a closer substitute for financial assets elsewhere in the hard currency area and thereby enable a larger outstanding stock to be absorbed at “tolerable” interest rates.

What the arithmetic in Table 1 shows, in fact, is simply that it is quite possible to have very different (stable) levels of public sector debt and budget deficits alongside comparable rates of growth in the monetary base. The reason that Belgium and France would not see their debt destabilized by the rigors of a hard currency regime is that both countries have already all but eliminated their recourse to monetary financing (as has the Netherlands—another high-debt/high-deficit country—see Table 2). This is not to say that all high-debt countries wishing to stablize their exchange rate can be complacent about their fiscal positions. In some countries, it might only be feasible to make the necessary reduction in monetary financing by cutting the budget deficit itself. ^7/ But it implies—and Table 2 confirms—that monetary and inflation rate convergence (hence exchange rate stability) can be consistent, both in practice and in principle, with a wide range of fiscal positions.

Table 2.

Selected Indicators for EC Countries

(1987-88 averages)

Sources: OECD Economic Outlook, IMF International Financial Statistics; and staff calculations.

^1/Percent of GNP.

^2/Defined as change in Central Bank holdings of government securities.

^3/1986 Q3-1988 Q3 annual average.

Table 2.

Selected Indicators for EC Countries

(1987-88 averages)

	CPI Inflation	General Government Deficit 1/		Net Debt	Current Account Balance 1/	Monetary Financing 1/2/		Exchange Rate Change (%) Vis-à-vis DM
Germany	0.7	1.9		23.1	4.0	zero		—
Netherlands	zero	7.3		53.8	2.0	0.2		0.1
Belgium	1.4	9.0		123.2	2.1	-1.0		-0.9
Ireland	2.7	7.8	3/	131.2	1.9	0.1		-4.2
France	2.9	1.5		25.7	-0.4	-0.1		-3.0
Denmark	4.3	-1.2		24.8	-2.5	-0.1		-1.4
U.K.	4.6	0.3		40.9	-2.0	-0.9		-0.7
Italy	4.9	10.9		91.5	-0.3	0.7	4/	-3.7
Spain	5.1	3.4		31.1	-0.6	0.3		-1.3
Portugal	9.6	n.a.		n.a.	-0.1	n.a.		-8.2
Greece	15.0	n.a.		n.a.	-2.3	n.a.		-10.6

Sources: OECD Economic Outlook, IMF International Financial Statistics; and staff calculations.

^1/Percent of GNP.

^2/Defined as change in Central Bank holdings of government securities.

^3/1986 Q3-1988 Q3 annual average.

View Table

Table 2.

Selected Indicators for EC Countries

(1987-88 averages)

	CPI Inflation	General Government Deficit 1/		Net Debt	Current Account Balance 1/	Monetary Financing 1/2/		Exchange Rate Change (%) Vis-à-vis DM
Germany	0.7	1.9		23.1	4.0	zero		—
Netherlands	zero	7.3		53.8	2.0	0.2		0.1
Belgium	1.4	9.0		123.2	2.1	-1.0		-0.9
Ireland	2.7	7.8	3/	131.2	1.9	0.1		-4.2
France	2.9	1.5		25.7	-0.4	-0.1		-3.0
Denmark	4.3	-1.2		24.8	-2.5	-0.1		-1.4
U.K.	4.6	0.3		40.9	-2.0	-0.9		-0.7
Italy	4.9	10.9		91.5	-0.3	0.7	4/	-3.7
Spain	5.1	3.4		31.1	-0.6	0.3		-1.3
Portugal	9.6	n.a.		n.a.	-0.1	n.a.		-8.2
Greece	15.0	n.a.		n.a.	-2.3	n.a.		-10.6

Sources: OECD Economic Outlook, IMF International Financial Statistics; and staff calculations.

^1/Percent of GNP.

^2/Defined as change in Central Bank holdings of government securities.

^3/1986 Q3-1988 Q3 annual average.

2. Current account constraints

Aside from domestic monetary constraints, fiscal policy is also often viewed as constrained by the need to keep the current account balance above some floor. Indeed, the existence of such a constraint is often thought to account for much of the close historical relationship between national savings and investment rates, despite capital mobility. That is, governments are viewed as having adjusted policies so as to minimize the gap between these otherwise largely independent variables (Bayoumi (1989)). Although the existence of such an external constraint does not hinge on the choice of exchange rate regime, this constraint is often viewed as particularly binding in the case of countries pursuing a hard currency policy because of the implications current account imbalances have for the core of the policy—the fixity of the exchange rate. Indeed, countries pursuing such policies do tend to keep a particularly careful watch on their external position, especially when the balance on current account is in deficit. This has been most evident in the case of France over the past few years, but applies to other members of the ERM with current account deficits as well. In this perspective, deficits are viewed as symptomatic of excess demand and underlying inflationary pressure; of future pressures on the exchange rate; and more generally of a country “living beyond its means”. Accordingly, a key objective of fiscal policy is taken to be the need to constrain the current account above some floor—often taken to be balance.

Almost all would agree that fiscal policy ought to be set in part with a view to avoiding excess demand, inflationary and exchange rate pressures, and long run solvency problems. What is less clear is the usefulness of a current account objective in this process. It is true of course that excess demand and inflationary pressures have traditionally led to a deterioration in current account positions, as indeed is to be expected on analytical grounds. However, current accounts are affected by a host of other factors as well—e.g., shifts in the productivity of capital or labor, demographic factors or portfolio preferences—that bear no simple relationship to the postulated policy concerns. For instance, it is not clear that an investment-induced deterioration in the current account is necessarily inflationary; nor does it have any clear implications for long-run solvency or national wealth (Ingram (1973)). Hence, there would seem to be little usefulness in simple current account rules to guide fiscal policy. Indeed, any such rules could well prove to be welfare-reducing—e.g., if an investment-induced worsening of the current account was choked-off through a tightening of fiscal policy, the result being a lower long-run capital stock (see Bean, 1989).

Leaving aside for the moment credibility issues (discussed below), current account objectives would seem particularly inappropriate for European countries participating in a hard currency arrangement, for two reasons. First, the hard currency option brings with it a stable, non-accommodating nominal framework for macro-management. ^8/ Given this framework, the current account balance would be expected to lose much of the usefulness it may once have had as an indicator of deficiencies in nominal demand management. Second, the process of economic and financial integration underway in Europe is likely to have a major, if highly uncertain, impact on national saving and investment propensities. Thus, while any deficiencies in nominal demand management would still show up in current accounts, the noise in the relationship would be such that it would be preferable for policies to focus directly on the indicators of interest—inflation, consumption and savings, etc.—rather than the current account.

3. Credibility considerations

The foregoing suggests that adoption of a hard currency policy in a country that has achieved inflation convergence has no necessary implications for its fiscal “room for maneuver.” This is a view which many would find counterintuitive. It could be argued, quite reasonably, that the problem with the analysis so far is that it does not take account of the intangible links between levels of indebtedness and current account balances, on the one hand, and confidence in the exchange rate objective, on the other. Confidence, or credibility, is a slippery concept: to the extent that it already exists, the impact of the so-called fundamentals on the exchange rate (or on the level of interest rates necessary to maintain a given exchange rate) disappears; yet it may not be possible to establish credibility in the first place unless the fundamentals are in order. For instance, while prevailing debt levels might be viewed as acceptable by markets, they could, in the face of unfavorable shocks, be perceived as increasing the incentive for the authorities to breach the exchange rate commitment and raise the domestic price level: the larger the stock of debt the greater the returns from monetization (see Giavazzi and Pagano (1989)). As a result, governments may see themselves as having to accept constraints on fiscal policy as an investment in credibility—even where, in a deterministic world, no such constraints should exist.

A second, closely related reason for governments to apply prior restraint on fiscal policy is to forestall the emergence of “unacceptable” increases in interest rates. Interest rates are the residual element of policy in a hard currency regime, being forced to endure the brunt of the adjustment to unexpected developments. Indeed, insofar as demand management policies do not adjust fully to the change in currency regime, it can be expected that the variability of exchange rates is simply transposed on to interest rates. This increased variability is not without its costs, and governments may well wish to forestall such costs by taking steps to diminish the currency’s exposure to shocks.

While such a view can be defended, a government committed to a hard currency policy would have to ask itself whether it would not be preferable—through a combination of repeated policy statements and determined use of interest rates—to “ride out” the exchange market pressures rather than continuously seeking to forestall them through more restrictive fiscal policies. The issues here are ultimately empirical, involving as they do a cost benefit analysis of the two strategies. Nevertheless, governments committed to a hard currency policy should perhaps not overestimate the risks on the interest rate side. Interest rates are not, in general, seen as having very large real effects on European economies. Moreover, these effects would presumably be even smaller under the postulated circumstances since given the commitment to a stable nominal macroeconomic framework, the rise in interest rates would be transitory and largely limited to short-term interest rates. This might well be a cost worth bearing if one considered that it could establish the underlying credibility of the exchange rate policy and hence diminish the probability of incurring such costs in the future. Secondly, it is worth noting that the country which has the longest experience with a hard currency regime free of capital controls (the Netherlands) is also the country which has experienced the least variability and the smallest differentials in interest rates vis-à-vis the anchor country (see Table 3). While this may be due in part to the Netherlands’ strong current account position, the fiscal accounts have not been nearly as solid. Moreover, the 1986 fall in oil prices constituted a major adverse and asymmetric shock for that economy in relation to other countries participating in the ERM. The implications would seem to be therefore that adoption of a hard currency policy, together with a willingness to adjust interest rates to whatever degree was required to maintain the exchange rate parity (as is the case with the Netherlands) can significantly lessen the interest rate risks confronting national authorities.

Table 3.

Behavior of Short-Term Interest Rates ^1/

(Based on monthly data, Jan. 1985 - Aug. 1989)

Source: IMF International Financial Statistics.

^1/Three-month interbank or Treasury bill rates.

^2/Up to July 1989 only.

Table 3.

Behavior of Short-Term Interest Rates ^1/

(Based on monthly data, Jan. 1985 - Aug. 1989)

		Own Rates		Differential vs. Germany
		Average	Standard deviation	Average	Standard deviation
Germany		4.9	1.0	—	—
Netherlands	2/	5.4	0.9	0.6	0.5
Belgium		7.8	1.2	3.0	1.1
Ireland		10.3	2.3	5.5	2.4
France		8.6	1.0	3.7	0.9
Denmark		n.a.	n.a.	n.a.	n.a.
U.K.		10.6	1.6	5.7	0.8
Italy		11.9	1.4	7.0	1.2
Spain		10.9	2.2	6.0	2.0
Portugal		n.a.	n.a.	n.a.	n.a.
Greece		16.8	0.6	12.2	1.2

Source: IMF International Financial Statistics.

^1/Three-month interbank or Treasury bill rates.

^2/Up to July 1989 only.

View Table

Table 3.

Behavior of Short-Term Interest Rates ^1/

(Based on monthly data, Jan. 1985 - Aug. 1989)

		Own Rates		Differential vs. Germany
		Average	Standard deviation	Average	Standard deviation
Germany		4.9	1.0	—	—
Netherlands	2/	5.4	0.9	0.6	0.5
Belgium		7.8	1.2	3.0	1.1
Ireland		10.3	2.3	5.5	2.4
France		8.6	1.0	3.7	0.9
Denmark		n.a.	n.a.	n.a.	n.a.
U.K.		10.6	1.6	5.7	0.8
Italy		11.9	1.4	7.0	1.2
Spain		10.9	2.2	6.0	2.0
Portugal		n.a.	n.a.	n.a.	n.a.
Greece		16.8	0.6	12.2	1.2

Source: IMF International Financial Statistics.

^1/Three-month interbank or Treasury bill rates.

^2/Up to July 1989 only.

1. Requirements for monetary stability

We have argued that financing constraints, whether of monetary or external origin, would not, in and of themselves, require a particular configuration of fiscal policies in countries pursuing a hard currency policy. A proponent of binding rules might respond that the arguments advanced—if they have any merit at all—apply only in the case of countries whose fiscal policies are seen by the markets to be clearly “under control.” That is, though deficits and debt ratios may be very different, they must be seen to be on a stable or declining path. This has been more or less the case, in recent years, for the two high-debt, strong-currency countries referred to so far: Belgium and the Netherlands. A major concern of the Delors Committee, however, is that a shift toward fixed exchange rates, and especially toward monetary union, would ease the normal constraints facing national fiscal authorities. This easing could encourage policies which do not meet the “stability” criterion and could thereby undermine the system unless neutralized by centrally determined binding rules on fiscal policy, designed to preserve overall monetary stability.

The easing of the constraints could take several forms. One element would be the elimination of the currency premium in interest rates. This would result in a once-and-for-all, step decrease in the debt service burden, which might cause governments to let up on control of noninterest expenditure, for a period at least. Another would be the easing of the current account constraint (and, thus, at one remove, of the fiscal constraint) because of the reduced dependence of the exchange rate on current account developments as the exchange rate regime hardened. Thirdly, as was suggested earlier, the financial integration of Europe—running alongside and given further impetus by moves toward monetary union—would broaden and deepen the market for government paper, making funding of larger deficits feasible. Moreover, the very creation of a European Monetary Union (EMU) would imply increased solidarity among its members, raising the prospect of bail-outs if pressures on member states should become too severe (Isard (1989)). Taken together, these elements could be expected at least to slow countries’ adjustment to adverse shocks and might even, in extreme circumstances, lead some into fiscal “adventurism.” One would, of course, expect market discipline to counteract at least some of these tendencies. However, in the view of the proponents of binding rules, market discipline could not be expected to be as effective on an ex ante basis as the discipline provided historically by the fear of devaluation. There is, therefore, a built-in threat of increased fiscal laxity and hence, ultimately, of monetary accommodation that can only be countered by imposing constraints on the fiscal policies of member countries.

While many of the points in this argument are valid, they do not reflect the whole story. Although the adoption of a hard currency policy can be expected to lower real interest rates and hence ease fiscal constraints in one respect, that easing is the result of the Government’s having accepted, in a form credible to the market, another constraint—not to inflate their problems away. There is, in the end, no net easing of discipline. To assume otherwise is to suppose that markets would accept a lower risk-adjusted expected return from governments that adopt a hard currency policy than from those that do not, which is implausible. Similarly, while the discipline of the exchange rate would obviously disappear in a monetary union, that symbol of the quality of countries’ policies would be replaced by others, e.g., credit ratings and interest rate spreads. In the same vein, the fact that financial integration broadens the market for debt does not mean that governments would be able to be any less responsible—in the sense of pursuing a more lax fiscal strategy. On the contrary, since domestic residents will face increased possibilities for placing their savings abroad, governments may end up having to compete for funds on the basis of a sound fiscal reputation.

Similarly, the evidence used to suggest that market discipline might be insufficient (Lamfalussy (1989), Thygesen (1989)) is not particularly convincing. It is not clear, for example, that the fact that there is only a 50 basis point differential in the borrowing rates of Canadian provinces (despite differences in fiscal positions that approach those of the EC) is relevant to the present situation in the EC. The Canadian provinces achieved monetary unification many years ago, and the differential in their borrowing rates today is unlikely to provide even an approximate notion of how markets would discriminate among EC borrowers as they progress through stages I, II and III. A more relevant measure of such market discipline might be the cost in terms of real interest rates that markets have exacted from those countries that have sought to pursue hard currency policies. This is generally deemed to have been high.

Finally, market discipline, though it is clearly a crucial factor, is not necessarily the only force motivating responsible fiscal behavior; governments’ (and their electorates’) own conceptions of the proper role of the state in modern economies and concerns about limiting the share of interest payments in the budget or the level of taxation are also likely to play their part. It is noteworthy, in this connection, that the evidence on the finances of local governments in federated states (Lamfalussy, 1989) does not suggest that these units have been particularly undisciplined. The bulk, albeit not all, of the evidence is rather in the opposite direction. Indeed, juxtaposition of the evidence on the finances of federated states cited in the Delors report with that of the finances and inflation performance of nation-states presented in International Financial Statistics suggests, if anything, that fiscal prudence is inversely proportional to the authorities’ leverage over monetary policy, i.e., their access to the inflation tax.

We are thus rather doubtful about the “lack of discipline” arguments. It remains, however, that a monetary union does raise free-rider issues—the incentive for members to use the system to shift their problems onto other members. It is possible to envisage a scenario whereby one member’s fiscal policies became so extreme—possibly in part as a result of a misperception of the true nature of the constraints—as to lead to pressures for a bail-out at the community level which in turn might generate an unwanted degree of monetary expansion and undermine the monetary stability of the system. ^9/ This is a possibility. But the solution is clear: monetary arrangements at the Community level that would insulate monetary policy from such developments. This could be achieved in a variety of ways, e.g., through restrictions on the Central Bank’s holdings of government paper and/or adoption of a binding monetary rule, e.g., regarding monetary growth or inflation. Given such arrangements, it should not even be necessary to impose a no-bail-out restriction since bail-outs (such as, for instance, that of the S&Ls in the United States) do not, in and of themselves, generate monetary accommodation. Only if the complete independence of the Central Bank could not be guaranteed, allowing the possibility that the pressures (e.g., on interest rates) generated by a bail-out could lead to monetary accommodation, would a no-bail-out restriction also be required. In either case, of course, bail-outs are a fiscal threat to members of a monetary union, just as they are to a country’s tax payers when a government bails out a corporation, for instance. It is therefore quite appropriate for participating countries to guard against such threats by explicitly prohibiting cross-guarantees regarding the public debt of member countries—an arrangement that would provide the basis for a standard “stabilization program” arrangement if a member should run into difficulties. Be this as it may, the basic point is that free-rider threats to the financial stability of a monetary union are best taken care of by clear monetary “rules of the game,” of the sort mentioned, than by additional restrictions on fiscal policy.

2. Fiscal coordination and the efficiency of policymaking

The second main strand in the argument for increased centralization of fiscal policies within a hard currency area can be viewed as an extension of the general case for policy coordination. This derives from the existence of significant spillover effects on the economy of one country resulting from policies pursued in another (see, for example, Corden (1986)). Where all countries are “small”, in the sense that no individual government can influence world relative prices through its policy actions, there is no need for coordination: by analogy with the perfect competition model and standard welfare theory, it can be deduced that decentralized and uncoordinated policymaking is efficient when governments are atomistic. This has been proved formally by Backus et al. (1988). But in the more pertinent case where countries are large enough to have perceptible effects on one another, there is a problem of externalities: a government may fail to give adequate (if any) weight in its policymaking to potential effects on partner countries. These externalities generate inefficiency—excessive or insufficient policy action relative to some optimum—which coordination is supposed to cure. It is this notion which underlies much of the case for fiscal coordination in Lamfalussy’s (1989) contribution to the Delors report.

A point which is sometimes overlooked in the debate on this subject is that it is not sufficient to demonstrate the existence of externalities from uncoordinated fiscal policies in order to make central coordination a necessary ingredient of a fixed exchange rate regime or monetary union. If the externalities would have existed in any case (i.e., with flexible or managed floating exchange rate regimes), and if they were not aggravated by the imposition of fixed exchange rates, then they would have no bearing on the present debate. In these circumstances, the merits and demerits of policy coordination and the prerequisites for an effective hard currency regime should properly be treated as separate issues. If, on the other hand, the move to fixed exchange rates could be shown to generate new or more significant spillovers between participating countries, then there would be grounds for requiring enhanced policy coordination as part of the new regime. Bearing this in mind, we consider two of the main potential sources of inefficiency which have been ascribed to uncoordinated fiscal policy: disturbances to saving, investment and real interest rates in partner countries; and incentives for beggar-thy-neighbor strategies in response to exogenous shocks.

On the first of these, the argument runs roughly as follows. With integrated capital markets and no exchange controls, fiscal expansion in country X will: (a) draw partly on the savings of country Y; (b) almost certainly reduce the two countries’ combined saving rate (though the level overseas may rise); and (c) tend to push up real interest rates in both countries, crowding out private sector investment. ^10/ This sequence of effects does not, in itself, imply any inefficiency in policy. Residents of country Y would only reduce domestic investment or raise saving (if at all) to the extent that they are compensated for doing so by higher interest rates. However, the policy of country X may generate costs for Y which are not internalized (Corden, 1986). First, there are likely to be adjustment costs imposed on Y, for example in labor markets, where it would be necessary to reduce the average capital-labor ratio in order to avoid increasing unemployment. The less flexible are relative factor prices in Y, the more difficult this adjustment will be. If X’s fiscal expansion were subsequently and unexpectedly reversed, the adjustment costs could be even higher, even though saving, investment, and real interest rates may return to previous levels. Secondly, over a longer horizon, higher real interest rates may adversely affect the collective potential growth rate, with implications for intertemporal or intergenerational welfare. Only if X’s government happened to be acting “as if” it were a competitive private entity—using the proceeds of its extra borrowing for productive capital investment at market-determined interest rates—would this problem not arise. ^11/

Clearly, these sorts of externalities could occur whatever exchange rate regime was in place. As regards a hard currency regime (and particularly a monetary union) one could follow Lamfalussy (1989) and others in suggesting that governments might be encouraged to be less careful about using borrowed funds efficiently and productively, on the assumption that other countries would be more likely to share the burden of any future financing difficulties. This was discussed in the previous section, and the same counterarguments used there can be applied. Alternatively, it is possible to argue that the (alleged) increase in the international substitutability of assets associated with a fixed exchange rate regime could raise the degree to which a given amount of government borrowing in one country diverts saving from other partner countries. Any externalities involved would therefore be exacerbated. This is a valid concern, but one which should not be overstated: increasing integration of European (and world) capital markets has been a prominent feature of the 1970s and 1980s, and the “hardening” of exchange rates could be seen as a continuation of that process rather than a sea-change. On this view, the implications for European fiscal policy—though pointing in the direction of more coordination—would be less radical than some have suggested.

A second source of externalities in policymaking, and one which has received more attention in the literature (Sachs (1983), McKibbin and Sachs (1988), Canzoneri and Henderson (1989)), is the existence of incentives for strategic beggar-thy-neighbor policies in response to exogenous shocks. Consider the case of two identical countries, with objectives for price and output stability, subject to a simultaneous and symmetric adverse supply shock (a rise in energy prices, for example). If each takes the setting of the other’s policy instruments as given, and the exchange rate between them is fully flexible, it is straightforward to show that both countries have an incentive to pursue a combination of relatively expansionary fiscal policy and tight control of monetary growth: in so doing, each will hope to force up its exchange rate relative to the other, thereby dampening the effect of the supply shock on domestic prices by “exporting inflation.” ^12/13/ Of course, with both countries acting in the same way, the pressures on the exchange rate will cancel out and the only result will be an inefficient distortion to the collective fiscal-monetary mix. Coordination of policy, aimed at maximizing a joint objective function, would eliminate this inefficiency, since movements in the exchange rate would be seen to have only redistributive effects. The benefits from such coordination would be stronger the more prevalent are symmetric rather than asymmetric shocks, since it is clear that the conflicts over the exchange rate will be much reduced (or possibly, in the simplest models, eliminated) if disturbances in the two countries tend more often to be in opposite directions. ^14/

In the simple two-country world described here, a shift from floating rates to a fixed exchange rate regime is sufficient to eliminate the negative externalities deriving from unilateral changes in a country’s fiscal-monetary mix (see McKibbin and Sachs (1988) for formal proof). It would be tempting to infer from this that fiscal coordination becomes less, not more, necessary as the flexibility of exchange rates is reduced. However, more recent studies (Cohen and Wyplosz (1989); Canzoneri and Henderson (1989)) have shown the importance of taking into account interactions with third countries, or groups of countries.

For ease of exposition, suppose that Europe consists of France and Germany, and that the U.S. comprises the rest of the world. If France and Germany agree to fix the franc-deutsche mark exchange rate (floating jointly vis-à-vis the U.S. dollar), they eliminate the intra-European externality, as we have said, but create a new one: a fiscal expansion in, say, France (for given monetary growth) will now tend to push up the joint exchange rate against the dollar, generating spillover effects on Germany. Cohen and Wyplosz show that, here, the resulting inefficiency will—in contrast to the case of flexible exchange rates—be most severe when shocks affecting the European partners are asymmetric. In order to dampen the impact on its domestic inflation rate of an adverse supply shock, France might wish to see an appreciation of the European currency vis-à-vis the U.S. dollar and would tend to loosen fiscal policy to bring this about. If, at the same time, Germany was subject to a beneficial supply shock, the French action would aggravate the disturbance to the German economy (push inflation down further); countervailing fiscal measures by the Germany authorities would then set in train an inefficient and fruitless conflict over the joint exchange rate. ^15/ In the case of a symmetric shock, conflict would be much reduced—since both countries would want the exchange rate to move in the same direction—and would derive solely from the incentive to “free ride” (each wanting the other to make the necessary policy adjustment, resulting in insufficient adjustment overall, relative to the optimum).

Consideration of these stylized scenarios is therefore, a priori, inconclusive on the question of whether fiscal coordination is more or less necessary under fixed than under floating exchange rates. Incentives for inefficient strategic policy conflicts exist under both regimes, only in rather different forms. However, there are two reasons for believing that the potential problems might actually be less severe in a hard currency regime or monetary union than under present arrangements. First, it is widely held that the European economies are more closely integrated with each other than they are with the rest of the world. Since the spillovers from policy are directly related to the degree of integration between economies, this would imply that the externality being eliminated (the intra-European externality) might be more significant in practice than that being created (conflict over the common exchange rate vis-à-vis the rest of the world). Secondly, some empirical work in Cohen and Wyplosz (1989) ^16/ conforms with intuition in finding that disturbances to the French and German economies from the mid-1960s to the mid-1980s have been more strongly symmetric (and sustained) than asymmetric. ^17/ As we have seen, this would also tend to favor fixed over adjustable exchange rates in Europe, in terms of minimizing the costs of inefficient policy conflicts. While there is a clear need for more research in this area—both theoretical (looking at a wider range of shocks and more sophisticated models) and empirical—it appears that, at present, the case for enhanced coordination of fiscal policies as a necessary adjunct to monetary union is unproven.