This Selected Issues paper addresses the question of what policy changes in France are needed under European Monetary Union (EMU), as regards the role of fiscal policy in stabilizing the economy. The fiscal strategy over the past two and a half decades is reviewed, and, against this background, an assessment is offered concerning the role and scope of fiscal stabilizers in France under EMU. The main conclusions is that over the past two and a half decades, fiscal policy operated in a clear countercyclical way in France, but this reflected essentially the functioning of automatic stabilizers.

Abstract

This Selected Issues paper addresses the question of what policy changes in France are needed under European Monetary Union (EMU), as regards the role of fiscal policy in stabilizing the economy. The fiscal strategy over the past two and a half decades is reviewed, and, against this background, an assessment is offered concerning the role and scope of fiscal stabilizers in France under EMU. The main conclusions is that over the past two and a half decades, fiscal policy operated in a clear countercyclical way in France, but this reflected essentially the functioning of automatic stabilizers.

I. France: Fiscal Stabilizers Under EMU1

A. Introduction and Summary

1. In the recent period France’s economic strategy was governed by a logic of transition. The essential goal of fiscal policy was to satisfy the deficit criterion of the Maastricht treaty so as to ensure France’s participation in the monetary union. Now that this goal is completed, the time is ripe for thinking about a new “fiscal philosophy” stating what France expects from its fiscal policy in the new, permanent, EMU regime.

2. The European Economic and Monetary Union will change the framework in which national fiscal policies are determined in two essential respects. First, monetary policy will be determined with reference to the economic conditions in the euro zone as a whole, rather than targeted toward national objectives (including in this connection the use of the French franc exchange rate with the deutsche mark as an intermediate target). Second, the extent to which governments can resort to fiscal deficits will be limited by the 3 percent deficit ceiling of the Stability and Growth Pact (SGP). As a number of economists have noted, this deficit ceiling could limit the scope for fiscal stabilizers to operate precisely at a time when fiscal policy might become more important to stabilize the economy.2

3. This paper addresses the question what policy changes in France—if any—are needed under EMU, as regards the role of fiscal policy in stabilizing the economy. In the sections that follow, fiscal strategy over the past two and a half decades is reviewed; and, against this background, an assessment is offered concerning the role and scope of fiscal stabilizers in France under EMU. The main conclusions of the paper are as follows:

  • Over the past two and a half decades, fiscal policy operated in a clear counter-cyclical way in France, but this reflected essentially the functioning of automatic stabilizers. There was no systematic and deliberate attempt by the fiscal authorities to stabilize the economy over the cycle by discretionary interventions—indeed, such initiatives were at times procyclical, as, for example, in the economic upswing of 1989–90.

  • There is no strong argument, in the case of France, for making fiscal policy more countercyclical because of EMU. The argument based on the loss of monetary autonomy seems weak in the case of France, where monetary policy was already geared towards an external objective before EMU, and the other arguments that can be invoked are speculative and difficult to quantify. Hence, the main challenge ahead is to leave enough room for the automatic stabilizers to operate under the constraints of the SGP.

  • Given the constraints of the SGP, the full and effective operation of automatic stabilizers requires the public finances to be close to equilibrium on average. In particular, the general government structural deficit would need to be reduced to a level very close to balance to avoid a significant risk of a breach in the deficit limit that would not appear to fall under the “exceptional circumstances” clause of the Pact. Thus the operation of fiscal stabilizers provides some ground for balancing the public finances over the medium term—in addition to other considerations outside the scope of this paper, such as the need to reduce the burden of the public debt ahead of impending demographic pressures on the public finances.

4. The remainder of this paper is structured as follows. Section B provides some stylized facts on the broad trends and cyclical fluctuations in fiscal policy during 1970-97. Section C scrutinizes the operation of fiscal stabilizers over the past two decades, focusing in particular on the question of whether automatic stabilizers were supplemented or mitigated by discretionary interventions. Section D investigates the implications of EMU for the stabilizing role of fiscal policy. Section E assesses the level of the structural deficit that would allow fiscal stabilizers to operate under the SGP constraint.

B. Fiscal Policy During 1970–97: Some Stylized Facts

5. The last quarter of a century was marked by a steep increase in the share of general government expenditures in GDP, which approached 55 percent in the 1990s (Figure I.1). Most of this growth can be ascribed to an increase in social expenditures in the 1970s and early 1980s (Figure I.2), following important extensions of the social insurance system in the 1970s, as well as mounting unemployment. The growth of expenditures tended to slow down after 1983, a phenomenon which—though often associated with the shift to more austere policies (tournant de la rigueur) in France—was evident more generally in the EU. The revenue side exhibits the same broad trends, with the share of general government revenues in GDP increasing at a rapid pace until the early 1980s—mainly under the pressure of social contributions—and then stabilizing on a plateau around 50 percent (Figure I.3).

FIGURE I.1
FIGURE I.1

FRANCE: General Government Revenues and Expenditures

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, Datafund.
FIGURE I.2
FIGURE I.2

FRANCE: General Government Expenditures

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, Datafund.
FIGURE I.3
FIGURE I.3

FRANCE: General Government Revenues

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, Datafund.

6. The increase in revenues and expenditures was accompanied by a steady deterioration of the fiscal balance, which fell from a state of surplus in the early 1970s to a deficit in excess of 3 percent of GDP during most of the 1990s (Figure I.4). Under the cumulative effect of widening deficits, the level of the public debt surged—from about 20 percent to almost 60 percent of GDP between 1980 and 1997 (Figure I.5), and interest payments became an important component of public expenditures—accounting for about one-fifth of the rise in the share of public expenditures in GDP over the period.

FIGURE I.4
FIGURE I.4

FRANCE: General Government Balance

(In Percent of GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Sources: IMF, Datafund; and EC, European Economy.
FIGURE I.5
FIGURE I.5

FRANCE: General Government Consolidated Gross Debt

(In Percentage Points of GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: EC, European Economy.

7. The share of general government expenditures—and to a lesser extent revenues—in GDP was also significantly influenced by cyclical fluctuations in economic activity. The troughs of the business cycle—in 1975, 1985 and 1993—were systematically associated with increases in the ratio of expenditures to GDP, which were not completely offset in the following peaks. By contrast, the share of revenues in GDP seems to have been less responsive to the cycle. Overall, fiscal policy has behaved in a consistently countercyclical manner, associating periods of lower output with higher fiscal deficits (Figure I.6).

FIGURE I.6
FIGURE I.6

FRANCE: General Government Balance and Output Gap

(In Percent of GDP and Potential GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, World Economic Outlook.

8. The cyclical properties of fiscal policy can be investigated at a more disaggregated level by looking at the correlation of the components of general government revenues and expenditures with cyclical fluctuations in output. For this purpose, the methodology presented by Blanchard and Fischer (1989/1) was used to establish the correlation between innovations obtained from estimation of univariate ARIMA processes for the growth rate in real GDP and the real level of fiscal variables.3 Tables I.1 and I.2 report the results for general government revenues and expenditures, respectively. Each table gives the share of the fiscal component in GDP as of 1997, the correlation between the innovations in the fiscal component and the innovations in real GDP, and the coefficient from a regression of the innovation in the component on the contemporaneous innovation in real GDP. While the latter regression should not be interpreted as causal, the regression coefficient provides some indication how far particular fiscal components move with GDP.

Table I.1.

France: Co-movements in GDP and General Government Revenues

article image
Source: INSEE, Quarterly National Accounts.

Significant at the 10 percent level.

Significant at the 5 percent level.

Significant at 1 percent level.

In 1997.

Correlation between the residuals from ARIMA processes for GDP and the components of general government revenues (deflated by the GDP deflator). Annual data, from 1970 to 1997. A dummy variable for the period 1970–82, denoted by δ, was included when significant:

Δgdp = 1.59 x 10-2 + 0.306 Δgdp(-1) + e where gdp is the logarithm of real GDP.

Δggr = 2.34 x 10-2 + 2.13 x 10-2δ + e where ggr is the logarithm of the real level of general government revenues.

Δggt = 2.96 x 10-2 + e where ggt is the logarithm of the real level of taxes.

Δtva = 0.74 Δtva(-1) + e - 1.30 e(-1) where tva is the logarithm of the real level of VAT receipts.

Δipt = 5.08 x 10-2 - 0.25 Δipt(-1) + e where ipt is the logarithm of the real level of income and property taxes.

Δssc = 2.25 x 10-2 x3.80 x 10-2δ + e where ssc is the logarithm of the real level of social security contributions.

For social security contributions, with one lag in the innovation in real GDP.

Coefficient from a regression of the innovation in the component of general government revenues on the innovation in real GDP.

Table I.2.

France: Co-movements in GDP and General Government Expenditures

article image
Source: INSEE, Quarterly National Accounts.

Significant at the 10 percent level.

Significant at the 5 percent level.

In 1997.

Correlation between the residuals from ARIMA processes for GDP and the components of general government expenditures (deflated by the GDP deflator). Annual data, from 1970 to 1997.

Δst = 2.66 x 10-2 + 3.36 x 10-2δ + e - 0.38 e(-1) where st is the logarithm of the real level of social transfers.

Δec = 1.76 x 10-2 + 4.16 x 10-2δ + e + 0.64 e(-1) where ec is the logarithm of the real level of general government compensation of employees.

Δcoe = 3.10 x 10-2 + e where coe is the logarithm of the real level of general government consumption and other current expenditures.

Δce = 0.56 x 10-2 + e where ce is the logarithm of the real level of general government capital expenditures.

For wage, with one lag in the innovation in real GDP.

Coefficient from a regression of the innovation in the component of general government expenditures on the innovation in real GDP.

9. The results indicate, as might be expected, a significant positive correlation between output and the components of general government revenues. Taxes are very responsive to cyclical fluctuations in output, an effect that is more marked for direct taxes than indirect ones, presumably because of the progressivity in direct tax rates. The elasticity of VAT receipts is not significantly different from 1, as would be anticipated in the case of a tax that is roughly proportional to value added. Social security contributions are less significantly correlated with the cycle, which may be explained by their regressivity—resulting from ceilings on individual social contributions—and the fact that they are based mostly on labor income, which, like employment and wages, is less closely correlated with output than total income.

10. There is a significant negative correlation between output and general government primary expenditures, which reflects in large part the countercyclicality of social transfers. A number of components of social transfers are more or less directly responsive to the level of economic activity: unemployment benefits, but also other transfers linked to employment status or income levels, such as the minimum benefit (revenu minimum d’insertion) or family allowances (allocations familiales) (Bismut, 1997). Even pension benefits may be indirectly affected by cyclical conditions because of early retirement mechanisms which, in France, act in a sizeable measure as substitute to layoffs during economic downswings (Blanchet and Pelé, 1997). There is some evidence—although not statistically significant—that capital expenditures are procyclical, a result that may be due to a tendency of the government to respond to cyclical increases in its deficit by cutting or postponing public investment. The compensation of public employees is negatively correlated with output—a correlation that is not significant and subsumes opposite cyclical behaviors of the hiring and wage policies of the public sector. The number of employees in the public sector is significantly countercyclical (even though the low value of the elasticity coefficient shows this effect to be small) suggesting an attempt by the general government to offset cyclical variations in the level of private employment. On the other hand, the average real wage in the public sector is procyclical, which may reflect the general tendency of wages to grow more during economic upswings than during downswings.

11. Overall, the cyclical properties of fiscal policy seem to conflate a number of different mechanisms by which fiscal variables adjust—or are adjusted—to cyclical fluctuations in output. Some of these mechanisms, such as those involving the passive response of taxes and social contributions to cyclical fluctuations in income, clearly fall in the category of automatic stabilizers. Others—those related to the hiring and investment policies of the public sector, in particular—reflect the discretionary behavior of the fiscal authorities. Between the two lies a grey zone of “semi-automatic” stabilizers, which may not reflect a deliberate attempt by the authorities to stabilize the economy, but at the same time do not function mechanically as automatic stabilizers stricto sensu. Social benefits that are indirectly linked to unemployment, such as early retirement schemes, would be an example of this intermediate category. Another example is the sensitivity of public expenditures to macroeconomic variables, such as the real wage or interest rates, which are systematically correlated with the cycle through mechanisms that do not involve the active use of fiscal policy. While the estimates presented here point to the existence of such a “semi-automatic” component of fiscal stabilizers, they do not allow a conclusion on their statistical and economic significance.

12. In order to assess the options that will be open to the authorities in the operation of fiscal stabilizers under EMU, it is important to draw an operational borderline between the discretionary and automatic components of fiscal policy. This is a delicate exercise, but also a necessary one, and it is considered in the following section of this paper. The correlations and elasticities reported in Tables I.1 and I.2, while producing useful information on the cyclical properties of fiscal policy, cannot be used for that purpose because they mix the automatic and discretionary components and suffer from a simultaneity bias—output being partly endogenous to fiscal policy. However, they do indicate the need—when formulating medium-term fiscal objectives—to assess and factor in any desired latitude for “semi-automatic” responses to fluctuations in the economy.

C. The Operation of Fiscal Stabilizers

13. The traditional approach to the assessment of automatic stabilizers is to perform a decomposition of fiscal developments into structural and cyclical components, along the lines of staff analyses used in the World Economic Outlook, and similar analyses made by the IMF, the OECD, and the European Commission.4 While some of the assumptions underlying these estimates are questionable (Blanchard, 1990), they offer the advantage of being structural, in the sense that they are based on the cyclical elasticities of the different component of revenues and expenditures as they can be computed from the prevailing fiscal structure, and not from statistical correlations observed in the past. A next step in the analysis, therefore, is to review the operation of fiscal stabilizers in France over the past two decades in the terms of this conventionalized staff measure of automatic stabilizers.

14. Following the conventional approach, the observed fiscal balance is decomposed into two components, the structural balance and the cyclical balance:

bt=bts+btc

where the variables are expressed as ratios to GDP. The cyclical or automatic stabilizer component captures the built-in response of the government balance to cyclical output fluctuations. It is assumed to be proportional to the output gap (expressed in percentage points of GDP):

btc=αgapt

where coefficient α measures the cyclical responsiveness of the fiscal balance, that is, the percentage point change in the balance-to-GDP ratio caused by a 1 percentage point change in the output gap.5

15. The staff estimate of the cyclical responsiveness is 0.6 for France, broadly in line with the estimates used by the OECD and the EC.6 The assumptions underlying this estimate are broadly consistent with the cyclical behavior of fiscal variables documented in the previous section. The tax and social contribution components of public revenues are assumed to be automatically responsive to cyclical variations in output, with elasticities that depend on the progressivity or degressivity of tax rates and the cyclical responsiveness of the tax base. The aggregate responsiveness of government revenues is then computed as the average of the different tax elasticities, weighted by the share of each tax in total revenues. On the expenditure side, the only component that is assumed to be automatically responsive to the cycle is unemployment benefit payments. This assumption does not take into account other social transfers that may be sensitive to the cycle as well as other mechanisms involved in the “semi-automatic” stabilizers, but it is not clear, on the basis of the estimates presented above, that ignoring these effects biases the estimate of the cyclical component of fiscal policy to an important extent.

16. The structural—or cyclically adjusted—balance is the hypothetical value that the balance would take if the output gap was zero. It is estimated as a residual, by subtracting the automatic stabilizer component from the observed balance:

bts=btαgapt

The structural balance reflects changes in interest payments on the public debt as well as new fiscal measures. The evolution of the structural balance net of interest payments, the primary structural balance, gives a rough indication of structural changes in fiscal policy.

17. Figure I.7 shows the evolution of the primary structural balance and the output gap over the past two decades. If the fiscal authorities were systematically attempting to reinforce the automatic stabilizers by discretionary fiscal interventions, the primary structural balance and the output gap would tend to move in the same direction. Conversely, if the government was trying to insulate the deficit from cyclical influences a negative correlation would appear between the two variables. There is no visible stable pattern of correlation in Figure I.7, a finding that is confirmed by regressing the change in the primary structural balance on the change in the output gap:

Δbtps=0.215(0.144)Δgapt
FIGURE I.7
FIGURE I.7

FRANCE: General Government Primary Structural Balance and Output Gap

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, World Economic Outlook.

The regression coefficient is not statistically different from zero at standard levels of confidence. Thus, a crude analysis of the data attributes the cyclical properties of fiscal policy to the operation of automatic stabilizers, which do not seem to have been significantly augmented—or mitigated—by discretionary interventions over the past two decades.

18. The results of these regressions should be interpreted with care, as fiscal policy has undergone several changes in regime since the 1970s. The beginning of the period was characterized by two Keynesian reflationary experiments that were quickly reversed under the pressure of their consequences: first, the reflationary measures adopted by the Chirac government in 1975, followed by the austerity plan formulated by the Barre government in 1978; and second, the expansionary fiscal policy of the first socialist government in 1981, followed by a shift to fiscal rigor with the Fabius government in 1983. Another regime break occurred in the last few years of the sample period, when a sizeable fiscal retrenchment—involving mostly the revenue side—was implemented by the Juppé government to satisfy the convergence criteria of the Maastricht treaty.

19. The period from the mid-1980s to the mid-1990s may be more revealing of the true preferences of fiscal authorities than the rest of the sample period since, on the one hand, it was free of the reflationary Keynesian experiments of the late 1970s and early 1980s, and on the other hand, it was not yet constrained by the Maastricht treaty. Regressing the change in the primary structural balance on the change in the output gap over 1983–95 gives:

Δbtps=0.218(0.109)Δgapt

As before the coefficient is negative, but it is now different from zero at the 10 percent level of confidence, suggesting a significant procyclical behavior of the fiscal authorities during this period.7 Thus, it may be interesting to look in more details at the forces and constraints that were important in shaping fiscal policy outcomes and its adjustment to the cycle from 1983 to 1995.

20. The period from 1983 to 1987 was dominated by fiscal consolidation, in a context of stable but slightly lower than potential output. The central government budget for 1985 was the first one to endorse explicitly the political commitment in 1983, under the Mitterrand presidency, to lower the share of taxes and social contributions in GDP; and the budget of 1986 had similar orientations, with the declared objective of containing central government budget deficit under 3 percent of GDP. On the expenditure side, reductions dominated, particularly as a result of substantial cuts in transfers to social insurance funds, as well as in subsidies and capital transfers to nationalized enterprises. The new conservative government that came into office after spring 1986 general elections was keen to reduce even further the share of the government deficit and public debt in GDP. One major move consisted of privatizing a number of state-owned companies and using the proceeds of the sale to reduce central government debt.8

21. Fiscal policy took a new turn in 1988, when the structural deficit embarked in a path of relative deterioration. This change in fiscal orientations finds its origin in a string of expenditure-increasing measures initiated by the government of Prime Minister Rocard—and pursued by its socialist successors—including notably the renegotiation of public employees’ wages, the institution of a new minimum income scheme (revenu minimum d’insertion), and the adoption of a major program in favor of higher education. At the same time, the strong recovery of 1988-1990 allowed the authorities to reduce a number of taxes—the personal and profit tax rates, in particular, were lowered—without visible consequences for the public deficit. The extent of the underlying deterioration of the public finances, however, became blatant in the economic slowdown of 1992-93, when the general government deficit came close to 6 percent of GDP. The fiscal policy of the conservative government of Prime Minister Balladur that took office in the spring of 1993 was affected by its changing perceptions of the extent of the recession. The government first took a number of measures to stop the degradation of public financial conditions—excise taxes, the Contribution Sociale Généralisée and employees’ social contributions were raised—but then shifted to a more expansionary stance, as the recession appeared even deeper than anticipated, with a one-off measure to the benefit of private firms. As a result of these different measures in 1995 the structural primary balance was about at the same level as in 1991.

22. Overall, the historical record does not suggest that the French fiscal authorities resorted to discretionary interventions in order to stabilize the economy, even when they were free to do so. Between 1983 and 1994 the determination of fiscal policy was subject to a complex set of influences, in which the desire to stabilize output, without being absent from policymakers’ preoccupations, were often outweighed by other considerations. In the years leading up to the cyclical peak of 1989-90 the structural stance of fiscal policy was in fact procyclical—the cyclical improvement in the public finances being used to develop the supply of public goods and reduce taxes. The recession certainly became an important concern of the French fiscal authorities in 1993, but their ability to stimulate the economy using fiscal policy was constrained at that time by the size of the deficit—and it is difficult to conjecture how fiscal policy would have been used in the absence of this constraint.

23. While the record of this subperiod does not demonstrate a clear counter-cyclical pattern in discretionary policy actions (a conclusion consistent with the findings of Section B above), one important lesson of these years is the danger of failing to aim for sufficiently ambitious fiscal goals during a period of economic recovery, when the actual fiscal balance performs deceptively favorably, and thus of constraining the scope for policy to operate freely in a subsequent downturn.

D. Should Fiscal Policy Become More Countercyclical Under EMU?

24. EMU involves a major change in regime, and taking the statistical regularities observed in the past as a benchmark to analyze the future raises a Lucas-type critique. In order to address this critique one needs to consider the different ways in which the economic environment of fiscal policy might be changed by EMU. A widespread argument is that the loss of their national monetary autonomy will induce countries adopting the common currency to shift the burden of stabilization to fiscal policy. EMU may also affect more broadly the nature of the shocks hitting the economy, and the response of economic agents to them.

25. It is often argued that because monetary policy is no longer available as a national policy tool, fiscal policies should become more countercyclical under EMU. This argument, however, assumes that countries exploit their monetary independence in the period leading up to the monetary union—for example, in order to support economic activity at times of a significant output gap. This was not the case in France, which—under the auspices of the “franc fort” policy—has kept its EMS central parity unchanged since 1987, and has maintained the objective of holding the French franc/deutsche mark rate within a narrow band. French monetary policy was geared toward an external objective for more than a decade before the transition to EMU, and the completion of the monetary union should not significantly shift the allocation of the stabilization burden from monetary to fiscal policy.

26. In fact, monetary policy may, if anything, become more tuned to the French economic conditions under EMU, not less. In the past decade, French monetary policy was determined by German interest rates plus shocks to the credibility of the franc/mark parity.9 Under EMU, credibility shocks will disappear, and monetary policy will be determined with reference to the economic conditions in the whole euro zone. As Figure I.8 shows,10 the French cycle has in the past been more closely correlated with the euro zone cycle than with the German one—and to that extent, monetary policy may become more in phase with the French cycle under EMU than before.

FIGURE I.8
FIGURE I.8

FRANCE: Output Gaps in France, Germany, and The Euro Zone

(In Percent of Potential GDP)

Citation: IMF Staff Country Reports 1998, 132; 10.5089/9781451813494.002.A001

Source: IMF, World Economic Outlook.

27. Second, EMU may affect the nature of the shocks hitting the economy—a question that has been addressed extensively, if not conclusively, by the literature on EMU as an Optimum Currency Area (Eichengreen, 1993). A common currency will obviously remove the shocks to relative prices that were previously made possible by devaluations of the currencies of the trading partners concerned. Closer international trade with other members of the union may also significantly affect the nature of national business cycles in a way that could result, from a theoretical point of view, in either tighter or looser correlations of national business cycles. On the one hand, trade integration may favor the specialization of countries in the goods in which they have a comparative advantage, making them more sensitive to asymmetric sector specific shocks (Krugman, 1993). On the other hand, most of the European trade is intra-industry, rather than being based on specialization, so that closer integration might lead to more diversified portfolio of industries in each country, and less asymmetry in national shocks. Which effects will dominate is an empirical matter. The study of Frankel and Rose (1997) lends support to the view that closer trade integration will make shocks more symmetric, not less, in the euro zone.

28. The Lucas critique also suggests considering carefully the way the behavior of economic agents may be affected by the change in policy regime associated with the monetary union. Some authors, like Cotis et al (1998), have argued that households might become less responsive to fiscal impulses, as the SGP will make them more conscious of the future tax consequences of current fiscal policy. In this less Keynesian environment, fiscal policy would have to become more active to obtain the same degree of output stabilization. On the other hand, it has been argued that under EMU the working of European labor markets may improve in ways that make stabilizing fiscal policy less necessary. The common currency may make labor more mobile internationally, and the impossibility to address national unemployment problems using monetary policy will increase the pressure on policymakers to make labor markets more flexible. Whether or not the monetary union will produce these effects is difficult to predict, and quantifying them seems even more difficult.

29. To summarize, the classical argument for making fiscal policy more countercyclical under EMU—the loss of monetary autonomy—seems rather weak in the case of France, and the other arguments that can be invoked are highly speculative and difficult to quantify. Overall, it would appear that the objective of leaving enough room for the automatic stabilizers to operate may provide a reasonable initial basis for assessing the structural goal for fiscal policy under EMU.

E. What is the Size of the Structural Deficit That Would Allow the Fiscal Stabilizers to Operate Effectively Under the Stability and Growth Pact (SGP)?

30. The SGP clarifies the provisions of the Maastricht Treaty regarding the 3 percent reference value for budget deficits and the Excessive Deficit Procedure.11 Countries that breach the deficit limit may be subject to an excessive deficit finding leading to financial sanctions if the breach is not corrected in time, but the Pact also allows significant scope for discretion in the application of the Excessive Deficit Procedure. In particular, a country would not be subject to an excessive deficit finding if the breach of the 3 percent deficit limit is small, temporary, and can be attributed to “exceptional circumstances.” These exceptional circumstances are defined to encompass unusual events outside the control of the member state in question, or a severe economic downturn. While the Pact defines a severe economic downturn as an output decline of 2 percent or more, it also allows member states to argue that smaller output declines can be considered exceptional, notably on the basis of evidence concerning the abruptness of the slowdown or cumulative loss in output relative to past trends. Countries have agreed not to take advantage of this provision for annual output declines of less than ¾ percent. Further room for discretion is provided by the clause that countries subject to an excessive deficit finding can be given an extended period of adjustment if there are “special” circumstances, which the Pact leaves undefined.12

31. What is the size of the structural deficit that would allow fiscal stabilizers to operate effectively under the SGP? This question can be approached first from a retrospective point of view, and reformulated in terms of the level of the structural deficit that would have prevented excessive deficits in the past (Buti, Franco and Ongena, 1997). This, of course, involves assessing past policy ex post in terms of a framework and incentive structure that did not then exist—inevitably a somewhat artificial procedure. This said, the only period in which the deficit limit would have been a serious concern for France appears to be 1992-96, when, for five years in a row, the deficit was markedly larger than 3 percent of GDP (Table I.3).

Table I.3.

France: Fiscal Developments, 1992–96

(In percent)

article image
Sources: World Economic Outlook, and Fund staff calculations.

32. Hypothetically, had the SGP then applied, France would likely have been subject to an excessive deficit finding in 1993 on the basis of the breach in the deficit limit in 1992, if indeed this was judged to be large, unlikely to be temporary, and not justified by a recession or other exceptional circumstances. If corrective action based on the Council’s recommendations had not then been taken, a financial penalty, in the form of a non-interest bearing deposit of 0.28 percent of GDP, would have been applied to France by the end of the year. While the French government might have invoked special circumstances in 1994, on the basis of the 1993 recession, the Council, hypothetically, would have been entitled to increase the deposit by 0.26 percent of GDP, and convert it into a fine in 1995.13 This scenario assumes that France would have made no fiscal effort to meet the 3 percent deficit limit.

33. The usefulness of this example is that it provides a practical illustration with which to assess the level of the structural deficit that would have preserved France from an Excessive Deficit Procedure in the economic downswing of 1992–95—in other words, the maximum structural deficit that would have kept the actual deficit below 3 percent from 1992 to 1995, except in 1993, where the recession might justify—temporarily—a slightly higher deficit limit. As the last row of Table I.3 shows, a structural deficit of 1 percent of GDP would have been just sufficient for France to escape an excessive deficit finding. This figure is slightly higher than the 0.8 percent estimate derived in IMF (1998b) under the assumption that the deficit ceiling should have been strictly satisfied also in 1993.

34. Retrospectively, it is always possible to specify a level of the structural deficit that would just have prevented a breach in the deficit limit—but the result will hinge on the historical accidents of the period under consideration, here the 1993 recession. Prospectively, however, uncertainties surrounding future output developments make it more difficult to determine a given level of the structural deficit to ensure that the deficit limit will not be breached. The choice of a lower structural deficit, from this point of view, may be interpreted as an insurance against the risk of excessive deficit—an insurance, however, that does not cover the most extreme output fluctuations and entails some cost in terms of fiscal adjustment. How can one characterize, in the case of France, the tradeoff between the level of the structural balance and the risk of excessive deficit?

35. The frequency with which the deficit limit will be violated depends on the long-run statistical distribution of the output gap. Given a constant structural deficit bs the deficit limit is violated when the output gap falls below a threshold level, gap*, defined by:

bs+αgap*=3

and the frequency of breaches is given by:

p=F(gap*)=F((3+bs)/α)

where F(.) denotes the long-run (unconditional) cumulative distribution function of the output gap.

36. Over 1970–97 the output gap had a mean not significantly different from zero and a volatility of 2.32 percent; a Kurtosis test does not reject that it was normally distributed. While these statistics may give an imperfect measure of the long-run distribution of the output gap—in particular in view of the short duration of the sample period over which they are estimated—they nevertheless allow some empirical content to be given to the relationship between the level of the structural deficit and the frequency of breaches in the deficit limit. Table I.4 gives the frequency of breaches, p, for different levels of the structural balance, bs assuming that the past behavior of the output gap reflects its long-run distribution. As expected, the frequency of breaches in the deficit limit increases with the level of the structural deficit. It is noteworthy that a structural deficit of 1 percent of GDP, which was found to be sufficient to prevent an excessive deficit finding in the retrospective application of the SGP, now implies that the deficit limit is breached with a frequency of 7.5 percent, or every thirteen years in average. It is not clear that these breaches could be justified, as a rule, by the exceptional circumstances clause of the SGP.

Table I.4.

France: Frequency of Breaches in the Three Percent Deficit Limit: Univariate Estimates

article image
Sources: World Economic Outlook, and Fund staff calculations.

Unconditional probability of a general government deficit higher than 3 percent of GDP based on the univariate statistical distribution of the output gap.

37. The exceptional circumstances clause of the SGP applies automatically in case of an exceptionally severe recession involving a negative growth rate of 2 percent or more, but governments breaching the deficit limit are entitled to invoke it for milder recessions with growth rates between -2 and -¾ percent. This might seem to imply that the admissible threshold for the frequency of breaches in the deficit ceiling can be made equal to the frequency of recessions. This conjecture is not completely correct, however, because there is no guarantee that the largest deficits systematically coincide with recessions. If actual growth is lower than potential growth several periods in a row the cyclical deficit may become large even though no recession is observed, and conversely a recession may not lead to a large cyclical deficit if the economy was initially above potential.

38. Thus, the determination of the optimal structural deficit must be based on a model of the co-movements of the growth rate and the cyclical deficit. In the absence of obvious theoretical restrictions on these co-movements, we estimated a bivariate VAR for the output gap and the real GDP growth rate including one lag of the endogenous variables and a constant, and then used Monte Carlo simulations to assess the extent to which breaches of the 3 percent deficit limit could be justified by exceptional circumstances.14 Table I.5 reports the frequency—in terms of the average number of years per century—with which the 3 percent deficit limit is breached, as well as the frequency with which these breaches are associated with recessions that might allow the authorities to invoke exceptional circumstances.15 The difference between these two variables is the frequency of “unjustifiable” deficits, which are presumed to trigger excessive deficit findings.

Table I.5.

France: Frequency of Breaches in the Three Percent Deficit Limit

article image
Sources: World Economic Outlook, and Fund staff calculations.

Average number of years per century in which the general government deficit is higher than 3 percent of GDP.

Average number of years per century in which the general government deficit is higher than 3 percent of GDP and the growth rate in real GDP is below -¾ percent.

39. As Table I.5 shows, the frequency of unjustifiable excessive deficits increases with the size of the structural deficit. If the public finances are structurally in equilibrium, an unjustifiable excessive deficit would occur less than once per century in average. On the other hand, if the structural deficit amounts to 1 percent of GDP, the breaches in the deficit limit—which occur about every thirteen years in average—can be justified by a recession less than one-fourth of the time, implying, hypothetically, that France would be subject to an excessive deficit finding about every twenty years in average. As a linear interpolation of the values given in Table I.5 shows, the structural deficit needs to be lower than 0.1 percent of GDP for the frequency of unjustifiable excessive deficits to remain below 1 percent. More generally, keeping the risk of unjustifiable excessive deficits at a low level requires the finances of the general government to be structurally close to balance.

40. As noted above, these calculations are based on past fiscal performance, under a policy regime different from that which will prevail under EMU. Nonetheless, they appear to provide some interesting and useful insights concerning appropriate goals for fiscal policy in light of the constraints set out in the SGP. At the same time, in framing such objectives, two considerations underscored in earlier sections of this paper should doubtless be borne in mind when setting and implementing medium-term fiscal goals: (1) the need to factor in any desired use, in the future, of the “semi-automatic” stabilizers that operated at times in the past; and (2) the importance of forming a prudent assessment of the cyclically adjusted fiscal position at times when the actual balance is favorably affected by strong economic activity. More broadly, of course, an important lesson of recent years is the scope for structural reforms in labor and product market policies to enhance the flexibility of the economy in complementary ways, so that the burden on fiscal policy in this regard is more limited.

REFERENCES

  • Artus, P., 1998, “Le Pacte de Stabilité Est-il un Accord Efficace?,Revue d’Economie Financière, Vol. 45, pp. 15365.

  • Bismut, C., 1997, “Trends and Cycles in Government Expenditure in France: is Growing Government Share Reversible?,Cahier de Recherche No. 1997–02 (CAPE, Université de Montpellier 1).

    • Search Google Scholar
    • Export Citation
  • Blanchard, O.J., 1990, “Suggestions for a New Set of Fiscal Indicators,OECD Department of Economics and Statistics Working Paper No. 79 (Paris: OECD).

    • Search Google Scholar
    • Export Citation
  • Blanchard, O.J., and S. Fischer, 1989, Lectures on Macroeconomics (Cambridge, MA: MIT Press).

  • Blanchet, D., and L.-P. Pelé, 1997, “Social Security and Retirement in France,NBER Working Paper No. 6214 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Buti, M., D. Franco, and H. Ongena, 1997, “Budgetary Policies during Recessions - Retrospective Application of the “Stability and Growth Pact” to the Post-War Period,EC Economic Papers, No. 121, European Commission, Directorate-General for Economic and Financial Affairs.

    • Search Google Scholar
    • Export Citation
  • Commission of the European Community, 1995, “Technical Note: the Commission Services’ Method for the Cyclical Adjustment of Government Budget Balances,European Economy, No. 60, pp. 3591.

    • Search Google Scholar
    • Export Citation
  • Cotis, J.-P., B. Crepon, Y. L’Horty, and B. Meary, 1998, “Les Stabilisateurs Automatiques Sont-ils Encore Efficaces? Le Cas de la France dans les Années Quatre-Vingt-Dix,Revue d’Economie Financièrs Vol. 45, pp. 95118.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B., 1993, “European Monetary Unification,Journal of Economic Literature, Vol. 31, pp. 13211357.

  • Eichengreen, B., and C. Wyplosz, 1998, “The Stability Pact: More than a Minor Nuisance?,Economic Policy, Vol. 26, pp. 67113.

  • Frankel, J., and A. Rose, 1997, “The Endogeneity of the Optimum Currency-Area Criteria,Swedish Economic Policy Review 4, pp. 487512.

    • Search Google Scholar
    • Export Citation
  • Giorno, C., P. Richardson, D. Roseveare, and P. Van den Noord, 1995, “Potential Output, Output Gaps and Structural Budget Balances,OECD Economic Studies, No. 24, pp. 167209 (Paris, OECD).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 1993, “Structural Budget Indicators for the Major Industrial Countries,” in World Economic Outlook, Annex I, pp. 99103 (Washington: International Monetary Fund, October).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 1998a, “Fiscal Policy under EMU,Germany—Selected Issues and Statistical Appendix, SM/98/209, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 1998b, “Medium-Term Framework for Fiscal Policies,” in Economic Policy Challenges Facing the Euro Area and the External Implications of the Euro—Annex I, EBS/98/134, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Krugman, P., 1993, “Lessons of Massachusetts for EMU,” in The Transition to Economic and Monetary Union in Europe, ed. by F. Giavazzi and F. Torres, pp. 24161 (New York: Cambridge University Press).

    • Search Google Scholar
    • Export Citation
  • Le Cacheux, J., 1994, “Fiscal Policy in France: Assessing the Possibilities for Reducing Budget Deficits in the EMU Perspective,European Economy, Vol. 3, pp. 185218.

    • Search Google Scholar
    • Export Citation
  • Levy, J., and I. Halikias, 1997, “Aspects of the Monetary Transmission Mechanism under Exchange Rate Targeting: The Case of France,IMF Working Paper No. 97/44 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
1

Prepared by Olivier Jeanne (RES).

3

This approach thus offers a broader insight into cyclical fluctuations in the fiscal position by comparison with the conventional analysis of automatic stabilizers used by the staff, which is presented in Section C below.

4

The methods used by the IMF, the OECD and the EC to construct these estimates are presented in IMF (1993), Giorno et al (1995) and Commission of the European Community (1995), respectively.

5

For expositional simplicity, lags in the response of the balance to the cycle are ignored, as are the (small) changes in coefficient a resulting from the evolution of the fiscal structure over time.

6

The OECD estimate is the same as that of the IMF staff. The Commission estimate is slightly lower (0.5), because of small differences in the computation of the tax elasticities.

7

IMF (1998a) finds that in Germany procyclical discretionary interventions appear to have largely offset the operation of automatic stabilizers since the end of the 1970’s.

8

A more detailed account of fiscal policy during this period may be found in Le Cacheux (1994).

9

For a discussion, see Levy and Halikias, IMF Working Paper No. 97/44.

10

Figure I.8 depicts the output gap in France, Germany and the Euro zone. The index for the latter was computed as a GDP-weighted average, in which Finland was excluded because of the lack of data.

11

The SGP was agreed at the June 1997 meeting of the Council in Amsterdam. It consists of Council Regulation (EC) No. 1466/97 on “the strengthening of the surveillance of budgetary positions and the surveillance and coordination of economic policies,” Council Regulation No. 1467/97 on “speeding up and clarifying the implementation of the excessive deficit procedure,” which were adopted on July 7, 1997, and the Resolution of the European Council of 17 June 1997 on the SGP. Some analysis of the Pact implications is provided in Eichengreen and Wyplosz (1998), Buti, Franco and Ongena (1997) and IMF (1998b).

12

The normal timing is for a finding to be made in the spring of the year following that in which the excessive deficit arose, and then for measures to be adopted in that year (consistent with the Council’s advice) to correct the deficit in the next year.

13

The first year of the breach the amount of the deposit comprises a fixed component equal to 0.2 percent of GDP and a variable component equal to one tenth of the difference between the deficit as a percentage of GDP in the preceding year and the reference value of 3 percent of GDP. Each following year, the additional deposit, if it is decided, is computed according to the same formula as the variable component in the first year.

14

The estimated VAR was simulated over periods of one hundred years one thousand times, and each time the series for the fiscal balance was computed for different levels of the structural balance.

15

To simplify the illustration, the cut-off period of an output decline of ¾ percent of GDP has been applied mechanically, although this is not the intent of the SGP. The frequency of breaches reported in the first column of Table I.5 is slightly different from that reported in Table I.4 because it is based on a bivariate—as opposed to univariate—estimate of the long-run distribution of the output gap.

France: Selected Issues
Author: International Monetary Fund
  • View in gallery

    FRANCE: General Government Revenues and Expenditures

    (In Percent of GDP)

  • View in gallery

    FRANCE: General Government Expenditures

    (In Percent of GDP)

  • View in gallery

    FRANCE: General Government Revenues

    (In Percent of GDP)

  • View in gallery

    FRANCE: General Government Balance

    (In Percent of GDP)

  • View in gallery

    FRANCE: General Government Consolidated Gross Debt

    (In Percentage Points of GDP)

  • View in gallery

    FRANCE: General Government Balance and Output Gap

    (In Percent of GDP and Potential GDP)

  • View in gallery

    FRANCE: General Government Primary Structural Balance and Output Gap

  • View in gallery

    FRANCE: Output Gaps in France, Germany, and The Euro Zone

    (In Percent of Potential GDP)