II. France: Lessons from Past Fiscal Consolidation Plans1
A. Introduction: A Long History of Deficits
1. Over more than 30 years, France has incurred chronic fiscal deficits, and only a few attempts to revert this pattern were made. Of the total of four relatively sizeable consolidation plans (“Plan Barre” of 1976-80, “Virage de la rigueur” of 1983-86 2, the first five-year budgetary plan of 1994-97 prepared for the EMU accession process, and the Excessive Deficit Procedure of 2003-07 under the Stability and Growth Pact) this note will only focus on the two most recent ones. Earlier attempts were hampered by the absence of any budgetary planning that would span several years and resulted either in a quick reversal of the consolidation plan due to deteriorating macroeconomic conditions (the first attempt) or in a painful procyclical consolidation in which growth underperformance largely offset the structural consolidation effort (the second attempt).
France: Evolution of General Government Balance and Debt Ratios, 1970-10
Source: INSEE and IMF staff calculations.
2. As compared to past fiscal consolidation episodes in other large economies, the French episodes exhibit similar patterns in terms of timing and achieving the announced objectives. Germany and the UK had similarly timed early consolidation episodes (1976-78, 1980-84 and 1980-86, respectively) but, unlike France, they relied on multiyear plans since the outset. The panel chart below illustrates the sizeable deterioration of the fiscal stance in each of the six large economies in the early 90s, as a consequence of the global economic recession. This triggered a global fiscal consolidation attempt that started with Germany and the UK (1992), continued with France and the US (1993), and was joined by Canada (1994). In Japan, the return to positive, but rather tepid, growth delayed the start of consolidation until 1997; meanwhile, the public debt kept accumulating.
France: History of Fiscal Consolidations: General Government Revenues and Expenditures (left axis, in percent of GDP) and Real Growth Rates (right axis), 1976-09
Source: WEO database, various national sources, and IMF staff calculations.
3. In France, each level of government contributed to the trend deterioration of public finances, yet for different reasons. Of the three levels of public administration, as of 2008, central government spending accounted for about 38 percent of total spending, as compared to 42 percent for the social security administrations and 20 percent for local governments. For the latter, the well-contained deficits since the early 90s reflect the Golden Rule that restricts borrowing only to investment purposes (about one-quarter of expenditures). In fact, as a result of fiscal decentralization, further transfers of responsibilities, and additional sources of fiscal revenues, the share of the local government spending in GDP has doubled. The graph below highlights another French specificity, namely a sizeable correlation between the deficit patterns of the central government and the social security systems that have almost identical tax bases and are very sensitive to the cyclical conditions. Both the social security administrations and local governments have been operating under rather soft budget constraints over the period under consideration and have relied on ever increasing transfers of resources from the central government.
France: Fiscal Balance Trajectory by Subsector of the General Government, 1978-2008
4. Our main findings are in line with the existing literature on fiscal consolidations.3 Every consolidation attempt in France was triggered by: (i) a prior severe deterioration of fiscal ratios (marked by the dashed lines in the graph above); (ii) a potentially high reward from consolidation (or a high punishment for not achieving one) and hence, a strong political will; and (iii) a shared resolve for consolidation at all levels of government, including local governments and the social security administrations. The latter is best illustrated during the mid-90s episode, during which the central government’s efforts were accompanied by a period of declines in real spending on healthcare (which accounts for over 8 percent of GDP, or about a third of the total social security spending in France, and contributes about half to its deficit). As soon as healthcare spending started to increase again (2000-04), this time due to a double shock from the hospital tariff adjustment and the introduction of the 35-hour work week, the gains from the previous consolidation were partly undone. The activation of the Excessive Deficit Procedure of 2003-07 entailed a subsequent reform of the healthcare system that, in turn, helped during the 2004-06 consolidation.
B. A First Attempt at a Medium-term Consolidation in 1994-1997 4
The First Five-Year Budgetary Plan: The 1994 Reference Law for Public Finances
5. In response to the deterioration of the fiscal situation in the early 1990s, the French authorities adopted in early 1994 a five-year reference law on public finance control (“Loi d’orientation quinquennale relative à la maîtrise des finances publiques”). This law, which constituted a first attempt at medium-term budgeting, set the stabilization, and then reduction, of government debt as the main goal of fiscal policy. Meeting this goal was expected to create additional fiscal space, avoid crowding-out of private investment, and allow France to meet the recently adopted Maastricht criteria (i.e. an overall deficit of less than 3 percent of GDP and a public debt of less 60 percent of GDP) by 1997, in order to become a member of the European Monetary Union (EMU). In that sense, the reference law can be viewed as an implementation tool of the convergence program presented jointly by France and Germany to the European Council in November 1993.
6. In order to meet its overarching goal, the reference law set a number of quantitative medium-term objectives. These included a central government deficit target of 2.5 percent of GDP by 1997 and a limit on overall spending growth that implied a ½ percent per year reduction in real terms. In order to comply with the Maastricht criteria set for the general government, the law assumed a gradual improvement in the balances of local governments and social security administrations. Revenues (including privatization receipts) were expected to grow at 2.8 percent per year over 1995-97, on average, in line with nominal GDP. Any additional revenues from better-than-expected macroeconomic developments were to be saved or used to finance a reduction of the tax burden. In addition, draft budget laws were to be accompanied by reports presenting a trajectory of five-year budgetary projections that targeted a deficit reduction by about 0.5 percent of GDP per year. The companion budget law for 1994 was consistent with these objectives and stipulated a reduction in the number of civil servants, wage moderation, a substantial decline in investment spending, and a deceleration in transfers to local governments, as well as a slight decline of the tax burden, owing primarily to a reform of personal income taxation, and a rise in privatization receipts.
A Mixed Start: 1994-1995
7. While the 1994 deficit ended up in line with the budgetary objectives, the underlying spending was significantly higher than planned. Contrary to what was envisaged under the five-year reference law, rather than saving the additional revenues resulting from higher-than-expected growth, these were used to finance a number of additional expenditure items, including increases in school start allowances, social spending, labor market measures, and peace-keeping operations.
8. In 1995, the cycle of ex-ante expenditure restraint and ex-post spending overruns continued. The budget law provided for a further decline in the deficit, to 3.5 percent of GDP (including privatization receipts), based primarily on expenditure restraint and increases in excises. However, the new government that was formed after the May 1995 presidential elections passed a supplementary budget providing for additional spending in support of employment, social housing, and small- and medium-sized enterprises. These new outlays, along with the spending overruns observed during the first half of the year, were to be financed through (temporary) increases in taxes and some savings on non-priority spending. The weakening of economic activity during the second half of the year, as well as spending overruns, necessitated the adoption of a second supplementary budget in November. At this time, extra expenditure cuts and the mobilization of nontax revenue allowed the government to meet its deficit target.
Expenditure Restraint at Last: 1996–1997
9. The 1996 budget law was accompanied by an updated medium-term budget plan, reiterating the deficit objectives and commitments to future expenditure restraint. This updated plan covered the period of 1996-99 and, in line with the original plan, aimed at a gradual reduction (of about ½ percent a year) of the deficit from 4.1 percent of GDP (excluding privatization receipts) in 1995, to 3 percent in 1997 and 2 percent in 1999.5 Here too, the adjustment was to be achieved through keeping spending constant in real terms and maintaining the tax-to-GDP ratio unchanged.
10. Contrary to previous years, expenditure restraint did actually happen in 1996. Consistent with the medium-term objectives, the 1996 budget law included a number of measures aimed at reining in expenditure growth, including a further decline in capital spending; a stability pact with local governments aimed at moderating central government transfers; the freezing of the pay scale of public sector employees; and cuts in defense spending and social transfers. Nevertheless, to meet the Maastricht criterion on the deficit, the authorities had to take some last-minute measures in 1997.
11. Although the authorities accomplished a significant consolidation, the central government deficit was not reduced as planned. The deficit declined by 1.6 percent between 1993 and 1997, 0.4 percent less than initially planned. To meet the Maastricht deficit criterion in 1997, one-off payments from France Telecom amounting to 0.45 percent of GDP had to be mobilized. Over 1993-97, the underperformance primarily reflected higher-than-planned increases in primary spending of about 0.8 percent of GDP and lower-than-expected economic growth (which contributed to an increase in the deficit-to-GDP ratio of 0.2 percent) that were only partly offset by an increase in the tax burden, which helped to reduce the deficit by 0.5 percent of GDP. As a result of higher deficits and lower growth, the debt-to-GDP ratio increased more than planned, to 44 percent of GDP instead of 42 percent in 1997.
12. This fiscal consolidation episode illustrated the institutional limits to the first attempt at medium-term budgeting. The fact that the quantitative objectives of the reference law were not legally binding allowed for considerable discretion in the conduct of fiscal policy, as long as the overall deficit stayed on a declining path. In particular, in the absence of binding limits on spending growth, the government did not implement the expenditure restraint that it had committed to, especially when revenues were higher than envisaged.
The 1994-97 Consolidation: Targets and Outcomes
Source: WEO database, national sources, and IMF staff calculations
C. A More Systematic Approach to Medium-Term Budgeting: Consolidation Experiences Under the SGP
13. France issued twelve annual stability programs (SPs) over 1998-2009, of which six envisaged reductions in the overall fiscal deficit of more than ½ percent of GDP per year, all in the context of excessive deficit procedures (EDP). The first episode of a significant planned consolidation included five SPs submitted from 2003 to 2007 that aimed to reduce the deficit below 3 percent of GDP and terminate the EDP that lasted from June 2003 to January 2007. The second episode is ongoing and includes the most recent SP submitted in January 2010 that targets a large reduction in the overall fiscal deficit under the EDP opened in February 2009. Both planned consolidation episodes followed economic downturns and significant deteriorations in public finances.
France: General Government Balance: Stability Program Targets and Outcomes
Sources: Stability Programs, WEO database, and IMF staff calculations.
14. While deficit targets set in the successive SPs have been frequently missed, France succeeded in reducing the deficit below 3 percent of GDP and therefore terminating the first EDP.6 That said, the deficit dipped only slightly below 3 percent of GDP in 2005 thanks, in part, to a number of one-off receipts. It then remained just below the SGP ceiling during the global economic boom of 2005-07, before exceeding it again when the recent global recession put public finances under severe pressure.
The Composition of Adjustment
15. French SPs have focused on expenditure control, while revenue ratios have been targeted to remain stable or decline slightly. During the 2003–07 EDP, key adjustment measures included a legally binding zero real growth rule for central government spending, as well as significant health and pension reforms. Overall, the SPs of 2003–04 targeted reductions of the deficit by 0.6–0.7 percent of GDP per year, on average, under the reference macroeconomic scenario, while the SPs of 2005–07 included a medium term objective (MTO) of balancing the budget.
16. The deficit targets were attained, in part, thanks to favorable macroeconomic developments over 2003–07. To assess the relative contributions of macroeconomic developments and fiscal effort to the overall performance, implementation discrepancies relative to SP targets are decomposed into parts corresponding to cyclically–adjusted revenues and primary expenditures, interest spending, and cyclical balances. This decomposition shows that over 2003–07 cyclical balances often exceeded expectations thanks to strong GDP growth, while interest spending was typically below targets owing to low interest rates. Over the same period, cyclically–adjusted revenues tended to fall short of targets, while cyclically–adjusted expenditures were sometimes lower than targets (although not in the outer years of the SPs). In sum, favorable macroeconomic developments helped offset revenue shortfalls and expenditure slippages relative to SP targets in the outer years during this consolidation episode. Once the recent global recession set in, cyclical balances fell sharply below expectations, while primary expenditures rose significantly above targets, reflecting both automatic stabilizers and discretionary stimulus measures.7
A Method of Decomposing Implementation Discrepancies
Implementation discrepancies relative to SP targets are decomposed into components corresponding to cyclically–adjusted revenues and primary expenditures, interest spending, and cyclical balances, as follows:
where b, r, g, and i stand, respectively, for the overall balance, revenues, primary expenditures, and interest spending (all relative to GDP), superscripts refer to actual (A), target (T), cyclical (C), and cyclically–adjusted (S), and subscripts denote years. Cyclically–adjusted revenues and primary expenditures are expressed as:1
France: General Government Implementation Discrepancies: Actual–Target, 2004–09
Source: Stability Programs, WEO database, and IMF staff calculations.
A Closer Look at Macroeconomic Assumptions
17. French SPs have not relied on conservative growth assumptions, but they benefitted from the favorable macroeconomic environment over 2003–07.8 The SPs have typically included two scenarios: a cautious, or low, (reference) scenario with real GDP growth of 2¼–2½ percent per year and a favorable, or high, scenario with real GDP growth of 3 percent per year. Even the cautious (reference) scenario has tended to be somewhat optimistic in comparison with consensus forecasts (CF) published just a couple of months before the SPs and covering similar time horizons. However, as the consolidation under the 2003–07 EDP took place in a period of global economic boom and relatively high growth rates in France, the SP projections during that time turned out close to actual growth rates.
A Closer Look at Spending by Levels of Government
18. Real spending growth has tended to overrun SP targets, often reflecting slippages by local governments and social security systems. The French SPs have targeted average annual real spending growth at all government levels, including the general government (GG), the central government (CG),9 the local governments (LG), and the social security systems (SS). The general government real spending growth has typically exceeded targets, with overruns often reflecting slippages by local governments and social security systems, which account for about 20 percent and 45 percent of total expenditures, respectively. The central government—bound by the zero real spending growth rule—managed to reduce its spending in real terms over the course of SP2003 and SP2004, on average.10
19. Strong GDP growth over 2003–07 helped offset spending overruns. Although the real spending growth of the general government has tended to exceed SP targets, it remained close to the real GDP growth from 2004 to 2007 and even dipped below it in 2006.11 Therefore, favorable macroeconomic conditions during the global economic boom that preceded the recent recession helped contain expenditure ratios.
France: General Government Real Spending Growth: Stability Program Targets and Outcomes
Sources: Stability Programs, WEO database, and IMF staff calculations.
20. Past episodes of fiscal consolidation highlight the importance of having both a realistic assessment of how much effort would be required and strong motivation to achieve the target. The former was, on the one hand, facilitated by a multi–year approach but, on the other hand, put at risk by somewhat sanguine macroeconomic assumptions. Compared to the period of global economic boom previous to the recent recession, the need for realistic macroeconomic assumptions is more acute at a time when economic growth is fragile. The realism of macroeconomic forecasts could be enhanced if they were to be prepared or validated by an independent body. Another crucial element of a successful consolidation is political resolve to set binding spending limits for each level of government within the multi–year budget framework. Strong peer and market pressure can help bolster such political determination. To ensure the credibility of political commitment, consolidation strategies could be cemented into a formal fiscal rule across all levels of government. Together, credible political commitment and realistic macroeconomic assumptions would help to ensure sustainable public finances in France.
Champsaur, P. and J. P.Cotis, 2010, “Rapport Sur La Situation Des Finances Publiques”, report, available at lesrapports.ladocumentationfrancaise.fr/BRP/104000234/0000.pdf.
Cotis, J. P., 2007, “Fiscal Consolidation: Lessons from Past Experience”, Chapter IV of the OECD Economic Outlook, May, 2007
European Commission, 2005, “New and Updated Budgetary Sensitivities for the EU Budgetary Surveillance”, Information Note for the Economic and Policy Committee, DG ECFIN, September.
European Commission, 2007, “How to Stick to Medium-Term Budgetary Plans”, Part III of Public Finances in EMU 2007, European Economy No. 3/2007, DG ECFIN, pp. 149–192.
Fedelino, A., A.Ivanova, and M.Horton, 2009, “Computing Cyclically-Adjusted Balances and Automatic Stabilizers”, IMF Technical Notes and Manuals 09/05, Fiscal Affairs Department, May.
Girouard, N., and C.André, 2005, “Measuring Cyclically-Adjusted Budget Balances for OECD Countries”, OECD Working Paper No. 434, Economics Department, July.
Jonung, L., and M.Larch, 2004, “Improving Fiscal Policy in the EU: The Case for Independent Forecasts”, European Commission Economic Paper No. 210, DG ECFIN, July.
Moulin, L. and P.Wierts, 2006, “How Credible are Multiannual Budgetary Plans in the EU?”, proceedings of the Banca d’Italia workshop on fiscal indicators,30 March to 1 April.
Prepared by Edouard Martin, Irina Tytell, and Irina Yakadina. This note is part of a research project on past fiscal adjustment plans across countries conducted at the IMF’s Fiscal Affairs Department. The authors thank seminar participants at the Ministère de l’économie, de l’Industrie, et de l’Emploi and the IMF’s Fiscal Affairs Department for valuable comments and Pierre Ecochard and Samuel De Lemos Peixoto of the European Commission’s DG ECFIN for help with some of the data. Anastasia Guscina provided excellent research assistance.
The focus of this consolidation episode was on reducing the current account deficit to regain competitiveness.)
See Chapter IV of the OECD Economic Outlook 2007 and references therein.
In this section, “central government” refers to the state (“État”), which is covered by the budget laws, and does not include the other central government units (“Organismes d’administration central.”)
Contrary to the initial plan, which recorded privatization receipts in revenue above the line, the updated plan, in line with the Maastricht definition, recorded privatization receipts as part of financing.
Notably, this analysis does not take into account the effect of the past recession on the revenue elasticity.
Optimistic biases in growth assumptions that underlie fiscal projections have been identified in a number of countries under the SGP (see Jonung and Larch, 2004 and Strauch, Hallerberg, and von Hagen, 2004).
The central government is defined to include the State but not central government agencies.
The average annual real spending growth targets have been set for years T+1 to T+3 in most past SPs. The outcomes are computed for the corresponding years through 2009. Inter–governmental transfers are excluded from this analysis.
The real spending growth is obtained from nominal expenditures using CPI minus tobacco as a deflator.