This Selected Issues paper applies a range of methods to assess Italy’s competitiveness gap versus other euro area members. All indicators point to a clear erosion of competitiveness in recent years, with related market share losses. Nonetheless, the estimated competitiveness gap, while appreciable, still remains quantitatively contained. The results suggest that the unwinding of the competitiveness gap will require reforms that boost productivity growth and continued wage moderation, as well as an adjustment by export firms in Italy.

Abstract

This Selected Issues paper applies a range of methods to assess Italy’s competitiveness gap versus other euro area members. All indicators point to a clear erosion of competitiveness in recent years, with related market share losses. Nonetheless, the estimated competitiveness gap, while appreciable, still remains quantitatively contained. The results suggest that the unwinding of the competitiveness gap will require reforms that boost productivity growth and continued wage moderation, as well as an adjustment by export firms in Italy.

II. Public Spending in Italy: A Crescendo6

Objective. To examine the determinants of the increase in Italy’s primary spending ratio over the past decade, including from a cross-country perspective.

Results. While Italy was subject to “exogenous” spending pressures, some of which can be inferred from EU-wide trends, its record of persistent slippages indicates country-specific policy shortcomings and missed opportunities.

Policy recommendations: (1) accord top priority to expenditure control, focusing on durable, efficiency-enhancing reforms in key spending areas; (2) formulate a rules-based framework to establish a track record of achieving annual budget targets; and (3) leverage growth-enhancing reforms and episodes of cyclical strength to help reduce the spending ratio.

1. Italy’s public finances have been its chief economic vulnerability over the past quarter-century. Chronic fiscal deficits, on average 11 percent of GDP in the 1980s, used to generate inflationary pressures and, at times, currency crises. Concurrently, the public debt ratio doubled between 1980 and mid-1990s, topping 120 percent, making Italy’s debt burden among the highest in the world. As the interest bill alone reached 12 percent of GDP in the early 1990s, Italy seemed mired in a vicious circle of exploding debt and deficit dynamics.

2. In the mid-1990s, fiscal consolidation and pension reforms seemed to break the vicious circle. After a sobering 1992 crisis, a major fiscal effort, helped by the Euro-qualification-related decline in interest rates, cut the deficit from 10 percent of GDP in 1993 to less than 3 percent in 1997, yielding a primary surplus of 6½ percent of GDP. The public debt ratio, after peaking in 1994, fell gradually but steadily, also reflecting privatization proceeds. And substantial pension reforms of 1992 and 1995 made Italy’s long-term aging spending projections more benign than in many other EU countries.7 The Maastricht requirements also seemed to catalyze strengthened fiscal discipline, both directly and through some, if limited, enhancements to the fiscal policy framework.

A02ufig01

Italy: Fiscal Developments, 1980–2005

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

3. But this consolidation was rolled back by underlying spending pressures. After bottoming out at under 1 percent of GDP in 2000, due in part to one-off factors, the deficit breached the key 3 percent limit the following year, and topped 4 percent in 2005, causing the decline in the public debt to reverse, while the primary surplus fell sharply. From an accounting perspective, the deficit was driven mainly by primary spending, whose growth in Italy outpaced that of other large euro area countries.

A02ufig02

Change in Primary Spending 2000–2005

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

A. Anatomy of Italy’s Deficit Slippages

4. Italy’s absolute and comparative performance under the EU’s Stability and Growth Pact (SGP) slipped progressively. While initially the original SGP’s 2003 difficulties occurred against the background of deficits run by Germany and France, Italy’s more favorable fiscal position at the time proved illusory, as its first-reported deficits had been systematically underestimated, and partly achieved via one-offs. Thus, while the first breach of the 3 percent threshold in Italy became apparent in 2005, it was only then revealed that the country had violated the ceiling back in 2001, earlier than Germany and France. By 2005, with the reduced scope for one-offs as major tax amnesties expired, even Italy’s first-reported deficits exceeded those of its key comparators.

A02ufig03

Overall Balance in Large Euro-area Countries

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

5. Much of the EU-wide analysis of the SGP’s problems centered on growth shortfalls relative to forecasts. The authorities of noncompliant countries claimed they had fulfilled original budget plans and that the shortfalls in headline fiscal outcomes were caused by unexpected growth slowdowns (though the EC, based on its readings of structural balances, argued the opposite (see Jonung and Larch, 2005)). These authors provided evidence to bridge the two views, concluding that the cyclically adjusted balance may fall short of the targeted ratio even if expenditure plans are fully implemented (in levels), because of the effect of the lower potential growth on the denominator of the spending ratio. Such growth-related considerations played an important role in the design of the SGP’s 2005 reform.

6. But growth shortfalls played only a subordinate role in Italy’s failure to meet its own deficit targets. Assuming a budget balance elasticity to real growth of one-half, less than half of the average headline deficit shortfall of 2.4 percentage points relative to the annual budget targets in 2001–05 was related to the overly optimistic (real) growth projections.8 This proportion shrinks to one-third if account is made for various “exogenous” factors that improved the deficit ratios (within-year measures, nominal GDP revisions, and interest rate changes). And while the relative role of growth appears to have been increasing more recently, especially in 2005, this would only hold only if these latest deficit outcomes are confirmed as final—something that has not happened in Italy since the 1990s.9

7. Revenue targets for 2001–05 were largely met. Surprisingly, neither revenue levels nor its growth rates were substantially lower than projected (except in 2002), despite considerable (real) growth shortfalls. Three reasons accounted for these outturns:

  • the revenue base for short-term projections seems “understated,” as receipts were generally adjusted upward ex-post, partly due to new information; in addition, some subsequent statistical reclassifications tended to raise both revenues and expenditures;

  • various tax amnesties (mostly included in capital revenues) led to better total revenue performance, with current revenue performing less well (but if these revenues are adjusted for the growth shortfall, even these slippages are eliminated);

  • inflation was somewhat underprojected in 2001–03.

A02ufig04

Italy: Budget Targets and Outcomes in 2001–05

(in percent unless otherwise indicated)

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

8. But expenditure consistently overshot. Italy’s recent budgets have largely and systematically underprojected spending in levels and growth rates. These slippages basically account for the estimated non-growth-related fiscal balance shortfall.

  • Total spending in 2001–05 was 3½ percent higher than the nominal budget target (Table 1), and average annual growth of total spending was 1.6 percentage points higher than budgeted. Current primary spending accounted for the bulk of the overruns.

  • This underprojection was spread over most spending categories for 2001–05, with the exception of the interest bill (systematically overprojected) and of social spending (no apparent bias in the projections).

  • The downward biases in the projections have been strongest in capital spending, “goods and services,” and “other” current primary spending. The underprojection bias on wage spending was also significant, albeit not as large in percentage terms.

Table 1.

Italy: Cumulative Underprojection Bias in Spending, 2001–05

article image

9. The expenditure-based nature of the slippages likely indicates fundamental problems with spending control, rather than merely with forecasting. First, the deviations occurred in discretionary spending — where recent policy efforts to curb expenditure tended to concentrate, while projections in other areas, including revenues, appeared to be surprisingly conservative. Second, in Italy, spending-based slippages are structural by definition, since the cyclical component in Italy’s public spending is negligible due to low unemployment benefits. In any case, Italy clearly did not fit a general EU pattern whereby deficits were missed but spending plans were fully implemented in nominal terms.

B. Taking Stock of Italy’s Spending: Stylized Facts

10. The level and overall structure of Italy’s public spending differs moderately from that of most other EU countries. With the high interest bill reflecting the debt burden, primary spending in Italy has been lower than in the euro-area on average (but the gap has been closing rapidly). At the same time, Italy’s level of primary spending is higher than in Anglo-Saxon countries and its Mediterranean peers, which pursue a similar social model).

A02ufig05

Primary Spending

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

A02ufig06

Primary Spending, 1995–2005

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

11. The structure of primary spending also exhibits some differences.

  • In terms of economic classification, spending on social protection, goods and services, and investment is relatively low, while the public wage bill is high.

  • By function, Italy spends relatively more on public order and safety and on the environment, but less on social protection; and housing and community amenities. Italy’s pension spending/GDP ratio is relatively high reflecting its aging demographics and relatively low effective retirement age, but other spending on social protection is much lower, more than offsetting the differential on pension spending. Public health care outlays were particularly low a decade ago (following an appreciable retrenchment in the early 1990s), but have recently caught up with the EU average and even somewhat exceeded it.

A02ufig07

Public Wage Bill

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

12. The key long-term trends in Italy’s spending have been (1) a sharp rise and fall of the interest bill; and (2) a steady upward drift in most categories of primary spending. The latter was uneven across sub-categories of spending. For example, the wage bill barely grew as a percent of GDP over the last 25 years, but the other main categories —social spending and spending on goods and services — surged over the same period. Still, much of the increase in social spending, which grew most over the long term, occurred before the significant pension reforms of the early and mid-1990s.

A02ufig09

Italy: Key Spending Items in percent of GDP: 1980-2005

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

A02ufig10

Key Primary Spending Items in Terms of GDP (1980=100)

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

13. Since the late 1990s, however, spending on goods and services and, to a lesser extent wages and investment, accelerated, while social spending slowed. Much of the increase was attributable to health care (in the purview of regional authorities) and other local government spending. While a major discretionary attempt to limit spending, mostly on goods and services and investment, was made with the so-called “expenditure-cutting” (“taglia-spese”) law at the end of 2002,10 it did not alter the medium-term dynamics. The public wage bill also resumed growing from 2001, indicating limits to the effectiveness of freezes on new hiring (which contained wage pressures in the 1990s) and the low effectiveness of efforts, pursued since the early 1990s, to link wage increases to productivity. More generally, these developments reinforce doubts on the sustainability of the discretionary elements of the squeeze on spending attempted in the mid-1990s.

A02ufig11

Italy: Medium-term Spending/GDP ratios (1999=100)

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

14. On balance, these stylized facts raise concerns over Italy’s prospects for controlling spending. The differences with other European countries have been moderate (indicating limited scope for fiscal savings, from this aggregated perspective, without efficiency-enhancing reforms), and recent trends have been toward converging to this structure (i.e., relatively slow growth of the wage bill but acceleration in goods and services). The largest difference with other countries—low aggregate spending on social protection (particularly on unemployment benefits)—has been linked to the traditional role of the family as an alternative provider of the social safety net in Italy (see (Bibbee and Golio, 2002)), but the persistence of this factor cannot be taken for granted and may imply an upward adjustment for Italy’s spending. From the time series perspective, the sharp fall in the interest bill offered a unique opportunity to adjust, but was used instead for other spending.

C. Empirical Analysis of Determinants of Spending

Broad Approach and Data

15. The literature on public spending focuses on a variety of economic, political, and demographic factors. For aggregate spending, much of the debate has centered on national political economy determinants,11 though several factors have tended to weaken their applicability recently in the EU (especially the Maastricht criteria). Various economic factors are also believed to drive aggregate public expenditure outcomes, such as (1) past and actual inflation via indexation; (2) economic growth/cycle; (3) convergence of public spending across countries and over time; 12 (4) per capita output, real or evaluated at PPP (the latter reflecting the theory that demand for public services increases with wealth). Demographic factors also play a role, especially in the rapidly-aging industrialized countries.

16. The economic and other determinants would vary substantially between current and capital spending. Thus, while the theoretical literature on the determinants of aggregate spending is quite broad-based, the causes of public infrastructure spending are generally believed to be economic in nature, and (Turrini, 2004) provides a reasonably concise framework for testing these empirically in an EU-wide panel. Thus, it may be useful to investigate the determinants of capital and current primary spending separately.

17. The econometrics are based on annual panel data, although some times series properties have been studied.13 The number of relevant countries is too small for purely cross-country exercises, while the availability of a consistent time series is also severely restricted. And the period of interest would generally be from the mid-1990s, in order to abstract from the significant changes in the underlying economic parameters that were entailed by the accession to (or qualification for) the euro area.

Cross-country Analysis

18. Regression analysis indicates convergence in the primary current spending ratios in the euro area over the past decade. 14 The negative relationship between the change in the primary current spending ratio and its initial level holds for various econometric specifications and, as per Figures below, becomes stronger with controls for the cross-country differences in growth. This convergence in spending is consistent with the recent literature on industrialized countries (Skidmore et al, 2004). It may also reflect greater economic and policy uniformity as EU integration advances. Beyond convergence, other econometric results suggest that current primary spending ratio growth/level depends negatively on (lagged) real growth and positively on inflation, while the measured political factors and demographic structure variables did not play a significant role. The latter result is interesting and may partly reflect the fact that the direct spending effect of higher population aging may be partly offset by increased political economy urgency for reforms.

A02ufig12

Convergence in Current Primary Spending, EU-11

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

A02ufig13

Conditional Convergence in Current Primary Spending, EU-11

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

19. A cross-country perspective points to the importance of several factors in assessing actual or potential drivers of primary current spending in Italy. This spending broadly followed that of other high-debt countries such as Greece and Belgium, where it also increased by about 3 percentage points of GDP since 1995. This was in line with the convergence pattern, which may at least partly be related to the rebound of primary spending in high-debt countries as the interest bill fell. But some countries have, for significant stretches, “bucked” the convergence trend, reducing primary current spending despite below-average levels. While this concerned high-growth countries, Germany’s example, whereby the country stabilized primary current spending despite low growth, is likely indicative of the role of a policy effort to contain public spending.

A02ufig14

Primary Current Spending

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

20. The empirical analysis of capital spending over the last decade highlights the role of interest rates. Consistent with recent empirical estimates (Turrini, 2004), (Valila et al., 2005), the lagged debt/GDP ratio has a significant and negative relationship with capital spending, with public investment acting so as to smooth out changes in public debt. The interest rate has an expected, and significant, negative relationship with capital spending (which does not hold in a longer EU sample covering 1972–2002 (see Valila et al., 2005). This suggests that the pronounced decline in long-term interest rates since 1995 played some role in facilitating investment spending, while for earlier periods cost-of-capital considerations possibly had a lesser role to play.

21. These empirical results seem consistent with different stylized country experiences. Gross fixed capital formation tended to rise in countries that most benefited from the decline in interest rates, such as Italy and Greece, in contrast to more “stable” countries like Germany. However, both Italy and Greece, whose economies had been subject to significant discretionary fiscal efforts to stay within the 3 percent threshold, experienced pronounced brief slumps in public investment spending.

A02ufig15

Public Gross Fixed Capital Formation

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

Time Series Analysis15

22. Spending is negatively correlated with real GDP growth and positively with inflation. As expected, there is co-movement between annual changes in the ratios of the main categories of spending to GDP, which is slightly lower for wages, possibly due to the idiosyncrasies of the bargaining process. This co-movement however does not hold for intra-year (quarterly) data in 1999–2006, based on a simple pair-wise regression with seasonal dummies. But the inverse relationship of the ratios with real growth and positive with inflation holds both for annual and quarterly data, indirectly pointing to the possible role of these in driving expenditures, subject to the caveats of reverse causality.

23. The elasticity of expenditure to nominal GDP has on average been higher than unity, and rising most recently for primary spending. Regressions of total expenditure and its subcomponents on output point to a long-term elasticity of 1.04, although it declined in the mid-90s and rose over the last few years. In the period 1999–2006Q2, quarterly data indicate that the elasticity for primary current spending was 1.25. It was the highest for “other current” spending, goods and services, and wages, and lower for social spending. This ordering is roughly consistent with that for budget slippages for these categories, confirming the sharp acceleration in such spending.

A02ufig16

Elasticity of Selected Spending Categories to Nominal GDP, 1999-2006

Citation: IMF Staff Country Reports 2007, 065; 10.5089/9781451820003.002.A002

24. Time series statistics suggest that 1995 marked a surprisingly large discretionary effort in expenditure-based consolidation. One-step Chow tests for recursive regressions for 1980–2005 for elasticities of spending to GDP, and simple autoregressive analysis of the expenditure/GDP ratio, suggest that the containment in primary current expenditure was “extraordinary” in 1995. (Incidentally, that year was characterized by relatively high real and nominal economic growth, which permitted sharp declines in all main primary current spending categories). At the same time, other annual fiscal consolidations, in particular the much-discussed 1997 budget, directed at meeting the Maastricht criteria, did not amount to a break in primary current expenditure dynamics, whether measured in nominal terms or as ratios to GDP. In fact, a series of consolidations in the second half of the 1990s only stabilized the spending ratio after its sharp decline in 1995.

25. Intra-year spending on wages, investment, and goods and services has recently tended to “seesaw,” whereby current growth depends negatively on the past period’s. Regressing the change in the main primary spending categories on its lagged rate of change (and seasonal dummies) for 1999–2006 yields negative and significant coefficients for wages, investment, spending on goods and services, and “other spending,” but a non-negative coefficient for social spending (indicating a “smoother” pattern). While such a pattern may be inevitable for wages due to bargaining delays and the consequent repayment of arrears, it is less logical for other categories. The exact explanation for the seesaw pattern is not clear, but, against the backdrop of fiscal overruns and high spending growth, it may denote inefficient stop-and-go cycles, whereby last-minute attempts to limit spending to meet aggregate targets are followed by costly rebounds. There is some evidence that overruns on some public investment projects may have been partly affected by these factors.

D. Conclusions

26. Part of Italy’s recent higher public spending was due to underlying, “exogenous” factors. These stemmed from a spending rebound after the “austerity” of the 1990s and convergence of primary spending to the higher EU-wide level. Regressions also suggest that low growth and above-average inflation complicated the task of curbing current primary spending, while the medium-term decline in interest rates has cut the cost of capital, fostering public fixed capital formation. The results should however be interpreted with caution given the small number of observations and the low power of many empirical tests.

27. But the spending surge also reflected policy weaknesses and lack of fundamental reforms. Obviously, the rebound of spending and its observed convergence also reflected policy choices, though disentangling these formally would be difficult (the measured political economy factors do not prove significant, probably due to data limitations). In any case, Italy’s expenditure targets were substantially exceeded in nominal terms and growth rates, while overruns occurred in areas where fundamental reforms have yet to take hold. Against this background, the relative stabilization of Italy’s pension-based social spending (which has been reformed), points to the key role of fundamental reforms in stabilizing spending.

28. There is no scope for further aggregate expenditure increases in Italy. Pension spending and the wage bill are already higher than those of the euro-area, while health care spending has caught up. Future age-related spending pressures on the debt ratio may exceed official projections, especially if growth rates continue to surprise on the downside. Italy’s traditionally low unemployment benefits may also be subject to upward pressures, partly if there is further euro-area-wide convergence of spending. Capital spending will likely be high due to intra-Italy regional differences and substantial public infrastructure needs. Thus, given the medium-term objective of a small overall budget surplus and Italy’s already-high fiscal burden, there is no alternative to expenditure-based adjustment. Absent effective curbs on spending, adjustment will likely take the form of “speed bumps,” in the form of financial tensions and possibly crises.

29. Expenditure reforms are thus needed to generate resources for productive spending and fiscal consolidation. In particular, these should:

  • Maximize performance of public spending in terms of micro-efficiency, as there is evidence that in Italy the outcome efficiency of spending could be improved, particularly in education (see Afonso and St. Aubyn, 2005)) and public infrastructure (Fedelino, 2006).

  • Ensure a rule-based approach to expenditure management, which should smooth spending patterns over time and avoid cost overruns, thus helping reduce longer-term spending pressures. In this regard, (EC, 2006) explicitly links more effective spending restraint to the presence of fiscal rules.

  • Establish a track record of at least meeting expenditure targets in nominal terms, which would go a long way toward attaining the targeted deficit ratios. Beyond the spending reforms, reform of budget procedures and greater transparency in the projections would help.

  • The strong effect of growth on the spending ratio suggests that episodes of cyclical strength be seized upon for adjustment, as expenditure cuts in high-growth environments seem to be the most effective and durable. But broad structural reforms to achieve durable improvement in growth (which include spillovers from expenditure rationalization) would be even more important for this link to operate fully.

References

  • Afonso A. and M. St. Aubyn (2005) “Assessing Education and health Efficiency in OECD Countries Using Alternative Input Measures.” pp. 361 –389 in “Public Expenditure,” Banca d’Italia, 2005.

    • Search Google Scholar
    • Export Citation
  • Alesina, A., N. Roubini, and G. Cohen, 1999, Political Cycles and the Macroeconomy (Cambridge, Massachusetts: MIT Press).

  • Balassone F., Franco D. and S. Zotteri (2006) “EMU fiscal indicators: a misleading compass?” Empirica, 2006, vol. 33, issue 2, pages 63 –87.

    • Search Google Scholar
    • Export Citation
  • Barro R. (1990) “Government Spending in a Simple Model of Endogenous Growth,” Journal of Political Economy, 98, pp 102 –25.

  • Bibbee A. and A. Golio (2002) “Public Spending in Italy: Policies to Enhance its Effectiveness,” OECD Economics Department Working paper, No 324, March.

    • Search Google Scholar
    • Export Citation
  • EC (2006) “Public Finances in the EMU,” EUROPEAN ECONOMY. No. 3. 2006. Office for Official Publications of the EC. Luxembourg.

  • ECB (2006) “The Importance of Public Expenditure Reforms for Economic Growth and Stability,” European Central Bank monthly bulletin, March, pp 61 –73.

    • Search Google Scholar
    • Export Citation
  • EPC (2006) Report by the Economic Policy Committee and the European Commission on the impact of ageing populations on public spending (February), http://ec.europa.eu/economy_finance/epc/documents/2006/ageingreport_en.pdf.

    • Search Google Scholar
    • Export Citation
  • Fedelino A. (2006) “Infrastructure Investment and PPPs,” IMF Selected Issues, 06/59, pp 86 –98.

  • Jonung L. and M. Larch (2005) “Improving Fiscal Policy in the EU: the Case for Independent Forecasts,” Economic Policy, September, pp. 1 –33.

    • Search Google Scholar
    • Export Citation
  • Milesi-Ferretti, G. M., R. Perotti, and M. Rostagno, 2002, “Electoral Systems and Public Spending,” Quarterly Journal of Economics, Vol. 117, No. 4, pp. 607 –57.

    • Search Google Scholar
    • Export Citation
  • Skidmore M., Toya H., and D. Merriman (2004) “Convergence in Government Spending: Theory and Cross-Country Evidence”. Kyklos, Vol. 57, No. 4, pp. 587 –620, November.

    • Search Google Scholar
    • Export Citation
  • Turrini A. (2004) “Public Investment and the EU fiscal framework,” European Commission, Directorate General for Economic Affairs, Economic Papers, 202.

    • Search Google Scholar
    • Export Citation
  • Valila T. and A. Mehrotra (2005) “Evolution and Determinants of Public Investment in Europe,” European Investment Bank, Economic and Financial Report 2005/01.

    • Search Google Scholar
    • Export Citation
6

Prepared by Bogdan Lissovolik (EUR), blissovolik@imf.org.

7

While Italy currently has the highest pension expenditure as a share of GDP in the EU, it is officially projected to increase minimally between now and 2050 (see EPC, 2006), also reflecting the impact of the recent 2004 reform. But this conclusion hinges crucially on full implementation of the legislated reforms (notably adjustment coefficients) and on long-term growth assumptions that appear optimistic for Italy (see IMF Country Reports for Italy for 2003–06).

8

The calculation aims to skirt the frequently revised estimates of the output gap. In any case, the results should not be affected significantly as long as the elasticity to real (actual or potential) growth remains around ½, which is consistent with the figure agreed by the authorities with the European Commission.

9

See (Balassone et al., 2006) for an analysis of ex-post revisions in Italy and EU-wide.

10

The 2002 law gave the Minister of Economy and Finance additional authority to cut spending allocations within the budget year.

11

See (Alesina et al., 1999) and (Milesi-Ferretti et al., 2002) and references therein.

12

The observed convergence was linked to diminishing returns to government activity on the production side (see Barro, 1990), falling marginal utility from government spending on the demand side, as well as globalization’s impact on the equalization in tax burdens.

13

Data have been taken from a variety of sources. For cross-country regressions, these have been Eurostat and OECD. For Italy, the main source is the official statistical agency ISTAT, which publishes data on economic and functional classification of expenditure. Two such updated aggregate time series of public spending data are available: (i) annual data on the economic classification of expenditure in 1980–2005; and (ii) quarterly data since the first quarter of 1999.

14

The analysis is based on a panel for 11 euro-area countries (excluding Luxembourg). Consistent with recent studies (EC, 2006), the focus is on structural factors, by expressing all fiscal variables in percent of trend GDP. For most regressions, the main specification is country fixed-effects (though other specifications were used where relevant to check for robustness). More detailed results are available from the author on request.

15

The analysis of the Italy-only time series is constrained by data availability. Since published updated annual and quarterly series contain some 25–30 observations, at best these allow to check very simple statistical properties and bivariate relationships. In addition, given that Italy’s prospects of EMU accession marked a structural break, the main period of interest would be post-1995.

Italy: Selected Issues
Author: International Monetary Fund