Italy: Selected Issues
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google ScholarClose
Publication Date:
18 Jul 2000
eISBN:
9781451819793
Language:
English
Keywords:
ISCR; CR; Ministero del Tesoro; regulatory environment; sc; national government; tax rate; government hope; programming document; financial institution; Health care spending; Total expenditures
Download PDF (2.1 MB)

The paper examines macroeconomic and structural factors potentially explaining the country's underperformance. A comparison between the reform and baseline policy scenarios underscores maintaining a strong fiscal position, early reductions in primary expenditures, and reducing fiscal vulnerability. Assigning the financing and management of the health care system to the regions may increase the efficiency and the productivity of the health care system. The information on Italy's economy and legal as well as regulatory environment is available on the worldwide web, and the paper lists the related sites.

Abstract

The paper examines macroeconomic and structural factors potentially explaining the country's underperformance. A comparison between the reform and baseline policy scenarios underscores maintaining a strong fiscal position, early reductions in primary expenditures, and reducing fiscal vulnerability. Assigning the financing and management of the health care system to the regions may increase the efficiency and the productivity of the health care system. The information on Italy's economy and legal as well as regulatory environment is available on the worldwide web, and the paper lists the related sites.

The Evolving Role of Regions in Italy: The Financing and Management of Health Care Services54

A. Introduction

111. The previous chapters argued that reducing the tax burden and improving public sector efficiency would have a central role to play in reinvigorating growth; and that these growth supporting fiscal policies called for reining in aging-related increases in public expenditure—a challenge far from unique, but particularly daunting for Italy in view of relatively adverse population aging trends. Following an assessment of the Italian pension system in last year’s background studies (Chapter I in SM/99/115, 5/20/99), this chapter focuses on the second major expenditure area affected by population aging, health care.

112. Health care currently accounts for some 13 percent of general government primary expenditures in Italy and could potentially increase by as much as 2½ percent of GDP as a result of the projected population aging. When the National Health Service (NHS) was established in 1978, the responsibility for providing public health care services was assigned to the regions; nonetheless, the NHS has in effect been centrally managed and financed. The recent reform of intergovernmental relations will bring about a major change in this regard, and increase the responsibility of regional governments in the managing, and their resources for financing, of the public health care system. This chapter focuses on the evolving role of regional governments, describes the recent reform of intergovernmental fiscal relations, and its implications for health care financing—assessing how these reform steps fit into the broader policy challenges outlined in the previous chapters. The discussion attempts to put Italy’s experience in a broader cross-country perspective, by comparing the institutional setup and the expenditure trends to those prevailing in other industrialized countries.

113. The rest of the chapter is organized as follows. Section III.B gives a brief description of the regions in Italy and Section III.C describes their expenditure assignments. Section concentrates on health care services, by far the most important expenditure assignment of the regions. Section III.E describes and assesses the tax assignment of regions and the future equalization mechanism. Section III.F gives a brief description of the rules limiting borrowing by regional governments, whose fiscal responsibilities are also covered by the Internal Stability Pact (Section III.G). Finally, Section III.H concludes.

B. Key Characteristics of Italian Regions

114. The principle of decentralization of powers was already established in Italy’s Constitution of 1948, which divided the Italian Republic into regions, provinces, and municipalities (Article 114). There are 20 regions (listed in Article 131), 5 of which, namely Sicilia, Sardegna, Trentino Alto Adige, Friuli-Venezia Giulia, and Valle d’Aosta are special statute regions, that is, regions that enjoy a higher degree of autonomy and whose statutes are approved with constitutional laws (Article 116). The remaining 15 are ordinary statute regions, and their statutes are approved with simple parliamentary laws.

Geographic and demographic characteristics

115. The 20 Italian regions form an extremely varied universe. They differ widely in size, ranging from the 25,000 sq km of Piemonte and Sicilia (each 8½ percent of the national territory) to the 3,000 sq km of Valle d’Aosta (1 percent of national territory). The population is also very unevenly distributed, with nearly 9 million people living in Lombardia (15½ percent of the total population) and less than½ of 1 million in Valle d’Aosta and Molise. Population density also varies considerably (see Table 1), and is on average 20 percent higher in the Center-North than in the South.

Table 1.

Italy: Key Characteristics of Regions

article image
Source: ISTAT.

Inhabitants for square kilometers.

116. The regions also differ significantly in the age distribution of their population—an important determinant of future outlays on health benefits. The South is relatively “younger,” with 15 percent of the population aged 65 and older compared to 19 percent in the Center- North (and to a nationwide average of 17½ percent). Liguria is the “oldest" region, with one- fourth of the population aged 65 and older, followed by Emilia-Romagna, Toscana, and Umbria with 21½ percent, whereas Campania is the “youngest,” with only 13 percent.

Economic characteristics

117. Recorded labor force participation rates are markedly higher for the Center-North (62 percent) than for the South (52 percent); Emilia-Romagna has the highest participation rate with 68 percent, whereas Puglia and Calabria are at the bottom, with participation rates barely above 50 percent, and Lazio is the only region in the Center-North with a participation rate below 60 percent (see Table 2). Perhaps the most dramatic difference between the Center-North and the South lies in the unemployment rate, which in 1999 averaged 22 percent for the South versus 7 percent for the Center-North.

Table 2.

Italy: Regional Labor Market Indicators

article image
Source: ISTAT (1996).

118. Other economic disparities across regions are equally significant. The characterization of Italy as a “dual economy” can easily be understood by noting that 75 percent of Italy’s GDP is produced in the Center-North (with Lombardia alone accounting for 20 percent) and only 25 percent in the South—this proportion has remained virtually unchanged over the last 20 years (Table 3).

Table 3.

Italy: Regions’ Contributions to GDP, 1980-96

article image
Source: ISTAT

119. Divergences in per capita GDP are equally stark. In 1996, when nationwide per capita income stood at Lit 32 million, four regions of the Center-North, that is, Valle d’Aosta, Lombardia, Trentino Alto Adige, and Emilia-Romagna, had a per capita income above Lit 40 million, while at the other end of the spectrum Calabria was the poorest region with just Lit 18 million, and Campania, Basilicata, and Sicilia were around Lit 20 million. Overall, the South had a per capita GDP of Lit 21 million, or only⅔ of the national average, whereas per capita income in the Center-North was 20 percent higher than the national average, at Lit 38 ½ million (Table 4).

Table 4.

Italy: Regional Per Capita GDP, 1996

article image
Source: ISTAT

C. Expenditure Assignments of Regions

Expenditure responsibilities of regions

Historical perspective

120. While the Constitution approved in 1948 already provided for the establishment of regions as autonomous entities and with directly elected regional governments, only special statute regions were quickly established. The first regional councils for ordinary statute regions were only elected in 1970, and two more years elapsed before the transfer of powers, personnel and funding necessary for these governments to operate took place. In 1977, the “616 decrees” transferred to the regions large portions of the state bureaucratic apparatus (with several thousands of employees being moved from the central administration to the regions), delegated to them legislative authority in important fields such as territorial planning and social agencies, and transferred to them almost full responsibility for provision of health care services. The regions thereby assumed responsibility for approximately one quarter of the national budget—in large part as a consequence of the transfer of responsibility over the health care sector, which by 1989 already accounted for over one-half of regional spending.’55

121. The competencies and responsibilities of the ordinary statute regions are determined by Article 117; they include health services, social assistance, labor training and professional instruction, transports of regional importance, tourism, road maintenance and construction, public works, and agriculture. The areas of action of the special statute regions are broader and vary from case to case; in addition to those of the ordinary regions they often include partial responsibility for primary and secondary education, subsidies to industry, commerce and agriculture, and financial support for culture and the arts. The regions can in turn delegate some of their administrative activities to the lower levels of governments, that is, provinces and municipalities (Article 118).

122. The regions are granted financial autonomy within the limits determined by the national laws, which coordinate the regions’ financial autonomy with that of the state, the provinces and the municipalities (Article 119). The regions are given (i) own revenues; and (ii) conditional and unconditional state transfers, in amounts sufficient for them to carry out their regular functions. Moreover, regions have their own patrimony. The state assigns by law special contributions to individual regions for specific purposes, in particular to foster development in the South and in the islands (Sicilia and Sardegna). Finally, regions are allowed to borrow to finance investment (see below).

The Bassanini reform

123. The recent reform of intergovernmental fiscal relations is part of an overall reform of the state in Italy, known as the Bassanini reform,56 comprising two framework laws, Laws 59 and 127 of 1997, and subsequent implementation decrees. The goal of the reform is to create a service-oriented state providing high quality services at the lowest cost (including compliance costs). The reform is to be carried out within the context of the existing constitution—though decentralization is a central part of the reform, the unitary nature of the Italian state will remain unchanged.

124. The basic idea behind the reform is to redefine the role of the state, divest noncore activities, and reallocate core activities across levels of government and different parts and agencies of the central government in order to achieve the above-stated goal. The reform involves decentralization of the retained functions of government, deregulation, simplification of administrative procedures, contracting out of government services, a new approach to human resource management, and a reform of the budget process.

125. Concerning decentralization, the first step has been to identify those functions that should stay with the central government. These include justice, national security, defense, international relations and trade, relations with international organizations and coordination of EU relations, macroeconomic policy, national energy policy, industry, telecommunication, transport, education, and research, as well as setting the national policy objectives and standards in areas such as agriculture, employment, public housing, public works, and consumer protection. The remaining core functions have been assigned to subnational governments according to the principle of subsidiarity. Trade and industrial development policies have been devolved to the regions. The final step in the process of decentralization, which has been designed as a five-year process, will be the transfer of financial and human resources to the subnational governments. The modalities of the process will be agreed upon at the central state-regions and the central state-municipalities conferences, established to foster cooperation between the different levels of government.

Overall trends in public expenditure at the subnational level

126. Expenditure responsibilities of subnational governments expanded substantially in the second half of the 1970s, when total expenditures of subnational governments came to account for about 10 percent of GDP, or roughly one-fourth of total expenditures of the general government (Figure 1). As the size of government expanded, so did the role played by subnational governments—within a few years their total expenditures increased to over 14 percent of GDP. Their share in total expenditures of the general government was close to 30 percent for much of the 1980s. As a percent of GDP, total expenditures of subnational governments declined over the following decade, from a peak of 15 percent in 1991 to about 13½ percent at the end of the 1990s.

Figure 1.
Figure 1.

Italy: Public Expenditures, 1977-99

(In percent of GDP)

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

127. The decline in expenditures over the 1990s reflected the contribution of subnational governments to Italy’s fiscal consolidation effort in the run-up to stage three of EMU: the overall deficit of subnational governments dropped from an average of 1 percent of GDP over 1981-90 to an average of 0.2 percent of GDP over 1991-98 (as the deficit of the consolidated general government declined from an average of 11 percent of GDP to an average of 7 percent of GDP). In 1990, the general government deficit stood at 11 percent of GDP, and subnational governments recorded a deficit of 1.3 percent; by 1997, the year when compliance with the criteria for EMU accession was assessed, the deficit of the general government had declined to 2.8 percent of GDP, and that of subnational governments to 0.3 percent (Figure 2)

Figure 2.
Figure 2.

Italy: Composition of General Government Deficit, 1977-99

(In percent of GDP)

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

128. The increase in total expenditures mentioned above was accompanied by a near- doubling of revenues from 7½ percent of GDP in 1976 to 14½ percent of GDP in 1991-92 (see Table 5, which summarizes the consolidated budget of subnational governments—regions, provinces, and municipalities—over the period 1976-98). Transfers from the state and other public institutions increased rapidly at the beginning of the period, from 4Y2 percent of GDP in 1976 to a peak of 10 percent in 1983; tax revenues increased steadily, from less than 1 percent of GDP in the late 1970s-early1980s to 3-3½ percent of GDP in the mid-1990s, and then jumped to 5½ percent of GDP in 1998 with the introduction of IRAP. On the expenditure side, wages and salaries quickly increased from 3½ percent of GDP in the late 1970s to an average of 4½ percent of GDP over 1980-98; social transfers, which accounted for only⅓ of 1 percent of GDP in the late 1970s, increased to 2-2½ percent in the 1980s and 1990s; whereas capital expenditures, which between 1980 and 1991 had hovered at about 3 percent of GDP, thereafter declined to between 2 and 2½ percent for the remainder of the decade. The share of social transfers in total expenditures quickly increased from just 3 percent of total expenditures in the late 1970s to about 15 percent on average over 1980-98 (Table 6).

Table 5.

Italy: Consolidated Fiscal Accounts of Subnational Governments, 1976-98

article image
Sources: Bank of Italy, Relazione Annuale, various issues.
Table 6.

Italy: Expenditures of Subnational Governments

(In percent of total)

article image
Sources: Bank of Italy, Relazione Annuale, various issues.

129. Provision of health care services is currently by far the single most important function in the regions’ budgets (Table 7).57 In 1999, outlays on health care (both current and capital) accounted for over 60 percent of regions’ total cash expenditures.58 Almost the totality of the regions’ expenditures on health care consist of transfers to local health units and hospitals.

Table 7.

Italy: Expenditures of Regions, 1997-991/

article image
Source: Ministry of the Treasury, Budget, and Economic Planning, Relazione Trimestrale di Cassa.

Data are on a cash basis.

D. The Management and Financing of the Public Health Care System

130. The Italian public health care system has been highly centralized and integrated—even though most of the responsibilities for providing health care services were assigned to the regions when the NHS was established in 1978. The central government (Ministry of Health) determined in detail the services all regions had to provide and the budget allocated earmarked resources based on the cost estimates of the Ministry of Health. The regions have enjoyed very little freedom in this respect, and consequently had very little incentive to contain costs. While this arrangement made the decentralization of the management of the NHS a notional one, it made it possible to maintain a highly uniform level of per capita expenditure across regions. This, however, did not ensure equal access to and quality of services, because the efficiency of local health units varied greatly across regions.

131. The system was also characterized by recurrent budgetary overruns on health outlays, often financed through accumulation of arrears to suppliers. The regions have argued that (i) expenditure overruns were the unavoidable consequence of unrealistic budget allocations set by the central government; and that (ii) given that health services were largely mandated by the state, they had little or no control over the evolution of health expenditures. The constitutional court agreed on the latter point, and a large part of the responsibility for repaying the outstanding arrears has been borne by the central government. More recently, the central government and the regions have jointly agreed on what they consider a more realistic budgetary allocation for health expenditures—a move that the government hopes will increase “ownership” on the part of the regional authorities, at the same time as a major reform of health care expenditure is underway.

132. Previous reforms of intergovernmental fiscal relations concerned almost exclusively the revenue assignments of the regions. The reform being presently carried out on the basis of the 1997 framework law completes this process. However, the reform includes a new element which could fundamentally change the way the NHS is managed: starting in 2004, the regions will be free to allocate their revenues and a new system of centralized monitoring of health care services delivered by the local health units will be set up. Regions failing to meet the defined minimum standards can be sanctioned; the transfer they get from the equalization fund can be reduced.

133. Within a cross-country framework, the remaining part of this section gives a brief analysis of public health care expenditure and financing in Italy. A detailed analysis of micro efficiency or incentive structures, even at the national level, is beyond the scope of the paper.59

Macroefficiency

134. The share of total health care expenditure in GDP in Italy grew rapidly (Figure 3), and reached the EU average by 1990 (Figure 4). During the 1990s, however, public health care expenditures in Italy slowed down compared to the rest of the EU. Driven by the need for fiscal consolidation, current public health care expenditure was reduced in real terms between 1990 and 1995 (Figure 5); as a result, the share of public health care expenditure in GDP fell by 1.2 percentage points between 1991 and 1995. Part of this reduction was offset by a rapid increase in private health care expenditure, a clear indication that the reduction was not demand driven.60 Besides the Scandinavian countries, no other country within the EU managed to reduce the share of health care expenditure in GDP during this period—between 1990 and 1995, it increased by 0.9 percentage point in the rest of the EU, and by 1.7 percentage points in Germany. This would suggest that Italy was rather successful in containing health care costs, in particular given the rapid increase in the elderly as a proportion of the population (Figure 6). In the 1990s, population aging was one of the most important factors explaining the continued increase in the share of health care expenditure within the EU (Figure 7), in spite of numerous attempts to arrest this tendency (OECD, 1994).

Figure 3.
Figure 3.

Italy: The Share of Health Care Expenditure in GDP, 1960-97

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.
Figure 4.
Figure 4.

Italy: The Share of Total Health Care Expenditure in GDP, 1970-97

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.
Figure 5.
Figure 5.

Italy: Average Annual Growth Rates of Real Current Health Care Expenditure, 1961-95

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.
Figure 6.
Figure 6.

Italy: Health Care Expenditure and Aging, 1970-97

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.
Figure 7.
Figure 7.

Italy: Health Care Expenditure and Aging in the Rest of the European Union, 1970-97

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

135. However, a part of the reduction in public health care expenditure achieved in the first half of the 1990s may not have been sustainable. The share of public health care expenditure in GDP started to increase after 1995, offsetting one-fourth of the reduction achieved earlier. This increase was partly related to weak growth and the continued increase in the share of the elderly population, but also to the way the reduction in health care expenditure was achieved and, as described below, the way health care costs were budgeted and financed. In 1996 and 1997, the overruns in public health care expenditure were sizable,61 mostly financed by unpaid obligations to private suppliers. As Table 8 shows, wage costs accelerated rapidly after 1995, clearly indicating that earlier wage cost reduction was not achieved in a sustainable manner; the same was true for contracted services and pharmaceuticals. The rapid increase in the arrears of the local health units and in expenditure62 on contracted health care services during this period were also very closely related developments, again indicating that the way the earlier reductions were achieved was not sustainable. The experience of 1995-97 suggests that it is not possible to achieve sustained reduction in public health care expenditure without properly designed reforms aimed at increasing efficiency.

Table 8.

Italy: The Structure and Growth of Regional Current Health Care Expenditure

article image
Source: Corte dei Conti (1999).

Microefficiency and regional pattern of public health care expenditure

136. A detailed analysis of microefficiency at the regional level is well beyond the scope of this paper. A partial analysis commissioned by the technical expert group on public expenditure of the Ministry of Treasury (CTSP, 1999) suggests that relatively uniform levels of public health care expenditure did not result in uniform levels of health services provided by the local health care units (output), or the same improvement in health outcome across the regions.

137. During 1980-97 regionally financed health care expenditure increased by 2.5 percent per year in real terms, considerably faster than real GDP. Correcting for population growth, which was faster in the South, and aging, which was more pronounced in the North, average growth in real expenditure was below average in the regions in the North East and the Center (Table 9). This reflected a catch-up process, as the levels of per capita health care expenditure in the regions with lower average growth rates were still above the national average in 1997 (Table 10).

Table 9.

Italy: Growth of Real Health Care Expenditure, 1980-97

article image
Source: Fund staff calculations based on data from CTSP (1999).
Table 10.

Italy: Regional Health Care Expenditure, 1994-97

article image
Source: Fund staff calculations based on data from Corte dei Conti (1999) and from MOP (1991).

138. The extent of regional inequality in public health care expenditure in Italy is limited, in particular when adjustment is made for differences in age structure and for the expenditure on hospital care for nonresident patients (Figure 8 and Table 10).63 At least as far as expenditure on health care (input) is concerned, the NHS did achieve its stated goal of providing uniform level of assistance (Law 833/78).

Figure 8.
Figure 8.

Italy: Inequality Among Regions in Income and Health Care Expenditure, 1997

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

139. A noteworthy fact is the rapid increase in health care expenditure in the northern special statute regions in the 1990s.64 Income levels in these regions are close to that in Lombardia (the richest ordinary statute region), and other characteristics are also similar to those of the rich northern regions. These regions, however, enjoy a much higher level of expenditure autonomy and have much larger tax bases because of their special tax sharing arrangements with the central government. There are pronounced differences in cost structures across regions (Table 11). The share of expenditure on services directly provided by the local health units is lower in the South than in the North, with regions in Central Italy closer to the regions in the South. Within the services directly provided by the local health units, the share of personnel costs in the South is significantly higher than in the North, reflecting a rapid growth of health personnel since the establishment of the NHS which resulted in low productivity, rather than increasing output. (Expenditure on pharmaceuticals is also significantly higher in the South than in the other regions, reflecting a higher per capita expenditure in the South partly explained by a higher share of exempt prescriptions.)65

Table 11.

Italy: Structure of Regional Health Care Expenditure in 1996

article image
Source: CTSP (1999).

140. More than half of the increase in NHS staff between 1981 and 1997 was attributable to the regions in the South, one -third of which to Sicily, where the increase was over 40 percent66 compared to a national average of 14 percent.67 At the same time, local health units in the South relied on contracted private hospital care and specialist services to a much larger extent than in the North, and patients from the regions in the South accounted for around 90 percent of outward hospital mobility (hospitalization of patients in other regions).68 While Southern regions had the lowest ratio of hospital beds to inhabitants, they had also the lowest bed-occupancy ratios. Empirical research (see Alesina and others, 1999) found a significantly lower incidence of recourse to certain public health care services69 and significantly lower rating of the quality of health care services by residents in the South.70 Improvements in health outcome were also significantly higher in the North.

Financing of the NHS

141. The financing of the NHS in Italy was until very recently characterized by deliberate ex ante underfunding (in the budget), large deficits, and frequent bailout operations. Underfunding was seen as a means to put downward pressure on health care expenditure, as it was difficult to identify and measure the factors which determined the demand for and the cost of delivering health care services71 and local health units had little incentive to control costs. With the exception of 1995, actual health care expenditure exceeded projections in each year between 1993 and 1997 (see CTSP, Table 12). The ensuing arrears and debts of the local health units were periodically taken over by the central government (eight times during the period 1980-1997), balancing health care financing and expenditure in an ex post manner. Thus the periodic reduction (elimination) of the accumulated liabilities of the local health care units became a structural element of health care finance in Italy.

Table 12.

Italy: Sources of Health Care Finance, 1993-97

(Total Actual Expenditure = 100)

article image
Source: CTSP (1999).

142. Table 12 shows the financing structure of the NHS for 1993-97. Until 1992, financing was exclusively from the central government budget through the national health fund (Fondo Sanitario Nazionale, FSN). To enhance fiscal responsibility at the regional level, the health care contributions were assigned to the regions in 1993 and regions were given the right to increase the rate of contribution—none of them actually exercised this right, however, even though all recorded a deficit. Moreover, ca-payments (ticket) were introduced. As described above, these steps failed to strengthen fiscal discipline. The 1997 tax reform replaced health contributions (as well as other local and central business taxes) with a local business tax (IRAP), and introduced a regional personal income tax. The most recent reform replaces the budget transfer to the health care fund (FSN) with a share of VAT revenue assigned to the regions.

143. Prior to the most recent reform, the allocation of the (ex-ante) available financing to the regions was based on a formula, and was decided by the Interministerial Committee on Economic Planning (CIPE). The formula, however, was changed practically every year, and the set of relevant variables changed five times during the last two decades. Between 1992 and 1996 the financing was practically distributed on a pro-capita basis. In 1997 an index of demand for health care services, reflecting the age structure and health condition of the population was introduced.

144. The actual outcome of the allocation process was a very even distribution of per capita (ex ante) financing. However, the actual deficits were very unevenly distributed across regions. The bail-out operations distributed ex post finance more or less in proportion to the accumulated liabilities; thus the regions in the North-East and the Center effectively managed to finance their historically higher levels of health care spending. The amount of (ex post) financing channeled to the regions through the bailout operations was considerable (CTSP, 1999). This resulted in a loss of credibility for the allocation mechanism through which ex ante financing levels were determined and in a very nontransparent and unstable system of ex post health care financing (Reviglio, 2000).

Potential problems of decentralization

145. The decentralization of the management of the public health care services will imply a major change in the role of regional governments. The initial positions of individual regions will be rather different, as local health units in the South are less efficient than those in the North. Due to the existing labor market regulations, the degree of overstaffing, and the size of the public sector wage premium72 in the South, it will be difficult to reduce employment and introduce more flexible labor contracts in order to lower unit labor cost. A weak local economy could make this process even slower.

146. Local health units in the South rely more on contracted private health care services and have to devote a higher share of expenditure to pharmaceuticals, cost items which tend to increase fast. Because of the large share and downward rigidity of wage costs, they will find it difficult to counterbalance this increase. As their capacity to raise additional own revenue is very limited (see Section III.E), they will have to either (i) increase co-payments and/or reduce the level of public health care services, or (ii) ask for higher transfers. This will put increasing pressure on the central government to assign a higher share of tax revenue to the regions.

147. While more pronounced in the South, the inefficiency of local health units is by no means restricted to the South. Moreover, local health units in Central and North-Eastern Italy were the ones that exceeded their budget allocations by the widest margins. Thus, governments in these regions will also find it difficult to contain health care expenditure.

148. The reduction in the share of public health care expenditure in GDP achieved in the first half of the 1990s appears unsustainable without further reforms aimed at increasing the efficiency of local health units and containing the costs of purchased services and pharmaceuticals. It is not evident that regional governments are better placed to design and carry out such reforms than the central government—especially given the rather limited administrative capacity of some regional governments. Moreover, due to the highly centralized nature of the system so far, regional governments have accumulated very little experience with managing a health care system and they are likely to have less power to contain costs using administrative pressure. That is, while the transfer of this responsibility to regional governments may be expected to result in better and more cost efficient public health care services in the longer run, the transition is likely to be a difficult one.

E. Revenues of Regional Governments

149. Prior to the recent reform, intergovernmental fiscal relations in Italy had been characterized by a high degree of vertical imbalance, ambiguity over responsibility for financing expenditure, and lack of stability and transparency.73 Concerning the ordinary statute regions, the process started in the late 1970s with the decentralization of the functions assigned to the regions in the constitution. This was however done without creating adequate tax bases for the regions, and their revenues, including transfers from the central government, were determined by ad hoc legislation.

150. Before 1992, ordinary regions had only low yielding own taxes amounting to less than 3 percent of their total revenue (Emiliani and others, 1997). Since then, several waves of tax reforms have gradually increased regional tax revenue (see Table 13).74 The first step was to attribute health care contributions and assign a motor vehicle tax and a surcharge tax on natural gas consumption to the regions in 1992. This was followed by the sharing of the excise on gasoline in 1995, and the assignment of a new regional business tax (IRAP) and a regional personal income tax (PIT) to the regions in 1997. By 1998, total tax revenue of the regions reached some 3.1 percent of GDP. Finally, the 1999 reform has introduced the sharing of VAT revenue, and increased the maximum regional PIT rate and the share of the gasoline excise revenue assigned to the regions.

Table 13.

Italy: Own Revenues of Ordinary Statute Regions

article image
Source: Fund staff-calculations based on data received from the authorities.

151. In spite of the gradual increase in own tax revenue, however, the actual degree of fiscal autonomy enjoyed by the regions was very limited until recently. Health care contributions attributed to the regions, and later IRAP, were fully earmarked to finance health expenditure. Thus, regional governments had practically no power to decide on the allocation of the largest part of their revenue. Moreover, they enjoyed only a very limited degree of tax autonomy. The major change the recent reform will bring about in this respect will be the phasing out of earmarking of revenues and of the restrictions on the allocation of health care expenditure. The regions will be given an increased (though still rather limited) degree of tax autonomy and an almost complete freedom to allocate expenditure, provided they comply with minimum standards on health care services to be defined by the central government.

Own revenues

152. The information on the taxes of ordinary statute regions75 is summarized in Table 14 and the main regional taxes are described in the Appendix. IRAP, the regional business tax introduced in 1998, is by far the most important source of own tax revenue for regional governments (Table 15). In 1998, it accounted for 74 percent of the total own tax revenue of the regions, or 2.2 percent of GDP.

Table 14.

Italy: Revenues of Ordinary Statute Regions

article image

IRAP—imposta regionale sulk attivita produttive.

Addizionale regionale IRPEF.

ARISGAM—Additionale regionale imposts di constorno gas metano.

Tasse sulle concession regionali.

Taking into account population, relative size of potential tax base, health care needs of population, relative standardized cost of providing nonhelath care public services.

Table 15.

Italy: Structure of Tax Revenues in 1998

article image
Source: Fund staff calculations based on data in Bordignon and Emiliani (1999).

153. In order to provide a cross-country perspective, Box 1 summarizes the current tax assignment arrangements for a selected group of EU countries.

154. Of some concern is that the tax base of IRAP is the most unequally distributed tax base across the regions (see Table 16 and Figure 9). The ratio between the region with the highest and lowest per capita IRAP revenue is higher than 3.5, reflecting the highly unequal distribution of productive activities covered by the tax net and also the significant differences in the regional structure of economic activity.76 The regional personal income tax was the third most important source of tax revenue for regions in 1998, amounting to some 6.5 percent of total own tax revenue, less than 10 percent of the revenue from IRAP. The distribution of per capita tax revenue from the regional PIT is much less unequal than that from TRAP, indicating the extent of income redistribution through the central government budget. Revenue from the motor vehicle tax and the shared excise on gasoline produced some 19.5 percent of total own tax revenues of the regions in 1998. Per capita tax revenues from these taxes are even less unequally distributed than the revenue from the regional PIT.

Table 16.

Italy: Per Capita Revenues of Regions in 1998

(Average of Regions = 100)

article image
Source: Fund staff calculations based on data in Bordignon and Emiliani (1999).
Figure 9.
Figure 9.

Italy: Interregional Inequality in Regional Tax Bases and Total Regional Budget Revenue, 1998

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

The new system of equalization

155. The 1999 framework law (Law 133/1999, Art. 10) abolished almost all previously existing state transfers to ordinary statute regions, including the transfer from the FSN. The regions were fully compensated with the tax measures described above, including the sharing of VAT. The law also introduced an equalization mechanism, which takes into account the existing differences among regions in fiscal capacity, need for health care services, and fiscal effort. At the beginning, the law allows only for the redistribution of the VAT revenue allocated to the regions, but in the future the shared excise may also be redistributed through the equalization fund. The law envisages a transition period, during which equalization will be (partly) based on historical levels of expenditure.

Tax Assignments and Tax Autonomy at the Intermediate Level of Subnational Governments in Selected EU Countries

No single rule can explain the existing tax assignments of intermediate level subnational governments in the EU countries. They have been determined by a large number of social, political, and historical factors; moreover, they also depend on the expenditure responsibilities of subnational governments, which also greatly vary among EU countries. This Box summarizes the existing arrangements in those EU countries where intermediate level subnational governments—states, regions, counties, or communities—have significant tax revenues or enjoy a considerable degree of tax autonomy.

Lander in Austria receives some 10 percent of general government tax revenue, almost exclusively through shared taxes. Sharing arrangements are fixed by federal laws for 4-5 years and changes to the arrangements are always negotiated between the federal government and the Lander. The major-shared taxes are VAT, wage tax, PIT, tax on interest (withheld), motor vehicle tax, and tax on mineral oil. Lander in Germany obtain some 21.5 percent of general government tax revenue mainly through shared taxes—PIT, CIT, and VAT—but also through own taxes, such as, taxes on wealth, estate, and transfer of title. They have no legal power to set tax rates or define tax bases, but have a strong influence on the law-making process due to their representation in the upper house of the parliament. The tax assignments of the communities and regions in Belgium are derived from the constitution and constitutional laws. Communities, the main responsibility of which is education, receive some 13 percent of general government tax revenue, almost entirely through shared taxes. The sharing arrangements can only be changed by a law. Regional governments, which have wide ranging economic policy responsibilities, receive some 10 percent of general government tax revenue. The most important taxes they collect are a surcharge on the national PIT, an inheritance tax, and transaction taxes.

In Denmark, counties obtain some 9 percent of general government tax revenue, dominantly through a county PIT, the rate of, which is set, each year depending on the level of outlays. The rest is from a property tax, the tax base and rate of which are determined by national legislation. Counties in Sweden are in a rather similar position, they receive some 10.5 percent of general government tax revenue almost entirely through a flat rate county PIT with an average rate close to 10 percent. The tax rate is determined by the local council, but the tax is collected by the central government and transferred to the counties.

Spain has 17 autonomous regions and 2 autonomous cities that were formed on the basis of the new democratic constitution enacted in 1978. The constitution limits the tax autonomy of the regions. They receive some 5 percent of general government tax revenue, through ceded taxes—on wealth, inheritance, transfer of title, and duties—governed by central government legislation which gives some tax autonomy to the regions, and through shared taxes. The most important shared tax is the PIT, 15 percent of which is transferred as block grant. Another 15 percent of PIT is assigned to the regions, for which they can set the tax rate and can introduce tax credits. The rules governing the sharing of other tax revenues may change depending on the transfer of new responsibilities to the regions.

156. The actual amount of VAT revenue allocated to a region is determined by a formula which is the sum of four components. The first one is the share of the region in the total VAT revenue assigned to the regions according to the size of its population. The second one deducts 90 percent—the solidarity coefficient—of the difference between the potential own tax revenue of the region (excluding VAT) and the potential tax revenue the region would have if it had the average per capita potential own tax base.77 It is important to point out that the formula is based on the potential tax revenue and not the actual one—thus it sanctions lack of fiscal effort at the regional level.78 The third component adds the difference between the eligible health care expenditure of the region and the health care expenditure the region would have if it had the average eligible per capita health care expenditure. The Ministry of Health will determine the level of eligible expenditure by taking into account the defined uniform minimum standards, the regional need for health care services, and the acceptable costs of service delivery. This in principle should provide incentives for the regions to keep health care costs below the level determined as acceptable by the Ministry of Health. The last component adds 70 percent of the difference between the standardized nonhealth care cost of the region and the standardized nonhealth care cost the region would have if it had the average standardized cost. The standardization is based on a semi-logarithmic function fitted to historical per capita nonhealth care costs of the regions (as a function of population) and the formula uses the fitted value for the regions. This component is based on the empirical finding that there is a sizable economies of scale involved in the provision of nonhealth care public services in Italy. It does not reward the inefficiency of an individual region, as it uses the fitted values from a regression.

157. In 2001, the first year of the operation of the equalization fund, the actual amount of VAT revenue allocated to a region will be determined by its historical expenditure, that is, there will be no change in the financial situation of any of the regions. To put it differently, the amount of VAT revenue allocated to a region will be a weighted average of the amount determined on the basis of historical expenditure and the amount determined on the basis of the new formula, where the second component gets zero weight in the first year. Over the subsequent 12 years, the weight of the formula-based amount will gradually be increased, and by 2014 it will fully determine the amount of VAT allocated to a region.

Potential problems

Tax assignment

158. The conventional prescription of the public economics model (Bird, 1999) is less likely to produce sustainable results for Italy because regional governments have sizable expenditure responsibilities. Nonetheless, there are some well-established basic principles that tax assignments in Italy should follow (Bird, 1999 and Norregaard, 1997).

159. The new system of regional taxes in Italy meets the two basic criteria for subnational taxes. It makes the most affluent region practically fiscally autonomous, even if regional tax rates are set at the minimum levels stipulated in the law (Figure 10). Second, it imposes fiscal responsibility at the margin, at least on regions in the North. Nonetheless, there are several potential problems with this system.

160. In spite of its appealing technical features (Bird, 1999), IRAP does not seem an optimal choice as the main source of revenue for regional governments. As pointed out earlier, MAP has the most unequally distributed tax base among the regional taxes.79 This creates a large built-in need for horizontal equalization, much larger than the existing degree of interregional income inequality would necessitate (Figure 11)—at the same time, the tax with the most equally distributed tax base, the VAT, is used for equalization. (Moreover, the fact that the equalization fund is limited to the shared VAT revenue effectively limits the extent to which vertical imbalances can be reduced.) Furthermore, the heavy reliance on a local business tax is somewhat questionable, as its share in total revenue of regional governments is clearly not in line with the share of regional expenditure benefiting local businesses. The benefit argument would clearly support a higher regional PIT (with a concomitant reduction in national PIT).80

Figure 10.
Figure 10.

Italy: Projected Revenues of Regions in the First Year of the Transition

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.
Figure 11.
Figure 11.

Italy: Interregional Inequality in Income and Budget Revenue, 1998

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

161. On the other hand, TRAP has some appealing characteristics (Bird, 1999). Given its base, it is neutral with respect to factor mix and financing structure. Thus, it is not biased against employment and it is less subject to base erosion resulting from thin capitalization. It is less distortionary than some of the taxes it replaced (municipal business tax and local payroll taxes) and other forms of subnational business taxes common in other countries (such as, for example, nonresidential property tax). Its base is fairly wide, not only because it is an income-based value added tax, but also because it is paid by nonincorporated businesses as well. However, the present IRAP rate (range) appears rather high when compared to the rates on similar taxes in other countries.

162. The actual extent of fiscal autonomy is very limited in the poorer regions. Under the present legislation, regional governments can increase their total revenue (including the revenue from VAT) by some 16 percent on average by increasing the regional tax rates (for IRAP and the regional PIT) up to the allowed limits (Figure 12). This would seem to provide considerable room for maneuver. However, due to the unequal distribution of regional tax bases, rich regions can increase their revenues by up to 23.5 percent (Lombardia), while poor regions by only around 8 percent.81

Figure 12.
Figure 12.

Italy: Capacity to Raise Additional Revenue from Regional Taxes by Increasing Tax Rate

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

163. It is not clear whether the new system can impose fiscal responsibility and/or preserve equal access to public health care services. An 8 percent increase in real expenditure would require the poor regions to fully exercise their right to increase the IRAP and the regional PIT rates,82 while rich regions could achieve the same increase by considerably lower rate increase.83 This would result in further distortions in business location and a further deterioration of the growth potential of the poor regions—the mobility of the tax base of IRAP may effectively limit tax rate increases in the poor regions. Given the uncertainties in projecting health care expenditure and the likely extent to which the growth of health care expenditure will exceed the growth of the regional tax base, poor regions may not be able to exercise fiscal responsibility. A continued sizable difference in growth performance among the regions would make this problem even more severe.

164. The administration of regional taxes is relatively easy. VAT and the excise on gasoline remain national taxes, thus they impose no extra administrative burden. IRAP is relatively easy to administer due to the subtraction method and can be administered centrally through the newly established single-tax return form. The regional PIT taxes the same base as the national PIT (IRPEF) and it is administered centrally, thus its collection involves limited additional administrative burden. With the exception of the motor vehicle tax, there is practically no additional compliance cost involved in paying local taxes. The fact that none of the regional taxes is deductible limits the potential for vertical spillovers.

Horizontal equalization mechanism

165. Figure 13 shows the total revenues that the regions will actually receive in 2001 (including VAT revenue), based on historical expenditure, and the total revenues that they would receive if the new formula was already in force. Both are highly equalized and the differences between the two do not seem large at first glance. Over time, however, some of the regions that will lose revenue because of the new formula may find it difficult to adjust, unless they reduce their expenditure. Veneto and Lazio have adequate own tax bases, thus even if they could not reduce expenditure it would be relatively easy for them to offset the loss of revenue due to the new formula by slightly increasing their TRAP and/or regional PIT rates. The same is not true for Campania and Calabria, which would have to increase their rates quite substantially. Over the next 13 years, therefore, these regions would need to either grow faster than the economy as a whole—thereby enlarging their tax base—or acquire the administrative capacity necessary to lower costs.

Figure 13.
Figure 13.

Italy: Projected Total Revenues of Regions in the First Year of Transition

Citation: IMF Staff Country Reports 2000, 082; 10.5089/9781451819793.002.A003

Source: Fund staff calculations.

166. The new equalization mechanism has some appealing features. It links the total amount of money allocated to the equalization fund to the tax base of VAT, which is a rather elastic one. It effectively limits the extent of inequality in per capita total revenue across regions, due to the very high solidarity coefficient set by the law, and automatically redistributes resources when individual regions are subject to asymmetric shocks.84 It takes into account fiscal need and economies of scale in providing nonhealth care related public services, and promotes fiscal effort at the regional level.

167. However, its implementation requires the specification of a number of concepts defined in the law, which may give rise to technical difficulties and to disagreement between regions and central government. The potentially most contentious one is the definition of eligible health care expenditures, used to determine the average eligible cost and thus the distance from the average. Under the existing law on the national health service (legislative decree 502/92), this should be based on a uniform level of services to be provided across the country and an estimate for each region of the per capita cost of providing these services.

168. If the history of the allocation mechanism under the previous system85 is an indication, it will be extremely difficult to come to an agreement concerning this aspect of the equalization mechanism. The allocation formula between 1980 and 1997 changed practically every year, with five major changes that concerned the variables included in the formula and a change in the coefficients in the other years. Such frequent changes clearly undermine the credibility of any formula-based redistribution of resources. The frequent changes are partly explained by the scarcity of sound research results available on the basis of which such formula could be devised in Italy (CTSP, 1999). Thus, unless the required empirical research is carried out soon, this component of the equalization formula is likely to be subject to frequent changes.

169. Significantly faster growth in the North for a longer period, a phenomenon observed during the last decade, would put some pressure on the mechanism, though in the coming decade this would require a change in the design only if Lombardia grew significantly faster than the national average. The increase in regional tax rates that poorer regions would have to undertake in this case would however create further distortions in business location and would thus reduce their growth potentials.

170. The increase in health care expenditures that is projected to ensue from population aging could, however, pose a major threat to the sustainability of the system. A yearly increase in real health care costs 1 percentage point higher than real GDP growth would create a sizable imbalance in the system within a decade.86 While the affluent Northern regions could raise the required extra revenue by increasing their TRAP and regional PIT rates, none of the Southern regions would be able to do so. This may lead to a continuous reform of the tax assignment and the equalization mechanism, which could reduce the credibility of the system and thus produce perverse incentives for the regions.

F. Borrowing by Regions

171. Regions can borrow from financial institutions, issue marketable securities, and borrow from the central government. Borrowing is allowed only to finance investments and total debt service cannot exceed 20 percent of own revenues. Historically, regions have borrowed from commercial banks and from the Cassa Depositi e Prestiti, a special government financial institution created to finance subnational governments and which refinances the extended loans issuing postal bonds in the retail market. In 1995, regions were allowed to issue bonds with a minimum maturity of five years—bonds can be denominated in foreign currency, but only if the exchange rate risk is fully hedged. The total amount of new bonds issued by the regions increased rapidly since then, reaching Lit 4 trillion in 1999. At the end of 1999, the total debt of the regions reached Lit 24.7 trillion, or 1.2 percent of GDP.

172. One of the central elements of the recent reforms of intergovernmental fiscal relations is the gradual replacement of government transfers with own revenues. This will substantially increase the borrowing capacity of regional governments. Given that the Internal Stability Pact limits the primary balance excluding investment spending (see below), this could leave room for a sizable fiscal expansion at the regional level.

G. The Internal Stability Pact

173. Given the important role of subnational governments in determining the developments of Italian public finances, as well as their increasing degree of fiscal autonomy, it was natural for the government to attribute to them formal responsibility to contribute to meeting the fiscal discipline obligations undertaken by Italy in the context of EMU. This was done in 1998 by introducing, in the context of the medium-term economic and financial programming document (DPEF) for 1999-2001, the so-called Internal Stability Pact, and the relevant provisions were set out in Article 28 of Law 448 of December 23, 1998.

174. This section provides an assessment of the Italian Internal Stability Pact, and discusses the extent to which it can help reconcile the fiscal commitments under EMU with the increasing degree of fiscal decentralization. Box 2 briefly describes the equivalent arrangements prevailing in other selected EMU countries with a degree of fiscal decentralization comparable to Italy’s—the cross-country comparison underscores the inherent tension between fiscal autonomy at the subnational level and the need to safeguard fiscal objectives at the general government level.

175. As originally formulated, the Internal Stability Pact applied to the three-year period 1999-2001, and stipulated that local governments should (i) improve their fiscal balance by Lit 2.2 trillion, or 0.1 percent of GDP in 1999; and (ii) maintain their balance at least at this improved level over 2000-01. The Lit 2.2 trillion improvement in the consolidated balance was distributed among the different levels of subnational governments—regions, provinces, and municipalities—in proportion to their respective levels of total expenditures. Within each level, individual subnational governments would have to contribute to the achievement of the objective in proportion to their level of primary current expenditure. The improvement in the fiscal balance was defined in reference to the overall balance projected on a current services (unchanged policies/legislation) basis, as estimated in the government’s programming documents, and it implied an improvement of only some Lit 600 billion over the 1998 outturn. The fiscal balance was defined as in the Maastricht rules, except that it excluded investment spending. It should be pointed out that no sanctions were stipulated for cases of noncompliance, with one exception: should Italy incur the sanctions envisaged under the EC excessive deficit procedures, Law 448 states that these sanctions will be distributed across those general government entities which failed to fulfill their commitments.

176. During its first year of operation, 1999, the Internal Stability Pact is estimated to have yielded only half of the envisaged savings. Several modifications to the pact were therefore approved in late 1999, in the context of the new medium term plan for 2000-03 and the 2000 Budget Law. First of all, it was stipulated that subnational governments would improve their fiscal balance by Lit 3.3 trillion, or 0.15 percent of GDP in 2000, of which (i) Lit 1.1 trillion corresponding to the shortfall in savings for 1999; and (ii) Lit 2.2 trillion corresponding to a more ambitious target for 2000 than originally envisaged. The improvement achieved over 1999-2000 should then be maintained over the following three years (2001-03). Second, the revised Internal Stability Pact stipulates some changes to the relevant definition of the overall balance, including, importantly, that receipts from the sale of real estate assets should now be excluded from the definition of revenues. Third, following the revisions to the relevant definition of the overall balance, subnational governments are given a choice of three options under which their compliance with the Internal Stability Pact could be determined: (i) they could have their 1999 performance measured according to the old criteria, and the 2000 performance according to the new criteria; (ii) they could have their 1999 performance reassessed according to the new criteria; or (iii) they could have their performance assessed on a cumulated basis for the 1999-2000 period according to the new criteria.

Internal Stability Pacts in Selected EU Countries

In most member states the fiscal target set out in the Treaty on the European Union and the Stability and Growth Pact implies a sizable fiscal adjustment which is difficult to achieve without some form of coordination between the central and subnational governments. The actual form of the coordination, the way in which the fiscal targets for the different levels of government are reached and documented, depend on a number of factors. In some countries it is a formal agreement between the central and subnational governments on how to distribute the burden of fiscal adjustment. The importance of such an agreement depends on the extent of the required adjustment and the size of subnational governments, while its legal form and technical details depend on the legal status of subnational governments and the existing legal limitations on budget deficit and borrowing at the subnational level.

The first arrangements between the national and regional governments in Spain were introduced in March 1992 in the context of the convergence program. Though some of the details have been changed since then, the arrangements presently in force for the period 1997-2001 are very similar to the original ones. The arrangements are agreements between the central government and each subnational government, both at the regional and local level—there is no legal basis for their enforcement. The overall target for the consolidated fiscal position of the subnational governments, as well as individual targets in terms of overall deficits and debt, were agreed upon through negotiations that took into account the evolving expenditure responsibilities and revenue assignments of the subnational governments involved. The agreements involve no sanctions or incentives, but debt issuance by subnational governments requires the approval of the central government, giving a leverage to the central government over those subnational governments that do not comply. Despite the lack of a formalized enforcement mechanism and sanctions, subnational governments have met their targets in recent years.

Regions and communities (so-called federated entities forming the intermediate level government) in Belgium are subject to the supervision of the High Finance Council, which issues an annual report on the financial situation of each entity, examining its financial needs and recommending a maximum ceiling for its deficit. In its 1999 annual report, the High Finance Council set out recommendations for the overall balances of the federated entities for the period 2001-10. For entities in deficit (all regions and communities except the Flemish Community), a gradual move towards a balanced fiscal positions was recommended. In late 1999, at the regular consultation of the federal government, the regions and the communities, these recommendations were accepted as policy targets for the period 2000-03. The agreement is only political in nature; legal sanctions can only be imposed lithe fiscal policy of an entity puts at risk the internal or external equilibrium of the national economy. Peer pressure among regions and communities is however an effective political mechanism through which agreements can be enforced. In addition, as mentioned above, regions and communities are subject to an annual examination by the High Financial Council. Local governments (provinces and municipalities) are subject by law to a balanced budget constraint.

In Finland, though subnational governments enjoy a very high degree of fiscal autonomy—their combined shares in general government revenue and expenditure are very high, and they face practically no restriction on borrowing—there is no formal agreement setting out targets for the fiscal positions of subnational governments. The reason is that local governments (practically the only existing level of subnational government) have traditionally pursued very conservative fiscal policies, thus the need for such agreement did not arise.

177. No new sanctions for noncompliance have been approved, however. As an incentive to comply, on the other hand, it has been decided that subnational governments meeting the objectives of the Internal Stability Pact will be granted a reduction on the interest on their outstanding debts to the Cassa Depositi e Prestiti. Even this incentive, however, seems undermined by the recent decision to grant to all regions an unconditional reduction. Against the background of increased devolution of fiscal responsibility to local governments, the Internal Stability Pact seems still to be interpreted more as a framework to foster cooperation between central and local governments rather than as an instrument for the former to impose fiscal discipline on the latter.

178. The fairly rigid rules imposed by the Maastricht Treaty, relying on numerical parameters that have to be observed ex post, inevitably creates a tension with the flexibility involved in delegating fiscal responsibility to subnational governments. An important question is therefore to what extent the Internal Stability Pact can help Italy reconcile the fiscal commitments undertaken with membership in EMU with the increasing degree of fiscal decentralization.87 In this regard, the Internal Stability Pact in its current form is less than ideal in several respects. First and foremost, the lack of strong and credible sanctions leaves wide room for free riding by individual subnational governments. As was mentioned above, the pact stipulates that, should Italy incur in the financial penalties implied by the excessive deficit procedure, the subnational governments that failed to comply with the Internal Stability Pact would have to share the penalty in proportion to their degree of noncompliance. However, the reputation cost that would ensue from breaching the Maastricht deficit ceiling would be felt disproportionately by the central government. This could generate in subnational governments the expectation that any overshooting on their part would be compensated by the state. Even the financial penalty may not be credible in the case that few subnational governments were responsible for the breach, as their share of the penalty could be excessive. Moreover, to the extent that a breach of the Maastricht ceiling appears unlikely, the current system fails to provide any sanctions against free riders. A second shortcoming of the Internal Stability Pact lies in the fact that it targets a definition of the fiscal balance different from the Maastricht definition, which is the one that must ultimately be kept under control at the general government level. A third limitation of the pact is that it does not attempt to deal with the impact of the cycle on the budgets of subnational governments, just as new changes to the tax assignment framework make these budgets more cyclically sensitive than ever before. Finally, distributing the required effort across subnational governments in proportion to their levels of primary current expenditures appears questionable: not only does a government which is already running a surplus have to contribute to the effort, but if its level of current primary expenditures is high, it may be required a larger effort than a government running a (perhaps large) deficit.

H. Conclusions

179. The reform of intergovernmental fiscal relations has so far concerned the tax assignment of the regions and the system of horizontal equalization. It has created a larger regional tax base and a transparent and, at least at the outset, adequately sized system of horizontal equalization. These steps could go some way towards strengthening transparency of intergovernmental relationships. However, the reform has not yet addressed the need for stemming the future expenditure demands related to population aging. Moreover, there are several elements of the new system of intergovernmental relations that may limit the positive impact of the reform.

180. The increased regional tax base is very unevenly distributed, because of the large share of IRAP in total own revenues of the regions, thus limiting the ability of poorer regions to raise additional revenues and forcing them to rely on the transfer from the equalization fund to a considerable extent. Given their initial expenditure structure, these regions will thus have very little real fiscal autonomy. At the same time, richer regions will be able to generate substantial amounts of additional revenue by exercising their tax autonomy, and will hence be able to provide better services, if their constituencies decide to do so. This could exacerbate the regional differences in provision of health care services—although the setting of minimum standards should still guarantee nationwide access to a set of core services.

181. The extent to which fiscal responsibility can be fostered in the poorer regions is also circumscribed by their limited capacity to finance an increase in expenditure. Continued slower growth in the South or a rapid increase in health care costs, two developments not unlikely to take place in the coming years, could further aggravate this problem. While in principle the system can cope with the consequences of continued slow growth in the South, the required regional tax rate increases would further reduce the growth potential of regions in the South and thus eventually undermine the stability of the system. Imposing fiscal responsibility upon the regions could be self-defeating in such a case.

182. Increasing the regional tax base or the size of the equalization fund to finance rising health-care costs could create two problems. First, it might reduce the credibility of the system. In order to foster fiscal responsibility at the regional level, the system has to establish hard budget constraints, that is, it has to limit the tax base assigned to the regions. An increase in the tax base may be justifiable only to the extent that the expenditure increase stems from factors beyond the regions’ control, and it may be hard to identify this in a credible manner. Second, without a change in the design, it will be technically difficult to increase the resources available to the regions in a way that properly targets the regions in need of additional revenue. Changing the design would, however, also raise the issue of credibility.

183. Assigning the responsibility of financing and managing the health care system to the regions may increase the allocative efficiency and the productivity of the health care system but—similarly to the decentralization of other fiscal activities—only if certain conditions are met (Tanzi, 1995). Decentralization will lead to an improvement in public services only if subnational governments have the necessary administrative capacity to manage the provisioning of the services in an efficient manner. The initial situation in some of the regions may be far from this (Putnam, 1993 and Tanzi, 1995). In fact, it seems that institutional capacities will be needed most in regions where they are least available. The quality of public expenditure management at the regional level, a particularly important aspect of the quality of regional administrations, is also a factor that may limit the benefits from decentralization (Tanzi and others, 1994). Finally, it is questionable whether the spatial characteristics of health care services make the regions the right level of government to assign this responsibility to (in this respect, however, there is very little choice because the responsibility for providing health care services is assigned to the regions in the constitution).

184. Decentralization is deepening at a time when the capacity of the regional budgets to contain the increase in costs is likely to be critical. As argued above, due to the way the health care system has been managed so far, even regions with strong administrations have little experience with this, compared to the central government, and there is a risk that regions may be less able to exert administrative pressure on the health care system to contain costs.

185. Therefore, in order to make the recent reform successful and to fully utilize the sizable potential gain from decentralization, measures aimed at building the necessary institutional capacity at the regional level are urgently needed. A radical improvement of human resource management in the public sector is an important building block of the Bassanini reform; it is of great importance to ensure that such improvement takes place also at the regional level, especially in the regions in the South.

186. A comprehensive reform of the health care system, which will now be designed and carried out at the regional level, is also needed. The reform should create strong incentives for the local health units to reduce costs. It should also create mechanisms through which best practices can be transferred to regions where local health units are less efficient. This should result in a higher level of transparency concerning the performance of individual local health units, allowing regional administrations to rely on nationwide benchmarks when assessing the performances of local health units in their regions.

187. Finally, faster and more evenly distributed growth is perhaps the best way of making the reform of intergovernmental fiscal relations successful. Policy measures aimed at achieving this—discussed in the main report—appear even more important in this light.

APPENDIX: The Main Regional Taxes

188. IRAP (imposta regionale sulle attivita produttive) is a regional tax on productive activities which was introduced in 1998 to replace a municipal business tax (ICIAP), a tax on dividend payments, a net worth tax, and a payroll health care contribution. It is levied on companies, partnerships and individuals carrying out business activity. Unlike the previously existing health care contributions, it is not deductible from the taxable base for the national corporate income tax (IRPEG). The tax base of IRAP is the value added (net of depreciation) produced in a region. In case a taxpayer carries out business activities in several regions, the value added is allocated according to the labor cost in each region. There are special rules for financial institutions. The uniform tax rate is 4.25 percent. Banks and insurance companies pay a higher rate, while agricultural businesses pay a preferential rate. Both nonstandard rates will be phased out gradually. Starting in 2001 regions will have the right to increase the rate of IRAP by up to 1 percentage point and also to establish differentiated tax rates for certain types of businesses.

189. The 1997 tax reform introduced a regional personal income tax (addizionale regionale IRPEF). The regional personal income tax (regional PIT) taxes the same base as the national progressive personal income tax (IRPEF), but at a flat rate which the regional governments could originally set between 0.5 and 0.9 percent. To create the room for the regional income tax, national income tax rates were reduced by 0.5 percentage point. The recent reform has increased the range to 0.9 to 1.4 percent, and reduced national rates accordingly (by 0.4 percentage points). The law gives the right to collect the regional income tax to the central government, which then transfers it to the regions, but tax payment withheld at source is directly transferred to the region. The regional PIT is not deductible under the national income tax.

190. Since 1995, the revenue from excise taxes on gasoline is shared between the central government and the regions. The framework law passed in 1999 allows for up to Lit 450 per liter to be allocated to the regions, but the government legislation implementing the law allocated only Lit 250, Lit 8 more than previously. Regions have the right to increase excise tax on gasoline by up to Lit 50 per liter.

191. Regions levy a tax on motor vehicles registered and operated by resident companies and individuals (Tassa regionale di circolazione). The tax base is the engine performance of the car and the rate set by the central government can be increased by up to 10 percent by the regions.

References

  • Alesina, A., S. Danninger, and M. Rostagno, 1999, “Redistribution Through Public Employment: The Case of Italy,” IMF Working Paper 99/177, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Balassone, F. and D. Franco, 1999, “Fiscal Federalism and the Stability and Growth Pact: A Difficult Union,” Working Paper (Bank of Italy).

    • Search Google Scholar
    • Export Citation

  • Bird, R. M., 1999, “Rethinking Tax Assignment: The Need for Better Subnational Taxes,” mimeo.

  • Bordignon, M. and N. Emiliani, 1999, “Federalismo e perequazione in Italia, a proposito dell’articolo 10 della legge 133/1999,” Rapporto preparato per it Consiglio Nazionale dell’Econornia e del Lavoro, Roma, September 1999.

    • Search Google Scholar
    • Export Citation

  • Corte dei Conti, 1999, Relazione sulfa gestione finanziara delle regioni a statuto ordinario per I’anno 1997 (Roma).

  • CTSP, 1999, “L’Alliocazione delle Risorse nel Servizio Sanitario Nazionale,” Commissione Tecnica per la Spesa Pubblica, Ministero del Tesoro, del Bilancio e della Programmazione Economica.

    • Search Google Scholar
    • Export Citation

  • Emiliani, N., S. Lugaresi, and E. Ruggiero, 1997, “Italy,” in Fiscal Federalism, ed. by T. Ter-Minassian, pp. 249–284.

  • Giarda, P., 2000, “Il federalism fiscale in attuazione della Legge n. 133/1999: aspetti tecnici. ragioni e problemi aperti,” revised version, January 2, 2000.

    • Search Google Scholar
    • Export Citation

  • Norregaard, J., 1997, “Tax assignment,” in Fiscal Federalism, ed. by T. Ter-Minassian, pp. 49–72.

  • OECD, 1994, “The Reform of Health Care Systems. A Review of Seventeen OECD Countries,” Health Policy Studies No. 5 (OECD).

  • Pintus, E., 2000, “Government Reform in Italy,” Working Papers (Joint Program Indes-European Union, Inter-American Development Bank).

    • Search Google Scholar
    • Export Citation

  • Putnam, R. D., 1993, Making democracy Work: Civic Traditions in Modern Italy (Princeton: Princeton University Press).

  • Reviglio, F., 2000, “A System in Need of Reform: Health Care in Italy,” mimeo, (Washington: International Monetary Fund).

  • SVIMEZ, 2000, I Conti Economici delle Regioni lialiane dal 1970 al 1998 (II Mulino).

  • Tanzi, V., 1995, “Fiscal Federalism and Decentralization: A Review of Some Efficiency and Macroeconomic Aspects,” Annual Bank Conference on Development Economics, The World Bank, Washington, D.C., May 1995.

    • Search Google Scholar
    • Export Citation

  • Tanzi, V., A. Premchand, B. Potter, S. Lugaresi, and D. Demekas, 1994, “Italy: The Management of Public, Spending,” IMF Fiscal Affairs Department.

    • Search Google Scholar
    • Export Citation

  • Ter-Minassian, T., ed., 1997, Fiscal Federalism, (Washington: International Monetary Fund).

54

Prepared by Marco Annunziata and Istvan Szekely.

56

For a detailed description of the Bassanini reform, see Presidenza Consiglio dei Ministri (2000). For an analytical assessment, see Pintus (2000).

57

Besides those of the regions, the table also includes the accounts of the “autonomous” provinces of Trento and Bolzano, which share some of the responsibilities of regions, notably for the provision of health services.

58

Health expenditures consist for about 60 percent of purchases of goods and services, and for about 40 percent of payments of wages and salaries.

59

For a general description of the Italian health care system, see OECD (1994).

60

The different trends in public and private health care expenditure between 1980 and 1985 also suggest that in periods of radical cuts in public health care expenditure the demand for private health care increases, partly offsetting the decline in public expenditure. The rapid increase in private health care expenditure in the first half of the 1990s reflected importantly an increase in co-payments.

61

Actual expenditures were 4.7 and 7 percent higher, respectively, than assumed in the budget.

62

Expenditure numbers for public health care costs are on an accrual basis, thus they include arrears to suppliers. The part of total expenditure, which was not financed by the budget, reached 2.7 and 8.8 percent in 1996 and 1997, respectively.

63

As the comparison of nonadjusted and adjusted relative per capita expenditures for the different regions show, regions in the South are net beneficiaries of hospital care provided to their residents in the northern-central regions. The total size of this expenditure is rather large.

64

For a discussion of the role of the special statute regions, see Chapter II in SM/99/115, 5120/99.

65

Per capita expenditure on pharmaceuticals dispensed by pharmacies was 56 percent higher in Campania than in Lombardia. In 1995-97, expenditure on pharmaceutical increased rapidly. Though the rapid increase was common to all regions, the increase in the South was significantly higher (see Corte dei Conti, 1999).

66

Overstaffing in public services is a widespread phenomenon in the South. Recent empirical research (see Alesina and others, 1999) suggests that excessive public employment in the South was an important mechanism through which budgetary resources have been channeled to the South and though which the government tried to reduce unemployment.

67

It is worth mentioning that the rate of increase in NHS personnel in the special statute regions in the North was in fact higher than in the South, and at present these regions have the highest per capita levels of public health care employment in Italy.

68

The outflow of hospital patients from Southern regions has increased steadily in recent years.

69

Medical tests taken in public laboratories and medical examinations carried out in public health care facilities (see Table 4 in Alesina and others, 1999).

70

Survey results on the subjective evaluation of the quality of public services show a strongly significant difference between the South and the North (see Table 5 in Alesina and others, 1999), with public health care services rated as one of the lowest quality public services in the South.

71

Reliable microlevel data, on which empirical analysis of these factors could have been (or could now be) based, were (and still are) not available, at least not to the experts outside the Ministry of Health care (see CTSP, 1999; Reviglio, 2000).

72

Empirical research (Alesina and others, 1999) suggests that the public sector wage premium was above 26 percent in the South, compared to 19 percent at the national level, mainly explained by the rather uniform level of public sector wages throughout the country and the substantial difference in cost of living between the South and the Center-North.

73

For a detailed discussion of intergovernmental fiscal relations prior to the recent reforms, see Emiliani and others (1997).

74

The table does not account for the different types of health care contributions, which were transferred to the regions in 1994, but were earmarked to finance health expenditure. In 1998, IRAP was also earmarked to finance health care expenditure.

75

In addition to the taxes shown in the table, there are a large number of minor regional taxes and fees, which are not discussed (such as for example, the fee on solid waste disposal, or the sharing of university enrollment fees). For a detailed list of taxes and fees collected by the individual regions, and for the revenue from these sources, see Corte dei Conti (1999), Volume II. Local health units collect a nonnegligible amount of coinsurance payment, which is basically a user fee for a service financed by the regional governments. Nonetheless, these payments are not recorded in the accounts of the regions.

76

Agricultural businesses are subject to lower IRAP and the share of agriculture in GDP is significantly higher in the poorer regions.

77

Average of the ordinary statute regions participating in the scheme.

78

So far, this only applies to the motor vehicle tax, as the other taxes are collected by a centralized tax administration, but in the future, a larger role is envisaged for the regions in the collection of MAP.

79

One of the newly established criteria for subnational taxes is that its tax base should be relatively evenly distributed (Norregaard, 1997).

80

A higher regional PIT minimum rate had initially been considered. However, a significantly higher PIT rate would have resulted in a net increase in own revenue for Lombardia higher than the phased-out transfers. Given the principle of revenue neutrality, this would have resulted in a net transfer of budgetary resources from the central government to the regions, even though there was no change in the expenditure responsibilities of the regions.

81

Calabria, the poorest ordinary statute region, by slightly more than 6 percent.

82

By 1 and 0.5 percentage point, to 5.25 and 1.4 percent, respectively. Calabria would not be able to reach such increase under the existing rules.

83

If it is assumed that they increase TRAP and IRPEF revenue proportional to the tax base, the increase in IRAP and IRPEF surcharge rates would be 0.5 and 0.1 percentage point, respectively.

84

Simulation results (Giarda, 2000) show that the equalization mechanism can cope with rather sizable negative and positive asymmetric shocks to individual regional economies.

85

Which was based on transfers from the budget playing practically the same role as the redistributed VAT revenue

86

Results referred to here are from a simulation exercise which assumes the same GDP growth rat F, as the baseline scenario (for the period 2001-10) in the medium-term simulations presented in Chapter II.

87

For a more detailed discussion of the interaction between international fiscal obligations and domestic intergovernmental fiscal relationships, see Balassonne and Franco (1999).

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Italy: Selected Issues
Author:
International Monetary Fund
Volume/Issue:
Volume 2000: Issue 082
Publisher:
International Monetary Fund
ISBN:
9781451819793
ISSN:
1934-7685
Pages:
141
DOI:
https://doi.org/10.5089/9781451819793.002
  • Figure 1.
    View in gallery
    Figure 1.

    Italy: Public Expenditures, 1977-99

    (In percent of GDP)

  • Figure 2.
    View in gallery
    Figure 2.

    Italy: Composition of General Government Deficit, 1977-99

    (In percent of GDP)

  • Figure 3.
    View in gallery
    Figure 3.

    Italy: The Share of Health Care Expenditure in GDP, 1960-97

  • Figure 4.
    View in gallery
    Figure 4.

    Italy: The Share of Total Health Care Expenditure in GDP, 1970-97

  • Figure 5.
    View in gallery
    Figure 5.

    Italy: Average Annual Growth Rates of Real Current Health Care Expenditure, 1961-95

  • Figure 6.
    View in gallery
    Figure 6.

    Italy: Health Care Expenditure and Aging, 1970-97

  • Figure 7.
    View in gallery
    Figure 7.

    Italy: Health Care Expenditure and Aging in the Rest of the European Union, 1970-97

  • Figure 8.
    View in gallery
    Figure 8.

    Italy: Inequality Among Regions in Income and Health Care Expenditure, 1997

  • Figure 9.
    View in gallery
    Figure 9.

    Italy: Interregional Inequality in Regional Tax Bases and Total Regional Budget Revenue, 1998

  • Figure 10.
    View in gallery
    Figure 10.

    Italy: Projected Revenues of Regions in the First Year of the Transition

  • Figure 11.
    View in gallery
    Figure 11.

    Italy: Interregional Inequality in Income and Budget Revenue, 1998

  • Figure 12.
    View in gallery
    Figure 12.

    Italy: Capacity to Raise Additional Revenue from Regional Taxes by Increasing Tax Rate

  • Figure 13.
    View in gallery
    Figure 13.

    Italy: Projected Total Revenues of Regions in the First Year of Transition

Figure 1.

Italy: Public Expenditures, 1977-99

(In percent of GDP)
Figure 2.

Italy: Composition of General Government Deficit, 1977-99

(In percent of GDP)
Figure 3.

Italy: The Share of Health Care Expenditure in GDP, 1960-97
Figure 4.

Italy: The Share of Total Health Care Expenditure in GDP, 1970-97
Figure 5.

Italy: Average Annual Growth Rates of Real Current Health Care Expenditure, 1961-95
Figure 6.

Italy: Health Care Expenditure and Aging, 1970-97
Figure 7.

Italy: Health Care Expenditure and Aging in the Rest of the European Union, 1970-97
Figure 8.

Italy: Inequality Among Regions in Income and Health Care Expenditure, 1997
Figure 9.

Italy: Interregional Inequality in Regional Tax Bases and Total Regional Budget Revenue, 1998
Figure 10.

Italy: Projected Revenues of Regions in the First Year of the Transition
Figure 11.

Italy: Interregional Inequality in Income and Budget Revenue, 1998
Figure 12.

Italy: Capacity to Raise Additional Revenue from Regional Taxes by Increasing Tax Rate
Figure 13.

Italy: Projected Total Revenues of Regions in the First Year of Transition