Documento de Trabajo - 2018/09

Ambition beyond feasibility? Equalization transfers to regional and local governments in Italy

Giorgio Brosio (Universita´di Torino)

fedea

Las opiniones recogidas en este documento son las de sus autores y no coinciden necesariamente con las de FEDEA.

Ambition beyond feasibility? Equalization transfers to regional and local governments in Italy

Giorgio Brosio Dipartimento Economia e Statistica. Universita’ di Torino. Italy

July 2018

Abstract

The paper illustrates the present system of intergovernmental transfers in Italy with a focus on equalization grants. There are presently two separate systems operating in the country. The first one allocates funds to the regions, while the second one finances municipalities. The constitution of 2001 sets the principles for both systems. It introduces the obligation for the state to determine and ensure, by providing adequate financing, uniform essential levels, i.e. standards, of provision across the whole country for a set of basic services assigned to regions and local governments. The implementing legislation (the so called Fiscal Federalism Law of 2009) mandates the use of standard expenditure needs and of fiscal capacity for the allocation of equalization transfers. These are extremely ambitious aims, only partially achieved.

The Regional Health Fund is the dominant regional fund and is allocated, fundamentally, on a slightly modified per capita basis, upon agreement reached among regions and the central government within the State-Regional Governments Conference. The Municipal Solidarity Fund, in existence since 2014, shows more determination to implement the legal mandates. Allocations are determined on the basis of standardized expenditure and fiscal capacity. The Fund has also been subject to almost yearly changes and evolves in a framework of restructuring of the finances of the public sector characterized by continuous changes that make the system extremely difficult to manage, and also to understand.

The paper presents the various steps for the determination of individual allocations and evaluates the merits, but also the difficulties and incongruities of the procedures adopted.

Key words Intergovernmental fiscal relations Fiscal federalism in Italy Equalization transfers

Prepared for the jornada sobre Financiación territorial en España y otros países. Instituto de Estudios Fiscales y FEDEA. April 13, 2018. Instituto de Estudios Fiscales 218Avda. Cardenal Herrera Oria 378. Madrid

Introduction

This paper illustrates the present system of intergovernmental transfers in Italy with a focus on equalization grants. There are presently two separate systems operating in the country. The first one allocates funds to the regions, while the second finances municipalities. The constitution of 2001 sets the principles for both systems. It introduces, more specifically, the obligation for the state to determine and ensure, by providing adequate financing, uniform essential levels, i.e. standards, of provision across the whole country for a set of basic services assigned to regions and local governments. The implementing legislation (the so called Fiscal Federalism Law of 2009) mandates the use of standard expenditure needs and of fiscal capacity for the allocation of equalization transfers.

This is a quite ambitious mandate. Determining standard expenditure needs according to these legal mandates implies, first, the definition of standards of service provision for all the governments units concerned. It requires, as a second step, the definition of standard costs for each of these standard levels of service provision. This is a difficult task for complex sectors, such as health, which is the main responsibility of Italian regions. The task becomes even more difficult in the case of municipalities, considering their huge number (more than 8000 units), the range of services they provide and their diversity.

Actual equalization transfer systems for regions and municipalities are very different one from the other. They are distant from the model envisaged in the legal mandates. This applies especially to the regional Health Fund that is the dominant regional fund and is allocated, fundamentally, on a slightly modified per capita basis, upon agreement reached among regions and the central government within the State-Regional Governments Conference. The Municipal Solidarity Fund, in existence since 2014, shows more determination to implement the legal mandates. However, the implementation of standards needs – the most ambitious aim of the legal discipline – is rather partial. The allocation of the Fund is done with the participation of the Association of Municipalities and with the use of standard statistical techniques. The Fund has also been subject to almost yearly changes and evolves in a framework of restructuring of the finances of the public sector characterized by continuous changes that make the system extremely difficult to manage, and also to understand.

The paper is focused on equalization transfers with only essential references to the other sources of revenue. The paper is structured along five main sections. The first one illustrates the legal framework for the allocation of grants. The second one provides the analytical framework by presenting the principle of interjurisdictional equity. The fourth and the fifth sections illustrate the equalization transfers systems for municipalities and regions. The fifth one presents the stakeholders and the system of conferences.

1. The legal framework for equalization grants

The Italian, rather complex, story begins with the constitutional review of 2001. The revised constitution assigns (Article 117) to the central government the determination of the basic level of benefits relating to civil and social entitlements to be guaranteed throughout the national territory. Determining basic level of benefits implies, explicitly, the introduction of standards of service provision. No specific services are mentioned. However, the constitution has clearly in mind health and education, which in turn are assigned to the regional governments. Education has not been taken over by the Regions until presently. The constitution also delegates (Article 117.m) the central government to determine the fundamental functions of the Municipalities, Provinces and Metropolitan Cities.

The constitution introduces grants of various types. Art. 119, paragraph 5) introduces a first type of grant by mandating, specifically, the creation on an equalization fund to the benefit of the territories having lower per capita taxable capacity.

A second type of grant is targeted to promote economic development along with social cohesion and solidarity. It includes grants from EU programs and the corresponding Italian co-financing transfers. It includes also specific grants from the central government targeted at filling regional disparities in growth.

The third category of grants, the most interesting to our purposes, derives implicitly by the combination of the above mentioned introduction of basic level of benefits and the general provision that revenues raised from various sources shall enable municipalities, provinces, metropolitan cities and regions to fully finance the public functions attributed to them. As a matter of fact, reference to basic levels of benefits (standards) amounts to the recognition that the grant for these functions has to be based on standard expenditure needs in addition to taxable capacity (as for the first type).

Subsequent national legislation, basically Law 42 of 2009 and Law 23 of 2011, implements the constitutional mandates, with some innovation, but also with some deviation and complication. It defines the fundamental functions of municipalities. They are administration, local police, civil register, provision of urban services, such as street cleaning, lighting, sanitation, garbage collection and disposal. They, also, include building and management of school premises, and very poorly defined social services (whose assignment to municipalities is not specified in the constitution).

Law 42 also mandates the introduction of a municipal equalization fund aimed at filling, with reference to fundamental functions, the difference between “standard needs” and (standardized) fiscal capacity. For the remaining (non-fundamental) functions the law mandates that equalization will fill only differences in fiscal capacity (with no consideration to expenditure needs). There is no definition of standards needs, but the idea seems that their determination should consider two distinct elements: a) standards of service provision (basic level of benefits) and, b) standard costs associated to the basic level of benefits.

A short reference to the literature, more specifically to the principle of interjurisdictional equity may help to understand the rationale, or even better the ambitions, of the mandated municipal equalization fund.

2. Inter-jurisdictional equity

A general formulation of the principle says that persons in comparable circumstances, for example same age and same income, should have access to comparable public services in all localities. In other words, in an intergovernmental framework equity implies that residence should not create differences between citizens in their access to the public services and to the cost of access.

The principle is represented in the following formula that refers, for simplicity sake, to the case where only one service, such as education, is provided locally and is financed by only a revenue instrument, such as an income tax.

\[\frac {E _ {c , d , e f , \ldots j}}{R _ {j , w y}} = k, \quad \text { for each local jurisdiction }\]

Where:

E is the expenditure for the service;

c, d, e, f, …, is a set of characteristics that define the quality and quantity of the service and impact on the expenditure needed for its provision;

R is the revenue source assigned to the financing of the service.

w and y are parameters applied to the tax or the levy financing the service. One example would be the tax rate, or the exemption of basic income;

k is the equity parameter.

Inter-jurisdictional equity is reached when the parameter k is equal across all the jurisdictions. This means that individuals residing in different localities will be subject to the same proportional difference between what they receive in terms of health care and what they pay for it. To ensure full equalization, a transfer, ?, is needed that is equal to the difference between expenditure and revenue, as determined by the parameters.

\[T = E _ {c, d, e f,.. j} - R _ {j, w y}\]

The role of parameters c, d, e, f referred to the expenditure and w and y referred to the tax is essential to determine the comparability of personal circumstances. To make an example, c may be the number of pupils per class, d the provision of a free lunch to students, w the tax rate levied. The transfer will then cover the difference between the cost of providing the same level of service everywhere and the tax, whose burden is proportional to income, also everywhere. Filling this difference is the main target of equalization grants to Italian municipalities that are illustrated in the next section

The Municipal Solidarity Fund

We have to wait until 2013 to see the implementation, in a revised form, of the municipal equalization fund envisaged by the law 42 of 2009. More precisely, the so-called Stability Law for 2013 (Law 228 of 2012) mandates the creation of the Municipal Solidarity Fund (MSF). Its present functioning is, however, determined by the subsequent Stability Law for 2015.

The restructuring of the public finances has contributed to shaping it. This applies, especially, to the high contributions asked to municipalities and implemented through recurrent and always deeper changes in central government grants. The MSF is now mostly a horizontal equalization fund fed by a varying percentage - 22,43% for 2017- of the municipal property tax (IMU/TASI) with almost no supplementary contribution from the central government. Its equalization intensity has been also watered down, although, in principle, it is going to be increased during the transition period supposed to be extended until 2021.

The MSF has three components:

a) the first one is the so-called historical share, It is based, for each municipality, on the amount of central transfers allocated to it before the introduction of the MSF, minus its share of IMU/TASI going to the Fund. This component may be negative, and this happens in the reality, as we will see later.

b) the second one is, in principle, the crucial component. It is aiming at filling, for the fundamental functions, the gap between standard needs and fiscal capacity. Let’s call it where is standard needs and is standard capacity and i stands for municipality.

c) the third component aims a equalizing fiscal capacity with reference to non fundamental functions , where ?? is the average fiscal capacity for the all country.

The total fund accruing to each municipality is the weighted sum of the three components . The weights ?, ? and ? add up to one and are determined applying a number, exactly, four parameters, listed in table 1, whose function is to ensure a gradual transition from the old to the new system, to smooth the intensity of the equalization performed by the Fund and to determine the relative size of the second and third component.

Table 1. Structure of Municipal Solidarity Fund 2015-2021

Source: Marchionni, Pollastri and Zanardi (2017).

A first parameter, α , determines the size of the historical component. It means that each municipality is entitled to keep 80 per cent of its historical component (obviously, if positive). It has set up at an initial very high (although over time decreasing) level, 80%, crystallizing de facto the present situation. It should reach a zero level after 2021. A second parameter, works in the same direction. It determines the percentage of fiscal capacity, of each municipality, that is taken into consideration in the equalization process. Its value is 45,8 percent. When β is combined with α, it raises the total weight of the historical component to a level that absorbs most of the whole fund, as shown in Table 1 above. For example in 2017 α and β, when combined together make up exactly 80 percent of the total fund. The third and the fourth parameter, ϒ and 1- ϒ determine, respectively, the relative shares of the two equalization components. More precisely, ϒ, the weight assigned to the basic functions, is set at 70 percent. It defines the weight, 1- ϒ, assigned to the remaining (non basic) functions for which only fisca capacity is taken into account.

Standard expenditure needs are determined according the so-called Regression Cost-Based Approach (RCA), according to which needs are defined with reference to past expenditure. As a matter of fact, this approach does not determine standard needs, but rather standardized expenditure, because it does not rely on government determined standards of quality or quality of expenditure, but rather on the factors that contribute to determine the actual expenditure made by local governments. For example, standardized expenditure for education is what a municipality would spend for the provision of this service taking into account the factors that in the practice of the country impact on education, such as the size of classes and the elevation over sea level of schools, and the average efficiency (they way municipalities respond to the impact of these factors). The distinction is not simply terminological. The determination of standard needs would require, it is better to stress it again, the determination by the central government for each individual service of (minimum? full?) standards of service provision and the costs associated to these standards. However, this has not been done and it is unlikely it will ever be done, considering the political and information complexity of the issue.

According to the Regression Cost-Based Approach standard expenditures of each municipality are the expected value of the model. There are three steps. The first one is a typical expenditure determinant model that tries to explain variation in local government expenditure on the basis of a large set of variables. The second step is the selection, by the agencies in charge, of the variables that are considered to reflect standardized expenditure. This step leads to sterilization, through outright elimination, of a number of variables that impact on actual expenditure, but are considered to not reflect standardized expenditure. This is the case of regions. Sterilization introduces a questionable degree of discretion in the process. The third step calculates standardized expenditure for each municipality by inserting in the regression equation the value of the independent variables.

Presently, three distinct approaches serve to determine standardized expenditure, depending on the possibility of identifying and measuring the provision of service.

They are the following:

a) “pure” expenditure functions. With these functions output cannot be measured. Hence no output indicators are used, as independent variables, to estimate the expenditure.

\[\gamma = \alpha_ {0} + \alpha^ {\prime} X + \gamma^ {\prime} W + \delta^ {\prime} Z + \eta^ {\prime} T + \varepsilon\]

where:

X are demand and supply variables;

W are cost and price (normalizing) variables;

Z represents choices about provision (direct production, local utilities, consortiums);

T are regional dummies (efficiency varies between regions).

b) “Increased” expenditure functions. Again here, output cannot measured, but the provision of service is identified.

\[+ \gamma = \alpha_ {0} + \alpha^ {\prime} X + \gamma^ {\prime} W + \delta^ {\prime} Z + \eta^ {\prime} T + \lambda^ {\prime} D + \zeta\]

where: the same as before, with D indicating when the service is provided.

c) cost functions, when output is identified

\[\gamma = \alpha 0 + \alpha^ {\prime} X + \gamma^ {\prime} W + \delta^ {\prime} Z + \eta^ {\prime} T + \theta^ {\prime} C + \psi\]

where: in addition to above,

W represent output indicators;

C are clusters of demand variables.

In all the three cases standardized expenditure is not determined according to actual provision of services. The method, rather, determines the amount of money made available (after deduction of fisca capacity) to each municipality according to the factors that determine actual expenditure across all municipalities. In other words, the grant allows each municipality to provide a level of service that is similar to that of other municipalities, provided that it uses the money received with an average level of efficiency. This method is a current occurrence across the world.

A final remark refers to the use since 2016 of cluster variables. They are dummy variables that identify clusters of municipalities that are homogenous with reference to a set of socio-economic and spatial characteristics impacting on the provision of services.

The results of the procedure are showed in Table 2 that follows. More specifically, the table lists, by columns, the expenditure functions for which standard expenditure is estimated. By rows the table lists the categories of independent variables used for estimating the expenditure and cost functions and, then, for determining the standard expenditure, using the results provided by these functions. The specific variables are detailed in the footnotes of the table and the percentage values, in the table, represent the weight assigned to each category of independent variables. Hence, for example, population variables (they are: resident population, population, population density, share of elderly population) have a weight of 43 per cent in the determination of the standard expenditure for the administration function. Physical/territorial variables are assigned a weight of 19 per cent in the determination of standard expenditure for administration, and so on (as we can easily check the weights of five categories (in bold) sum up vertically to 100 % for each function).

Table 2. Municipal Solidarity Fund. Determinants of expenditure and weights assigned to them in the estimation of standardized expenditure

Individual variables 1. Resident population, population, population density, share of elderly population. 2. Area, seismic risk, roads length, altitude, number of schools, etc. 3. Number of buildings, vehicles, number of firms, disabled pupils,number of public lights, deprivation index, etc. 4. Number of employees in restaurants and hotels, number of tourists, of visitors to museums etc. 5. Direct production, school canteen, subcontracting for kindergartens, and residences for the elderly, use of private security. 6. Population 3-14 years, disabled pupils, pupils in municipal schools, pupils in kindergartens, teachers per pupils in kindergartens, area of kindergartens, share of recycling, etc. 7. 8 clusters of municipalities used for education and kindergartens; 23 clusters for garbage disposal 8. Costs of renting, private sector wages, average salary cost of municipal personnel, rice of gasoline 9. Dummy variables for regions. Source: IFEL (2018).

There are, surely, problems with the use of regression analysis for the identification of standard expenditure needs. I am mentioning only two. The first one is a general problem, and lies in the huge (non fully justified) amount of discretion involved in the exercise. It refers, for example and especially, to the weight assigned to each category of variables (and to the each single variable within each category). Discretion stems from the fact that weights cannot be derived directly and uniquely from regression analysis, leaving room for decision-making to people/agencies responsible for the exercise. This shows, as experts perfectly know, that use of regression analysis for determination of standard expenditure and, more generally, of transfers is not a fully objective method.

The second problem is the elimination of variables that impact on expenditure, but are considered not to impact on standardized expenditure). A concrete example refers to the modes of provision. Suppose that the use of a separated agency, a local public company, leads to a reduction of the expenditure, reflecting a more efficient way of operation. The coefficient associated to this dummy would be negative in the regression. If we keep the dummy in the calculation of standard expenditure, we will reduce the transfer to municipalities that make use of a local public company, because they spend less and have lower standard needs. However, we introduce a perverse incentive. If we sterilize the dummy, meaning we exclude its impact from the estimate of standard expenditure, we eliminate the perverse incentive, but we don’t follow anymore the logics of the regression analysis, because we discard arbitrarily (some of) its results. There is the concrete risk that government would discard the results that it does not like, introducing an unacceptable amount of arbitrariness in the exercise.

The determination of fiscal capacity

The second crucial component of the MSF is fiscal capacity, or standard revenue. Its determination, for each local government, implies to identify the revenue that would derive from applying the tax effort applied on average by all other municipalities to its own standard tax base. Standard tax effort can be determined, either with reference to the statutory tax rate, or with reference to the average effective national tax rate (the ratio of collections to the effective tax base). The effective tax base differs from the actual tax base by the value of evaded tax base and by locally granted exemptions. In other words, it determines the full potential of taxation, if there were no evasion, or erosion, of the tax base. The actual tax base is the assessed tax base. Reference to the effective tax base and its difference with the actual/assessed tax base points to the difficulty of estimating tax capacity, since the information on effective tax base is, in most cases and especially in Italy, missing. Two methods are used for the Italian municipalities.

The first one is officially a (imperfect) replica of the representative tax system¹ used currently, among other countries, in Canada. The representative tax system is used to determine fiscal capacity in countries, where subnational governments, such as the Canadian provinces, make use of different tax instruments. In this setting, fiscal capacity can only be determined with reference to a common set of taxes, and, more precisely, to a standard tax base for each of them. Setting up a representative tax system does this. It consists of a set of taxes that includes all types of taxes actually used by Canadian provinces, but not necessarily by every province. For these taxes the standard tax base is calculated and the average tax rates applied are identified (measuring the average national, or standard tax effort).

The representative tax system method is applied in Italy to the property tax and the personal income tax (PIT) surcharge. It has to be noted that, while until recently municipal governments had the power to determine their own tax rates for both taxes, this power is presently frozen, making in the reality fiscal capacity equal to actual revenue. Estimates of fiscal capacity, in lieu of actual revenue, are justified by the fact that before the freeze municipalities were levying different tax rates. Hence, using actual revenue would overestimate fiscal capacity, where higher than average tax rates were previously levied.

¹ Martinez Vasquez and Boex (1997) provide a good illustration of this concept.

In the case of the property tax the tax base, which in principle is the cadastral value of properties, is determined deductively, by dividing actual collections by the tax rate chosen by each municipality. Then the standard legal (0,76%) rate is applied according to the following formula:

Standard property tax revenue = 0,76% x actual collections / chosen tax rate

As determined, the actual tax base underestimate effective fiscal capacity. A correction is introduced to take into account evasion and erosion. However, since no information referred to individual municipalities is available, an equal across the board correction is made for all, amounting to no correction at all.

For the PIT surcharge, the standard revenue is determined by applying the average surcharge tax rate, 0,4 (the range of the applicable rates varied from 0,0 to 0,8) to the centrally assessed tax base.

The garbage collection fee is extremely important in terms of revenue and, as a matter of fact, it plays the role of a property tax on the occupants of properties. Despite this fact, the fee it is not included in the determination of fiscal capacity, on the (debatable) ground that the existing constraint for municipalities to set the fee at the level that ensures the full financing of the service does not leave discretion to municipalities and makes useless the determination of fiscal capacity with reference to the fee.

Finally, for all the remaining fees and taxes (they include taxes on tourism, advertising, and the occupation of public space), the fiscal capacity is determined with the use of regression analysis.

\[R F C _ {i} = a + a X i + b M i + c P i + f S i + g T i\]

Where for each municipality :

RFC is per capita residual fiscal capacity;

X is the average rent from buildings;

M is the average price of buildings;

is the daily number of commuters;

is the number of secondary residences;

Ti is the number of tourists.

Table 3. The components of fiscal capacity of the Italian municipalities

Figure 1 that follows provides a graphic representation of the spatial distribution of fiscal capacity with a concentration of high values in the Northern areas. It is also worth remarking that higher values (intense red spots) are more frequent along the Tyrrhenian and Adriatic coast and on the Northern, alpine, borders. All these are tourist areas with a huge concentration of highly taxed secondary residences. Touristic municipalities are the most affected by the MSF, since their expenditure needs are not adequately considered (variables related to tourism are considered only for a few expenditure categories), while their higher tax capacity is correctly determined.

Figure 1, Italy. Tax capacity by clusters of municipalities.

Data collection and elaboration of information to determine needs and capacity

The entire exercise of determining standard needs and fiscal capacity requires a huge amount of information and statistical work. Data is collected with use of two extremely large surveys to be filled by municipalities. The first survey refers to service provision and is 13 pages long; the second one refers to financial and personnel data and is 20 pages long (samples of surveys on request). The direct collection of such a big amount of information from beneficiaries/losers is a very impressive effort. It raises, however, a few problems. In principle, the system has a strong built-in incentive to manipulate information. Obviously, municipal officials deny this possibility, but not the possibility that municipalities provide information in a self-rewarding way. Clearly, administration capacity (that is normally related also to size of municipalities) matters and it can affect the allocation of the funds.

In connection to this, there is the problem of control and validation of data. Presently, control is done statistically with the singling out of outliers and the subsequent request of additional information. The system also imposes a high burden on small municipalities than may not have the capacity of filling the surveys. These municipalities could be penalized in terms of smaller grants, because they have not been able to provide the right information.

Elaboration of data and determination of individual transfers is operated by a joint venture between three distinct agencies. They are: a) SOSE (Solutions for the Economic System), a company jointly owned the Ministry and Finance that is specialized in tax and tax compliance issues; b) the Directorate for Local Finance of the Ministry of Finance and, c) IFEL, a foundation wholly owned by the Association of Italian Municipalities that is in charge of research, technical support of municipalities and training.

Determination of individual allocations

It is very simply done. Allocation coefficients for expenditure, are calculated, first, by summing up for each municipality, , standard expenditure calculated for each category, , and then by dividing the sum by the corresponding national total.

\[\frac {\sum S N j i}{S N J} = a _ {s n i}\]

The same procedure is followed for fiscal capacity, getting allocation coefficient, , for revenue

\[\frac {\sum F C j i}{F C J} = b _ {s n i}\]

Net allocation is then the difference:

Assessing the redistribution impact

There are different ways of observing the redistribution of the operation of the MSF. The first, and most simple, way is to compare the number of gainers and losers. For 2017 there are 2.497 losers and 4.897 winners.

To come closer to operation of the MSF, we can distinguish between two groups of losers. The first group includes the municipalities, whose historical component, is negative, meaning that they loose at the very moment they enter the MSF, because their share of IMU/TARI is larger than the historical grant and the equalization component, is also negative. They are shown in Figure 2 with read areas. They comprise the largest group. Municipalities, for which the equalization component is larger than the historical component is positive, but is more than compensated by a negative equalization component, form the second group. One could say that that the municipalities in second group could become gainers, if the equalization mechanism were somewhat modified. They are shown in Figure 2 as the yellow areas. Gainers are shown in green.

The second way would look at the intensity of the equalization. The simplest way it to look at the per capita amounts, as done in Table 4 for a sample of municipalities comprising the largest losers and gainers. Loss can be extremely large, as in the case of two tourist cities, Rapallo and San Remo that would loose, respectively, 617 euros and 373 euros per capita, out of a per capita expenditure of 1300 and 1800 euros, again respectively. Clearly, in the case of Rapallo the allocation is untenable. It derives from the incapacity of the system to take into account the expenditure needs from tourism, while it is able to take into account the impact of tourism on tax capacity.

Table 4. The impact of the Municipal Solidarity Fund on a sample of cities. 2017

Source: Taddei and Falcone (2017).

Regional equalization

Constitutional mandates referred to equalization are very partially implemented in the case of regions. These governments are responsible for the provision of four categories of basic functions for which standard costs minus fiscal capacity equalization grants should be paid, as in the case of municipalities. They are: health, education, social assistance and local transportation. However, only health expenditure, representing almost two thirds of regional expenditure, is subject to an equalization mechanism implemented through the National Health Fund. Although in principle, the Fund should allocate funds on the basis of standard levels of service provision and of standards costs, in the practice allocations derive from a negotiation process among the regional governments and their central government counterparts (the ministries of Health and of Finance) that takes place within the State-Regions Conference.

Standard levels of service (called LEAs, or Levels of essential assistance) determine the various types of health care, such as access to specialist care, emergency units, various kind of surgical operations that patients are entitled to receive (free and/or with individual contributions) and that, correspondingly, regions have to provide within their own territory. Over the years, the ministry of Health has proceeded to formulate and update the list of LEAs and (at least officially) to estimate, accordingly, their cost and, through factoring, the total cost of health expenditure, including preventive care². The total amount of expenditure is divided into six main categories of health care, whose relative shares have been maintained stable over the years. Allocations to individual regions are made according to population and weighed population, as illustrated in Table 5.

The use of the age structure of the population is standard in the literature on health care and in actual governmental practice. This method aims at transforming the effective population of every region into a “virtual” population, taking into account the differences in demand for the health care services within different age groups. For instance, the weight assigned, for the hospital care, to children of less than 1 year of age is 3,122, while the weight assigned to people older than 75 years, is 2,906. The weights are calculated by the ministry of Health in cooperation with the ministry of finance on the basis of observed expenditure per patients and are periodically revised. They allow defining the so-called equivalent expenditure per capita, as the total health care expenditure divided by the weighted population, rather than by the effective population.

Table 5. Allocation of health expenditure by main areas of activity

² LEAs are listed in

http://www.salute.gov.it/portale/temi/p2_6.jsp?id=1300&area=programmazioneSanitariaLea&menu=lea

Source: Ministero della Salute (2017).

There is no estimate of standard costs and, as a consequence, of expenditure needs. Since 2013 this estimate has been replaced by a (partial) system of benchmarking. The system consists in selecting, according to a set of efficiency and quality criteria, a small set of regions, whose expenditure for the big areas of activity is considered to represent standard costs. In principle, these standard costs should be used for determining the allocations to regions. In the reality, since the total amounts are determined through negotiations and their allocation is done according to the population indicators, benchmarks are not applied. Instead, benchmarks are used for the allocation of a small share of the total fund that is used to foster efficiency. They are also used for monitoring the level of expenditure of regions with an unbalanced budget and in need of financia restructuring.

Finally, mention has to be made of interregional mobility. Although there is no full portability of health care, patients can select where they want to receive health care. Also, there is a dense network of agreements among regions on exchange of facilities and patients. A mobility fund reimburses regions according to the amount of health services provided to patients from the other regions. This fund amounts to almost four percent of the total. The incidence of the mechanism reaches a much larger share in some regions, where reimbursements to other regions absorb a large share of revenue due to the fact that patients give low valuation to the quality of health care received in loco. In principle, compensated interregional mobility in health care should foster efficiency, because patients exporting regions will loose resources. In a negotiated approach, the efficiency impact is diluted, since loosing regions are asking, and are able to get, partial refund of the cost of compensation.

The Fund is financed with revenue from IRAP (a regional tax on business, a direct type VAT), the regional surcharge on PIT, and a few other minor regional taxes, plus a negotiated contribution from the central government, under the appearance of a share of VAT collections. Collections from regional taxes are standardized, despite the fact that regions have presently no more tax rates determining authority. Standardization is done by applying the legal rates (plural, because they vary by sectors) to the tax base, and it takes into account the fact that, otherwise, regions that levied, before the freezing, higher than average tax rates would be penalized. Also, regions undergoing a process of restructuring their finances are presently forced to charge a higher than statutory PIT surcharge tax rate.

The National Health Fund has a strong equalizing impact, as shown in Table 5 below. Regional per capita allocations are practically the same all over the country.

Table 5. Allocation of the National Health Fund. 2017

Sources: NHF: Ministry of Health; population: Istat.

The stakeholders

The municipalities

There almost 8,000 municipalities in Italy, more precisely, as of today, 7954. Their size, in terms of population, varies from less than 50 inhabitants to more than 3 million. The size distribution of Italian municipalities still reflects the one that prevailed in the pre-unitary States (that is, before the creation of the country). While in the Northern regions a huge fragmentation prevails, the municipalities located in the Southern regions have a much larger size. A typical Sicilian rural municipality has from 30,000 to 50,000 inhabitants. This amount corresponds, more or less, to the size of the small provincial capita cities in the North. Obviously, there is a huge diversity between the two sets of municipalities in the urban functions they perform, despite the similarity in their population. Variance of income and wealth conditions is also considerable. The average per capita tax base of the municipalities of Calabria - the poorest region - is barely one third of that of the richest one, namely Lombardy. Also climatic conditions are extremely different, due to the diversity of latitude and altitude and of their combination.

The overwhelming majority of municipalities (meaning their political appointees and their administration officials) have been traditionally, at best, lukewarm towards any reform of the grant system. In fact, the old system based on historical expenditure gives them certainty of revenue and looks more attractive than the prospect of a reform, where, considering the present climate of fiscal restructuring, losses could outweigh gains. In other words, officials and politicians show a risk-averse behavior and try to do their best to avoid losses coming from reforms.

Their association

The National Association of Municipalities (ANCI) represents municipalities. It is subdivided into regional branches. The chairman of ANCI is traditionally elected among the representatives of small and medium sized municipalities, mainly because of their large number. As a consequence, the association has a clear small-medium sized municipality orientation. Big municipalities have mainly a veto power inside ANCI and have direct access to the central government. Party cleavages do not play a fundamental role within the association, both because of the bipartisan tradition of Italian politics, where government and opposition parties were often associated in financial decisions, and because of the increasingly higher variance of electoral fortunes of political parties at the municipal level.

The reform of the grant system does not traditionally rank high among ANCI priorities. Considering the huge disparities between distinct municipalities, assembling a large enough and united front for the reform would be a quite cumbersome and risky task. ANCI prefers, by large, to support individual requests for marginal changes to the present system and to negotiate, every year, with the central government a substantial increase of the previous total allocation. Most of ANCI's activity concerning municipal finances is on a day-to-day basis and is concentrated on asking compensation far every central decision that impacts negatively on the finances of municipalities. ANCI created a research and training arm, IFEL that has ben involved in the determination of grants from the MCF and has developed good analytical and statistical skills. Clearly, IFEL has also developed a vested interest in the continuation of the MCF, although its principal, ANCI, is becoming increasingly skeptical.

The central government

There are two main players at the central level, namely the Ministry of Interior (MoI) and the Ministry of Finance (MoF). MoI is the traditional tutor of Italian municipalities. In the former centralized system, its main functions consisted, in addition to finance, in monitoring and control.

Monitoring and control have been replaced by sponsoring. In other words, MoI represents the interests of municipalities at the central government level (included the Parliament). Over the years a close working relationship has been forged between MoI and ANCI. The latter prepares the requests, which are then brought by MoI to the attention of the competent Ministry, or are translated into parliament bills, when the need arises.

As in most countries, MoF is keen of replacing MoI in every aspect of the intergovernmental transfers policy. Intergovernmental grants being a considerable item in the central budget, MoF likes to play an active role, hoping that a more efficient system of allocation would bring savings in their total amount. Moreover, MoF rightly believes that controlling the grant lever could ease the present difficulties of restructuring the finances of the public sector.

Regional governments

Northern and Central regions, being much wealthier than the Southern ones, are rather cold towards intense redistribution of resources through the grants system. Different political orientations play a smaller role, while there is substantial animosity between ordinary and special statute regions, these latter ones being hugely privileged in the distribution of resources.

Despite their public utterances, all regions are not very inclined to use tax autonomy. They rather claim and favor a system of revenue sharing. Regional governments, especially those situated in Northern and Central Italy, have increasingly asked to be made responsible of allocating grants to municipalities (given, of course, the provision of the corresponding finance by the central government). However, those pleas made very little inroad in the minds of municipal officials. As is usually observed across the world, municipalities prefer to interact with a more distant payer, that is, with the central government, rather than with the regions - partly because the former is considered financially more viable and politically fair. The 2009 law implementing the constitution of 2001 has partially accepted the regional requests by making regions responsible for allocating general grants to the small municipalities. However, also this ruling has not yet been implemented.

Intergovernmental consultative bodies

As mentioned in the introduction, Italy has developed three intergovernmental coordinating bodies. They are the Conferenza Stato-Regioni; the Conferenza Stato-città ed autonomie locali and the Conferenza Unificata that puts together the previous two conferences.

All central government decisions and laws impacting on subnational government have to be presented for evaluation to these bodies. Both conferences (Conferenza Stato-Regioni and Autonomous Provinces and the Conferenza Stato-città) have also to agree, at unanimity, on the yearly allocation of grants to each level of government. Due to the incremental nature of the system of allocation, both conferences usually reach easily an agreement. The Conferenza Stato-Regioni and Autonomous Provinces had also reached, after long lasting debates about the intensity of equalization and the timing of the reform, a substantial agreement on the allocation of the National Health Fund and formed a common front in negotiations with the central government. This makes the prospects of reform unlikely.

Presently agreements reached within the State-Regions Conference are almost automatically validated by the Parliament. There is, however, growing unease and resentment against this rubberstamping procedure.

Conclusions

Since the constitutional review of 2001 Italy has embarked on a rather ambitious process of reforming its system for financing subnational governments. The constitution assigns to the central government the responsibility of determining, and guaranteeing across the whole nation, essentia (minimum) levels of provision of a set of basic services assigned to regions and local governments. The implementing legislation has mandated the introduction of a municipal equalization fund aimed at filling, with reference to fundamental functions, the difference between standard needs and standardized fiscal capacity. For the remaining (non-fundamental) functions the law mandates that equalization will fill only differences in fiscal capacity. The determination of standards needs implies consideration of two distinct elements: a) standards of service provision (basic level of benefits) and, b) the standard cost associated to the basic levels of benefits. This is really a Herculean task, particularly in the case of municipalities, given their huge number and variety. It does not come as a surprise that the result, the present Municipal Solidarity Fund, has achieved only partially the aims assigned to it. Possibly it will never overcome them.

In any case, a substantial effort as been devoted to the determination of standardized expenditure and fiscal capacity/standardized revenue. Results are more or less in line with what achieved in other countries engaged in the allocation of similar equalization grants schemes. At the same time, real standard needs and costs have not been defined, and I really wonder how this can be achieved in the practice.

The use of statistical methods, such as regression analysis, making use of very detailed statistical information for the determination of transfers filling the gap between standardized expenditure and revenue is an increasing popular practice, but it can be questioned on many grounds. One can question the difficulty of ensuring transparence deriving from its sheer complexity, and the possibility of manipulation both by the agency that makes the calculations and by the local governments that provide the information. However, there are merits also. The main one it that it forces to make detailed analyses of the provision of services, of the use of revenue sources and of the various factors that impact on these activities. Also in Italy, despite the shortcomings of the procedures illustrated in this paper, one has also to recognize that there has been an improvement over time in the estimates and that an impressive work of collection of relevant statistical information has been started.

A viable alternative is a method, such as that used for the National Health Fund, relying on a very restricted number of variables, all focused on the impact of population and its age structure on expenditure and leaving aside fiscal capacity. Simplicity can lead to greater transparency, but it also implies a lesser capacity to take into account factors that may have a substantial impact on expenditure needs. This is a likely occurrence, when the number of beneficiary governments and variation of their essential characteristics is huge, as in the case of Italian (and also Spanish) municipalities.

References

Brosio, Giorgio (2016). Intergovernmental Equalization Transfers with National Standards, in Geißler, Knüpling, Kropp, Wieland (eds.) Das Teilen beherrschen. Analysen zur Reform des Finanzausgleichs, Nomos, Berlin.

IFEL (2018). Perequazione comunale Italiana: fabbisogni e capacità fiscali standard. Roma,

Ministero della Salute (2017). Riparto disponibilita’ finanziarie per il Servizio Sanitario nazionale nell'anno 2017. Richiesta di Intesa alla Conferenza Stato-Regioni. Roma.

Marchionni,Emilia, Corrado Pollastri e Alberto Zanardi (2017). Fabbisogni standard e capacity a fiscali nel sistema perequativo dei Comuni. Ufficio Parlamentare di Bilancio, Roma.

Martinez Vasquez, Jorge and Jameson Boex (1997). Fiscal capacity an overview of concocts ad measurements issues and their applicability in the Russian Federation, Working paper. Andrew school of public policy, Atlanta.

Ministero della Salute (2017). Riparto disponibilita’ finanziarie per il Servizio sanitario nazionale nell'anno 2017. Richiesta di Intesa alla Conferenza Stato-Regioni. Roma.

Taddei, Andrea and Giulia Falcone (2017). Il Fondo di Solidarietà: effetti e sostenibilità per i comuni italiani. Anci Firenze.