VI Fiscal Decentralization in Spain
Author:
Mr. Gonzalo C Pastor Campos
Search for other papers by Mr. Gonzalo C Pastor Campos in
Current site
Google Scholar Close
Language:
English
Print Publication Date:
15 Jan 1993
Keywords:
OP; market; government; transition process; exchange rate.; income; income tax; government finance statistics yearbook; Comunidades Autónomas de regimen Común; revenue collection; government finance statistics data; authority finance; Subnational tax; Public employment
Download PDF (4.6 MB)

Abstract

Fiscal decentralization is rather a new phenomenon in Spain. In contrast to recent developments in other countries of the Organization of Economic Cooperation and Development (OECD) where governments have embarked on a process of growing fiscal centralization, in Spain, the weight of regional governments in general government total expenditure has increased from less than 2 percent in 1981 to 7.5 percent in 1986 and to around 12 percent in 1988.1 At the same time, regional governments have become an important source of public employment. For example, in 1986, they employed 340,000 people, twice as many as in 1984. The rapid growth in the size of regional government employment represented an increase in its share in total public sector employment from around 10 percent in 1983 to 23 percent in 1988. Employment by regional governments is now more than half the size of the state civil service.

Fiscal decentralization is rather a new phenomenon in Spain. In contrast to recent developments in other countries of the Organization of Economic Cooperation and Development (OECD) where governments have embarked on a process of growing fiscal centralization, in Spain, the weight of regional governments in general government total expenditure has increased from less than 2 percent in 1981 to 7.5 percent in 1986 and to around 12 percent in 1988.1 At the same time, regional governments have become an important source of public employment. For example, in 1986, they employed 340,000 people, twice as many as in 1984. The rapid growth in the size of regional government employment represented an increase in its share in total public sector employment from around 10 percent in 1983 to 23 percent in 1988. Employment by regional governments is now more than half the size of the state civil service.

The process of decentralization has proceeded very rapidly with regional governments obtaining responsibility for a vast array of public services in less than ten years, a transformation that in other countries took decades. Moreover, this trend toward decentralization promises to continue, as the Socialist Party, in its campaign for the 1991 regional elections, proposed to transform Spain into a full-fledged federal state.

Since 1978, a complete array of fiscal instruments has been developed to finance regional governments’ new responsibilities. A key criterion in developing these instruments was to ensure that the regions had sufficient resources to cover the cost of their new activities. At the same time, the system was designed to provide incentives for improving tax collections by the regions. Another important consideration was to establish a system that would work to reduce regional disparities. To this end, a system of capital transfers was developed to redistribute public investment across the Spanish territory. The method, however, by which the central government allocates grants to regional authorities has reduced the regions’ financial autonomy and has led to growing regional governments’ fiscal imbalances. As a result, regional governments have increased their recourse to debt financing, which, if not addressed, could have important implications for the economy in the near future.

The Size and Scope of Regional Governments

The 1978 Constitution divided the Spanish state in 17 regional governments (Comunidades Autónomas) invested with administrative and legislative powers. Rather than specifying a definite state of affairs, the Constitution sketched an open process of fiscal decentralization. The decisions on the speed at which fiscal responsibilities were to be transferred from the central administration to the regions remained in the hands of the regional authorities. In principle, all regional authorities were made responsible for the structure of their institutions, their financial relationships with local governments in their territory, and investment decisions on infrastructure development, such as public works, housing, transportation (roads, railways, harbors). Other responsibilities included the provision of education, health, and welfare services, as well as environmental policy implementation, cultural development, tourism, and police.2 Consequently, the scope of the decentralization process has varied substantially across regional governments depending mainly on two elements: (1) the intensity of each region’s political movement for independence and autonomy and (2) the degree of sophistication of each region’s institutions and tax administrative capacity.

As a result of this changing nature of the Spanish state, regional governments have become an important sector in the Spanish economy in the last ten years. From a base of 0.6 percent of GDP in 1981 regional governments’ total expenditure increased to 3.2 percent of GDP in 1984, and to 5.2 percent in 1988.3 The increase in regional authorities’ operations did not translate into higher total public sector expenditure, but rather represented a substitution of regional authority’s expenditure for central government’s expenditure.4 This was particularly true for the period 1984–90, when total government expenditure leveled off at an average of around 41 percent of GDP, notwithstanding the rapid increase in regional government expenditure. During this period, the weight of central government operations in total public sector’s total expenditure declined from around 82 percent in 1984 to 76.3 percent in 1988 (Table 16).

Table 16.

Participation of Different levels of Government in Total Public Sector Expendiure

(In Percent of total public sector expenditure) 1

article image
Source: International Monetary Fund, Government Finance Statistics Yearbook, various issues.

Refers to unconsolidated public sector expenditure.

Includes the operations of the state, the administrative autonomous agencies, and the social security system.

Excludes expenditures associated with the transfer of health (Insalud) services from the social security system to the regional governments.

The decline in the weight of central government operations in total public sector expenditure in Spain contrasts with recent trends in the main OECD countries (Table 17). Between 1977 and 1987, both federal and unitary states in the OECD show a tendency toward a more centralized public sector. In the case of federal countries, the average weight of central government expenditure in total public expenditure rose from around 57 percent in 1972 to around 59 percent in 1985–87. Similarly, in the case of unitary countries, this centralization ratio went from 67 percent in 1972 to some 70 percent in 1987. Interestingly, in the United States a sharp increase in the degree of centralization occurred during 1972–87 despite calls for a policy of “New Federalism.” Although the U.S. Congress delegated to the states new responsibilities, there were no provisions to allow for a parallel increase in state revenues.

Table 17.

Particiption of Central central Government Expenditure in Total Public Sector Expenditure in Selected Countries

(In percent of total public expenditure) 1

article image
Sources: International Monetary Fund, Government Finance Statistics Yearbook; and IMF staff estimates.

Except for Australia, figures are with respect to unconsolidated public sector expenditure.

Refers to 1988.

In 1980, Government Finance Statistics data for Australia contain a break in the time series starting 1980

Notwithstanding the progress to date, much remains to be done to reduce the weight of central government operations in total public sector expenditure in Spain to levels comparable to other OECD countries. The weight of central government operations in total public sector expenditure in Spain remains 18 percent higher than the average for federal states and 8 percent higher than the average for unitary states (Table 17).

Instruments of Regional Authority Finance

Regional authorities in Spain are classified in two broad categories: regional governments under the regular financing scheme (Comunidades Autónomas de Regimen Común) and regional governments under the forum system (Comunidades Autónomas de Regimen Foral). The Basque Provinces and Navarra are the only two regional governments under the latter system, which has evolved from these two regions’ unique historical relationship with the central authorities. Within their territory, these two regional authorities are fully responsible for the legislation and administration of a large number of state taxes (Impuestos Concertados), which are negotiated annually between the state and the regional authorities. These regional authorities are then obligated to transfer back to the central government a specific amount of resources (Cupo) in lieu of those public services that are still being provided by the state. In 1986, revenue from Impuestos Concertados, net of Cupo, represented 93 percent and 78 percent of total revenues of Cataluña and Navarra, respectively.

The other 15 regional governments are all included under the regular financing scheme (Comunidades Autónomas de Regimen Común) and represent the core of the process of fiscal decentralization in Spain. In terms of their responsibilities, they can be classified in two groups: regional authorities with limited responsibilities and regional authorities with a broad mandate, including the provision of education and health services. The first group is usually referred to in Spain as regional authorities under Article 143 of the Constitution. It includes ten regional governments: Aragón, Asturias, Baleares, Cantabria, Castilla-La Mancha, Castilla-León, Extremadura, Madrid, Murcia, and La Rioja. The second group is known as the regional authorities under Article 151 of the Constitution. This group includes five regional governments: Andalucía, Canarias, Cataluña, Galicia, and Valencia. Except for Andalucía, regions under Article 151 of the Constitution are the richest regions in Spain.

In financing expenditures, regional governments under Article 143 and 151 of the Constitution derive resources from a number of instruments, which have been developed to finance regional governments’ responsibilities and reduce regional differences of income and wealth in the last ten years. The instruments of regional authority finance are (1) regional governments’ own taxes and fees; (2) tax sharing of specific state taxes collected in their territory; (3) general grants from the state; (4) conditional grants; and (5) the Inter-territorial Compensatory Fund (Table 18). General grants are the most important source of finance for these regional authorities. However, the weight of conditional grants in total revenue rises sharply for these regional governments (included under Article 151) administering the provision of education and health (INSALUD) services.

Table 18.

Total Revenue of Regional Governments1

article image
Sources: Spain, Ministry of Economy and Finance; and IMF staff estimates.

Refers to comunidades autonómas de regimen común: Andalucía, Canarias, Cataluña, Galicia, Valencia, Aragón, Asturias, Baleares, Cantabria, Castilla-La Mancha, Castilla-León, Extremadura, Madrid, Murcia, and Rioja.

Excludes conditional grants for the financing of education and health (INSALUD) services. As of 1987, only two regional governments (Cataluña and Andalucía) had received administrative powers in this area. The data for these cases were in billions of pesetas: Cataluña (Ptas 203.4 billion), Andalucía (Ptas 206.2 billion).

Interterritorial Compensatory Fund.

Regional Authorities’ Own Taxes and Fees

Regional authorities fully administer and regulate their own taxes and fees. There are two kinds: (1) taxes and fees that were originally administered by the state, but have been turned over to regional governments; and (2) new taxes and fees created by the regional authorities. The first group covers fees and duties collected for public services; the second includes, for example, the tax surcharge on gambling created by the regional authority of Cataluña. Although revenues from these two sources have steadily increased in recent years, they represent the least important source of income in regional authorities’ total revenues (Table 18).

Tax Sharing

Tax sharing pertains to a small number of state taxes earmarked for distribution to the regional authorities. The central government decides the taxes to be shared, the rate to be imposed, and the formula for allocating receipts. Regional authorities are responsible for collecting these taxes in their territory. The tax revenues included under this heading are inheritance tax, wealth tax, tax on luxury items, tax on gambling, and the tax on legal services (transmisiones patrimoniales intervivos). Through 1985, the most important of these taxes was the tax on luxury items, which represented around 33 percent of total receipts; however, in 1986 receipts from the tax on luxury items dwindled as the introduction of the value-added tax abolished this source of revenue. Since then, the tax on gambling and the tax on legal services account for the bulk of the revenue from this source.

General Grants

In 1989, general grants represent around 60 percent of regional authorities’ total revenue (Table 18). They mainly finance the operational cost of public services that were originally in the hands of the central administration, but which are now the responsibility of the regional authority. The system of general grants was reformed in November 1986, as the estimation process of the amount of general grants to be transferred to regions lacked the degree of automaticity envisaged in the 1978 Constitution. Also, the original system favored the financing of current expenditures and did not provide enough resources to cover new investments in public services. The new methodology, agreed for the period 1987–91, gave more automaticity to the system, broadened the coverage of the system to incorporate financing of new investment in public services, established precise rules for the evolution of general grants for the period 1987–91, and reformed the criteria used for grant distribution between regions.

Under the original system, the amount of central government general grants to each regional authority was defined as a share of state tax revenue collections. This share, in turn, was expressed as the ratio of the regional authority’s operational balance to total state revenue collections, lagged by one year. The operational balance was defined as the difference between receipts from those state taxes earmarked for distribution to the regional authorities and the regional authority’s “effective cost” of providing services that were originally in the hands of the state. The concept of effective cost included outlays on wages and salaries, goods and services, and equipment maintenance associated with the regions’ new responsibilities.

The original system of general grants was inefficient or flawed in several respects. First, it lacked automatism in estimating the amount of grants to be received by each regional authority. Second, it introduced incentives for inefficient collection of those state taxes that had been relinquished to regions for their collection. This was a major disincentive for a better administration of state taxes in the hands of the regional authorities. Third, it did not provide regional governments with resources to finance investment projects associated with their new responsibilities.5 This situation generated two additional distortions. On one hand, financing for these investment projects had to come from sources that were originally targeted for different purposes. For example, many regional governments had to draw resources from the Interterritorial Compensatory Fund, which had been designed to correct regional differences in Spain and not to finance expenditures stemming from regional authorities’ new responsibilities. On the other hand, by focusing only on the financing of operational expenditures, the system perpetuated an unequal provision of public services across the regions.

Lengthy negotiations between the central administration and the regional governments produced a reform of the general grant system in November 1986 (see the appendix for the general analytical framework). At that time, a new methodology for general grant financing was agreed upon for the period 1987–91. To provide regions with sufficient resources to finance current and capital expenditures associated with their new responsibilities, the new system set a three-step estimation process. First, it developed the concept of “financial need.” This concept was defined as total amount of resources used by all regional governments to finance the provision of public services in 1986.6 These resources included general grants, conditional grants for the financing of subsidies for free education, funds received under the Interterritorial Compensatory Fund (FCI), and all additional state financing to regional authorities. Also, the new system developed a measure of “potential” tax revenue collections from those state taxes that had been turned over to regional governments.7 This was the revenue that regional authorities would have raised by a standard “effort” in collecting state taxes relinquished to them.

Second, the new system of general grants modified the criteria used for distribution of grants in 1986 across regional authorities. In place of the concept of “effective cost,” general grants were now to be distributed on the basis of each region’s population, territorial size, administrative structure, fiscal effort, and relative wealth.8 Under the new system, the rather large weight given to the variable population in the new distribution formula has been instrumental in reducing differences in per capita grants across regional governments with similar level of responsibilities.

Third, to avoid the extensive negotiations of the past, the new system established that, for 1987–91, the level of grants should increase in line with the growth rate of “structurally adjusted state taxes” (Ingresos Tributarios Ajustados Estructuralmente). This concept was defined as the sum of those state taxes that had not been relinquished to regions (or to the European Community) and the total amount of social security and unemployment contributions.9 However, as social security and unemployment contributions are particularly affected by the different phases of the business cycle, reliance on them would create an unstable source of revenue to the regions. Consequently, to reduce the variability of regional governments’ total income, the new methodology provided for upper and lower limits for the growth of general grants. The upper limit was defined as the rate of growth of nominal GDP. The lower limit was defined as the rate of growth of total expenditure (gastos equivalentes) by the state and other administrative autonomous agencies that provide large transfers to regions.10

Conditional Grants

Conditional grants provide central government aid to regional authorities for specific services of national interest. The activities covered by these grants are specified by the central administration. They include mainly regional authorities’ administrative outlays (autogobierno y gestión) and subsidies for free education. In the case of regional governments under Article 151 of the Constitution, conditional grants also cover out-lays from the provision of health and social services transferred from the social security system. However, as of 1987, only Andalucía and Cataluña had received responsibilities for these areas. For these two regions, the share of conditional grants in their total revenues was as high as 53 percent in 1987. For the other regions, conditional grants represented around 15 percent of their total revenues.

In general, conditional grants are provided on the ground that the benefits of many public services “spill over” from the community in which they are performed to other communities. For example, a person may be educated under one regional government and migrate to another when entering the labor force. In such circumstances, investment in education will be too low if it is financed entirely by regional authority resources, because each region would be willing to pay only for those benefits likely to accrue to its citizens. State assistance would be needed to increase the level of such expenditures to the optimum from the national standpoint.

To give regional governments more independence in their actions, the reform of the system of general grants in November 1986 transformed most conditional grants into general grants. Grants to finance the provision of health and social services, however, remained as conditional grants.

The Interterritorial Compensatory Fund

The Interterritorial Compensatory Fund (Fondo de Compensation Interterritorial) (FCI) consists of a system of capital transfers from the state to the regional governments for the financing of investment projects in the areas of public works, water supply, rural housing, transport, and communications. Its importance in regional governments’ total revenue is considerable (8 percent of total revenues in 1989), although it has been declining steadily since 1986 when the new system of general grants transformed some of the FCI funds into general grants (Table 18).

The FCI was established by the 1978 Constitution to redistribute public investment across the Spanish territory and thus reduce regional differences in income and wealth. In practice, however, this has not always been the case. Through 1989, the Fund was used for regional redistribution of investment, but also as a way to finance investment projects associated with services transferred to the regional authority and excluded from the “effective” cost concept. This situation was problematic, however, because the redistribution criteria and the criteria of sufficiency for the financing of investment projects were not always consistent. Based on the criterion for sufficiency, the distribution of FCI funds benefited those regions (usually the richest regions) with most services transferred. Whereas, based on the redistribution criteria, the distribution of funds was inversely related to the degree of economic development of the regions. To the extent that the criteria for sufficiency had, at times, prominence over the redistribution criteria, the capacity of the FCI to act as an instrument of regional equalization was thus significantly diminished. To solve this problem, it was reformed in 1990 and made into an exclusive instrument of regional policy.11

Through 1989, the size of the FCI totaled 30 percent of central government’s investment in civil activities and its resources were distributed across all regional governments in Spain. The criteria used to distribute FCI funds between regional authorities were composed of the inverse of the region’s per capita income, migration patterns during the previous ten years, unemployment rate, territorial size, and degree of isolation. In determining the precise distribution formula, the 1981 FCI Law gave particular prominence to income per capita and migration patterns.

It is possible to identify two broad periods in the functioning of the original FCI. The first phase, through 1986, was characterized by an active role of the FCI in correcting interterritorial imbalances. During these years, around 70 percent of FCI funds were transferred to the seven poorest Spanish regions (Extremadura, Andalucía, Castilla-La Mancha, Galicia, Castilla-León, Murcia, and Canarias), which contained 45 percent of the Spanish population and generated 35 percent of Spain’s national income. During 1987–89, however, this trend was reversed and the capacity of the FCI to correct regional imbalances was reduced significantly. The seven poorest regions received only 65 percent of FCI resources, with Extremadura and Andalucfa, which are both among the poorest regions in Spain, losing the largest magnitude of funds. At the same time, the share of some of the richest Spanish regions in the FCI rose significantly. This was particularly true for the case of Cataluña and the Basque Provinces. An illustrative example of the situation reached in 1989 was that the Basque Provinces, which is the region with the third largest income per capita in Spain, received a larger FCI per capita than that of Andalucfa, Castilla-León, Murcia, or Asturias, which are among the poorest regions in Spain.12

The deviation of FCI resources from the poorest to the richest regions resulted from the rather high weight given to migration patterns in the formula governing the allocation of funds between regions, in a context where the patterns of labor migration had been changing sharply. Before the 1979 oil shock, there were large migratory movements from the poorest regions to the rich industrial centers of Madrid and the northeastern regions of Cantabria and the Basque Provinces. After 1979, however, with the process of industrial reconversion in Spain, there was a rather limited migratory flow to industrial centers and in some cases, the patterns of migration reverted. Indeed, during 1982–86, regions like Extremadura, Andalucía, Castilla-La Mancha, and Galicia, after having been labor-exporting regions for years, had large population inflows. By contrast, the decline of the steel and naval industries in Cantabria and the Basque Provinces resulted in population outflows.

In 1990, the scope of the FCI was substantially revised and made into a more effective instrument of regional policy. First, the number of regions receiving support from it was reduced from 17 in 1989 to only 9. The regions selected were those whose average income per capita was lower than 75 percent of the EC average. These regions are also known as regions under Objective 1 of EC Structural Funds policy. They include Extremadura, Andalucía, Castilla-La Mancha, Galicia, Murcia, Canarias, Castilla-León, Valencia, and Asturias. Second, the total level of the new FCI was redefined as 30 percent of central government investment, adjusted by the relative size of each region’s population and its gross value added per capita. Third, the criteria for distribution were set in two steps: first, the fund was distributed across the nine regions selected so that each received the same FCI per capita grants; then, among the selected regions, adjustments were made to redistribute funds from the richest regions to the poorest ones.13 The criteria for redistribution of funds between the rich and the poor regions were based on the differences between the average gross value added for all the nine regions and the gross value added of the specific region. Finally, to provide investment financing to those regions that were excluded under the new FCI, special funds were allocated under a transitory budget (compensation transitoria) to provide some degree of continuity with the previous system of financing.

Analysis of the Process of Fiscal Decentralization

A key theme emerging from the study of fiscal de-centralization in Spain is the importance of the process of decentralization as much as its product. The decentralization of public sector activity has met with a number of problems. The negotiations between the central administration and the regional governments regarding the transfer of responsibilities and its financing proved to be very difficult. A huge amount of resources has been used during the bargaining process. Conflicting interests between the state and local authorities, between rich and poor regions, and between different political parties had to be accommodated within the bargaining process. The resulting compromises from this process led to several changes in the instruments of regional expenditure finance and the institutions within which policy making has taken place.

Autonomy

A conventional argument in federal finance literature is that, in principle, both central and regional (local) governments must each have under its own independent control financial resources sufficient to perform its exclusive functions.14 In this ideal situation, the “own sources” revenues of each level of government would sufficiently finance the expenditure for which it is responsible without recourse to intergovernmental grants (transfers). Only then can the desired position of local autonomy of action and local responsibility of those actions be achieved. Fiscal balance in this sense requires the assignment to each level of government separate and independent revenue sources sufficient to finance its respective expenditures.

The ideal fiscal balance proposed by the literature is rarely achieved, however. A “vertical fiscal imbalance” in the revenues and the expenditures of different levels of government has persisted in federal and non-federal states for years, and the resulting “fiscal gap” of regional and local levels of government has been closed by intergovernmental grants. Nonetheless, during the 1980s, the dependence of subnational governments on state grants has somewhat declined and the role of taxes in regional governments’ expenditure finance has become more important in each of the six advanced industrial federal countries and in all main nonfederal countries. Certainly, there are numerous economic and political reasons for the increased role of taxes in these countries. No single fiscal indicator is likely to tell us much about what is really going on in the public sector of any country.15 But, regardless of the specific reasons for each of these developments, there can be no doubt of the advances toward a situation where each jurisdiction has under its control sufficient resources to finance its expenditure responsibilities.

In Spain, the degree of vertical fiscal imbalance differs sharply between regional governments under the forum system and the regional governments under the regular financing scheme. For the first group of regional governments (the Basque Provinces and Navarra), tax revenue (from Impuestos Concertados) represents around 95 percent of regional governments’ total revenue, giving these regional governments a large degree of financial autonomy. By contrast, for regional governments under the regular financing scheme, tax revenue represents on average only 14-percent of their total revenue, and therefore, their degree of dependency on the central administration is much higher than in the case of the Basque Provinces and Navarra. The level of their tax revenue is way below corresponding figures in other OECD countries.

The limited role of tax revenue collections in regional governments covered under the regular financing scheme responds to three broad forces. First, the process of fiscal decentralization in Spain is rather at an early stage of development and many regional governments, especially those included under Article 143 of the Constitution, still need to develop their tax administration machinery. Second, the policy of tax sharing of state taxes collected in each region’s territory has centered around tax figures with rather low income (GDP) elasticities. Moreover, the elimination of the tax on luxury items, following the introduction of the value-added tax in January 1986, emptied an important source of tax revenue for regional authorities. Finally, fear of losing commerce and industry to other regions and of discouraging the entry of new business has restrained regional authorities from increasing taxes and fees.

Accountability

The rather high degree of vertical fiscal imbalance of the Spanish case has been useful to both state and regional governments. The state has managed to keep its taxing power over the main tax figures (i.e., value-added tax, income tax) and to establish a significant degree of financial control over regional governments through the system of general and conditional grants. At the same time, the regions have achieved financial sufficiency in the provision of public services without having to face the political costs of imposing higher local taxes.

Nonetheless, the system has many disadvantages. First, and perhaps most important, the regions are less than fully accountable to their citizens for expenditure decisions, as the regional authorities need not resort to increasing taxes to finance growing local services. Under these circumstances, regional governments have been tempted to provide services beyond the point where the costs of an extra unit of service equal the benefits from that unit. Second, under the current system, the central administration has gained control over public policy areas that are generally under the control of subnational governments in other European countries. Control is exerted over the level and direction of regional expenditure, as well as over financial practices (i.e., matching funds, cofinancing strategies). In addition, regional governments have not been motivated to take an active role in the government’s campaign against tax fraud, in particular, in cases where tax enforcement could affect important and influential citizens in their communities. This was particularly true for the case of the “Primas Unicas.” Despite regional governments’ responsibilities in the collection of taxes on wealth (under the tax-sharing system), regions like Cataluña passively promote the development of this financial instrument of fiscal opacity.

Several alternatives exist for reducing the dependency of regional governments from state grants. One would be to increase the number of taxes that have been relinquished to regional authorities for collection in their territory. For example, collections from the value-added tax at the retail level could be delegated to regions. Another option, much in line with practices in federal countries, would be to substitute grant financing for regional governments’ surcharges to the federal income tax schedule. These two alternatives would increase the weight of regions’ own resources in total revenue and foster improvement in the regions’ tax administration. The second alternative would also establish a greater link between expenditures and taxes in each region, obligating regional authorities to be more accountable for their actions.

Equity Between Regions

Another issue relevant for a discussion of fiscal de-centralization has to do with the role of equalization grants. This type of grant enables poor regional governments to provide a level of public services similar to those supplied by rich regions without having to set their tax rates at much higher levels than in wealthier areas. In Spain, there is not a system of grant equalization per se, but it could be argued that over the last ten years much has been done in terms of equity between regional governments. Indeed, much of central government help to regions under the system of general grants and the FCI has been allocated among regions using distributive (per capita) criteria, with the main objective to reduce inequality between regions.

In Spain, central government grants have helped equalize the fiscal resources of the regional governments, though problems remain. When the regions are ordered by the size of per capita gross value added, the per capita revenue obtained from tax sources (regions’ own taxes and rates and state taxes relinquished to regions for their collection) increases from low to high GDP regions. In 1987, the three regions with the lowest GDP per capita raised only Ptas 5,053 in taxes per head, whereas the three regions with the highest income raised Ptas 8,474 per head. This situation is reversed, however, after revenues from grants are taken into account. Indeed, after grants, the per capita total revenue of the poorest regions shoots up to Ptas 36,747 against a per capita total revenue of Ptas 26,182 for the three wealthiest regions. This reversal of the original horizontal inequality is particularly worrisome since concern has been expressed from time to time on the allegedly higher “expenditure” needs of the more densely populated (and richest) urban areas.16 The “overequalizing” effect is more marked for regions included under Article 143 of the Constitution than for regions under Article 151.

Macroeconomic Effects

One of the most significant effects of the process of fiscal decentralization has been the rapid deterioration of the overall balance of the regional governments. From a situation of virtual equilibrium in 1987, the overall balance of the regional governments shifted into a deficit equivalent to 1.1 percent of GDP in 1990. Although this deterioration occurred at a time when the overall fiscal deficit of the central government was cut from 3.1 percent of GDP in 1987 to 2.4 percent in 1990, the growing magnitude of regional governments’ fiscal deficits is of concern and complicates fiscal management.

The deterioration of regional governments’ finances has resulted in a rapid growth of regional governments’ indebtedness, which, although small when compared with the level of indebtedness of the rest of the general government, could become a source of concern over the medium term. There are only binding rules for the growth of regional governments’ long-term borrowing, leaving the regions considerable freedom to borrow short term. According to the Constitution and Law of Regional Governments’ Finance (LOFCA), the annual amount of interest payments plus amortization paid by regional governments on their long-term (bonded) debt cannot exceed 25 percent of the regions’ current revenues. Although regional governments require prior authorization from the state to issue bonds and borrow in international money markets, the law does not preclude borrowing from domestic commercial banks and savings banks, which indeed has increased sharply in recent years. The regional debt is concentrated in few regional governments, notwithstanding the presence of all regions in the market. In 1986, three communities (Cataluña, the Basque Provinces, and Andalucía) absorbed about half of the total debt. In general, there seems to be a positive correlation between the size of the debt of the region and income per capita of the region and the inverse of the number of provincial autonomies. Also, regional authorities under the control of political parties other than the Socialist Party seem to be more inclined to borrow in the financial markets.

Before the 1986 reform of the system of general grants, it was argued that borrowing by regions was only a transitory phenomenon.17 It was argued that the reasons for borrowing were mainly (1) the limited financing available from the state owing to the rather narrow definition of the concept of effective cost under the original general grant system, which covered only current expenditures stemming from those services transferred from the state to the regional authorities; and (2) the rather excessive role of conditional grants in regional authority finance, whose earmarked funds reduced regions’ ability to attend alternative outlays and forced them to borrow in the financial market. Also, the rapid growth of the state public debt during the first half of the 1980s had perhaps a demonstrative effect that did little to encourage more conservative behavior on the part of the regional authorities.

Recent debt statistics, however, suggest that these studies were overoptimistic in expecting a slowing down of the growth of regional debt after the reform of the general grant system. Regional debt has continued to grow rapidly in recent years and appears to have become a more permanent phenomenon. In many regions, it may reflect a need for the region to assert its autonomy from the central administration.

To complement more traditional sources of borrowing, several regional governments have been taking advantage of their special fiscal status to issue financial instruments with “fiscal opacity.” Specifically, the jurisdictions of Vizcaya, Guipúzcoa, and Alava in the Basque Provinces have been issuing what is called in Spain Pagarés Forales. These Pagarés are short-term financial instruments with a special fiscal status: holders are exempted from the payment of withholding tax and the regional authorities do not have to inform the central administration of the identity of the buyer. Hence, they are generally considered as a refuge to “black” money. Although precise figures on the amount of Pagarés Forales are still being prepared by the Bank of Spain, it is generally believed that the total is around 1.5 percent of GDP in 1990. In 1991, the central administration devised a process of fiscal regularization that met with a relative success at the beginning of 1992.

Conclusion

Since the establishment of the regional governments by the 1978 Constitution, the transfer of responsibilities from the central administration to the regional governments has been rapid. Over these years, the process of decentralization has evolved to correct some of the initial drawbacks. First, there has been a move away from a system of conditional grants toward general grants that give regional governments increased leeway in the use of state funding. Second, the scope of the general grant system has been broadened to provide regional governments with sufficient resources to cover current and capital outlays connected with their new responsibilities. This has corrected initial practices, where grant financing covered only current expenditures and where funding for investment projects to improve the quality of public services had to come from sources that were originally targeted for different purposes (i.e., correction of interterritorial differences). The wider scope of the general grant system has been particularly helpful for those regional governments trying to revert past trends characterized by an underprovision of public services. Third, since 1986, the general grant system has effectively introduced the concept of “tax potential” in allocating funds across regions. This concept has been an effective way to provide incentives for a better tax administration of those state taxes that had been relinquished to regional authorities. In addition, the general grant system is now based on precise rules governing the annual growth of state help to regions, thus preventing the need for extensive negotiations, which characterized the early years of the system.

The Spanish system of regional government finance also puts particular emphasis on reducing regional disparities of income and wealth. To address this concern, the FCI was created by the 1978 Constitution. Initially, the objectives of the FCI were twofold: to redistribute public investment and to finance a portion of the investment projects associated with the regions’ new responsibilities. This led to operational problems and made the FCI less than fully effective in addressing problems of regional disparities. As a result, in 1990 the FCI was modified and made into an exclusive instrument of regional policy by narrowing its focus. Under the new system, the regions selected are in line with the benchmarks developed by the EC under its Structural Funds Policy.

Problems still remain, however. The central administration has retained taxing power, despite having delegated numerous public services to regional authorities. This has led regions to become increasingly dependent on state grants, a dependency that is rather high by international standards. This vertical fiscal imbalance has reduced autonomy of regional governments and has occurred despite growing recognition of the benefits of a more decentralized tax authority. In such a system, regional authorities would be more accountable to their citizens for the growing costs of public services. Nonetheless, the present division of responsibilities has benefited both the central administration and the regional authorities. The central government, through the general grant system, has been able to keep some control over regional authorities. At the same time, the regions have not had to face the political cost of higher taxes for better public services.

With regard to the objective of regional equalization, much has been done to supply poor regions with sufficient resources to support “standard levels” of public services, without them having to resort to excessive increases in tax rates and fees. However, the current system of grants could be criticized in that it has “over-equalized” regions, reverting initial conditions of in-equality of per capita revenues to a situation where per capita revenues of the richest regions are more than one third lower than that of the poorest regions. This is particularly problematic as, in order to support an adequate level of public services, the largest (richer) urban areas require per capita revenues similar to the poorer regions. This, in turn, has triggered a rapid growth of short-term borrowing by the richer regions. In addition, many of the poorest regions lack the administrative capacity to adequately administer the large amount of funds coming from the state.

Finally, perhaps one of the most significant implications of the process of fiscal decentralization has been the rapid deterioration of regional governments’ financial position. This has occurred at a time when the central government has embarked on a process of increased budgetary discipline. From virtual equilibrium in 1987, the overall balance of the regional governments shifted into a deficit equivalent to 1.1 percent of GDP in 1990, which has been financed with short-term commercial debt. This deterioration occurred despite the generous increase in central government grants to the regional authority. Initially, it was thought that the increase in regional governments’ debt would be a transitory phenomenon. Several studies on the topic argued that the growing level of regional indebtedness largely reflected the incapacity of the grant system to provide enough resources for the financing of new investment projects associated with regions’ new responsibilities. Experience, however, has shown that these studies were overoptimistic and that the growth of regional governments’ debt could become a permanent phenomenon. This could lead to increasing problems in the fiscal and monetary management of the economy. It could also impede the convergence process toward European economic and monetary union (EMU). Consequently, the next step in the Spanish decentralization process must be to effectively secure budgetary discipline at the level of the regional governments. In this regard, a step in the right direction is the agreement to establish a new system of regional government finance for the period 1992–96 concluded between all regional governments and the central administration at the outset of 1992. In general, the new system preserves the main features of that in place during the last five years, though, given the need of a greater fiscal convergence within the EC context, some major steps are taken toward a better coordination between the fiscal stance of regional governments and that of the central government. In particular, the central government agreed with the regions to phase out most of their respective budget deficits by 1996 according to a path defined in the Convergence Plan, and to apply a system of controls on the execution of regional budgets. Moreover, prior authorizations were established for the issuance of bonds in both domestic and foreign financial markets.

Appendix Analytical Framework of the New System of General Grants

Systemic Restriction

In a public sector composed of the central government and the regional governments, the public sector’s budget constraint can be defined as follows:

GT + GG = ITX + RTX p + D , ( 1 )

where GT refers to the total amount of central government expenditures that has been relinquished to regions; GG refers to central government expenditures that continue to be administered by the central administration; ITX refers to central government’s total income tax collections; OTX refers to other central govern-ment’s taxes and revenue; RTXP is an estimate of “potential” tax receipts from those state taxes transferred to regions for the collection; and D is the central government’s deficit.

Equation (1) could be rearranged as follows:

( GT – RTX p ) + GG = I T X + OTX + D , ( 2 )

where the (GT—RTXP) equals those expenditure responsibilities transferred to the regional authority, but which are not being financed with taxes. These expenditures need to be financed with central government grants to regions.

Financing Arrangements of Individual Regional Governments

Assigned Revenues

Assuming that regional government i, to finance its total outlays, receives those income tax collections collected in its territory plus a portion of the whole public sector financing resources, we can write

ITX i + a i ⁢ ( OTX + RTX p + D ) , ( 3 )

where ai is defined as the “index of financial need” of regional government i, and Σ αi = 1.

Expression (3) would be the amount of assigned revenues to regional government i, if this regional authority would not have any taxes transferred from the state for its collection (RTXip = 0). Now, since each regional government has some taxes transferred from the state, whose potential revenue collections are estimated as R⁢T⁢Xip, the amount of net resources that the state needs to transfer to the regional government for its adequate functioning is only

R i = TIX i + a i ( OTX + RTX p + D ) – RTX i p ⁢ . ( 4 )

Assigned Expenditures

If the state would transfer Ri to each regional government without receiving any transfer from each of them, the state would exhaust its resources and would be unable to finance those central government expenditures that continue to be administered by the central administration (GG). Therefore, at the same time that the central administration transfers resources to regional governments for the financing of their activities, each region is obligated to finance some of GG. The share of regional government i in the financing of GG is defined as follows:

G i = b i ⁢ GG , ( 5 )

where bi is defined as the “index of economic capacity,” where Σ bi = 1.

Regional Government’s Net Financing from the State

Each region receives net financing from the state (NFi) defined as follows:

NF i = R i – G i , ( 6 )

or using equations (4) and (5) (6)

NF i = ITX i + a i ( OTX + RTX p + D ) – RTX i p - b i ⁢ GG ⁢ . ( 7 )

Then using equation (1) and some algebra we get

NF i = a i ⁢ GT + ( a I - b i ) ⁢ ( GG - ITX ) + ( ITX i / ITX - b i ) ⁢ ITX - RTX i p ⁢ . ( 8 )

Therefore, according to equation (8), each region receives net resources to finance its responsibilities, as a share of the total amount of central government expenditures that have been relinquished to regions (GT) based on the regions’ “index of financial needs” (ai). This is the term ai GT in the right-hand side of equation (8).

Also, each regional government receives resources for an amount equal to (ai - bi (GG - ITX), which is positive if ai > bi and negative if ai < bi; the term (GG - ITX) is assumed to be always positive: historical data show that GG is always higher than ITX. The term (ai—bi) (GG—ITX) represents the “redistributive” element of the system of general grants. Note that

Σ ( a i - b i ) ⁢ ( GG - ITX ) = 0. ( 9 )

For the relatively wealthy regions, ai < bi and therefore the system calls for a net withdrawal of resources. For the relatively poor regions, ai > bi and the system calls for a net transfer of resources. Thus, the coefficient (ai—bi) could be interpreted as an index of relative poverty.

The third term on the right-hand side of equation (8), (ITXi/ITX - bi), is a measure of “fiscal effort” of regional authority i as regards income tax collections in its territory. For this term, it holds that

Σ ⁢ ( ITX i / ITX - b i ) ITX = 0. ( 10 )

Regional Government’s Total Financing

Equation (8) represents the flow of resources from the state to regional government i. Each regional government, however, obtains a certain amount of resources from the taxes transferred from the state RTXi.

Therefore, the amount of regional government’s total financing from the state is

F i = a i ⁢ GT + ( a i – b i ) ⁢ ( GG – ITX ) + ( ITX i / ITX – b i ) , ⁢ ITX + ( RTX i – RTX i p ) ( 11 )

where the term (RTXi–RTXip), represents an incentive for better administration of state taxes relinquished to regions. If the administration of these taxes is good, then RTXi>RTXip and the difference benefits the regional government. Otherwise, the difference reduces the amount of total resources available to the regional government.

The Redistributive Criteria

In its simplest version, the model could become operative imposing specific parameters in equation (8). That equation represents the amount of net financing received by each regional government from the state. In practice, however, the distribution of resources across regions is defined in terms of the gross amount of resources transferred to the regional authorities, that is, without taking into account the estimate of “potential” tax receipts from state taxes transferred to regions (RTXP). Accordingly, the amount of gross financing is defined as

GF i = a i ⁢ GT + ( a i – b i ) ⁢ ( GG – ITX ) + ( ITX i / ITX – b i ) ⁢ ITX ⁢ . ( 12 )

Equation (12) shows the core of the distributional criteria of the system of general grants. It highlights the role of the index of financing needs ai, the index of relative poverty (ai—bi), and the index of fiscal effort of the community i (ITXi/ITX - bi).

The model agreed in November 1986 defined ai as the relative population of each community (Pi/P) and bi as the relative GDP of the community (GDPi/ GNP). Using these definitions, we can write equation (12) as

GF i = ( P i / P ) ⁢ GT + ( P i / P – GDP i / GDP ) ⁢ ( GG – ITX ) + ( ITX i / ITX – GDP i / GDP ) ⁢ ITX ⁢ . ( 13 )

Appendix Sectoral Aspects of Price Rigidities in Spain

Thierry Pujol

Divergence in inflation rates between sectors is a striking feature in many countries (see Table 19). It reflects differences in market structures, productivity, labor and intermediate consumption costs, and import contents. In most instances, between 1985 and 1989, the cumulated difference in inflation in the total consumer price index (CPI) and its services component ranged between 4 and 6 percentage points; it represented, very roughly, the productivity growth differential between services and other sectors. In Spain (and Italy), however, both inflation and the price differential are higher than in other countries. This raises questions about the pass-through of costs to prices in the different sectors of the economy.

Table 19.

Total CPI and Its Services Component in Selected European Countries, 1985–89

article image
Source: Organization for Economic Cooperation and Development (OECD), Main Economic Indicators, 1990.

Sectoral Price and Cost Indicators During the 1980s

After a period of deceleration, inflation in Spain has picked up in the last few years in almost all sectors (Table 20), with the main exception of industry. Recent developments in the service sector can be seen as especially worrying as the inflation in this sector stands well above average. This is a marked reversal of what happened during the early 1980s when prices changed at the same pace in the manufacturing sector and the service sector. These differences have been heightened by the opening up of the Spanish economy after 1986, which has intensified competitive pressure on industry.

Table 20.

Consumption Prices

(Year-on-year percentage change)

article image
Source: Bank of Spain, Statistical Bulletin.

It is difficult to draw firm conclusions about the influence of market structure on sectoral price behavior from a casual examination of the preceding table, for various reasons. First, as import content varies among sectors, the dampening effect of lower import prices has been more beneficial for the manufactured goods component of the CPI since 1986. Second, relative prices might also differ because of divergences in factor endowment and in their relative costs. This is particularly true for the energy-intensive sectors, which were helped by the fall in oil prices in 1986.

Import Prices and Domestic Prices

A number of interesting conclusions may be drawn from the comparison between sectoral GDP deflators (Table 21). First, agricultural and industrial price inflation has markedly cooled down since 1986 by more than 8 percentage points in the industrial sector and by 4 percentage points in agriculture. Second, inflation has also subsided in the service sector, although less markedly. At the same time, prices in the construction sector have registered an alarming upswing. Third, a longer-term view suggests that this inflation differential is, by no means, exceptional. Indeed, the price gap between industry and services averaged 3½ percentage points between 1965 and 1982 and can be related to fundamental discrepancies in productivity.1

Table 21.

Sectoral GDP Deflators

(Average annual increase in percentage)

article image
Sources: Malo de Molina and Ortega (1985); and National Institute of Statistics (INE).

Since 1987, import prices (Table 22) had, indeed, dampened the prices of food, energy, and consumption goods. Over the last four years, import prices for consumption goods and food have been almost stable, while import costs of energy have dramatically declined.

Table 22.

Imports Prices

(Year-on-year percentage change)

article image
Source: National Institute of Statistics (INE), Boletin Trimestral de Coyuntura.

One may distinguish various channels through which foreign prices influence domestic prices. First, they have a direct and immediate impact on the CPI through imported consumption goods. Second, import prices can also alter the domestic price setting, since one would expect the law of one price to hold in the long run. Third, import prices of energy, raw materials and equipment affect firms’ costs and prices. The first channel is likely to be more rapid and propitious to disinflation when the currency appreciates.

The beneficial influence of imports upon the CPI remains limited, however, as the openness of the Spanish economy still stands below the average of EC countries. In 1990, the import-to-GDP ratio equaled 17.8 percent in Spain and 23.3 percent on average for EC members. In some small open economies, like the Netherlands and Belgium, it reached, respectively, 45.6 percent and 61.6 percent. In addition, imports of consumption goods represented a relatively small share of total imports and only 6 percent of overall consumption in 1989, although this ratio had doubled between 1985 and 1989.

Domestic Costs and Output Prices

Is it possible to explain sectoral inflation differential by divergences in costs? As mentioned earlier, since productivity gains are, on average, higher in the manufacturing sector, one would expect prices to grow at a slower pace in a competitive environment. The same applies to intermediate consumption costs, since the use of energy or other raw materials is more intensive in industry.

Table 23 summarizes sectoral developments in wages and intermediate consumption costs over the 1980s.2 The evolution of wages was quite similar in the industrial and construction sectors,3 where wage increases decelerated until 1988 or 1989, and accelerated sharply in 1990. By contrast, in the service sector, wage increases gained momentum as early as 1987. Over the period 1986–90, average wage increases have not been significantly different among sectors. Thus, sectoral dispersion of changes in unit labor costs reflects mainly differential of productivity gains. In industry, large productivity gains contained unit labor costs between 1981 and 1985. Since then, productivity growth has slowed considerably, despite strong output growth. In the construction and service sectors, productivity gains were feeble or even negative, which, in the latter, reflects the increasing share of labor intensive activities. One may also note that intermediate consumption costs have increased at a slower pace reflecting the decline in the use of energy and raw materials.

Table 23.

Sectoral Price and Cost Developments

article image
Sources: National Institute of Statistics (INE); and IMF staff calculations. Note: IC price = Intermediate Consumption price; PPI = Producer Price Index for industry, GDP deflator elsewhere.

Econometric Analysis of Price and Cost Developments

Some producer price and CPI equations were estimated for the manufacturing sector and the sheltered sectors to test how import price and cost changes in the Spanish economy are passed on to production and consumption prices.

A Simple Model of Price Setting

Price setting is modeled as a two-stage process that depends on the behavior of producers and that of wholesalers and retailers.

Producer prices (P) are linked to unit variable costs (uc) in a simple mark-up model. Unit variable costs represent the sum of unit wage costs (ulc) and unit intermediate consumption costs (uic):

P = ( 1 + m ) ⁢ ⁢ ( ulc + uic ) ⁢ . ( 1 )

Note that for gross profits to be positive, the mark-up, m, has to be positive. In equation (1) the mark-up factor varies depending on degree of market competition. Accordingly, m can differ across sectors.

In an open economy like Spain, it is also apparent that foreign competitors influence domestic producers’ margins in the nonsheltered sector. For given unit costs, domestic producers increase their prices when import prices (pm) rise:

P = [ ( 1 + m ) ⁢ uc ] a ⁢ 1 ⁢ ⁢ Pm a ⁢ 2 , ( 1 ′ )

where a1 + a2 = 1.

Note that this model is also consistent with a less restrictive formulation where the mark-up itself depends upon foreign prices. Equations (1) and (1’) only capture relatively short-term behavior, since it is implicitly assumed that the capital stock is a quasifixed factor. More fundamentally, in the long run, new entrants in a given market progressively reduce profits, which implies that m tends to zero. The mark-up might then fluctuate depending upon demand. For instance, margins should increase when demand increases for firms facing increasing returns to scale. The opposite holds for decreasing returns to scale.

Also, firms may not react immediately to changes in unit costs for various reasons. One explanation is the cost of adjusting prices in the short run. Other models may imply that, possibly because of scant information, a firm facing cost changes waits until it knows whether other firms modify their prices. Consequently, such firms may possibly react differently to macro shocks (e.g., a shock in energy price) than to micro shocks. Similarly, a change in wages (w) or productivity (q) can lead to different results in the very short run. however, in the long term, the response of prices to changes in variable costs depends on the relative share of each factor in total variable costs. This constraint is imposed in the modeling of price-cost adjustments presented hereafter.

For a given sector, changes in producer prices are linked to changes in wages, productivity, unit intermediate costs, and possibly import prices as follows:

Δ log ( P ) = Σ t = 0 n a t Δ log ( w – t ) – b t Δ log ( q – t ) + c 1 Δ log ( uic – t ) – d t log [ P / ( ulc + uic ) ] – n – 1 + e Dem + f . ( 2 )
Δ log ( P ) = Σ t = 0 n a t Δ log ( w – t ) – b t Δ log ( q – t ) + c 1 Δ log ( uic – t ) d 1 Δ log ( Pm – t ) – e t [ log ( P ) – a 1 log ( Pm ) – n – 1 – a 2 log ( uc ) ] + f Dem + g ( 2 ′ )

In equations (2) and (2’) the error-correction term is respectively consistent with equations (1) and (1’). In addition, a proxy (Dem) intends to capture demand pressure. These equations were estimated imposing various constraints on lagged coefficients to overcome the lack of data. The selection of lags was decided on a purely empirical basis and was not based on a priori restrictions.

Changes in retail prices (CPI) (see equation (3)) are determined by a weighted average of changes in producer prices (P) and import prices (Pm) and additional variable costs, in particular those borne by the distribution sector (e.g., wages (w)). They also respond to the ratio of retail to suppliers’ price (Ps). Here again, the CPI for a group of products is unlikely to adjust immediately to changes in costs. The existence of buffer stocks, for instance, provides one explanation for lags in the adjustment of retail prices to changes in import prices.

Δ ⁢ log ( CPI ) = Σ t = 0 n ⁢ a t ⁢ Δ ⁢ log ( P – t ) – b t ⁢ Δ ⁢ log ( PM – t ) + c 1 ⁢ Δ ⁢ log ( w – t ) – d ⁢ log ( CPI / Ps ) – n – 1 + e ⁢ Dem + f ⁢ . ( 3 )

Note that equation (3) only captures the direct impact of import prices on the CPI. The producer price equation describes, instead, the indirect impact of import prices through unit intermediate consumption costs.

The Results

For the manufacturing sector, two alternative equations were estimated corresponding, respectively, to equations (2) and (2’) of the price setting model presented above. For both cases, our econometric results suggest that retail prices in manufacturing adjust only gradually to changes in costs. The specification of our estimated equations is as follows:

Producer price index, 1982 Q1-1990 Q3

Δ log ( P ) = 0.10 ( 3.2 ) Δ log ( w ) – 2 + 0.22 Δ log ( uc ) – 3 ( 3.9 ) – 0.18 ( 4.5 ) log ( P / uc ) – 4 – 0.006 ( 3.1 ) Dem + 0.10 ( 5.5 ) ( 4 )

R2 = 0.84 h = 1.7 σ = .49% F = 20.5

Dum = 1 after 1986Q1 and 0 before.

Producer price index, 1982 Q1-1990 Q3

Δ ⁢ log ( P ) = 0.25 ( 4.8 ) ⁢ Δ ⁢ log ( uc ) + 0.27 ⁢ Δ ⁢ log ( uc ) – 1 ( 8.6 ) + 0.09 ( 2.6 ) ⁢ Δ ⁢ log ( Pm ) + 0.09 ⁢ Δ log ( 2.5 ) ⁢ ( Pm ) – 1 + 0.005 ( 2.3 ) – 0.43 ( 4.5 ) ⁢ EC - 2 ( 4 ′ )

R2 = 0.71 h = 0.9 σ = .56% F = 15.2

where EC represents the residuals of the steady-state model:

log ( P ) = 0.84 ( 32.3 ) ⁢ log ( uc ) + 0.23 ( 9.3 ) ⁢ log ( Pm ) – 0.026 ( 0.5 )
R 2 = 0.99. D . W . = 1.3 σ = 1.1 % F = 6433.7

Simulation exercises using our estimated equation (4) show that a change of wages and intermediate consumption costs by 1 percent leads to a change in producer price of around 0.85 percent after one year. These findings are comparable with those of Dolado, Malo de Molina, and Ortega.4 The pass-through of changes in suppliers’ prices to the CPI is quite slow. Finally, the impact on the CPI amounts to one fourth of the initial change in costs after two years.

In the sectors other than manufacturing, price setting seems to be even stickier than in the production industries (this could be partly due to statistical problems; for instance, the definition of “Services” in the Input-Output table does not coincide with that used for the CPI). In services, where the distinction between producers and retailers is immaterial, only one third of the change in cost is passed on to prices after a year. In the food processing industry, the CPI lags well behind the evolution of costs. The specification of our estimated equation is as follows:

Services

Consumer price index, 1982 Q1-1990 Q3

Δ ⁢ log ( CPI ) = 0.18 ( 2 ⁢ 6 ) ⁢ Δ ⁢ log ( uc ) + 0.13 ⁢ Δ ⁢ log ( uc – 2 ) ( 2 ⁢ 0 ) - 0.06 ( 3.0 ) ⁢ log ( CPI / uc ) – 4 – 0.007 ( 3.3 ) – Dum ⁢ 8488 + 0.009 ( 5 )

R2 = 0.87 h = 0.3 σ = .45% F = 21.7

Food processing industry

Producer price index, 1982 Q1-1990 Q3

Δ ⁢ log ( P ) = 0.22 ( 3.3 ) ⁢ Δ ⁢ log ( uc ) + 0.26 ( 4.0 ) ⁢ Δ ⁢ log ( ulc ) – 2 – 0.17 ( 2.7 ) ⁢ log ( P / uc ) - 4 + 0.08 ( 2.8 ) ( 6 )

R2 = 0.66 h = 0.7 σ = .86% F = 9.2

Consumer price index, 1992 Q1-1990 Q3

Δ ⁢ log ( CPI ) = 0.7 ( 4.2 ) ⁢ log ( Ps ) – 0.04 ( 2.3 ) ⁢ Δ ⁢ log ( Pm ) – 0.004 ( 0.3 ) ⁢ log ( CPI / Ps ) – 1 + 0.03 ( 7.4 ) ⁢ Dum ⁢ 861 + 0.013 ( 3.4 ) ⁢ Dem ⁢ 862 + 0.011 ( 5.4 ) ( 7 )

R2 = 0.91 D.W. = 1.24 h = 1.2 σ = .38%

F = 39.2 Dum862 = 1 in 86Q2 and 0 elsewhere.

Agriculture

Consumer price index, 1981 Q1-1990 Q3

Δ ⁢ log ( CPI ) = 0.44 ( 4.2 ) ⁢ Δ ⁢ log ( Ps ) – 0.07 ( 2.0 ) ⁢ log ( CPI / Ps ) - 1 – 0.05 ( 2.9 ) ⁢ Dem ⁢ 833 + 0.044 ( 2.3 ) ( 8 )

R2 = 0.81 h = 0.5 σ = 1.66% F = 23.2

Dum833 = 1 in 83Q3 and 0 elsewhere.

One may also note that according to our econometric estimations, for the services sector, the food processing industry, and the agricultural sector the impact of demand pressure is weak or not statistically significant. This result has obvious policy implications since it would imply that demand has to be restrained longer to secure progress in disinflation. There are, however, many statistical problems that make demand pressure difficult to pick up econometrically. (Capacity utilization rates are surprisingly flat in Spain given the large swings in investment and growth experienced over the last years.) Firms are also likely to react in a discontinuous fashion to changes in demand, for instance, in order to acquire information about other firms’ reactions. In the construction sector, the surge in demand resulting from the large projects in Sevilla and Barcelona has been reflected locally in higher inflation and wages.

Conclusion

This appendix has provided some evidence about the stickiness of prices in Spain, especially in the nonindustrial sectors. Indications are that it is unlikely that wage policy alone could bring inflation down. In other European countries, the disinflation process may have been initiated through a temporary reduction in nominal wage increases (or equivalently some form of deindexation of wages to prices). Immediate, substantial transmission of this reduction in labor costs to prices would have minimized the reduction in real wages. By contrast, in Spain, our evidence suggests that wage reductions have had a weak direct impact on the CPI, involving a correspondingly large reduction in real wages and attempts at a wage catch-up process later on. One should keep in mind, however, that the estimates here presented underestimate the indirect price impact of wage cuts through their effect on the price for intermediate consumption goods. Second, the strong peseta, which has been instrumental in the progress of disinflation since 1986, also affected producers’ margins in industry. Indeed, firms respond quite promptly to changes in foreign competitors’ prices. As a stylized fact, one could say that prices in industry adjust quickly to foreign competition, while prices in the sheltered sector are sticky. Hence, a strong peseta may squeeze profits in industry without significantly reducing overall inflationary pressure. There is a risk that the exposed sector might suffer damages while sheltered sectors maintain their margins unchanged.

To understand the resilience of inflation it may well be interesting to focus on market structures, in particular in the distribution sector. Indeed, it would be beneficial to speed up the pass-through of suppliers’ price changes to the CPI, particularly for food. A survey by the National Institute of Statistics (1990) suggests that competition is lacking in some sectors where the ratio of margins to sales is excessively high (e.g., clothing, baked goods).
1

Preliminary information suggests that this share increased to around 20 percent in 1990.

2

The Constitution also defined those responsibilities that were to remain with the central administration, but indicated that some of them might be transferred to the regional authorities in the future.

3

Preliminary data suggest that regional governments’ total expenditure reached 8 percent of GDP in 1990.

4

In Spain, the central government comprises the state, the administrative autonomous agencies, and the social security system.

5

The system also did not provide resources to finance regional governments’ subsidies for free education. These subsidies, although not directly linked to the services transferred to regions, represented an important component in regional governments’ budgets.

6

The year 1986 was used as a benchmark year for setting the new system of grants.

7

For purposes of general grant financing, the concept of “potential” tax revenue collections, rather than “actual” collections of regional authority taxes was first introduced in the 1986 budget and had already been in place for one year at the time of the reform of the system of general grants.

8

The concept of fiscal effort was defined as the difference between the region’s share in total tax collections of personal income tax and the region’s share in total GDP. Similarly, the concept of relative wealth was defined as the difference between the region’s share in total population and its share in total GDP.

9

During 1987–90, social security and unemployment contributions represented about one third of the “structurally adjusted state taxes.”

10

To provide some degree of continuity with the previous system, it established that, for any regional government, financing under the new system would not be lower (in real terms) than the one received in 1986.

11

Another element leading to change in the original FCI was the 1989 reform of the system of EC Structural Funds. Hitherto, each regional authority in Spain had access to EC Structural Funds for an equivalent up to 30 percent of its FCI funds. Under the new system of EC Structural Funds, however, the decision capacity for the allocation of these funds across regions passed from the Spanish State to the EC Council.

12

Referred by Borrell and Zabalza (1990).

13

The idea of a new FCI where each of the nine regions selected would receive the same per capita grants was rejected because of the vast differences in the degree of economic development within this group of regions. For example, during 1985–90, the per capita gross value added of the poorest region in the group (Extremadura) was consistently 40 percent lower than the per capita gross value added of the richest region in the group (Asturias).

14

See Bird (1986) and Commission of the European Communities (1977).

15

For example, in the United States, the federal administration’s policy to reduce nondefense federal outlays has resulted in consistent withdrawal of federal funding from the intergovernmental financial system. The administration has followed a policy of “New Federalism” with two main elements: (1) increased independence to states and local governments in the administration of their own affairs; and (2) reduced federal fiscal assistance to these subnational tiers of governments. This situation has forced states and local governments either to find alternative means of raising revenues or to reduce their service provisions. In the United Kingdom, however, the slow growth of grants to local authorities since 1979 has been consistent with the central government policy of reducing local expenditure to affordable levels.

16

For example, Bird in Federal Finance cites a study by the United States Advisory Commission on Intergovernmental Relations emphasizing the problem of dealing with the fiscal problems in urban areas. He also refers to a study by J.S.H. Hunter (Federalism and Fiscal Balance, Canberra, Australian University Press, 1977) that stress the inadequacy of the Australian system with respect to large urban areas. According to the 1989–93 Regional Development Plan, in 1986 the nine regional governments, included under Objective 1 of the EC Structural Funds, covered 76.1 percent of Spain’s total territory and contained 58.2 percent of Spain’s total population.

17

See Mendel (1987).

1

According to Malo de Molina and Ortega (1985), prices, labor costs, and productivity have evolved between 1965 and 1982 as follows.

article image

Note that the average productivity differential between industry and services (3.3 percent) compares with the price differential (3.5 percent).

2

The method to build these indicators is as follows. Aggregating an input-output table for Spain provides a structure of variable costs for various sectors and subsectors (agriculture, food processing, industry, energy, construction, raw materials, semimanufactured goods, consumption goods, investment goods, private services, public services). For these sectors a breakdown of intermediate consumption between domestic and imported goods is available. Thus, by combining producer prices and import prices, it is possible to build a proxy of intermediate consumption prices. This method, however, is subject to criticism. Implicitly, it assumes that the Input-Output matrix has been constant since 1980. Yet, it has been shown that the consumption of fuel for instance has decreased since then (see Sanz (1986)).

3

Wage differentiation is indeed feeble in Spain, although not nonexistent. In 1990, average wages were equal in the manufacturing and services sectors. They were 20 percent lower in construction. Albarracin and Artola (1990) suggest that wage inequality, depending upon skills, may have increased between 1981 and 1988. A study by the Ministry of Economy and Finance (1990), based on a sample of large firms, also shows that profitable firms bestow higher wage increases.

4

Dolado, Malo de Molina, and Ortega (1985).

Cited By

Within Same Series

Other IMF Content

Other Publishers Content

  • Collapse
  • Expand
Cover Spain
Spain
Converging with the European Community
ISBN:
9781557753199
ISSN:
0251-6365
Publication Date:
15 Jan 1993
Publisher:
International Monetary Fund
DOI:
https://doi.org/10.5089/9781557753199.084
Pages:
65

  • Adams, Charles, and David T. Coe “A Systems Approach to Estimating the Natural Rate of Unemployment and Potential Output for the United States,” Staff Papers, International Monetary Fund (Washington), Vol. 37 (June 1990), pp. 232–93.

    • Search Google Scholar
    • Export Citation

  • Albarracin, J., and C. Artola “El Crecimiento de los Salarios y el Deslizamiento Salarial en el Periodo 1981 a 1988,” Banco de España, Research Memoranda No. 9001 (1990).

    • Search Google Scholar
    • Export Citation

  • Andres, J., and others, “La Inversion en España: Un Enfoque Macroeconómico,” in Moneda y Crédito, Segunda Epoca (1989), pp. 67–97.

    • Search Google Scholar
    • Export Citation

  • Artus, J. R., “Exchange Rate Stability and Managed Floating: The Experience of the Federal Republic of Germany,” Staff Papers, International Monetary Fund (Washington), Vol. 83 (July 1976), pp. 312–34.

    • Search Google Scholar
    • Export Citation

  • Baldwin, Richard, “On the Growth Effects of 1992,” NBER Working Paper, No. 3119 (Cambridge, Massachusetts: National Bureau of Economic Research, September 1989).

    • Search Google Scholar
    • Export Citation

  • Bentolila, S., and J. Blanchard “Spanish Unemployment,” Economic Policy: A European Forum, (April 1990), pp. 234–81.

  • Bergés, A., E. Ontiveros, and F. J. Valero “Estructura Comparada de la Gran Banca de la CEE,” Suplementos del Sistema Financiero, No. 28, Fundación Fondo para la Investigatión Ecónomica y Social (1989).

    • Search Google Scholar
    • Export Citation

  • Bird, Richard M., Federal Finance in Comparative Perspective, Financing Canadian Federation: 6, Canadian Tax Foundation (1986).

  • Borrell, José, and Antonio Zabalza Marti El Nuevo FCI: Un Instrumentode Politica Regional, D–90010 (Madrid: Ministry of Economics and Finance, Spain, September 1990).

    • Search Google Scholar
    • Export Citation

  • Bruneel, D., and others, “Nominal and Real Interest Rates in France, Recent Development, Determinants and Macroeconomic Effects,” Bank for International Settlements, CB384, March 1985.

    • Search Google Scholar
    • Export Citation

  • Caminal, R., J. Gual, and X. Vives “Competition in Spanish Banking,” Discussion Paper Series, No. 314 (London: Center for Economic Policy Research, July 1989).

    • Search Google Scholar
    • Export Citation

  • Cecchini, Paulo, Michel Catinat, and Alex Jacquemin The European Challenge, 1992: The Benefits of a Single Market (Aldershot, Hants, England, 1988).

    • Search Google Scholar
    • Export Citation

  • Cuervo, Alvaro, “La Crisis Bancaria en España, 1975–85,” Ariel Economía (1988).

  • De Pinies Bianchi, J., “La Regularización de la Liberalizatión Financiera Española,” Economistas, 40 (October-November 1989).

  • De La Dehesa, G., J.J. Ruiz, and A. Torres “Spain,” in Liberalizing Foreign Trade, Vol. 6, Part II, ed. by A. Choksi, M. Michaely and D. Papageorgiou (Oxford; New York: Basil Blackwell, 1989).

    • Search Google Scholar
    • Export Citation

  • Dolado, J. J., J. L. Malo de Molina, and E. Ortega “Respuestas en el Deflactor del Valor Añadido en la Industria ante Variaciones en los Costos Laborales Unitarios,” Banco de España, Research Memoranda, No. 8513 (1985).

    • Search Google Scholar
    • Export Citation

  • Dooley, M. P., and P. Isard “The Portfolio Balance Model of Exchange Rates,” International Finance Discussion Papers, No. 141, Board of Governors of the Federal Reserve System (1979).

    • Search Google Scholar
    • Export Citation

  • Fama, Eugene F., “Foward and Spot Exchange Rates,” Journal of Monetary Economics, Vol. 14 (November 1984), pp. 319–38.

  • Galy, Michel, “Le Marché des changes est-il efficient?” Revue Economique, Vol 39 (September 1988), pp. 913–20.

  • Gamir, L., “Política Arancelaria,” Revista de Economía, Información Comercial Española, No. 676–677 (December 1989–January 1990).

    • Search Google Scholar
    • Export Citation

  • Garcia Diez, J.A., “Política de Ajuste,” Información Comercial Española, No. 676–677 (December 1989-January 1990).

  • Giavazzi, Francesco, and Alberto Giovannini Limiting Exchange Rate Flexibility: The European Monetary System (Cambridge, Massachusetts: MIT Press, 1989).

    • Search Google Scholar
    • Export Citation

  • Gonzalez-Paramo, J. M., “La Autarqía Económica Bajo el Regimen del General France: Una Visión desde la Teoria de los Derechos de Propiedad,” Revista de Economía, Información Comercial Española, No. 676–677 (December 1989–January 1990).

    • Search Google Scholar
    • Export Citation

  • Gonzalez-Paramo, J. M., “The Role of the Public Sector in Spain: Structural Characteristics and Policy Adjustment on the Road to the EEC Internal Market,” Universidad Complutense de Madrid, Fundacion FIES-CECA (October 1989).

    • Search Google Scholar
    • Export Citation

  • Gordon, James, “Structural Funds and the 1992 Program in the European Community,” IMF Working Paper, WP/91/65 (Washington: International Monetary Fund, 1991).

    • Search Google Scholar
    • Export Citation

  • Hansen, L. P., and R. J. Hodrick “Forward Exchange Rates as Optimal Predictors of Future Spot Rates: An Econometric Analysis,” Journal of Political Economy, Vol. 88 (October 1980).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, World Economic Outlook, May 1992, World Economic and Financial Surveys (Washington, 1992).

  • International Monetary Fund, Government Finance Statistics Yearbook (Washington, various issues).

  • Judge, G.G., and others, The Theory and Practice of Econometrics (New York: Wiley, 1980).

  • Kouri Pentti, J. K., and M. G. Porter “International Capital Flows and Portfolio Equilibrium,” Journal of Political Economy, Vol. 82 (May-June 1974).

    • Search Google Scholar
    • Export Citation

  • Lazaro Araujo, L., “Política Regional Comunitaria, Evolution y Reforma del Feder,” Ministerio de Economía y Hacienda, Documentos de Trabajo, No:SGFEDERD-91006, 1992.

    • Search Google Scholar
    • Export Citation

  • Lopez-Claros, A., “The Search for Efficiency in the Adjustment Process: Spain in the 1980s,” IMF Occasional Paper, No. 57 (Washington: International Monetary Fund, February 1988).

    • Search Google Scholar
    • Export Citation

  • Malo de Molina, J.L., and E. Ortega “Estructuras de Ponderación y de Precios Relativos entre los Deflactores de la Contabilidad National,” Banco de Espana, Research Memoranda, No. 8505 (1985).

    • Search Google Scholar
    • Export Citation

  • Manas, Luis A., “The Impact of 1992 Financial Liberalization on the Spanish Financial Industry,” CEPR-European Commission Project on Integration in an Enlarged Community (mimeo, 1989).

    • Search Google Scholar
    • Export Citation

  • Mendel Camara, Braulio, “La Dimension Autonómica de la Deuda Pública,” in Papeles de Economía Española, No. 33 (Madrid: Fundación Fondo para la Investigatión Ecónomica y Social, 1987), pp. 329–37.

    • Search Google Scholar
    • Export Citation

  • Ministerio de Industria y Energía, “Informe sobre la Industria Española, 1989,” Centro de Publicaciones del Ministerio de Industria y Energía (1991).

    • Search Google Scholar
    • Export Citation

  • Ministry of Economy and Finance, Informe sobre la Reforma de la Impositión Personal sobre la Renta y el Patrimonio (Madrid, June 1990).

    • Search Google Scholar
    • Export Citation

  • Ministry of Economy and Finance, “La Negotiatión Colectiva en las Grandes Empresas,” Direction General de la Politica Economica (1990).

    • Search Google Scholar
    • Export Citation

  • Molinas, C., and others, MOISSES: Un Modelo de Investigation y Simulation de la Economia Espanola (1990).

  • Obstfeld, M., “Exchange Rates, Inflation, and the Sterilization Problem,” European Economic Review, Vol. 21 (1983).

  • Pietrefesa, N. “L’armonizzazione delle imposte dirette a la tassazione dele attivita finanziaria: analisi giuridica,” paper presented at the conference on “Le imposte del 1992,” held in Pavia, Italy on October 14–15, 1988.

    • Search Google Scholar
    • Export Citation

  • Plummer, Michael O., “Efficiency Effects of the Accession of Spain and Portugal to the EC,” Journal of Comparative Economics, Vol. 29 (March 1991), pp. 317–25.

    • Search Google Scholar
    • Export Citation

  • Sanz, R., “Efectos de Variaciones en los Precios Energeticos Sobre los Precios Sectorales y de la Demanda Final de Nuestra Economía,” Banco de Espana, Research Memoranda No. 8605 (1986).

    • Search Google Scholar
    • Export Citation

  • Spitäller, Erich, and Michel Galy “Spain: Landmarks in Economic Development, 1939–92,” IMF Working Paper, WP/92/78 (Washington, 1992).

    • Search Google Scholar
    • Export Citation

  • Termes, Rafael, “Los Resultados de la Banca en 1989,” in Papeles de Economía Española, Fundación para la Investigatión Económica y Social, No. 44 (1990), pp. 73–92.

    • Search Google Scholar
    • Export Citation

  • Tobin, J., “A General Equilibrium Approach to Monetary Theory,” Journal of Money, Credit and Banking, Vol. 1 (February 1989).

  • Trujillo, J. A., C. Cuervo-Arango, and F. Vargas “El Sistema Finánciero Español,” Ariel Economía (1988).

  • Viñals, J., and others, “Spain and the ‘EC cum 1992’ Shock,” in Unity with Diversity in the European Economy: The Community’s Southern Frontier, ed. by C. Bliss and J. B. de Macedo (Cambridge; New York: Cambridge University Press, 1990).

    • Search Google Scholar
    • Export Citation

  • Zellner, A. “An Efficient Method of Estimating Seemingly Unrelated Regression and Tests of Aggregation Bias,” Journal of the American Statistical Association, Vol. 57 (1962).

    • Search Google Scholar
    • Export Citation