The Impact of Member State Policies on...

The Impact of Member State Policies on Cohesion Final Report

1

The Impact of Member State Policies on Cohesion

FINAL REPORT

CONTENTS

Acknowledgements ............................................................................................................... 3 Summary ............................................................................................................................... 4 1. Introduction and background....................................................................................... 19

1.1 Context ...................................................................................................................... 20 1.2 Aims and approach of the study ................................................................................ 21 1.3 Structure of report...................................................................................................... 22

2. Conceptual framework and lessons from literature review ......................................... 23 2.1 Policy aims ................................................................................................................ 26 2.2 Policy instruments and their impact on cohesion ...................................................... 27 2.3 Convergence, technology policy and cohesion ......................................................... 30 2.4 Evidence on Member State policies .......................................................................... 32 2.5 Coordination between national and European policies ............................................. 40 2.6 Commentary .............................................................................................................. 44

3. Macroeconomic policies.................................................................................................. 45 3.1 Macroeconomic policy and ‘EU cohesion’ ............................................................... 46 3.2 Internal cohesion ....................................................................................................... 49 3.3 EMU, trade and competition ..................................................................................... 52 3.4 Commentary and key findings................................................................................... 53

4. Public expenditure and public transfers........................................................................... 54 4.1 Differing national systems......................................................................................... 55 4.2 Criteria for resource allocation.................................................................................. 56 4.3 Composition of public spending................................................................................ 61 4.4 Impact on cohesion.................................................................................................... 63 4.5 Commentary and key findings................................................................................... 67

5. State aid ........................................................................................................................... 69 5.1 Which public measures are state aid?........................................................................ 70 5.2 Impact of state aid on cohesion: empirical tests ........................................................ 74 5.3 Country studies.......................................................................................................... 75 5.4 Commentary and key findings................................................................................... 83

6. Territorial policies ........................................................................................................... 96 6.1 Decentralised policies: Austria, Belgium and Denmark............................................ 97


2

6.2 Central funding, but devolved implementation: Finland, Greece, and Spain............ 98 6.3 Complementing Community policy: Ireland and Portugal........................................ 99 6.4 Dealing with the dual economy: Italy and Germany................................................. 99 6.5 France ...................................................................................................................... 100 6.6 Netherlands.............................................................................................................. 100 6.7 UK ........................................................................................................................... 100 6.8 Commentary and key findings................................................................................. 101

7. National employment policies and regional development............................................. 107 7.1 The Mediterranean Member States.......................................................................... 112 7.2 The Nordic Member States...................................................................................... 115 7.3 The Central EU Member States............................................................................... 117 7.4 The United Kingdom............................................................................................... 119 7.5 Commentary and key findings................................................................................. 121

8. Technology policies....................................................................................................... 124 8.1 Spatial Distribution of Total R&D Expenditures .................................................... 125 8.2 Spatial Distribution of Patent Applications ............................................................. 125 8.3 Spatial Distribution of Government R&D expenditures ......................................... 126 8.4 Spatial distribution of technology indicators and cohesion..................................... 130 8.5 Spatial distribution of technology indicators and economic growth ....................... 132 8.6 Regional R&D policies............................................................................................ 135 8.7 Commentary and key findings................................................................................. 137

9. Regional Distribution of National FDI and its Impact on Cohesion ............................. 142 9.1 Influences on the location of FDI............................................................................ 142 9.2 Impacts on FDI of EU versus National Regional Policies ...................................... 146 9.3 Commentary and key findings................................................................................. 150

10. Conclusions ................................................................................................................. 154 10.1 Macroeconomic and budgetary policies ................................................................ 154 10.2 Economic development policies............................................................................ 156 10.3 Issues for cohesion policy ..................................................................................... 157

Bibliography...................................................................................................................... 159 Annex 1: DATA COLLECTION TEMPLATE AND QUESTIONNAIRE...................... 173

A General information on cohesion .......................................................................... 173 B Specific policies and their effects on cohesion...................................................... 173

Annex 2: EPRC Report on Regional Incentives................................................................ 178 Annex 3: Regional Profiles ............................................................................................... 187


3

Acknowledgements This study could not have been carried out without the help of all the officials from Unit A1 in DG Regio of the Commission who were directly involved in its planning and execution, and the assistance of a large number of officials in each of the Member States who contributed to the research by providing data and information on policies. We are grateful to all who have helped us.

In particular, we wish to thank Matthew Brooke (who served as the project officer on behalf of DG Regio), Nicola de Michelis and Vasco Cal for their support and encouragement.

We are also grateful to John Bachtler and his colleagues at the European Policies Research Centre at the University of Strathclyde for providing updated tables on Regional Policy Developments and for permission to reproduce the material in annex 2 of this report.

This study was prepared as one of the background studies for the Third Cohesion Report by a multinational team directed by Iain Begg, European Institute, London School of Economics and Political Science. The members of the team consisted of a core group who took primary responsibility for thematic studies and for providing reports on their respective Member States, and national correspondents who were responsible for producing reports covering their countries.

The core group leaders and their topics were: Gianfranco Viesti, CERPEM and Universita degli Studi di Bari – Employment Policy; Italy

Francisco Torres and Miguel Lebre de Freitas, Universidade de Aveiro – Macroeconomic Trends; Portugal

Carmela Martin, Carlos Mulas-Granados and Ismael Sanz, European Economy Group (EEG), Universidad Complutense de Madrid - Technology policies; Spain

El Mouhoub Mouhoud and Sébastien Dupuch, CEPN, Université Paris 13 – Literature Review; France

Phedon Nicolaides, European Institute of Public Administration, Maastricht – State Aids

Frank Barry, University College Dublin – Foreign Direct Investment; Ireland

Iain Begg and Vassilis Monastiriotis, London School of Economics – Public Expenditure and Public Transfers; Territorial Policies; UK

The national correspondent teams were led by: Gabriele Tondl, Institute for European Affairs, Wirtschaftsuniversität Wien – Austria

Norbert Van Hove and Laurent Van Depoele, Studiegroep voor Europese Politiek, Brussels - Belgium

Per Køngshoj Madsen, Institut for Statskundskab, Københans Universitet – Denmark

David Mayes, Suomen Pankki, Helsinki – Finland

Gaby Umbach, Lehrstühl Wolfgang Wessels, Universität zu Köln - Germany

Sarantis Kalyvitis, Athens University of Economics and Business – Greece

Dirk Strijker, Faculteit der Economische Wetenschappen, Rijksunversiteit Groningen - The Netherlands

Johan Lundberg, Centrum för Regionalvetenskap, Umeä Universitet - Sweden


4

Summary 1. Disparities between regions are affected by a wide range of factors that contribute to the

performance of economies and to the distribution of income among groups in society. Both Community and Member State policies play a role in these processes, and consequently have an impact on cohesion.

2. This report presents the findings of a study designed to assess the main categories of Member State policies that bear on cohesion. It was conducted by a multinational team comprising partners in 14 of the 15 Member States of the EU before its enlargement to 25 Members on the 1st of May 2004. The study findings contributed to the 3rd Cohesion Report adopted by the Commission in February 2004. The terms of reference were ‘to clarify to what extent – and how – national policies reduce economic and social disparities, and enhance cohesion’.

3. The research conducted included an extensive literature survey, together with collection of data on regional trends, policies and their incidence. It was, however, recognised from the outset that differences in national circumstances meant that there was no realistic prospect of obtaining data defined in a fully harmonised manner. Rather, the object of the data collection was to try identify the main influences on cohesion in the Member State and to assess the relative prominence of different types of policy.

4. This study focuses on the impact of those Member State policies expected to bear on cohesion. As can be seen from table S1, these policies can be expected to affect cohesion in diverse ways.

5. With an average of 47% of GDP devoted to public expenditure in the EU, national policies are bound to have a substantial impact, but it is no surprise that their impact on cohesion varies hugely. Some public policies are explicitly geared towards reducing regional disparities, some do so indirectly even though their primary objective is different, while some can aggravate regional divergence, especially those which target segments of economic activity identified as priorities for national economic development.

6. An underlying question is whether Member State policies compete with, or complement, Community policies and thus whether there needs to be fresh thinking on who does what. This study confirms the conclusion of similar work done in the course of preparing the 1st Cohesion Report which found that ‘national policies mostly concern aspects of social cohesion and tend to impact on regional cohesion indirectly’ and that the bulk of the transfers within Member States go to support consumption rather than investment..

The elusive concept of ‘cohesion’ 7. Cohesion can be an elusive concept and may be interpreted in a number of different ways.

It embraces: convergence in the sense of fostering the catch-up of more backward economies; economic development or regeneration for economies that have encountered difficulties; redistribution at the level of the household or the individual; and the assurance of a spatial balance across territories.

8. Member State policies affecting cohesion have to be considered at two levels. The first is the impact of policies on disparities within the Member State and the second, the consequences for regions compared across the EU as a whole. Member State policies will do little to affect disparities between Member States unless it is the case that the conduct of policies is significantly more effective in one country than elsewhere, with the result that it boosts overall performance. However, if they steer resources unevenly to certain regions, whether explicitly or indirectly, national policies can have a pronounced effect on the relative standing of different regions.


5

Table S1 Summary of impacts on cohesion of different policy areas Policy area Instruments or

mechanisms Direct effects Indirect effects Expected impact

on cohesion Scale of impact on cohesion or

disparities Macroeconomic policy

Common monetary policy; single market

Stabilises economic environment

Induces change in behaviour, e.g. in labour market

Positive for inflation-prone and previously volatile economies

Potentially large, but uncertain

Public expenditure

National programmes of current spending

Assures common levels of provision

Concerns about dependency

Positive in the short-term, but may have negative dynamic effect

Large, but can be two-edged

Public transfers Inter-governmental grants; social transfers

Large equalising effect on household current incomes

Depends on balance between stability and dependency

Positive in short-term, less certain longer-term

Large

Territorial policies

Subsidies; infrastructure enhancement; spatial planning

Competitiveness Long-run economic development

Positive, but depends on scale

Not great, except in rare cases; effects only visible long-run

Technology, R&D & innovation

Support for new technologies; innovation strategies

Stimulates new industries; product & process innovation

Helps to transform and modernise economic structure

Negative because most effort goes to few core regions

Small polarising effect

FDI Incentives and support packages

Adds to the export-base of host regions; net new jobs

Technology transfer; other externalities for host economy

Can be positive, but main impacts on national economy

Depends on balance between national and regional priorities

Labour market & employment policies

Functioning of labour market; raising employability

Better matching; improves employment rate

Higher labour productivity; raises attractiveness to investors.

Can improve growth trajectory; potential for ‘catch-up’

Primary impact on national economy; some benefits for high unemployment regions

State aids Broad range of subventions for companies: sectoral, ad hoc, horizontal

Retention of jobs and economic activity

Risk of slowing restructuring; can be anti-competitive; has opportunity cost

Often has perverse effect, especially between countries; unintended inter-regional impacts

Linked to ability to pay; incidence of costs not always evident

9. The Treaty, in article 2 TEU, elevates the pursuit of economic and social cohesion to being a fundamental aim of the European Union; then, in article 158 TEC, it highlights ‘reducing disparities between the levels of development of the various regions and the backwardness of the least favoured regions or islands, including rural areas’. Clearly, this wording signals that a key element of ‘cohesion’ is the promotion of ‘convergence’ in the sense of reducing disparities in GDP per head, a facet of cohesion that can be interpreted as primarily economic.

10. Social cohesion is also central to the aims of the Union and embraces not only traditional concerns about the distribution of income, but also employment levels and standards, as well as aspects of social policy that bear on the quality of life and equality of opportunity. In addition, there is an interest in ensuring that there is territorial cohesion in a number of senses, one of which is ensuring that the economic development of the Union does not become too polarised.

11. Conceptually, three very different categories of cohesion policy instruments can be distinguished: those that deal predominantly with the long-term competitiveness of regions by supporting transformation of the supply-side of the economy and thus which equate with economic cohesion; redistributive or palliative measures which enhance the


6

current standard of living of residents or attempt to deal with specific social problems; and policies which help to stabilise the demand-side of the economy. In practice, the divisions between these three types of policy are fungible.

12. Inevitably, measuring a concept as broad as cohesion is fraught with methodological difficulties. In practice, economic cohesion can largely be equated with differences in economic performance using simple macroeconomic indicators: income per head or the unemployment rate are the most used. Clearly, however, economic structures are also relevant, especially in thinking about dynamic aspects of the concept.

13. Social cohesion is less easily measured and attempts to do so typically require greater normative input. Nevertheless, methods for measuring inequality are well established and there has been extensive effort to construct robust objective indicators that can be used across the EU, with a list having been agreed (at the 2001 Laeken European Council) for social inclusion policy. A further concept of territorial cohesion concerns the balance of the spatial pattern of economic development. The challenge here is to deal with the simultaneous presence of excessive concentrations of activity and inaccessibility.

14. A further important distinction is between policies which result in improvements for regions relative to their respective national average (which we can label internal cohesion) and policies which affect the region’s position relative to the EU average (call this EU cohesion), whether or not the Member State’s position changes (figure S1).

DIMENSIONS OF COHESION

Social

Economic Territorial

IntraMember State

BetweenMember States

Figure S1 15. Pronounced institutional and political differences between Member also bear on cohesion

policy and its impact. Among the key dimensions are the scale of the public sector, the degree of decentralisation of authority, the competencies assigned to different levels of government and the mix of revenue sources. Rights and obligations in social protection systems can potentially have a marked effect on flows of resources, and the size and ownership of the national debt can impinge on which regions ‘win’ or lose’ from public policy.

16. One example is Portugal, where most government expenditure that takes place at the regional and local levels largely reflects policies determined nationally and where there tends not to be an explicit regional dimension to policy-making. Exceptions are: transfers


7

from central government, some employment policies and state aid for the autonomous regions. Moreover, regional authorities’ discretion (except in the two autonomous regions) over the way they spend the budget on the provision of services received from central government is very limited.

17. Many Member States have changed the institutional structures that bear on cohesion. Italy, for example, is in the process of a radical recasting of its regional structures and, as a direct consequence, of responsibilities for the assignment among tiers of government of policies that bear on cohesion. Territorial policies, traditionally based on a top down approach and the large use of financial incentives to backward regions, have also been changing since the mid-1990s. Public investment for the provision of ‘external economies’ and a complex process of institution building and vertical institutional cooperation is now centre-stage.

Cohesion challenges confronting Member State policies 18. The nature of the cohesion challenge varies greatly among the Member States and this too

is germane to the analysis of policy impacts. As is well-known, in two larger Member States – Germany and Italy – there is a marked geographical divergence between the lagging regions and the rest of the country. The remote sparsely populated regions of Finland and Sweden are the main problem in these countries, while in Denmark, the dichotomy is more between the rural and urban areas. In France and Spain, the legacy of agricultural decline has left a problem of development and diversification, whereas in Belgium and the UK, industrial decline coupled with urban problems have been evident. Austria, too has regional challenges that result from structural problems.

19. Greece, Ireland and Portugal have, with varying degrees of success, sought to ‘catch-up’ and have generally succeeded in converging towards the EU average in terms of GDP per head. They have, though all seen a degree of polarisation in which –especially - the more dynamic capital regions have seen the greater part of the economic development, and although nearly all other regions have also converged towards the EU average, most have done so to a lesser extent than the capitals.

20. Regional dynamics and the design of regional policies continue to have strong national characteristics and regional inequalities vary enormously. However, common features can be seen. Several countries experienced periods of regional convergence until the end of the 1970s, but internal disparities have since widened, with the paradox that disparities increase while support from national and supra-national authorities has been substantial and growing.

21. Many indicators of regional potential reveal a regional imbalance. Thus, inward foreign direct investment (FDI) has tended to gravitate towards the richer regions in most Member States, while the research and development effort of the EU is heavily skewed towards a relatively small number of regions which benefit from disproportionate shares of both private and public R&D spending.

22. There is a large gulf between European regions that invested most in total R&D in 2000, where figures in excess of 4% GDP are observed – attaining more than 7% in Braunschweig - and those that invested least – mainly in Southern Europe - where the ratio is below a quarter of a percentage point. The spatial distribution of patent applications – a measure of the output of research effort - is even more unevenly distributed across regions. Many regions, especially (again) in Southern Europe, filed fewer than 4 patent applications per million people in 2000, while at the other extreme, were regions which filed more than 180 applications per million people, the highest rates being in Oberbayern (441). However, although important, these indicators of technology performance tell only part of the story, as innovations can be adopted by firms which do not themselves invest directly in R&D or in patenting.


8

The macroeconomic dimension 23. Macroeconomic policy mechanisms that affect cohesion include the interest rate and the

exchange rate, the scale and mix of public expenditure, and stability in the trajectory of output and inflation. Essentially, the impact of macroeconomic policy will be ‘top-down’ to the extent that the effects on individual regions will depend on the overall stance and effectiveness of national policies, but there may be asymmetric impacts – both positive and negative - on different regions. For example, the reduction in the Italian debt service burden, consequent on full participation in stage 3 of EMU, indirectly benefits poorer regions. This is because they tend to have fewer holders of the debt than richer regions, yet have to share proportionately in raising the taxes to pay for the debt service.

24. In many policy areas, national provision is aimed at maintaining a common standard of public services, and thus has a pronounced effect in preventing the widening of disparities in living standards. Typically, major policy areas such as education, health, welfare provision and pension payments have this characteristic. If a region has a low productive base or a substantial dependent population, it will tend to receive net inflows of public funds, whereas regions that collect high taxes because of economic success are net payers.

25. Although low income regions tend to be net recipients of public flows, there can also be sizeable net flows to richer regions, especially if they are home to pensioners who receive substantial support or are accorded higher levels of spending for historical or political reasons. As a result, there is not necessarily a direct association between net inflows and relative GDP per head. Examples are Liguria in Italy and Scotland in the UK, both of which receive higher net inflows than relatively poorer regions in their respective Member States.

26. Public funds to support national economic development, by contrast, will often flow to richer regions because it is believed that is where they can be most productively used to enhance national competitiveness. Moreover, the companies or research institutes that secure such funds will tend to be located disproportionately in richer areas.

27. Member State policies towards FDI tend to focus primarily on attracting the investment to the country, rather than a specific location. Similarly, national and European funds for R&D initiatives are distributed across regions on the basis of ‘excellence criteria’, and the main recipients of these funds tend to be those regions which have a technological comparative advantage, typically in the most developed countries of the EU. Consequently, support for national excellence and competitiveness aims can be at odds with cohesion aims.

28. More generally, aggregate public investment as a proportion of GDP in EU-15 has averaged just over 3% in the last decade and of this, only a relatively small proportion has explicitly been geared towards the formation of capital directly linked to economic cohesion.

The impact of budgetary transfers 29. Net public budgetary transfers tend to increase the disposable income of weaker regions

in all Member States, while reducing that of more prosperous regions. Two distinct sorts of mechanism (figure S2) lie behind these transfers: explicit equalisation schemes that ‘tax’ richer regions to support poorer ones (‘horizontal’ transfers, of which the German länderfinanzausgleich is the best known); and the interplay of national taxation and social charges, social protection payments and grants from central government to lower tiers (‘vertical’ transfers, of which some for the purpose of financing sub-national tiers of government can amount to large proportions of recipient region income).

30. In some countries, the differences come about largely on the tax side, while in others it is the distribution of spending which matters most. These transfers are different in character from population related public expenditure flows, but analytically they are very similar in


9

their impact. Although there are vast differences between Member States in the institutional arrangements, the data for all countries show that the combined effects of these mechanisms has a marked effect in reducing regional disparities in living standards.

CEN TR A L G O V ER N M EN T

RIC H ER SU B -N A TIO N A L

A REA S

PO O RER SU B-N A TIO N A L

A REA S

H orizontalEqualisation

Per capitaPublic

Expenditure: C ore Services

Per capitaPublic

Expenditure: C ore Services

D iscretionaryspending

Taxes andsocial charges

Figure S2 Fiscal flows affecting regional income relativities 31. Because most public spending and transfers mainly affect current consumption, they

advance cohesion primarily by redistribution, but do not necessarily add to the productive potential of net recipient regions. Much of the first round effect will be to boost (reduce) activity in public services or population related private goods and services.

32. The net impact of public spending and transfers on regions is the balance between two large gross flows: public receipts and public expenditure. There are also awkward methodological issues in calculating impacts. For example, where the disparities between regions in per capita GDP levels are very large, the choice of denominator to benchmark the data becomes crucial. Low per capita receipts may actually constitute high proportions of regional GDP. As an illustration, in Italy, public spending (less interest payments) as a proportion of GDP is highest in Sardegna, and above the national average in all the Southern regions. But measured in terms of per capita receipts, the South seems to be poorly treated.

33. The powerful impact of public transfers on regional disparities can be seen from the orders of magnitude involved. In Sweden, one equalisation scheme reduces the county level coefficient of variation of income from 18.5% to 13.0%, with the income of the two most assisted northerly counties raised by around a third, while the transfer from Skåne in Southern Sweden is 9.7% of GDP and that of Stockholm 4%. Similarly, identifiable public expenditure in the fiscal year 2000/01 ranged from a low of 26.4% of GDP in the South East of the UK to as much as 63.9% in Northern Ireland, compared with a UK average of 35.9%.

34. An interesting dimension of net transfers is how fiscal consolidation affects them. Here, recent changes in Finland illustrate the impact. In 1996, all Finnish regions received net transfers from central government, and the scale of the net transfer for the three poorer regions was very substantial, with gross receipts exceeding gross tax paid by a ratio of 3:2. In 2000, the rapid fiscal retrenchment meant that the prosperous Uusimaa region, which received a small net transfer in 1996, was paying some 30% more tax than it received in central government spending. Meanwhile the three poorer regions saw their net gains shrink with expenditure much closer to taxes paid.


10

State aids and cohesion 35. Despite efforts to curb its use, all countries provide state aid to enterprises in their

jurisdictions, and it is estimated that the total spending on these state aids is roughly double that on the Structural Funds, that is close to 1% of GDP. The bulk of state aid is not directly related to regional development although it is usual for higher intensities of aid to be granted to undertakings in assisted areas.

36. There appears to be no overwhelming evidence that state aid either correlates with reduced regional disparities or that Member States grant more of it when disparities grow larger (box S1). Although the overall amounts of state aid in the EU have recently declined, some regions have received larger amounts of aid. The amounts of state aid received by each region fluctuate considerably from year to year. Some types of state aid like R&D aid are inversely related to regional income, with the richer regions receiving larger amounts of such aid.

37. By and large, most regional state aid goes to poorer regions. However, when examining only poorer regions, there appears to be no precise correspondence between regional income and either the overall amount of state aid or regional aid received by the poorer regions. In Member States with federal structures, regional authorities grant significant amounts of state aid. Since richer regions can afford to grant larger amounts of aid, the policy decisions of sub-national governments may also have a considerable impact on national cohesion.

Box S1 Key findings on state aids

Annual spending on state aids is more than double the outlays of the Structural Funds

Most state aids are not regionally targeted, but regionally targeted state aids go mainly to poorer regions. In the year 2000, regional state aid was between 0.4% and 9% of total state aid granted in the Member States, except in Greece and Ireland where it approaches a third of the total

Econometric tests found no evidence that state aids promote regional convergence

In Member States where regional authorities have substantial budgetary autonomy, sizeable state aids can be granted in richer regions

Some types of sectoral or industrial state aid are inversely related to regional prosperity

Regional receipts of state aids fluctuate considerably from year to year

Regional and other spatially targeted ‘territorial’ policies 38. Spatially targeted economic development policies generally have the aim of raising the

competitiveness of assisted economies. These ‘territorial policies’ comprise regional policies, local policies and urban and rural policies, and use a range of policy instruments. In many Member States, there have been marked changes in policy approach in recent years.

39. Such policies are typically relatively small in scale (figure S3), especially for some of the very specific measures, and can have conflicting objectives. For example, an urban regeneration policy may, quite legitimately, steer resources to a deprived district of a city in a rich region, yet in so doing compete with regional policy that supports poorer regions. Three factors seem to shape the nature and scale of regional and other territorial policies: the extent of internal disparities (especially Germany and Italy), the demands for co-financing of Structural Funds programmes, and the identification (especially in richer Member States) of specific, usually narrowly drawn, problem areas.


11

INTENSITY OF REGIONAL AIDMost heavily supported region

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Austria Finland France Germany Greece Italy Spain UK

% of GDP

Figure S3 40. In terms of total outlays, Germany and Italy account for more than half of all such

regional aid in the EU. Just under a quarter is accounted for by the four cohesion countries, with the rest split among the other Member States. Regional aid as a proportion of GDP is highest in three of the four cohesion countries (Greece, Ireland and Portugal), followed by Germany and Italy, with Spain lagging somewhat behind. By contrast, Sweden today has little more than assistance to new business start-ups as a form of incentive.

41. It is also important to note that responsibility for the funding, instruments and implementation of territorial policies varies with the institutional structure of the Member State. The provincial tier of government in Austria, for example, is largely responsible, while in Greece it is top-down, though with lower tier influence. In some Member States territorial policies simply reproduce the patterns of Community cohesion policies while in some others, territorial policies may conflict with EU spending.

42. The application of other regional policies varies across countries depending on their size, level of development, extent of duality, and history of economic restructuring. For example, Germany and Italy focus on providing support to more backward regions, while smaller countries with less acute regional divisions seem to focus more on smaller area policies (table S2). In some Member States territorial policies simply reproduce the patterns of EU regional and cohesion policies, while in others they are even in conflict.

Table S2 Main emphases in ‘territorial’ economic development policies Territorial priorities Link to Community policy Cohesion countries Lagging regions; selected

special areas Very strong co-ordination

Germany and Italy The lagging regions; special cases

Linked for objective 1 areas

Belgium and Austria Decentralised, emphasis on local problems

Not great

Finland, France, & Sweden Sparsely populated and economically un-diversified areas; spatial balance

Limited

Netherlands and UK Emphasis on problem urban areas and weaker regions

Some complementarity


12

43. Territorial policies in the Member States have some common characteristics but also exhibit key differences. In Austria, Belgium, and Denmark, policies are overseen by the central governments, but sub-national governments are responsible for the planning and financing of spatial policies. Finland, Greece, and Spain have a wide spatial coverage though with an emphasis on backward regions and policies are the responsibility of the central governments, which also provides the vast majority of funding, with smaller-scale policies (e.g., urban regeneration) not being particularly prominent.

44. Ireland, Portugal, and Sweden have few explicit territorial policies, although national state aids remain sizeable in Ireland. In Italy policies focus on the development of the lagging regions, mainly through public investment, are designed and financed by the central government, though there are doubts about their effectiveness. German territorial policies (both regional and urban) are directly designed and implemented at the Land level and are largely compatible with EU regional policy although their effectiveness is limited by their reliance on co-financing

45. Both France and The Netherlands have important, centrally administered, spatial planning programmes which address regional issues on the basis of the perceived needs of the country as a whole, with only limited explicit regional or urban policies. In the UK, long-term goals and priorities are designed (and financed) centrally, with emphasis on clusters, competitiveness, and urban regeneration, although specific territorial policies are designed and delivered at the regional and sub-regional levels.

Employment and labour market policies 46. Employment policies, potentially, make a big difference to the aggregate economic

development effort, depending on the structure and priorities of the social protection system, the degree of labour market imbalance and the institutional framework. Member State policies can be split between passive measures that provide income support and active measures that aim to reintegrate unemployed individuals into the labour market and promote human capital development.

47. The impact on regional cohesion of national employment policies is influenced by the different institutional setting in Member States with respect to welfare systems, the amount of both active and passive labour market expenditures, the extent of regional labour market imbalances, and the governance framework (table S3). Available data regarding regional expenditures and the number of regional beneficiaries (compared to target population) are not, however, always available, so that some of the conclusions in this policy domain must be seen as tentative and open to discussion.

Table S3 Attributes of broad clusters of employment/welfare models Model Scale of spending

on employment measures

Emphasis of employment

measures

Character of social protection

Mediterranean Well-below EU average

Mainly passive Limited and has gaps in coverage

Nordic High Strong role for active measures

Generous and inclusive, but

activating Central EU Fairly high Passive, but

with increase in active

Bismarckian with strong entitlements

UK Low Becoming more active

Safety net approach, high in some areas


13

48. The main finding is that different models co-exist within the EU, with different implications for regional cohesion, differentiated according to the scale of social protection outlays, the mix between active and passive labour market policies, and the role of public services, including employment services. These different patterns have been characterised as: a ‘Mediterranean model’, with relevant similarities among Spain, Portugal, Italy and Greece; a Nordic model, that may be described with reference to Finland, Sweden, Denmark and the Netherlands; a ‘Central European models’, that may be seen as the one in force in France, Germany, Belgium and Austria. The situation in the United Kingdom seems quite unique, although Ireland comes closest to it.

49. As can be seen from figure S4, there are large differences in the scale of spending on employment measures as a percentage of GDP, with a range from 0.89% for Greece to 4.74% for Denmark. Expenditure on all employment policies fell during the late 1990s, partly because relatively robust growth reduced the cyclical demand for passive policy spending, but partly also because overall pressures on public budgets in some countries led to curbs on active measures.

50. Within the European Employment Strategy, the regional dimension did not initially figure very prominently, although it has subsequently been accorded a higher weight and recent trends in a number of Member States point to a shift towards more decentralised employment strategies. One interesting innovation has been the establishment (for example in Austria, Italy and Portugal) of Territorial Employment Pacts. However, as many of these initiatives are very recent, other than measuring inputs in terms of public expenditure, it is difficult to ascertain how effective policy has been or to judge its impact on cohesion.

EXPENDITURE ON LABOUR MARKET POLICIES

00.5

11.5

22.5

33.5

44.5

5

Greece

United

Kingdo

mIta

ly

Portug

al

Austria

Spain

Sweden

France

Finlan

d

German

y

Netherl

ands

Belgium

Denmark

% of GDP

ACTIVE PASSIVE

Figure S4

Technology policies and their effects on cohesion 51. With public funding contributing around 1-1.5% of GDP, technology policies are a much

more substantial contribution to economic development than territorial policies. The weight of evidence is that, in most Member States, technology policies exacerbate rather


14

than reduce regional GDP disparities (figure S5) and tend to add to the competitive advantage of richer regions. This conclusion is most pronounced for countries where the policy is largely set and funded at national level; in countries where regional or municipal authorities have greater control, the impact of technology policies is more evenly spread.

REGIONAL DISPARITIES IN R&D SPENDINGRegional Spread and Member State Values

0

1

2

3

4

5

6

7

Belgium

Austria

German

ySpa

in

Denmark

Finlan

d

France

Greece

Irelan

dIta

ly

Netherl

ands

Portug

al

Sweden UK

EU-15

% of GDP1999

Figure S5

52. The experience of the nineties shows that while the distribution of total R&D spending became more unequal (led by an unequal expansion of its private component) it was partly offset by public R&D spending and patent applications. Even so, the disparities in regional R&D and patent activity are much greater than disparities in GDP per head. As the distribution of patents and public R&D converged, income per capita converged too. Since the only indicator that can be directly affected by policy-makers is the share of public funds that they dedicate to R&D activities, a tentative conclusion is that government R&D helped during the 1990s to mitigate the economic disparities that other technology indicators promoted.

53. Total R&D spending increases economic growth, especially if this R&D activity is quickly transformed into new patent applications. However, neither R&D figures nor patents tell the whole story, since innovation is the real key for economic growth. It is only when R&D spending is translated into innovation that economic performance improves. This positive effect on growth depends on the willingness of companies to adopt new ideas and methods, and to embrace new ways of organizing their business processes, as well as product innovation.

54. The overall trend (box S2) appears to be towards greater regional balance in the spread of technology policies, but from a very uneven starting point.Therefore, while technology policy based on pure excellence and efficiency criteria should remain as a policy tool for economic growth, this policy may need to be complemented by European and national regional policies which transfer funds to the least developed regions to maintain a minimum degree of economic cohesion.


15

Box S2 Summary of key points on technology policies

More investment in R&D is associated with more rapid growth, although it is also important to acknowledge the importance of diffusion and take-up of innovation.

Member States spend substantially more on technology policies than on spatially targeted economic development policies.

Technology policies typically focus on national competitiveness objectives, with the result that they tend to favour regions with an established lead in technological capability.

Private R&D is very heavily concentrated in favoured regions. Public R&D spending is less regionally concentrated, to some extent offsetting the impact of private, although the overall disparities are very pronounced, and the disparities in regional R&D and patent activity are much greater than disparities in GDP per head.

Inward foreign direct investment 55. In some countries, foreign direct investment (FDI) accounts for a high – though volatile -

proportion of overall investment (as measured by gross domestic fixed capital formation - GDFCF). In recent years, the range has been around 5% in Italy and 2% in Greece to very high percentages in the Netherlands, Sweden and Ireland (where the stock of foreign investment in 2000/1 was over 130% of GDP, compared with just 11% in Italy and Greece).

56. Data on FDI into EU regions suggest that inflows are disproportionately attracted towards richer regions in nearly all Member States (box S3). For example, in Germany, the new Länder, except for Eastern Berlin, received only about 2% of total FDI into Germany between 1998 and 2000 (about 22% of GDFCF over the period), while FDI was largely concentrated in a limited number of Länder (Nordrhein-Westphalia, Hesse and Baden-Württemberg, accounting for 71% from 1998 to 2000, with Bayern and Hamburg absorbing a further 17%).

57. Econometric work found that regional state aids have some small effects on distributing FDI towards less competitive, often peripheral regions. Only in a small number of cases, such as Ireland in the 1970s and the UK in the 1980s, has the regional distribution been affected strongly enough for FDI inflows to be a possible force for regional convergence.

Box S3 Summary of key points on FDI

There are very big differences among the Member States in the contribution of FDI to total investment, and in most Member States, the bulk of the FDI occurs in richer regions.

Regional state aids have a minor effect in steering FDI to less-favoured regions.

For most Member States, the key policy objective is to attract companies to their country rather than elsewhere, with the regional distribution of the investment seen as secondary.

Econometric work on Ireland showed that national regional grant policy did not have a statistically significant impact on the distribution of jobs provided by multinationals across the two broad regions

58. Even this latter finding ignores quality differences between the types of FDI attracted to core and to peripheral regions. Attracting high-quality FDI to lagging regions will remain difficult in the absence of integrated regional development programmes which combine infrastructural and human-capital development with carefully targeted innovation strategies.


16

59. A tension exists between the lure of pre-existing agglomerations and the part played by regional policies in attempting to attract FDI to more peripheral regions. Regional state aids, which are often co-financed by EU structural policies and which are consistent with EU competition policy, do appear to attempt to do this, though they are rarely successful. Though regional policies in some countries have been found to have had an effect on the location of FDI, in no case were the effects strong enough (or the resources devoted to regional policy large enough) to overcome the attractiveness of pre-existing agglomerations. The inference to draw is that the major emphasis of national policy towards FDI is on attracting projects to the country rather than attempting to distribute them across the regions.

60. Development of transportation infrastructure alone in lagging regions is unlikely to be sufficient to promote convergence in FDI shares. Instead, it is clear that other complementary policies – from human capital development to the promotion of innovation – have to be in place in such regions for FDI inflows to be able to contribute to regional convergence, emphasising the challenges from the finding that state aids in the form of assistance towards the development of R&D are allocated disproportionately towards already richer regions.

Overview: the impact of Member State policies on cohesion 61. The range of Member State policies bearing on different facets of cohesion is

simultaneously very broad and very diverse. Together, these policies have, in combination with Community policies, undoubtedly reduced the divergence that would otherwise have occurred in the living standards of residents of the poorest regions. In this respect they have manifestly promoted social cohesion. By contrast, their impact on the long-term potential for economic development of weaker regions, and thus advancing economic cohesion, is less visible. The effects of national policies on territorial cohesion are uneven, with some Member States having territorial balance as an explicit policy objective, while in others it is, at best, an indirect aim.

62. The impact of macroeconomic policies on cohesion is, potentially, a very powerful one, albeit one that operates primarily at the Member State level and especially for countries that have opportunities to achieve ‘catch-up’ growth. Equally, macroeconomic problems can exacerbate regional economic problems and the relative stagnation of the German and Italian economies has been inauspicious for the lagging regions in those two countries.

63. By increasing current consumption, public expenditure programmes and net fiscal transfers can create employment, while contributing to long-term growth by activating resources and by boosting local economic activity, even if they have no direct impact on competitiveness. But they can also have a negative effect on incentives and entrepreneurship and thus on the long-run capacity of the recipient economy to raise its competitiveness. Therefore there is a generally positive effect on social cohesion and a more questionable impact on economic cohesion.

64. National studies revealed that although there are manifestly common patterns to public intervention, there are also substantial differences in approach and policy ambitions. In many countries, too, the scale and direction of change has been great. An example is in France where policy no longer targets the reduction of territorial inequalities, but aims instead at promoting the sustainable development of the national economic space, taking into account the diversity and heterogeneity of regional circumstances. European funds are the more often larger than expenditures devoted to regional planning contracts (CPER), the main regional policy instruments.

65. Problems are also apparent in implementation. Thus, in Italy, the goals stated for public investment in the South of Italy are very far from being met. Sectoral policies remain dominant: examination of labour market policies, state aid for regional development and R&D policies reveals quite different territorial impacts. A key message of this report is


17

that the effect of national policies on internal cohesion in Italy is changing profoundly, and the positive impact of these changes on weaker regions cannot be taken for granted.

66. In all the Member States, public finance systems redistribute current income towards lower income regions (though not exclusively so, notably where historical compromises or other factors result in higher per capita receipts. However, they do so in very diverse ways and to considerably differing degrees that reflect the character of the public finance systems and the distributions of competencies among the tiers of government.

67. Many different national policies have the potential to affect the supply-side of the economy and thus to shape the competitiveness of regions. There can, however, be a tension at the heart of these policies, which is that they often have to reconcile national competitiveness and cohesion objectives. As a result their impact is ambiguous and even their underlying aims in relation to cohesion uncertain. This applies, notably, to state aids, technology policies and efforts to attract FDI. Similarly, despite the strong commitment by Member States to employment policies, there is little evidence that they contribute significantly to regional convergence.

68. Economic development policies that target problem (or priority) areas as defined by the Member State are used in all countries. Overall, their scale measured as a proportion of national budgets is very limited, although those regions which receive the highest levels can obtain support that can make a visible difference. In most Member States, the thrust of territorial policies does contribute to economic cohesion, although with the caveat that different policy initiatives can, on occasion, pull in different directions depending on the precise spatial targets of the policy (table S4).

69. A general conclusion to draw is that although Member State policies have a large and positive impact on aspects of cohesion, there continues to be an important role for Community policies to assist in promoting convergence, especially for the least competitive regions in poorer Member States.

Table S4 Summary of impacts of Member State policies on cohesion Policy area Economic Cohesion Social Cohesion Territorial Cohesion

Macroeconomic No clear effect, but ‘good’ policy helps

growth

Depends on fiscal policy choices

No direct effect

Public spending Some indirect effect via higher public

consumption

Strongly positive Positive effect in sustaining public

services Public transfers Some effect via

higher private consumption

Positive in most cases

Positive by supporting incomes

State aids Not systematic, but does not directly

favour poorer regions

Has some effect, but national, rather than

regional

May help to support weaker economic

bases, not systematic Territorial economic development

Helpful to cohesion, but has multiple

targets

Depends on precise targeting

Quite pronounced in some cases

Employment Some benefits Positive, mainly for high unemployment

areas

Can be helpful. Depending on

national structures Technology Negative: favours

richer areas No direct impact Tends not to be

supportive Attraction of inward investment

Negative: favours richer areas

No direct impact Depends on Member State priorities


18

70. Member State policies and Community policies are, on the whole, complementary in their objectives and, for the most part, in their impact. The former focus, on the one hand, on maintaining levels of public services across spatial boundaries and on dealing with specific territorial priorities – often political or social rather than economic. On the other hand, where national policies focus on competitiveness or on sectoral aims, their main impact is felt at the national level and their regional impact is not only indirect, but may also be at odds with economic cohesion aims. Moreover, to the extent that they support richer regions, some state aids may hinder efforts to stimulate the development of lagging regions. Community policies, by contrast, are specifically geared towards altering the economic development trajectory of eligible regions.

71. It is also noteworthy that in the Member States that receive substantial resources from Community policy, Community and national policy towards economic development have become quite closely inter-twined. The implication is that the approach required by the Structural Funds and the Cohesion Fund has spilled over to shape national policy on cohesion-related matters.


19

1. Introduction and background Within all Member States, there are substantial disparities between the most and least prosperous regions, reflected in a range of indicators of current economic performance and potential for future growth. Measured at regional level (NUTS 2 disaggregation), the biggest spreads are in Italy and Germany, but marked spatial differences exist everywhere, although the nature of the spatial disadvantage varies according to national circumstances. In some countries, it is evident that there is a gap between developed and more backward regions, while in other the spatial problem is associated with the decline of key industries, possibly associated with urban decline.

These disparities exist despite the fact that, in all countries, there are policies which act to mitigate them, either through direct targeting of the less prosperous regions or as an indirect consequence of the operation of the policy. These policies are extremely diverse in character, having been shaped by institutional and political legacies, developmental priorities and the economic circumstances of the Member State. Recent changes in the public sector, especially the trend towards devolution in many previously centralised economies, have also played a part. Without such policies, the differences between localities would almost certainly be much more pronounced, even if there are sometimes doubts about whether the mix of policies affecting a region always or necessarily improves its economic potential.

Other policies that have primarily national objectives may, by contrast, exacerbate regional disparities. This is because, although they may support the aim of raising national efficiency or competitiveness, their spatial incidence is predominantly on richer, more developed regions. As a result, even if such policies can be shown to boost the GDP of the Member State, it is possible that they do so either without raising the GDP of the least prosperous regions or by doing so less than in richer areas. Where this is the case, it can be argued that the impact of the policy is to diminish cohesion.

Although EU policies, notably the Structural Funds and the Cohesion Fund, play a substantial role in fostering economic and social cohesion, the sheer scale of Member State public expenditure compared with the limited resources available to the Community level inevitably means that, in most countries, they play the dominant role. In aggregate, national public expenditure ranges from 35% of GDP to nearly 60%, with an average of around 47 % in 2002. Consequently, EU spending at just 1% of Community GDP is just 2.5% of total public expenditure in the EU.

Complex processes and flows shape the effect of Member State policies on cohesion and these have, moreover, to be considered at two levels. The first is the impact of policies on disparities within the Member State and the second, the consequences for regions compared across the EU as a whole. A reasonable assumption is that certain categories – possibly even most - of Member State policies will do little to affect disparities between Member States, except insofar as a country which adopts ‘good’ policy will achieve better outcomes than a country which runs ‘bad’ policy. But the intra-Member State effects of policies can make a marked difference to the relative standing at EU level of different regions.

On the whole, the full array of cohesion-related national policies will tend to lessen disparities within a Member State, precisely because internal political processes tend to mean the better-off regions within that country are net contributors, while the poorer regions are net recipients. Some simple arithmetic illustrates the point. If a richer region has a GDP per head of 125% of the national average, but the country in question has a ratio of only 80% of the Community average, then the rich region will have a GDP per head exactly equal to the Community average. National policies which impose a net ‘tax’ on the said rich region equivalent to 10% of its GDP would result in its disposable income falling to 90% of the Community average.


20

Concepts Although the notion of cohesion is both intuitively obvious and formally linked in the Treaty to regional disparities, it embraces a diversity of concepts. These can yield different results and point to conflicting policy conclusions depending on which facets of cohesion are under scrutiny. Because Member State policies have a variety of aims, it is important at the outset to clarify these concepts and what is being measured. In addition, it has to be recognised that despite policy efforts, there may be little or no measurable change in indicators of outcomes or regional performance, especially if what the policy does is simply to counter what would otherwise be a widening of disparities. Instead, the most that can realistically be done is to estimate the ex-ante effect of the policies, principally by assessing the magnitude of the resources devoted to cohesion policies.

A pro-cohesive policy can therefore be defined as one which, other things being equal, increases the resources at the disposal of the region, thereby giving it the opportunity to reduce the gap between its performance and the national mean. Whether or not it achieves such a reduction will depend on the effectiveness of the policy and cannot be guaranteed, so that the emphasis is on inputs rather than outcomes. In similar vein, an anti-cohesive policy is one that steers resources to regions that are already favoured. Even then, distinctions need to be drawn between measures which primarily target social cohesion, that is improving the lot of individuals, policies which raise the current income of regions and help to maintain an economic base (territorial cohesion) and those which enhance the long-run economic prospects of a region, thereby engendering economic cohesion. Yet, although these different conceptions of cohesion may be clear, the practice is usually much more muddied: a given policy may well have an impact on all three types of cohesion, even if its main purpose is to act on one or the other.

It is, however, important to stress immediately that there are different dimensions of ‘cohesion’ and that within the range of policies, some bear only on certain facets of cohesion. National policies, especially those which target current incomes and of recipient persons and regions, cover a much broader range of ‘cohesion’ related objectives than Community policies. In spite of some overlap, however, the main thrust of Community policies is to promote economic cohesion.

Aims of the report This report presents the findings of a study designed to assess the main categories of Member State policies that bear on cohesion. The project was one of several background studies for DG Regio of the European Commission, for which calls for tender were launched in the summer of 2002. It was conducted by a multinational team comprising partners in 14 of the 15 Member States of the EU before its enlargement to 25 Members on the 1st of May 2004. The study findings contributed to the 3rd Cohesion Report adopted by the Commission in February 2004. The terms of reference, as published in the Official Journal in call No. 2002/S 132-102933 invited contractors ‘to clarify to what extent – and how – national policies reduce economic and social disparities, and enhance cohesion’.

The research conducted included an extensive literature survey which sought to identify both material produced in national research and studies and broader contributions to the academic and policy literature. Data on regional trends, policies and their incidence were collected for all Member States with the exception of Luxembourg, using a common questionnaire. It was, however, recognised from the outset that differences in national circumstances meant that there was no realistic prospect of obtaining data defined in a fully harmonised manner. Rather, the object of the data collection was to try identify the main influences on cohesion in the Member State and to assess the relative prominence of different types of policy.

1.1 Context There has been an abundance of ad hoc work on the impact of specific policies on different facets of economic and social cohesion in different settings, but the only comprehensive


21

previous attempt to assess how Member State policies affect cohesion was conducted as a background study for the 1st Cohesion Report. That study, by the EPRC and L’OEIL (1996)1, examined a range of policies and provides an excellent foundation for the present study. It concentrated on five main categories of policies: Member States’ regional policies; other explicit spatial policies; ‘horizontal’ policies with a potential, if not necessarily intentional spatial impact; employment policies; and inter-regional transfers. While there have been important changes since in the policy framework and in the scale and thrust of particular policies, the principal empirical findings are likely to remain broadly valid.

Previous work on regional disparities and on the impacts of different sorts of policies can provide some insights into the balance between national and Community policies in advancing cohesion. Theoretical contributions offer some pointers, but it is mainly empirical studies that are useful. The latter includes the various analyses done at Community level (including successive Periodic Reports and the first two Cohesion Reports), evaluation looking at the incidence of particular types of policy or developments in particular regions, and comparative work, including with the US.

The 2004 enlargement of the European Union to ten Central and Eastern European Countries (CEEC’s) increased regional heterogeneity within the EU. However, the reduction of territorial disparities has been an objective of the European authorities since the adoption of the Treaty of Rome in 1957. The EU’s approach to cohesion is to encourage convergence in per capita GDP and unemployment rates between European regions. On these two criteria, regional differences in Europe are great. One quarter of the population of EU-15 lives in a region which has a per capita GDP below 75 per cent of the EU average, the Objective 1 threshold. If the same criterion were applied in the United States, only two states containing 2% of the population would be eligible (Puga, 2001). Moreover, there has been convergence across Member States since the beginning of the 80’s, but regional disparities within EU countries have not narrowed and even widened. This result can be regarded as puzzling, given that resources aimed at reducing disparities have increased substantially. EU structural operations amount to only 0.4% of Community GDP, but this amount is low compared to Member State public spending averaging 47% of GDP. Although the bulk of this spending is not explicit regional policy, the interaction of fiscal revenues and public expenditures favours the most disadvantaged regions which contributes less to the state budget and receive more because of more acute social needs.

The various instruments of explicit regional policy are a public policy response to the threat that economic integration will aggravate regional disparities. However, regional policies have been criticised as ineffective and some critics point to the role of the Member States and the fact that national factors in regional development are not well understood (Fayolle & Lécuyer, 2002). The literature on this topic remains scarce because of the lack of available empirical data. Moreover, the diversity of national institutions, the specific problems in each Member State and the variety of national strategies of regional development explain why such a study is rather complex. Nevertheless, national transfers mechanisms, state aids, inter-regional transfers, together with national co-financing of Community regional policies are manifestly much bigger than Community interventions.

1.2 Aims and approach of the study The present study aimed to cover similar ground to the 1996 report, but to offer an updating and extension, and to reflect the key policy developments in the last seven years. With twelve of the EU-15 participating fully in monetary union, the establishment of the European Employment Strategy, and the articulation of the ‘Lisbon’ objectives covering 1 European Policies Research Centre and Laboratoire d’Observation de l’Economie et des Institutions Locales (1996)


22

competitiveness and social cohesion, it is clear that the framework for policy has evolved markedly. EMU, in particular, changes the policy environment by introducing a new macroeconomic policy regime. For some Member States, the immediate impact of monetary union on the principal macroeconomic variables that shape economic performance has been considerable, and, in turn, has affected the rate at which the lower income Member States have converged

The study brought together national teams representing fourteen Member States (Luxembourg was not researched), several of whom also took responsibility for integrating the work on key policy areas. A work plan was agreed at a kick-off meeting that brought together the core team and Commission officials, held at the Commission offices in Brussels on December 5th 2002. Further meetings were held on the 13th of March (the project co-ordinator and Commission officials) and the 12th of June (core team and Commission officials). Early work concentrated on elucidating concepts and constructing a ‘questionnaire/data template’ (a copy of which is appended as Annex 1) to guide the data collection carried out by national experts in 14 Member States. As an input to the 3rd Cohesion Report, initial findings were submitted to the Commission at the end of August 2003, and a draft report in November 2003.

1.3 Structure of report Although the aims of cohesion policy at the Community level are explicit in the Treaty, different factors come into play in looking at cohesion at the Member State level. These are explored in Chapter 2 which draws on previous work to set the scene by exploring some of the conceptual and political economy considerations that bear on cohesion policy. The third chapter examines macroeconomic developments and policies, which manifestly play a large part in shaping the trajectory of economies. Chapter 4 then examines the role of public expenditure and intra Member State inter-governmental transfers, the dimension of Member State policies that, quantitatively, has the greatest potential to correct disparities in regional living standards. The focus in the subsequent five chapters shifts to different sorts of supply-side policies that have the potential to affect cohesion, either by acting directly on regional capacity and performance, or by establishing conditions that favour certain classes of regions. The policies examined, which do not necessarily promote greater cohesion, are: state aids, ‘territorial’ polices, employment measures, technology policies and inducements to inward investment. Conclusions and a comprehensive bibliography complete the report.


23

2. Conceptual framework and lessons from literature review Cohesion is not an easy notion to define and, although there is often a tacit understanding of what it means, it is open to a variety of interpretations. It arguably embraces inequalities, whether in income, living standards, employment or environmental conditions, and also has to be seen in terms of opportunities as well as outcomes. The related notion of ‘convergence’ is also germane, though with the emphasis on ‘real’ variables rather than the nominal criteria that were embodied in the obligations Member States had to fulfil to be eligible for stage 3 of EMU. Convergence or divergence, however, are long-term processes that reflect both history and the effects of recent trends that shape the ability of a local economy or social group to compete. Cohesion in this latter sense is about ensuring that the least well-off are able to share in the benefits of economic integration, possibly disproportionately.

The processes that give rise to disparities are, themselves, dynamic ones, and include the impact of economic integration. A ‘state of cohesion’ may exist at a point in time, yet be based on measured disparities that would be unacceptable at a different time: what is acceptable or tolerable today may no longer be so in the future, especially in the context of an ‘ever closer union’. This dimension of cohesion can be expected to loom larger once the anticipated accession of central and eastern European countries (CEECs) takes place. The disparities they exhibit today may become increasingly incompatible with the aims of the Union if they are not reduced. Yet it is clear that defining precise thresholds for what constitutes a state of cohesion, or targets for phased reductions of ‘non-cohesion’ requires difficult normative judgements.

Other secular trends are also important, and the roles of important drivers such as technology or investment have to be taken into consideration. It can be argued that, with the full implementation of EMU, new risks and opportunities will arise, some of which may be adequately dealt with by Member State policies, while others require fresh analysis. In particular, EMU can be seen as a new and significantly different regime that will induce important macroeconomic changes. These will, moreover, interact with other processes that bear on cohesion.

Empirically, cohesion embraces both short-term comparisons that highlight divergence in the economic circumstances of economic entities – with targets including territories and social groups – and longer-term dynamic changes relating to the realisation of potential. There are elements of countering disadvantage in the concept, mixed up with provision of opportunities. Dynamic considerations also point to the difficulty of constructing indicators that can used to monitor cohesion, especially over time, and to identify the policy responses that might be required to promote it. In the same vein, a wide range of policies bear on the achievement of cohesion, some of which are explicitly aimed at assuring cohesion, while others have effects that are tangential, indirect or - on occasion - contrary to cohesion.

In assessing the impact of Member State policies, it is consequently useful to examine whether the aim of the policy is directly to affect cohesion or whether its incidence is largely indirect, the mechanisms by which different categories of policy do so and the facets of cohesion addressed, the expected magnitude and direction of the policy’s impact, and the configuration of policy instruments. Economic and social cohesion are often lumped together, yet plainly relate to very different underlying societal processes and, a fortiori, policy problems. In practice, economic cohesion can largely be equated with differences in economic performance using simple macroeconomic indicators: income per head or the unemployment rate are the most used.2 Clearly, however, economic structures are also relevant, especially in thinking about dynamic aspects of the concept.

2 Although such indicators should, in principle, be easily comparable across territories, differences in statistical conventions, methodologies or quality – even with supposedly harmonised data – often make it hard to compare like with like.


24

Social cohesion is less easily measured and attempts to do so typically require greater normative input, even though many analytic techniques are available. Methods for measuring inequality are well established and there has been extensive effort to construct robust objective indicators that can be used across the EU (Atkinson et al., 2002), with a list having been agreed (at the Laeken European Council) for social inclusion policy. Relative unemployment rates can also be readily calculated for different segments of the working population and there is a growing number of hard indicators for different facets of social exclusion. But other variables, such as the poverty rate which requires agreement on a benchmark rate for the ‘poverty line’ are more open to dispute.

A further concept of territorial cohesion can be derived by looking at the challenges from a geographical and spatial planning standpoint. If economic trends result in a spatially unbalanced pattern of economic development, action to redress the imbalances may be warranted. Spatial balance leads to the further important distinction between policies which result in improvements for regions relative to the national average (which we can label internal cohesion) and policies which affect the region’s position relative to the EU average (call this EU cohesion). How the Member State as a whole performs will have a very large bearing on the latter, whereas it is national policies that will dominate the former. Figure 2.1 brings together these different effects.

DIMENSIONS OF COHESION

Social

Economic Territorial

IntraMember State

BetweenMember States

Figure 2.1 Building on the conventional distinction in the theory of public finance between allocation, distribution and stabilisation – explained notably in the work of Musgrave (1959) – three very different strands of cohesion policy instruments can be distinguished. A first set deal predominantly with the long-term competitiveness of regions by supporting transformation of the supply-side of the economy: their target is allocation. Second, there are measures which enhance the current standard of living of residents or attempt to deal with specific social problems, both of which are, thus, essentially redistributive or palliative in their impact. Third, there are policies which help to stabilise the demand-side of the economy. In practice, the divisions between these three types of policy are fungible. Supply-side measures will also have some redistributive effect, though not necessarily or always from rich to poor regions, just as redistributive measures are often the means by which stabilisation is achieved.


25

In addition, a distinction can be drawn between policies that seek directly to alter relativities between regions and others that do so only as a by-product of their primary objectives, even if these are cohesion-related. In the former category, the most obvious examples are programmes that are only open to regions which fulfil some eligibility criterion, such as state aids that are only available (or set at a higher rate) in regions with unemployment rates over a certain threshold. A good example of indirect regional impacts (which can be both positive and negative) is the incidence of the social protection system, which – to simplify greatly - collects resources from those in work and distributes to the inactive or the needy. Where a principal recipient group is the involuntarily inactive (notably the unemployed) the policy will tend to be pro-cohesive in relation to regional income disparities, since it can be assumed that regions with high proportions of the inactive will have lower GDP. But to the extent that the system transfers to pensioners in rich regions, the impact could appear to be anti-cohesive (although clearly fulfilling an important role in promoting social cohesion).

Policies with national allocative objectives can also have a skewed regional impact which can, moreover, be anti-cohesive. This is often the case with national competitiveness-promoting policies such as support for new technologies or innovation: numerous studies have shown both that there is a strong spatial concentration of R&D and related activity, and that it tends to cluster in richer regions.

In the 1st Cohesion report, it was noted that ‘national policies mostly concern aspects of social cohesion and tend to impact on regional cohesion indirectly’ and that the bulk of the transfers within Member States go to support consumption rather than investment. In this regard, national policies are aimed at a different objective from Community policies. The empirical evidence assembled in this report confirms that the weight of Member State cohesion policies remains firmly on support for current consumption.

Institutional structure and change There are very pronounced differences between Member States in the character of the public sector and the form of political institutions, and these differences inevitably bear on cohesion policy and its impact. Among the key dimensions are the scale of the public sector, the degree of decentralisation of authority, the competencies assigned to different levels of government and the mix of revenue sources. Rights and obligations in social protection systems can potentially have a marked effect on flows of resources, and the size and ownership of the national debt can impinge on which regions ‘win’ or lose’ from public policy.

One of the main consequences of different institutional circumstances is that where competencies for economic development policies are extensively devolved to lower tiers of government, the financial resources for this purpose are part of the general budget of the devolved government. Thus, a German or Austrian Land government may have extensive policies in this domain, but the only measurable flow is the general fiscal transfer to the region. Depending on the obligations they have to meet, the same flow could be used by different Land government in very diverse ways, with one choosing to favour consumption (i.e. social cohesion), while another gives much more weights to economic development (and thus economic cohesion). By contrast, top-down policies administered by central governments can more readily be split between different categories of policy that impinge on cohesion. Clearly, these differences make identification of common policies problematic, if not impossible in some cases.

Scale of policies The revenues of the public sector as a whole in EU countries accounted, according to the most recent data for 2002, for between 34.1% (Ireland) and 58.9% (Sweden) of GDP, with an EU-15 average of 45.7%. Figures for total public spending show that the range in 2002 was from 35.2% of GDP in Ireland to 57.6% in Sweden, and an EU-15 figure of 47.6%. It is also important to note that over a period of a few years there can be marked changes in these levels. Within these totals, there are significant difference in the compositions of both


26

revenues and expenditures. In particular, public gross fixed capital formation in 2002 varied between 4.5% of GDP in Ireland and 4.8% in Luxembourg, to just 1.3% in the UK and 1.2% in Austria. Adding other capital expenditure and capital transfers narrows these disparities somewhat, but also shows that there is no obvious relationship between the size of the public sector and its willingness to invest. There are statistical limitations3 in these figures that have to be borne in mind, but they do signal the scale of public activity.

2.1 Policy aims A wide range of cohesion policy objectives can be identified, some of which are only really capable of being implemented through action at the Member State level, while others may involve both Community and national efforts. In what follows the main focus is on economic cohesion or, put differently, the ‘regional problem’. Three broad categories of policy objectives can be enumerated.

The first concerns competitiveness, itself a slippery concept. Several of the policy areas that bear on cohesion can be construed as competitiveness enhancing measures. Their purpose is to correct shortcomings on the supply-side of the economy that result in a deficient economic performance. Policies aimed at boosting regional competitiveness though raising the long-run level of productivity include public investment (widely defined to include ‘soft’ as well as ‘hard’ infrastructure), support for entrepreneurship and SMEs, training and other labour market measures and the fostering of R&D, innovation or other technology-related initiatives. Regulatory interventions such as universal service provisions in network industries, as well as direct public expenditure on supporting a specific investment, can have significant potential for affecting cohesion under this heading.

A second category of cohesion measures is those that boost demand in an area and thus act in a preventative manner to forestall the emergence of disparities. It is also recognised that stable demand favours longer-term economic development. In the chapter in the 1st Cohesion Report that dealt with national policies, great weight was attached to the importance of macroeconomic policies in providing the platform for ‘catching-up’ and it is clear that the overall conduct of macroeconomic policy is bound to loom large in the trajectory of demand for any region. Interest rates, the volatility of prices and of output, and the real exchange rate vis-à-vis major trading partners will all bear on demand. From a regional perspective, national wage-setting behaviour and the effects of macroeconomic factors on asset prices can shape both demand and input cost competitiveness. The inter-play of public expenditure and taxation, especially, performs an important role autonomously in stabilising regional demand. Regions with below average economic activity will tend to receive proportionally higher inflows of public expenditure, while dynamic regions pay more in tax revenue. It is well-established that measures of this sort attenuate short- to medium-term fluctuations in regional demand that might otherwise give rise to problems of cohesion.

Third there are measures that deal, above all, with the consequences of a lack of cohesion. These may be primarily intended to have solidaristic and palliative roles, but can also complement action to reinvigorate the supply-side and, indeed, to stabilise demand. Much of the contemporary effort to counter social exclusion is predicated on the simultaneous offer of assistance and incentives towards self-improvement. Certain labour market initiatives, in particular, span the two aims, highlighting the fact that a policy can have impacts of diverse sorts. One of the key decisions of the Lisbon European Council, in this regard, was to seek to promote social protection as a productive factor.

3 In particular, during the period 2000-02, the impact of UMTS licence receipts is very pronounced. These count as ‘negative’ spending and have been added back to the totals shown in figure 2.3, below.


27

In practice, Community policies to promote cohesion (through the Structural Funds and the Cohesion Fund) are principally focused on the first of the three sorts of factors, insofar as they target the longer-term upgrading of less competitive regional economies, implying that this is the main focus of interest with respect to cohesion. Even then, it is clear that EU support can help those countries that received the highest levels of transfers to pursue other aims. The other dimensions of cohesion policy also matter and manifestly should not be neglected in a study of the impact of national policies. Because the Community budget, at just over 1% of EU GDP, is small and cannot, in any case, be varied, it is not able to provide a meaningful macroeconomic stabilisation function by responding to cyclical variations in regional circumstances. Member State policy, by contrast, manifestly does. Nor does Community policy have any direct role in redistribution where, again, the Member State is the key actor, although once again the indirect effects of Community policy may have an influence.

Where policies only affect cohesion indirectly, the primary aim of the policy will be to achieve some other target. Yet the potential of public spending programmes to shape regional economic performance is clear, especially, when the policy in question has substantial budgets. It is important to note, moreover, that certain significant Member State policies may have an adverse effect on cohesion. Examples include major infrastructure developments that either directly benefit more prosperous regions or ease market access to less prosperous regions in a way that undermines the competitive position of indigenous firms; and some facets of defence spending, especially procurement of high technology products and of research skills.

2.2 Policy instruments and their impact on cohesion The mechanisms by which different policy instruments impinge on cohesion will inevitably shape the assessment of their impact. In some cases, the mechanisms are self-evident, while for others they are more subtle. Here we review them and a summary of expected effects is presented in table 2.1.

Macroeconomic conditions play a substantial part in determining the underlying growth rate, the employment rate and the extent of unemployment in the Member State economy. The impact of macroeconomic policy, generally, and specifically of changes wrought by EMU on cohesion will come through a number of channels. These include the interest rate and the exchange rate, the scale of public expenditure, and stability in the trajectory of output and inflation. Essentially, the impact of macroeconomic policy will be ‘top-down’ one, such that the effects on individual regions will depend on the overall stance and effectiveness of national policies.

There may, however, be asymmetric impacts on different regions that are shaped by the indirect spatial incidence of major public expenditure programmes. Broadly, such programmes will grow as the economy grows, so that if there is a systematic bias in favour of, or against, certain regions, an effect on cohesion will follow. Transport spending, for instance, will often be geared towards national networks in the interests of national efficiency with the result that its primary impact is on the most congested (typically also the most prosperous) regions. But even if network investments improve the accessibility of more peripheral regions, the effects can be two-edged if the outcome is to diminish a barrier that protects local economic activity. The resulting increase in competition may well lead to long-term gains by accelerating structural change, but could have damaging consequences in the short-term.4

Explicit territorial policies (regional policy, rural development, urban policy and local economic development) encompass a variety of measures that target the supply-side of the economy. There have been changing fashions within these policies, but they can be summed-up as consisting of a number of broad categories:

4 This is reminiscent of the well knowing saying that the operation was a success, but the patient died


28

Table 2.1 Impacts on cohesion of different policy areas Policy area Sub-area Instruments or

mechanisms Direct effects Indirect effects Expected impact

on cohesion Scale of impact on cohesion or

disparities Policy regime (especially EMU)

Common monetary policy; single market

Stabilises economic environment

Induces change in behaviour, e.g. in labour market

Positive for inflation-prone and previously volatile economies

Potentially large, but uncertain

National public consumption; automatic stabilisers

Maintains public consumption; can sustain level of activity

Concerns about dependency; better market access, can pose competitive threats

Positive in the short-term, but may have negative dynamic effect

Large, but can be two-edged

Macroeconomic policy

Public expenditure

Infrastructure & other public investment

Improves services; helps competitiveness

Eases market access for local firms, but also for competitors

Ambivalent: key is balance between better access and induced entry

Unpredictable

Public transfers Inter-governmental grants; social transfers

Large equalising effect on household current incomes

Depends on balance between stability and dependency

Positive in short-term, less certain longer-term

Large

Regional Subsidies; infrastructure enhancement; spatial planning

Competitiveness Long-run economic development

Positive, but depends on scale

Not great, except in rare cases; effects only visible long-run

Urban or rural

Property initiatives; city marketing; economic diversification

Accelerates change in economy; higher investment

Often helpful for localities, but may have perverse inter-regional effects

Small

Territorial policies

Local development

Support for clustering; capacity building; social cohesion; governance

More competitive local firms

Shapes structure of local economy; can assist development or regeneration of recipient areas

Depends on activism of local authorities; risk of reinforcing advantage of richest

Minor to negligible

Technology, R&D & innovation

Support for new technologies; innovation strategies

Stimulates new industries; product & process innovation

Helps to transform and modernise economic structure

Negative because most effort goes to few core regions

Small polarising effect

FDI Incentives and support packages to attract footloose investment

Adds to the export-base of host regions; net new jobs

Technology transfer; other externalities for host economy

Potentially positive, but risk of regional competition

Depends on balance between national and regional priorities

Regulatory Functioning of labour market; bargaining systems

Better matching; less inflation-prone bargaining system

Potentially strong for national economy

Labour market & employment

Employment Raising employability & adaptability; training

Less unemployment; employment rate rises

Higher labour productivity; more attractive to investors. Supply-side changes can foster more demand for labour

Can improve growth trajectory; potential for ‘catch-up’; but national rather than regional emphasis Moderate gains

for regions with unemployment or lack of jobs

State aids NOTE: Can be found in many of the foregoing categories

Broad range of subventions for companies: sectoral, ad hoc, horizontal

Retention of jobs and economic activity

Risk of slowing restructuring; can be anti-competitive; has opportunity cost

Often has perverse effect, especially between countries; unintended inter-regional impacts

Linked to ability to pay; incidence of costs not always evident


29

Subsidies to support employment or the operation of firms

Packages of incentives and support to attract inward investment

Measures to stimulate the creation and growth of SMEs

Initiatives to enhance the supply of industrial and commercial property

Marketing of the locality and improvements in the physical, social and cultural environment aimed at making it a more attractive location

Publicly mediated transfers can be very substantial depending on the nature of the tax and expenditure system in the Member State and the sensitivity of the regional economy to public finances. Where the share of the public sector in the national economy is high, the potential impact of public transfers on regional income will, ex-ante, be relatively higher, especially if there is a sizeable gulf between the average GDP of the most and least prosperous regions In extreme cases, the net fiscal transfer can reach very high levels, comfortably exceeding 10% of regional income. The impact on cohesion will depend above all on the elasticity of the net transfer to GDP changes, shown to be substantial in empirical studies.

State aids continue to account for large amounts of public spending in some countries. However, whether or not they advance cohesion depends in large part on their sectoral and spatial distribution – as a first order effect – and, second, on the degree to which they distort the market and whether such distortions work in favour of or against less-favoured regions. In addition, state aids, by definition, are public expenditure, so that the opportunity cost of taxes raised needs to be taken into account. The incidence of the Common Agricultural Policy is also felt through similar channels.

Inward investment is attracted by a mix of cost factors, supply of inputs, incentives and regulatory or fiscal conditions. The immediate incidence of FDI on economic activity is to create net new economic activity in the country as a whole and in the region where it is located, especially where the investment is targeted at the export base and does not directly compete with indigenous companies. But FDI has potentially important externalities. On the one hand, it can be a means of achieving technology transfer towards the host region and be a source of upgrading by introducing new management and organisation, as well as offering training. On the other hand, inward investment, depending on the nature of the work offered, may engender a degree of deskilling and do little to embed itself in the local productive structure, or might bid up factor prices by offering prime wages or terms for sub-contractors. The balance of these effects is an empirical matter.

Labour market policies that increase the flexibility and responsiveness of the labour market will influence regional competitiveness, but also bear on social cohesion. Unit labour costs as well as wage rates will matter, pointing to a link to productivity. Factors that bear on these variables include: the policy influences on the generation of skill or, more generally, human capital; the character of labour market regulation, much of which is differentiated nationally, but with the possibility there will be region-specific differences.

Employment policies that improve the attributes of the labour force can contribute to cohesion in two distinct ways. First, by facilitating adaptability and entrepreneurship, and helping to make individuals more employable, such policies give a boost to the competitiveness of the region. However, whether this Member State policy overall contributes to cohesion by improving the relative position of less-favoured regions will depend on whether the policy is concentrated on areas of high unemployment or below average employment rates. The second means by which employment policy advances cohesion is though its incidence on social cohesion.

Technology, R&D and innovation policies are, manifestly, important in fostering ‘new’ industries, stimulating productivity gains and helping to develop the knowledge-intensive economy. However, it is clear that only few regions have attained strong competitive positions in such industries and that support for national excellence can be at odds with


30

cohesion aims. There can, too, be a tension between attempts to reinforce national competitive advantage (whether through clusters or other means) and the desire to spread the benefits of high-technology.

2.3 Convergence, technology policy and cohesion Given the range of dimensions of cohesion discussed above, it is to be expected that different sorts of policies will vary in how they affect different aspects of cohesion. Most economic development policies focus mainly on economic cohesion, and the underlying rationale for them is that policy can help to raise the growth rate. Here we review specifically how the relationship between technology policy and growth can be evaluated.

Technology policies are very difficult to measure quantitatively, and therefore their analysis has to rely on a set of technology indicators that approximate different phases of these policies, assuming that they follow a certain input-output sequence. It has to be recognized that this approach to measurement of technology variables will not necessarily capture broader concepts of technology and innovation that shape company performance. The latter include forms of corporate organization and strategic decision-making by companies. For examples, there is evidence that multi-location companies may concentrate research facilities in a few specific locations, but then use the fruits of that research effort to disseminate innovative approaches much more widely across the company, or indeed among partners.

To assess the impact of technology policies standard measures of economic convergence were applied to a panel of 205 European regions at NUTSII level for the period 1989-2000 and for three technology indicators: total R&D spending, patent applications, and government R&D spending. The purpose is to describe and compare the evolution of a measure of technology input (total R&D spending), a measure of technology output (total patent applications), and a measure of public technology policy (government R&D spending). Following common practice, total R&D expenditures by all sectors in % of GDP (TERD) is used as the technology input indicator. The idea that total expenditures in R&D is a good indicator of technological innovation derives from the so-called linear model of innovation5, which assumes that investment in basic research is strongly positively correlated with technological innovation in the market place. Independently of whether this assumption holds or not, this indicator provides a robust measure of resources allocated to R&D in a particular region.

An indicator of technology output commonly used is the number of patent applications per million people. This so-called inventiveness coefficient ‘should be interpreted with care for the Southern European regions, since they are - for cultural, technical and administrative reasons- much less inclined to file patents for innovative products of processes, the index tends to underestimate their real technology investments’ (European Commission, 1997: 349). With this caveat, this is the best indicator to give an idea of the technology output intensity in a particular region.6 To appraise the impact of publicly funded support, government R&D expenditures (GERD), which is in itself a portion of the more general total R&D spending by all sectors7, is analysed separately. In chapter 8, trends in the indicators are examined, and a systematic convergence analysis of economic and technological convergence is conducted.

5 For a summary, see Soete and Arundel (1993). 6 Data for all technology and economic indicators used in this analysis comes from the New Chronos database of the European Commission. For R&D expenditures at regional level there are many gaps in the series for Belgium, Ireland, the Netherlands, Sweden and the UK between 1989 and 1994. Where possible gaps in the patents and R&D data have been filled by means of simple estimation techniques. In the case of the Innovation Index, data are only available for year 2002 and have been obtained from the 2nd European Scoreboard on Innovation (2002). 7 The other sectors being private R&D spending and R&D spending by higher education institutions.


31

Although there is continuing debate about the merits of different convergence measures8, the two most popular ones are beta-convergence and sigma-convergence. The former implies that the poor countries (regions) grow faster than richer ones and it is generally tested by regressing the growth in per capita GDP on its initial level for a given cross-section of countries (regions). In turn, this beta-convergence covers two types of convergence: absolute and conditional (on a factor or a set of factors in addition to the initial level of per capita GDP). Sigma-convergence looks at the reduction of per capita GDP dispersion within a sample of countries (regions) (see Barro and Sala-i-Martin (1995:11) for further details).

The simplest indicator is the absolute beta-convergence index using the well-known 'Barro type regression':

( ) ( ) ( ) ittfiffitfifit TITITI εβα ++=− −− 1,1, lnlnln (1)

where:

TIft: is the Technology Indicator (patents, R&D, etc.) in year t.

i: 205 regions of the EU at the NUTS II level of disaggregation for regional convergence

t: all the years in the period 1989-2000

αfi: country dummy.

βf: coefficient reflecting the existence and the speed of convergence.

According to this equation, if the coefficient β takes a negative and significant value, there has been a convergence process in this technology indicator. There would be absolute convergence in two cases. The first is if the GLS estimator is unbiased and hence we do not include any other variable apart from the previous year’s value as an explanatory variable for the change of rate. The second if only the within estimator is unbiased, but we can not reject the hypothesis of country dummies being equal for all the countries (De la Fuente, 2000). In this case all the regions will converge to the same steady state.

Because the existence of beta-convergence is a necessary but not sufficient condition for convergence (Barro and Sala-i-Martin, 1992), the standard deviation of the logarithm of each technology indicator was also computed. This sigma-convergence explores whether the dispersion among the different measures of technology inputs or outputs among European regions has been reduced. Computing the standard deviation of the different indicators gives the possibility of comparing the dispersion effect of the different policy initiatives.

The beta-convergence analysis shows that there was convergence in both the R&D and patent indicators during the 1990s, but that the main impetus came from public R&D which converged most. Sigma-convergence confirms that patent and public R&D convergence occurred, but finds that overall R&D spending did not. These two results are not incompatible, because the existence of beta–convergence is a necessary, but not a sufficient, condition for sigma-convergence (De la Fuente, 2000). Random shocks may have increased temporarily the dispersion of total R&D expenditures even in the presence of beta-convergence or regions may be approaching their steady state shares (conditional convergence) with higher dispersion than at the beginning of the period. The results are elaborated in Chapter 8.

8 For references on this debate, see Baumol, Nelson and Wolff (1994); Barro and Sala-i-Martin (1995); Quah (1993, 1996); and Boyle and McCarthy (1997, 1999).


32

2.4 Evidence on Member State policies A perennial difficulty in appraising policies that are intended to affect regional disparities is that it is often hard to distinguish between policy instruments that have explicit regional aims and others. Most public policies have some spatial dimension, even if they do not explicitly have a territorial cohesion objective. Thus, sectoral policies that support a particular industrial sector will often indirectly have a spatial dimension. Measures that favour agriculture clearly help rural areas, support for fishing favours coastal areas, while spending on education or research has impacts that are concentrated in urban areas. More specific measures (such as fiscal exemptions, subsidies or public investments) that contribute explicitly to the development of less favoured regions will be considered as narrow regional policy.

However, many other factors can influence regional development, often to greater effect than narrow regional policies. Examples are targeted aids or sector specific support. Analysis of policies within the EU Member States is influenced by specific national contexts – see table 2.2. Economic development is not uniform and the problems that are tackled differ from one country to another. In ‘cohesion countries’, the main concern is countering backwardness vis-à-vis the richer Northern European countries. In Germany or Italy, regional policies attempt to lessen regional disparities between the West and the East in the first case, and the North and the South in the second case. France and the United Kingdom face high and persistent unemployment and exclusion problems that have arisen from industrial restructuring in some regions. Finally, the Nordic countries focus on policies aimed at populations concentrated in a small number of areas for geographic and climactic reasons (Hallet, 1997). With such diverse aims, the efficiency and effectiveness of regional policies cannot easily be evaluated in a uniform way. A broad overview is, nevertheless provided in this section

Sweden Compared to other countries, public intervention in Sweden plays a particularly active role in economic development. Based on the principle of wealth redistribution more than a market-based allocation of resources, the Scandinavian welfare state, endowed with a large public sector plays an important part in cohesion policy in Sweden. Government action is guided by the idea that the reduction of regional disparities will be beneficial for growth but its role has weakened in the last years because of extensive privatisations. Hence, public spending should minimise regional disparities and thwart the negative effects of market forces.

The low density of population is an obvious feature of Sweden. Outside urban areas, the country has relatively few inhabitants. In peripheral areas, the public sector is the leading employer and the main form of economic activity. Hence, it contributed to a relatively well-balanced development path. Public transfers play an important role, especially in lagging Northern regions, in the reduction of spatial differences, the implementation of new activities in peripheral regions and in countering trends towards migration to urban areas.

Consequently, regional disparities are quite weak. GDP per head in Northern regions is relatively high, thanks to the importance of the public sector and natural-resources-based industries, such as mining, wood industries and hydroelectric plants. Differences with the South, which has a higher density of population, are relatively weak.

Recently, the main objective of regional policy has shifted to give less weight to cohesion and to focus more on achieiving higher national growth. The emergence of high unemployment, industrial restructuring and European integration explain this shift in national public policies, which have diminished in value. Sectoral aids now concentrates on the private sector, especially in reinforcing clusters and growth poles. They are the same, whatever the region, but areas which are not targeted by other regional policies or Structural Funds, are favoured.


33

Table 2.2 : Member State characteristics and policies

Geographical position Institutions European Policy Cohesion

Belgium Centre Transition to federal state now complete

Intermediate Structural Fund receipts

Predominance of the capital region

Denmark North Strong role for local level

Very low Structural Fund receipts

Limited economic disparities

Germany Centre Federal State

Substantial and concentrated Structural Fund receipts

Strong gap between the West and the East since reunification

Greece South Central state Eligible for Cohesion Fund,

Isolated from the rest of Europe Importance of agriculture

Spain South Decentralized State

Eligible for Cohesion Fund

Increasing disparities Importance of the agriculture sector in some regions

France Centre Central State Below average Structural Fund receipts

Dominance of the capital region Some regions with old declining industries

Ireland North Very centralised Eligible for Cohesion Fund Dominance of capital region;

Italy Centre South

Progressively becoming more decentralised

Substantial and concentrated Structural Fund receipts

Strong and increasing disparities Strong gap between the North and the South

Netherlands Centre Central but with growing role for local level

Low Structural Fund receipts

Strong central regions, problems in periphery

Austria Centre East Federal State Low Structural

Fund receipts Dominance of the capital region

Portugal South Central but with growing role for local level

Eligible for Cohesion Fund

Predominance of the capital region Importance of agriculture in several regions

Finland North Strong role for local level


Sparsely populated and remote regions

Sweden North Strong role for local level

Low Structural Fund receipts

Extensive welfare state limits disparities Sparsely populated and remote regions

United Kingdom Centre North

Centralised, but progressive devolution being enacted


North-South problem Dominant South East Some regions with old declining industries

Finland Finland is the Nordic country in which regional disparities are most pronounced, with the Northern and Eastern regions least favoured. In these regions, there is still a high dependence on natural resource based activities (wood, agriculture), but substantial restructuring of industry has been taking place. Traditionally, Finland has been characterised by a strong specialisation in wood industries, although, some large firms have recently emerged in ICT industries, notably the Nokia phenomenon. In the South, where more people live, economic difficulties arose from industrial restructuring engendering a high and persistent level of unemployment.

After a period of strong regional growth and convergence in the 1960’s and 1970’s, a shift occurred in the beginning of the eighties, when Finnish regions started to diverge (Pekkala, 1999). Regional policies evolved in response to the steep recession in the beginning of the 1990’s. The objective of balanced development faded away and the necessity of reaching


34

positive national growth and fighting against unemployment became paramount. As in other Nordic countries, the public sector supported employment in the least developed regions. However, as the population was ageing and unemployment persisted, public job creation was stopped and the national authorities concentrated their action on supporting the private sector, especially small and medium-sized enterprises. Services and ITC have become the main job creators. Finally, migration flows accelerated. They have become an important factor in cohesion policies and have revived the debate on regional disparities in recent years. European initiatives are today central to Finnish regional policy. National support is linked to the Structural Funds.

Denmark Regional disparities essentially stem from the differences between the cities and rural areas: on one hand, an urban centre around the capital, Copenhagen; on the other hand, a less industrialised and more rural periphery. The first legal framework in terms of regional development was implemented in the late 1950s to confront high unemployment in rural areas, as agriculture became increasingly mechanised. Regional disparities diminished in the 1970s: first, industries in the centre were then declining slightly and had to face competition from new industries located in formerly less favoured regions; second, the expansion of the public sector as the welfare state developed helped to balance the distribution of wealth across regions (Nordregio, 2002).

In the beginning of the 1990’s, public intervention to promote regional development came to an end. Only a few programmes targeted at less developed areas and national co-financing for the small number of Structural Fund projects remained. National growth became the government’s main objective as unemployment rose in the whole country and large public deficits emerged. Market forces and private initiatives were given priority. Public support concentrated on private enterprises’ competitiveness, while innovation policies operated mainly to the advantage of developed regions.

Recently, the regional question has reappeared in public debate. Indeed, development operates in a selective way, prompting public intervention in support of activities in the least developed regions. Cohesion has become a main objective of the government. New industrial development strategies were also implemented at the regional level, involving the private sector and all the levels of public institutions. In 2001, regional policy consisted mainly of three elements: support for the private sector; to a lesser extent, support for areas in great difficulty; and co-financing of European programmes (ERDF and ESF). Halkier (2001) makes the case that Denmark is a radical, if overlooked, example of the general Western European trend away from national top-down schemes towards a situation where regional policy is a multi-level activity in which European and regional actors are heavily involved.

Germany Funke & Strulik (1999) have shown that regional disparities in West Germany narrowed between 1970 and the mid-eighties, but experienced a trend reversal in the late 1980s and that strong differences persist. The poorest Länder have disappointing growth performances despite the implementation of economic policies aiming at reducing the gap and the fact that, within Germany’s federal structure large inter-regional transfers regions play a substantial role in assuring ‘automatic’ stabilisation. Regional transfers concentrated in the last decade on the catching-up of the Eastern new Länder. Around DM 1700 billion of public transfers went from the West to the East.

With the aim of reaching western living standards, three main objectives were adopted when reunification occurred: the implementation of an efficient public administration, the modernisation of public infrastructure and the creation of a legal social security regime. These objectives were quickly and successfully reached (Lallement, 2002). Similarly, some elements of policy towards regions have been, in some respects, successful, especially as regards transport and communication networks. The integration of Eastern people in the


35

social system has also been quick and should be recognised as well. Finally, innovation policy led to the establishment of high quality scientific infrastructure in the new Länder.

Results in terms of catching-up are not encouraging. Although productivity gains were quicker than in other transition economies, convergence seems to have stopped since the mid-1990s. Despite massive investment, partly associated with privatisation at the beginning of the 1990s, growth has remained slow and differences compared with the western part of the country remain large. The ongoing convergence process should rely on the production of higher quality and more technology and skilled-intensive goods (Quehenberger, 2000).

Italy Differences between the North and the South decreased markedly in Italy between 1960 and 1996. Per capita GDP in Southern regions was 58% of the Northern ones at the end of the period, but only 47% at the beginning of the 1960s. However, disparities remain high in spite of massive public investments and large transfers from the North to the South. Faster convergence occurred in the sixties. Since 1971, periods of divergence and convergence have alternated, but divergence has dominated since the mid-80s: during this period Southern per capita GDP relative to the North declined from 62% to 58% in 1996. Moreover, the unemployment rate still remains close to 20% while Northern Italian regions experienced near-full employment. There are also large disparities in growth rates within the Mezzogiorno: Abruzzo is the region which benefited the most from regional development programmes, whereas Calabria and Sicilia have faced great difficulties.

The North-South gap has characterised Italy since its unification during the 19th century and in the early 1950s the Mezzogiorno was the poorest region in Western Europe with a low level of industrialisation and a high reliance on agriculture (Helg et al., 2000). Concerted efforts to promote its development began in the 1950s, initially through investment in infrastructure and then via an industrialisation strategy, based on private investment incentives and investment by public enterprises. Publicly-owned plants were created in the chemicals, metals and steel industries. The catching-up process was stimulated by large investments and significant cost-advantages. The labour market experienced the joint effect of increased labour force participation and large-scale emigration, especially to the North of Italy.

After the oil shocks and recession in the 1970s, support for investment dried-up and, at the same time, the Mezzogiorno was particularly hit by increasing energy costs. The public sector had to bail-out collapsing private industries in the energy sector. Falling emigration coupled with rising labour force participation saw the unemployment rate rise rapidly to peak at 20% of the active population. Dependence on the public sector and the ensuing collusion between the public enterprises and political power, together with escalating corruption, largely explain the stagnation of Southern regions and increasing disparities relative to the North. After 1992, the EMU convergence requirements saw a marked shift in budgetary policy, mass privatisation of the public sector and a profound recasting of regional policy instruments. Investment in the South is still low and jobs continue to be lost.

Quarella and Tullio (2001) provide two main explanations for the persistence of North-South differences in Italy: counterproductive economic policies and the growing influence of organized crime and corruption. The positive effects of economic policies implemented in the sixties vanished and became negative in the 1980’s and the 1990’s. They stress the view that national authorities should ‘concentrate on speeding up the cases pending in front of the courts in the South, increase the resources of the judicial system and the police there, fight more convincingly corruption and organized crime in the area and pass special laws limited to the most infiltrated areas’. Finally, it is suggested that investment should be postponed until links between political power and public administration on one hand, and organised crime networks on the other hand have been broken.


36

France France is characterized by a marked gap between Ile-de-France - the largest and most advanced region - and the rest of the country where disparities are relatively narrow. Regional policies are mainly a response to this gulf between between Paris and la ‘Province’. At the beginning of the 1950’s, the French government decided to set up a new active and ambitious policy in order to counter these inequalities. This policy introduced a new approach to public administration sector: curbing Parisian expansion and industrialising regions south-west of a line from Le Havre (North) to Marseille (South) while also seeking to stop migration from the regions to Paris and developing new and large agglomerations.

A first set of reforms was institutional, with the creation of bodies to plan development. notably the ‘Commissariat Général du Plan’ ‘DATAR’ (Délégation à l’Aménagement du Territoire et à l’Action Régionale). The 22 NUTS-II Regions were established in 1964 and the first contracts between the State and some local authorities were adopted. A programme of investments in infrastructure was also initiated, focusing on ‘grands projets’, new towns, and transport and communication networks. In the manufacturing sector, policy encouraged the relocation of head offices from Paris to the regions and promoted industrial concentration. At the territorial level, a principle of ‘positive discrimination’ was adopted to boost weaker areas. This first period of territorial planning proved successful in the 1970s: the growth of the Parisian agglomeration slowed, the regions of the West achieved industrialisation and the decline of rural economies was alleviated.

After this interventionist period, a new market oriented and less ‘state-based’ policy was implemented (1975-1995), in a context of internationalisation and the furthering of European market unification, and state intervention and territorial public planning came to be regarded as inefficient. Moreover, the long French tradition of central policy eroded: central planning by the state was stopped and decentralisation was boosted. As a result, the regional policy philosophy moved from an industrial logic to an employment targeting logic. Efforts were concentrated on areas in difficulty, but public intervention was no longer targeted at reducing territorial inequalities. The objective became the promotion of indigenous potential based on the comparative advantages of each region. Regional policy now had three main dimensions: to identify the most pertinent levels for public action (towns, villages…) and to use zoning; to support initiatives in the field of transport, R&D and territorial policy; and to rethink inter-governmental links in a context of decentralisation and European structural funds, and the adoption of planning contracts between the State and the Regions (CPER, Contrat de Plan Etat-Région).

Available data and different studies show that regional disparities have widened since the beginning of the 1980’s. For example, Davezies (2001) provides a measure of sigma-convergence applied to regional GDP per head. The coefficient of variation has an upward parallel trend with the weight of Ile-de-France in total French production. Fayolle & Lécuyer (2000) compute a beta-convergence equation over the period 1986-1996. France seems to conform to the convergence equation. But when the overseas departments and Corsica are left out of the calculation, the French pattern looks like a special case: virtually all regions exhibit a divergence from the EU mean, even those initially lagging behind. In a similar vein, Carluer and Gaulier (2001) show that in terms of productivity, differences across regions narrowed in the two last decades but strong disparities remain. At an aggregate level, convergence can be explained in a large part by the shift of agricultural employment towards manufacturing and tertiary sectors.

United Kingdom Since the early 1990s the UK has seen significant changes in the design and implementation of its policies towards regional economic development. With the establishment of integrated government offices and, subsequently, the creation of devolved administrations in Scotland and Wales, the degree of institutional reform has been considerable. These devolved bodies have brought together a range of economic development functions from different central


37

government functions into a single office. In addition, in England, regional development agencies have also been set up as a means of channelling the delivery of policy support, using both common and customised policy instruments.

In the 1980s, policy focused more on macro-economic stabilisation policies to remove obstacles to the free operation of market forces, like rigidities in the labour market and inefficiencies in traditional sectors of production. This focus meant territorial policies in general and regional policies in particular assumed a secondary role and declined substantially in extent and magnitude.

Since the late 1990s, there have been various attempts to re-organise the targeting and delivery of territorial policies and a stronger commitment to foster a more balanced economic development across the UK. The current strategy has a strong focus on regionalisation, including a target of achieving real convergence in the trend growth rates of the regions by 2006. Yet it is also evident that the strategy has been ‘leaning-against the wind’ given the continuing stront gravitational pull of the South-East of England.

Spain In the latter half of the 1970s, Spain experienced the twin effects of transition to democracy and the shift towards a decentralized state. In the decade to 1985, the Spanish economy lost ground relatively, with per capita GDP falling from 79% of the EU average in 1975 to 69% in 1985. A prior decline of regional disparities also came to an end. Subsequently, after accession to the EU, the Spanish economy experienced a period of catching-up, which was especially rapid between 1985 and 1991 and growth has alos been above the EU average in recent years.

The industrial regions and tertiary-intensive regions had the highest growth rates, while for many, tourism was also an important factor. Regions traditionally lagging behind mainly in the South and West experienced weak economic performance and unemployment rose to high levels. Thus, Asturias, facing acute industrial restructuring and the very agricultural Galicia, had the lowest growth rates, whereas the more dynamic regions in the centre such as Madrid and in the North-East of the country have the best economic performance. Hence, regional disparities that were in the past contra-cyclical have become pro-cyclical as the poorest regions have the lowest growth rates.

Endogenous growth models help to explain the persistence of regional differences: for example, the concentration of R&D activities in some growth poles and external economies. Inward FDI and public investments also influence the reversal of regional trends. Another explanation as regards structural change can be given: convergence in the post-war period depends on the homogenisation of productive structures. Between 1950 and 1970, important migration flows occurred from agricultural regions to more dynamic areas, which implied large transfers of workers from agriculture to manufacturing and services. These migration flows have contributed to increase productivity in agricultural regions and favoured income convergence. This process of industrial adjustment stopped at the end of the 1970’s when migration flows slowed. Then, the labour force has shifted instead from the manufacturing sectors to the services where productivity is weaker (Rodriguez-Pose, 2000).

Hence, there is the paradox that the slowing down of the convergence process among Spanish regions occurred while resources directed to the reduction of regional differences grew. In the two last decades, regional policies aimed at developing economic activities and reinforcing economic and social cohesion were implemented at different levels: regional, national and supra-national.

Political transition in Spain and EU accession lead to the birth of two new actors of regional and economic policies: the regional government and European authorities gained competencies as regards regional economic development. Before 1982, regional development was a competence of the Spanish State. As a consequence of the creation of new decision levels, the institutional framework for regional policy became more complex. Regional


38

governments are now the core of the system. Each autonomous community has its own structures and approach, departments were created in order to promote the development and the attractiveness of regions, to provide financial support to SME’s and agencies are charged with enhancing regional competitiveness.

The national government and the EU also have competencies as regards regional policy. The State intervenes mainly through public investment and an inter-territorial compensation Fund which has the constitutional aim of reducing regional disparities. Since 1986, EU policies have also become important. Structural Funds co-finance regional development programmes and European initiatives are implemented in order to enhance cross-border cooperation and the restructuring of agricultural regions.

Regional support programmes aimed at the development of economic activities (horizontal development) often work in parallel with industrial policies (vertical development). The complexity of the institutional framework and the multiplication of policy instruments can often lead to jurisdictional conflicts between the different actors. Some measures are detrimental to the efficiency of other measures, so that it is rather difficult to evaluate particular development programmes or industrial support policies.

Portugal Thanks to fast economic growth, Portugal experienced strong convergence in the first years of its membership of the EU (1986-1990) but this growth abated subsequently. Due to slow growth, per capita GDP has diverged since 2000. Portugal, when compared to other cohesion EU countries, exhibits a high employment rate and a very low productivity level.

The country is characterised by the dominance of the capital region (Lisbon). However, disparities are not that high among Portuguese regions. Convergence has been achieved, in the sense that the poorest regions have grown at a faster rate than the richest regions, although the rate of convergence has been found to be very low (Mateus, 2001). The author also points to a deceleration of convergence in the 1990s and even divergence from 1995 onwards. Between 1995 and 2000, almost all the most heavily assisted regions have seen their GDP per capita levels fall with reference both to the EU average and to the national average. Only Madeira and, marginally, Açores improved in this period. Norte and Centro are the regions that suffered the most during the last business cycle.

Portuguese regional policy is closely integrated with EU cohesion policy and the use of structural funds. The main national regional policy tool, the territorial valuation programme (Programa de Valorização Territorial), aims, in conjunction with other ministries and the local authorities, to concentrate investments in defined areas in order to lessen the propensity to invest in the coastal areas. Basically, it intends to foster regional development throughout of Portugal, trying to reduce regional asymmetries and metropolitan concentration.

Greece Greece went through a long period of macroeconomic difficulties during the 1980s and 1990s for a number of reasons. First, Greece is a small and very fragmented country and industry essentially rests on very small sized enterprises, so that the opportunities for achieving economies of scale are limited. Services activities are concentrated in the two large cities: Athens and Thessalonica which represents more than 40% of total population. Second, Greece share common characteristics with other less developed such as: a high share of agriculture and an under-developed manufacturing sector, specialised in traditional sectors and unskilled-labour-intensive (food products, textiles and clothing). Hence, Greece is doubly disadvantaged: on one hand by a relatively weakly competitive industry, on the other hand, by the competition from much lower-wage countries in its main activities.

In addition, Greece suffers from its relative geographical isolation which hampers its economic and trade structures. Remote from the rest of the EU, Greek products have a limited access to large European markets. Moreover, its borders were real trade barriers because of


39

the cold war. This implied limited access for its exports to neighbouring countries and large core European countries (Ioannides & Petrakos, 2000).

Until the mid-1970s, Greece tended to converge with the rest of the EU. Then, it diverged from 1980 to 1995, as a result of a lower growth rate before reversing the trend again by the end of the 1990s. As in Portugal, the capital region is in many ways dominant. At NUTS-III level, convergence was interrupted at the beginning of the eighties as the most advanced regions have attracted more capital-intensive activities. The establishment of new firms strongly depends on initial conditions: industrialisation level, average growth rate, density of population, infrastructures, skilled labour availability. To a greater extent than elsewhere, social cohesion in Greece has been a problem, with inequalities tending to grow in boom times and challenges from growing ethnic minority populations.

Regional policies in Greece are essentially defined by the central government, as a result of a much centralised administration. Intermediate levels of administration levels (at the regional level) appeared in the mid-eighties, but lacked real financial autonomy or decision power. Prefectures and cities have much power locally. The distribution of responsibilities across the different decision levels is not clearly defined and remains subordinated to the central State. Despite a far from perfect institutional framework, significant initiatives were taken as regards regional development. Public aids to private investments are the key instrument. They include grants, loans, subsidies, tax allowances and are granted if different eligible criteria relative to the investment project and the location area are satisfied. Generally, the less favoured regions are privileged. Public support for infrastructure investment is another regional policy instrument which remains defined by the central government.

Unsurprisingly, EU policies are crucial for regional development. Despite different objectives across the regions, they often share the same priorities: development of infrastructure, human capital and tourism. National programmes are closely tied to EU funding and generally benefit the whole economy rather than the lessening of regional inequalities.

Ireland Ireland is a special case because of its spectacular convergence relative to the EU average in the last decade. Irish per capita GDP reached around 120% of the EU average in 2001 (although it is recognised that GNP is much lower because of net remittances abroad), having been only 71% in 1990. This success story results from the conjunction of several factors: changing specialisation towards high-technology intensive activities, high levels of inward FDI, European regional policy and a successful social partnership. At the national level, there was in general a trend towards policies which favoured development rather than regionally differentiated policies. However, the current spatial pattern of economic activity is not merely the outcome of regional/industrial policy. Rather it is the product of interaction between such policies and economic factors, such as comparative advantage, market conditions and agglomeration economies. Thus the more successful Irish regional economies were the ones which grew up around the larger urban centres where clusters and therefore increasing returns are more likely to develop. This has come about in spite of the long-term aim of industrial policy to disperse the benefits of FDI throughout the country.

As a consequence, regional disparities have remained high and have increased recently, despite high national growth. The lagging region is Border, Midlands and West (BMW) and the richer region is the South and East (S&E). The BMW region remains predominantly rural. More than half of Ireland’s farms are located within the region and much of the land is defined as ‘severely or less severely handicapped’. The S&E region, despite the fact that it is predominantly urban in character, accounts for almost two-thirds of national output in agriculture, forestry and fishing, reflecting the presence of generally good farm structures.

At the NUTS 3 (Regional Assembly) level, income convergence can be observed until the late 1980’s but divergence occurred since then. There has been significant income redistribution across the regions through the social welfare system, but the dominant force in driving output


40

differences between the regions for the period 1991-1996 is productivity differentials that are not primarily determined by differences in sectoral employment shares. The presence of agglomeration economies also appears as a significant factor.

Austria Austria can be characterised as a small, open economy. Its macro-economic development is closely inter-linked with that of neighbouring countries, especially Germany, by far its biggest trading partner. Recently, several macro-economic events have affected Austria: EU accession in 1995 together with the fall of the iron curtain supported cross-border integration. The advent of EMU led to efforts in budget deficit reduction, while the effects of the single currency are more difficult to determine. However, it is the impact of those events on structural change which may be more important in explaining regional growth patterns. Disparities across regions are primarily due to structural factors, whereas the business cycles as such, as well as inflation rates, are broadly in line with that of southern Germany. In comparison to Germany, Italy and Switzerland, Austria seems to benefit from competitive advantages and a stronger exposure to CEE countries. This is reflected in GDP growth slightly above Germany’s but mostly below the euro area average.

Weak support for the convergence hypothesis can be found in Austria during the period 1960-1989 (Höfer & Wörgötter, 1997). Using Barro-type convergence equations, the cross-sectional approach shows no strong relationship between growth rates and initial income level. After a period of diminishing regional differences, the gap between the regions enlarged in the 1980’s. Per capita incomes in Austria widely differ across regions. In 1999, Vienna’s per capita GDP is well above the national average (135) while Burgenland is well below (64). In general, the western regions of Austria achieved better economic performance than their eastern and southern counterparts, with the notable exception of Vienna. Economic growth, however, is comparatively faster in the poorer provinces – with Burgenland, Lower Austria, Styria and Carinthia exhibiting convergence to national performance indicators throughout the 1980s and 1990s. Since the second half of the 1990s, Carinthia’s growth lost momentum, whereas Upper Austria and the Tyrol managed growth rates similar to those of Burgenland.

Austria lacks a clear constitutional assignment of territorial policy competences. Most high volume measures with large regional impact (as well as certain regional policy programmes) are administered by the federal government. The federal chancellery is responsible for the coordination of regional policies by EU programmes, federal ministries and provinces. In addition, it manages important schemes concerning regional innovation and regional management. Provincial governments hold the constitutional competence for spatial planning. They are also responsible for regional economic development schemes relating to the EU ‘objective’ regions. During recent years, the increasing focus on innovation led to the adoption of coherent technology development concepts in most provinces. Indeed, during the second half of the last decade, regional development schemes experienced major transformation due to two factors. First, EU accession forced the bulk of federal/provincial aid programmes to consider regional aspects, initiating a restructuring of the aid scheme system. Second, a focus on endogenous development and innovation emerged, led by the federal chancellery’s initiatives which induced a shift towards general R&D support in federal ministries and provincial governments.

2.5 Coordination between national and European policies Member State policies and Community policies often have common objectives, but employ different instruments and follow different policy philosophies. For economic development, especially, there tends only to be a limited degree of co-ordination between Community policy and the range of policies implemented by Member States. The continuing extensive use of state aids and the ambivalence about their objectives illustrate the challenges.


41

State aids and cohesion There is a large amount of literature on whether, when and how governments should aid their companies, industries or regions.9 By and large, this literature concludes that, on efficiency grounds, aid may be justified when it is intended to correct market failure. The typical reasons cited for market failure are externalities (plus public goods), economies of scale and asymmetric information. However, in markets with multiple distortions, addressing only one source of distortion by intervening in the market creates its own distortion, which may have the apparently paradoxical effect of reducing welfare. For instance, subsidies to attract companies to certain regions may also worsen environmental pollution, increase congestion or put pressure on weak transport systems. Public policy may also generate negative regional (and cross-border) spillovers. This is particularly relevant to cohesion, as aid to stimulate creation of employment in one area may simply result in no net increase in jobs but simply in a shift of economic activity from one area to another or competition between regions. This is explicitly recognised in the Commission’s new multisectoral framework for regional aid which reduces the intensity of allowable aid for large projects because large companies are more mobile and because they can exert more pressure on regional authorities.10

Although EU competition policy frowns on state aids, economic theory does not establish an over-whelming case provides no grounds for a complete banon them. Some public assistance is necessary to remedy market failure. Although a preference for horizontal aid is a move in the right direction, economic distortions will not necessarily be eliminated. In practice not all state aid is intended to remedy market failure or maximise overall or regional economic output. Nor, are conflicts between different policies totally avoided (e.g. competition v cohesion). The granting of state aid requires judgement and balancing of opposing policy aims. That is why it is important to examine the actual impact of state aid on fundamental Community objectives such as cohesion.

In a study by Midelfart-Knarvik & Overman (2002), it is argued that state aid to manufacturing sectors has no effect on the poorest regions’ catching-up and has no influence on location choice. However, the Structural Funds can influence the industrial location by attracting R&D intensive activities to regions poorly endowed in skilled labour. Although, the evidence suggests that state aids often go against national and regional comparative advantages (with the exception of Ireland), their impact on cohesion is not always fully brought out. Some restrictions on their use in richer areas may, nevertheless, be warranted since Structural Funds do not play a similar role. A guidning principle, arguably, is that national intervention should not contradict Community support.

Generally, coordination between regional, national and supra-national institutions is still far from perfect as regards the elaboration of European financed projects, with a detrimental effect on their efficiency (Fayolle & Lécuyer, 2000). The implementation of these programmes often remains too centralised at the State level and local authorities play a minor role in their management.

Inward investment The international evidence suggests that there is a positive relationship between FDI inflows and regional or national economic growth. The direction of causation is a matter of debate however.11 It is clear that FDI is attracted to more successful economies, while even in more peripheral ones a certain threshold level of local capabilities – in terms of the stock of human 9 See, for example, Bishop (1997), Collie (2000), Ehlermann (1995) Ioannis and Reiner (2001) Lehner et al. (1991) Neven (1994), Nicolaides and Bilal (1999) and Wishlade (2003) 10 See the Notice on the Multisectoral Framework on Regional Aid for Large Investment Projects, OJ C107, 7/4/1998. The Framework has been extended until 31/12/2003. A new Framework [OJ C70, 19/3/2002] will come into effect on 1/1/2004 and will be valid until 31/12/2009. 11 Blomstrom et al . (1994) claim to find a causal link, though Dutt (1997) finds no such evidence.


42

capital for example – appears to be required before the growth effects can be in any way substantial; E. Borensztein et al. (1995).

This suggests that attracting FDI to lagging regions will be difficult, that the FDI attracted will not necessarily be of the highest calibre (i.e. is likely to be located in low-skill sectors with relatively poor spillover benefits), and that other policies and factors need to be in place before the impact can be sufficient to promote convergence. Nor, in regional economies, is it likely to be sufficient to ensure an adequate educational throughput, since, with inter-regional labour mobility, the educated labour may be attracted to the core regions where FDI is likely to locate, rather than FDI being attracted to the regions where the labour is educated; Markusen (1988). Such a ‘brain drain’ from peripheral regions will strengthen divergence tendencies.

This perspective suggests then that strong national FDI inflows, unless they are disproportionately dispersed to peripheral regions, are more likely to generate divergence rather than convergence. Only if other complementary policies are in place to increase the attractiveness of peripheral regions to FDI inflows are national inflows likely to be associated with regional convergence. This conclusion indeed mirrors a view increasingly voiced about the efficacy of regional aid inflows, i.e. that they impact positively on convergence prospects only if a series of other criteria are also being met.12

Rodriguez-Pose (2000) illustrates this by comparing the fortunes of two Spanish regions, Navarre and Galicia. Convergence theory suggests that poorer regions, all else equal, should grow more rapidly than richer regions. Navarre, however, despite being richer than the Spanish average at all times since the early 1980s at least, and despite being located in the relatively declining Northern fringe of the country, has grown at a more rapid pace than the rest of Spain, while Galicia, one of the poorer regions, has suffered a relative decline over the period. He argues that Navarre has excelled particularly in terms of its ability to attract FDI, but that skills and accessibility have been more important factors behind this success than subsidies and incentives. He faults Galicia for pouring all its resources into infrastructural development while ignoring the simultaneous need to develop its human capital and adopt targeted innovation strategies.

An inference to draw is that an integrated regional approach is required for success in the convergence stakes, and for FDI inflows to make a contribution to that success. The next section details the tension that exists between the lure of pre-existing agglomerations, which promotes divergence, and the role of regional policies in attracting FDI in order to promote convergence.

Regional Locational Determinants of FDI Some broad implications for the development potential of particular regions can be drawn from the changing levels of concentration and dispersion of industries across national locations. Midelfart-Knarvik et al. (2000) find, for example, that some of the forces encouraging medium and high increasing-returns to scale industries to locate in central regions are diminishing, arguably as a result of generally declining transport costs and other improvements in local infrastructure. Many high-tech industries for which increasing-returns are less important have become less concentrated in core regions over time, being attracted to those more peripheral locations which have good infrastructure and adequate supplies of skilled labour. At the same time, industrial linkages are encouraging some industries – for

12 In the regional aid case, these other requirements include high institutional quality and low corruption; Ederveen, S., H. de Groot and R. Nahuis (2002). Barry (2003) emphasises the importance of the simultaneity of the series of benefecial shocks and policy changes that ushered in the Celtic Tiger era in the Irish case and promoted convergence to average EU living standards.


43

example those with high shares of intermediate goods in production – to move into central locations.

Amongst the factors generally agreed to be of importance in determining the regional distribution of FDI within nation states are (i) manufacturing agglomerations, (ii) population density, and (iii) local infrastructure. As the Midelfart-Knarvik analysis suggests however, the importance of these various factors differs according to industry-type. Head, Ries and Ruckman (1998) explore the question of why Japanese investors in the US tend to locate affiliates near concentrations of US and Japanese establishments in their own industry. They find that agglomeration forces are stronger in natural resource industries and in industries that use their own sector’s output intensively. They also find that Japanese manufacturers with high transport costs display greater tendencies to cluster, perhaps around geographically concentrated downstream purchasers of their products.

Much of the literature on the regional determinants of FDI is concerned with exploring these types of effects as well as those of regionally-differentiated tax holidays and regional grants payments. In a study of the regional location of French MNCs, Ferrer (1998) finds that French firms locate in the most industrialised regions of a given country and in the regions where similar activities are already concentrated, with inter-industrial and intra-industrial economies both playing a major role in locational choice. Proximity of major airports also displays great significance, while, ceteris paribus, lower wage locations tend to be favoured. Regional financial assistance is not found to be an important determinant of the regional location of these French firms, in that it cannot counterbalance (or is generally too small to counterbalance) weak regional comparative advantages. More pro-active regional policies directed towards innovation, however, proxied in Ferrer (1998) by the number of research centres, universities, chambers of trade and industry and associative networks, do have positive effects however.

These findings on the low level of importance of regional financial assistance are in contrast to those of Meyler and Strobl (2000) who show that over the period 1973-82 when Irish industrial policy explicitly encouraged job generation in certain designated areas (via, amongst other things, preferential grant treatment, job targets and the building of advance factories), an annual 27 per cent of the job generation in the designated areas could be attributed to the explicit regional industrial policy.13 More recently Barrios, Gorg and Strobl (2003) have shown that since 1982 when regional dispersion became of secondary importance to the goal of attracting high-tech foreign firms to Ireland, regional policy was effective only in attracting low-tech firms to the disadvantaged areas. Urbanisation economies were for high-tech firms a more important locational determinant than public incentives.

One possible difference between the results for French MNCs and for MNCs in Ireland is that Ireland is such a geographically small unit that poor infrastructure in the lagging regions may have had less important effects on regional development there. This accords with the point made by Guimaraes, Rolfe and Woodward (1998) concerning the emerging consensus that tax-holiday effects (and by extension, those of regionally-differentiated grants) are stronger when competition consists of small geographical units sharing many other similar characteristics.14

Hill and Munday (1992) present evidence which throws further light on this issue. They find that FDI inflows to the UK over the course of the 1980s showed an increasing tendency to locate between the core and peripheral regions of the UK, with Wales, the West Midlands and 13 Guimaraes, Rolfe and Woodward (1998) in a simulation exercise show that regionally-differentiated tax holidays in Puerto Rico generated shift of about 7 percent of manufacturing plant openings from the island’s capital. 14 Another possibility is that the French MNCs analysed by Ferrer (1998) may be more high-tech than the industries attracted to the more peripheral Irish regions, or simply that the Irish grants to encourge dispersion were higher than the grants being offered elsewhere.


44

the North attracting shares of new investment well out of proportion to their size. They find that infrastructural spending and regional preferential assistance combined were important in influencing new firm location. The relatively successful regions, furthermore, represented relatively low cost locations for standardised production, with access to adjacent UK and EU markets.

Friedman et al. (1992) analyse the location decisions of all MNCs, Japanese MNCs and European MNCs across US states. They find that there is a degree of consensus in the literature on this issue that factors such as market size and market access (which is related to transportation infrastructure), the manufacturing wage rate, and state promotional activates designed to attract FDI are all of importance. They find furthermore that state and local taxes are significant, while Coughlin and Segev (2000) add educational attainment levels to the list of significant variables. The earlier study also finds that the decision determinants for Japanese and European MNCs differ, with European MNCs less influenced by wages and local labour market conditions. They speculate that this may be due to the different industries in which the Japanese and European firms are located, with European firms more heavily concentrated in capital-intensive chemicals and Japanese firms in less capital-intensive electrical components and transport equipment.

2.6 Commentary Convergence occurs at a national level, thanks to higher growth rates in cohesion countries, but may be at the expense of intra-national inequalities that persist and in some cases enlarge. Hence, the least favoured regions come under international competition and could experience, in some areas a deindustrialisation process because of declining economic structure and/or unfavourable geography.

This overview demonstrates that regional dynamics and the design of regional policies continue to have strong national characteristics. While each region benefits from national performance, other things being equal, regional inequalities vary enormously from one country to another. However, some common features can be identified. Several countries experienced periods of regional convergence up to the end of the 1970s, but have since witnessed a deepening of internal disparities that reflect local specific factors (natural resources, geographical situation, accessibility, institutions). Hence, a paradox is often reported: disparities increase over a period when support from national and supra-national authorities has been substantial and growing. As regards the conduct of regional policies, their administration often seems disorganised: the different decision levels are imperfectly coordinated in the definition and the financing of projects, all of which hampers their efficiency.


45

3. Macroeconomic policies Macroeconomic conditions play a substantial part in ensuring that countries are able to achieve and improve their potential growth rates and have a marked impact on the employment rate and the extent of unemployment in any Member State economy and, a fortiori, its regions. The primary empirical question that has been asked in relation to cohesion is whether rich and poor regions converge. The answer has been tentatively in the affirmative. Small rates of convergence have been observed by several studies15 and are confirmed in new work done as part of the present study. But there are also very pronounced disparities that demonstrate that regions which are supported by Community policies do not uniformly benefit, with the implication that national policies make the difference. In a computation of relative performance in the last decade, of the twenty five regions that diverge most from a regression line of mean performance, nine were objective 1 regions: four comfortably out-performed the expected values, while five were well below.

The impact on cohesion of macroeconomic policy, generally, and specifically of changes brought about by EMU, will come through a number of channels. These include the interest rate and the exchange rate, the scale and mix of public expenditure, and stability in the trajectory of output and inflation. Essentially, the impact of macroeconomic policy will be ‘top-down’ to the extent that the effects on individual regions will depend on the overall stance and effectiveness of national policies. There may, however, be asymmetric impacts on different regions (see Ardy et al., 2002; Fatas, 2003)

In the chapter of the 1st Cohesion Report which dealt with national policies, great weight was attributed to the importance of macroeconomic policies in providing the platform for ‘catching-up’. From a regional perspective, stability in economic conditions, the impact of national wage-setting behaviour and the effects of macroeconomic factors on asset prices are all factors that can shape both aggregate demand for the region’s output and its input cost competitiveness. The inter-play of public expenditure and taxation, in particular, performs an important role by itself in stabilising regional demand. Regions with below average economic activity will tend to attract proportionally higher inflows of public expenditure, while dynamic regions pay more in tax revenue. The direct effect of such flows on cohesion through effects on living standards is discussed in the next section; here the focus is on the aggregate trajectory of the regional economy.

It is important in discussing the effects of macroeconomic policy to distinguish between their incidence on ‘internal’ and ‘EU’ cohesion as discussed in the preceding section. Macroeconomic policy will promote ‘internal cohesion’ if its effects are to reduce disparities within a country, and ‘EU cohesion’ if regions below the EU average converge. It is entirely possible that faster growth for the country as a whole (leading to ‘EU cohesion’) will be accompanied by widening internal disparities, even if higher growth ‘trickles down’; equally, the opposite could apply (see figure 3.1).

Figure 3.1 Macroeconomic trends and cohesion - examples EU Cohesion

Internal cohesion

Pro-cohesive outcomes Anti-cohesive outcomes

Pro-cohesive effects of policies Some Spanish regions since 1997

Some Italian and German regions since

mid-1990s Anti-cohesive effects of policies Ireland and Portugal in

the 1990s Some regions in Greece during 1980s and early

1990s

15 See Barro and Sala-i-Martin (1992); de la Fuente (2000) and references therein.


46

3.1 Macroeconomic policy and ‘EU cohesion’ Recent years have seen a substantial re-orientation of macroeconomic policies towards what might be termed stability-orientated policy. Major changes that bear on cohesion have occurred as EMU has been consolidated and reflect a process that was engaged well before the formal introduction of the euro in 1999. For example, it is generally accepted that Ireland was an early beneficiary of the shift to stability orientated macroeconomic policy that began in the late 1980s, while Greece, despite not being in the first wave of members of the euro area, has seen a progressive transformation of its macroeconomic environment since 1996. But it can be argued that the changes apparent in Danish, Swedish and UK policies follow much the same lines, implying that it is a wider paradigm shift in policy thinking.

The degree to which countries have altered their policies, and might have altered the relationship between macroeconomic policy and cohesion, depends on two sets of criteria: first, the extent of the policy shift from the previous regime; and, second, the macroeconomic impact of EMU membership, especially, on performance. For those countries that were, de facto, part of a DM zone, it can be argued that the change in policy stance has been limited. Thus, apart from Germany, for Austria, the Benelux and possibly also for France – certainly since the espousal of the franc fort policy – the character of macroeconomic policy has not changed greatly. By contrast, Italy, the Iberian countries and Greece, and possibly also Finland, have undergone quite a shift in the last decade. Sweden and the UK have made a series of reforms that have progressively led to a greater focus on stability, yet have still witnessed sizeable exchange rate movements, while Denmark and Ireland – having acted to consolidate public finances sooner than others – have been ahead of the game.

The reconfiguration of the policy framework comprises the unification of monetary policy, fiscal policy subject to the constraints of the Stability and Growth Pact (SGP), and the soft coordination of employment and social policies to stimulate reform. One classical source of asymmetric shocks, changes in exchange rates, has been removed between euro area members. The fact that, since 1999, it is the cohesion countries, collectively, that have enjoyed the most rapid economic growth suggests that macroeconomic conditions have indeed promoted ‘EU cohesion’. A number of explanations can be advanced for this observation. The first is that membership of the euro and acclimatisation to the new policy environment changes the incidence and response to country or region-specific asymmetric shocks. Second, there are likely to have been ‘windfall’ gains in credibility that ease constraints, while a third explanation is that the conjunction of policy settings – interest rate, real exchange rate and so on – was asymmetrically favourable to the cohesion countries.

However, interest rate policies can no longer be differentiated and a uniform interest rate policy might well become a source of asymmetric shocks. This could happen either because of continuing differences in inflation rates - which mean that a given nominal interest rate results in different real rates between countries - or because of underlying characteristics affecting the transmission of monetary policy, such as differences in the balance between short and long-term credit with variable or fixed rates. Already, since 1999, it can be noted that real interest rates have been low or even negative in some cohesion countries, but high in Germany.

The process of nominal convergence required for entry into stage 3 of EMU involved the stabilisation of exchange rates, the consolidation of public finances, and reduction of the rate of inflation which led to lower nominal long-term interest rates. Efforts to curb government borrowing have been a characteristic of all Member States in recent years, either (for the twelve euro area members), initially in order to meet the convergence criteria for EMU membership, then to abide by the terms of the Stability and Growth Pact, or (in the case of the other Member States) because of self-imposed rules. The shift to stability orientated policy has a number of ramifications for cohesion.

Empirically, most of the reining-in in public deficits has been accomplished by reducing public expenditure. Typically, cuts in spending fall on areas that are easiest to change such as


47

public investment or discretionary rather than statutory programmes. To the extent that public spending has been more carefully managed, it can be argued that the pressure on budgets may have led to greater efficiency. But it is also likely that the cuts in public spending will fall disproportionately on weaker regions, since they will generally – other things being equal – generate less public revenue and receive more public spending relative to indigenous income.

Total public expenditure declined between 1995 and 2002 from 53% of EU GDP to 47%. While government expenditure was reduced markedly across the EU, government revenue from taxes and other sources declined only slightly in relation to GDP, implying the broad maintenance of tax rates. Except in Ireland, where the growth of GDP was exceptionally high, in no Member State did revenue fall by more than 2% of GDP and in 6 countries, it increased. Budget deficits were, therefore, reduced throughout the Union and, in a number of cases, transformed into surpluses. In 1995, the budget was in surplus only in Luxembourg and in 5 Member States, the budget deficit was over 7% of GDP, in Germany and Spain, over 10% ; in 2002, 6 countries had a surplus and only in Germany and France was the budget deficit over 3% of GDP and then only slightly. Figure 3.2 summarises these changes.

BUDGET DEFICITS

-6-4-202468

1012

Belgium

Denmark

German

y

Greece

Spain

France

Irelan

dIta

ly

Luxe

mbourg

Netherl

ands

Austria

Portug

al

Finlan

d

Sweden UK

Euro ar

eaEU-15

% of GDP

1995 2001 2005(f )

Figure 3.2 Note: Negative values represent budget surpluses

The interest rate effect became a positive asymmetric shock for those Member States, especially the cohesion countries, which had high levels of public debt or high nominal interest rates. The impact of macroeconomics on the holding and servicing of national debt is a specific channel that has been significantly affected by recent macroeconomic developments. The fall in debt service costs, as figure 3.3 shows, was dramatic for Greece (from 11.2% in 1995 to 5.5% in 2002) and Italy (from 11.5% to 5.7%), but also marked for Spain (5.2% to 2.9%) and Portugal (6.3% to 3.0%). By contrast, German and French servicing costs fell by just 0.5 and 0.7 percentage points over the same period. Especially for the highly indebted Member States, this macroeconomic change created considerably more room for manoeuvre in fiscal policy.

Price stability has also improved in recent years as countries, initially, sought to meet the ‘Maastricht’ criteria and subsequently as the impact of the single monetary policy has made itself felt. Separately from the nominal interest rate effect just described, there is tentative evidence that lower inflation has encouraged investment in the lower income countries, again fostering cohesion. However, whether this has a further impact on poorer regions, either


48

because of so-called Balassa-Samuelson effects or by making conditions for private investors more stable can only be the subject of conjecture.

GOVERNMENT DEBT SERVICE PAYMENTS (% of GDP)

0.0

2.0

4.0

6.0

8.0

10.0

12.0

Germany France Ireland Spain Portugal Belgium Greece Italy EU-15 Euro area

1995-97 1998-2000 2001-03

Figure 3.3 The fillip to economic activity provided by euro area membership allowed expanding output where resources were unemployed and via the ‘windfall’ gain to public finances, notably in Spain, Greece and Ireland, though only to a lesser extent in Portugal. Elsewhere, demand was constrained by reductions in the government deficits and/or the expansion of surpluses. Inflation and the balance of payments are the other means of accommodating higher nominal demand. Initially inflation accelerated in some countries, although as growth slowed in 2002 inflationary pressures abated. This inflation may be regarded as part of the normal adjustment process in EMU. Fast growing, competitive countries and regions will have their performance dampened by relatively higher price inflation, while slower growing countries will see their competitiveness improve because of a lower rate of inflation.16

The eventual impact of short-term acclimatisation could translate into lasting pro-cohesive effects to the extent that there are persistence mechanisms at work that lead to virtuous circles: the experience of Ireland gives reason for optimism. While little is known about the region-specific working of these persistence mechanisms17, even the short-term effects of EMU must not be dismissed as neutral as regards cohesion.

A striking feature of public finances is the relative shifts in two key areas of expenditure. Public gross fixed capital formation has generally fallen in the EU as a whole over the last two decades, although with the apparent paradox that it is the poorer Member States that have most successfully sustained the investment effort, in contrast to the richer ones (see figure 3.4). Second, the burden of interest payments has declined steeply since the early 1990s, reflecting both the fall in the size of the public debt and the steady decline in nominal interest rates. In this respect, EMU has manifestly made it much easier for the poorer Member States to devote more national resources to strategic expenditure identified in the 1st Cohesion Report as a vital contribution to catching-up with the rest of the Union (figure 3.5).

16 See Alesina et al. (2001) 17 Discussed in Balakrishnan and Michelacci (2001); Blanchard (1998); Blanchard and Woolfers (2000); Obstfeld and Peri (1998) and Karanassou et al. (2002)


49

Public investment, 2000

0

1

2

3

4

5

6

7

Belgium

Denmark

German

y

Greece

Spain

France

Irelan

dIta

ly

Luxe

mbourg

Netherl

ands

Austria

Portug

al

Finlan

d

Sweden UK

EU-15

% of GDP

GFCF Other capital

Figure 3.4

PUBLIC INVESTMENT

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Germany France Ireland Spain Portugal Belgium Greece Italy EU-15 Euro area

% o

f GD

P

1995-97 1998-2000 2001-03

Figure 3.5

3.2 Internal cohesion Both theoretically and empirically, the impact of improved macroeconomic policy on individual regions is much less clear. Globally, as the examples cited in figure 3.1 indicate, there have been cases of all four possible combinations of outcomes for cohesion. A problem in interpreting the outcomes, however, is that they reflect not just macroeconomic policies, but also the success (or failure) of structural and other policies, whether targeted at cohesion or not. Some general propositions can, nevertheless, be advanced.

Macroeconomic stability affects growth mainly through its impact on investment. Uncertainty measured in various ways seems to reduce long run fixed capital investment but that effect


50

varies across industries and types of capital goods.18 In the EMU system, national fiscal policies have a vital role to play in stabilisation19, although their effectiveness depends on there being room for them to work by holding the medium-term budgetary position close to balance. In this regard, the sheer size of the national budget matters: the larger it is the greater its potential to respond to national asymmetric shocks.20 The operation of these mechanisms can compensate for the absence of nationally diversified asset holdings and savings of households and firms, and of cross border credit. But if public finances are under pressure, retrenchment will tend to be damaging for cohesion if an obligation to cut spending means that programmes that have a pro-cohesive stabilising impact have to be scaled-back. This macroeconomic impact has to be distinguished from the redistributive impact discussed in subsequent sections.

The Danish economy is an example of a favourable macroeconomic setting, and exhibits many of the attributes that also helped Ireland to grow so spectacularly. The upswing in the Danish economy since 1993 has several roots: it is, first, a standard example of demand-driven growth. An expansion of fiscal policy was allowed in 1993-94. Then came falling international interest rates, rising prices of houses and a credit reform allowing home-owners to convert the fall in long-term interest rates into lower housing costs. Private demand was strongly stimulated. In 1994 alone, private consumption grew by 7 per cent in real terms. Investment in housing accelerated. The same happened, after a while, with private investment in general, all according to the standard economic textbook. A further important observation is that the change from economic stagnation to economic growth and rising employment has been possible without the outburst of wage inflation, which normally follows a dramatic fall in unemployment. There are several explanations for this development:

• The introduction of a number of reforms of labour market policy from 1994 and onwards.

• Changing attitudes of trade unions paying more attention to the effects of wage increases on international competitiveness and related probably to still firmer expectations of a fixed exchange rate between the Danish currency and the Euro.

• The lowering of the international rates of inflation, which – under the regime of a fixed exchange rate – were directly transmitted into the national economy

The steady growth since 1993 seems to have benefited all regions equally (with the exception of the island of Bornholm).

Private capital markets can also help to cushion regional specific shocks. However, cross border ownership of assets in the euro area remains low and the scope for cross border borrowing is restricted by the primarily national nature of credit markets. Some of the evidence in recent studies for the Commission (Giannetti et al, 2002) suggests that further integration of EU financial markets will enhance the role of private financial flows in regional stabilisation. Although this could improve the overall adjustment capacity of regions and thus be pro-cohesive, an adverse impact on cohesion could occur for two reasons. First, there tends to be a distributional bias in private insurance mechanisms which largely favour households and firms with internationally diversified portfolios of assets and liabilities, but have little impact on low-income and wage-dependent households. A negative effect on social cohesion could, therefore, arise. Second, it is likely that there would be further agglomeration in financial services that would concentrate economic activity in the sector at the expense of financial intermediaries in poorer regions.

18 See Carruth et al (2000); and Pain (2002) 19 The impact of automatic stabilisers over the 1990s is found to be significant, dampening around 25-30% of fluctuations in major EU Member States, namely Germany, Italy and the UK - see Buti and Sapir (1998), and P. van den Noord (2000) 20 Fatás and Mihov (2001)


51

Internal cohesion is also likely to be affected by changes in the debt service resulting from macroeconomic policy shifts. One interpretation is that the interest payments on national debt should be considered separately from political choices on the distribution of public expenditure, since they are not the result of decisions on economic policies, but depend on the stock of regional wealth. They are also a first call on public resources in the sense that governments have to provide for the debt service before considering other spending programmes. To the extent that national debt is owned disproportionately by residents of wealthier regions – a reasonable conjecture – a fall in the costs of debt service should favour poorer regions. In the context, first, of the process of convergence to reach stage 3 of EMU, then the SGP (at least initially) governments have, in addition to curb aggregate spending. The usual targets for such cutbacks are social programmes (and thus social cohesion policies) and public investment (often hampering economic cohesion). It is an open question whether the freedom of manoeuvre from lower national debt interest burdens will be used to loosen the shackles on cohesion related expenditure.

Firm evidence to support the contention that debt servicing favours richer regions is scarce, but figures for Italy suggest that the magnitudes may well be significant. Italian data are sufficiently detailed to permit a regional breakdown of where interest on the national debt accrues. Almost 40% is paid to Lombardia alone (1992 euros per capita in the year 2000, more than 10 times the per capita amount paid to all Southern regions except Abruzzo, and amounting to 12% of the region’s GDP). The minimum value is in Calabria with 115 euros, only 5.8% of the amount paid per capita to Lombardia. Given the decline in debt servicing shown in figure 3.3 above, the pro-cohesive effect of EMU in this respect is huge. In the year 2000, debt service was 6.5% of Italian GDP, compared with 11.5% five years earlier. Assuming the same pattern of payments to regions, the gross transfer to Lombardia would have been around 20% of its GDP – see figure 3.6.21

Macroeconomic imperatives have also led to the imposition of rules on the fiscal positions of sub-national governments in several Member States. A good example is in Austria, where the Österreichischer Stabilitätspakt 2001 (the pact currently in force), requires provinces to attain a surplus of 0.75% of GDP (shared pro-rata to population, with the richer regions taking a slightly heavier burden) and municipalities to achieve a balanced budget – allowing the federal government to retain a low deficit in order to arrive at a balanced budget of the public sector as a whole. Provinces may undershoot their target by 0.1% of GDP (i.e. the corresponding amount for a single province) whereas municipalities may run a deficit of 0.15% of GDP in total. If the target is not met but the deficit/surplus lies within the prescribed range, the authority in question has to regain that amount in the subsequent fiscal year.22 Consequently the scope for regional and local governments to engage in deficit spending is severely restrained and it is often cohesion-related policies which are the first to be cut.

21 These inferences have to be treated with some caution because the figures have to be interpreted as spending ‘in’ the region rather than ‘for the benefit’ of the region and it is likely that the concentration of financial intermediaries located in the region significantly overstates the impact on residents of Lombardia. Nevertheless they are striking. 22 Matzinger, 2002, pp.84-86.


52

PAYMENT OF INTEREST ON ITALIAN NATIONAL DEBT, 2000

0

500

1000

1500

2000

2500

Piemon

te

Valle d

'Aosta

Lomba

rdia

Trentin

o Alto

Adige

Veneto

Friuli V

enez

ia Giul

iaLig

uria

Emilia R

omag

na

Tosca

na

Umbria

Marche

Lazio

Abruzz

oMoli

se

Campa

niaPug

lia

Basilic

ata

Calabri

aSicil

ia

Sardeg

na

Italy:

regio

nalize

d

Euros per capita

Figure 3.6

Public investment and cohesion Finally, as regards investment, Greek evidence suggests that public investment does not necessarily favour internal cohesion. Filippaios and Kottaridi (2003) show that there has been a concentration of investment in the large metropolitan cities (Athens and Thessaloniki). Interestingly, Community policy seems to have been influential via its effect on variables related with economic development, in contrast with the Greek Development Law of 1992, which had little impact. Lambrinidis et al. (2001) extend research by Christodoulakis and Segoura (1994) and conclude that in contrast to its important role in the evolution of manufacturing output at the national level, public infrastructure capital has not been a significant determinant of the differentiation of manufacturing output at the regional level.

3.3 EMU, trade and competition The medium- and longer-term effects of monetary union may also contribute to convergence, although empirical evidence on effects to date is limited and constrained by the absence of information on inter-regional trade. The principal channels are:

More intensive cross-border exchanges of goods and services

Greater labour and capital mobility

Closer correlation of fluctuations in output, employment and inflation that reduce the probability of asymmetric shocks hitting regions.

An important adjustment mechanism is the labour market and it is arguably the case that most of the lagging regions in the EU are in Member States that have relatively heavily regulated product and labour markets. Consequently, as ‘Cardiff’ and ‘Lisbon’ measures are progressively adopted, their impact on competitive structures will cumulate. It is difficult to arrive at unambiguous conclusions about the regional impact of such measures, although some indirect evidence can be adduced from what has happened in Member States that have been in the vanguard. In Denmark, the Netherlands and the UK, all of which engaged in far-reaching labour market reforms – albeit following diverse routes – the variance in regional unemployment rates is much lower than in Germany or Italy.

Petrakos and Saratsis (2000) use β- and σ-convergence methods to test for the pattern of regional inequalities in Greece at the NUTS III level. An important finding is that regional inequalities are pro-cyclical, increasing during expansions and decreasing during recessions.


53

Thus, the reduction of inequalities that occurred during the 1980s is probably due to the prolonged recession of that decade.

A study by ESRI (1996) on Economic Implications for Ireland of EMU classified industrial sectors in terms of (a) exposure to sterling fluctuations, with Ireland in EMU and the UK outside, and (b) sensitivity to high interest rates. The sectors most vulnerable on both counts tended to be the low-technology sectors consisting largely of Food, Drink and Tobacco (NACE 15-16) and Textiles, Clothing and Footwear (NACE 17-19). Their distribution, in terms of the share of regional manufacturing employment in low-tech sectors for the year 2000, was 41% of regional manufacturing employment in the BMW region and 39% in SE. One must judge then that while the lagging BMW region gained most from the lower interest rate environment represented by EMU, it has suffered most in terms of the increased exposure to sterling fluctuations that EMU membership has entailed.

3.4 Commentary and key findings For some observers, macroeconomic policies are central to cohesion. Certainly, their enabling role has to be stressed, as an absence of stability can severely undermine the prospects for cohesion. By definition, the major impact of sound macroeconomic policies is on output and employment at the level of the Member State. On the one hand, the growth rate achieved becomes the benchmark for all regions within the Member State, so that if the economy as a whole does well, it can reasonably be expected that this performance will percolate down to all regions. But there may also be a built-in asymmetry that militates against less competitive regions. In this regard, Antikainen (2001) reports the interesting finding that the recession of the early lasted three years longer in Finnish Objective 1 regions than in the larger urban regions of southern Finland. On the other hand, it tends to be easier to pursue other policy priorities when the macroeconomic setting is favourable. Structural changes can more readily be implemented when there is robust growth, but may engender resistance when incomes and jobs are under threat. Similarly, a dynamic macro-economy simultaneously faces fewer demands for solidaristic and palliative cohesion policies, yet can afford higher levels of public expenditure without risking budgetary imbalance.

From a regional perspective and looking beyond the enabling function of macroeconomic policies two further aspects of this policy area deserve to be highlighted. First, it is common to find that public investment – including on policies to promote economic cohesion - is squeezed when there are budgetary problems associated with inappropriate macroeconomic policies. Since weaker regions tend to depend more heavily on such public investment, precisely because they are less able to attract private funds, ‘EU’ economic cohesion, as defined above, may depend partly on national growth performance. But it remains an open question whether intra Member State disparities narrow or widen. Second, the impact of macroeconomic policies on the national debt has been shown in the analysis in this section to be potentially a powerful influence on income disparities and thus on social and territorial cohesion.

Key findings are: Cohesion countries have, on the whole benefited from the macroeconomic stability

afforded by EMU.

However, slow growth in the core euro area countries has dampened demand

EMU has substantially reduced the debt service burden for countries with high levels of public debt, a change that would be expected to favour poorer regions. Italian data suggest that the regional impact of this change in debt service has transferred income form richer to poorer regions, an income transfer that arguably favours social cohesion.

Public investment has fallen overall, but has been most successfully maintained in cohesion countries. An implication is that Community policies have helped in this regard.


54

4. Public expenditure and public transfers Public expenditure in EU countries is, as noted in the preceding section, substantial as a proportion of GDP, so that its distribution by region will inevitably have a very pronounced impact on relative living standards. Even relatively small variations in the scale of flows to different regions – whatever their rationale – will affect the resources at the region’s disposal. The amounts accruing to different regions depend on political decisions, constitutional obligations and entitlements that derive from the rules governing particular programmes.

As was stressed in the First Cohesion Report, public expenditure is a key factor in assuring social cohesion. Most governments commit themselves to high levels of service provision, common standards and universal access which mean that even in the weakest regional economies, the population can be assured of a standard of living that reflects social policy objectives. Public spending on health, education and basic infrastructure has the further effect of providing a level of investment in human and physical capital that, in parallel, means that the productive base of all regional economies is supported, thereby promoting economic cohesion. Especially in education, where public spending typically accounts for between 4% and 6% of Member State GDP, this spending substantially exceeds Community spending on structural operations.

Net public budgetary transfers can also result in substantial flows to or from regions, typically increasing the disposable income of weaker regions in all Member States, while reducing that of more prosperous regions. Two distinct sorts of mechanism lie behind these transfers (figure 4.1): explicit equalisation schemes that ‘tax’ richer regions to support poorer ones (‘horizontal’ transfers, of which the German länderfinanzausgleich is the best known); and the interplay of national taxation and social charges, social protection payments and grants from central government to lower tiers (‘vertical’ transfers). In some countries, the differences come about largely on the tax side, while in others it is the distribution of spending which matters most.

CENTRAL GOVERNMENT

RICHER SUB-NATIONAL

AREAS

POORER SUB-NATIONAL

AREAS

HorizontalEqualisation

Per capitaPublic

Expenditure: Core Services

Per capitaPublic

Expenditure: Core Services

Discretionaryspending

Taxes andsocial charges

Figure 4.1 Types of inter-governmental flows


55

These transfers are different in character from the public expenditure flows discussed above, but analytically they are very similar in their impact. Although there are vast differences between Member States in the institutional arrangements, the data for all countries show that the combined effects of these mechanisms has a marked effect in reducing regional disparities in living standards.

Fiscal flows affecting regional income relativities Because most public spending and transfers mainly affect current consumption, they advance cohesion primarily by redistribution, but do not necessarily add to the productive potential of net recipient regions. Much of the first round effect will be to boost (reduce) activity in public services or population related private goods and services. The aggregate redistributive effect depends on the structure of the receipts, that is on the mechanisms underlying taxation and in particular on the degree of tax progressiveness. It also depends on the functional structure of public expenditure and the criteria determining its territorial allocation. Particular spending initiatives can also make a difference, a striking current example being the Olympic Games in Athens which are absorbing a high share of Greek public investment. Finally, it depends on the organisation of the public sector, in particular the distribution of spending competences (and the funding mechanisms) between the different government levels. Other things being equal, a country with a large public sector and substantial regional disparities will tend to have a greater degree of redistribution through this channel.

These institutional differences also make it difficult to arrive at a common definition of public budgetary flows and the research conducted for the background study suggests that there is little to be gained by trying to shoehorn data constructed on such varying bases into a single empirical measure. Indeed, in some Member States, it proved to be unrealistic to construct data in an appropriate form at all23 and it is also important to note that in most Member States, major changes have taken place, or are in prospect, that make comparison of the cohesive effect over time and across countries problematic.

It is also important to note that there is a conceptual difference between expenditure that takes place ‘in’ a region and spending which ‘benefits’ a region wherever it takes place. Both are valid from different political economy perspectives, but for the most part the data that can be collected are confined to the location of the spending rather than its ultimate impact and evidence suggests that the difference is not that great in practice (McLean et al., 2003). Recently completed projects in the UK and Italy illustrate some of the problems and the corresponding difficulties in using readily available published data. In Italy, a project initiated by the Tesoro (2003) to establish Territorial Public Accounts (TPA) has built up a concept of the enlarged public sector, adding social security, the activities of public corporations and certain other public functions to conventional public expenditure, then netting out intra-public sector transfers. The study in the UK (McLean et al, 2003) illustrated the range of methodological pitfalls in establishing a coherent measure of the territorial distribution of public spending.

4.1 Differing national systems Some systematic features of the different systems at present can be identified. Decentralisation is written into the constitution in countries with federal structures, such as Belgium, Germany and Austria. Strong local government has long been a characteristic of the three Nordic Member States, in contrast to those where the central government has been dominant, such as France and the UK, even if recent changes have seen some shifts. On the whole, unitary states control more resources centrally and transfer higher proportions of GDP to lower tiers of government than federal ones, not surprisingly when the point of

23 Thus, the Dutch rapporteur stated that there is no consolidated overview of the territorial incidence of the policies of the central government in the Netherlands for recent years. However, there is an outstanding inventory for an earlier period: 1979-1986 (Stoffelsma, 1992).


56

decentralisation is to give both power (including command of budgetary resources) and responsibility to lower tiers of government.

For historical and cultural reasons, or to accommodate nationalities within the Member State, Italy, Spain and the UK accord special status to certain regions, giving rise to greater devolution of powers and, in most cases, different funding formulae. This is true to a lesser extent in Finland, where the Åland region has special status not only in relation to the rest of the country, but also in relation to EU membership, while Portugal also gives special fiscal and funding terms to its two island groupings, the Açores and Madeira. Two of the German city states – Berlin and Bremen – receive special treatment, as do Corse and the ‘DOM/TOMs’ in France. For some of these special status regions, public transfers provide net benefits, even though they are comparatively rich regions in the national context: Åland and Trentino Alto Adige are examples. On a purely economic calculation, such payments might be regarded as anti-cohesive, but they clearly serve a purpose in relation to political dimensions of cohesion.

Institutional change and public transfers In several countries, extensive changes in the structure of sub-national government can be expected to alter the manner in which transfers affect cohesion. The trend towards decentralisation which tends to result in sub-national governments having both greater fiscal resources and a more extensive range of competencies is critical. Belgium, for example, now has far greater devolved fiscal policy, with both the regions and the French linguistic community controlling substantial amounts. Interestingly, in the Belgian system, transfers from the regions to the lower tiers of government are now small amounts. German unification caused a significant rethinking of the finanzausgleich after opposition from both net contributors and net recipients in West Germany, with a shift from pure horizontal equalisation to a system in which federal resources topped up the fund.

Portugal has also recently (a law passed in 1999) recast the links between central and sub-national government by replacing a single Fundo de Equilíbrio Financeiro with four new funds. Between them, they assure equitable ‘vertical’ burden sharing between, on the one hand, central and sub-national authorities in funding the latter’s expenditure, and on the other, ‘horizontal’ fairness between sub-national units with differing fiscal capacities.

An extreme case in prospect is Italy, where a succession of changes – once complete - will make it the most decentralised system in the OECD apart from Canada. Re-assignment of taxes and spending competencies from higher to lower tiers of government have the potential greatly to change the potential cohesive effect of public transfers. The trouble is that because the terms remain under discussion, the effects cannot fully be anticipated. Under the most favourable scenario (for them) the weakest regions will obtain the same level of expenditure, while in all the others the constitutional reform will produce an unfavourable effect on territorial cohesion. A number of new measures have been introduced or are in the pipeline. One of these, the National Equalisation Fund, which is supposed to compensate regions for low tax yields, uses criteria which the Mezzogiorno regions claim has discriminated against them.

4.2 Criteria for resource allocation Transfers operate on a variety of ‘keys’ and systems. In some countries there is a ‘needs’ assessment that allocates resources according to a range of social and other indicators. Such formulae determine the degree of redistributive impact because they explicitly factor into the transfer the normative choices on national cohesion policy. Table 4.1 shows the character of some of the principal budgetary mechanisms in the Member States and the basis on which they operate. Thus in Finland, there is a standardised level of services which constitutes the benchmark for the funding of all regions. Although the municipalities have a sizeable tax base, they do not have the power to determine tax rates. The redistributive effect comes about because richer regions are not able to retain all the tax proceeds, while poorer one receive net


57

transfers and, in addition, there are discretionary national budget lines which flow disproportionately to poorer regions.

Table 4.1 Principal Member State Budgetary Transfer Mechanisms Member State Nature of mechanism Criteria Belgium Central allocation federal budget lines

and social security; horizontal sharing of VAT receipts

Decided politically at central government level

Equalisation mechanism to reallocate income tax

Range of need indicators Denmark

Social security Demand driven Germany Horizontal equalisation Formulae Greece PIP fund to allocate major programmes

to regions Linked to Structural Funds, but with additional national criteria

Spain Inter-territorial compensation plan Only to Objective 1 regions France Central funding of regional budgets Special rates for selected

regions Ireland Inter-play between tax payments, social

transfers and grants to local authorities Indirect effects from national rates or social priorities

Italy Higher public spending in many richer regions offset by lower tax yield in South

Combination of per capita and special factors

The Netherlands Central government funding of lower tiers

Standard service for standard tax rate

Austria Finanzverfassungsgesetz Formulae Portugal Range of funds to transfers from centre

to sub-national governments Mix of per capita and cohesion related indicators

Finland Transfer from central government and correction of local income tax yield

Based on standardised per capita services, and imputed taxes and costs of services

Sweden Equalisation payments and special spending programmes

Indicators of need

UK Central government funding of sub-national budgets

Combinations of formulae including standard services and weighting for ‘nations’

In several countries, it is striking that it is the local or municipal tier of government, not the region, that is the level at which the net redistributive impact of national expenditure and transfer policies is calibrated. The provinces and municipalities in the Netherlands spend about 12 % of Dutch GDP (municipalities: 11%; provinces: 1%) and employ around 22% of the country’s civil servants. Financial relations between governments in the Netherlands are based upon the so-called ‘third aspiration level’, which says that a local authority should be able the make available to its citizens an equal level of services at an equal level of taxes. Although local and regional authorities can vary service provision by adjusting the rates on taxes assigned to them, most of their income consists of transfers from central government. These grants are based on a complicated framework that has just one central goal: equalisation (spatial and personal cohesion) and comprise a mix of general grants, which can be use more or less freely, and targeted grants that can only be used for specific tasks. General grants are calculated on the basis of a large number of indicators: size, population density, soil quality, social structure, degree of urbanisation, etc., and the local tax capacity.

In Portugal, a general fund allocates resources to the three NUTS level 1 regions, largely on a per capita basis, but with additional criteria that benefit the two island regions. This general fund also uses a range of criteria to determine allocations to municipalities within each region.


58

A second fund, with explicit cohesion aims, is limited to less developed municipalities, while two other funds ensure that the municipalities have adequate basic resources. Broadly, the public transfers are inversely correlated with income per head and thus have a pro-cohesive impact. In continental Portugal, Lisboa e Vale do Tejo, the region with the highest GDP per capita, receives less than a third of the per capita transfers of Alentejo, which has the lowest GDP per head. As a percentage of GDP, the figure for the capital region is barely 1%, whereas for Alentejo it is 6.1%.

In the UK, there are overlapping mechanisms: aggregate public expenditure in the territories (i.e. the four constituents of the union) is based on the ‘Barnett formula’, named after a former Treasury Minister, which originally aimed to give per capita figures higher than the national average to Scotland, Wales and Northern Ireland, with England correspondingly lower. A parallel standard spending assessment, similar in character to the Finnish and Dutch ones, determines flows to local government in England (local government in Scotland, Wales and Northern Ireland being a competence now devolved to the regional level) and health care is also based on a needs assessment. According to McLean et al. (2003), the health and local authority budgets plus the small amounts allocated to regional development within identifiable public expenditure account for half the total and can be regarded as formula driven. Of the remaining identifiable public expenditure, some 38% is social protection which they classify as demand driven, with the balance arguably discretionary.

The budget of the French regional authorities are financed mainly through transfers from the State. The French overseas territories have regional per capita public spending roughly three times the average for metropolitan France and Corse 3.5 times the national average. Most other regions are within a relatively narrow range, although the tendency is for poorer regions to receive relatively more. There are, however, several anomalies and the correlation between income per head and public spending is weak. Thus, the Ile de France also receives a premium over the national average – arguably because of higher service delivery costs - while Lorraine is well below it.

Some two-thirds of total public expenditures can be considered to have a regional impact in Spain (de la Fuente, 2001; Castells et al. 2000). Of the spending by the central government that can be considered to have a territorial effect, 80% is channelled through personal redistribution mechanisms (de la Fuente 2001; Correa and Manzanedo 2002). Public spending with a regional impact in Spain is inversely correlated with income per capita, and net fiscal balances reflect the combined effect of the tax and social protection systems (de la Fuente, 2001). Two notable exceptions are the Basque Country and Navarra which receive net inflows in spite of their high income per capita levels, due to the special treatment that they receive in the system for political reasons. There is an explicit corrective factor inversely proportional to the per capita income in each territory, and special compensation for the insular territories. In recent years, this correction has had a considerable redistributive impact in favour of Andalusia, Galicia and Extremadura. This additional money was primarily invested in those regions to finance the construction of roads, housing and education.

Inter-regional equalisation systems Horizontal equalisation systems are becoming more common as countries adopt decentralisation. The German system (see box 4.2) was originally based on the principle that richer regions pay in to the equalisation fund and poorer ones receive from it, with no explicit role for central government other than indirectly mediating the flows. In its current form dates from 1995 when the separate systems in East and West Germany were merged under the Solidarpakt and now comprises a mix of pure horizontal equalisation and federal topping-up. The computation of the respective positions of each Land works through a number of stage. First, it takes account of tax capacity based on tax instruments that are either exclusively ‘owned’ by the Land level, or shared with the Federal level. The system aims to compensate for differences in aggregate revenue, then to allocate the tax revenues pro rata to population. However, it gives a special position to the three city regions, Hamburg, Bremen and Berlin on


59

the grounds that they have special requirements. The aim is to raise enough to ensure that no Land has a fiscal capacity lower than 95% of the national average. Supplementary grants from the federal level then raise these länder to 99.5% of the national average.

Box 4.2: Equalisation in Germany

The principal channels through which money is transferred between the Länder and between the Federal level and the Länder in Germany is the German Financial equalisation system, the Länderfinanzausgleich which has its legal basis in the Finanzausgleichsgesetz. Following unification, the system developed in it current form when the separate systems in East and West Germany were merged in 1995 under the Solidarpakt.

These revenues are allocated between the federal level and Länder and among the Länder through a series of procedures. A primary allocation concerns taxes. Much the largest component of German government revenue (roughly 70%) is Gemeinschaftssteuern (shared taxes), e.g. income tax and other profit taxes, turnover tax which, in 2001, were shared as follows:

BUND LÄNDER KOMMUNEN Income tax 42,5 % 42,5 % 15 % Corporate tax 50 % 50 % - Capital gains tax 50 % 50 % -

Within the turnover tax, 75% of the Länder share is apportioned by population, with the balances reserved for ‘financially frail’ states. This ensures that the fiscal resources of each Land are raised to at least 92 % of the average. There is then a ‘secondary’ financial equalisation which corrects the primary tax distribution to guarantee equal per capita tax distribution among the Länder. A further stage then provides for transfers from Federal level to the Land designed to top up further for those Länder that are below the average or face special circumstances.

Results of specific horizontal equalisation scheme (Billion DM) (Positive: received; negative: paid)

1990 1994 1995 1999 2000 2001 Trend since 1995

Hessen -1,4 -1,8 -2,2 -4,7 -5,4 -5,1 - Baden-Württemberg -2,5 -0,4 -2,8 -3,4 -3,9 -4,2 - Bayern 0 -0,7 -2,5 -3,2 -3,7 -4,5 - Nordrhein- Westfalen -0,1 0,2 -3,4 -2,6 -2,2 -0,5 o Hamburg 0 0,1 -0,1 -0,7 -1,1 -0,5 o Schleswig- Holstein 0,6 0,1 -0,1 0,2 0,4 0,1 o Saarland 0,4 0,4 0,2 0,3 0,3 0,3 o Rheinland-Pfalz 0,5 0,7 0,2 0,4 0,8 0,4 o Bremen 0,6 0,6 0,6 0,7 0,9 0,8 + Mecklenburg- Vorpommern - 0,0 0,8 0,9 1,0 0,8 + Niedersachsen 1,9 1,0 0,5 1,0 1,1 1,9 + Brandenburg - 0,0 0,9 1,1 1,3 1,0 + Thüringen - 0,1 1,0 1,2 1,3 1,1 + Sachsen-Anhalt - 0,1 1,1 1,3 1,4 1,2 + Sachsen - -0,1 1,8 2,1 2,3 2,0 + Berlin - - 4,2 5,3 5,5 5,2 + Total yield of balancing 4,0 2,9 11,2 14,6 16,3 14,8 +


60

The transfers are substantial. In 2000, Berlin, the Land which receives the most from the finanzausgleich, had its fiscal resources increased from 68.6% of the national average to 95%. Berlin’s fiscal capacity is increased by a further 4.5 percentage points by supplementary federal grants, yielding a total worth 6.35% of the region’s GDP in that year. However, because it benefits greatly from a special supplementary programme for regeneration, Bremen receives an even higher percentage of GDP: 6.5%. Net transfers to the eastern Länder range from 4.82% to Brandenburg to 5.33% for MV. For Hessen - the Land which pays proportionally most in finanzausgleich - the effect is to reduce fiscal capacity from 126% of the national average to 106%, equivalent to a reduction of 1.5% of the Land’s actual GDP.

Sweden, where the public sector is the largest in the EU, also has a form of horizontal transfers. One reason for the expansion of the public sector is that the national government has delegated and imposed new obligations on local government as the public sector has grown. The municipalities’ main income comes from the local personal income tax and the transfer system aims to compensate those municipalities with weaker fiscal capacity: in short, ‘wealthier’ municipalities pay a fee to ‘poorer’ municipalities. As in Germany, it is supplemented by national grant-in-aid programmes, but it is the horizontal transfers which are the most important and it is noteworthy that the transfers occur at a more disaggregated (county) level than in Germany. Central government policy has also sought to discriminate in favour of less competitive economies by relocating certain public functions – government agencies, universities - to those municipalities, thereby boosting the underlying tax base.

Similar arrangements operate in Denmark, although here the level at which the equalization functions is still more disaggregated. Denmark is a single Level II region in the NUTS classification, but internally is split into fourteen counties, plus two special status regions (Copenhagen and Frederiksberg) and 271 municipalities. This structure is due to change soon: a Strukturkommissionen is preparing a number of proposals for far-reaching changes in the structure of the public sector and of the division of tasks and responsibilities between the different levels of Government, and is expected to finalise its proposals by the end of 2003. Both the regions and the local municipalities collect their own income and property taxes and consequently have a rather high level of autonomy.

In Austria, the main component of sub-national spending is derived from the Verbundsteuern (shared taxes or compound taxes) through which the proceeds of all major national taxes is divided (vertically and horizontally) by a fixed system. This is, negotiated periodically for four to five years and embodied into a single federal law (the Finanzausgleichs-Gesetz) which regulates most public transfer payments as well. Data on the regional incidence of social security spending is fragmented: the impact of the contributions they receive could be estimated analogously to income taxes, whereas their expenditure by region is only available for some instruments. As for taxes and federal spending, no current study examines the general regional impact of these institutions. There is no current work estimating the regional impact of federal expenditures.

Dovetailing with Community support In countries which receive substantial amounts from the Structural Funds, some national policies are quite closely tied to EU funding. Thus in Spain, the inter-territorial compensation fund allocates complementary funding only to Objective 1 regions, although there is also a special ‘Teruel’ fund which provides support to that part of the Aragon region, even though Aragon as a whole is not designated under Community regulations. The main instrument of government policy for promoting economic and social cohesion in Greece is the Public Investment Programme (PIP) which finances large projects in infrastructure, education, health, culture and other key sectors of the economy at national and regional level, with most of the funds allocated by the PIP going through CSF III. Those regions which receive the highest per capita Community transfers under the current CSF (Dytiki Makedonia and Voreio Aigaio), receive five and six times as much from the Greek state as Attica, which constitutes a


61

substantial, pro-cohesive transfer. Ireland, too allocates proportionally more from the national budget to the BMW region to match funding for Structural Fund programmes.

4.3 Composition of public spending Although governments tend to have surprisingly similar patterns of public expenditure when measured in terms of broad categories, much greater differences emerge in the detail. Table 4.2, adapted from a recent OECD report, presents an interesting distinction between merit goods – those that could be regarded as publicly funded provision of services – and income transfers that are more explicitly about social cohesion. This way of looking at public spending differs somewhat from the customary presentation of social protection which includes health and social services – here classified as merit goods – to focus more specifically on transfers that directly support incomes.

The merit goods category averages 12.8% of GDP in the EU-15, but is generally higher in the richer Member States, while the lowest levels are in the four cohesion countries (Greece, Ireland, Portugal and Spain) and in Italy. Relatively low spending on merit goods is also a characteristic of the three countries with the highest levels of public debt: Belgium, Greece and Italy. It is a reasonable conjecture that the spending on these merit goods is determined principally by population and demographic structure, rather than relative prosperity, because it is generally the case in a Member State that national policy preferences shape per capita expenditure. There is, consequently, unlikely to be a systematic effect of such spending on cohesion. If, for example, a region has a higher than average proportion of school-age children, it would be expected to receive above average expenditure on education, irrespective of its relative prosperity.

There are sizeable differences between Member States in the components of merit goods. Denmark, Sweden and (to a lesser extent) Finland devote substantial resources to the two categories of support services, whereas Belgium, Italy and Spain give these services only very limited resources. Spending on education exhibits less variability, although Greece stands out for its low proportion, while the two Scandinavian countries are well above the average and Portugal too is shown to favour education24. Germany also has a low education share, but here the demographic structure has to be taken into account. Public spending on health appears to be correlated with relative prosperity, being lowest in the cohesion countries and higher in richer countries, although the highest ratios in Germany and France can be considered to reflect social cohesion priorities.

Public spending on income transfers, in contrast to merit goods, tends to be sensitive to different needs. Aggregate spending on income transfers seems to reflect a combination of relative prosperity and the social model of the Member State. They are lowest in Ireland and well below the EU average in Spain and Portugal, but Greece is only marginally below the average, ahead of the Netherlands. France and Italy have the highest levels of income transfers and most of the other countries are quite closely bunched around the average.

24 Some caution has to be exercised in looking at proportions of GDP, especially for countries with a GDP relatively far from the EU average such as Portugal, since the absolute amounts spent may tell a different story. In Ireland – a country known for its long-standing commitment to education - the apparently low level of education spending is partly explained by using GDP rather than the much lower GNP as the denominator.


62

Table 4.2 Public expenditure by major functions as a % of GDP, 1999

Merit goods Income transfers

TOTAL

Education

Health

Services for the

elderly and disabled people

Family services

TOTAL

Old-age pensions

Survivors pensions

Disability

Sickness

Family cash benefits

Unemployment benefit

Austria 13.8 6 5.8 0.9 1.1 19 9.9 2.9 2.3 0.2 1.9 1.4

Belgium 11.4 5 6.1 0.1 0.2 18.1 7.3 2.5 1.7 0.4 2 3.8

Denmark 18.8 6.8 6.8 3 2.2 17.8 6.8 0 2 0.7 1.5 5

Finland 14 5.7 5.3 1.5 1.4 18.3 7 1 3.1 0.4 1.9 3.9

France 15 5.9 7.3 0.7 1.2 19.7 10.6 1.6 1.1 0.5 1.5 3.1

Germany 13.7 4.4 7.8 0.7 0.8 18 10.5 0.5 1.4 0.3 1.9 2.6

Greece 9.1 3.4 4.7 0.3 0.7 16.9 10.2 2 1.1 0.8 1.2 0.7

Ireland 9.5 4.3 4.6 0.4 0.2 10.6 2.5 0.9 0.8 0.7 1.6 2.9

Italy 10.8 4.8 5.5 0.2 0.3 19.1 12.8 2.6 1 0.7 0.6 1.4

Luxembourg .. .. 5.4 0.5 0.4 15.2 7.8 0.9 2.4 0.7 2.3 0.8

Netherlands 12.1 4.5 5.9 1.3 0.4 16.1 6.2 0.8 2.4 1 0.8 3.9

Portugal 11.3 5.6 5.1 0.3 0.3 12.5 6.3 1.4 1.9 0.5 0.6 1.6

Spain 10.2 4.4 5.3 0.3 0.1 13.9 8.1 0.8 1.3 0.9 0.3 2.2

Sweden 18.6 6.6 6.6 3.7 1.7 18.9 7.5 0.7 2.4 1.1 1.6 3.9

United Kingdom

11.6 4.6 5.6 0.8 0.5 17.8 9.8 1 2.7 0.1 1.7 0.6

EU average 12.8 4.9 6.4 0.7 0.7 17.9 9.9 1.3 1.6 0.5 1.4 2.3

Source: OECD (2003)


63

Perhaps surprisingly, the composition of income transfers varies enormously. Italy has a very large commitment to old-age pensioners (12.8% of GDP, some two percentage points above the EU average), but offers little support to the unemployed, with outlays on unemployment benefit of only two thirds of the EU average, despite the high level of unemployment. Several countries spend relatively high amounts on transfers to the disabled, with Finland and Austria paying out most, whereas France, Ireland and Italy are the laggards. Family benefits are highest in Belgium, Luxembourg and Germany, but are very low in Italy, Portugal and Spain, in the last of which outlays are less than a quarter of the EU average. Sweden devotes 1.1% of GDP to sickness benefits and the Netherlands 1%, ten times as much as the UK. Belgium and France clearly transfer much more to the unemployed than Italy, although the unemployment rates in all three countries in 1999 were comparable, while Denmark and the Netherlands manifestly are much more generous. The background to these disparities is taken up in more detail in the section on labour market policies.

The implications for cohesion of these national differences in public expenditure patterns cannot easily be appraised. Old-age pensions (and possibly also survivors’ pensions), amounting to roughly half of income transfers for the EU as a whole, will accrue to regions in which the pensioners are resident and not necessarily to lower income regions, or to regions facing economic difficulties. Indeed, those pensioners who are geographically mobile may well gravitate towards attractive richer regions where the quality of the local environment and public services is superior. It follows that the impact of old age pensions on regional prosperity does not necessarily reflect either the region’s underlying circumstances or explicit policy aims. One view of old-age pensions is that they are simply institutionalised savings that will primarily benefit residents of regions attractive to retired individuals and may, consequently, not be linked at all to regional per capita GDP. These transfers manifestly derive from social policy choices about the distribution of national welfare policies, with social groups as the targets of policy, but may well have little or no correspondence with regional income disparities. Family benefits are also very varied in scale, ranging from 2.3% of GDP in Luxembourg, to just 0.3% in Spain and 0.6% in Italy: these are big differences.

Other elements of income transfers, in contrast to the depiction of old-age pensions as institutionalised savings, are more explicitly linked to social problems although a proportion of old-age pensions is, in any case, a form of social assistance, though more so in some countries than in others. Disability, sickness and (especially) unemployment benefits are transfers motivated above all by ‘solidarity’ concerns. They are, consequently, most explicitly cohesion related transfers and it is also reasonable to assume that the recipients of such assistance will be relatively more concentrated in regions facing economic difficulties. However, the key for expenditure patterns will be indicators such as unemployment rates, rather than (necessarily) income per head.

Totting up the three headings of disability, sickness and unemployment to obtain a sub-aggregate of transfers that are most explicitly social solidarity, the figures show that Denmark spends three times as much (7.5%) as Greece. Italy and the UK also redistribute smaller amounts under these headings, while Belgium, the Netherlands, Finland and Sweden are generous. The scale of these transfers means that the cohesion effort of these Member State policies is very much greater in the generous countries.

4.4 Impact on cohesion The net impact of public spending and transfers on regions is the balance between two large gross flows: public receipts and public expenditure.25 Both gross flows are, in addition, subject to methodological problems and there are other are awkward methodological issues in calculating impacts. For example, where the disparities between regions in GDP levels are

25 Social security payments are typically much less easy to assign by region, and may (as in Belgium) be politically sensitive because governments do not like to be too transparent about winners and losers.


64

very large, the choice of denominator to benchmark the data becomes crucial. Low per capita receipts may actually constitute high proportions of regional GDP. This is the case in Italy where current public spending (less interest payments) as a proportion of GDP is highest in Sardegna, and above the national average in all the Southern regions, though also well over average in Liguria, Valle d’Aosta and Lazio. But measured per capita, the South seems to be poorly treated.

Assigning taxes to regions is also problematic for a number of reasons. Where they are collected is not necessarily where the tax burden lies and there are many tax instruments for which the amount generated in a region can only be imputed, and even where official statistics exist, they have to be treated with some caution. The fiscal capacity of a region bears on the cohesion equation, as regions that are less prosperous often end up contributing the least to national fiscal receipts. This is manifestly the case in Italy, again, where the estimated balance of payments deficits of the Southern regions signal that even though most parts of the Centre-North tend to receive more public expenditure per capita, the tax yield in the South is comparatively so much lower that it engenders a net fiscal inflow (see figure 4.1). Taxable income per inhabitant in Italian regions ranges from a maximum of 136.1 in Lombardia to a minimum of 52.4 in Calabria, with wider disparities at the level of the province.

Net imports of Italian regions 1995-2000

-20.0

-15.0

-10.0

-5.0

0.0

5.0

10.0

15.0

20.0

1995 1996 1997 1998 1999 2000

% of regional GDP

North-West

North-East

Centre

South

Figure 4.1 With these caveats in mind, it is nevertheless instructive to look at some of the estimates. Information on the nature and scale of a range of mechanisms - some of which operate simultaneously within a single country - used in selected Member States is presented in table 4.3. What the table demonstrates is that these are potentially powerful instruments for altering regional disparities, but that their impact can vary substantially depending on the setting.

In Sweden, the data are most comprehensively available at the more disaggregated level of Län (counties). The powerful equalisation scheme in force makes a substantial correction that reduces the county level coefficient of variation from 18.5% to 13.0%, with the income of the two most assisted northerly counties raised by around a third, while the transfer from Skåne in Southern Sweden is 9.7% of GDP and that of Stockholm 4%. Bearing in mind that this is only one of several pro-cohesive mechanisms, the degree of redistribution is very large.


65

Table 4.3 Redistributive impact of selected budgetary mechanisms Mechanism

involved Net transfer to most assisted region (% of actual* GDP)

Net transfer from most prosperous region (% of actual* GDP)

Impact on coefficient of variation**

Finanzausgleich 3.74% 1.50% Germany Supplementary federal payments

4.50% -- Reduced from 30.9% to 29.9% #

Finland Equalising transfer to municipalities

6.5% 0.5% Reduced from 27.8% to 25.6%

UK Net fiscal transfer estimated by allocating all flows

29.5% 12.6% Reduced from 27.2% to 16.8%

Belgium Aggregate impact of financing between three regions

6.7 3.6 n.a (for only three regions)

Ireland Net fiscal transfer estimated by allocating all flows

9.53 2.26 n.a. (for only two regions)

Sweden Equalisation of county tax bases

36.8 9.7 Reduced from 18.5% to 13.0%

Notes: * The actual GDP includes the impact of the transfer on the region. ** The difference in the coefficient shows the percentage difference between the actual co-

efficient of variation and a hypothetical one calculated by removing the effect of the transfer. Because the coefficient is sensitive to the number of regions in a Member State, caution needs to be exercised in comparing the implicit cohesion effect across countries

# The equalisation sharply increases the GDP of Bremen, already well above the average, dampening the apparent pro-cohesive effect as measured by the CoV. However, because it is a city state that attracts inward commuters to the workforce, Bremen’s measured GDP per head overstates the true income of its residents. Excluding Bremen from the calculation makes the pro-cohesive impact noticeably stronger.

In the UK, identifiable public expenditure in the fiscal year 2000/01 ranged from a low of 26.4% of GDP in the South East to as much as 63.9% in Northern Ireland, compared with a UK average of 35.9%. Although the various formulae allocate significantly more money automatically to Wales, Scotland and Northern Ireland, the effects of public spending and net transfers allocations are predominantly pro-cohesive in that it favours lower income English regions. Thus, the impact on GDP is greater in the relatively poor North East and clearly transfers net resources away from the three most prosperous regions - Eastern, London and South East. Nearly half the identifiable public expenditure in Northern Ireland is a net fiscal transfer, equivalent to about 29% of the regional GDP. At the other extreme, the net fiscal outflow from the prosperous South East is close to half the identifiable expenditure in that region (12.6% of its GDP).26

The net transfers arise on both sides of the account, with the lower taxes raised in poorer regions reinforcing the effects of public spending, and with social security payments playing

26 The linkages between London, which provides services consumed by South East residents, and the encompassing South East region complicate interpretation of these numbers.


66

an especially pronounced role in shaping differences between English regions.27 However, the data also reveal that the most generously supported region (Northern Ireland) receives more per capita spending than English regions on most public functions, the only exceptions being culture and media and transport, both relatively small budget lines. London stands out among the English regions for the high level of spending it receives under several headings, although the big disparity for London in the ‘law order and protective services’ category is arguably explained by the fact that these are national services located in London, rather than authentic benefits ‘for’ London. Without the net fiscal transfer, the coefficient of variation for regional GDP per head would be 27.2 compared with an actual value of 16.8, implying that the net transfer is associated with a reduction of just over a third in the ex-ante disparities between regions. This is, clearly, a very big impact. But given that nearly all the equalising effect comes through publicly-funded current consumption, the effect on economic as opposed to social cohesion is inevitably limited.

Ireland, admittedly only recently split into two NUTS II regions, exemplifies the top-down model of public transfers. Only a small proportion of overall taxes are ‘own resources’ of sub-national government, and the principal budgetary flows are to and from the central government. These include personal taxes paid by residents to central government and social transfers received, transfers to local government, producer subsidies and taxes and central government support for investment. Here we concentrate on the former two categories. The data reveal that the balance of these personal social transfers and taxes, plus local authority grants yields a per capita gain of €214 to the BMW region and €–309 to SE Ireland. The principal reason for the difference is the much higher tax payments made in SE, explained by a higher employment rate and average incomes.

Belgian transfers are, overall, from the Flanders region to Wallonia (especially) and to Brussels-Capital. Three distinct mechanisms are at work: the social security system, top down grants from the federal level and a form of equalisation of the proceeds of VAT. As can be seen from the table, the impact on Walloon GDP is substantial, raising it by 6.7%, while Flanders’ income is reduced by a lesser amount. Similarly, in Finland, the main pro-cohesive public finance mechanism appears to be the transfer from central to local government, with both taxes collected and the explicit equalisation system also contributing in a pro-cohesive manner. Together these flows reduced the ex-ante income of the richest region – Uusimaa – by 1% of GDP, and raised that of Eastern Finland by nearly 8 percentage points in 2000.

An interesting dimension of net transfers is how fiscal consolidation affects them. Here, recent changes in Finland illustrate the impact. In 1996, all Finnish regions received net transfers from central government, and the scale of the net transfer for the three poorer regions was very substantial, with gross receipts exceeding gross tax paid by a ratio of 3:2. In 2000, the rapid fiscal retrenchment meant that the prosperous Uusimaa region, which received a small net transfer in 1996, was paying some 30% more tax than it received in central government spending. Meanwhile the three poorer regions saw their net gains shrink with expenditure much closer to taxes paid. While, analytically, the relative positions of regions may not have changed, the effect of fiscal consolidation is to alter the absolute flows of resources to regions considerably, with an ambiguous effect on cohesion. With several Member States having established internal stability pacts (Italy and Austria are good examples) and others likely to follow, the likelihood is that the scope for fiscal transfers to be pro-cohesive will be further attenuated.

Even after removing the very substantial debt servicing charge discussed in section 2, above, the data show that gross public spending is substantially biased in favour of richer Italian regions, with several elements of public spending appearing to run counter to immediate cohesion with only Sardegna (a special status region) among the Southern regions having per

27 In the UK system, social security as defined here excludes publicly funded health care, so that the definition is narrower than that of social protection


67

capita receipts above the national average, in stark contrast to Campania, with just 79.5%. Social security payments which favour the retired population (concentrated in the North) and discriminate against regions with younger demographic profiles such as Campania, again, are a key factor, but by no means the only one. The upshot is that the primary distribution of public spending is very much anti-cohesive. Thus, social security receipts are highest in Liguria – a richer region – and well below the national average in all the Southern regions. The explanation is that social security is a demand driven transfer and the biggest concentrations of claimants are in the North, not least because of the institutional design of the Italian system which favours those who have been in work, especially pensioners. In this regard, the UK presents a sharp contrast, with social security transfers more closely mirroring relative unemployment rates.

The Italian TPA project will also to cover receipts of the public sector, but has, so far, only produced provisional data which precludes calculation of a comprehensive balance difficult. Italian data on regional net imports do, however, provide a proxy variable: net imports can be conceived of as the difference between aggregate supply and demand in a region, with the presumption that the gap is filled by public transfers. The data show that the very high net imports of the Mezzogiorno regions (over 20% of GDP throughout the 1980s) fell sharply after the fiscal crisis of the early 1990s to 13%. Comparable data for the late 1990s are not available because of a change in statistical methodology, but according to data on the post-1995 methodology, the net import figure for the Mezzogiorno stabilised in the late 1990s at a figure around 17% of GDP.

4.5 Commentary and key findings Manifestly, public expenditure flows and transfers result in very large shifts in the disposable incomes of the most and least prosperous regions within all Member States. They do so by very diverse mechanisms that are often far from transparent, yet there can be little doubt that they are powerfully redistributive. Within the Member State, regions with high levels of output generate higher tax yields and, typically, have lower public spending as proportions of their GDP. In some cases, the revenues accrue directly to the central government and are then disbursed to sub-national governments; in others, the primary income may be retained by the sub-national government but is then topped up by either, or both of, inter-governmental grants from central government or direct horizontal equalisation payments from richer lower tier governments.

For the most heavily supported regions, the net transfer through these mechanisms can be very substantial, reaching a third of the region’s income in some cases. As a result, these regions are able to maintain standards of public services that they might not be able to afford if they had to rely on their own tax bases. A consequence is that the share of the public sector in the most supported regions can rise to very high levels. But although this implies that poorer regions are relatively well supported, there are instances where the absolute amount of public spending may still be higher in more prosperous regions.

There are mixed views about both the degree to which and the manner in which public expenditure and transfers advance cohesion. On the one hand, by increasing current consumption – both public and private – the transfers raise living standards in net recipient regions and thus achieve a degree of convergence between regions. Public expenditure programmes can also contribute to long-term growth by activating resources and by boosting local economic activity, even if they have no direct impact on competitiveness. However, they can also have a negative effect on incentives and entrepreneurship and thus on the long-run capacity of the recipient economy to raise its competitiveness. In this regard, a distinction can be drawn between the generally positive effect on social cohesion and a more questionable impact on economic cohesion.


68

Key findings are: Public transfers constitute the main channel through which social cohesion is advanced.

But a variety of different schemes co-exist, and in some Member States several mechanisms operate simultaneously.

The institutional structure of the Member State affects how transfers are made, but there is a degree of uniformity in the potency of the mechanisms.

In some Member States, tax collection is centralised then resources are distributed from central government to sub-national tiers. In others, especially the Nordic countries, ‘vertical’ equalisation takes place through central government, while Germany typifies ‘horizontal’ equalisation.

The extent of sub-national autonomy determines whether it is policies implemented by the Member State level that appear to support economic development and thus underpin economic cohesion, or whether the decisions of the sub-national tier affect the policy mix.

Very big shifts in regional per capita income are achieved through public budgets, with income in the poorest regions raised by as much as a third. The diminution of income in net contributor regions is smaller because they are usually economically bigger.

These policies play a big part in sustaining territorial cohesion by ensuring that public services are comparable across regions.

Because Member State expenditure and transfer policies support current income, it is more open to question whether they influence economic cohesion.


69

5. State aid Despite completion of the single market and persistent efforts by the Commission to curb their use, state aids continue to be widely granted by Member States. On average, national spending on state aids averages just under 1% of GDP, but data for the year 2000 assembled in the Commission’s state aid scoreboard reveal a range from 0.4% of GDP in the UK to over four times as much in Finland, with 1.7% - see figure 5.1. Given the obligation on Member States to conduct their economic policies and coordinate them in such a way as to attain the objectives of development, cohesion and reduction of regional disparities [Article 159], how national state aid policies affect is an important question in relation to cohesion. This section explains the concept of state aid and why state aid may differ from other public spending in its impact on national cohesion. It briefly appraises EC state aid practice and then examines available data, considers how the effect of state aid on national cohesion may be tested and carries out a number of tests. In addition, it brings together and compares national data gathered in the course of this study.

Spending on State Aids by Member States

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

UK E P S A NL F D I GR IRL B DK L FIN

% of GDP, 2000

Regional Horizontal Agriculture & transport

Figure 5.1 State aid may be thought of as one of a number of policies that have the aim of affecting the supply-side of the economy by altering the allocation of resources. As discussed in section 2 of the report, their incidence, generally, and how they bear specifically on cohesion will depend on the programmes for which they are granted. It is, then, essentially an empirical matter whether they have a positive or negative effect on convergence or on the long-run potential of regions. If the recipients of state aids are located in regions that are regarded as targets of economic cohesion policy, the expectation would be that state aids will contribute to the realisation of cohesion aims and vice-versa.

It is important to note, though, that part of the rationale for supply-side policies may be to promote national economic growth. Underlying this proposition is the idea that intervention through the public sector can counter market failure in a way that achieves a more favourable aggregate outcome for the economy, even if there are some adverse consequences. The most obvious is that state aids (and other such interventions) have a cost to the public purse, so that other parts of the economy have to be taxed to provide for them. Other things being equal, this may have effects on household incomes and thus on social cohesion. But if a second order effect of spending on state aids is to support productive activity that might otherwise


70

disappear, the aids may well result in higher total production, at least some of which will trickle-down to the less well-off.

Since state aid is part of national expenditure, it may be tempting to conclude that national spending in the regions has the same effects as expenditure which is categorised as state aid. This, however, would be a premature conclusion. Apart from the fact that national spending in the regions is many times larger than the amount of state aid, the latter also has different objectives or different targets from national expenditure that takes place in regions. That is trivially true in the sense that the largest proportion of Member States’ budgets is devoted to salaries, social welfare, public health, public works and defence and runs into trillions. By contrast, state aid granted by the fifteen Member States totals no more than eur 85 billion per year.28

5.1 Which public measures are state aid? Not all public expenditure is state aid. It is necessary, therefore, to define what is state aid. The first paragraph of Article 87 of the EC Treaty stipulates that:

‘Save as otherwise provided in this Treaty, any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the common market.’

This wording indicates that a measure of public support is classified as state aid only if the conditions defined in that paragraph are all satisfied simultaneously. Alternatively, not all measures of public support, even those that may involve public subsidies, are necessarily classified as state aid. Box 5.1 summarises the conditions that define state aid.

Types of state aid granted by the Member States In principle, all state aid must be authorised by the Commission before it is put into effect. The Commission is obliged by EU law not to approve any aid scheme or individual award of aid unless it falls into one of the categories for which the Treaty allows an exception. Most aid authorised by the Commission falls into just two categories: 87(3)(a) and 87(3)(c). The guidelines issued by the Commission are a good indication on the kinds of state aid that may be exempted under Article 87(3)(a) & (c). These guidelines cover the following themes:29

Transactions by public authorities:

Government Capital Injections, Financial Transfers to Public Enterprises, State Guarantees, Risk Capital, Public Land Sales, Export Credit Insurance, Fiscal Aid (Direct Business Taxation).

Horizontal schemes:

Research and Development Aid, Environmental Aid, Rescue and Restructuring Aid, Regional Aid.

Sectoral schemes:

28 See the Commission’s State Aid Scoreboard which contains data submitted by the Member States. It can be accessed at: ‘http://europa.eu.int/comm/competition/state_aid/scoreboard/’. It does not necessarily include all state aid granted by national central and regional authorities. About 15-20% of all cases of state aid investigated by the Commission each year concern non-notified aid. This suggests that if Member States fail, for whatever reason, to notify all cases of state aid, it is also likely that they fail to report the true amount of state aid granted by public authorities within their territories. 29 They can be accessed at ‘http://europa.eu.int/comm/competition/state_aid/’.


71

Synthetic Fibres, Motor Vehicle Industry, Shipbuilding, Steel, Coal, Road, Rail and Inland Waterway Transport, Maritime Transport, Air Transport, Agriculture, Fisheries, Electricity, Broadcasting.

Box 5.1 What is state aid?

The five conditions that must all hold at the same time are the following:

1. aid must favour or confer an advantage to the recipient undertakings;

2. this advantage must be granted by the state or through state resources;

3. the advantage must favour certain (selected) undertakings or economic activities;

4. it must affect trade between Member States; and

5. it must distort competition in the common market.

The five conditions, together imply that the following types of measures are not state aid:

financial assistance from one private company to another

public assistance to individuals

financing of entities which are not undertakings, such as hospitals or schools

financing of projects which are not related to undertakings, such as infrastructure

regulatory measures that confer advantages to undertakings without transferring public money or without increasing state liabilities

general measures of economic policy that do not favour any particular undertakings, such as reduction of the standard rate of corporate taxation

measures that do not distort competition, such as incentives for early retirement of professionals

measures that have a purely local effect, such as the funding of municipal recreational facilities.

In relation to the last point, small amounts of public subsidies are not considered to have an effect on trade and competition and, for this reason, they are not classified as state aid. This kind of aid is the so-called ‘de minimis’ aid and is defined as aid that does not exceed eur 100,000 per beneficiary undertaking over a three-year period.

Lastly, funding that is classified as state aid must come from national sources (including regional and other sub-national authorities), not the EU budget. Community money that is used to finance the same project together with national money is still Community money (but, confusingly, it is taken into account in calculating the intensity of aid received by those projects).

In addition there are three Regulations on training aid, employment aid and aid to SMEs Aid granted in compliance with the provisions of these regulations need not be submitted for prior approval by the Commission before it is put into effect.

It is obvious that only regional aid has a direct relationship with assisted areas for the simple reason that it may not be granted to any other area. However, aid for R&D, training, employment and environmental protection and aid to SMEs also allow for higher intensities when the aid is granted within assisted areas. The problem is that it is not known how much of the aid that falls in these categories is granted within assisted areas and how much in other areas. Even if higher intensity rates are allowed for these kinds of aid, the total amount that may be granted within non-assisted areas may overwhelm the effect of any aid in assisted areas for the simple reason that most R&D, training and creation of new jobs may take place


72

outside assisted areas. The end result could be regional economic divergence incompatible with economic cohesion aims, rather than convergence.

Data collected by the Commission and published in its State Aid Surveys and the various editions of the State Aid Scoreboard indicate that there are large variations not just in the overall scale, as noted above, but also in the mix of aid granted by each Member State. For example, aid to manufacturing makes up between 15% and 35% of the total aid granted by each Member State. Aid to transport [mostly railways] varies between 15% and 70%, while aid to agriculture varies between 10% and 60%. For the EU as a whole, in 2001, manufacturing received 25% of total aid, transport 45% and agriculture 25%. The two biggest beneficiaries of state aid are, therefore, transport and agriculture which take up 70% of all aid. Over the years, these two sectors have consistently absorbed the majority of state aid. See Table 5.1 for the latest recorded amounts of state aid granted by each Member State expressed in absolute terms, as a percentage of GDP and in amounts per capita.

Table 5.1 Eligibility for state aid by Member State Member State % of eligible population Type of 87(3) area

B 30.9 [c] DK 17.1 [c] D 34.9 [a+c] GR 100 [a] E 79.2 [a+c] F 36.7 [a+c] IRL 100 [a+c] I 43.6 [a+c] L 32 [c] NL 15 [c] A 27.5 [a+c] P 100 [a+c] FIN 42.2 [a+c] S 15.9 [c] UK 28.7 [a+c] EC 42.7

This raises again the question about the impact of state aid on cohesion. Does state aid in agriculture and transport benefit less prosperous regions? There may be positive effects. Many poor regions also rely more on agriculture as their main source of income. In the case of transport, good links also make subsequent investment in remote regions more attractive. But there may be also negative effects. If state aid in agriculture follows the pattern of the CAP expenditure, then most aid may be granted to large farms. In the field of transport, most investment may be intended to improve connections between large urban centres.

In conclusion, the largest proportion of state aid is not directly related to regional development. However, most types of aid allow for higher intensities when the aid is granted to undertakings in assisted areas.

Differences between national spending in the regions and state aid State aid differs from general spending in regions in a number of ways. The first is that the objectives of state aid are broader than regional development. State aid may finance, for example, the rescue or restructuring of an enterprise in difficulty. Rescue aid is operating aid that does not promote new investments or jobs. Moreover, rescue and restructuring aid may benefit enterprises located in areas which are not eligible for regional assistance. The motivation may be laudable and, especially where there is a risk of concentrated


73

unemployment, could be regarded as a contribution to social cohesion, as discussed in section 2. Of the estimated total amount of eur 85 billion of state aid granted by the Member States only about eur 8 billion are explicitly earmarked for regional development [‘87(3)(a)’ areas]. It corresponds to about 25% of aid to manufacturing. Perhaps not surprisingly, Germany and Italy account for over 50% of the total regional aid granted by the Member States.

Second, current rules on state aid allow the granting of higher intensities of aid in assisted areas even when the aid in question does not have regional development as its primary objective and here the motivation is economic cohesion. This is the case, for example, with aid for R&D or environmental protection. By contrast, national spending is not normally modulated on a regional basis. Third, not all national spending in the regions is regional state aid. For example, national funding of infrastructural projects is not classified as state aid. Fourth, not all national spending in the regions is intended for regional development and, therefore, it may have a different multiplier from regional state aid. For example, spending of €1 million on the construction of a school may have less impact on the local economy than the spending of €1 million on the construction of a new factory, and will also bear on different facets of cohesion: the former affecting social cohesion, the latter economic.

Fifth, national spending in the regions may also be combined with EU structural funds and again have different multiplier effects due to different co-financing arrangements. One euro of national public money may be combined with half a euro of EU funds, while one euro of state aid may have to be matched with four euros of private money. Sixth, the funds for national spending in the regions come from public budgets. This means that they have an opportunity cost that reduces the net impact on both the national economy and regional economies. This opportunity cost is made up of the cost of tax collection and the related distortions of taxation plus other direct distortions caused by public expenditure.

Regional state aid State aid for the purpose of regional development is allowed by Article 87(3)(a) & (c). Article 87(3)(a) exempts aid intended ‘to promote the economic development of areas where the standard of living is abnormally low or where there is serious underemployment’. This condition covers areas with per capita income at PPS that falls below 75% of the EU average. This is exactly the same criterion as for Objective 1 regions as settled in the Structural Fund regulations for 2000-06. Areas that qualify for aid under Article 87(3)(a) are defined at NUTS II level, which is also the same for objective 1.

However, in practice Objective 1 regions and 87(3)(a) areas do not coincide precisely so there are some differences in the areas covered by national funds and areas covered by EU funds. For example, Sweden has a large Objective 1 region which is not classified as an ‘Article 87(3)(a) area’. This means that Swedish authorities are prevented from granting state aid to their Objective 1 region at the higher rates of intensity that are allowed for 87(3)(a) areas.

Regional aid is also allowed by Article 87(3)(c) because it provides for ‘aid to facilitate the development of ... certain economic areas, where such aid does not adversely affect trading conditions to an extent contrary to the common interest’. These areas are defined by the Commission in cooperation with the Member States. In principle, areas that qualify for this kind of aid should coincide with Objective 2 regions defined at NUTS III level. However, here too there is a small degree of discrepancy.

Under the 1994-99 financial perspective, 6.6% of the EU population lived in regions receiving structural funds but precluded from obtaining state aid and 2.7% in regions which were covered by national regional aid schemes but were not eligible under the structural funds. For the 2000-06 period, the coincidence of regions that the Commission was seeking between national and Community designations was not completely achieved. Some areas are eligible for structural funds but not for state aid and vice versa. The current situation is as follows:


74

Eligible population across the EU

87(3)(a) = 19.8% but Objective 1 = 22%

87(3)(c) = 22.9% but Objective 2 = 18%

This means that while only 40% of the EU population qualifies for EU structural action, 43% are eligible for regional state aid. Notwithstanding these problems, the situation within each Member State is currently as shown below. There are five Member States without areas eligible for aid under Article 87(3)(a) – see table 5.2.

Table 5.2 Eligibility for state aids Member state % of eligible population type of 87(3) area] Belgium [c] 30.9 Denmark [c] 17.1 Germany [a+c] 34.9 Greece [a] 100 Spain [a+c] 79.2 France [a+c] 36.7 Ireland [a+c] 100 Italy [a+c] 43.6 Luxembourg [c] 32 Netherlands [c] 15 Austria [a+c] 27.5 Portugal [a+c] 100 Finland [a+c] 42.2 Sweden [c] 15.9 United Kingdom [a+c] 28.7 EC 42.7

5.2 Impact of state aid on cohesion: empirical tests As explained earlier, national regional spending and state aid have a different impact on cohesion within Member States. What then should we expect the impact of state aid to be? Only regional aid may be expected unambiguously to be positively correlated with regional development and thus economic cohesion. Although other types of aid may allow for higher intensities of aid in assisted areas, the absolute amounts of aid granted in those areas may be significantly smaller than the amount of similar aid granted in non-assisted areas. There are no consistent data on the regional distribution of the different types of state aid, although some insights have been obtained from the national data collected as part of this study.

On the whole, however, the empirical literature has found considerable evidence that there is some positive relation between regional investment and convergence of regional incomes. This suggests that in general neo-classical factors outweigh agglomeration factors. The implication of this is that if we find a negative relationship between state aid and cohesion, it is likely to be the result of too much state aid to rich regions, which overwhelms the higher marginal productivity of poorer regions.

Generally, we expect regional state aid and horizontal state aid at the margin to have a positive effect on cohesion because most types of state aid allow for higher intensities in assisted areas. Certain sectoral state aid, such as for shipbuilding, automobiles and textiles, is also likely to have a positive impact, at least on employment, because most of the recipient industries are based in the old industrial regions many of which are assisted areas. But if the target is emerging industries, the reverse may hold. In terms of the opportunity cost of state


75

aid, since most aid is financed by revenue from taxation and since richer areas contribute a larger amount of tax revenue, state aid must also have an indirect positive effect on assisted areas because it imposes a smaller tax burden on them.

If a negative relationship is found between non-regional types of state aid and cohesion, then there must also be a policy conflict. On the one hand, Member States grant regional aid so as to promote the development of less prosperous regions. On the other, they grant aid for other purposes which may indirectly have a negative impact on cohesion because it makes other regions more attractive. If this is the case, then Member States themselves neutralise the effectiveness of their own policies.

Testing the impact of state aids The aim in the empirical work reported here was to test the hypothesis that state aid is positively related to convergence in regional income (or negatively related to disparities). Data on state aid are divided into four categories, each of which was tested separately:

• Regional state aid

• Horizontal state aid other than regional aid (e.g. environmental aid),

• Transport and agricultural (plus fisheries) aid,

• Total state aid.

Two types of regressions were run: over time for each Member State for period 1990 to 2000; and a pooled for all the Member States for two periods: 1999 and 2000. The independent variable (i.e. state aid) is measured as state aid per capita and state aid as a share of GDP (both expressed in real figures) to account for the fact that larger countries are more populous and grant larger amounts of state aid in absolute terms. Differences in regional income in terms of deviations from national average income per capita expressed in PPS are the measure of cohesion.

Discontinuities in data posed problems, with the result that for the time-series regressions, Austria, Finland and Sweden are omitted, because for them data on state aid start in the mid-1990s. Denmark, Ireland and Luxembourg are also omitted, because they were single NUTS level 2 regions up to 1999, with Ireland only being split into two regions from 1999. For the same reason, Denmark and Luxembourg were excluded from the pooled regressions.

Regression results All the regression results on time series are shown in Appendix table Table 5.A.1 and on cross-country data in Table 5.A.2. The time series results are summarised in box 5.2.

5.3 Country studies The regression results obtained above are based on analysis of data from Commission sources – primarily the State Aid Scoreboard. The project team has also gathered data directly from national and regional sources. On the whole, the main difficulties encountered by the country rapporteurs were (a) to ensure the consistency of available data with the definition of state aid and (b) to classify them in categories which are comparable to those used by the Commission in the Scoreboard. All of the rapporteurs faced the problem that there is no single national source of state aid, nor a single national register of state aid.

As will be seen, the data from countries for which national sources exist do not match those reported in the Commission’s Scoreboard. Some times our own numbers exceed those of the Scoreboard, some times they fall below those of the Scoreboard. By and large, categories of state aid do not match either across countries or with the Scoreboard.


76

Box 5.2 Summary of regression findings

For the time-series regressions, the following key points emerge:

In many cases, the sign of the slope of the independent variable is positive instead of negative, implying that larger amounts of state aid go hand in hand with larger regional disparities, so that state aids act against economic cohesion

The coefficient of correlation (R-squared) is in many cases very low, which means that the explanatory power of the equation used is low. In other words, the possible adverse impact of state aids on cohesion is rather muted.

Regional aid has relatively high R-squared values only in Belgium and Spain; yet in both countries the independent variable has the wrong sign. This implies that the effect of regional state aid is generally very low, but results in these two countries in greater divergence rather than convergence and thus does not promote economic cohesion.

Horizontal aid appears to be positively and significantly related to economic cohesion (or negatively related to disparities) in Belgium, Greece and the Netherlands.

Agricultural and transport aid has a negative sign and is relatively significant only in Spain. Thus it has a minimal impact on cohesion.

Total state aid appears to be significant and positively related to cohesion only in Belgium and Greece. Where it is negatively related to cohesion (i.e. it has a positive sign) it does not appear to be significant, except perhaps in the case of Germany.

The cross-country results for 1999 and 2000 indicate the following:

In most cases, the sign of the slope of the independent variable is positive instead of negative, implying that larger amounts of state aid go hand in hand with larger regional disparities

However, the coefficient of correlation (R-squared) is in almost all cases very low and lower than in the time series

Regional aid is negatively related to regional disparities, but the values of R-squared are very low.

The data reported below reveal three features of national state aid schemes which are common in most countries.

• First, the amounts of state aid granted in each region vary considerably from year to year. This holds even when the national amounts appear to be either stable or declining over time. The implication is that the overall reduction of state aid reported in the Scoreboard does not apply to all regions.

• Second, the regional distribution of state aid varies, depending on the type of aid. Some types of aid such as aid to R&D appear to be granted mostly in richer regions. Some other types such as aid for regional investment go mostly to poorer regions.

• Third, although most regional aid is granted to poorer regions, within the groups of these regions there is no precise correspondence between the allocated amount of aid and the need of each region as indicated by its level of income. This means that regional aid is not concentrated in the regions that need it the most.

Austria The Commission’s Scoreboard puts Austria’s state aid at EUR 2.06 billion in 2001 (at 2000 prices). More than 70% of that aid goes to agriculture, fisheries and transport. Only 5% of total aid or 20% of non-farm aid aims to promote regional development (table 5.3). Given the


77

federal structure of the country, state aid may be granted by the federal authorities, Länder authorities, and by local authorities such as municipalities. Länder authorities also co-finance expenditure under the European Union’s common agricultural policy and structural funds.

Table 5.3 Austria: Total state aid (EUR, million) Region GDP/head

(2000, PPS)

Total state aid Total state aid per head

Total state aid as % of regional

GDP 2000 2001 2000 2001 2000 2001 Wien 157 277.57 224.24 172.51 141.54 0.49 0.39 Salzburg 131 138.72 122.78 268.32 242.37 0.91 0.80 Vorarlberg 118 119.09 118.17 341.22 345.03 1.28 1.27 Tirol 113 241.01 218.71 359.72 333.04 1.41 1.28 Oberösterreich 109 355.13 420.25 257.52 312.69 1.04 1.23 Niederösterreich 97 353.33 354.03 229.14 233.75 1.05 1.05 Kärnten 96 187.34 222.40 332.75 400.80 1.53 1.82 Steiermark 96 350.84 306.22 291.88 259.09 1.35 1.18 Burgenland 73 185.71 197.42 668.02 721.02 4.04 4.29

Total 2208.74 2184.23 272.38 274.43 1.05 1.04

As revealed by appendix tables 5.A.4 to 5.A.6, Länder governments grant more state aid than the federal government. If richer Länder are able to grant larger amounts of state aid, this raises an important question concerning the impact of aid by lower tiers of government on regional cohesion within Austria. During the past few years, Länder authorities have tended to establish entities with the status of corporations for the purpose of managing funds, such as investment aid, channelled to the private sector. The country rapporteur has observed that this trend makes it more difficult to measure the precise amount of state aid granted.

Finland All the regions of the country at NUTS II level (five regions plus Åland Islands) have areas eligible for regional state aid. The population coverage for the 2000-6 period has been reduced to 42.2%. The Scoreboard indicates that in 2001 Finland granted a total of EUR 2,074 million of state aid (in 2000 prices) and was the Member State that used state aids most. More than 80% of that aid goes to farmers (65%) and transport (17%). The data gathered by the national rapporteur reveal a different picture. Aid for purposes other than agriculture is almost as large as aid to agriculture. Admittedly, however, the data may underestimate the true amount of state aid because it was not possible to determine the aid element in the EUR 770 million of loans and loan guarantees granted by public authorities or agencies under the control of the state.

In terms of regional distribution of the different types of aid, the data present a conflicting picture. Aid to agriculture goes to northern regions in proportion of almost three to one (table 5.4). Most aid for investment goes to the three poorest regions (60%). Investment aid accounts for 65% of all horizontal aid except R&D. The case of R&D is quite different. More than 80% of that kind of aid goes to the two richest regions.

Table 5.4. Finland: Regional distribution of aid to agriculture (EUR, million)

2000 2001 Aid for northern agriculture 349 337 Aid for southern agriculture 138 125

Total agricultural aid 576 554


78

It follows, therefore, that the regional distribution of aid varies depending on the purpose for which it is granted. Since most companies undertaking substantial R&D are likely to be located in the most economically active areas, R&D is concentrated in the richer Finnish regions. By contrast, larger amounts of investment aid are granted to the poorer regions, even though the second richest region also receives substantial investment aid (table 5.5).

Table 5.5. Finland: Regional distribution of state aid (EUR, million; 2002) Regions at NUTS II (excluding Åland)

GDP/head (PPS, 2000)

Aid for

R&D

Aid for other

horizontal objectives

(investment aid)

R&D aid / head

Other hor. obj. aid /

head

Total aid / head

Uusimaa 143 80.4 6.9 (0.5) 57.71 4.95 62.67 South 96 30.8 38 (24.6) 16.86 20.81 37.67 North 91 11.3 20.2 (14.6) 20.20 36.11 56.31 Middle 83 4.8 16.1 (10.6) 6.78 22.75 29.53 East 74 6.4 33.4 (24.5) 9.32 48.65 57.97

Total - 133.7 114.6 (74.8) Regions at NUTS II (excluding Åland)

GDP/head

(PPS, 2000)

R&D aid as % of

regional GDP

Other hor. aid as % of

regional GDP

Total aid as % of

regional GDP

Uusimaa 143 0.18 0.02 0.19 South 96 0.08 0.10 0.17 North 91 0.10 0.18 0.27 Middle 83 0.04 0.12 0.16 East 74 0.06 0.29 0.35

Total - - - -

Germany Germany is the largest grantor of state aid in the EU in terms of total outlays, if not in relation to GDP. According to the Scoreboard, the total amount of state aid provided in 2001 was EUR 23,274 million (in 2000 prices) with almost 50% going to agriculture and transport. Given the federal structure of the country, state aid may be given by the federal government, the Länder governments and local or municipal authorities (table 5.6).

Table 5.6 Germany: State aid by the three levels of government (€, billion) 2000 2001 Federal 10.1 9.5 Länder 11.2 11.2 Local 1.6 1.6

Total 22.9 22.3

The table reveals that Länder governments grant larger amounts of state aid than the federal authorities. Just as in the case of Austria, if richer German Länder are able to grant larger amounts of aid, this raises an important question concerning the impact of regional cohesion within Germany. In terms of the regional distribution of state aid, the breakdown between ‘new’ and ‘old’ Länder shown in table 5.7 were obtained. Given the fact that the ‘new’ Länder are poorer than the ‘old’ ones, this kind of aid is, therefore, largely directed to the less prosperous regions by a factor of about eight to one. Given that the population of the new


79

Länder is under a quarter of the German total, the intensity of the assistance in per capita terms is multiplied. Substantial financial assistance to the new Länder is also granted in the context of support for social housing but it is not possible to determine the proportion, if any, of state aid that it may include. Overall, there is a reduction in the amount of state aid granted by the federal authorities over time.

Table 5.7 Germany: Subsidies to investment undertaken by companies (EUR, million)

1999 2000 2001 (budgeted)

‘New’ Länder 899 800 697 ‘Old’ Länder 100 96 115

France The Scoreboard indicates that France grants the second largest amount of state aid in the EU. In 2001, it reached EUR 15.84 billion (in 2000 prices). Two thirds of that aid goes to agriculture and transport. Only EUR 0.7 billion or 4% of total aid or 12% of non-farm and non-transport aid is granted explicitly for regional development. Data were collected from DATAR on state aid granted to firms located in areas of ‘difficulty’, the so-called ‘zones PAT’ (Prime d'Aménagement du Territoire). These areas are defined at NUTS IV level. The amounts shown in table 5.8 below are very small; EUR 75.4 million over the 3-year period of 1999-2002. Even with these relatively small amounts of state aid, there is considerable variation in their regional distribution, as indicated by the amount of aid per head.

Table 5.8 France: Regional distribution of state aid (EUR, million)

Region GDP/head (2000, PPS)

State aid State aid per head (euros)

1995-1998 1999-2002 1995-1998 1999-2002 Ile-de-France 158 0.00 0.00 0.00 0.00 Rhône-Alpes 103 3.09 4.05 0.55 0.72 Alsace 103 1.75 4.48 1.01 2.58 Haute-Normandie 95 5.48 2.65 3.08 1.49 Champagne-Ardenne 95 1.96 0.45 1.46 0.33 Bourgogne 93 0.77 1.30 0.48 0.80 Centre 91 0.58 0.97 0.24 0.40 Provence-Alpes-Côte d'Azur 91 15.14 3.79 3.36 0.84 Aquitaine 90 1.78 2.87 0.61 0.99 Pays de la Loire 90 5.11 7.96 1.59 2.47 Midi-Pyrénées 89 4.31 4.52 1.69 1.77 Franche-Comté 88 0.78 3.28 0.70 2.94 Auvergne 87 2.24 3.63 1.71 2.77 Bretagne 86 6.44 4.33 2.22 1.49 Basse-Normandie 85 2.72 5.09 1.91 3.58 Lorraine 84 12.74 4.31 5.51 1.86 Poitou-Charentes 83 2.02 3.18 1.23 1.94 Picardie 82 1.68 1.32 0.91 0.71 Limousin 82 2.51 1.47 3.53 2.06 Nord-Pas-de-Calais 81 10.44 1.42 2.61 2.86 Languedoc-Roussillon 78 0.98 4.54 0.43 1.98 Corse 76 0.17 Na 0.67 na Total 101 82.25 75.35 1.41 1.29 Source: DATAR


80

Greece The Scoreboard indicates that for 2001, Greece granted EUR 1.3 billion of state aid. Agriculture, fisheries and transport absorb EUR 840 million. Of the remaining aid, EUR 419 million or 90% went to regional development. Indeed the country rapporteur notes that the main state aid instrument in Greece is Law 2601/98 on investment incentives for economic and regional development. Under that law, 420 and 465 investment projects were approved in 2000 and 2001, respectively. However, when the data reported by country rapporteur for state aid granted through the approved investment projects, the reported figures show only EUR 201.07 milion and EUR 220.92 million for 2000 and 2001, respectively. That is about half of the amount shown in the Scoreboard (table 5.9).

Table 5.9 Greece: Aid to investment and regional development (incentives granted through Law 2601/98) (EUR, million) Region GDP/head

(2000, PPS) State aid

State aid per

head State aid as % of

regional GDP 2000 2001 2000 2001 2000 2001 Notio Aigaio 80 8.36 11.05 30.62 44.81 0.17 0.22 Attiki 77 39.80 13.14 11.52 3.26 0.07 0.02 Sterea Ellada 76 45.22 23.02 68.10 49.97 0.39 0.20 Kentr. Makedonia 68 21.04 42.33 11.63 23.69 0.08 0.15 Dytiki Makedonia 67 1.80 1.43 5.92 5.38 0.04 0.03 Voreio Aigaio 66 5.45 5.22 29.78 28.84 0.20 0.19 Kriti 66 13.91 12.30 24.58 23.79 0.16 0.15 Thessalia 61 13.32 7.90 17.90 11.49 0.13 0.08 Ionia Nisia 59 4.34 5.19 21.17 29.44 0.16 0.19 Peloponnisos 57 3.98 14.02 5.93 27.37 0.05 0.16 Anat. Mak.-Thr. 55 22.70 62.51 40.25 111.11 0.33 0.90 Dytiki Ellada 51 11.21 9.91 15.13 15.95 0.13 0.12 Ipeiros 47 9.94 12.91 26.44 43.12 0.25 0.32

The table indicates that some regions experience considerable annual variability in the amount of state aid. The table also indicates that most investment incentives for regional development actually go to relatively richer regions which are economically more active.

Ireland The Scoreboard shows that in 2001, Ireland granted a total of EUR 1.3 billion in state aid (in 2000 prices). Agriculture, fisheries and transport absorbed almost half of that aid (46%). Regional aid accounted for 33% of total aid or 60% of non-farm and non-trasnport aid. This is a very high proportion of aid devoted to regional development. Of the 15 EU Member States, only Greece grants a higher proportion of regional aid.

Table 5.10 Ireland: Incentive schemes (grants and equity - EUR, million) Region GDP/head

(2000, PPS) Grants & equity

Grants & equity per

head Grants & equity as % of regional GDP

2000 2001 2000 2001 2000 2001 S&E 126 214.38 215.67 76.65 76.29 0.27 0.27 BM&W 84 96.22 65.56 96.03 64.79 0.51 0.35

Total 310.60 281.23 81.76 73.26 0.31 0.28

Until 1999, Ireland comprised a single NUTS II region. Since 2000 it has been divided into two NUTS II regions: the Southern & Eastern region with per capita income (PPS) at 126% of the EU average (2000) and the Border, Midland and Western region with per capita income at 84% of the EU average. Data on state aids collected from the agencies responsible for attracting investors are shown in table 5.10.


81

Italy The Commission’s Scoreboard indicates that in 2001 Italy granted €4.11 billion of aid (in 2000 prices) in sectors other than agriculture, transport and fisheries. The country rapporteur records a much higher amount for the same year, of €5.2 billion. The largest category of aid was for ‘reduction of territorial inequalities’ (46%). In this connection, it is worth noting that Germany and Italy accounted for 50% of all regional state aid granted in the EU in 2001.

The second and third largest categories of aid in Italy, other than for agriculture and transport, were to R&D and to investment, accounting for 24% and 10% of total aid to manufacturing and services, respectively. Italy has introduced ‘automatic’ incentives for R&D, investment and purchasing of new equipment. This raises the question whether all of these incentives can count as state aid. If they are state aid, then this may also explain the higher amount of aid recorded by the country rapporteur. Of the aid that was granted to business in 2001, 62% went to SMEs. However, in the south, SMEs received 67% of aid while in the centre and north they accounted for only 53% of aid. This suggests that in richer regions of Italy, a higher proportion of aid is absorbed by large companies.

Table 5.12 presents the amounts of aid to businesses per region in the years 2000 and 2001. As for other countries, larger amounts of aid are granted in the poor regions than in the richer regions. This is also indicated by the significance of the aid to the recipient regions. The proportion of aid in relation to non-farm valued-added was 0.20% in the centre and north while in the south it rose to 1.10%. However, and in common again with other countries, there is significant variation of the amount of aid within poor and within rich regions, with no close correlation between the income of the region and the amount of state aid within each group. There is also variation from year to year in the reported amount of aid granted in each region.

Table 5.12. Italy: Aid to business (EUR, million) Region GDP/head

(2000, PPS) Aid to business Business aid per

head Business aid as % of regional GDP

2000 2001 2001 2001 2000 2001 Trentino 136 11 8 11.68 8.42 0.04 0.03 Lombardia 135 309 541 33.92 59.01 0.11 0.20 Emilia-Romag. 129 180 214 44.97 53.12 0.15 0.18 Piemonte 120 196 230 45.75 53.54 0.17 0.20 Valle d’Aosta 123 1 2 8.33 16.49 0.03 0.06 Veneto 119 192 161 42.35 35.26 0.16 0.13 Friuli-Venezia 114 158 124 133.11 104.17 0.52 0.41 Toscana 114 147 144 41.49 40.49 0.16 0.16 Lazio 113 169 184 31.92 34.53 0.13 0.14 Liguria 108 102 173 63.00 107.06 0.26 0.44 Marche 102 75 68 51.12 46.05 0.22 0.20 Umbria 101 48 41 57.21 48.58 0.25 0.21 Abruzzo 84 143 123 111.81 95.70 0.59 0.51 Molise 79 49 40 149.85 122.27 0.84 0.69 Sardegna 76 218 246 132.44 149.13 0.78 0.88 Basilicata 73 101 147 167.22 243.15 1.01 1.47 Puglia 67 405 482 99.24 117.70 0.65 0.78 Campania 65 558 782 96.64 134.85 0.65 0.92 Sicilia 65 379 667 74.75 131.26 0.51 0.89 Calabria 62 224 382 109.80 187.01 0.78 1.33

Spain In Spain, state aid may be granted by public authorities at different levels of government. Unfortunately, the country rapporteur has not been able to find data on state aid granted by


82

each of Spain’s autonomous regions. The regional distribution of regional state aid is shown in the table below. Two things stand out from the table. First, for some regions the amount of aid varies considerably from year to year. Second, there is no strict correlation between the prosperity of each region and the amount of state aid.

Table 5.13. Spain: Regional aid (EUR, million, 2001, 2002)

Region GDP/head

(PPS, 2000)

Regional aid

Regional aid per head

Regional aid as % of regional

GDP

2001 2002 2001 2002 2001 2002 Madrid 110 Navarra 105 Pais Vasco 102 Catalunia 100 Balearic Ils 98 Rioja 91 Aragon 88 18.5 8.8 15.83 7.54 0.08 0.04 Cantabria 80 15.7 10.3 29.73 19.41 0.16 0.11 Valencia 79 33.6 34.9 8.32 8.43 0.05 0.05 Canary Ils 78 27.6 33.3 16.34 18.73 1.28 1.54 Castilla Leon 76 38.0 26.7 15.38 10.86 0.09 0.06 Asturias 71 36.2 15.3 34.38 14.58 0.21 0.09 Murcia 69 12.9 48.0 11.47 41.61 0.07 0.27 Ceuta&Melilla 68 0 0 0.00 0.00 0.00 0.00 Castilla Mancha 67 4.6 7.7 2.69 4.45 0.02 0.03 Galicia 65 18.0 31.7 6.63 11.60 0.05 0.08 Andalucia 61 57.7 42.0 7.97 5.72 0.06 0.04 Extremadura 53 5.8 9.8 5.40 9.06 0.05 0.08

United Kingdom Most of the relatively small amount of state aid in the UK is granted to agriculture, transport and fisheries (75% in 2001). Of the aid that is granted to sectors other than agriculture, fisheries and transport, most of it goes to horizontal objectives (90%) such as training, R&D and for SMEs. For the 2000-6 period, the UK has four 87(3)(a) areas at NUTS II level.

Table 5.14. UK: Regional selective assistance and enterprise grants (GBP, million; April 2001- March 2002) Region GDP/head

(PPS, 2000) RSA EG Regional aid

per head in euro (2002)

Regional aid as % of

regional GDP London 147 0.84 0.38 0.27 0.00 South East 111 3.71 0.58 0.85 0.00 East 104 0.71 0.22 0.27 0.00 East Midlands 94 6.07 1.23 2.78 0.01 West Midlands 92 10.57 1.39 3.58 0.02 South West 91 5.82 0.57 2.06 0.01 Yorkshire 88 6.19 1.46 2.42 0.01 North West 87 31.44 2.23 7.81 0.04 North East 77 32.65 2.16 21.50 0.12 Scotland 97 42.55 -- 13.33 0.06 Wales 81 52.16 -- 28.27 0.16 Northern Ireland 78 -- -- -- --


83

Northern Ireland has a special status as ‘exceptional region’ enjoying higher rates of aid intensity than other 87(3)(c) areas. The most important regional instrument is the Regional Selective Assistance (RSA). England also operates the Enterprise Grant (EG). The table below shows payments under these two schemes for fiscal year April 2001 to March 2002. As can be seen, there is a rough concentration of regional aid to areas with lower per capita income. However, the amounts of aid do not correlate strictly with the need of each area as indicated by its per capita income.

5.4 Commentary and key findings State aids could, in principle, be a potent policy instrument for advancing economic cohesion. However, the evidence from examination of the institutional structures and the data, together with the econometric analyses conducted, suggests that the impact of state aids on cohesion has, at best, been ambiguous. The main reason for this is that the bulk of state aid is granted for reasons other than promoting the development of poorer regions.

Key findings are: • There appears to be no overwhelming evidence that state aid either correlates with

reduced regional disparities or that Member States grant more of it when disparities grow larger.

• Although the overall amounts of state aid in the EU have recently declined, some regions have received larger amounts of aid.

• The amounts of state aid received by each region fluctuate considerably from year to year.

• Some types of state aid like R&D aid are inversely related to regional income, with the richer regions receiving larger amounts of such aid.

• By and large, most regional state aid goes to poorer regions.

• However, when examining only poorer regions, there appears to be no precise correspondence between regional income and either the overall amount of state aid or regional aid received by the poorer regions.

• In Member States with federal structures, regional authorities grant significant amounts of state aid. Since richer regions can afford to grant larger amounts of aid, the policy decisions of sub-national governments may also have a considerable impact on national cohesion.


84

State Aids: Data Annex This annex explains the sources of information used in this section and presents more detailed tables.

The figures on state aid in Tables 5.A.1, 5.16 and 5.17 have been drawn from the online country-specific tables in the Commission’s online state aid Scoreboard.30 These tables provide data categorised into different types of aid for the years 1997-2001 in € millions. The three categories of state aid used in the Tables (regional, horizontal and agricultural, fisheries and transport aid) do not sum up precisely to the total amount of state aid shown in the fourth column of the Tables because total aid also includes sectoral aid. More precisely aid to shipbuilding, coal, etc. is omitted.

All figures are expressed in 2000 prices, the base year for the calculations. The state aid data for the periods 1996-1994, 1993-1992 and 1991-1990 are extracted from the Sixth, Fifth and Fourth State Aid Surveys in the EU, respectively. The Surveys use three-year averages, because the Commission wanted to iron out the difference between payments and commitments and the differences caused by inflation. There are no state aid data available for Austria, Finland and Sweden prior to their accession to the EU in 1995.

To convert the state aid figures prior to 1997 into real 2000 prices, rates of inflation reported in the OECD main economic indicators were used. But for Table 5.17, Eurostat’s harmonised indices of consumer prices (HICPs) from the Eurostat statistical yearbook 2002 were used.

The Gross Domestic Product (GDP) of the Member States has been extracted from the same yearbook. Since they are expressed in current prices, real GDP has been calculated with the OECD deflator for Table 5.16 and with HICPs for Table 5.17. The aid per capita is calculated from the population numbers, also given in the Eurostat statistical yearbook 2002.

The regional income disparities were drawn from the Commission’s Second Progress Report on Economic and Social Cohesion 2003 (table 2).31 Denmark and Luxembourg have been omitted from our calculations because they do not have NUTS II regions.32 Ireland has also been omitted because data on Irish income disparities are only available from 1995 onwards.

The country-specific state aid figures are drawn from the reports of the country rapporteurs. The regional population numbers have been drawn from Eurostat’s main regional indicators. The regional GDP figures have been extracted from Eurostat’s statistics in focus publication.33 Regional GDP figures for 2001 are not yet available so we have used the 2000 GDP figures to calculate aid as % of regional GDP for 2001 and 2002. In the case of the UK, the amounts of state aid have been converted into euro according to the ECB’s average exchange reference rate in that period (1£ = 0,62€).

30 http://europa.eu.int/comm/competition/state_aid/scoreboard/ 31 Second progress report on economic and social cohesion, Brussels, 30.1.2003, COM(2003) 34 final 32 See: European Regional Statistics: Reference Guide, Luxembourg: Office for Official Publications of the European Communities 2003 33 Behrens, A. (2003), Regional Gross Domestic Product in the European Union 2000, Luxembourg: Eurostat


85

Table 5.A.1 State Aid in the EU

Country

State Aid in 2001 (euro million, 2000 prices) State Aid as percentage of GDP in 2000

Regional Horizontal Agr. + Trans. Total Regional Horizontal Agr. + Trans. Total A 108.2 403.7 1,509.5 2,056 0.05 0.15 0.77 0.99 B 273.8 495.2 2,551.9 3,330.3 0.11 0.21 1.04 1.37

DK 9.1 1,136.4 1,188.5 2,369 0.0 0.68 0.69 1.51 D 2,107.4 5,405.7 11,420.3 23,273.5 0.11 0.32 0.56 1.26 E 408.2 1,046.2 2,035.8 4,658.6 0.07 0.23 0.35 0.88

FIN 47.6 315.7 1,698.4 2,073.9 0.05 0.26 1.40 1.76 F 704.7 2,501.4 9,739.4 15,844.3 0.08 0.19 0.71 1.22

GR 419.2 42.9 840.4 1,305.7 0.48 0.02 0.56 1.07 I 684.8 3,252.7 7,878.1 11,998.6 0.12 0.24 0.66 1.06

IRL 429.9 117.1 595 1,300.7 0.48 0.06 0.62 1.33 L 10.1 18.7 240.3 273.3 0.09 0.10 1.20 1.41

NL 48 516.6 3,362.8 3,992.2 0.04 0.12 0.81 0.98 P 64.3 306.5 323.3 1,225.1 0.07 0.28 0.30 1.17 S 19.1 374.2 1,356.8 1,862.9 0.01 0.16 0.55 0.77

UK 526.4 1,874.5 7,910.8 10,550.1 0.05 0.13 0.21 0.41


86

Table 5.A.1 continued.

Country State Aid per capita in 2001 Regional Horizontal Agri. & Trans. Total

A 13.32 49.71 185.88 253.17 B 26.68 48.26 248.67 324.53

DK 1.70 212.45 222.19 442.89 D 25.64 65.77 138.94 283.16 E 10.34 26.49 51.55 117.97

FIN 9.19 60.93 327.81 400.29 F 11.84 42.03 163.63 266.20

GR 39.68 4.06 79.55 123.59 I 11.84 56.23 136.20 207.43

IRL 112.54 30.65 155.76 340.50 L 22.90 4240 544.90 619.50

NL 3.00 32.32 210.40 249.78 P 6.42 30.58 32.26 122.23 S 2.15 42.13 152.74 209.72

UK 8.80 31.33 132.33 176.33


87

Table 5.A.2 Income Disparity(t+1) = α + β(State Aid)(t) + ε (1990-2000) [per capita GDP at PPS vs real state aid as % of GDP & real state aid per capita]

Country Type of state aid

Intercept Independent Variable R-squared

Belgium Regional 0.796 258.440 0.547

10.964 0.787 0.313 Horizontal 50.217 -41.147 0.959

51.304 - 0.203 0.960 Agric. & Transport 59.920 -26.134 0.094

29.179 2.28E-02 0.003 Total 65.738 - 18.650 0.936

66.705 - 8.7E-02 0.870 France

Regional 31.384 - 29.423 0.207 30.718 - 0.100 0.149

Horizontal 26.256 10.244 0.401 26.964 2.94E-02 0.285

Agric. & Transport 27.419 1.598 0.037 27.498 6.72E-03 0.033

Total 26.457 1.510 0.106 26.643 6.06E-03 0.094

Germany (1991-2000) Regional 24.140 7.053 0.216

25.295 2.82E-02 0.115 Horizontal 33.777 - 32.639 0.039

38.477 - 0.219 0.073 Agric. & Transport 20.147 10.302 0.175

26.556 1.37E-02 0.010 Total 15.66 6.017 0.524

16.133 2.78E-02 0.342

Greece Regional 6.832 4.312 0.110

5.734 6.80E-02 0.327 Horizontal 10.146 - 3.492 0.821

10.085 - 3.6E-02 0.830 Agric. & Transport 7.176 3.426 0.255

7.351 2.45E-02 0.149 Total 14.397 - 4.172 0.722

13.707 - 3.8E-02 0.567

Italy Regional 28.699 - 3.343 0.216

29.284 - 0.024 0.418 Horizontal 30.613 - 11.590 0.213

32.329 - 0.090 0.557 Agric. & Transport 30.489 - 4.099 0.448

30.132 - 0.019 0.605 Total 30.204 - 1.711 0.351


88

30.618 - 0.010 0.598

Netherlands Regional 14.761 - 29.531 0.149

14.304 - 0.109 0.106 Horizontal 18.831 - 28.385 0.804

17.486 - 0.094 0.685 Agric. & Transport 6.765 11.957 0.389

7.568 0.0453 0.386 Total 16.068 - 3.125 0.038

14.830 - 0.0084 0.020

Portugal Regional 17.488 - 14.780 0.178

15.300 - 0.0159 0.002 Horizontal 16.930 - 6.462 0.065

15.015 0.0256 0.008 Agric. & Transport 16.142 - 1.106 0.007

13.578 0.0483 0.060 Total 13.826 1.469 0.151

13.217 0.0190 0.246

Spain Regional 11.263 93.673 0.477

11.435 0.672 0.606 Horizontal 15.091 10.296 0.318

15.349 0.0560 0.147 Agric. & Transport 18.916 - 4.113 0.692

18.931 - 0.032 0.617 Total 21.279 - 3.990 0.203

21.335 - 0.030 0.439

UK Regional 48.925 - 212.176 0.371

39.259 - 0.604 0.241 Horizontal 29.893 - 18.105 0.024

29.108 - 0.079 0.026 Agric. & Transport 26.464 4.791 0.001

18.118 0.159 0.100 Total 18.945 16.407 0.062

25.907 0.0116 0.004


89

Table 5.A.3 Income Disparity(t+1) = α + β(State Aid)(t) + ε (1999, 2000) [per capita GDP at PPS vs real state aid as % of GDP & real state aid per capita]

Type of state aid Intercept Independent Variable R-squared 2000

Regional 26.134 - 18.294 0.107 24.410 - 0.03 0.020

Horizontal 14.987 46.526 0.185 12.35 0.29 0.379

Agric. & Transport 20.100 4.919 0.041 18.94 0.03 0.112

Total 22.418 0.836 0.001 24.61 - 0.01 0.023 1999

Regional 24.946 - 9.584 0.059 23.71 - 0.02 0.006

Horizontal 16.639 37.896 0.162 14.83 0.24 0.289

Agric. & Transport 20.060 4.477 0.042 19.14 0.03 0.105

Total 20.667 1.980 0.007 15.39 0.03 0.145


90

Table 5.A.4 Austria: Federal state aid (EUR, million) Region GDP/head

(2000, PPS) Technology aid Technology aid

per head Technology aid as

% of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 104.48 85.83 64.93 54.18 0.18 0.15 Salzburg 131 22.73 17.26 43.97 34.07 0.15 0.11 Vorarlberg 118 36.48 32.37 104.53 94.51 0.39 0.35 Tirol 113 53.25 19.38 79.48 29.51 0.31 0.11 Oberösterreich 109 91.33 81.12 66.23 60.36 0.27 0.24 Niederösterreich 97 37.18 31.89 24.11 21.06 0.11 0.09 Kärnten 96 23.27 21.16 41.33 38.13 0.19 0.17 Steiermark 96 65.23 44.27 54.27 37.46 0.25 0.17 Burgenland 73 4.50 18.22 16.19 66.54 0.10 0.40 Total 438.51 351.51 54.1 44.16 0.21 0.17

Region GDP/head

(2000, PPS) Regional aid Regional aid per

head Regional aid as % of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 0 0 0.00 0.00 0.00 0.00 Salzburg 131 0.33 0.45 0.64 0.89 0.00 0.00 Vorarlberg 118 0 0 0.00 0.00 0.00 0.00 Tirol 113 5.17 3.21 7.72 4.89 0.03 0.02 Oberösterreich 109 3.21 24.64 2.33 18.33 0.01 0.07 Niederösterreich 97 12.72 34.09 8.25 22.51 0.04 0.10 Kärnten 96 13.79 28.32 24.49 51.04 0.11 0.23 Steiermark 96 24.86 22.24 20.68 18.82 0.10 0.09 Burgenland 73 25.24 49.96 90.79 182.47 0.55 1.09 Total 85.32 162.91 10.52 20.47 0.04 0.08

Region GDP/head

(2000, PPS) Environment aid Environment aid

per head Environment aid as % of regional

GDP 2000 2001 2000 2001 2000 2001 Wien 157 1.16 0.58 0.72 0.37 0.00 0.00 Salzburg 131 4.27 2.09 8.26 4.13 0.03 0.01 Vorarlberg 118 3.92 2.53 11.23 7.39 0.04 0.03 Tirol 113 6.44 5.15 9.61 7.84 0.04 0.03 Oberösterreich 109 10.41 8.52 7.55 6.34 0.03 0.02 Niederösterreich 97 11.98 17.56 7.77 11.59 0.04 0.05 Kärnten 96 2.35 3.01 4.17 5.42 0.02 0.02 Steiermark 96 18.24 12.49 15.17 10.57 0.07 0.05 Burgenland 73 0.79 0.32 2.84 1.17 0.02 0.01 Total 59.56 52.25 7.34 6.56 0.03 0.02


91


SME aid SME aid per head SME aid as % of regional GDP

2001 2001 2000 2001 2000 2001 Wien 157 30.41 27.20 16.90 17.17 0.05 0.05 Salzburg 131 15.93 13.78 26.65 27.20 0.10 0.09 Vorarlberg 118 10.44 7.56 21.66 22.07 0.11 0.08 Tirol 113 22.14 20.64 30.81 31.43 0.13 0.12 Oberösterreich 109 33.41 43.69 31.68 32.51 0.10 0.13 Niederösterreich 97 36.54 23.33 15.13 15.40 0.11 0.07 Kärnten 96 13.75 17.86 31.72 32.19 0.11 0.15 Steiermark 96 22.06 16.87 14.03 14.27 0.08 0.06 Burgenland 73 7.41 8.27 29.75 30.20 0.16 0.18 Total 192.09 179.20 23.69 22.51 0.09 0.09

Region GDP/head

(2000, PPS) Total aid Total aid per head Total aid as % of

regional GDP 2000 2001 2000 2001 2000 2001 Wien 157 132.84 113.61 82.56 71.71 0.24 0.20 Salzburg 131 41.11 33.58 79.52 66.29 0.28 0.22 Vorarlberg 118 47.96 42.46 137.42 123.97 0.55 0.46 Tirol 113 85.50 48.38 127.61 73.67 0.51 0.28 Oberösterreich 109 148.64 157.97 107.79 117.54 0.41 0.46 Niederösterreich 97 85.21 106.87 55.26 70.56 0.29 0.32 Kärnten 96 57.27 70.35 101.72 126.78 0.44 0.58 Steiermark 96 125.20 95.87 104.16 81.12 0.50 0.37 Burgenland 73 38.80 76.77 139.57 280.39 0.82 1.67 Total 770.08 748.28 95.62 93.71 0.37 0.36


92

Table 5.A.5 Austria: Länder state aid (EUR, million) Region GDP/head

(2000, PPS) Aid to energy Aid to energy per

head Aid to energy as

% of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 0 0 0.00 0.00 0.00 0.00 Salzburg 131 1.286 1.805 2.49 3.56 0.01 0.01 Vorarlberg 118 0 0 0.00 0.00 0.00 0.00 Tirol 113 0 0 0.00 0.00 0.00 0.00 Oberösterreich 109 4.146 3.823 3.01 2.84 0.01 0.01 Niederösterreich 97 0.103 0.401 0.07 0.26 0.00 0.00 Kärnten 96 5.54 6.647 9.84 11.98 0.05 0.05 Steiermark 96 1.805 1.811 1.50 1.53 0.01 0.01 Burgenland 73 0.766 0.973 2.76 3.55 0.02 0.02

Total 13.646 15.460 1.68 1.94 0.01 0.01 Region GDP/head

(2000, PPS) Aid to tourism Aid to tourism per

head Aid to tourism as

% of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 13.837 13.948 8.60 8.80 0.02 0.02 Salzburg 131 9.419 8.655 18.22 17.08 0.06 0.06 Vorarlberg 118 4.331 4.384 12.41 12.80 0.05 0.05 Tirol 113 41.075 45.738 61.31 69.65 0.24 0.27 Oberösterreich 109 18.008 19.307 13.06 14.37 0.05 0.06 Niederösterreich 97 19.369 20.202 12.56 13.34 0.06 0.06 Kärnten 96 1.581 0.246 2.81 0.44 0.01 0.00 Steiermark 96 10.816 19.844 9.00 16.79 0.04 0.08 Burgenland 73 15.493 15.652 55.73 57.17 0.34 0.34

Total 133.929 147.976 16.52 18.59 0.06 0.07 Region GDP/head

(2000, PPS) Aid to trade &

industry Aid to trade &

industry per head Aid to trade &

industry as % of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 126.341 94.072 78.52 59.38 0.22 0.16 Salzburg 131 15.289 10.917 29.57 21.55 0.10 0.07 Vorarlberg 118 11.667 13.768 33.43 40.20 0.13 0.15 Tirol 113 24.131 30.394 36.02 46.28 0.14 0.18 Oberösterreich 109 52.493 86.579 38.07 64.42 0.15 0.25 Niederösterreich 97 30.748 32.607 19.94 21.53 0.09 0.10 Kärnten 96 37.324 46.217 66.29 83.29 0.31 0.38 Steiermark 96 38.468 25.372 32.00 21.47 0.15 0.10 Burgenland 73 61.239 48.599 220.28 177.50 1.33 1.06

Total 397.700 388.525 49.04 48.81 0.19 0.19


93


Agricultural aid Agricultural aid per head

Agricultural aid as % of regional

GDP 2000 2001 2000 2001 2000 2001 Wien 157 1.364 2.612 0.85 1.65 0.00 0.00 Salzburg 131 41.78 44.235 80.81 87.32 0.27 0.29 Vorarlberg 118 27.492 28.426 78.77 83.00 0.30 0.31 Tirol 113 49.609 52.702 74.04 80.25 0.29 0.31 Oberösterreich 109 102.382 118.061 74.24 87.84 0.30 0.35 Niederösterreich 97 144.489 144.542 93.70 95.43 0.43 0.43 Kärnten 96 49.119 52.27 87.25 94.20 0.40 0.43 Steiermark 96 100.474 85.863 83.59 72.65 0.39 0.33 Burgenland 73 57.094 41.476 205.37 151.48 1.24 0.90

Total 573.803 570.187 70.76 71.64 0.27 0.27


Total aid Total aid per head Total aid as % of regional GDP

2000 2001 2000 2001 2000 2001 Wien 157 141.524 110.632 87.96 69.83 0.25 0.19 Salzburg 131 67.774 65.666 131.09 129.62 0.44 0.43 Vorarlberg 118 43.490 46.578 124.61 135.99 0.47 0.50 Tirol 113 131.226 147.634 195.86 224.81 0.77 0.86 Oberösterreich 109 177.029 227.770 128.37 169.47 0.52 0.67 Niederösterreich 97 194.709 197.752 126.27 130.56 0.58 0.59 Kärnten 96 93.564 / 105.380 166.19 189.91 0.77 0.86 Steiermark 96 151.563 132.890 126.09 112.44 0.58 0.51 Burgenland 73 134.592 106.700 484.14 389.70 2.92 2.32

Total 1135.471 1141.002 140.03 143.35 0.54 0.54


94

Table 5.A.6 Austria: Municipal state aid (EUR, million) Region GDP/head

(2000, PPS) Agricultural aid Agricultural aid

per head Agricultural aid as % or regional

GDP 2000 2001 2000 2001 2000 2001 Salzburg 131 2.962 2.655 5.73 5.24 0.02 0.02 Vorarlberg 118 2.486 3.272 7.12 9.55 0.03 0.04 Tirol 113 3.996 3.852 5.96 5.87 0.02 0.02 Oberösterreich 109 4.682 4.961 3.40 3.69 0.01 0.01 Niederösterreich 97 12.060 14.774 7.82 9.75 0.04 0.04 Kärnten 96 5.006 6.693 8.89 12.06 0.04 0.05 Steiermark 96 22.587 19.632 18.79 16.61 0.09 0.08 Burgenland 73 7.836 8.300 28.19 30.31 0.17 0.18 Region GDP/head

(2000, PPS) Energy aid Energy aid per

head Energy aid as % of regional GDP

2000 2001 2000 2001 2000 2001 Salzburg 131 0.044 0.049 0.09 0.10 0.00 0.00 Vorarlberg 118 0.257 0.040 0.74 0.12 0.00 0.00 Tirol 113 0.043 0.035 0.06 0.05 0.00 0.00 Oberösterreich 109 0.074 0.107 0.05 0.08 0.00 0.00 Niederösterreich 97 1.470 1.374 0.95 0.91 0.00 0.00 Kärnten 96 0.140 0.108 0.25 0.19 0.00 0.00 Steiermark 96 0.092 0.110 0.08 0.09 0.00 0.00 Burgenland 73 0.072 0.090 0.26 0.33 0.00 0.00 Region GDP/head

(2000, PPS) Tourism promotion

aid Tourism

promotion aid per head

Tourism promotion aid as

% of regional GDP

2000 2001 2000 2001 2000 2001 Salzburg 131 20.504 16.816 39.66 33.19 0.13 0.11 Vorarlberg 118 19.362 22.411 55.48 65.43 0.21 0.24 Tirol 113 11.117 9.146 16.59 13.93 0.07 0.05 Oberösterreich 109 9.165 6.957 6.65 5.18 0.03 0.02 Niederösterreich 97 15.660 15.581 10.16 10.29 0.05 0.05 Kärnten 96 22.545 24.260 40.04 43.72 0.18 0.20 Steiermark 96 18.723 19.300 15.58 16.33 0.07 0.07 Burgenland 73 2.123 2.166 7.64 7.91 0.05 0.05 Region GDP/head

(2000, PPS) Trade & Industry aid Trade & Industry

aid per head Trade & Industry

aid as % of regional GDP

2000 2001 2000 2001 2000 2001 Salzburg 131 4.175 4.018 8.08 7.93 0.03 0.03 Vorarlberg 118 2.650 3.412 7.59 9.96 0.03 0.04 Tirol 113 7.629 9.658 11.39 14.71 0.04 0.06 Oberösterreich 109 25.816 22.489 18.72 16.73 0.08 0.07 Niederösterreich 97 31.015 17.683 20.11 11.68 0.09 0.05 Kärnten 96 12.921 15.613 22.95 28.14 0.11 0.13 Steiermark 96 27.488 38.417 22.87 32.50 0.11 0.15 Burgenland 73 3.146 3.390 11.32 12.38 0.07 0.07


95


Total economic development aid

Total economic development aid

per head

Total economic development aid as % of regional

GDP 2000 2001 2000 2001 2000 2001 Salzburg 131 27.685 23.538 53.55 46.46 0.18 0.15 Vorarlberg 118 24.755 29.135 70.93 85.07 0.27 0.31 Tirol 113 22.785 22.691 34.01 34.55 0.13 0.13 Oberösterreich 109 39.737 34.514 28.82 25.68 0.12 0.10 Niederösterreich 97 60.205 49.412 39.04 32.62 0.18 0.15 Kärnten 96 40.612 46.674 72.13 84.11 0.33 0.38 Steiermark 96 68.890 77.459 57.31 65.54 0.26 0.30 Burgenland 73 13.177 13.946 47.40 50.93 0.29 0.30


96

6. Territorial policies Spatially targeted economic development policies are widely used in the Member States and generally have the aim of raising the competitiveness of assisted economies. These ‘territorial policies’ - that is, policies that are aimed at geographically defined targets - comprise regional policies, local policies and urban and rural policies, and use a range of policy instruments. Two common characteristics of national policies (see, also, the summary in section 2.3 of this report) are that these policies are typically relatively small in scale, especially for some of the very specific measures, and often have conflicting objectives. For example, an urban regeneration policy funded by central government may, quite legitimately, steer resources to a deprived district of a city in a rich region, yet in so doing conflict with regional policy that seeks to improve the position of poor regions.

In many Member States, spatially targeted policies have undergone considerable change in recent years. One of the few common characteristics of national territorial policies across Member States is expenditure on regional incentives, although evidence suggests that they have been declining over the years34. Annex 2 to this report presents the most recent overview by EPRC of regional incentives, showing both magnitudes and changes in the philosophy of policy. The EPRC assessment also shows the diversity of policy instruments for granting regional incentives. As noted in section 4, this type of regional assistance is designed to stimulate investment in areas normally covered by Article 87(3)(a) and Article 87(3)(c) status, with the objective of safeguarding or expanding employment and economic activity. Partly due to their size, Germany and Italy account for more than half of regional aid in the EU (see Figure 6.1).

Figure 6.1 Average annual regional aid to Article 87(3)(a) areas by Member State, 1997-01

Greece Spain

Portugal

Ireland

Germany

UK

Italy

Austria

France

Source: State Aid Scoreboard, DG COMP

Just under a quarter is accounted for by the four cohesion countries, with the rest split among the other Member States. This is also reflected in the share of regional aid to GDP (Figure 6.2). Three of the four cohesion countries have relatively high shares of regional expenditure, 34 See successive issues of European Regional Incentives by the team at the EPRC, University of Strathclyde.


97

followed by Germany and Italy, with Spain lagging somewhat behind. Sweden, today, has little more than assistance to new business start-up as a form of incentive: the scheme is only open to the unemployed in richer, urban areas, but to all elsewhere. Companies – especially those in difficulty - can apply for support, but it is only granted on an ad hoc, rather than regional basis: that is non-regional state aid.

Figure 6.2. Average annual GDP share of Article 87(3)(a) aid by Member State, 1997-2001

0.02%

0.03%

0.04%

0.12%

0.27%

0.27%

0.60%

0.85%

0.89%

Austria

UK

France

Spain

Italy

Germany

Greece

Portugal

Ireland

Source: State Aid Scoreboard, DG COMP

Data collected in the course of this study35 show that the amounts directed to regional state aid are not particularly significant, measured as shares of regional GDP. Thus, it can be tentatively argued that the impact of such policies on regional cohesion is limited. On the other hand, the regional allocation of aid generally favours poorer areas. This is particularly so in countries with more centralised structures, whereas in countries with greater devolution of financing, the pro-cohesive transfers of public money occur upstream of the choice of development policies. Three factors seem to shape the nature and scale of regional aid policies: the extent of internal disparities (especially Germany and Italy), the demands for co-financing of Structural Funds programmes, and the identification (especially in richer Member States) of specific, usually narrowly drawn, problem areas.

Many of the impacts of regional aids have already been covered in the preceding section on state aids. An overview of the aid intensities at regional level is provided in table 6.1 and in what follows, the focus is largely on institutional aspects.

6.1 Decentralised policies: Austria, Belgium and Denmark Regional policies in these countries are overseen by the central governments, but sub-national governments are responsible for the planning and financing of spatial policies. In this way, spending is mainly tied to the overall financing ability of each provincial government, such that disparities between regions (provinces, counties or municipalities) are not necessarily central aims. In Austria, priorities are the provision of physical and social infrastructure, the engagement of local actors in the implementation of policies, and the development of

35 The data are not directly comparable across countries, so that summary figures cannot be produced – see table 6.A.1 appended to this chapter.


98

technology and agglomeration economies. There is some emphasis on territorial cohesion (assisting areas like Burgenland and Carinthia), although the limited magnitude of spending (the main regional aid instruments – ERP and RIF-2000 account for 0.1% of national GDP) limits the possible effectiveness of such policies.

In Belgium, regional policy is less prominent, as it forms part of the Economic Expansion Laws. The main focus is on assisted areas, with emphasis on job creation in large employers in Wallonia and leading technology (‘strategic’) investment in Flanders. Although regional aid follows the regional distribution of EU spending in the country, co-financing seems to generate a system where more money is directed into the richer parts of the country. In Denmark, the design and implementation of some territorial policies is even more devolved, at the municipality level, which is also the level at which resources are drawn.

6.2 Central funding, but devolved implementation: Finland, Greece, and Spain Regional policies are the responsibility of the central governments, who also provide the vast majority of funding. Most policies have a wide spatial coverage, although there is an explicit bias towards policies addressing specific problems of more backward regions (a zoning system operates in all countries, which determines maximum amounts of regional aid, although practically these zones cover the whole of each country, so that policy can be seen as promoting development throughout). Smaller-scale policies (such as urban regeneration) are not so prominent as in other Member States, given the prioritising towards regional development at large. Tohmo et al. (2001) investigated the effects of incentives to individuals and firms in slowing down migration from sparsely populated areas in Finland. They found that such support is comparatively efficient and the effect at the regional level in strong. The conclusion of the research is that individual support incentives have potential but many unsolved problems still remain.

Finland has a series of instruments outside regional aid (POMO+, SEUTU, OSKU), which focus on the modernisation of the rural economies and the creation of clusters within regions, that are expected to act as growth poles for larger areas. Although such policies target the poorer regions and smaller areas, their small scale (around 0.02% of national GDP) restricts their pro-cohesive effects, although their impact on the targeted areas within a region can be significantly larger. On the other hand, regional aid, which accounts for 0.3% of regional GDP follows a less pro-cohesive pattern, with the more developed parts of the country managing to attract disproportionately large amounts of aid. Rosenquist (2002) shows that an integrated rural policy is hindered by the competing demands of regional or urban development discourses, agricultural and forest-sector discourses, and notes that rural policy remains separated from regional and sector policy.

Despite having a similarly centralised structure, regional policy in Greece is more skewed towards weaker regions, thus promoting economic cohesion. The main beneficiaries of national regional policy seem to be regions like Eastern Macedonia and Thrace, Ipeiros, and the Aegean islands, although some smaller scale policies (with a focus on small-area support and intra-regional disparities) seem to direct money to richer regions (e.g., Thessaly). Regional aid is at 0.18% of GDP, while other territorial policies reach around 0.28% (not including co-financing of CSF projects). Apart from the general objective of assisting regional development, emphasis is given to urban development and improvements in the physical and social infrastructure of isolated areas (e.g., islands).

Spanish regional policy is also very similar to EU regional policy, with a strong emphasis on support for the poorer regions. The focus is on regional convergence and economic restructuring, but there is relatively limited attention to intra-regional inequalities (with the notable exception of the Teruel Fund). The scale of direct regional incentives is rather small (0.045% of GDP), but together all regional policies are much more significant (almost 0.2% of GDP).


99

6.3 Complementing Community policy: Ireland and Portugal Explicit territorial policies are not a significant part of national policies in these countries, so that the main instrument that steers resources to the regions seems to be regional aid. It is particularly the case in Ireland, where policies are effectively limited to the allocation of EU structural funds and state aid. In general, policies are centrally administered and cover practically the whole of the country. The allocation of regional aid (at 0.23% of GDP) has a bias towards the less developed areas, but aid is quantitatively much larger in the more developed areas, reflecting the disparities in GDP levels. Regional support reproduces the patterns of EU spending and there seem to be no clear local-specific or urban/rural programmes. Portugal has a Territorial Improvement Programme, which aims at reducing regional asymmetries and concentration in metropolitan areas. The programme is centrally financed and administered but the focus of policies is at the sub-regional and metropolitan level. The main areas that benefit are Madeira, Alentejo, and the Algarve.

6.4 Dealing with the dual economy: Italy and Germany These two countries account for around half of all national spending on territorial policies in the EU, partly because of the significant duality that exists between their regions (North-South in Italy; East-West in Germany). Territorial policies in Italy are administered and funded by the central government and focus largely on capacity building through public investment instead of incentives to businesses, as in the past. Although regional incentives to companies still go disproportionately to the South, the material assembled in the course of the present study shows that the public investment programmes towards which Italian policy has shifted often favour regions in the North. Therefore, there seems to be a conflict between the main national territorial policy on the one hand and EU policies and the distribution of regional aid in the country. The latter is much more pro-cohesive, directing resources (0.47% of GDP) mainly to the southern regions (where aid exceeds 1% of GDP) and a couple of regions in the centre-north (Venezia, Liguria). The regions receiving less aid, in relative terms, are Trentino and Aosta. Small-area policies, which involve local actors and aim at generating local partnerships and networks, are also centrally designed, but early evidence suggests they have been less successful than expected, as is the case with the effectiveness of the ‘DEP’ programme.

In Germany, the federal government has a strategic influence on territorial policies, but most of these policies are directly designed and implemented at the Land level. The main instruments of regional policy are organised around a number of wide-coverage programmes (the ‘Common Tasks’), which are co-financed (50%) by the federal and Länder governments, but the latter have the discretion to implement specific projects, based on their specific needs and priorities. The same structure is reproduced for smaller-area and urban policies (which account for 0.02% of GDP), while the central administration has the responsibility for the Unification Fund programme, which directs financial transfers to the East German Länder. The main focus of the Common Tasks is on economic development in the east, economic restructuring in the regions with structural problems, and support of agriculture in the rural regions. Regional aid accounts for 0.11% of national GDP and generally favours the Länder of Brandenburg, Saxony Anhalt, Saxony, and Mecklenburg and Western Pomerania, while the Länder receiving the least support are Rhineland and Palatinate, Bremen, Bavaria, Hamburg, and Saarland. However, Bremen and Saarland benefit from special industrial restructuring programmes, highlighting the importance of looking at the full range of policies.

Overall, the national territorial policies are compatible with EU regional policy and some targeted policies even seem to support cohesion even more strongly. However, a significant limitation to the effectiveness of German territorial policy is related to co-financing (so that richer regions attract higher funding), while problems seem to exist also with the co-ordination of policies, which sometimes seem to lead to direct inter-regional competition (e.g., for the attraction of investment).


100

6.5 France Much of regional policy in France is related to co-financing of Community policies, with national funds providing almost twice as much as support from the Structural Funds.There is also an extensive territorial policy around ‘Planning Contracts’ (CPER) which the central government agrees with the regional governments. Direct regional aid is rather limited (1.3 euro per capita – less than 0.01% of GDP) and even in the most heavily supported regions (Basse-Normandie, Nord-Pas-de-Calais) does not exceed 0.02% of regional GDP. Generally speaking, national resources, mainly outside regional aid, are directed disproportionately to regions not heavily supported by Community policies, but complement rather than conflict with the latter. Thus, the CPER have the same time-span and share five out of six stated objectives. The main focus of CPER is on enhancing the attractiveness and competitiveness of the regions. Smaller-scale policies are delivered within this scheme, but because of their low level of funding they play only a limited part in French territorial policies.

6.6 Netherlands The Netherlands, like France, has an important, spatial planning programme administered by central government, which addresses regional issues on the basis of the perceived needs of the country as a whole. There is a moderate expenditure on territorial policies with the coverage emphasising the three northern regions (Groningen, Friesland, and Drenthe) and 30 urban areas. The main instruments of policy that have a specifically territorial focus are the Kompas programme for the North, the IPR, and the Policy for Large Cities. Regional spending is based on co-financing, with the regions’ contribution at around 30%, and there is also an expectation that substantial private funding will be levered-in. The principal thrust of policy is on fostering development of small areas with specific economic problems (mainly urban areas with high unemployment), so that regional and urban policy is largely integrated into other national policy schemes (notably, active labour market policies). The policy programmes bring together labour market and other measures as well as direct regional incentives, so that in philosophy, national territorial policy is largely consistent with (and, indeed operates alongside) Community policy.

In the Kompas programme, labour market measures are the largest single element, accounting for about a quarter of spending, whereas incentives for investment are just 12% (€ 370 million over the period 2000-06). A slightly smaller amount for investment incentives is provided through the IPR programme. The GSB urban programme is more substantial, attaining some €3.3 billion in 2002, equivalent to 8% of total spending by the designated large cities. There are nine policy-themes: labour market, economic competition, education, residential use, environment, infrastructure, social relations, safety, deprived areas. GSB can be seen as a territorial policy, but to a large extent these instruments do no more than regroup target grants.

6.7 UK Despite recent initiatives to devolve economic development policy, the design and funding of long-term goals and priorities has remained with the central government. Current policy goals include achieving regional convergence in trend growth rates, recognising that some regional imbalances (e.g., the deviation of the London/SE economy from the rest of the country) are structural and may not necessarily require policy intervention. There are also emphases on the creation of clusters as a means to enhance regional and business competitiveness, and on urban regeneration. Under these general priorities, specific territorial policies are designed and delivered at the regional and sub-regional levels. At this level policy falls largely under the responsibility of the ‘Regional Development Agencies’. Given this devolved structure of regional policy, spending on local-specific programmes is significantly larger than Community support and regional incentives: the annual budget for the Regional Coordination Unit which distributes resources to the RDAs, is over 13 billion euros, while direct expenditure on Regional Selective Assistance incentives is 0.75 billion euros. The latter amounts to 0.05% of national GDP, varying between 0.001% in the South East and Greater


101

London to 0.18% in Scotland (the figure is much higher for N. Ireland, but due to its special status, the definition of aid for the Province is not directly comparable). Thus, regional aid significantly favours economic cohesion, allocating resources to the poorer regions. Overall, although the system allows co-ordination at the national level, the focus of policy is on smaller areas. As a result, it remains possible for regional disparities to widen, because the effect of policy on entire regions is uncertain. Nevertheless, this structure seems rather effective in dealing with specific competitive weaknesses of problem areas within regions.

6.8 Commentary and key findings There is a wide variability in the content, targets, and structure of national territorial policies. In some Member States, territorial policies simply reproduce the patterns of Community cohesion policies. In others, territorial policies may conflict with EU spending. The pro-cohesive impact of the national regional policies is also quite variable.

The degree to which national regional policies effectively redistribute resources for economic development and, by so doing, implicitly contribute to economic cohesion, is also quite variable. Figure 6.3 presents a picture of the extent of redistribution achieved by spending on territorial aid policies by the Member-States.

Figure 6.3 Redistribution capacity of national regional policies (weighted range)

0 1 2 3 4 5 6

Ireland (Regional Aid)

Belgium (Regional spending)

Finland (Regional Aid)

Finland (Regional policies)

France (Regional Aid)

Greece (EPTA, etc)

Germany (Urban policy)

Italy (Regional Aid)

Greece (Regional Aid)

Germany (Regional Aid)

Austria (Regional Aid)

Spain (Regional Aid)

UK (Regional Aid)

The measure used looks at the GDP share of aid received by each region within each country and calculates the distance between the most and least supported region, standardised by the average GDP share of aid in the country (across regions).36 The information in Figure 6.3 can be read as follows: for Spain, the amount of support received by the most assisted region is

36 In technical terms, the measure is constructed as a simple standardised range of regional GDP shares of aid:

−=}{

}{}{Reii

iiii

XMeanXMinXMaxbiasondistributi where

i

ii GDP

AIDX =

and i indexes regions. Other measures (eg., coefficient of variation), gave almost identical results. Moreover, the results were identical also across indicators (for example, when the distribution of levels of spending, or of per capita spending, was used instead of the GDP shares).


102

over five times larger than that received by the average region in the country; for Italy, this amount is less than three times larger, implying that the redistributive impact of regional aid in Italy is lower than that in Spain.

Overall, as can be seen, Member States achieve very different levels of redistribution of resources for economic development across their regions. Spain, Austria, Germany and the UK seem to have a much more progressive system of redistribution, compared to France or Finland (results for Belgium and Ireland cannot reliably be compared due to small sample size), while Italy and Greece are close to the average level. Of course, the picture observed in terms of regional redistribution does not necessarily account for the differences in the main emphasis of territorial policies in each Member State. Thus, as the country analyses suggest, some countries focus more on local-specific problems, others focus on convergence, while a third group puts more emphasis on regional development across the board.

Largely such differences seem to reflect differences in the needs and circumstances of each Member State. The absence of a harmonised environment for the design and implementation of national territorial policies within the EU might be considered to be generating some long-run inefficiencies at the EU level. It is, however, consistent not just with subsidiarity, but also with a multi-level response to the cohesion challenges confronting the Member States in economic development in which national and sub-national programmes complement Community policy.

Key findings are: There is a wide range of territorial policies, but for the most part they are relatively small

in scale and not easily comparable across countries. In addition, the focus and instruments of policy have changed considerably in recent years.

The impact on economic cohesion is influenced by the institutional structure and the form and extent of disparities within the Member State. For the cohesion countries, especially, the nature of policy is strongly influenced by the procedures and priorities of the Structural Funds.

In countries where territorial policy is devolved (as in Austria), giving sub-national government the main responsibility, the impact tends to be more on intra-regional cohesion than inter-regional.

Similarly, if there are large inter-regional differences as in Italy, it is more likely that national policy will explicitly address these disparities.

Spatially targeted economic development policies have an impact on territorial cohesion by focusing on specific spatial imbalances, but the priorities vary greatly from Member State to Member State.


103

Table 6.A.1 Regional aid and regional policies in the EU Member States

country Nuts region % of GDP € per capita % of total 000s € Austria (Regional Aid) 0.093% 23.8 100.00% 193,222 AT11 Burgenland 0.626% 102.4 15.00% 28,463 AT21 Carinthia 0.279% 59.7 17.00% 33,648 AT12 Lower Austria 0.130% 28.1 22.00% 43,326 AT31 Upper Austria 0.140% 34.2 24.00% 47,168 AT32 Salzburg 0.001% 0.3 0.00% 168 AT22 Styria 0.141% 30.2 19.00% 36,343 AT33 Tyrol 0.024% 6.1 2.00% 4,106 AT34 Vorarlberg 0.000% 0.0 0.00% 0 AT13 Vienna 0.000% 0.0 0.00% 0Belgium (Regional spending) n/a n/a 100.00% n/a BE2 Flanders 0.218% 52.0 n/a 309,000 BE3 Wallonia 0.254% 44.6 n/a 149,000 BE1 Brussels n/a n/a n/a n/a Finland (Regional Aid) 0.265% 67.1 100.00% 345,503 FI16 Uusimaa 0.235% 81.9 32.88% 113,617 FI17 South 0.276% 64.7 34.04% 117,601 FI13 East 0.334% 60.8 12.02% 41,530 FI14 Mid-Finland 0.208% 42.7 8.71% 30,110 FI15 North 0.343% 76.6 12.34% 42,645 Finland (Regional policies) 0.018% 4.6 100.00% 23,918 FI16 Uusimaa 0.007% 2.6 32.88% 3,583 FI17 South 0.020% 4.6 34.04% 8,378 FI13 East 0.048% 8.8 12.02% 6,003 FI14 Mid-Finland 0.024% 5.0 8.71% 3,498 FI15 North 0.020% 4.4 12.34% 2,456 France (Regional Aid) 0.005% 1.3 100.00% 75,350 FR61 Aquitaine 0.005% 1.0 3.81% 2,870 FR42 Alsace 0.011% 2.6 5.95% 4,480 FR72 Auvergne 0.014% 2.8 4.82% 3,630 FR25 Basse-Normandie 0.018% 3.6 6.76% 5,090 FR26 Bourgogne 0.004% 0.8 1.73% 1,300 FR52 Bretagne 0.001% 0.1 0.57% 433 FR24 Centre 0.002% 0.4 1.29% 970 FR21 Champagne-Ardenne 0.002% 0.3 0.60% 450 FR83 Corse n/a n/a n/a n/a FR43 Franche-Comté 0.014% 2.9 4.35% 3,280 FR23 Haute-Normandie 0.007% 1.5 3.52% 2,650 FR1 Ile-de-France 0.000% 0.0 0.00% 0 FR81 Languedoc-Roussillon 0.011% 2.0 6.03% 4,540 FR63 Limousin 0.011% 2.1 1.95% 1,470 FR41 Lorraine 0.010% 1.9 5.72% 4,310 FR62 Midi-Pyrénées 0.009% 1.8 6.00% 4,520 FR3 Nord-Pas-de-Calais 0.015% 2.8 15.16% 11,420 FR82 PACA 0.004% 0.8 5.03% 3,790


104

FR51 Pays de Loire 0.012% 2.4 10.56% 7,960 FR22 Picardie 0.004% 0.7 1.75% 1,320 FR53 Poitou-Charentes 0.010% 1.9 4.22% 3,180 FR71 Rhône-Alpes 0.003% 0.7 5.37% 4,050 Germany (Regional Aid) 0.111% 27.3 100.00% 2,244,398 DE1 Baden Württemberg n/a n/a n/a n/a DE2 Bavaria 0.006% 1.8 0.96% 21,467 DE3 Berlin 0.215% 47.8 7.20% 161,700 DE4 Brandenburg 0.799% 128.7 14.91% 334,733 DE5 Bremen 0.069% 22.9 0.67% 15,133 DE6 Hamburg n/a n/a n/a n/a DE7 Hesse 0.016% 4.9 1.31% 29,500 DE8 Mecklenburg West.Pomerania 1.041% 167.5 13.31% 298,700 DE9 Lower Saxony 0.059% 13.2 4.64% 104,100 DEA North Rhine Westphalia 0.018% 4.6 3.70% 82,933 DEB Rhineland Palatinate 0.010% 2.2 0.40% 9,067 DEC Saarland 0.052% 11.7 0.56% 12,567 DED Saxony 0.878% 143.5 28.40% 637,300 DEE Saxony Anhalt 1.226% 194.6 22.83% 512,467 DEF Schleswig Holstein 0.039% 8.7 1.08% 24,300 DEG Thuringia 0.001% 0.2 0.02% 431 Germany (Urban policy) 0.018% 4.3 100.00% 357,507 DE1 Baden Württemberg 0.005% 1.3 3.78% 13,520 DE2 Bavaria 0.005% 1.3 4.46% 15,937 DE3 Berlin 0.036% 8.1 7.67% 27,416 DE4 Brandenburg 0.105% 16.9 12.32% 44,027 DE5 Bremen 0.006% 1.9 0.35% 1,240 DE6 Hamburg 0.004% 1.5 0.72% 2,557 DE7 Hesse 0.005% 1.4 2.31% 8,265 DE8 Mecklenburg West.Pomerania 0.106% 17.0 8.50% 30,387 DE9 Lower Saxony 0.006% 1.3 2.97% 10,607 DEA North Rhine Westphalia 0.005% 1.3 6.76% 24,180 DEB Rhineland Palatinate 0.006% 1.4 1.61% 5,757 DEC Saarland 0.007% 1.5 0.44% 1,568 DED Saxony 0.110% 18.0 22.32% 79,779 DEE Saxony Anhalt 0.111% 17.6 12.94% 46,256 DEF Schleswig Holstein 0.007% 1.5 1.20% 4,297 DEG Thuringia 0.106% 17.1 11.67% 41,714 Greece (Regional Aid) 0.183% 21.3 100.00% 225,141 GR11 Anatoliki Makedonia, Thraki 0.723% 68.0 17.04% 38,373 GR12 Kentriki Makedonia 0.119% 13.9 11.21% 25,232 GR13 Dytiki Makedonia 0.110% 12.7 1.72% 3,867 GR14 Thessalia 0.290% 30.6 10.12% 22,792 GR21 Ipeiros 0.472% 38.3 6.39% 14,388 GR22 Ionia Nisia 0.146% 14.8 1.35% 3,048 GR23 Dytiki Ellada 0.302% 26.6 8.76% 19,721 GR24 Sterea Ellada 0.049% 6.4 1.90% 4,273 GR25 Peloponnisos 0.131% 13.0 3.89% 8,752


105

GR3 Attiki 0.031% 4.2 6.42% 14,457 GR41 Voreio Aigaio 0.078% 8.9 0.72% 1,623 GR42 Notio Aigaio 0.251% 34.5 4.19% 9,428 GR43 Kriti 0.322% 36.6 9.21% 20,728 Greece (EPTA, AKRITAS & POLITEIA) 0.284% 33.1 100.00% 349,551 GR11 Anatoliki Makedonia, Thraki 0.535% 50.3 8.12% 28,395 GR12 Kentriki Makedonia 0.238% 27.9 14.44% 50,460 GR13 Dytiki Makedonia 0.456% 52.6 4.58% 16,013 GR14 Thessalia 0.277% 29.3 6.25% 21,839 GR21 Ipeiros 1.203% 97.6 10.50% 36,691 GR22 Ionia Nisia 0.520% 53.0 3.12% 10,889 GR23 Dytiki Ellada 0.490% 43.1 9.13% 31,928 GR24 Sterea Ellada 0.299% 39.3 7.47% 26,100 GR25 Peloponnisos 0.466% 46.3 8.90% 31,094 GR3 Attiki 0.069% 9.2 9.08% 31,741 GR41 Voreio Aigaio 0.813% 91.9 4.82% 16,834 GR42 Notio Aigaio 0.450% 61.8 4.83% 16,899 GR43 Kriti 0.476% 54.2 8.77% 30,668 Ireland (Regional Aid) 0.228% 61.8 100.00% 234,655 IE01 BMW 0.245% 48.2 20.60% 48,346 IE02 S+E 0.224% 66.6 79.40% 186,309 Italy (Regional Aid) 0.408% 82.4 100.00% 4,757,800 IT11 Piemonte 0.227% 53.7 4.84% 230,100 IT12 Valle d'Aosta 0.055% 13.3 0.03% 1,600 IT2 Lombardia 0.223% 59.3 11.36% 540,600 IT31 Trentino Alto Adige 0.031% 8.4 0.17% 7,900 IT32 Veneto 0.151% 35.6 3.39% 161,200 IT33 Friuli-Venezia-Giulia 0.464% 104.7 2.61% 124,300 IT13 Liguria 0.501% 106.9 3.64% 173,100 IT4 Emilia-Romagna 0.209% 53.3 4.49% 213,500 IT51 Toscana 0.181% 40.6 3.02% 143,700 IT52 Umbria 0.248% 49.3 0.87% 41,400 IT53 Marche 0.228% 46.0 1.42% 67,500 IT6 Lazio 0.156% 34.8 3.87% 184,000 IT71 Abruzzo 0.579% 95.7 2.57% 122,500 IT72 Molise 0.794% 123.7 0.85% 40,400 IT8 Campania 1.049% 135.4 16.43% 781,700 IT91 Puglia 0.890% 118.1 10.13% 481,800 IT92 Basilicata 1.682% 244.1 3.10% 147,400 IT93 Calabria 1.525% 187.3 8.03% 382,200 ITA Sicilia 1.017% 131.5 14.01% 666,800 ITB Sardegna 1.002% 149.5 5.17% 246,100 Spain (Regional Aid) 0.044% 6.7 100.00% 268,500 ES61 Andalucía 0.051% 5.8 15.64% 42,000 ES24 Aragón 0.046% 7.5 3.28% 8,800 ES12 Principado De Asturias 0.110% 14.5 5.70% 15,300 ES53 Islas Baleares 0.000% 0.0 0.00% 0 ES7 Canarias 0.137% 19.7 12.40% 33300


106

ES13 Cantabria 0.131% 19.5 3.84% 10300 ES41 Castilla Y León 0.077% 10.8 9.94% 26700 ES42 Castilla-La Mancha 0.036% 4.5 2.87% 7700 ES51 Cataluña 0.000% 0.0 0.00% 0 ES63 Ceuta Y Melilla 0.000% 0.0 0.00% 0 ES43 Extremadura 0.093% 9.1 3.65% 9800 ES11 Galicia 0.097% 11.7 11.81% 31700 ES3 Comunidad De Madrid 0.000% 0.0 0.00% 0 ES62 Región De Murcia 0.335% 42.7 17.88% 48000 ES22 Comunidad Foral De Navarra 0.000% 0.0 0.00% 0 ES21 Pais Vasco 0.000% 0.0 0.00% 0 ES23 La Rioja 0.000% 0.0 0.00% 0 ES52 Comunidad Valenciana 0.059% 8.6 13.00% 34900UK (Regional Aid) 0.048% 12.5 100.00% 747,910 UKC North 0.072% 15.7 5.00% 37,410 UKD North West 0.027% 6.0 5.72% 42,775 UKE Yorkshire & the Humber 0.011% 2.6 1.76% 13,195 UKF East Midlands 0.008% 2.0 1.12% 8,410 UKG West Midlands 0.040% 9.7 6.94% 51,910 UKH Eastern 0.001% 0.2 0.17% 1,305 UKI London 0.001% 0.3 0.25% 1,885 UKJ South East 0.003% 0.7 0.79% 5,945 UKK South West 0.005% 1.2 0.78% 5,800 UKL Wales 0.253% 53.1 21.00% 157,035 UKM Scotland 0.178% 45.0 30.77% 230,115 UKN Northern Ireland 0.559% 113.1 25.69% 192,125 Data details: Austria: Federal state aid to the business sector in 2001 under the Regional Development Policy (ERP Fund and RIF-2000). Belgium: All awards (incentives) made in 2001 under the Economic Expansion Laws. Finland: (a) Total regionally identifiable state aid made in 2001 (by Tekes and T-EC); (b) Government funding of regional programmes in 2001 (POMO+, Community initiatives, Appropriations to island municipalities, RCDP, and OSKU). France: Regional aid directed to PAT areas only (average 1999-2002). Germany: (a) Approved Federal funds (2000-2002, annualised) under the ‘Common Task for the Development of the Regional Economic Structure’; (b) Targeted Federal support (budget) in 2001 for Urban Structural Development. Greece: (a) Regional allocation of state aid in 2002; (b) National financing of AKRITAS, EPTA and POLITEIA programmes in 2000 (excludes national contribution to ROP-CSF). Ireland: Agency Incentive Payments (grants and equity) in 2001. Italy: Total regionally identifiable state aid made in 2002. Spain: Payments made in 2001 under the Regional Incentives (regional state aid) programme. UK: Expenditure on Regional Selective Assistance in 2000/01 (excludes national contribution to European Regional Incentives); incentives in N.I. are not directly comparable with the rest of the UK regions.


107

7. National employment policies and regional development Employment policies, potentially, make a big difference to the aggregate economic development effort, as well as serving social objectives. A low employment rate means that there is unused productive potential, so that both economic and social cohesion aims can be served by increasing the level of employment. Indeed, in extreme cases such as those regions in the south of Italy and Spain where the employment rate is barely over 40%, the prospective contribution to GDP catch-up from higher employment is very substantial. In trying to assess the impact on regional cohesion of national employment policies of Member States, before dealing with specific national cases, it is necessary to take into account three general issues. These are:

First, there are marked differences in welfare systems and, within welfare policies, in the scale, mix of active and passive policies, and philosophies of labour market programmes, as well as their financial sustainability in the light of budgetary problems of several Member States.

Second there are evident differences in regional labour market imbalances within Member States. In some countries regional issues are crucial (with large differences in the employment and unemployment rates), while in others they are much less important.

The third concerns governance, that is the actors (national, regional, local) involved in labour policies planning and implementation, their responsibilities and powers, and the co-ordination mechanisms. Governance bears on the coherence of policy design and implementation, how well it copes with different regional situations, and the integration of employment and other regional development strategies.

Labour market policy expenditure in the framework of the welfare system The size and internal composition of the welfare systems in Europe, vary hugely, even though there are common elements that characterise the European social model (Esping Andersen, 1990; European Commission, 1995). One can distinguish, following Buti, Pench and Sestito (1998), four different models, taking into account total social expenditure as a percentage of GDP and the mix between in-kind services and cash transfers. The Nordic pattern (Sweden, Denmark and, as a borderline case, Finland), is characterised by a high level of social expenditure, and by a public policy against social risks based both on income support schemes and active labour market policies. Public services (PES) here play a crucial role. The Continental European pattern (Benelux, Germany, France and, as borderline cases, Italy and Austria and Ireland) shows high levels of social expenditures, though lower than in the Nordic case, and a policy mix based both on public services and benefits to individuals. The Mediterranean model (Spain and Greece, with Italy and Portugal as borderline cases), on the contrary is characterised by a much lower level of social expenditure and by a marginal role of public services. Finally, the United Kingdom shows a different pattern, with a low level of expenditure and a comparatively higher weight of services relative to benefits. Moreover, to show the very different patterns, one has at least to take into account the differences existing in Europe in terms of relationships between the generosity of unemployment benefits and the relevance of employment regulation.

Comparing summary measures of both Unemployment Benefits (UB) and Employment Protection Legislation (EPL) (Buti, Pench, Sestito 1998), as in Figure 7.1, one can see different grouping of countries: Spain, Portugal, Italy and Greece, where social protection is mostly based on the defence of existing employment (and where social benefits are generally linked to previous employment); in Denmark, Netherlands, Finland and Sweden the opposite is found, as employment protection is lower, but generosity of unemployment benefits higher; Belgium, France, Germany, Austria and Ireland are somewhere between, while the UK shows low levels of both employment protection and unemployment benefits, as in other ‘Anglo-Saxon’ countries. However, to complete the picture it is important to note that, in several EU


108

countries, invalidity and early retirement schemes operate in a manner that can formally exclude from the labour market workers who would otherwise be counted as unemployed. Figure 7.1 - EPL ranking (1985-93) and UB summary measure of generosity (av. 1990s)

0

10

20

30

40

50

60

70

0 5 10 15 20 25

Employment Protection Legislation (EPL)

Une

mpl

oym

ent B

enef

it (U

B) -

sum

mar

y m

easu

re o

f gen

eros

ity

*Adapted by the authors Source: Buti, Pench, Sestito (1998) based on OECD data

Dk

B

NL

FinNo F

D A

Gr

E P

IreNZ

Can Aus Swi UK

JapUS

It*

S*

In any case, workers’ protection is obtained by different combinations of job security and unemployment compensation. It is far from clear whether these different models have regional implications per se: however, it has been noted (Buti, Pench, Sestito, 1998), that the low UB-rigid EPL combination ‘may be highly inefficient in the event of sectoral, technological and organisational shocks entailing changes in comparative advantages and requiring swift labour reallocation between firms and sectors’. In the same vein it can be inferred that this model (prevailing in Mediterranean countries) may hinder regional reallocation of production and jobs, preserving the regional pattern of employment.

Tables 7.A.1 and 7.A.2, appended to this chapter, present more recent data on labour market programmes (LMP) expenditures, respectively for 2000-01 and 1995-96. They confirm the different magnitudes of employment policies within overall social expenditure in EU Member States: as a percentage of GDP, LMP expenditure varies from 0.89% for Greece to 4.74% for Denmark. Figure 7.2 maps EU countries (as of end-1990s) in terms of their expenditure on active (ALMP) and passive labour policies, confirming the Nordic pattern of higher expenditure in both of them, with some national variation, and the Mediterranean (plus UK and Austria) pattern of lower expenditure.

0.5

0.7

0.9

1.1

1.3

1.5

1.7

1.9

2.1

2.3

2.5

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 19987

7.5

8

8.5

9

9.5

10

10.5

11

Note: Unweighted dataSource: Montanino and Sestito (2003) based on Eurostat data

% GDP Employment rateFigure 7.3 - Labour market policies expenditure in the EU

--------------- Active Policies ________Passive Policies ____ Unemployment rate

Figure 7.2 - Labour market policies expenditure in the EU-15 (end


109

Table 7.1 measures the ‘intensity’ of LMP dividing the LMP expenditure/GDP ratio by an index of unemployment, confirming the overall picture. Comparing 1995-96 with 2000-01, one can see a general decrease of LMP expenditure: in evaluating this one has to consider that passive policies are highly cyclical, showing a strong, obvious, correlation with unemployment rates. Figure 7.3 shows how passive policies were more intense in the mid-1990s, when unemployment in Europe was higher. However, after 1996, overall active LMP expenditure decreased in Europe, following a long period of increase, probably because of national budgetary constraints. In particular (Tables 7.A.1 and 7.A.2), ALMP expenditure decreased in Portugal, Germany, Denmark, Finland and Sweden, increased in Greece, Austria and the Netherlands, remaining basically unchanged in other countries. The size of LMP expenditure is, and will be in the future, influenced by the need of EU Member States to respect budget deficit limits, reducing public expenditures.

It is important to note that the level of expenditure does not necessarily reflect the impact of LMP. In its 2003 Joint Employment Report (JER), the European Commission (EC 2003, page 14), states that ‘although Member States have developed many employment measures, assessment of their impact is generally lacking’. In sum, one may note that:

LMP expenditure is widely different in size in EU Member States, in the framework of very different models of welfare states and social protection;

In some Member States LMP expenditure decreased in more recent years, while in others it increased.

Labour market policies expenditures in the regions Regional labour market disparities, including employment rates, are quite large in the EU, prompting the Commission to observe that ‘regional disparities remain a serious issue’ (European Commission 2003). Yet the regional dimension was largely ignored in the early days of the European Employment Strategy (EES): none of the several Guidelines proposed by the Commission specifically mentioned taking account of either territorial disparities or

0.5

0.7

0.9

1.1

1.3

1.5

1.7

1.9

2.1

2.3

2.5

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 19987

7.5

8

8.5

9

9.5

10

10.5

11

Note: Unweighted dataSource: Montanino and Sestito (2003) based on Eurostat data

% GDP Employment rateFigure 7.3 - Labour market policies expenditure in the EU

--------------- Active Policies ________Passive Policies ____ Unemployment rate


110

the need to co-ordinate LMP with regional development policies. In 2001, however, the regional and local dimension was introduced in the EES through ‘horizontal objective’ D, emphasising the need ‘to take into account the regional dimension and regional disparities in terms of differentiated policies and targets, while fully respecting the attainment of national targets and the principle of equal treatment’, while Guideline 11 was reformulated, calling for a strengthening of regional and local action for employment (European Commission 2002). According to the European Commission, ‘the development of the territorial dimension of the EES has entered a phase of consolidation’ (European Commission, 2003).

Table 7.1 - Intensity of labour market policies (1)

(second half of 1990s)

PASSIVE ACTIVE POLICIES POLICIES

Spain 9.3 2.3Portugal 12.3 11.0Italy 6.9 3.5Belgium 25.7 13.1France 13.8 10.1Germany 15.2 13.6Austria 22.4 9.1Ireland 19.0 12.6United Kingdom 7.0 4.8Netherlands 55.3 21.9Denmark 32.6 14Finland 14.5 6.8Sweden 21.0 21.3 (1) LMP expenditure / unemployment rate =

GDP 1-(unemployment rate) Source: Montanino and Sestito (2003)

Also, due to the limited attention paid to the regional dimension of the EES, there is a lack of data and analysis, at European level, on regional impact of Member State employment policy. In the 1996 evaluation of the impact of Member State policies on cohesion (EPRC-LOEIL, 1996), the same point was raised: ‘there is, in practice a serious lack of empirical evidence on the effects of most of the measures included in the employment programmes being followed by Member States’.

In what follows, data from different national sources will be presented; they refer to the LMP expenditure by region and/or to the member of beneficiaries of LMP by region, when possible compared to target population (unemployed, labour force). Generally speaking, national sources provide more data where regional imbalances are larger: this is particularly the case for Spain, Italy, Germany, Belgium and Finland, and to a lesser extent, for Portugal, Greece, Ireland and the United Kingdom.

The conclusion of the 1996 evaluation (EPRC-LOEIL 1996) was that ‘while the effect of present policies on employment growth is uncertain, particularly in terms of its scale, their impact on cohesion is less questionable. A common characteristic of most of the measures in operation or being planned is to reduce disparities in employment opportunities whether between individuals, social groups or regions. While most measures do not have a specific


111

regional dimension, the very fact that many of them are aimed at the unemployed – and within this group at the most disadvantaged of those without work – is likely to mean that they benefit regions with the most serious problems of unemployment and inadequate rates of job creation more than others, so tending to narrow regional disparities’.

The tentative conclusion of the present survey is less straightforward. First of all, conclusions are tentative because almost in no case it was possible to have a complete picture, adding expenditures by all levels of Government and comparing them with target groups; good data on the regional breakdown could only be collected for selected countries. No study trying to assess the impact of national policies on regional circumstances was found.

However, as detailed below, only in the case of Finland does there seem to be an overall effect of LMPs in favour of regional cohesion. In Portugal, Spain and Italy, rules governing unemployment benefits tend to produce a situation in which the coverage of unemployment policies is higher in regions with lower unemployment. Elsewhere, the evidence is mixed. In most cases it is true that ALMP seem to be more intense in high unemployment regions, being specifically addressed to those regions or related to their needs. However, it is very difficult to state if their intensity is enough to guarantee a reduction of regional disparities. The 2002 JER contains an analysis by the European Commission of Member States’ LMPs as reported in their National Action Plans (NAPs) and expresses concern with respect to some possible impacts of Member States policies on high unemployment regions in Spain, Italy and Germany. The JER called for more intense policies in favour of regional cohesion in those countries, as well as in Portugal and Finland, also expressing fears of increasing disparities (though with much lower levels of both unemployment and disparities) in Austria, Ireland and the United Kingdom.

The governance of labour market programmes The main governance issues to take into account are the relationships between national, regional and local authorities in planning and implementing LMP. This is linked to the need for regional LMP to be closely linked to overall regional development strategies; thereby increasing their effectiveness and contributing to the success of the latter.

In general a ‘lack of institutional variety within national employment systems’ was noted in the past (Buti, Pench, Sestito, 1998): ‘a mismatch between institutions, mostly shaped at the national level and very often unduly homogeneous inside a given country, and the economic environment prevailing in some of its regions (…). Excess unemployment arises when a country is strongly differentiated on regional lines (…) and its political and social institutions are uniformly shaped from the centre’.

Again, Member States vary: ‘the degree of autonomy and influence of public administration at local and regional levels differs substantially across the EU. In Germany, the Länder and local authorities have flexibility to develop their own LMP programmes, while in Greece they do not (European Commission 2002). Principal differences between the Member States relate to the role of local authorities delivering social welfare (relevant in Denmark, Finland, France, Germany, Belgium, the Netherlands and the United Kingdom), in the use of territorial partnership (such as in Denmark , Ireland, Finland, Portugal) the relevance of regional PES (as in Austria) (European Commission, 2002).

According to the European Commission, in general ‘there is a growing tendency towards decentralisation of employment policies’. In fact, in 2002:

- together with National Action Plans (NAP) as agreed in the EES, regional action plans were in place in three Member States (Belgium, Portugal and Spain), and were introduced in Finland from 2002 onwards;

- Regional authorities have increased competence not only in the implementation of LMP but also in their strategic planning: in next paragraphs data and information will be


112

provided, with particular reference to the cases of Portugal, Spain, Finland, Germany, Austria, the United Kingdom;

- public employment services have been decentralised in several Member States;

- Territorial Employment Pacts have been developed in several Member States (particularly in Austria, Italy, Portugal) to design labour market programmes tailored to local conditions and linked to regional development plans (see also: Ecotec 2002).

However, ‘participation of local and regional authorities in the NAP process is in most Member States limited to the implementation phase’ (European Commission, 2003).

In the five-year evaluation of the EES carried out by the Commission ‘a general lack of mechanisms for reconciling local, regional, and national policies’ was noted: ‘in most Member States the policy-design process remains resolutely top-down, even where significant encouragement is being given to local and regional strategic plans for employment and development. In no Member States there is evidence of clear, intensive, structured and permanent local or regional involvement in the monitoring and evaluation of the NAPs’.

In some countries, as detailed below, problems of horizontal (between regions) and vertical co-ordination of employment policies may be emerging. Moreover, the welcome process of decentralisation of LMPs must be analysed in the light of the growing decentralisation of regional revenues in some Member States. In particular, there is a risk that less prosperous regions with lower fiscal receipts my struggle to continue to finance higher levels of LMP, so that decentralisation and its effects on regional cohesion must be monitored. Better co-ordination of a traditional sectoral policy, such as employment policy, with different regional development strategies would be welcome. Experiences in regional planning of LMP and development policies as well as actions such as Territorial Employment Pacts (TEPs) are very interesting and are definitely worthy of a careful impact evaluation. European regions can learn from each other’s best practices, exchanging their experiences and innovations.

7.1 The Mediterranean Member States In the Mediterranean countries, the welfare system is traditionally based on the protection of employment of household heads (usually male); unemployment benefits are usually confined to those previously employed; active policies normally available to those who lost employment; while universally available measures to guarantee minimum income and to help find new jobs are generally quite weak. Social protection is more ‘on the job’, with regulations for lay-offs, and less ‘in the market’ (Figure 7.1). Regional disparities in unemployment are very high in Spain and Italy, significant in Portugal, but low in Greece.

Overall labour market programmes expenditure is lower than EU average, with a minimum level (as a % of GDP) in Greece (0.89%, the lowest in the EU-15) and higher values for Spain. Comparing 1995-96 and 2000-01 data shows a decrease of expenditure in all countries except Greece, with, in particular, a strong reduction of unemployment compensation in Spain. Unemployment compensation represents around half of the total expenditure (two thirds in Spain). However rules governing unemployment compensation do not guarantee a sufficient coverage – also due to the quite large amount of unemployed in those countries – and create problems of territorial cohesion.

In Spain, net coverage (compensated unemployed over unemployed, excluding agriculture) is 68.5%. Coverage tends to be higher in some richer Autonomous Communities such as Cataluña (85.9%), and below 60% in some of the poorer ones, where the bulk of the unemployment is concentrated. In Portugal, gross coverage (including agriculture) goes from 29.4% in Alentejo to 76.9% in the Centro. In Italy the total coverage of passive policies, if calculated relative to the total member of jobseekers, is as low as 24%, and the figure is just 18.6% in the South and 12.9% for females in the South (Tables 7.2a-c).


113

Table 7.2a - Spain: unemployment compensation net rate of coverage, 2001 (1)

(selected regions) Baleares 118.0 Cataluña 85.9 C. Valenciana 80.1 Murcia 70.4 Madrid 68.7 SPAIN 68.5 C.La Mancha 59.2 Cantabria 57.4 Castilla y Leon 56.2 Ceuta 55.9 Melilla 46.8 (1) Ratio between compensated unemployed (excluded partial unemployed, subsidized casual agriculturar labourers and fixed agricultural labourers) and industry, construction and services unemployed. Source: Ministry of Labour

Table 7.2b - Portugal: participants of unemployment benefits, 2001

(gross rate of coverage (1)) Norte 53.5 Centro 76.9 Lisboa and VdT 64.3 Alentajo 29.4 Algarve 45.0 (1) Ratio between compensated unemployed and unemployed Source: MSST

Table 7.2c - Italy: total passive policies coverage, 2001 (1)

(% of total number of jobseekers) Centre-North 33.5 South 18.6 Italy 24.0 (1) extra-agricultural ordinary unemployment benefits with full requisites, building industry UB, mobility allowance, extraordinary redundancy fund, social serviceable workers Source: Ministry of Labour


114

However, in all countries the coverage of unemployment benefits increased in the second half of the 1990s. The limited coverage is due to the fact that a large share of the unemployed do not qualify for benefits due to eligibility criteria such as insufficient social security contributions (Portugal) and to the fact the benefits are only given to the employed who lose their job (Italy). Unemployment benefits are traditionally complemented by special schemes for agriculture that provide limited income support (mostly in poorer regions), such as unemployment social benefits (particularly important in Alentejo and Algarve in Portugal where they cover around 75% of the unemployed), the subsidy for Casual Agriculture Labourers (reaching 230,000 people in Andalucia and Extremadura in Spain) and the Italian agricultural employment scheme, for 502,000 workers in Southern Italy. Portugal introduced in 1997 the guaranteed minimum income, expenditure on which is increasing.

Expenditure for active policies is much smaller in the Mediterranean Member States than in the rest of the EU, with a minimum of 0.41% of GDP (1997-98) for Greece. Data for training are incomplete. However, they show that a very limited percentage of the unemployed is enrolled in Portugal (with a slightly higher figure for Alentejo) as well as in Italy, where more people in the Centre-North than in the South benefited from the vocational training, mostly financed by the ESF, over the period 1994-99. In Spain around 260,000 unemployed people were involved in training in 1999.

Subsidies to labour demand are important in Spain and in Italy, where self-entrepreneurship incentives are also provided. In Southern Italy the percentage of the jobseekers covered by ‘mixed cause’ contracts, recruitment and self-entrepreneurship incentives went up from 15% to 40% (1996-2001). However 40% in the South still has to be contrasted with 100% in the Centre-North. Though effective in increasing total employment, direct subsidies have to be carefully targeted: otherwise they can have the unwelcome effect of promoting employment only among those with better skills and qualifications, leaving untackled the weakest areas of labour supply, especially the long-term unemployed.

A key issue in all Mediterranean Member States is devolution of powers concerning ALMP to regional and local authorities. In Spain the process is almost complete, in Italy it has gone a considerable way and in Portugal several regional programmes are in action, particularly Regional Networks. The Portuguese National Employment Plan subscribed to the logic that promoting employment and fighting unemployment is more effective when employment policies are implemented at regional and local levels. Some attempts have been made to use ‘local employment pacts’, especially in Portugal and Italy but also in Greece.

The use of such Pacts is, however, more likely to have a positive impact if regional authorities are able to achieve better integration of labour market policies with overall territorial development policies, trying to design and implement strategies that include measures (such as training, self-employment, public services for the unemployed) specifically tailored to local market conditions and that can best fit into development dynamics. However, there are potential pitfalls. On one hand, resources allocated to regional authorities must be adequate; on the other hand, the expenditure and the results achieved must be checked. In this respect both Spain and Italy seem still to lack adequate data. Moreover, the efficiency and efficacy of local and regional actions cannot be taken for granted. In Spain there is the fear that a more decentralised system may penalise less wealthy and dynamic regions; in Italy, available data on Public Employment Services (PES) show that they work better in richer regions, where unemployment is much lower. Decentralisation involves the need for a continuous and thorough institution building process at the local and regional levels and careful monitoring, co-ordination and evaluation of performances.

Finally, in all Mediterranean Member States, labour market regulations are becoming more flexible. Again, the territorial effects of labour market flexibility must be carefully assessed, to ensure that they help to create sustainable job opportunities in less developed regions. One key dimension of flexibility is labour costs. In Spain we see quite large differences between Canarias, Castilla La Mancia, Extremadura and Murcia (around 12 euros per hour in 2002)


115

and Pais Vasco (17.75), mostly due to salaries. In Italy, regional differences in labour costs were mostly due, in the past, to social contribution exemptions in the Mezzogiorno. With these exemptions now gone, larger differences in real compensation have emerged, with the manufacturing average in the South falling from 92% of the Centre-North in 1995 and 87% in 2000. In Greece, wage rates for white collar workerss (1998, national average=100) go from 92.1 in Ipeiros to 114 in Sterea Ellada.

Looking at the overall picture, the Joint Employment Report 2002 (European Commission 2003) states that in Spain ‘no additional measures have been adopted in order to reduce regional disparities in employment and unemployment. Given the scale of regional disparities, which are, although improving, still among the most pronounced in the EU, a greater effort is needed’. In particular it seems important ‘to increase co-ordination between regional employment services’ and ‘adequate balance between regional development and functional and geographical mobility seems needed’.

In the case of Portugal ‘the reduction of employment disparities between regions requires that the PES continue their efforts to provide adequate responses’. For Italy, the JER states that ‘doubts remain on whether the reforms will improve the labour market in the South’; ‘the policy mix could lead to marginalisation of disadvantaged groups and generally in the South, also due to the reliance on tax credits which do not seem to favour unemployed people in the area’, so that ‘particular attention seems necessary for the security and quality aspects in order to avoid the risk of marginalisation for the long-term unemployed and women, especially in the South. Key issues remain the reduction of tax burden on labour, especially for low-paid workers, and the role of the PES at regional level in providing active and preventive measures’. In the case of Greece, the JER recommends ‘measures to reduce the non-wage labour costs (notably for the low-paid)’.

7.2 The Nordic Member States The three Nordic Member States, together with the Netherlands, have long had the largest labour market programme expenditure as a proportion of GDP. In 2000-01 Denmark spent 4.74% of GDP on those policies, with Finland at 3.04% and Sweden at 2.49%, while the Netherlands is at 3.53%. The ‘intensity’ of LMP (Table 7.A.1) in these countries is the highest in the EU.

Comparing 1995-96 with 2000-01, there has been an overall reduction of expenditure for both active and passive measures, with the exception of active policies in the Netherlands, although both remain large (Figure 7.2). Unemployment compensation ranges from 1.26% to 2.02% of GDP; in Denmark and Finland large early retirement schemes exist as well. Active labour market policy expenditures also range from 0.97% to 1.57% of GDP. Active policies are based on actions performed at the local level by public employment services, such as massive training programmes (up to 0.92% of GDP in Denmark) and counselling. Subsidised employment is important, as well as measures for the disabled, especially in the Netherlands, Denmark and Sweden. Wages and labour conditions are bargained centrally, with minor regional differences.

Finland is the most interesting case for this analysis, because it presents a quite large variation of unemployment rates among regions, with peripheral regions suffering more than the coastal cities in the South. The intensity of most employment policies is stronger in less developed regions, notably Eastern and Northern Finland. The regional breakdown of funds for active labour policies, in fact, shows a higher incidence of policies in the regions with higher unemployment: this happens for adults in training (as a percentage of regional labour force: 1.27% in Eastern Finland compared to 0.48% in Uusimaa), for the participants in youth measures (3.6% versus 0.83%), for subsidised employment (5% vs. 0.99%) and for people participating in measures for the disabled (0.9% vs. 0.42%) (Table 7.3). Moreover it was decided to waive social security contributions for a trial period (2003-2005) in 14 municipalities in Northern Finland in order to raise employment. In 2002 the general criterion was that if the unemployment rate exceeded the national average by more than 1.6 times,


116

special measures for subsidised employment were started: and this was the case of Eastern and Northern Finland.

The picture for passive policies is rather different (Table 7.3). Data for unemployment assistance show a more limited regional differentiation; as far as early retirement is concerned, Southern Finland comes first, followed by Eastern Finland. Policies are basically implemented on a regional basis, with Regional Employment and Economic Development Centres (TE-Centres), playing a central role. Regional targets exist for the maximum number of people left unemployed after labour market training and subsidised employment.

Table 7.3 - Finland: selected labour policy measures (beneficiaries as a percentage of labour force)

adults youth subsidized unemployment in training (1) measures employment assistance

Uusimaa 0.5 0.8 1.3 1.7South F. 0.9 1.9 2.6 1.9East F. 1.3 3.6 5.1 2.0Mid F. 0.9 2.6 2.9 1.8North F. 1.1 3.1 5.3 1.6Aland 0.1 0.0 0.1 0.1Finland 0.8 2.0 2.8 1.8 (1) unemployed during training Source: Ministry of Labour

The reduction in Danish unemployment since 1993 has benefited all regions. While this cannot be attributed to a set of selective territorial policies, there is an important regional element in this development due to the fact that labour market policy was strongly decentralised with the labour market reform in 1993. The reform made tri-partite regional labour market boards responsible for the implementation of active measures.

In the Netherlands, policies follow a national strategy, the only exception being the Kompas program for the North of the country which has a specific employment target. There are neither specific employment policies of regional or local authorities, nor regional differences in the agreements concerning labour markets. Subsidised labour, in terms of both the so-called WSW-jobs (social labour provision) and the so-called WIW-job (subsidised employment) are relatively more important, as a percentage of employment, in the Northern provinces of Groningen, Friesland and Drenthe, as well as in Overijssel and Limburg in the South, where unemployment rates are higher. The share of WIW-jobs is as high as 3% in Drenthe and 2.7% in Groningen, compared with a national average of 1.4%.

Regional issues are not raised in the chapters of the Joint Employment Report 2002 (European Commission 2003) for Denmark, the Netherlands and Sweden. According to the JER, problems emerge in the Netherlands for the considerable share of working age population depending on social security benefits, in particular the disability scheme, and in Sweden for the rapid increase in long-term sick leave, with possible regional implications. For Finland, the JER notes that ‘in order to make regional policies more effective, Finland is introducing new legislative and operational tools, whose impact needs to be monitored’; ‘there is a trend towards higher participation of relevant actors in the NAP process at all


117

levels. However, the regional and local dimension of the process needs to be consolidated’. The large regional variations of unemployment ‘call for further action’.

7.3 The Central EU Member States Expenditure on labour market programmes in 2000/01 in the Central EU States ranges between 3 and 3.5% of GDP (France, Belgium, Germany), but is only 1.59% in Austria (Table 7.A.1). Their overall social expenditure is also quite high (even if lower than in the Nordic countries), and comprises a mix of benefits to individuals and public services (Buti, Pench, Sestito, 1998). The ‘intensity’ of LMPs is halfway between the Nordic and the Mediterranean countries.

Passive policies, in particular unemployment compensation, represent the largest share of the expenditure, but active policies are still substantial at around 1.3% of GDP (0.52% for Austria). In the second half of the 1990s, LMP spending decreased, especially in Belgium and in Germany, with passive policy expenditure decreasing particularly (Tables 7.A.1 and 7.A.2). In France, unemployment compensation and early retirement are important and stable, comparing 1995-96 and 1999-00. Active policies cost 1.34% of GDP, with training, youth measures and subsidised employment most prominent.

Belgium is interesting because of relevant regional disparities. In Belgium both unemployment compensation and early retirement allowances represent large parts of expenditures (1.75% and 0.51% of GDP respectively), with subsidised employment traditionally being the most important tool of active policies. Regional data are available as far as participation in training programmes is concerned (National Action Plan 2002), showing a higher rate of participation in the Flanders (where unemployment rate is lower) than in Wallonia. Regions are important in labour market programmes in Belgium. While action for unemployment is managed at a federal level, regions have substantial powers concerning employment programmes, and local Communities have powers concerning vocational training; each region (except for Wallonia) has its own actions plan for 2002 and the National Action Plan identifies regional targets for the employment rate, reflecting the different points of departure.

However governance seems to be a major issue. In fact, the EU Joint Employment Report 2002 (European Commission 2003) states that notwithstanding initiatives taken to harmonise the high number of federal activation measures, ‘further integration of federal and regional labour market policies is needed’. In particular, ‘in view of the substantial regional and sub-regional disparities in labour market performance, better co-ordination and co-operation among the regional public employment services seems to be called for’.

In Germany, labour market programmes fell to 3.13% of GDP in 2000-01 compared to 3.84% in 1995-96. Unemployment compensation remains a substantial expenditure, with 1.89% of GDP (the highest share in the EU-15, after the Netherlands). Active policies go particularly towards public employment services, training, subsidised employment (in particular direct public and no-profit-job creation) and measures for the disabled.

However, due to budget constraints, the Government is planning to reduce substantially the spending on earnings-related and unemployment benefits, from 40 to 13 billions, via stronger incentives for a quick job placement, support for self-employment and stronger use of temporary employment. The high level of early retirement still creates problems for the German social security system and for the reduction of non-wage labour costs. Main reform priorities are the reduction of youth and long-term unemployment, enhancement of training efforts, promotion of the dual apprenticeship (Job –Aqtiv Act and the so-called Hartz concept). According to the EU Joint Employment Report 2002 (European Commission 2003), ‘reform of the PES represents an opportunity to strengthen the guidance and placement functions and give a greater focus ton regional needs’.

National policies, as stated in German NAP, aim at the reduction of the persisting regional disparities between the old and the new Länder. Germany presents a partial differentiation of


118

labour regimes between the East and the West, with regards to working hours in industry, subsidised wages, working hours flexibility and wage differentiation; 2002 hourly wage in industry varies from 10.30 euros in Thüringen to 16.77 in Bremen.

The regional dimension of labour market programmes is particularly important, with the Länder enjoying substantial powers in education, training and employment policies. As outlined in the German NAP 2002, regional employment policies have different focal points and priorities, with a number of differences in regional policy measures (Table 7.4). Since 1998, all Länder have put greater emphasis on active labour market measures and the development of more innovative strategies. All Länder try to define their employment policies according to their own strengths and weaknesses and to particular problems. However, employment policies in the Eastern Länder still focus on publicly funded direct job creation measures as well as on subsidised employment. There, structural changes (such as the closing down of shipyards and mining activities) are partly addressed by national employment measures. As a result of these policies, subsidised employment measures are more costly in Eastern Germany, while the expenditure for apprenticeship (youth measures) and for disadvantaged persons is more even (Table 7.4).

Table 7.4 - Expenditure of the Federal employment service

(composition by Lander, 2000 and 2001)

PES

(2002) training youth subsidiseddisabled

(2002) unemployment TotalBaden-Wuttenberg 46.5 1.9 1.5 2.2 3.8 44.2 100.0Bayern 52.7 2.2 1.8 2.5 3.6 37.2 100.0Berlin 40.0 5.7 2.5 11.2 6.8 33.8 100.0Brandeburg 48.4 3.2 4.6 6.1 2.0 35.8 100.0Bremen 37.0 2.5 22.6 3.9 2.0 32.0 100.0Hamburg 48.9 3.1 2.2 5.1 2.4 38.3 100.0Hessen 50.2 2.6 2.2 4.5 3.0 37.5 100.0Meckenburg-V. 48.1 3.3 5.0 7.0 1.8 34.9 100.0Niedersachsen 48.1 2.8 2.5 6.4 2.9 37.4 100.0Nordhein-V. 49.1 2.8 2.4 3.0 2.9 39.7 100.0Rheinland-P. 43.4 2.5 2.1 16.4 3.4 32.3 100.0Saarland 49.9 3.1 2.7 3.6 3.7 37.0 100.0Sachsen 47.9 2.7 3.9 8.0 1.7 35.8 100.0Sachsen-A. 41.3 3.4 2.8 6.9 1.9 43.7 100.0Schleswig-H. 52.0 2.9 3.1 3.4 0.8 37.8 100.0Thüringen 49.8 3.7 3.7 6.4 2.4 34.0 100.0 Source: Bundesanstalt für Arbeit, Aktuelle Arbeitmarktdaten (2003)

However, according to the Joint Employment Report 2002 (European Commission 2003), ‘more efforts are required to address the structural problems of the Eastern Länder, and to improve the consistency between employment and inclusion policies’. In fact, ‘regional disparities on the labour market could increase further. A comprehensive approach in the Eastern Länder combining regional labour market and structural policies seems needed to support employment creation and reduce inflows into long-term unemployment’.

In Austria, labour markets exhibit fewer problems than in other EU countries; nonetheless, wages suffer from a heavy tax burden and the employment rate of over-55s is low, partly because of a generous early-retirement scheme. The global amount of public expenditure on labour market programmes is quite low, if compared to other Central European countries, at


119

1.59% of GDP in 2000/01; with passive policies at 1.07% of GDP, active policies are particularly small, among the lowest in the EU-15. However, comparing 1995-96 and 2000-01, there is a clear evidence of a decreasing role of passive action and of an increasing role of active measures, with a progressive phasing-out of the generous early-retirement scheme. Labour market policies are managed by the Arbeits marktservice (AMS), which is, according to the Austrian Ministry of Economic Affairs, the most regionalised employment agency in the EU.

A considerable share of active policies is in fact implemented through a network of territorial employment pacts (TEP), which cover the whole territory of Austria. The TEPs are regarded as key elements of the Austrian NAP strategy. The Joint Employment Report 2002 (European Commission 2003), states that the TEPs ‘seem to be the key to unlocking regional growth potential’.

Due to the different problems of regional labour markets, with higher unemployment rates both in Vienna and in the two lowest income regions (Burgenland and Carinthia), but widely different problems related to unemployment (structural labour shortages, inbound commuting, excess labour supply unemployment of older people - Alteneder et al., 2002), regional strategies are quite different. Hardship allowance recipients (persons eligible for unemployment benefits anymore) as a proportion of the workforce are as high as 3.9% in Vienna, as compared to 2.1 on average, while apprenticeships are more diffused in Burgenland and Carinthia, and enterprise integration subsidy beneficiaries in Carinthia. By contrast, senior part-time beneficiaries represent a higher share of the unemployed in Upper Austria, where the overall unemployment rate is the lowest in the country.

In general the emphasis on older workers and on the qualifications of the low skilled represent the major differences among regional strategies. However, Austria is the only Member State (apart from Germany), in which regional disparities are widening for both employment and unemployment (although in Austria it is still at a very low level - European Commission, 2003, page 27).

In Ireland labour market programme expenditure is quite high. As far as their regional impact is concerned, the JER notes that ‘the delay in finalising the National Spatial Strategy does not allow for a complete assessment to be made. Although it is recognised that policies adopted have had an impact in reducing regional imbalance, it is disappointing to note that the slowdown in economic growth and consequent increase in unemployment has had greatest impact in the Border, Midlands and Western region, the poorer of the two regions of Ireland’.

However, according to the JER ‘social partnership is a strong factor in employment success and could be further strengthened by increasing the regional and local dimension’. So ‘it is expected that the imminent National Spatial Strategy will contain policies to address the problem of regional imbalance’.

7.4 The United Kingdom Labour market programme expenditure in the United Kingdom is relatively low compared with other EU countries. In 1998-00 it was less than 1% of GDP, a figure comparable to that of Greece, the lowest EU spender. If compared to 1994-96, the importance of labour market programmes has substantially decreased; passive policies in particular were reduced, with their weight in terms of GDP falling from 1.34% to 0.60%.

In fact, employment policies in the UK have seen radical changes over the last two decades. Throughout the 1980s and 1990s the UK introduced a number of measures aimed at enhancing labour market flexibility and labour mobility, with the objective of improving the operation of its labour market(s) and enhancing adaptability, as well as employment prospects. Especially in the 1990s, UK employment policy put increasing emphasis on Active Labour Market Policies (ALMPs), probably to a larger extent than any other European country outside Scandinavia. In this vein, unemployment benefits were replaced by the Job


120

Seekers Allowance (JSA), which makes the payment of the benefit tied to active job-searching.

The main instrument for the implementation of the UK employment policy is the New Deal programme, based on a range of partnerships, mainly co-ordinated by Jobcentre Plus, between local and regional authorities, government bodies, training and educational institutions, employers, and voluntary organisations. The New Deal programme is divided into six initiatives, each targeting specific segments of the labour force and thus configured to reflect their specific needs, so as to increase efficiency in the provision of support.

Outside the area of job-brokering and unemployment insurance, the Employment Service has, since 2001, introduced various schemes that focus on the skills needs of job-seekers. The purpose is to identify areas of training provision and, in collaboration with the local LSCs, to provide vocational training on specific skills. Due to the emphasis on local partnerships in the implementation of job-brokering and re-training policies, there is a large variation across local areas in the range and quality of the non-financial elements of support provided by the employment institutions in the UK. Although this might allow for perpetuation of regional differences in employment opportunities, the autonomy enjoyed by the employment service agencies at the local level allows a higher degree of flexibility and thus a higher sensitivity to local conditions and needs.

At the regional level, responsibility for the design of employment strategies is shared between the RDAs, the GOs and the local authorities, which together participate in the designing of the Frameworks for Regional Employment and Skills Action (FRESAs) and various employment-generating schemes. At a central level, the government runs the New Deal for Communities (a 10-year initiative which started in 1998), which provides special job-brokering and retraining support in 39 high-unemployment neighbourhoods in England. The DWP ran for the last three years (April 2000 – March 2003) an Employment Zones programme in 15 areas of high long-term unemployment.

The various New Deal programmes are organised on a regional basis. Thus, training provision and job placements are run by different companies in different regions, with the implication that not all schemes are available in all areas in the country. Although there are significant differences in the provision and effectiveness of training programmes across the UK regions, these to a large extent reflect the differences in the economic circumstances of these regions. In the UK, there are substantial earnings differentials, not only between the greater South East and the rest of the country, but also between other regions

The regional distribution of recipients of unemployment insurance (JSA) reflects that of unemployment in the regions. Significant variations exist also in terms of types of benefit entitlement. Outside London, the east and south areas of the country have very high shares of contribution-based JSA recipients and the lower levels of JSA claimants with no benefits. Such variations reflect specific economic conditions in the regions as well as differences in unemployment duration and the age composition of the unemployed.

Similarly, such differences explain much of the regional variation of participation in New Deal and other training programmes (a simple correlation between ILO unemployment and adult participation rates is around 50%). However, significant differences remain. For example, the South West has one of the highest participation rates, despite being a relatively low-unemployment area. There are also significant regional variations in the type of programme followed: for example, London has higher relative participation in the Gateway programme and very low levels of subsidised employment schemes, which seem to concentrate significantly in Wales.

In its assessment of labour policies, the Joint Employment Report notes that in the case of the UK, ‘further challenges include tackling the increasing employment disparities particularly within regions’.


121

7.5 Commentary and key findings In some Member States, especially Italy, Spain and Germany, very low employment rates can be observed in lagging regions, suggesting opportunities for employment policies to advance economic cohesion. Active policies, targeted at improving employability and the adaptability of workers and thus boosting human capital, are the instrument that can help in this regard, rather than ‘passive’ measures which maintain the income levels of the unemployed.

There are very pronounced differences among the Member States in both the overall scale, and in the balance between, active and passive measures in employment policies. Labour market policies are in flux in many countries with reforms simultaneously being seen as essential yet politically very tricky to enact. In Germany, resistance to efforts to improve the effectiveness of public employment services exemplifies the problems. Regional priorities were not emphasised in the early years of the EES, but have subsequently been given greater prominence. Nevertheless, the main focus of employment policy is on national objectives, with the implication that economic cohesion in the sense of regional differences is a secondary objective.

As with other policy areas affecting cohesion, institutional structure matters. In some countries, such as Austria, there is substantial decentralisation of employment policy: according to the Austrian Ministry of Economic Affairs, the Arbeits marktservice (AMS) is the most regionalised employment agency in the EU. Even in the UK, despite its centralised tradition, the implementation of the various ‘new deal’ programmes – a flagship policy of the Labour government – has been devolved to regional offices of central government in England (and to the devolved authorities in Scotland, Wales and Northern Ireland).

Key findings are: Expenditure on employment policies declined marginally in the second half of the 1990s

as a proportion of GDP, principally because the intensity of passive policies declined in response to the improved macroeconomic position.

However, in some Member States there was also a decline in active labour market policy spending, with the imperative of fiscal consolidation evident in budget cuts.

From the evidence collected, only in Finland does there seem to be a marked direct impact on regional disparities of national employment policies, and thus on economic cohesion.

Because of rules governing eligibility for unemployment benefit, coverage of benefits in Portugal, Spain and Italy tends to favour richer regions, and could therefore be in conflict with both social and economic cohesion. This point was highlighted in the 2002 Joint Employment Report.

Employment subsidies that favour regions with high unemployment are used in some Member States, such as the Netherlands (where the WSW and WIW schemes cover twice as many workers in the Northern regions as the national average). In Italy, by contrast, a subsidy that reduced social charges in the South has been discontinued and the evidence suggests that this has led to lower compensation for workers, potentially aggravating regional income disparities.


122

Table 7.A.1 - Public expenditure in labour market programmes, average 2000-01

(% of GDP)

public empl l. market youth subsidized measures ACTIVE unemployment early PASSIVE

services training measures employment disabled POLICIES compensation retirement POLICIES TOTAL

Spain 0.09 0.15 0.07 0.43 0.03 0.77 1.34 1.34 2.11

Portugal** 0.11 0.22 0.25 0.1 0.04 0.71 0.67 0.18 0.86 1.57

Italy**** n.a. 0.06 0.24 0.30 n.a. 0.67 0.54 0.12 0.66 1.33

Greece* 0.09 0.14 0.10 0.07 0.01 0.41 0.48 0.48 0.89

Belgium ** 0.18 0.24 0.78 0.12 1.32 1.75 0.51 2.25 3.57

France** 0.18 0.27 0.41 0.4 0.09 1.34 1.42 0.29 1.7 3.04

Germany 0.23 0.34 0.09 0.29 0.28 1.22 1.89 0.02 1.91 3.13

Austria 0.14 0.19 0.04 0.11 0.5 0.52 1.02 0.05 1.07 1.59

Netherlands 0.26 0.31 0.04 0.4 0.57 1.57 2.02 1.96 3.53

Ireland 0

United Kingdom*** 0.13 0.05 0.14 0.01 0.02 0.35 0.6 0.6 0.95

Denmark** 0.12 0.92 0.11 0.20 0.33 1.67 1.39 1.68 3.07 4.74

Finland 0.12 0.30 0.17 0.30 0.09 0.97 1.58 0.49 2.07 3.04

Sweden 0.24 0.31 0.02 0.26 0.42 1.23 1.26 1.26 2.49

* data are for the years 1997-98; ** 1999-00; ***1998-00;**** OECD and national sources

Source: Own calculations based on Oecd Employment Outlook


123

Table 7.A.2 - Public expenditure in labour market programmes, average 1995-96

(% of GDP)

public empl l. market youth subsidized measures ACTIVE unemployment early PASSIVE

services training measures employment disabled POLICIES compensation retirement POLICIES TOTAL

Spain 0.09 0.33 0.09 0.23 0.01 0.74 2.29 2.29 3.03

Portugal 0.11 0.28 0.34 0.11 0.06 0.89 0.88 0.11 0.99 1.88

Italy 0.04 0.01 0.4 0.65 n.a. 1.11 0.68 0.2 0.87 1.98

Greece 0.12 0.09 0.03 0.07 0.01 0.32 0.44 0.44 0.76

Belgium 0.22 0.28 0.05 0.76 0.14 1.45 2.11 0.65 2.77 4.2

France 0.16 0.37 0.26 0.44 0.09 1.31 1.44 0.36 1.8 3.11

Germany 0.24 0.42 0.06 0.43 0.27 1.4 2.23 0.22 2.44 3.84

Austria 0.13 0.13 0.01 0.05 0.05 0.37 1.29 0.13 1.42 1.8

Netherlands 0.35 0.15 0.1 0.19 0.54 1.33 3.6 3.6 4.93

Ireland 0.26 0.22 0.25 0.87 0.07 1.67 2.43 0.14 2.56 4.23

United Kingdom* 0.21 0.12 0.13 0.03 0.03 0.50 1.34 1.34 1.84

Denmark 0.12 1.07 0.12 0.34 0.3 1.95 2.8 1.68 4.48 6.42

Finland 0.16 0.51 0.19 0.67 0.14 1.65 3.42 0.43 3.85 5.5

Sweden* 0.26 0.66 0.13 0.86 0.76 2.68 2.39 0.02 2.39 5.07

* data are referred to years 1994-96

Source: Own calculations based on OECD Employment Outlook


124

8. Technology policies The goal, announced at the Lisbon European Council in 2000 of becoming the most competitive knowledge-based economy in the world by 2010 was predicated on the view that policy can positively affect the long-run growth rate of the economy through incentives for capital accumulation and the promotion of technological innovations. Such a conviction relies on the postulates of endogenous growth models (Romer, 1986 and 1990; Lucas 1988) and has motivated the proliferation of numerous national and European technology programs over the last decades.37 The idea behind these programs is the following: R&D generates innovation and new technologies, and innovation and new technologies then generate economic growth. This will happen because new technologies increase productivity and therefore have a positive supply side effect on the growth potential of the economy.

Since both national and European funds for R&D initiatives are usually distributed across regions on the basis of ‘excellence criteria’, the main recipients of these funds tend to be those regions which have a technological comparative advantage, typically in the most developed countries of the EU or the more advanced regions in other countries. Thus, while technology, R&D and innovation policies could be very important in fostering new industries, stimulating productivity gains and helping to develop the knowledge-intensive economy, it is clear that only few regions have attained strong competitive positions in such industries. Consequently, support for national excellence can be at odds with cohesion aims and the impact of technology policies on economic cohesion will often be negative, insofar as they promote economic activity in richer, more advanced regions and result, at least in first round effects, in divergence.38

Yet in much the same way as discussed above in relation to state aids, there is a dilemma if technology policies agreed at the national level facilitate more rapid technological advance that boosts national growth. By doing so – even if regional divergence increases, and economic cohesion (and possibly territorial cohesion) is compromised – it may be that growth trickles-down to poorer regions such that the effect of the policy is ultimately beneficial for social cohesion, albeit in a manner that, if positive, is indirect rather than immediate.

Although all Member States have some form of technology or innovation policy, most often explicitly linked to the competitiveness agenda, the regional dimension is much less systematic and it is not uncommon for the policy framework to evolve rapidly. For example, in the Netherlands, there were regional innovation centres, which merged in 1998 with the information institute of small and medium-sized enterprises into one organisation named Syntens with 15 establishments throughout the country. Its main purpose was to mediate between companies and knowledge institutes. However, Syntens has been superseded by an organisation connected to the Ministry of Economic Affairs called Senter39 (www.senter.nl) whose main task it is to stimulate innovations in the Dutch business sector. Now, there are no innovation policies devoted to specific regions, although regions with an industry structure biased towards manufacturing and large enterprises may benefit more from R&D policies

37 While neo-classical growth models (Solow, 1956; Mankiw, Romer and Weil, 1992) consider that economic integration would assure convergence between poor and rich countries (regions) due to capital accumulation in poorer regions that offer higher returns to capital, more sophisticated endogenous growth models (Romer, 1986 and 1990; Lucas 1988) and new economic geography models (Krugman, 1991; Ottaviano and Puga, 1998) show that income convergence need not occur as a result of economic integration. Consequently, pro-active public policy has a role to play in promoting economic convergence between poorer and richer countries or regions. For a more detailed summary of growth theories and the convergence-divergence debate, see Martin and Sanz (2003). 38 It is sometimes asserted – in a not entirely tongue in cheek manner - that the Community Framework Programmes are Structural Funds for the rich. 39 For the former, see www.rik.nl/syntens.html; and for the latter, www.senter.nl


125

than regions that do not. Larger firms have more opportunities to engage in the process of obtaining innovation subsidies than small firms.

8.1 Spatial Distribution of Total R&D Expenditures A look at the distribution of European regions that invested most in total R&D in 2000, shows that there are marked disparities. Among the regions that invested most in R&D are Braunschweig (7.19%), Stuttgart (4.92%), Oberbayern (4.79%), Pohjois-Suomi (4.73%); Pohjois Suomi (4.73%), Uusimaa (4.09%) and Tübingen (4.31%). Those that invested least include Calabria (0.24%), Castilla-la Mancha (0.22%), Sterea (0.18%), Dykiti Makedonia (0.07%) and Notio Aigaio (0.06%). In 2000, the EU’s average regional R&D spending was 1.22% of GDP with a standard deviation of 1.01.

That German regions are the top three investors in R&D is no surprise, since these are often quoted in the specialized literature as prime examples of innovative regions with high technological potential. Typically, these regions have a high capacity to create and absorb new knowledge and transform it into commercial products or into some form of competitive advantage. In addition, these regions have comparatively high levels of economic activity (GDP per capita), as well as being endowed with high levels of human capital. There are large numbers of qualified scientists in both higher education institutions and public laboratories, and R&D personnel in the business sector. Large metropolitan areas such as Ile de France, Uusimaa region around Helsinki, and Oberbayern, provide firms with a thriving business environment due to economies of scope and scale, and a political power centre.

To put the spatial distribution of total R&D expenditures in context, it is very important to note that statistics at the regional level show the more severe disparities between regions that remain hidden in statistics at national level. This is especially true for data on technology indicators. For example, while regional differences in income per capita are only twice the economic differences between Member States, disparities in technology input (TERD/GDP) and output (patents per million people) are, respectively, as much as 20 and 55 times higher at regional than at the national level.

8.2 Spatial Distribution of Patent Applications The spatial distribution of patent applications is even more unevenly distributed across regions than total R&D expenditures. There are many regions which, in 2000, filed fewer than 4 patent applications per million people, for example: Dessau (3.3), Andalusia (2.8), Molise (1.9), Galicia (1.5) or Calabria (0.9). At the other extreme, regions which filed more than 180 applications per million people included Koln (189.3), Berkshire (197.1), Stockholm (219.7), Noord-Brabant (266.8), Darmstadt (306.6) and Oberbayern (441.95). Although the EU’s average of patent applications per million people at regional level in the year 2000 was 152.8, there is a very high standard deviation of 147.9.

A simple comparison between the distribution of the input and the output technology indicators shows that the technology gap measured by patent applications was wider in 2000 than initially suggested by the R&D indicators. In relation to this technology gap, the Member States of the EU can be divided into three groups. First, there are the countries in which there tends to be a small set of extremely patent intensive regions (relative to the country average). In six of these countries the patent intensity is highest in a single region that, statistically, is an outlier (Noord Brabant in the Netherlands, East Anglia in the UK, Vorarlberg in Austria, and Ile de France, Rhône Alpes and Alsace in France). Second, there is a group of countries with negligible patenting activity (Greece, Portugal and Spain). In a third group are countries with a larger number of patent intensive regions (Germany, Italy and Belgium), but where patent activity is dominated by a leading group of regions that performs outstandingly.

After controlling for outliers, the regional disparity in technological development is not so high. This is because patenting activity is Europe is dominated by a small set of regions (an ‘Archipelago’ of ten regions as suggested by Hilpert,1992), with all others making only a marginal contribution. Most of these regions belong to the group of the very rich and all are in


126

the EU’ old industrial core. Despite the emergence of new high tech regions (such as Alsace or Barcelona) which are on their way up, it is unlikely that these new catching-up regions will change the domination of the north over the south in patent applications. There is also evidence of specialization: Oberbayern concentrates on automobile, aerospace and electronics, obtaining 30% of its patents in the latter field; while the Brabant Wallon region requested half of its patents in the chemistry and metallurgy sectors.

8.3 Spatial Distribution of Government R&D expenditures When compared to the spatial distributions of the two technology indicators, disparities in regional public R&D as a % of GDP lie in-between. In 2000 the (unweighted) average level of regional public R&D in the EU spending was 0.19% of GDP with a standard deviation of 0.27. It is interesting to observe that the top regions in terms of the different technology indicators do not always coincide. For example, in Germany, Baden Württemberg outperforms all other Länder (i.e. Level I regions) in terms of patents, but Berlin receives the highest amount of public R&D funds. The opposite is visible in France where Ile de France stands out on all three indicators: R&D, patents and GDP per capita.

The evolution of technology indicators over time Turning to trends over time, beta-convergence analysis shows that there has been absolute convergence in R&D figures and in total patent applications. This overall convergence has, however, been stronger in terms of patent applications than in total R&D expenditures40. As the coefficients in tables 8.1 to 8.4 show, the same has occurred with government R&D expenditures which have converged at a higher speed than the other technology indicators.

Table 8.1 β-convergence of the distribution of total R&D expenditure among EU regions, 1990-2000

Absolute value of T-statistics in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%

40 Data on patent applications do not identify the nationality of the company or the university where the innovation to be patented was produced. Therefore multinationals from big advanced economies may be producing innovations that belong to them, but filing the patents in the country where they are going to use that new technology. Many of the current convergence in patent applications only responds to this process. The nationality of the ‘inventors’ may remain the same, while the distribution of patents applications becomes more equally distributed across regions only as result of the expansion of these multinationals. Unfortunately it is impossible to separate out this effect given the data available. 41 Note that the difference between estimation (3) and (4) is that in model (4) variables have been transformed in relation to the country average, in order to control that convergence could have a national origin or could be a purely regional effect.

Dependent variable: Ln(yi,t)- Ln(yi,t-1) where yi,t is Total R&D Expenditure as a % of GDP in region i and year t.

GLS Within Group(1)

Within Group(2)

Within Group (3)

Within Group (4)41

Lag -0.160*** (-8.85)

-0.406*** (-6.93)

-0.405*** (-6.91)

-0.406*** (-6.92)

-0.413*** (-6.89)

Constant -0.375*** (-8.31)

-1.974*** (-7.07)

-1.973*** (-7.08)

-0.982*** (-6.95)

-0.961*** (-7.01)

Significance of Region Dummies

-- 3.30*** F(151, 1078)

3.06*** F(151, 1068)

2.49*** F(144, 168)

3.01*** F(144, 168)

Significance of Time Dummies -- -- 0.91 F(10, 1068)

0.92 F(10, 1068)

0.95 F(10, 1068)

Significance of country dummies

-- -- -- 6.78*** F(10, 1068)

5.98*** F(10, 1068)

R2 0.082 0.284 0.291 0.290 0.324 Observations 1487 1234 1234 1234 1234 Arellano and Bover (1990) test 1.879 F(1, 2676)


127

Table 8.2 β-convergence of the distribution of Patent applications among EU Regions, 1990-2000


Table 8.3 β-convergence of the distribution of Government R&D Expenditure among EU Regions, 1990-2000


Dependent variable: Ln(yi,t)- Ln(yi,t-1) where yi,t is Patent applications per million people in region i and year t.

GLS Within Group

(1)

Within Group

(2)

Within Group

(3)

Within Group

(4)

Lag -0.04***

(-4.48)

-0.31***

(-9.36)

-0.58***

(-13.25)

-0.58***

(-13.25)

-0.60***

(-12.68)

Constant 0.23***

(5.96)

0.83***

(2.71)

1.79***

(5.69)

2.44***

(13,40)

-0.05***

(-2,41)


- 2.17***

F(203,1835)

1.86***

F(204,1835)

1.74***

F(192,1835)

1.76***

F(192, 1835)

Significance of Time Dummies - - 32.93***

F(9,1835)

32.93***

F(9, 1835)

1.83**

F(9, 1835)


- - - 13.36***

F(12, 1835)

9.30***

F(12, 1835)

R2 0.04 0.19 0.39 0.39 0.8

Observations 2050 2050 2050 2050 2050

Arellano and Bover (1990) test 4040.17 F(1,4302)

Dependent variable: Ln(yi,t)- Ln(yi,t-1) where yi,t is Gov’t R&D Expenditure as % of GDP in region i and year t.

GLS Within Group

(1)

Within Group

(2)

Within Group

(3)

Within Group

(4)

Lag -0.240***

(-10.28)

-0.567***

(-8.76)

-0.668***

(-9.99)

-0.666***

(-9.99)

-0.691***

(-12.11)

Constant -0.012

(-0.78)

0.575***

(7.06)

0.363***

(5.06)

0.632***

(8.89)

0.641***

(8.92)


-- 4.55***

F(156, 1020)

3.08***

F(156, 1010)

2.84***

F(145, 1010)

3.31***

F(145, 1010)

Significance of Time Dummies -- -- 7.95***

F(10, 1010)

7.96***

F(10, 1010)

9.96***

F(10, 1010)


-- -- 9.68***

F(11,1010)

9.75***

F(11,1010)

R2 0.135 0.324 0.372 0.373 0.392

Observations 1400 1178 1178 1178 1178

Arellano and Bover (1990) test 1.678 F(1, 2455)


128

Table 8.4 β-convergence of the distribution of Income per capita among EU Regions, 1990-2000

Absolute value of T-statistics in parentheses

* significant at 10%; ** significant at 5%; *** significant at 1%

Trends in total and government R&D diverged over the 1990-2000 period, largely because of the rapid expansion of private R&D spending as a share of total R&D expenditures. During the second half of the nineties while private R&D investment grew, public R&D expenditures stagnated as governments struggled to consolidate public budgets.

Sigma-convergence analysis yielded very similar results to those provided by the previous beta-convergence analysis, with only one exception (see figure 8.1). While both the beta and sigma- convergence analyses point to a convergence in patent applications and public R&D expenditures, particularly between 1996 and 2000, the picture for the evolution of total R&D expenditures is more heterogeneous. The findings suggest beta-convergence and sigma-divergence over time. The beta convergence indicates that regions with lower shares of total R&D in 1990 have increased their R&D expenditures at higher rates than those regions which started at higher levels. But sigma-divergence implies that the dispersion from the average share of total R&D spending has increased. The evidence of beta-convergence may reflect Galton’s fallacy, i.e. the tendency for regions to regress towards the mean (Quah, 1993).

To complement and illustrate the results provided by the beta-convergence and sigma-convergence analyses, we also plot Tukey’s box-and whisker plots for each technology indicator. The Tukey’s box-and-whisker plot is a histogram-like method of displaying data, where the box ends at the quartiles Q1 and Q3, and the statistical median is represented by a line that crosses the box. The farthest points that are not outliers (i.e. that are within 3/2 times the interquartile range of Q1 and Q3) are connected to the box by the ‘whiskers’, and for every point that is more than 3/2 further away the end of the box, we draw a dot.

Just by looking at figure 8.1 it is easy to arrive at a very interesting finding: at the beginning of the nineties the fact of measuring the technology gap using different indicators really made a difference. In 1990, the existing technology gap measured by the sigma in patent applications (1.6) was twice the technology gap if the indicator to be used was total R&D

Dependent variable: Ln(yi,t)- Ln(yi,t-1) where yi,t is GDP per capita in region i and year t.

GLS Within Group

(1)

Within Group

(2)

Within Group

(3)

Within Group

(4)

Lag -0.02***

(-3.91)

-0.03***

(-2.97)

-0.26***

(-6.79)

-0.26***

(-6.79)

-0.19***

(-8.33)

Constant 0.23***

(4.85)

0.35***

(3.32)

2.68***

(6.93)

2.63***

(7.06)

28.33***

(8.32)


- 1.30

F(204,2049)

2.18

F(204,1835)

1.37

F(192,1835)

0.40

F(193,1835)

Significance of Time Dummies - - 40.49

F(9,1835)

40.49

F(9,1835)

145.90***

F(9,1835)


- - 4.91

F(12,1835)

8.19

F(12,1835)

R2 0.03 0.17 0.51 0.51 0.59

Observations 2050 2050 2050 2050 2050

Arellano and Bover (1990) test 1643.44 F(1,4097)


129

expenditures (0.8). In 2000 the technology gap that both indicators measure is much closer, since in that year the sigma for patent applications was 1.6 while the sigma for total R&D expenditures was 1.3. The dynamic evolution of the different distributions under study that was described in previous paragraphs is confirmed again when the three Tukey’s box and whisker figures are plotted.

As can be seen in the first plot of figure 8.2, average total R&D spending has increased over time, as well as its degree of dispersion. However, the average level of public R&D has remained almost constant over the past decade and so has its degree of dispersion (plot 2). Finally, the average number of patent applications has increased slightly in the last decade, while its dispersion diminished, notably in 1995 and again in 2000.

Figure 8.2 Tukey’s box and whisker plots of Total R&D, Government R&D and Patent Applications, 1990-2000

Figure 8.1. Sigma Convergence of Technology Indicators among EU Regions (1990-2000)

0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

1,8

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Years

Standard Deviation of the Logarithms

Patent Applications

Gov't R&D Exp

Total R&D Exp

.02

7.195 Total Expenditure in R&D by all

19891990

19911992

19931994

19951996

19971998

19992000

0

3.075 Government Expenditure in R&D

19891990

19911992

19931994

19951996

19971998

19992000


130

It is interesting to analyze the shape of the different Tukey’s box plots because they offer some new information on the sources of the existing disparities in the distribution of each technology indicator. The fact that dots are above the upper whiskers in the plots for total and public R&D expenditures implies that most regional disparities in R&D expenditures originate in regions that are almost always above the regional average. For patent applications, the problem is the opposite: there is a significant number of regions that file very few patent applications and are therefore way below the regional average. Interestingly (figure 8.4), income disparities seem to be in between.

The convergence analysis provides empirical evidence showing that public and total spending in R&D have followed different dynamics over the last decades. As figure 3 shows, the main reduction in regional income disparities occurred at the beginning of the nineties. The implications for cohesion and growth are now examined.

8.4 Spatial distribution of technology indicators and cohesion Table 8.1 summarises beta and sigma convergence measures for income per capita at regional level in the EU, and reaffirms that there has been convergence. If the beta coefficients in table

-20

24

6

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Patents Applications l l l

-20

24

6

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

-20

24

6

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

-20

24

6

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Patents Applications l l l Patents Applications l l l

Figure 8.3. Sigma Convergence in the Distribution of Income per capita among EU

0 0

0 1

0 1

0 2

0 2

0 3

0 3

0 4

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Years

Standard Deviation of the Logarithms Income per capita


131

8.4 are compared to those in tables 8.1-8.3, we see that convergence in income per capita has been weaker than convergence in some technology indicators. The main reduction in regional income disparities occurred at the beginning of the nineties, but did not alter much subsequently. Examination of the data suggests that the reduction of income disparities across regions arose mainly because of the convergence of poorer regions to the EU average.

Figure 8.4 Tukey’s box and whisker plot of Income per capita, 1990-2000

The relative relationship between the parallel evolution of regional disparities in technology indicators and regional disparities in income per capita can be analyzed by computing Pseudo-Lorenz curves and Pseudo-Gini indices. The Pseudo-Lorenz curve is a useful tool for analysing the regional dimension of technology indicators vis á vis the distribution of income per capita. It compares the accumulated income of the 205 European regions, arranged in descending order of per capita income (the x-axis), with the levels of R&D expenditures or patent applications, also expressed in cumulative terms (in the y-axis). From this curve, it is possible to infer whether the distribution of technology indicators has, implicitly, contributed to economic cohesion (when the curve gets closer or even surpasses the 45º line) or not.

A quick look at figures 8.A.1-8.A.4 shows that while the Pseudo-Lorenz curve (PLC) of total R&D expenditures moved further away from the 45º line between 1995 and 2000, the PLCs of total patent applications and, especially, of government R&D expenditures have moved in the opposite direction. Focusing on the two measures of R&D spending and comparing their evolution, it can be seen that the movements in their PLCs, too, have been in opposite directions. The PLC of government R&D expenditures has moved closer to the 45º line, signaling that the distribution of this technology indicator has contributed more to territorial equity than has the indicator of total R&D spending.

These results are confirmed by the evolution of the Pseudo-Gini indices in figures 8.A.5 and 8.A.6 The distribution of total R&D expenditures has become more unequal over time, in contrast to the trend towards more equal government R&D expenditures that coincided with greater equality in the distribution of income (figure 8.5).

To assess the impact of changes in the regional dispersion of technology indicators on the regional dispersion of income per capita, the following equation, where all dependent and independent variables are transformed into their sigma-dispersion indexes (logarithm of their respective standard deviations), was estimated.

SigmaINCOME = SigmaPATENTS + SigmaTERD + SigmaGERD + ε (1)

8.5

99.

510

10.5

11

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Income per Capita PPSl

8.5

99.

510

10.5

11

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

8.5

99.

510

10.5

11

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Income per Capita PPSl Income per Capita PPSl


132

The equation is estimated by OLS and all results are reported in table 8.5. These results show that any increase in the dispersion of total R&D spending or in the dispersion of patent applications increases the dispersion in the income distribution among regions in all EU and among regions by country. The influence of public R&D spending on economic cohesion is weaker42 but works in a similar direction.

A 10% increase in the dispersion of total R&D expenditures such as occurred between 1998 and 2000, produces a 0.3% increase in the dispersion of income distribution across regions. However, a 10% decrease in the dispersion of public R&D expenditures (as occurred between 1993 and 1995 and again between 1997 and 1999) produced a reduction of 0.1% in both cases in income dispersion across regions in Europe.

The experience of the nineties shows that while the distribution of total R&D spending became more unequal (led by an unequal expansion of its private component) it was offset by public R&D spending and patent applications. As the distribution of patents and public R&D converged, income per capita converged too. Since the only indicator that can be directly affected by policy-makers is the share of public funds dedicated to R&D activities, a tentative conclusion is that government R&D helped somewhat during the 1990s to mitigate the divergence in economic performance resulting from the impact of other technology variables.

8.5 Spatial distribution of technology indicators and economic growth The relationship between the three technology indicators and economic growth is examined, first, by analysis of correlations between the three technology indicators and income per capita, then by estimating some simple regressions to look for one-way causality between these variables; and finally by a multiple regression for the impact of technology on economic growth (measured as the annual change in GDP per capita).

The correlation findings are clear. As table 8.A.1 shows (Spearman’s and Kendall’s) income per capita and patent applications are very strongly correlated (0.7) and persistent over time, as are income per capita and total R&D spending (0.4). But the correlations between income per capita and public spending R&D are much weaker (0.15) and fall over time. Additionally, the correlation between total and public R&D expenditures, although persistently strong (0.7), has weakened in the last ten years. This confirms again the progressive divergence between R&D decisions made by the public sector and those made by the private sector.

42 Coefficients for public R&D spending are only statistically significant at the 80% confidence level.

Figure 8.5 Gini Index of the Distribution of Income per capita among

0,145

0,15

0,155

0,16

0,165

0,17

0,175

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Income per

capita


133

The strong correlation between patent applications and income per capita fades away in a simple one-way test of causality between patent applications and the growth of income per capita (see figure 8.6). Although positive, the relationship is not statistically significant. In fact, this weak causal relationship between patents and economic growth after such strong bilateral correlations implies that the relationship goes the other way around, that is, from high rates of economic growth to patent applications. This suggests that innovation and patent applications mainly take place in regions which are already rich in relative terms.

-.0

4-.0

20

.02

.04

.06

Gro

wth

of L

og o

f Inc

ome

per c

apita

(199

9-20

00)

-6 -4 -2 0 2Log of Patent Applications per million people (1999)

coef = .00094027, se = .00078655, t = 1.2

Source: European Commission: Regio Database and own elaboration

Income growth and Patent Applications among EU Regions

Figure 8.6 Simple regression Patents and Economic growth The causal relationship between total R&D spending and economic growth is much clearer (see figure 8.7). Without controlling for any other factor, about 10% of the total variation found at regional level in the economic growth between 1999 and 2000 is explained by the differences observed in total R&D expenditures in 1999. But R&D investment, although necessary, is not a sufficient condition for growth. It is only when basic research improves the level of technical innovation in a region that growth appears immediately in that region.

As figure 8.7 shows, in 2002 almost 40% of the total variation observed in income per capita levels among different EU regions was explained by the degree of relative innovation that took place in these regions. This confirms the findings of the European Commission which having used the same indicator for a related analysis affirmed; ‘this suggests a positive relation between a region’s innovative performance as measured by its Revealed Regional Summary Innovation Index43 and its economic performance.’ (EIS, 2002: 12)

43 The Revealed Regional Innovation Index (RRSII) tries to take into account both the region’s innovative performance relative to the EU mean and the region’s relative performance within the country. Because this index is only available for years 2001 and 2002, this informative innovation index has only been used as an indicator of technology policy for this simple regression analysis, but has been excluded from the dynamic convergence analysis of section 3 and the panel analysis of the final regressions in this last section.


134

-.04

-.02

0.0

2.0

4.0

6G

row

th o

f Log

of I

ncom

e pe

r cap

ita (1

999-

2000

)

-2 0 2 4 6Log of R&D expenditures (% GDP, 1999)

coef = .00125895, se = .0010832, t = 1.16

Source: European Commission: Regio Database and own elaboration

Income growth and R&D expenditures among EU Regions

Figure 8.7 Simple regression: Total R&D and Economic growth Finally, the joint role that all those technology indicators have in explaining economic growth measured by the annual change in income per capita can be discerned by estimating the following equation:

( ) ( ) ( ) tiipttgigigtfiftlilit cohesionTITIGDPGDP εφωδβββα +++++++=∆ −−− 1,1,1,0 * (2)

where:

itGDP∆ : is the annual change in income per capita

βf: is the coefficient of the effect that all technology indicators (patent applications, total R&D spending, and public R&D spending) have on regional economic growth.

βg: is the coefficient of the effect that all technology indicators (patent applications, total R&D spending, and public R&D spending) have in regional economic growth in cohesion countries (Spain, Greece, Ireland and Portugal).

tδ : is the time dummy ; iφ : is the region dummy; pω : is the country dummy

This equation is estimated through GLS with robust standard errors. Results are reported in table 8.6.

As can be observed, real convergence is once again confirmed: the lower the existing regional income per capita in t-1, the higher the subsequent economic growth. In addition, the contribution of patent applications in t-1 to subsequent economic growth in year t is very positive. An increase of 1% in patent applications produces an increase in regional economic growth of 0.017. However, this effect is exactly the opposite for cohesion countries. This can be interpreted as follows: where the stock of patents is very low, one additional patent is not sufficient to start economic growth. Instead, the innovative effort required to produce an isolated patent could divert resources from more productive activities and thus be economically damaging.


135

Table 8.6 The influence of Patents and R&D on Economic Growth Change in

GDP per capita Change in GDP per capita

Income t-1 -0.007 -0.011 (-0.54) (-0.80) Patents t-1 0.008** 0.017*** (2.20) (3.35) Patents t-1*Cohesion country 0.013** (2.05) Total R&D Expenditure t-1 0.006* 0.005* (1.82) (1.74) Total R&D Expenditure t-1*Cohesion Country 0.003 (0.16) Gov´t R&D Expenditure t-1 -0.019 (0.041***) -0.014 (0.032**) (-1.06) (3.41) (-0.62) (2.10) Gov´t R&D Expenditure t-1*Cohesion Country -0.023 (0.028*) (-0.54) (2.02) Constant 0.113 0.096 (0.89) (0.72) Observations 1244 1244 R-squared 0.199 0.205 F (164, 1079); F (167, 1076) 1.90 2.04 Prob>F 0.000 0.000

Absolute value of t-statistics in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%

Note 1: Regression with robust standard errors. Coefficients for time, region and country dummies were included in the regression but are not reported here. All variables in Logarithms.

Note 2: In cursive letter we report the 4-year lag coefficients of public R&D expenditures.

More importantly, the role of total R&D expenditures is also strongly positive for all countries (including cohesion ones). The crucial impact of total R&D for economic growth is somewhat at odds with the insignificant effect that public R&D spending has on short run economic performance. However, this weak (even negative) short run impact of public R&D on economic growth, turns into a very strong and positive influence in the medium run. An increase of 1% in public R&D spending today is likely to increase the rate of growth by 0.04% in four years from now. This 4-years lagged positive effect of public R&D on economic growth holds also for cohesion countries, an important finding given that public R&D spending has to compensate for low private R&D initiatives in these regions.

8.6 Regional R&D policies At the national level, there have been divergent innovation strategies, in spite of a common tendency to increase investment in all types of R&D and innovation activities. Since both national and European funds for R&D initiatives are usually distributed across regions on the basis of ‘excellence criteria’, the main recipients of these funds are the small number of regions which have a technological comparative advantage, typically in the most developed countries of the EU. It is also clear that only few regions have attained strong competitive positions in high-tech industries and that support for national excellence can be at odds with cohesion aims. Therefore, the presumption is usually that the impact of technology policies on economic cohesion is most likely to be negative.

Consistent with ‘Lisbon’ aims, most countries have targets for raising the level of R&D intensity and have launched ambitious policy measures with the aim of increasing R&D funding and technology development. A good example is Austria, which is aiming to increase the R&D/GDP ratio to 2.5% by 2005. This goal could be achieved by a combination of public R&D funding and stimulation of private investment, and by increasing research activities in


136

polytechnics and the funding of competence centres. As a consequence of this policy, the corresponding provincial R&D budgets (6.4% of national R&D expenditures) nearly doubled between 1989 and 2001 to reach 0.12% of GDP in 2002. Spending per capita lies in a range from 7 Euro in Lower Austria and Burgenland to 50 Euro or 1.6% of GDP in Vorarlberg.44 Particularly noteworthy is the rapid growth of Carinthian R&D expenditures and a surge in Upper Austrian R&D spending in 2000. Provincial R&D funds still flow mainly to the public sector (public hospitals absorb a high share), and to universities.

Denmark plans to boost the R&D/GDP ratio to 3% by 2010, of which 40% should be financed through public funding, including an increase in the basis funding of research institutions. In Belgium, substantial public investment in R&D has helped to achieve an R&D intensity target of 2% by 2002. And, in Finland, higher public funding has contributed to an R&D intensity exceeding 3.2% of GDP by 2002. Despite these efforts, R&D investments within EU countries remain highly concentrated in central locations in each country, suggesting that intra-national differences may largely explain intra-EU disparities in R&D, technology and innovation activities.

Clustering Austria provides a good illustration of the clustering concept. In Styria, the provincial government overhauled its development strategy in 1996 and created an institutionalized ‘cluster’ network linking and synchronizing the Styrian automotive industry, which proved extremely successful. Upper Austria followed suit with a comprehensive provincial strategy and incremental increases in technology and networking subsidies. This provoked all provincial governments and their economic policy authorities to re-orient their support policies and focus on regional cooperation in innovation. The move was complemented by the introduction of specialized polytechnic colleges and by federal activities to create ‘impulse/competence centers’45. Today, every province hosts ‘clusters’ to a varying degrees (from small company networks to proper technology districts) and most regional authorities implement coherent development concepts in this respect.

In Germany, the InnoRegio programme concentrates on financing innovations in the Eastern Länder. It promotes innovative initiatives based on regional clusters of competence and production. So far 9 support for core growth areas has been agreed (40 million €). Additionally, about 40 laboratories for innovation and foundation were established in universities and research institutes in the Eastern Länder States (25.6 million €); and innovation forums with a total number of 444 regional initiatives applied for support.46 But this tendency to concentrate technological R&D and innovation activities in some intra-national ´clusters´ is not exclusive to federal states. In countries such as Italy or the UK, where the central authorities manage the design of technology policies and the regional allocation of funds, the polarization strategy has been very similar.

Nevertheless, the spatial balance is very uneven. Italy’s FAR programme, one of the two main funds for supporting both national and regional R&D, allocated 90% of its resources to Centre-North regions in 2000-2001. The regional concentration of funding is also very high: a third went to Lombardia alone, and 72.2% to just 5 regions (Lombardia, Emilia-Romagna, Veneto, Lazio and Piemonte). A second fund (FIT) is still more concentrated, with 94.3% of funding going to the Centre-North and, again, five regions (Lombardia, Piemonte, Veneto, Emilia-Romagna and Lazio) benefiting most. During the entire funding period of FIT, only 131 projects (3.4% of the total) were funded in the Mezzogiorno against 3768 in the Center-North.

44 See, Hutschenreiter and Sieber (2002: 25-34). 45 Steiner (2002: 752-755). 46 http://www.innoregio.de/foerderung.php


137

The UK government has recognized in recent policy documents both the need and the importance of R&D for regional economies. The North East, Yorkshire, London, Wales and Scotland receive a disproportionate amount of the R&D spending for higher education, while the South West, West Midlands and Scotland receive a disproportionate amount of government R&D spending (partly related to spending on defence R&D). In terms of business R&D, London seems to perform relatively poorly, as do Scotland and Wales, whereas the highest concentrations are in East Anglia and the South East (the areas surrounding London) and the North West. R&D spending in services is almost completely concentrated around Greater London.

8.7 Commentary and key findings A number of interesting conclusions about the relationship between technology policy and cohesion emerge from this analysis. With public funding contributing around 1-1.5% of GDP, technology policies are a much more substantial contribution to economic development than territorial policies. Yet large disparities persist and it is inauspicious for convergence that the poorer Member States tend to commit less to such policies.

It is clearly the case that the prosperous regions continue to dominate technological activity and that, other things being equal, this gives them an advantage that could widen GDP disparities in the medium- to long-term. The weight of evidence is that, in most Member States, technology policies exacerbate rather than reduce regional GDP disparities and tend to add to the competitive advantage of richer regions. However, the role of policy in this area has been ambiguous. Some technology indicators converged among regions during the nineties (especially public R&D spending), and this ran parallel to a real (though softer) convergence in income per capita levels. Yet total R&D expenditures have diverged across regions over time, as a result of an asymmetric expansion of private R&D activities during the second half of the nineties.

Second, total R&D spending increases economic growth, especially if this R&D activity is quickly transformed into new patent applications. Since innovation is the real key for economic growth, it is only when R&D spending is translated into new patent applications and innovation is extensively diffused that economic performance improves. This positive effect on growth is not exclusive of total R&D expenditures, but also applies with a 4-year delay to public R&D initiatives.

Key findings are: More investment in R&D is associated with more rapid growth, although it is also

important to acknowledge the importance of diffusion and take-up of innovation.

Member States spend substantially more on technology policies than on spatially targeted economic development policies.

Technology policies typically focus on national competitiveness objectives, with the result that they tend to favour regions with an established lead in technological capability.

Private R&D is very heavily concentrated in favoured regions. Public R&D spending is less regionally concentrated, to some extent offsetting the impact of private, although the overall disparities are very pronounced, and the disparities in regional R&D and patent activity are much greater than disparities in GDP per head.


138

Figure 5. Pseudo-Lorenz Curve of the Distribution of Total R&D Expenditures among EU regions (1995 and 2000)-

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9

Accumulated Population

Accu

mm

ulat

edTo

tal E

xpen

ditu

re in

R&

D(%

GD

P)

%terd1995

%terd200045ª-line

1

Figure 5. Pseudo-Lorenz Curve of the Distribution of Total R&D Expenditures among EU regions (1995 and 2000)-

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9


Accu

mm

ulat

edTo

tal E

xpen

ditu

re in

R&

D(%

GD

P)

%terd1995

%terd200045ª-line

1

Figure 6. Pseudo-Lorenz Curve of the Distrubution of Patent Applications among EU Regions (1990,1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1


19901995200045º Line

Ac c

u mul

a te d

Pat

e nt s

Figure 6. Pseudo-Lorenz Curve of the Distrubution of Patent Applications among EU Regions (1990,1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1


19901995200045º Line

Ac c

u mul

a te d

Pat

e nt s

Figure 7. Pseudo-Lorenz Curve of the Distribution of Government Expenditures among EU regions (1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9


Acc

umul

ated

Gov

ernm

enEx

pend

iture

(%G

DP)t

%terd1995

%terd2000

45ª - line

1

Pseudo-Lorenz Curve of the Distribution of Government Expenditures among EU regions (1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9


Acc

umul

ated

Gov

ernm

enEx

pend

iture

(%G

DP)t

%terd1995

%terd2000

45ª - line

1

Figure 7. Pseudo-Lorenz Curve of the Distribution of Government Expenditures among EU regions (1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9


Acc

umul

ated

Gov

ernm

enEx

pend

iture

(%G

DP)t

%terd1995

%terd2000

45ª - line

1

Pseudo-Lorenz Curve of the Distribution of Government Expenditures among EU regions (1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0, 7 0,8 0,9


Acc

umul

ated

Gov

ernm

enEx

pend

iture

(%G

DP)t

%terd1995

%terd2000

45ª - line

1

Figure 8.A.2

Figure 8.A.3

Figure 8.A.1


139

Figure 8.A.5. Gini Index of Technology Indicators among EU Regions( )

0

0 1

0 2

0 3

0 4

0 5

0 6

0 7

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Patent ApplicaGov't R&D ExpTotal R&D Exp

Figure 8.A.6 . Gini Index of the Distribution of Income per capita among EU Regions (1990-2000)

0,145

0,15

0,155

0,16

0,165

0,17

0,175

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Income per cap

Figure 8. Lorenz Curve of the Income Distrubution among EU Regions (1990,1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1


Acc

umul

ated

Inco

me

19901995200045º lIne

Figure 8. Lorenz Curve of the Income Distrubution among EU Regions (1990,1995 and 2000)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1


Acc

umul

ated

Inco

me

19901995200045º lIne

Figure 8.A.4


140

Table 8.A.1 Spearman and Kendall’s correlations

PATENTS AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Spearman Rho 0.674*** 0.693*** 0.649*** 0.688*** 0.713*** 0.738*** 0.755*** 0.743*** 0.726*** 0.723*** 0.722*** 0.716***

Regions 205 205 205 205 205 205 205 205 205 205 205 205

TOTAL R&D EXPENDITURE AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Spearman Rho -- 0.224** 0.479*** 0.358*** 0.298*** 0.348*** 0.433*** 0.300*** 0.401*** 0.382*** 0.394*** 0.393***

Regions -- 63 93 94 96 116 161 160 161 161 161 161

GOVERNMENT R&D EXPENDITURE AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Spearman Rho -- 0.249** 0.256** 0.226** 0.148* 0.166* 0.144* 0.134* 0.136* 0.102 0.106 0.086

Regions -- 65 110 111 117 117 165 165 165 165 173 173

TOTAL R&D EXPENDITURE AND GOVERNMENT R&D EXPENDITURE Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Spearman Rho -- 0.777*** 0.718*** 0.727*** 0.756*** 0.764*** 0.718*** 0.544*** 0.678*** 0.685*** 0.678*** 0.665***

Regions -- 62 88 87 93 113 161 157 161 161 161 160

PATENTS AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000


141

Kendall’s Tau 0.489*** 0.507*** 0.472*** 0.511*** 0.534*** 0.552*** 0.564*** 0.562*** 0.549*** 0.541*** 0.540*** 0.534***

Regions 205 205 205 205 205 205 205 205 205 205 205 205

TOTAL R&D EXPENDITURE AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Kendall’s Tau -- 0.149* 0.335*** 0.258*** 0.222*** 0.252*** 0.302*** 0.207*** 0.278*** 0.265*** 0.273*** 0.269***

Regions -- 63 93 94 96 116 161 160 161 161 161 161

GOVERNMENT R&D EXPENDITURE AND INCOME PER CAPITA Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Kendall’s Tau -- 0.171** 0.174** 0.152** 0.095 0.112** 0.091* 0.087* 0.086* 0.067 0.069 0.056

Regions -- 65 65 110 111 117 117 165 165 165 165 173

TOTAL R&D EXPENDITURE AND GOVERNMENT R&D EXPENDITURE Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Kendall’s Tau -- 0.565*** 0.526*** 0.542*** 0.572*** 0.590*** 0.543*** 0.389*** 0.499*** 0.503*** 0.500*** 0.487***

Regions -- 62 88 87 93 113 161 157 161 161 161 160

* significant at 10%; ** significant at 5%; *** significant at 1%


142

9. Regional Distribution of National FDI and its Impact on Cohesion In some Member States, inward foreign direct investment accounts for a large share of the capital stock and of annual capital formation, although there are surprisingly large variations between Member States in the importance of foreign capital. As can be seen from table 9.1 the spread for the annual inflow in recent years is from just 2% of gross domestic fixed capital formation (GFDCF) to over 100%, while the stock of inward investment ranges from just 11% of GDP in Italy and Greece to a peak of 130% in Ireland.47 The degree to which regions attract FDI is, consequently, likely to be a factor shaping economic cohesion, so that the incidence of policies affecting FDI will bear on development prospects. In many Member States, considerable efforts are devoted to drawing in FDI and it is not unusual to find both national and sub-national agencies involved in so doing.

Table 9.1 Scale of inward foreign direct investment

Flow of inward FDI as a proportion of GDFCF, %

Average 1998-2002

Stock of inward FDI as a proportion of GDP, %

Average 2000-01 Austria* 9 17 Belgium & Luxembourg* 148 81 Denmark 43 42 Finland 22 21 France 17 21 Germany 17 24 Greece 2 11 Ireland 76 130 Italy 5 11 Netherlands 52 32 Portugal 14 29 Spain 16 27 Sweden 60 41 UK 31 35

Source: UNCTAD – Foreign Direct Investment database

Note: * = Average for 1998-2001. The figure for Belgium/Luxembourg may be somewhat misleading because of flows associated with the fund management activities of Luxembourg financial intermediaries.

9.1 Influences on the location of FDI As indicated in chapter 2, section 5, above, industrial agglomerations appear to be important in attracting inward FDI in particular industries. These are the sectors which tend to remain concentrated in the core EU regions (hereafter, ‘CC’ sectors) over time. These include Motor Vehicles, Aircraft and Chemicals (other than Industrial Chemicals and Pharmaceuticals). Regional policy is probably incapable of overturning the importance of Marshallian agglomerations for these sectors (unless regional grant levels were to be pitched at extremely high levels, which is both politically infeasible and economically inefficient).

There are certain industrial sectors for which pre-existing agglomerations are of less importance however and, not surprisingly, these sectors are the ones currently of primary importance in peripheral regions. Indeed the European Commission report One Market, One Money (1990)

47 There is considerable annual volatility in these figures and, especially for the smaller countries, the data for a single year can be strongly affected by the timing of a few large inflows. The GDFCF figure can exceed 100% because the numerator is a gross flow and the latter a value-added concept. It is also important to note that the countries with large inflows typically also have large outflows.


143

reported that firms in peripheral regions identified infrastructural deficiencies in the areas of education and training, transport and communications, and the supply and cost of energy as more important impediments to their development than geographical aspects of peripherality such as the proximity of suppliers and customers.

These are of two types of sectors for which pre-existing agglomerations are of less importance: (i) the initially dispersed industries which have become more concentrated in the peripheral EU regions over time. These we term the DC sectors (i.e. formerly Dispersed sectors which have become more Concentrated). These include Textiles, Clothing, Footwear and Furniture, sectors which are characterised by low increasing returns, are low tech and of low skill intensity and which have a high share of resource inputs; and (ii) the formerly concentrated sectors which have become more dispersed (the CD sectors). These include Office and Computing Machinery, Radio TV and Communications, Machinery and Equipment, and Professional Instruments. Increasing returns and intra-industry linkages are not as important for these sectors as they are for the ones which have remained concentrated in the core, but these sectors are nevertheless quite skill intensive.

What is the relationship between FDI and these particular sectors? To analyse this, we look at the share of each group of sectors in foreign employment (Fi/F) in the periphery countries for which we have data (Ireland, Spain and Portugal), relative to its share in total employment (Ti/T). This is shown in Table 9.2.

Table 9.2: Share of sector i in foreign employment (Fi/F) relative to its share in total employment (Ti/T)

Ireland Spain Portugal CC 1.53 2.08 3.9 CD 1.69 1.26 2.1 DC 0.82 0.33 0.5 Source: Barrios, Barry and Strobl (2003).

The data contained in this table illustrate that sectors which have remained concentrated (largely in the core) tend, in the periphery, to be under foreign ownership. The same holds true of those which were formerly concentrated in the core but which have now become more dispersed. The lower-tech sectors which were formerly dispersed but which are now more concentrated in the periphery, by contrast, tend to be primarily under domestic ownership.

The Irish data suggest that grant payments in lagging regions will impact primarily on their ability to attract the latter (DC) sectors. By contrast, attracting the FDI-intensive CD sectors is more dependent on adequate levels of infrastructure and human capital as well as regional innovation policies. These are precisely the areas towards which much EU regional aid is directed.48 Accordingly, it appears to be the case that EU regional aid, to the extent to which it directs a greater share of resources towards lagging regions than purely national programmes do, will be particularly effective in the long run in drawing FDI towards the lagging regions.49

Some of the results discussed in chapter 2, section 5 suggest that, at least in some countries and in some time periods, regional policies can have statistically significant effects in promoting a regional dispersal of FDI. The more important question relates to their economic rather than their statistical significance. Are the policy packages in place large enough or their effects strong

48 Barry (2003) shows that EU regional aid has facilitated a general convergence in human-capital stocks in the EU periphery countries, has facilitated some convergence in innovation levels and has at least prevented further divergence in terms of physical infrastructure. 49 The state aids work analysed in chapter 5 suggest that EU regional aid does indeed direct a greater share of resources towards lagging regions than purely national programmes do.


144

enough to offset the divergence-promoting effects of FDI being attracted to pre-existing agglomerations?

To explore these issues, data were assembled on the regional distribution of FDI within national territories. These data are of two types. Some countries such as Spain, Germany and Portugal report on the regional distribution of FDI in monetary terms, while others – including France, Finland, Italy and Ireland provide data on the distribution of employment in foreign-owned manufacturing sectors across regions. For the UK we have data on employment in foreign-owned manufacturing and non-manufacturing activities.50

We also have, from the Regio database, evidence on the distribution of total industrial employment across regions. Accordingly we calculate the ratio of the regional share of national FDI relative to each region’s share of national industrial employment. If this ratio is high, it suggests that the region is a disproportionately attractive location for foreign direct investment. These data, along with regional GDP per head (in PPP terms) relative to the EU average for the year 2000 are presented in Table 9.A.1

We next propose to determine how this left hand side variable, which measures a region’s disproportionate attractiveness for FDI, is correlated with its level of income per head. A negative correlation would suggest that factors other than the tendency of foreign industry to locate close to existing agglomerations in richer regions is driving the location of foreign industry. This could be taken to imply that policies oriented towards the poorer regions were operating successfully in terms of attracting disproportionate FDI shares. A positive correlation on the other hand would suggest that the regional policies in place were insufficient or powerless to offset these tendencies. These correlations are reported in Table 9.3.

Table 9.3 The association between income and FDI

Correlation coefficients Germany 0.612757 Spain 0.574435 France 0.42678 Ireland 1 Italy 0.676933 Portugal 0.885175 Finland 0.307289 UK 0.891081 All countries in sample 0.612757

When all regions in the data set are grouped together, a correlation co-efficient of 0.6 emerges, suggesting that FDI inflows are disproportionately attracted towards richer regions. It would be dangerous to read too much into the differences in correlation co-efficients that appear for each national economy, given the differences in how FDI is measured.51 In each individual country case, however, the co-efficient is also positive.

What if causation is believed to run in the opposite direction however, from FDI presence to regional growth? This perspective proposes that a region’s success in attracting a 50 The data coverage for the various countries is as follows: Spain 1996-2001; Germany 1994-2000; France 1993 and 1999; Finland 2001; Ireland 2000; Italy 2000; UK 1998 and 2000, and Portugal 1996-1998. 51 As noted above, the data for some countries refer to FDI stocks or flows; in others, the data are for employment in foreign-owned industry. Furthermore, in some cases the FDI data refer to both manufacturing and services, while the foreign-sector employment data typically refer only to manufacturing.


145

disproportionate share of FDI will give it a high level of GDP per head, undermining the argument that it is the level of economic development of a region that makes it an attractive location for FDI. To analyse this we ran correlations between GDP per head measured at some time in the past (1988) with the variable representing current FDI success. Positive correlations between these variables would suggest that historically richer regions attract disproportionate shares of FDI today.

Table 9.4, Correlations based on historical GDP per capita data Correlation coefficients

Germany 0.863357 Spain 0.456755 France 0.475252 Ireland Italy 0.742775 Portugal 0.863661 Finland 0.164954 UK 0.952593 All countries in sample 0.567326

Once again we see that all the correlations are positive, illustrating that the historically richer regions record greater successes at attracting high FDI shares today. The co-efficients on the whole sample in the two different eras cannot be compared, as Ireland is dropped from the historical sample (having been regionalised only in 1999), while post-unification Germany obviously contains a greater and more diverse number of regions than heretofore. The latter correlations are higher than the former ones for Germany, France, Italy and the UK, while they are lower for Spain, Portugal and Finland. This should not necessarily be taken to indicate that regional policies in these countries have attained some degree of success in distributing FDI towards historically poorer regions, as a number of other factors, including infrastructure and human capital stocks (supported to some extent at least by EU aid) have changed in the interim.

The overall conclusion of this analysis is that FDI clearly tends to flow disproportionately towards richer regions, and must therefore be seen as a factor working in the direction of ‘cumulative causation’, resulting in regional divergence rather than convergence, and thus detracting from economic cohesion. Thus we can conclude that policies to encourage a more balanced regional distribution of FDI – to the extent to which such policies operate – have not had effects which have been sufficiently strong to promote convergence.

It is important to note however that while national regional policies might not have succeeded in generating a more balanced regional distribution of FDI, they may nevertheless have succeeded in counteracting some of the tendencies towards divergence. To see whether this is the case, we regressed our measure of a region’s ability to attract a disproportionate share of FDI on two variables (i) regional income per head, and (ii) regional aid per industrial job.

Table 9.5 Nature of FDI data Spain Sum of regional aid in 2001 and 2002 France Regional distribution of state aid 1999-02 Italy Regional distribution of aid to business 2000 + 2001 Finland Regional investment aid UK Regional Selective Assistance + Enterprise Grants, April 01-March 02

The quality of data available only allowed this analysis to done for Spain, France, Italy, Finland and the UK. The measure of regional aid is taken from the state aids data reported in chapter 5. The regional aid variables employed are shown in table 9.5. We ran the following regression:


146

Regional FDI share/Industrial Employment share =

Constant + a1 (Regional GDP per head) + a2 (Regional Aid/Industrial Employment)

The regression results are reported below in Table 9.6.

Table 9.6 Regression results for analysis of FDI and state aid (I)

Coefficients t Stat Spain

a1 0.043549 2.450448

a2 0.346683 0.18441 France a1 0.00368 0.739733 a2 -19.4226 -2.21348Italy a1 0.019336 1.831516 a2 0.01058 0.063937Finland a1 0.017123 5.349721 a2 0.412701 0.842369UK a1 0.032477 7.271256 a2 3.714881 2.302998

In most cases the regional income variable is significant, and is in all cases positive. The regional aid variable is positive in all cases except France. It is (statistically) significantly positive in the case of the UK and statistically significantly negative in the case of France.

These results suggest that state regional aid does not generally have much of an impact on the regional distribution of FDI. What of the French and UK cases however? In the UK case it has a positive effect. Does this work towards the promotion of FDI in richer or in poorer regions? This will depend on whether type of region receives disproportionate aid shares. The correlation coefficient between the aid and income variables is –0.53995, indicating that the aid goes disproportionately towards poorer regions, and that it does therefore serve to redistribute FDI towards the poorer regions.

And what of France? In the French case the correlation co-efficient is also negative, attaining a value of –0.45957, which suggests that French regional aid policy may work against the poorer regions. To check that this is not simply due to its acting primarily to generate domestic rather than foreign employment in the poorer regions (thereby raising the denominator in the FDI/Industrial Employment variable) we also ran the regression using Regional FDI share (rather than FDI Share/Industrial Employment Share) with the same right hand side variables.

The regression results are reported in table 9.7, confirming that French regional policy does indeed appear to operate in a perverse way. We conclude then that, of the countries studied, only in the UK case do national regional policies appear to work towards redistributing FDI towards the poorer regions.

9.2 Impacts on FDI of EU versus National Regional Policies In this section we explore whether national policies have possibly interacted with EU policies in redistributing FDI towards the more peripheral regions. Each EU Member State has territorial, state aid or employment policies that discriminate in favour of less-developed regions and that would also be thought to influence the regional dispersion of FDI. For example, as Tondl (2001) points out, investment subsidy limits (co-financed under the SF programmes and permitted under EU competition law) are generally quite high in Objective 1 regions. As noted in chapter 5, above, the distribution of state aid across regions depends on the type of aid. Some aid such as


147

R&D aid is granted mostly in richer regions, although regional state aid is granted mostly to poorer regions. However, within poorer regions, there are considerable variations in the amount of state aid. These amounts do not appear to correlate with the need of regions as measured by their relative income.

Table 9.7 Regression results for analysis of FDI and state aid (II)

Coefficients t Stat Spain

a1 0.43257 2.041786

a2 -4.44944 -0.19854France a1 0.187175 6.405927 a2 -125.991 -2.44478Italy a1 0.314259 1.83519 a2 2.184651 0.813981Finland a1 0.23709 0.634667 a2 -45.5121 -0.79593UK a1 0.206413 4.051012 a2 -12.1192 -0.65858

It is naturally difficult to try to disentangle the effects of national and European policy measures in this areas. Identification would require, at the minimum, temporally-varying data on the distribution of FDI within countries and information on expenditure levels across regions by regional/national and EU policymakers. These demanding minimum requirements for data are met, according to the data at our disposal, only for Ireland. Irish industrial policy has traditionally contained a strong regional component, with certain ‘disadvantaged’ areas being given preferential grant treatment, especially up to the early 1980s. Furthermore, since 1999 there has been some favouritism in the allocation of European funding towards the BMW Objective 1 region (i.e. the Border, Midlands and West region of Ireland). For example, under the 2000 National Development Plan, EU funds of about €62.4 million are allocated for the BMW region compared to €22.1 million for the rest of Ireland.52

Using data from the (Irish state agency) Forfás employment survey and data collected on grant expenditures by Irish policy makers we compiled summary statistics on national grant expenditures (in thousand of 1995 Irish pounds) and the location of FDI, measured both as employment share and as a share of new firm start-ups, across the two regional groups in Tables 9.8 and 9.9.

We calculated the means for three separate periods:

1972-1983, when regional job creation was a priority

1984-1998, where strategic industry support was prioritised over regional support

1999-2000, when the national strategic industry approach was still to the forefront but European funding was reallocated towards the lagging BMW region.

52 One should also note, that while the nationally targeted ‘designated’ areas do not all coincide with this broader breakdown, almost of the BMW region consists of such ‘designated’ areas.


148

Table 9.8: Summary Statistics for Ireland (I)

Region Period Nat. Exp / Emp.

FDI Share (Emp.)

FDI Share (Births)

BMW 1972-1983 6.80 25.66 29.55 BMW 1984-1998 4.12 22.86 20.49 BMW 1999-2000 3.97 23.39 19.45 Rest 1972-1983 4.91 74.34 70.45 Rest 1984-1998 4.31 77.14 79.51 Rest 1999-2000 4.09 76.61 80.55 Notes: Nat. Exp / Emp is in ‘000s of 1995 IEP

Table 9.9: Summary Statistics for Ireland (II)

Region period Nat. Exp / Emp.

FDI Share (Emp.)

FDI Share (Births)

BMW/ Rest

1972-1983 1.384929 0.345171 0.419446

BMW/ Rest

1984-1998 0.955916 0.296344 0.257703

BMW/ Rest

1999-2000 0.97066 0.305313 0.241465

Clearly, support for the BMW region was relatively much higher prior to 1984 when, per manufacturing job, expenditures in the BMW region were 1.38 times those going to the rest of the country. In the 1984-98 period this ratio fell to 0.96, before rising slightly to a figure of 0.97 over the 1999-2000 period.

Examining the share of FDI employment in the two regions, prior to 1984 the BMW region attracted 35 percent of jobs. Between then and 1998 this declined to 30 percent before rising marginally to 31 percent over the 1999-2000 period. In terms of new start-ups, the earliest period saw the BMW region attract a 42 percent share, declining to 26 percent and declining further to 24 percent over the more recent periods. Our indicators are shown in Figures 9.1 and 9.2.

Figure 9.1 confirms the summary figures from Tables 9.8 and 9.9 – there is a clear difference in national grant support for firms locating in the BMW regions prior to 1984, with the picture becoming less clear-cut since then. Yearly variations in the share of FDI start-ups and total employment in the BMW region show that the share of jobs created by multinationals in the BMW regions rose until the early 1980s, but has since stabilized. The BMW region has, since the early 1970s generally received a decreasing share of multinational start-ups.

Examining the share of FDI employment in the two regions, prior to 1984 the BMW region attracted 35 percent of jobs. Between then and 1998 this declined to 30 percent before rising marginally to 31 percent over the 1999-2000 period. In terms of new start-ups, the earliest period saw the BMW region attract a 42 percent share, declining to 26 percent and declining further to 24 percent over the more recent periods. Our indicators are depicted graphically in Figures 9.1 and 9.2.

Figure 9.1 confirms the summary figures from Tables 9.8 and 9.9 – there is a clear difference in national grant support for firms locating in the BMW regions prior to 1984, with the picture becoming less clear-cut since then. Yearly variations in the share of FDI start-ups and total employment in the BMW region show that the share of jobs created by multinationals in the BMW regions rose until the early 1980s, but has since stabilized. The BMW region has, since the early 1970s generally received a decreasing share of multinational start-ups.


149

Figure 9.1: National Grant Expenditure per Employee in 1995 IEP (1000’s)

Year

rest bmw

1972 1983 2000

2.96702

9.41926

To disentangle statistically the link between these trends in the distribution of FDI in the BMW region and the national and EU policy changes over the sample period, we estimated a simple econometric model where we regressed the shares of FDI employment and start-ups, alternatively, on proxies of national and industry policy measures. As a measure of national regional industrial policy we simply utilized the grant expenditure per job. The choice of an EU policy proxy proved more difficult. At hand is the total amount of expenditure by the NDP for a full six year period from 2000 onwards, which does not allow any control for temporal variation. To deal with this, the EU expenditure figures were simply normalized by total employment in the respective regions and this measure was held constant across 1999-2000. We also, however, experimented with a simple dummy variable that takes on a value of unity since the implementation of the EU policy change in 1999. Finally, given the time and panel nature of our pooled data set, we controlled for fixed and yearly effects in the standard way.

The results of this exercise are given in Table 9.10, where the first three columns show results pertaining to the FDI employment share, while the remaining ones refer to the estimations using FDI start-up shares as the dependent variable. As shown in the first three columns, national regional grant policy, as proxied by grant expenditure per job, did not have a statistically significant impact on the distribution of jobs provided by multinationals across the two broad regions in Ireland (though earlier work by Meyler and Strobl indicates that it did have such an effect in the 1970s, when regionally-differentiated grant levels, as seen earlier, were far higher than in later years). Similarly no impact was found from the recent change in EU regional policy, as indicated either by expenditures or by the introduction of a dummy variable. We also experimented with interacting the EU dummy variable with the national policy variable in order to see whether the EU policy change had an impact on the effect of Irish national regional policy, but as can be seen in the third column, there is no statistically significant evidence of this.

In contrast, in terms of new multinational plant location we find, as shown in the last three columns, that grant provision by Irish industrial policy makers has had a significant influence on where FDI has located within Ireland since the early 1970s. In contrast, there is no apparent effect on multinational location via the change in EU policy. EU policy also appears not to have changed how national grant provision affects the distribution of new FDI projects within Ireland.


150

Figure 9.2: Share by BMW Region in total foreign employment and new foreign plants

Year

emp_share new_plants_share

1972 1983 2000

.073684

.449438

Table 9.10: Regression Results for Ireland (1) (2) (3) (4) (5) (6) DEP VAR: FDI EMP FDI EMP FDI EMP FDI Births FDI Births FDI Births Nat. Grant/EMP 0.002 0.002 0.002 0.056*** 0.055** 0.054** (0.009) (0.009) (0.009) (0.019) (0.020) (0.020) EU Exp. / Emp -2,626.423 -

10,143.826

(12,693.620)

(27,372.187)

EU Dummy -0.010 0.051 (0.088) (0.190) EUD*Nat. Grant / Emp

0.028 0.165

(0.120) (0.257) Constant 0.486*** 0.488*** 0.489*** 0.115 0.121 0.131 (0.073) (0.075) (0.076) (0.158) (0.162) (0.162) Observations 58 58 58 58 58 58 R-squared 0.00 0.00 0.00 0.24 0.24 0.25

Notes: (a) Fixed Effects regression results. (b) standard errors in parentheses. (c) **, * signify five and ten per cent significance levels. (d) time dummies included.

9.3 Commentary and key findings A tension exists between the lure of pre-existing agglomerations and the part played by regional policies in attempting to attract FDI to more peripheral regions. Regional state aids, which are often co-financed by EU structural policies and which are consistent with EU competition policy, do appear to attempt to do this, though they are rarely successful.


151

Several instances where such regional policies have had some effect were, however, identified. This arose in the Irish case up to the policy changes introduced in the early 1980s when national strategic-sector targeting came more to the fore. Irish regional policy continues to draw disproportionate numbers of foreign-firm births to peripheral regions, but this is no longer reflected in disproportionate employment levels. The latter continues to be the case in the UK however. French regional policy by contrast appears to have perverse effects, operating as it does against the attraction of FDI to peripheral regions. The FDI attracted to more peripheral regions generally tends to be in lower-tech sectors than that which is attracted to more advanced regions, and beneficial spillover effects from this type of FDI are also likely to be lower.53

While some foreign-owned high-tech sectors have been found to have deconcentrated from core to peripheral EU countries, they tend, in the latter, to locate in the most advanced regions. It might be thought that this could give rise to backward and forward linkages that would aid these economies’ more peripheral regions. These sectors are particularly intensive users of Finance, Insurance and Real Estate services, however, which is the most human-capital-intensive of the market services sectors.54 As this is the services sector which has been deconcentrating most rapidly, the appearance of the high-tech downstream sectors in less advanced economies may promote further regional divergence in these economies, e.g. through ‘brain drain’ effects.

Though regional policies in some countries have been found to have had an effect on the location of FDI, in no case were the effects strong enough (or the resources devoted to regional policy large enough) to overcome the attractiveness of pre-existing agglomerations. We must conclude from this that the major emphasis of national policy towards FDI is on attracting projects to the country rather than attempting to distribute them across the regions.

It furthermore appears that the development of transportation infrastructure alone in peripheral regions is unlikely to be sufficient to promote convergence in FDI shares. This kind of effect is well known in the economic geography literature (see e.g. Puga, 2002) and an analysis of the contrasting experiences of two Spanish regions by Rodriguez-Pose (2000) is illustrative in this regard. Thus, it is clear that only if a range of other complementary policies – from human capital development to the promotion of innovation – are in place in peripheral regions, that national FDI inflows are likely to be associated with regional convergence.

Key findings are: There are very big differences among the Member States in the contribution of FDI to total

investment, and in most Member States, the bulk of the FDI occurs in richer regions.

Regional state aids have a minor effect in steering FDI to less-favoured regions.

For most Member States, the key policy objective is to attract companies to their country rather than elsewhere, with the regional distribution of the investment seen as secondary.

Econometric work on Ireland showed that national regional grant policy did not have a statistically significant impact on the distribution of jobs provided by multinationals across the two broad regions

53 Thus for the sectors for which data is given, 20 percent of foreign employment in the lagging BMW region in Ireland is in low-tech sectors, compared to only 11 percent in the more prosperous South and East region (where the high-tech sectors are defined as Chemicals, Machinery and Equipment, Electrical and Optical and Transport). 54 See Midelfart Knarvik et al. (2000) on the financial-services intensity of these sectors, and Prais (1995) on the human capital intensity of the latter.


152

Table 9.A.1

Foreign /

natl GDP per cap Foreign /

natl GDP per cap in 2000 in 2000 Germany UK Baden-W 0.76 122.00 North-East 0.94 77.42 Bayern 0.57 123.98 North-West 0.83 87.16 Berlin 1.36 95.64 Yorks 0.71 88.15 Branden 0.15 69.44 East Mid 0.83 93.86 Bremen 1.35 142.89 West Mid 1.00 91.96 Hamburg 6.26 181.50 Eastern 0.72 103.93 Hessen 3.15 129.40 London 2.72 146.99 Mecklen. 0.16 69.37 South E 1.22 110.56 Niedersach. 0.52 95.96 South W 0.77 91.03 Nordrhein 1.48 108.64 Wales 0.79 80.58 Rheinl 0.41 96.67 Scotland 0.95 97.23 Saarland 0.45 96.84 Portugal Sachsen 0.07 70.38 Norte 0.17 56.00 Sachsen Anh 0.35 68.38 Centro 0.45 54.19 Schlesw 0.67 96.18 Lisboa 2.90 90.87 Thuringe 0.55 69.57 Alente 0.36 54.47 Spain Algarv 1.10 65.97 Galicia 0.07 64.70 Acores 0.01 51.69 Princip 0.47 70.87 Madeira 0.47 74.45 Canta 0.18 80.27 FINLAND Pais V 0.82 101.48 Ita-S 0.72 74.54 Comu 0.73 105.30 Vali-S 0.70 84.42 La Rio 0.14 91.20 Pohjoi 0.74 91.49 Aragon 0.30 87.90 Uusim 1.66 143.19 Comunic 5.33 109.96 Etela-S 0.87 95.99 Castill y Leon 0.06 75.90 Aland 0.29 139.20 Castill La Man 0.04 66.75 ITALY Extre 0.05 53.00 Piem 1.50 119.56 Catalu 0.93 99.49 Valle 2.29 123.13 Comu 0.18 79.22 Liguria 0.92 108.05 Illes Bal 0.49 98.31 Lombard 1.85 134.50 Andal 0.18 61.16 Trentin 0.88 136.29 Region 0.05 68.69 Venet 0.41 119.01 Ceuta 0.12 68.15 Friuli 1.80 114.12 Canarias 0.97 77.54 Emilia-R 0.83 129.11 FRANCE Tosc 0.60 113.53 Ile de F 1.31 158.31 Unbria 0.72 100.58 Cham 1.41 94.57 March 0.19 102.05 Picar 1.52 82.32 Lazio 1.52 112.87 Haute- 1.44 95.22 Abruz 1.46 83.69 Centre 1.28 90.78 Molise 0.07 78.78 Basse 0.81 85.32 Campan 0.20 65.30 Bourg 1.29 92.71 Puglia 0.17 67.13 Nord-P 1.05 80.64 Basilic 0.22 73.40 Lorr 1.10 83.50 Calar 0.00 62.15 Alsace 1.66 102.86 Sicilia 0.19 65.43


153

Franch 0.70 87.62 Sardegn 0.23 75.51 Pays 0.78 90.04 IRELAND Bretag 0.31 86.18 BMW 0.87 83.80 Poitou 0.58 82.92 South and East 1.05 126.41 Aquita 0.75 90.36 Midi-P 0.57 88.54 Limou 0.61 81.97 Rhone 0.98 103.13 Auverge 0.77 86.50 Langu 0.57 77.69 Proven 0.67 90.80 Corse 0.03 76.05


154

10. Conclusions The range of Member State policies that bear on different facets of cohesion is simultaneously very broad and very diverse in the different Member States. Together, these policies have, in combination with Community policies, undoubtedly reduced the divergence of living standards of the poorest regions compared with what would have happened had the policies not been in place. This is true both for divergence from the Community average and, in most Member States, from the respective national averages. But the principal manner in which the national policies have been pro-cohesive has been by boosting current consumption, whether by individuals or (often to a greater degree), by municipalities and other devolved agencies of government.

The inference to draw is that the principal impact of national policies has been on assuring social cohesion, and that they have done so by providing comparable levels of public services as well as by guaranteeing income for economically inactive individuals. In terms of the facets of cohesion and the mix of policies discussed in section 2, it is much less obvious that the overall impact of Member State policies aimed at economic development, rather than redistribution, has promoted either territorial cohesion or economic cohesion.

In all Member States, the most significant policy in terms of the scale of public expenditure continues to be transfers, largely through inter-governmental grants and the social protection system. These transfers support the current incomes of residents of net recipient regions and, because a sizeable proportion of current incomes are spent on goods and services supplied locally, these higher incomes necessarily boost current regional GDP.

10.1 Macroeconomic and budgetary policies Macroeconomic developments since the mid-1990s, when the 1st Cohesion Report was published, have mainly assisted convergence at the Member States level. GDP in most of the lower income countries has increased at a markedly higher rate than in the richer ones and the record from the start of stage 3 of EMU to the current forecast horizon shows that macroeconomic change has given a fillip to the relative performance of the cohesion countries. It can reasonably be asserted that macroeconomic policies have facilitated a reduction in disparities between the poorest regions and the EU average. By the same token, the relative stagnation of the German and Italian economies has been inauspicious for the lagging regions in those two countries.

The impact of macroeconomic policies on cohesion is, arguably, a very powerful one, albeit one that operates primarily at the Member State level. The first few years of monetary union, in particular, have seen substantial convergence in key macroeconomic indicators, notably income per head and inflation rates. There are diverse explanations for these trends: EU structural policies have borne fruit; the fall in nominal interest rates has stimulated national economies where they were previously much higher and imposed a heavy burden on public finances; macroeconomic stability has facilitated catch-up; and investment has been stimulated.

It can also be argued that the recent re-orientations of macroeconomic policy have not only benefited the cohesion countries overall, but have also enabled public finance developments that will either directly reduce disparities within those countries or give their governments the scope to implement more pro-cohesive policies. A clear example of the first mechanism is the reduced debt service burden that stems from lower nominal interest rates and falls in the debt:GDP ratio. This benefits the less wealthy regions to the extent that debt is owned disproportionately by residents of richer regions: cuts in debt service reduce the flow of public funds to richer regions at the expense of poorer ones. There is also tentative evidence that the cohesion countries have been able to maintain levels of public investment, notwithstanding significant pressures to achieve fiscal consolidation.

The longer term effects of the impact of macroeconomic policies consistent with EMU are more speculative. A more stable macroeconomic environment is generally expected to promote continued catch-up by poorer Member States and should also facilitate the development of the


155

least competitive regions. However, the impact of closer economic integration on the least competitive regions is less predictable with both pro- and anti-cohesive mechanisms at work.

Public expenditure and public transfers In all the Member States, public finance systems redistribute current income towards lower income regions. However, they do so in very diverse ways and to considerably differing degrees that reflect the character of the public finance systems and the distributions of competencies among the tiers of government. Arithmetically, these public finance flows dominate all other pro-cohesive policies, with evidence that the regions that receive the highest net fiscal transfers have their GDP increased by a quarter or more, while net contributor regions may see their GDP reduced by up to 10 percentage points. The resulting gains are, typically, channelled either through public services or through increased consumers’ expenditure. In this regard, they have an obvious impact on social cohesion, but do not necessarily affect the underlying competitiveness of the region and may, indeed, ‘crowd-out’ the ‘productive’ sector.

Four different sorts of mechanisms contribute to the public transfer effect and these, in turn, are governed by a wide range of formulae and entitlement rules:

The rate at which revenue is raised in the region (taxes, social charges and user fees);

The extent of demand for ‘automatically’ allocated public services, such as pension payments or health care;

Discretionary public service provision;

Equalisation payments that may be either horizontal (i.e. effectively transfers from richer regions to poorer regions) or vertical (payments from central to lower tier public bodies).

The overall conclusion on public finance flows is that they continue to be the Member State policies that have the biggest impact on cohesion.

Table 10.1 Summary of impacts of Member State policies on cohesion

Policy area Economic Cohesion Social Cohesion Territorial Cohesion Macroeconomic No clear effect, but

‘good’ policy helps growth

Depends on fiscal policy choices

No direct effect

Public spending Some indirect effect via higher public

consumption

Strongly positive Positive effect in sustaining public

services Public transfers Some effect via

higher private consumption

Positive in most cases

Positive by supporting incomes

State aids Not systematic, but does not directly

favour poorer regions

Has some effect, but national, rather than

regional

May help to support weaker economic

bases, not systematic Territorial economic development

Helpful to cohesion, but has multiple

targets

Depends on precise targeting

Quite pronounced in some cases

Employment Some benefits Positive, mainly for high unemployment

areas

Can be helpful. Depending on

national structures Technology Negative: favours

richer areas No direct impact Tends not to be

supportive Attraction of inward investment

Negative: favours richer areas

No direct impact Depends on Member State priorities


156

10.2 Economic development policies Many different national policies have the potential to affect the supply-side of the economy and thus to shape the competitiveness of regions. There can, however, be a tension at the heart of these policies, which is that they often have to reconcile national competitiveness and cohesion objectives. As a result, their impact is ambiguous and even their underlying aims in relation to cohesion uncertain. Although they are conceptually distinctive, the different sorts of economic policies overlap in practice and it is common to find multiple policies affecting the same targets.

State aids Despite the extensive use of state aids in many Member States, only a small proportion of the aid granted to companies is targeted at lagging regions. Instead the bulk of state aids’ expenditure goes on sectoral and horizontal policies. An econometric analysis of the impact of state aids found that overall they do not act pro-cohesively. The main findings are as follows:

• Neither overall state aid, nor regional aid appear to be systematically related to cohesion as measured by regional income disparities.

• The distribution of state aid across regions depends on the type of aid. Some aid such as R&D aid is granted mostly in richer regions.

• Regional state aid is granted mostly to poorer regions. However, within poorer regions, there are considerable variations in the amount of state aid. These amounts do not appear to correlate with the needs of regions as measured by their relative income.

Territorial policies Economic development policies that target problem (or priority) areas, as defined (whether it is urban areas, rural development zones or whatever) by the Member State, are used in all countries. Overall, their scale measured as a proportion of national budgets is very limited, although those regions which receive the highest levels can obtain support that can make a visible difference. In most Member States, the thrust of territorial policies does contribute to economic cohesion, although with the caveat that different policy initiatives can, on occasion pull in different directions. Some urban policies, for example, can steer resources to regions which are otherwise regarded as favoured.

Employment measures Despite the strong commitment by Member States to employment policies, there is little evidence that they contribute significantly to regional convergence. Finland and Denmark are the only countries where it seems that the overall effect of labour market policies favours regional cohesion. In Portugal, Spain and Italy, rules governing unemployment benefits tend to produce a situation in which the coverage of unemployment policies is higher in regions with lower unemployment. In other cases and countries the evidence is mixed. Although active labour market policies are more intense in high unemployment regions, it is very difficult to say whether they are sufficiently so to assure a reduction in regional disparities.

Technology policies The empirical evidence on R&D expenditure and on technology-intensive activities is that they are most concentrated in richer regions of Member States and that few regions stand out in this regard. The implication is that technology policies have tended to favour richer regions and thus acted to exacerbate intra-country disparities. However, it is also apparent that there is a tension between national competitiveness aims and cohesion.

Policies to attract FDI A similar conclusion emerges about FDI in that the bulk of the policy effort appears to be devoted to attracting FDI to the Member State. In addition, the evidence suggests that only


157

certain industries are plausible targets for more diversified FDI that might benefit lagging regions.

10.3 Issues for cohesion policy Cohesion is the subject of parallel action by the Member States and the Community. While what the Community does is quite closely defined and delimited, the impact of Member State policies on cohesion is more diffuse. A general conclusion to draw is that although Member State policies have a large and positive impact on aspects of cohesion, there continues to be an important role for Community policies to assist in promoting convergence, especially for the least competitive regions in poorer Member States.

Member State policies and Community policies are, on the whole, complementary in their objectives and, for the most part, in their impact. The former focus, on the one hand, on maintaining levels of public services across spatial boundaries and on dealing with specific territorial priorities – often political or social rather than economic. On the other hand, where national policies focus on competitiveness or on sectoral aims, their main impact is felt at the national level and their regional impact is not only indirect, but may also be at odds with economic cohesion aims.

Moreover, to the extent that they support richer regions, some state aids may hinder efforts to stimulate the development of lagging regions. The usual rationale for regulation of state aids is rooted in competition policy aimed at ensuring that the internal market is not at risk, but it can also be argued that a cohesion rationale should be adopted. If state aids – especially, but not only, those granted by regional or local tiers of government – make enterprises in the richer region more competitive, the effect may be to detract from economic cohesion by damaging the prospects of firms in poorer regions. Community policies, by contrast, are specifically geared towards altering the economic development trajectory of eligible regions and are, therefore, firmly focused on economic cohesion.

It is also noteworthy that in the Member States that receive substantial resources from Community policy, Community and national policy towards economic development have become quite closely inter-twined. The implication is that the approach required by the Structural Funds and the Cohesion Fund has spilled over to shape national policy on cohesion-related matters. For the future, this poses a number of challenges about what should be expected of national policies and, subsidiarity notwithstanding, the degree to which Community policies are needed to reinforce the overall commitment to cohesion. In part this is because some Member State policies – state aids being a good example – may undermine, rather than support cohesion.

In Member States where technology policy is largely set and funded at national level, it tends to favour regions with established technological capacity, whereas in countries where regional or municipal authorities have greater control, the impact of technology policies is more evenly spread. The overall trend appears to be towards greater regional balance in the spread of technology policies, but from a very uneven starting point. While technology policy based on pure excellence and efficiency criteria should remain as a policy tool for economic growth, the evidence suggests a continuing role for both Community and national policies in supporting technological development to the least developed regions to promote economic cohesion.

A further influence is institutional reform, a phenomenon that has already occurred in several Member States and is in progress elsewhere. As an illustration, a large-scale reform of Danish regional and local government is under way. The reform will cut the number of local municipalities and establish a new division of tasks between the different levels of government. The drift towards devolution of powers places a greater onus on sub-national authorities to earmark sufficient resources for economic development aimed at promoting cohesion. For richer regions (for example in Austria), this is not a problem, insofar as they can make choices about how to spend. But regions with lesser fiscal means will depend on equalisation mechanisms and, in this regard, the evidence from Italy suggests that the new mechanisms may cause problems for cohesion. The problem is that with devolution of both revenue-raising and expenditure, there is


158

no guarantee that levels of public investment associated with economic development will be maintained, either because the devolved government lacks resources overall or because it opts to prefer current expenditure.

Cohesion is a fundamental aim of the EU and Member State policies plainly do much to advance it. Nevertheless, the policies that impinge on cohesion are often rather opaque and their impact is often indirect rather than explicit. This report has sought to elucidate the policy mechanisms and to assess their impact and shows that the effects of national policies can be powerful. But it also highlights difficulties and potential conflicts in policy aims that need to be reconciled if economic, social and territorial cohesion are to be dealt with convincingly.


159

Bibliography In this bibliography the references cited in the text of the report are listed, together with additional academic and other literature referred to by the authors in the course of the study and reported in complementary reports from this project.

AFII (Agence Française pour les Investissements Internationaux), 2003, Bilan 2002 des investissements internationaux en France, www.afii.fr

Alesina A., Blanchard O., Galí, J., Giavazzi F. and Uhlig H. (2001) Defining a Macroeconomic Framework for the Euro Area, Monitoring the European Central Bank No. 3, London: CEPR.

Alberola, E. and J. M. Marqués (1999) ‘On the revelance and nature of regional inflation differentials: The case of Spain’, Studies on the Spanish Economy 35, FEDEA.

Allers M.A. & E. Gerritsen, 2003. Decentrale overheden. In: C.A. De Kam & A. P. Ros, (eds.), Jaarboek Overheidsfinanciën 2003. Den Haag, SDU

Alteneder W. et al. (2002), Engpässe und Überhänge auf Österreichs regionalen Arbeitsmärkten: prognosen für 2002 und 2003, Wien

Anci, Formez, Legautonomie, Uncem, Upi (2003), Secondo Rapporto sullo stato delle autonomie locali, Rome

Antikainen, J. (2001): “Urban regions as nodes of regional development in eastern and northern Finland”. Terra 113: 2, 89-103.

Arachi G., Zanardi A. (2000a), “Il federalismo fiscale regionale: opportunità e limiti”, in Bernardi L. (ed), La finanza pubblica italiana. Rapporto 2000, Il Mulino, Bologna;

Arachi G., Zanardi A. (2000b), “La riforma del finanziamento delle Regioni italiane: problemi e proposte”, in Economia e politica, n. 1;

Arachi G., Zanardi A. (2001), “La devoluzione nel paese del dualismo” in Bernardi L., Zanardi A. (eds), La finanza pubblica italiana. Rapporto 2001, Il Mulino, Bologna;

Ardy, B., Begg, I., Schelkle, W. and Torres, F. (2002) ‘How will EMU affect cohesion?’ Intereconomics 300-314

Armstrong D. and D. Blackaby (1998), “Regional labour markets and institutions in the United Kingdom”, ch.5 in L. van der Laan and S. Ruesga (eds), Institutions and Regional Labour Markets in Europe, Ashgate, UK.

Aronica A., Flaccadoro E. (2000), “La politica per l’industria”, in CER/IRS, Fra decentramento amministrativo e integrazione europea, il Mulino;

Aronica A., Padovani R., Servidio G. (2000), “Incentivi e politica industriale nel Mezzogiorno” in Rivista Economica del Mezzogiorno, n. 4, Il Mulino, Bologna;

Atzmüller, R., Gaubitsch, R., Sturm, R. (2000) Innovation – Qualifikation – Aspekte regionaler Entwicklung, AMS report No.17, Wien: Arbeitsmarktservice Österreich.

Audretsch, D. B., Feldman M. P. (1996), ‘Spillovers and Geography of Innovation and Production’, American Economic Review, vol. 86, pp. 630-640.

Autorità Garante per il Mercato e la Concorrenza, (2003), Relazione annuale, Rome;

Autorità per l’Energia (2003), Relazione Annuale, Rome;

Balakrishnan R.and Michelacci C. (2001) ‘Unemployment Dynamics across OECD Countries’, European-Economic-Review 45(1), 135-65


160

Balassone F., Franco D. (2001), “Fiscal Federalism and the Stability and Growth Pact: A difficult Union”, in Banca d’Italia (ed.), Fiscal Rules, Rome;

Baldini M., Bosi P. (2002), “La riforma dell’imposta sul reddito: aspetti di equità ed efficienza”, in Politica Economica n.3;

Banca d’Italia (2003), Relazione Annuale, Rome;

Barca F. (2001), “New trend in the Mezzogiorno and the change of development policies” in Padoa-Schioppa T., Graubard S.R. (eds), “Italy: Resilient and Vulnerable: The European Challenge”, Daedalus, Journal of the American Academy of Arts and Sciences, vol. 1 Spring;

Barrios, S., H. Gorg and E. Strobl (2003) “Multinationals’ Location Choice, Agglomeration Economies and Public Incentives”, CORE Discussion Paper 2003/17.

Barro, Robert and Xavier Sala-i-Martin (1992), “Convergence”. Journal of Political Economy, 100 (2): 223-51

Barry, F. (2003) “European Union Regional Aid and Irish Economic Development”, in B. Funck and L. Pizzati (eds.) European Integration, Regional Policy and Growth, Washington: World Bank.

Barry F. and Begg I. (2003) ‘EMU and Cohesion’ Journal of Common Market Studies 41.5, 781-96

Baumol, W.J., Nelson, R.R. and E.N. Wolff (1994). “Introduction: The Convergence of Productivity, Its Significance, and Its Varied Connotations”, in W.J. Baumol, Nelson Richard R. and E.N. Wolff (eds.) Convergence of Productivity, Oxford: Oxford University Press: 3-19

Begg, I. ed. (2002) Urban Competitiveness Bristol: Policy Press

Begg I. (2003) ‘Complementing EMU: rethinking cohesion policy’ Oxford Review of Economic Policy Vol. 19 No. 1, pp 161-79

Bishop S. (1997) ‘The European Commission's Policy Towards State Aid: A Role for Rigorous Competitive Analysis’ European Competition Law Review 18(2), pp. 84-86;

Blanchard, O. (1998) ‘Revisiting European Unemployment: Unemployment, Capital Accumulation and Factor Prices’ Boston: NBER Working Paper, No. 6566

Blanchard O. and Wolfers J. (2000) ‘The role of shocks and institutions in the rise of European unemployment: The aggregate evidence’ Economic Journal 110 (462), 1-33

Blomstrom, M. et al . (1995) What explains developing country growth?, NBER 4132.

Bobbio L. (2000), “Produzione di politiche a mezzo di contratti nella pubblica amministrazione italiana” in Stato e Mercato, n.58;

Boeri T. (2000), Uno stato asociale. Perché è fallito il welfare in Italia, Laterza, Rome-Bari;

E. Borensztein et al. (1995) How does FDI affect Economic Growth? NBER 5057.

Bottazzi L., Peri G. (2003). ‘Innovation and spillovers: evidence from European regions’, European Economic Review, vol. 47/4 pp. 687-710.

Boyle, G.E. and T.G. McCarthy (1997), “A Simple Measure of β-convergence”, Oxford Bulletin of Economics and Statistics: 59: 257-264.

Boyle, G.E. and T.G. McCarthy (1999), “Simple Measures of Convergence in per Capital GDP: a Note on Some Further International Evidence”, Applied Economics Letters, 6: 343-347.

Brancati R. (ed.) (2001), Le politiche industriali nelle regioni, Meridiana libri, Corigliano Calabro

Brancati R. (ed.) (2002), Le politiche per le attività produttive, Meridiana libri, Corigliano Calabro


161

Burroni L., Trigilia C., (2001), “Italy: economic development through local economies” in Crouch C., Le Galès P., Trigilia C., Voelzkow H., Local production systems in Europe: rise or demise?, Oxford U.P., Oxford;

Buti M., Pench L.R., Sestito P. (1998), European unemployment: contending theories and institutional complexities, European Investment Bank, Economic and Financial reports, n.1;

Buti, M. and Sapir A. (1998) Economic Policy in EMU. A Study by the European Commission Services Oxford: Clarendon Press

Cadiou Y. & Sigogneau A., 2002, Eléments de cadrage pour la production de tableaux de bord régionaux des activités de recherche et d’innovation, OST, juillet

Campbell M. (2001) Partnerships in Austria: Enhancing Regional Co-operation in a Decentralised Policy Framework, in: OECD (ed.), Local Partnerships for Better Governance, Territorial Economy vol.2001, no.17. p. 281-308, Paris: OECD

Carluer F. & Gaulier G., 2001, « Les productivités sectorielles des régions françaises sur moyenne période : une convergence de façade », Revue Economique, vol.52, n° 1, pp.147-166

Carruth A., Dickerson A. and Henley H. (2000) ‘What do we know about investment under uncertainty?’ Journal of Economic Surveys 14, 119-153

Casavola P., Sestito P. (2000), “Politiche di sviluppo e politiche del lavoro: che sta succedendo nel Mezzogiorno?”, in Lavoro e Relazioni Industriali, n.2;

Caswell, Dorte og Torben Pilegaard Jensen (2001): Lokale Koordinationsudvalg – på rette vej? [Local coordination committees – On the right track?], AKF

Catanzaro R., Piselli F., Ramella F., Trigilia C. (2002), Il cambiamento nei governi locali, il Mulino, Bologna;

Catin M. & Hendricks-Candéla C., 2002, Concentration et Spécialisation régionales des activités de R&D en France, communication aux journées d’étude « Connaissances technologiques : concentration, marché et valorisation », Centre Walras, 16-17 mai 2002

Cersosimo D., Wolleb G. (2001), “Politiche pubbliche e contesti istituzionali. Una ricerca sui patti territoriali” in Stato e Mercato, n. 63;

Christodoulakis, N. and Segoura (1994), ‘Inadequate infrastructure and rising cost: The decline of Greek manufacturing’, AUEB Discussion Paper 94-01

Christoffersen, Henrik (2003): Det danske bymønster og landdistrikterne [The Danish urban pattern and the rural districts], AKF, København

Chiri S., Pellegrini G., Sappino C. (1998), “L’attuazione dell’intervento per lo sviluppo degli investimenti nelle aree depresse”, in Rassegna Economica, n. 1;

Clarysse, Bart and Ugur Muldur (1999), `Regional Cohesion in Europe? An Analysis of How EU Public RTD Support Influences the Techno-Economic Regional Landscape’, Brussels: European Commission.

Collie D. (2000) ‘State aid in the European Union: The Prohibition of Subsidies in an Integrated Market’ International Journal of Industrial Organization 18(6), pp. 867-884;

Combes P.P, Duranton G. & Gobillon L., 2003, « Origine et ampleur des inégalités spatiales de salaire en France », in Compétitivité, rapport du Conseil d’Analyse Economique, n°40, la Documentation Française

Commissariat Général du Plan, 2002, « La Recherche Publique et les Politiques de la Science et de la Technologie », in La France dans l’Economie du Savoir, chap.4, La Documentation Française, Paris


162

Commissariat Général du Plan, 2002, La France dans l’économie du savoir. Pour une dynamique collective, La Documentation Française, Chapitre 6, pp.165-199

Conseil Economique et Social, 2003, “Les Fonds Structurels, un outil d’aménagement du territoire”, coll. Avis et Rapports, juillet

Correa María D. and J. Manzaneda, (2002): Política Regional Española Y Europea: Período 1983 – 1999, Secretaria de Estado. Ministerio de Hacienda. SGFCC-2002-05.

Cortelius, M. (2002): Europæisering af danske amter [The europeanization of Danish counties] , Aalborg Universitetsforlag. Aalborg

Coughlin, C., and E. Segev (2000) “Location Determinants of New Foreign-Owned Manufacturing Plants”, Journal of Regional Science, 40, 2, 323-351.

Crozet M., Mayer T. & Mucchielli J.L., 2003, How Do Firms Agglomerate, A Study of FDI in France, CEPR Discussion Paper n°3973

DARES, 2003, « Politique de l’emploi en 2001, Repli et Recentrage », Premières Informations et Premières Synthèses, n°8.1, février

DATAR/SEGESA, 2002, "Les objectifs comparés des DOCUP et des contrats de plan Etat/région 2000/2003", november

Davezies L., 2001, « Revenu et Territoires », in Aménagement du Territoire, rapport du Conseil d’Analyse Economique, n°31, La Documentation Française, Paris

Dell’Aringa C. (2001), “Federalismo e politiche del lavoro”, in Economia Italiana, n.3;

European Commission (1996), Economic and social cohesion in the European Union: the impact of member states own policies, Brussels;

DfES (2003), 21st Century Skills: Realising Our Potential, Stationary Office http://www.dfes.gov.uk/skillsstrategy/_pdfs/whitePaper_PDFID4.pdf

DGCL, 2003, Les Finances des Régions 2001, ministry of interior

Dijk, J. van & J. Oosterhaven, 1994. Past, present and future of Dutch regional policy. In: C.H.A. Verhaar & P.M. De Klaver (eds.), The functioning of economy and labour market in a peripheral region – the case of Friesland. Leeuwarden, Fryske Akademy, pp. 111-135

Dutt A. (1997) Direct Foreign Investment, TNCs and Growth, in R. Kozul-Wright and R. Rowthorn, TNCs and the Global Economy, London: Macmillan. Ecorys, 2003a. Mid term review Kompas voor het Noorden. Ecorys, Rotterdam

Ecorys, 2003b. Jaarboek Grotestedenbeleid. Ecorys, Rotterdam

Ecotec (2002), Thematic Evaluation of the Territorial Employment Pacts, Brussels;

Ederveen, S., H. de Groot and R. Nahuis (2002) “Fertile soil for Structural Funds? A Panel Data Analysis of the Conditional Effectiveness of European Cohesion Policy”, Netherlands Bureau for Economic Policy Analysis

Ehlermann C.-D. (1995) ‘State Aid Control in the European Union: Success or Failure?’ Fordham International Law Journal 18(4), pp. 1212-1229

Elhorst, J.P., J. Oosterhaven, F. Sijtsma & T.M. Stelder, (1999). Welfare effects of spatial deconcentration: A Scenario for the Netherlands. Tijdschrift voor Economische en Sociale

EPRC and L’Oeil (1996), Economic and Social Cohesion in the European Union: the Impact of Member States’ own policies, report to the EC-DGXVI, Glasgow;

Erhel C., 2003, « L’emploi et le chômage en France, au début des années 2000 », Regards sur l’actualité n°290, La Documentation Française, avril


163

Escorsa, P. (2003), ‘La región y el fomento a la innovación y la competitividad: experiencias en España y Europa en la construcción de Sistemas regionales de Innovación’, Seminario sobre Ciencia, Tecnología y Globalización, OEI (Organización de Estados Iberoamericanos).

Esping Andersen G. (1990), The three worlds of welfare capitalism, Princeton University Press;

Esterle L. & Laville F., 2001, Cartographie des compétences scientifiques et techniques régionales, OST, mars

European Commission (1995), Social protection in Europe, Brussels;

European Commission (1997), 2nd European Report on Science and Technology Indicators (ERSTI). Brussels European Commission.

European Commission (2000): The EU compendium of spatial planning systems and policies: Austria, in: European Union (2000): The EU compendium of spatial planning systems and policies, Regional development studies No. 28.

European Commission, 2001, Employment policies in the EU, Joint Report 2001, Directorate-General for Employment and Social Affairs

European Commission (2001), Second Report on Economic and Social Cohesion, Brussels;

European Commission (2002), Impact evaluation of the European Employment Strategy. Technical analysis, Brussels;

European Commission, 2002, European Innovation Scoreboard : Technical Paper n°3 EU Regions, http://trendchart.cordis.lu/Scoreboard2002/index.html

European Commission (2002), 2nd European Scoreboard on Innovation (ESI). Brussels: European Commission.

European Commission (2003), 3rd European Report on Science and Technology Indicators (ERSTI). Brussels: European Commission.

European Commission (2003), Joint Employment Report 2002, Brussels;

European Commission (2003a), The future of the European Employment Strategy (EES). “A strategy for full employment and better jobs for all”, Brussels;

EZ, (ministerie Economische Zaken), 1999. Nota Ruimtelijk Economisch Beleid. Den Haag

Fabiani S., Pellegrini G. (1997), Education, infrastructure, geography and growth: an empirical analysis of the development of Italian provinces, Banca D'Italia-Temi di discussione, n. 323;

Fagerberg, J. and B. Verspagen (1996), “Heading for Divergence? Regional Growth in Europe Reconsidered”, Journal of Common Market Studies, 34 (3): 431-448.

Fatás A. and Mihov I. (2001) ‘Government size and automatic stabilizers: international and intranational evidence’ Journal of International Economics 55.1, 3-28.

Fayolle J. & Lécuyer A., 2002, Regional growth, national context and the European structural funds: an empirical reappraisal, conference « Innovation and Growth: New challenges for the Regions », Nice-Sophia-Antipolis, January

Feldman, M. (1999), ‘The New Economics of Innovation, Spillovers and Agglomeration: a Review of Empirical Studies’, Economics of Innovation and New Technology, vol. 8, pp. 5-25.

Ferrer, C. (1998) “Patterns and Determinants of Location Decisions by French Multinationals in European Regions”, in A. Rugman and J-L. Mucchielli (eds.) Multinational Location Strategy, Volume 6 in Research in Global Strategic Management, London: JAI Press.

Ferrera M., Gualmini E. (1999), Salvati dall’Europa? Welfare e lavoro fra gli anni ’70 e gli anni ’90: le riforme già fatte e quelle che restano da fare, Il Mulino, Bologna;


164

Ferrera M., Gualmini E., Graziano P., Alti T. (2000), Labour market governance in the Italian regions: the experience of the 1990’s, Poleis-Università Bocconi, Quaderno di ricerca n. 27, Milan;

Filippaios F. and K. Kottaridi (2003), ‘Investment patterns and the competitiveness of greek regions’, Discussion Paper Series 03-01, Department of European and International Economic Studies, AUEB

Fotopoulos, G. and N. Spence (1999), ‘Spatial variations in new manufacturing plant openings: Some Empirical Evidence from Greece’, Regional Studies, 33, 1, 219-229.

Fotopoulos, G., D. Giannias and P. Liargovas (2002), ‘Economic development and convergence in Greek regions 1970-1994: Alternative methodological approaches’, Aeichoros, 1, 60-91 (in greek).

Frattini T., Zanardi A. (2003), “Federalismo regionale: un anno al palo”, in Guerra M.C., Zanardi A. (eds), La finanza pubblica italiana. Rapporto 2003, Il Mulino, Bologna;

Friedman, J., D. Gerlowski and J. Silberman (1992) “What Attracts Foreign Multinational Corporations? Evidence from Branch Plant Location in the US”, Journal of Regional Science, 32, 4, 403-418.

de la Fuente, Ángel. (2000) 'Convergence across countries and regions: theory and empirics', CEPR Discussion Paper Series, 2465.

Gardiner, B. (2003): ‘Regional Competitiveness Indicators for Europe - Audit, Database Construction and Analysis’, Regional Studies Association International Conference Pisa, 12-15 April, 2003

Gerhardter, G., Gruber M. (2001) Regionalförderung als Lernprozess, Evaluierung der Förderungen des Bundeskanzleramtes für eigenständige Regionalentwicklung, Schriften zur Regionalpolitik und Raumordnung Nr. 32, Wien: Bundeskanzleramt Abt. IV.

Giannias, D., P. Liargovas and G. Manolas (1998), ‘Development characteristics and regional inequalities in Greece’, Topos, 15, 67-82 (in greek).

Giannias, D., P. Liargovas and G. Manolas (1997), ‘Regional disparities in Greece, 1961-1991’, Topos, 13, 47-61 (in greek).

Giannini S., Guerra M.C. (2003), “Aspettando la riforma: il sistema fiscale fra promesse e vincoli”, in Guerra M.C., Zanardi A.(2003);

Giarda P. (2002), Le regole del federalismo fiscale nell'art. 119: un economista di fronte alla nuova costituzione, Siep Working paper;

Giunta A., Florio M. (2002), “L’esperienza dei Contratti di programma: una valutazione a metà percorso” in L’Industria, n. 2;

Granaglia E. (2001) “Federalismo fiscale e redistribuzione. Alcuni nodi irrisolti del Dlgs 56/2000 in materia di sanità pubblica” in Anania G. (ed.), Scelte pubbliche, strategie private e sviluppo economico in Calabria, Rubbettino;

Griliches Z. (1990), ‘Patent statistics as Economic Indicators: a survey’, Journal of Economic Literature 28, 1661-1707.

Gual J. (1998) ‘State Aid and Convergence in the European Union’

Guimaraes, P., R. J. Rolfe and D. P. Woodward (1998) “Regional Incentives and Industrial Location in Puerto Rico”, International Regional Science Review, 21, 2, 119-138.


165

Halkier, Henrik & John Evan Flockhart: Partnership Project. The Case of Denmark, European Research Unit, Aalborg University, 2002

Halkier, Henrik (2001): “Regional policy in transition : A multi-level governance perspective on the case of Denmark”, European Planning Studies Vol. 9 (3), pp. 323-338

Head, K., J. Ries and K. Ruckman (1998) “Industry Agglomeration and the Location of Foreign Affiliates”, in A. Rugman and J-L. Mucchielli (eds.) Multinational Location Strategy, Volume 6 in Research in Global Strategic Management, London: JAI Press.

Heikkilä, M., T. Rintala, I. Airio & S. Kainulainen (2002): “The welfare and future of rural and urban areas”, STAKES, Research Report 126.

Helg R., Peri G., Viesti G., (2000), Abruzzo and Sicily: catching up and lagging behind, European Investment Bank Papers, n. 5/1;

Hill, S., and M. Munday (1992) “The UK Regional Distribution of Foreign Direct Investment: Analysis and Determinants”, Regional Studies, 26, 6, 535-544.

Hilpert, U. (1992) Archipiélago Europe: Islands of Innovation. Síntesis Report

HM Treasury (2002), Employment Action Plan 2002, HM Treasury, London http://www.hm-treasury.gov.uk/documents/taxation_work_and_welfare/unemployment/tax_unem_plan02.cfm

HM Treasury (2002), Opportunity and Security for All: Investing in an enterprising, fairer Britain (2002 Spending Review), HM Treasury, London <http://www.hm-treasury.gov.uk/spending_review/spend_sr02/report/spend_sr02_repindex.cfm> HM Treasury (2002), Investing in Innovation: A Strategy for Science, Engineering and Technology, HM Treasury, London <http://www.hm-treasury.gov.uk/Spending_Review/spend_sr02/spend_sr02_science.cfm> HM Treasury (2002), Cross-Cutting Review of Science and Research: Final Report, HM Treasury, London <http://www.hm-treasury.gov.uk/media/3A7B0/science_crosscutter.pdf>

Honohan, P. and P.R. Lane, 2003, Divergent inflation rates in EMU, Economic Policy, 18, Issue 37

Houdebine M., 1999, “Concentration Géographique des Activités et Spécialisation des Départements Français”, Economie et Statistique, n°326-327, pp.189-204

Huber, P. (2001) Territorial Employment Pacts, Discussion paper, peer review programme of the European Employment Strategy, Wien, ÖSB-Unternehmensberatungs GmbH

Hüfner, F.P. and M. Schröder, 2002, Exchange rate pass-through to consumer prices: A European perspective, ZEW Discussion Paper 02-20

Hutschenreiter, G., Sieber, S. (2002) Benchmarking der F&E-Aktivitäten der Bundesländer, WIFO Presentation for: Vereinigung der österreichischen Industrie, Wien.

INSEE, La France et ses Régions, annual publication, various issues

Ioannis G. and Reiner M. (2001) ‘State Aid Control in the European Union - Rationale, Stylised Facts and Determining Factors’ Intereconomics 36(6), pp. 289-297

Isae (2003), Rapporto annuale sull’attuazione del federalismo, Rome;

Isfol (2002), Impact Evaluation of the European Employment Strategy, Rome;

Istat (2002), Conti economici territoriali, Rome;

Jimeno Juan F. (2001) Employment Policies in Spain (1975-2000): Lessons for Central Eastern European Countries Acceding to the EU Paper prepared for the conference ‘Labor, Employment and Social policies in the EU Enlargement Process: Changing Perspectives and Policy Options’, Baden/Vienna, June 28-30, 2001.

Karanassou M., Sala H. and Snower D. (2002) ‘Unemployment in the European Union: A Dynamic Reappraisal’, Bonn: IZA Discussion Paper No. 531, www.iza.org.


166

Krugman, Paul (1991) Geography and Trade, Cambridge, Mass.: MIT Press.

Lallement R., Mouhoud E.M. & Paillard S., 2002, « Polarisation et internationalisation des activités d'innovation : incidences sur la spécialisation technologique des nations », Région et Développement

Lambrinidis, M., I. Psicharis and E. Kazazis (2001), ‘Public infrastructure and regional development: The case of Greek manufacturing’, Topos, 16, 45-63 (in greek).

Lehner S., Meiklejohn R. and Louw A. (1991) Fair Competition in the Internal Market: Community State Aid Policy Luxembourg: OOPEC

Lolos, S. and A. Theodoulides (2001), ‘The role of EU Structural Funds in the development of lagging behind regions’, Economic and Financial Moelling, Spring, 29-46.

Lucas, Robert E. (1988) 'On the Mechanics of Economic Development', Journal of Monetary Economics, 22(1), 3-42.

McLean I. and A. McMillan (2002), “The Fiscal Crisis of the United Kingdom”, Nuffield College Working Papers in Politics No10

McLean I. and A. McMillan (2003), ‘The distribution of public expenditure across the UK regions’, Fiscal Studies 24 (1), pp. 45-71)

Madsen, Bjarne; Chris Jensen-Butler og Poul Uffe Dam (2001): The Line-Model, AKF, [Institute of Local Government Studies]

Mailand, Mikkel (2001): Den danske model lokalt og regionalt. Konsensus og samarbejde i arbejdsmarkedspolitiske netværk [The Danish Model at the regional and local level. Consensus and cooperation in networks of labour market policy formation], Jurist- og Økonomforbundets Forlag, København

Mäkinen, P. (2002): “The limitations and opportunities of industrial and regional policy in EU-Finland”, Pellervo Economic Research Institute, PTT-survey 2/2002.

Mankiw, Gregory N., Romer, David and David N. Weil (1992) 'A Contribution to the Empirics of Economic Growth', Quarterly Journal of Economics, 107 (2), 407-437.

Mariotti S., Mutinelli M. (1999), Gli investimenti diretti esteri nel Mezzogiorno: il passato e le tendenze attuali, in L’industria, n.2;

Markusen, J. (1988) “Production, Trade and Migration with Differentiated, Skilled Workers”, Canadian Journal of Economics, 21, 3, 492-506.

Martin, Carmela and Ismael Sanz (2003), “Real Convergence and European Integration: The Experience of the Less Developed EU Members”, Empirica, Vol. 30 (forthcoming).

Martin P. (1999). ‘Public Policies, regional inequalities and growth’, Journal of Public Economics, vol. 73 (1), pp. 85-105.

Martin R. (1993), “Remapping British regional policy: the end of the North-South divide?”, Regional Studies 27 (8), pp.797-805.

Matzinger, A. (2002) Finanzausgleich, in: Steger, G. (ed.), Öffentliche Haushalte in Österreich, Wien: Verlag Österreich. p. 51-94

Maurel F. & Sédillot B., 1999, “A measure of the Geographic Concentration in French Manufacturing Industries”, Regional Science and Urban Economics, n°29, pp.575-604

Mayerhofer, P., Palme, G. (coord.) (2001) Aspekte der regionalen Wettbewerbsfähigkeit, part of: Preparity - Strukturpolitik und Raumplanung in den Regionen an der mitteleuropäischen EU-Außengrenze zur Vorbereitung auf die EU-Osterweiterung, Teilprojekt 8, Wien: WIFO study for INTEREG IIc.


167

Meyler, A., and E. Strobl (2000) “Job Generation and Regional Industrial Policy in Ireland”, Economic and Social Review, 31, 2, 111-128. Midelfart-Knarvik, K. H., H. Overman, S. Redding and A. J. Venables (2000) The Location of European Industry, Economic Papers No. 142, European Commission: Directorate-General for Economic and Financial Affairs.

Ministerie van Binnenlandse Zaken en Koninkrijkrelaties, 2000. Overzicht specifieke uitkeringen. Den Haag, SDU

Ministry for Productive Activities (2001), Relazione sugli interventi di sostegno alle attività economiche e produttive, Rome;

Ministry of Economics (2003), Relazione generale sulla situazione economica del paese – 2002, Rome

Ministry of Economics - RGS (2002), La spesa statale regionalizzata. Anno 2000, Temi di Finanza Pubblica n. 23, Rome;

Ministry of Economics - DPS (2001), Terzo rapporto sullo sviluppo territoriale 1999-2000, Rome;

Ministry of Economics - DPS (2003), Quinto rapporto sullo sviluppo territoriale 2000-2001, Rome;

Ministry of Labour (2001, 2002, 2003), Monitoraggio delle politiche occupazionali e del lavoro, Rome;

Ministry of Labour (2001a, 2002a), National Action Plan for Employment, Rome;

Moisio, A. (2002): “Essays on Finnish municipal finance and intergovernmental grants”, Government Institute for Economic Research, VATT-Research Reports 93.

Montanino A., Sestito P. (2003), Le politiche del lavoro in Italia e in Europa in una logica di benchmarking, Rome;

Muñoz, E. (2001) “The Spanish system of research: Research and innovation in Spain” in Laredo Ph. and Mustar, P. Research and innovation policies in the new global economy: an international comparative analysis, Cheltenham: Edward Elgar, 359-397.

Nenci S. (2000), “La politica di attrazione degli investimenti diretti esteri nel Mezzogiorno ed il sostegno alo sviluppo delle aree arretrate” in Rassegna economica n. 3;

Nenonen, T. (2002) “Migration and regional policy - Special supports for individuals”. University of Jyväskylä, Faculty of Business Administration. Part of the writings: “Hannu Tervo 50 years anniversary: Economics, Regional development and Labour market”.

Neven D. (1994) ‘The Political Economy of State Aids in the European Community: Some Econometric Evidence’ Cahiers de Recherches Économique no. 9402 Lausanne: University of Lausanne

Neven, D.J., and C. Gouyette (1995), “Regional convergence in the European Community”, Journal of Common Market Studies, 33 (1): 47-65.

Nicolaides P. and Bilal S. (1999) ‘An Appraisal of the State Aid Rules of the European Community: Do they Promote Efficiency?’ Journal of World Trade 33(2), pp. 97-124

Ninni A, Silva F. (eds) (1997), La politica industriale, Laterza, Rome-Bari;

van den Noord P. (2000) ‘The size and role of automatic stabilizers in the 1990s and beyond’, Economics Department Working Papers No.230, Paris: OECD

Norstrand, Rolf og Anne Kaag Andersen (2002): Indkomster, flytninger og uddannelse [Income, mobility and education], AKF [Institute of Local Government Studies]


168

Norstrand, Rolf; Anne Kaag Andersen og Karsten Bo Larsen (2001): Regional udvikling af indkomst og beskæftigelse [Regional trends in incomes and employment], AKF, [Institute of Local Government Studies]

Obstfeld M. and Peri G. (1998) ‘Asymmetric shocks: Regional non-adjustment and fiscal policy’ Economic Policy 28, 206-259

OECD-TDPC (2001), Territorial Review: Italy, Paris;

OECD (2002), Employment Outlook, Paris;

OECD, 2003, Trends and Recent Developments in Foreign Direct Investment, Directorate for Financial, Fiscal and Enterprises Affairs, june

Onofri P. (2001), Un’economia sbloccata. La svolta degli anni ’90 e le politiche per il futuro, Il Mulino, Bologna;

Onofri P. (2003), “Un anno di governo della finanza pubblica”, in Guerra M.C., Zanardi A. (2003);

ONS (2002), Research and Development in UK Businesses: Business Monitor MA14, Stationery Office <http://www.statistics.gov.uk/downloads/theme_commerce/MA14_2001.pdf>

Oosterhaven, J., 1996. Dutch regional policy gets spatial. Regional Studies, vol. 30, nr. 5, pp. 527-532

Oosterhaven J., Elhorst J.P. (2003). Indirect economic benefits of transport infrastructure investments. In: W. Dullaert, B.A.M. Jourquin, J.B. Polak (eds.) Across the border; Building upon a quarter century of transport research in the Benelux, pp. 143-162. De Boeck, Antwerpen

Oosterhaven, J.& W.E. Romp, 2003. Indirect Economic Effects of New Infrastructure: A Comparison of Dutch High Speed Rail Variants. Tijdschrift voor Economische en Sociale Geografie, vol. 94, issue 4, pp. 439-452

Oosterhaven, J. & R.J. Stoffelsma, 1990 (a), The net effect of social security on the Dutch interregional income distribution. In: L. Anselin & M. Madden (eds.), New directions in regional science: integrated and multiregional approaches. London, Belhaven Press, pp. 236-252

Oosterhaven, J. & R.J. Stoffelsma, 1990 (b), Dutch social security benefits and interprovincial inequalities: a decomposition and a factor analysis. Tijdschrift voor Economische en Sociale Geografie, Vol. 81, issue 1, pp. 30-46

ÖROK (2002) 10. Raumordnungsbericht, Wien: Österreichische Raumordnungskonferenz.

OST, Science et technologie, Indicateurs 2002, Economica, Paris, 2002

Ottaviano, Gianmarco and Diego Puga (1998) 'Agglomeration in the Global Economy: A Survey of the New Economic Geography', World Economy, 21(6), 707-731.

Pain N. (2002) ‘EMU, Investment and Growth: Some Unresolved Issues’, National Institute Economic Review 180, 96-108.

Pavitt, K.(1998) “The inevitable limits of EU R&D funding”, Research Policy 27: 559-568.

Pelegrín, A. (2003) Regional distribution of foreign manufacturing investment in Spain. Do agglomeration economies matter? Institut d’ Economia de Barcelona 6, University of Barcelona.

Peterson, J, and M.Sharp (1998) Technology Policy in the European Union, London: Macmillan Press.

Petrakos, G. and Y. Saratsis (2000), ‘Regional inequalities in Greece’, Papers in Regional Science, 79, 57-74.

Petrakos, G. and Y. Ioannides (2000), ‘Regional disparities in Greece: The performance of Crete, Peloponese and Thessaly’, EIB Papers, 5, 31-58.


169

Pica F., Stammati S. (2002), “La finanza delle regioni deboli nel Titolo V della Costituzione Italiana” in Rivista economica del Mezzogiorno, n. 1-2;

Psicharis, I. and E. Kazazis (2002), ‘The analysis of changes in employment in the EU regions’, Topos, 18-19, 17-43 (in greek).

Prais, S. J. (1995) Productivity, Education and Training, Cambridge University Press.

Puga, D. (2002) European Regional Policies in light of recent location theories, Journal of Economic Geography, 2, 373-406.

Quah, Danny T. (1996) 'Convergence Empirics Across Economies with Some Capital Mobility', Journal of Economic Growth, 1(1), 95-124

Ramella F., Magnatti P., Trigilia C., Viesti G., Burroni L. (2003), La lezione dei patti territoriali per la progettazione integrata territoriale nel Mezzogiorno, report for MEF-DPS, Rome;

Regeringen (2003): Den regionale Vækststrategi [Regional Growth Strategy], København

Regeringen (2004a): Regionalpolitisk redegørelse 2004 [Regional Report 2004], København

Regeringen (2004b): Regionalpolitisk redegørelse 2004 – Analyser og baggrund [Regional Report 2004 – Analysis and Background], København

RFT (2002) Nationaler Forschungs- und Innovationsplan 2002, Wien: Rat für Forschungs- und Technologieentwicklung.

Rodriguez-Pose A. (2000) ‘Economic Convergence and Regional Development Strategies in Spain: The case of Galicia and Navarre’, European Investment Bank Papers, 5, 1, 2000, 89-116

Romer, Paul M. (1986) 'Increasing returns and long run growth', Journal of Political Economy, 94, 1002-1037.

Romer, Paul M. (1990) 'Endogenous Technological Change', Journal of Political Economy, 98, 5-21.

Rosenqvist, O. (2002): “Regional centre development policy and the other obstacles to an integrated rural policy in Finland”, Terra 114:2, p.59-67.

Rossi S. (2001), La politica economica italiana 1968-2000, Laterza, Rome-Bari;

Rossi S. (ed.) (2003), La nuova economia, Il Mulino, Bologna;

Rovolis, A. and N. Spence (2002), ‘Duality theory and cost function analysis in a regional context: The impact of public infrastructure capital in the Greek regions’, Annals of Regional Science, 36, 55-78.

Royer J.F., 2001, « Quelques tendances affectant la répartition spatiale de la population, de l’emploi et de la valeur ajoutée en France métropolitaine, entre 1982 et 1999 », in Aménagement du Territoire, rapport du Conseil d’Analyse Economique, n°31, La Documentation Française, Paris

Saarinen, J. & P. Kauppila (2002): “Evaluation of the regional economics of tourism: the economic impacts of tourism in Pelkosenniemi, eastern Lapland”. Terra 114:1, p. 25-36

Sala-i-Martin, Xavier (1995) 'Regional Cohesion: Evidence and Theories of Regional Growth and Convergence', CEPR Discussion Paper, 1075.

Saxenian, A. (1994), Regional Advantage, Harvard University Press, Cambridge.

Sérandon D., 2003, « Aménagement des 35 heures, harmonisation des SMIC et allégements de charges », Regards sur l’actualité n°290, La Documentation Française, avril


170

Schneider, M. (2002) Raumordnung, Betriebsansiedlung und Finanzausgleich, in: Theurl, E., Winner, H., Sausgruber, R. (ed.), Kompendium der österreichischen Finanzpolitik, Wien: Springer

Sestito P. (2002), Il mercato del lavoro in Italia, Laterza, Rome-Bari;

Simmie, J., Sennett, J., Wood, P. and Hart, D. (2002), ‘Innovation in Europe: A Tale of Networks, Knowledge and Trade in Five Cities’, Regional Studies, Vol. 36, pp 47-64.

Simopoulos, C. and D. Asteriou (1998), ‘Testing convergence across the Greek regions’, Regional Studies, 32, 6, 537-546.

Smith, V., A.L. Broberg and J. Overgaard (2000): ” Regional influence on R&D behaviour. Evidence from Danish firms”, Analyseinstitut for Forskning, Working Paper 2000/3 (http://www.afsk.au.dk/ftp/workingpapers/WP00_3.pdf)

Soete, L. and A. Arundel (eds.) (1993), An Integrated Approach to European Innovation and Technology Diffusion Policy: A Maastricht Memorandum, Brussels: European Commission.

Solow, Robert M. (1956) 'A Contribution to the Theory of Economic Growth', Quarterly Journal of Economics, 70(1), 65-94

Steiner, M. (2002) Die Rolle der Finanzpolitik in der regionalpolitischen Entwicklung: die Perspektive der Industrieökonomik, in: Theurl, E., Winner, H., Sausgruber, R. (ed.), Kompendium der österreichischen Finanzpolitik, Wien: Springer

Stoffelsma, R.J., 1992, Centrale overheid en regionale welvaart. Groningen, Stichting Ruimtelijke Economie Groningen

Stoffelsma, R.J. & J. Oosterhaven, 1989, Social security benefits and interregional income inequalities: the case of the Netherlands. The Annals of Reguional Science, 23, pp. 2223-240

Stoffelsma, R.J. & J. Oosterhaven, 1991, Social security and the income distribution: a spatio-temporal analysis for the Netherlands. Public Finance, Vol. 46, pp. 268-288

Strijker, D., J.P. Elhorst, J. Oosterhaven & A.S. Zeilstra, 2001. Kompas voor Kwaliteit. REG-publicatie 24. Groningen, Stichting REG

Strukturkommissionen (2004): Strukturkommissionens betænkning (nr. 1434) [Report from the commission on the future structure of the public sector], København

Svimez (2001), Rapporto 2001 sull’economia del Mezzogiorno, Il Mulino, Bologna;

Svimez (2002), Rapporto 2002 sull’economia del Mezzogiorno, Il Mulino, Bologna;

Svimez (2002a), Sul finanziamento delle funzioni pubbliche di regioni ed enti locali secondo le nuove norme costituzionali, Quaderno 15 di Informazioni Svimez, Rome;

Syriopoulos, K., L. Lambrianidis and D. Asteriou (1997), ‘Convergence of per capital GDP in the prefectures of Greece’, Topos, 13, 67-82 (in greek).

Tödtling-Schönhofer, H. et al. (1999) Regionale Innovationspolitik 2000, Wien: Österreichisches Institut für Raumplanung.

Tohmo, T., J. Ritsilä, T. Nenonen and M. Haapanen (2001) ‘Slowing down migration’ Municipal Development Institution Research 31/2001.)

Tondl, G. (2001) Regional Policy, in The Economics of the European Union: Policy and Analysis, 3rd edition. Oxford University Press.

Travers, T. (2002), “Decentralization London-style: The GLA and London Governance”, Regional Studies 36 (7), pp.779-788. Trigilia C. (1992), Sviluppo senza autonomia. Effetti perversi delle politiche nel Mezzogiorno, Il Mulino, Bologna;


171

Trigilia C., Viesti G. (2002), “Politiche per il Mezzogiorno cercasi” in il Mulino, n. 6;

Tsionas, E. (2002), ‘Another look at regional convergence in Greece’, Regional Studies, 36, 6, 603-609.

Vandelli L. (2002), Devolution e altre storie, Il Mulino, Bologna;

Verhagen, A.J.W.M., 2001. Criteria voor de verdeling van specifieke uitkeringen. Alphen a.d. Rijn, Samson.

Verspagen B. (1997), European Regional Clubs: do they exist and where are they heading? On Economic and Technological Differences between European Regions, in Adams J. and Pigliaru F. (Eds) Economic Growth and Change: National and Regional Patterns of Convergence and Divergence. Edward Elgar, Cheltenham

Viesti G. (2000), Come nascono i distretti industriali, Laterza, Rome-Bari;

Viesti G. (2001), “Decentramento dei poteri e questione territoriale in Italia” in Economia Italiana, n.3;

Viesti G. (2002) “Lavoro e politiche del lavoro nel Mezzogiorno” in Rivista Economica del Mezzogiorno, n.3;

Viesti G. (2003), Abolire il Mezzogiorno, Laterza, Rome-Bari;

VROM-raad (1999). Sterk en mooi platteland. Advies 015. Den Haag

Wishlade F. (2003) Regional State Aid and Competition Policy in the European Union European Monographs 43 London: Kluwer Law International

Wolsink, M., 2003, Reshaping the Dutch planning system: a learning process?. Environment and Planning A, vol. 35, pp. 705-723

WRR, 2003, Nederland Handelsland. Rapporten aan de regering 66. Den Haag, Staatsutigeverij.

Zanardi A.(2002), “Riforma fiscale, devoluzione e federalismo”, in Politica Economica n.3, il Mulino, Bologna;

Zhang, J. (2001): The Impact of Tourism on the Economies of Danish Regions, AKF [Institute of Local Government Studies]


172

Annexes

1 Questionnaire/data template

2. EPRC report on regional incentives

3. Regional profiles


173

Annex 1: DATA COLLECTION TEMPLATE AND QUESTIONNAIRE ALL QUESTIONS TO BE ANSWERED IN RELATION TO THE REGIONS IN YOUR COUNTRY. FOR REGIONAL DATA, THE DISAGGREGATION SHOULD BE TO NUTS LEVEL 2 UNLESS THERE ARE GOOD REASONS WHY THAT IT IS NOT POSSIBLE, IN WHICH CASE NUTS 1 FIGURES SHOULD BE PROVIDED.

A General information on cohesion 1 Literature review In all countries there is a wealth of academic and policy evaluation literature that conveys information on the impact of policies that affect cohesion. Please provide a concise review of recent studies that document the effects of national policies on reducing regional disparities or in promoting cohesion in other tangible ways.

For key studies, as a minimum, provide a short description that indicates the purpose of the study, the methodology employed and the principal findings.

If studies contain useful empirical data, please supply details and, if relevant, copies of the most important tables.

2 Regional profiles Building on the statistical data in the second update of the 2nd cohesion report, additional statistical material that will be supplied to you by DG Regio and national sources, provide a short profile (around one page) of each region that combines statistical material and qualitative assessments of the region, following the profile template in appendix 1.

B Specific policies and their effects on cohesion Many different policy areas have an influence on regional disparities and on the promotion of cohesion. This data collection exercise has two aims: first, to document the different national policy areas that impinge on cohesion and how they do so; and, second, to gather primary data that can be used to assess the extent to which the policy area contributes to cohesion.

1 Macroeconomic policy How has the approach to macroeconomic policy changed in recent years, and especially since the advent of EMU? Show the impact of these changes on macroeconomic variables that influence competitiveness (real and nominal interest rates; nominal labour costs; relative unit labour costs and other indicators of competitiveness) for the last 20 quarters.

How well does the macroeconomic policy of the Member State reflect the economic circumstances of the region? Answer in terms of judgements on the appropriateness of the principal instruments of policy, i.e. the interest rate, the fiscal policy stance and the development of wages and prices.

2 Public expenditure For those central government policies which have an identifiable regional incidence, supply data on the amount spent in each region. Please try to construct a table showing flows by policy area and by region for a recent ‘benchmark’ year and for an earlier year (five to ten years previously).


174

[Note: Typically, national policies with an identifiable regional incidence will include areas such as social protection, infrastructure development and education. It is recognised that there will not be a meaningful breakdown of the likes of defence or overseas aid spending.]

3 Territorial policies Character of policies. Describe the current mix of ‘territorial’ (regional, urban, rural, local and any special initiatives) policies, the eligibility criteria and how they are implemented, and their budgets and how they are financed. This should focus on national policies and exclude EU policies which are covered by other studies. However, recognising that EU policies and national policies are often closely linked, indicate where national policy plays an important part in complementing EU measures.

Provide data on the scale of expenditure on such policies by region. Distinguish between different categories of ‘territorial’ policy. Particular attention should be paid to identifying whether the financing of the policy is indigenous or comes from outside the ‘territory’. In addition, please indicate whether the cost of the policy is borne by a single funding agency or split between different stakeholders, including co-funding by the private sector.

Have there been significant changes in the last five years in the orientation or scale of territorial policies? If so, what are the key changes?

What are the principal policy instruments used to support ‘territorial’ economic development and how much is spent in total – and by region if available - on different policy measures?

4 Public sector transfers What are the principal channels through which money is transferred between regions? Distinguish between:

Grants from central government to lower tiers of government

Equalisation payments between regions

Gross tax flows from regions

Gross expenditure in regions

Any other relevant transfers, whether publicly mediated or organised by autonomous agencies such as social security funds not strictly controlled by governmental bodies

Provide data for the most recent possible ‘benchmark’ year (2000, 2001 or 2002) showing the values of these flows to and from regions. If time series data are readily available, provide annual figures for the last decade; if only occasional data are available, provide data for an earlier year between five and ten years before the benchmark figure.

For taxes and expenditure flows for which a regional incidence cannot be identified in accounting systems, are there any studies or indirect indicators that could provide information on the regional impact of the flows? For example, data on corporate profits by region might be used to estimate corporate taxes.

5 State aid Member States already report to the European Commission the amounts and kind of state aid they grant. The Commission then compiles a ‘scoreboard’. Although these harmonised data will be the basis for the analysis of state aids, data collected independently or from other sources will mainly be used for two purposes: to check the reliability of the statistics in the state aid


175

scoreboard and to refine what is already available. You are asked to supplement the Commission data in three ways:

i. By investigating and obtaining figures from a broader range of sources of state aid (regional and local authorities as well as central authorities)

ii. Distinguishing under the general heading of ‘regional state aid’ how much is spent on particular categories of policy assistance [e.g. aid for regional employment that may also be classified as general employment aid]

iii. Identifying the regional breakdown of state aid by recipient regions.

6 Employment policies Employment policies are defined here as those policy measures included in the EU Member States’ National Action Plans on employment, following the EU Employment Guidelines of the European Employment Strategy (EES)

First, provide a basic outline of national employment policy. This should draw, inter alia, on the latest National Action Plan and on information in the OECD Employment Outlook (table on ‘public expenditure and participant inflows in labour market programmes in OECD countries’)

Second, indicate to what degree regional issues are relevant in national employment policies?

• Provide a general assessment of how much national employment policies differ in different regions and of the governance of employment policies

• Describe national policies with different regional coverage; their aim and coverage

• What is the role of regional and local governments in employment policies (if any)? In particular if there are Regional Action Plans and Local Action Plans (see: Employment policies in the EU, Joint Report 2001, pages 21 and 32), how do these operate?

• Are there relevant differences between regions in the arrangements governing labour markets? Regional differences in salaries? Regional differences in non wage labour costs?

Regional data on employment policies. Supply data that documents employment policies: concentrate on expenditure and the number of participants affected by these policies, but also report other relevant information that bears on cohesion.

(Note that, in addition to national sources, useful summary information can be found in OECD, ‘Employment Outlook’ and European Commission, Employment and social affairs, ‘Employment policies in the EU and in Member States’ for national data).

Topics that fall under this heading include:

• Public employment services;

• Labour market training (training for unemployed adults; training for employed adults)

• Youth measures (unemployed youth; apprenticeships and related forms)

• Subsidised employment (subsidies to regular employment; support for unemployed starting enterprises; direct job creation)

• Measures for the disabled

• Unemployment compensation

• Early retirement for labour market reasons


176

7 Technology policies Describe the main national technology and innovation policies and explain whether, and if so how, they are applied at regional level.

Are there any specific regional technology strategies administered by

• National government bodies or agencies

• Sub-national authorities

For a recent benchmark year (2000, 2001 or 2002), provide information on the scale and regional distribution of spending on these policies.

8 Inward direct investment Are there any data on employment in foreign-owned versus indigenous industry at the regional level (provide a sectoral breakdown of such data if available)

Any data on Manufacturing (and Services) firm size by region?

Any data on regional job targets set by national agencies?

Any data on advance factory building (at regional or even national level) by national agencies?

Annual employment numbers by region will allow calculation of net job creation numbers. Is there any possibility of obtaining gross job creation and gross job destruction numbers?

We will need a list of ‘designated regions; i.e. those which the national authorities designate as warranting policy support and towards which they claim (nominally, at least) to direct extra resources and attention.


177

Appendix 1

Regional profile: [Region name]

GDP per head (2000) relative to EU-15 and to Member State

Employment rate (2000)

Unemployment rate (2002)

Technological standing (5 point scale)

Strengths [short description]

Weaknesses [short description]

Evolution in last decade [short description]

Designation for principal domestic policy support instruments (5 point scale)

Impact on ex-ante divergence from national average of support from EU funds (7 point scale)

Impact on ex-ante divergence from national average of domestic policies (7 point scale)

Change in standing of region relative to national average in last five and ten years

Estimated effect on regional circumstances of different domestic policies:

Strongly positive:

Slightly positive:

Neutral:

Slightly negative:

Strongly negative:


178

Annex 2: EPRC Report on Regional Incentives


179


180


181


182


183


184


185


186


187

Annex 3: Regional Profiles Introduction It is well documented that regional economic standings and performance in the EU show limited convergence despite the fact that at the national level economic disparities have declined markedly. Almost irrespective of economic evolutions at the national level, each Member State has both growing and declining regions relative to the EU average. In this section we briefly review the main changes in economic circumstances in the regions of Europe.

Economic evolutions During the late 1990s and early 2000s, most of the large economies in the EU - most significantly Germany - have performed relatively poorly. On the other hand, smaller countries - like Greece, Ireland, Austria, Finland but also Spain – have grown relatively fast and converged to the EU average. However, within each Member State there are significant degrees of variation in regional economic performance. Especially in countries like Germany, Italy, and the Netherlands, regional differences in economic performance were high and less so for countries like Spain, France, Greece, and Sweden. These countries also exhibit significant regional variation in the economic structures of their regions. The following two tables depict the regions that have out- or under-performed their national average in terms of economic performance (Table A.3.1) and their technological standing (Table A.3.2).

Table A.3.1. Regional profiles: evolution and relative performance (1990s)

Country Above-average Below-average Levels and within-country variation

Belgium Flanders Wallonia Level: poor Variability: low

Finland Mid-Finland, North East Level: good Variability: low-to-average

France Bretagne, Ile-de-France, Midi-Pyrenees, Pays de Loire, Rhone-Aples

Basse-Normandie, Champagne-Corse, Franche-Compte,Languedoc-Roussillon, PACA, Picardie

Level: average Variability: average

Germany Brandenburg, Mecklenburg-West. Pomerania, Sachsen,Sachsen-Anhalt, Thuringia

Niedersachsen, Nordrhein-Westfalen, Rheinland-Pfalz, Saarland, Schleswig Holstein

Level: poor Variability: average-to-high

Greece Dytiki Makedonia, Voreio &Notio Aigaio, Sterea Ellada,Kentriki Makedonia, Ipeiros

Thessalia, Kriti, Ionia Nissia,Anatoliki Makedonia-Thraki

Level: average-to-good Variability: average

Italy Basilicata, Lombardia, Toscana,Marche

Abruzzo, Aosta, E.-Romana,Lazio, Sardegna, Sicilia

Level: average Variability: high

Netherlands Groningen, Noord-Brabant,Friesland

Gelderland, Noord-Holland

Level: average Variability: high

Portugal Lisboa e Vale-do-Tejo, Madeira Alentejo, Algrave, Centro Level: poor Variability: low

Spain Asturias, Cantabria, Madrid,Navarra, Basque country

Andalucia, Baleares, Mancha Level: good Variability: average

Sweden Stockholm, West Central-north, East, North-central, North

Level: average-to-poor Variability: average

Relative to their national circumstances, regions like Flanders in Belgium, Bretagne, Ile-de-France, Midi-Pyrenees, and Rhone-Alpes in France, Central Macedonia in Greece, Lombardia in Italy, Groningen and North Brabant in the Netherlands, and the capital regions in Portugal, Spain, and Sweden, have had in the last 10 years or so a strong technological base and an above-average economic performance. In Germany, the strong technological centres have had rather poor relative economic performance and the areas that out-performed the country (Mecklenburg – Western Pomerania and Thuringia) were rather backward in relative terms. On the other hand,


188

a number of regions across Member States that have relatively weak technological bases have performed poorly also in relation to the national economic performance of their countries. Such regions include Wallonia in Belgium, the East region in Finland, Champagne-Ardennes, Corse and Picardie in France, Saarland in Germany, East Macedonia-Thrace and the Ionian islands in Greece, Aosta, Sardinia and Sicily in Italy, Algarve in Portugal, Andalucia, La Mancha and the Baleares islands in Spain, and most of the northern part of Sweden.

Table A.3.2. Regional profiles: technological standing Country Above-average Below-average Levels and within-country

variation Belgium Brussels, Flanders Wallonia Level: average

Variability: low France Alsace, Bretagne, Franche-

Compte, Ile-de-France, Midi-Pyrenees, Rhone-Aples

Champagne-Corse, Ardennes,Limousin, Nord-pas-de-Calais,Picardie, Poitou-Charentes

Level: average-to-high Variability: high

Germany Baden-Württemberg, Bayern,Hamburg, Nordrhein-Westfalen,Rheinland-Pfalz, Hessen (north)

Berlin, Brandenburg, Saarland,Sachsen, Sachsen-Anhalt,Hessen (south)

Level: high Variability: average

Greece Kentriki Makedonia, Attiki Anatoliki Makedonia-Thraki,Dytiki Makedonia, Ipeiros,Ionia Nissia, Sterea Ellada,Peloponissos, Voreio Aigaio,Notio Aigaio

Level: low Variability: high

Italy Friuli/Venezia, Lazio,Lombardia, Piemonte

Calabria, Campania, Molise,Sardegna, Sicilia, Aosta

Level: low-to-average Variability: high

Netherlands Groningen, Noord-Holland,Noord-Brabant

Friesland, Drenthe Level: average-to-high Variability: high

Portugal Centro, Lisboa / Vale-do-Tejo Norte, Algrave, Acores Level: low Variability: average

Spain Catalunia, Madrid, Navarra Andalucia, Asturias, Baleares,Canarias, Mancha, Rioja,Extremadure, Galicia, Murcia

Level: low Variability: average

Sweden Stockholm, East-central, West Smaland, North-central,Central-north, South, North

Level: high Variability: high

The impact of cohesion policies These within-country differences in economic standings and evolutions are, of course, the underlying factors that give prominence to the issue of economic convergence and cohesion policies. The following two tables illustrate the regions that are mostly affected by such policies. Table A.3.3 considers the designation of the regions for national policy support. Table A.3.4 considers the impact on the regions of cohesion policies delivered by the EU (columns 1 and 2) and the Member States (columns 3 and 4) and presents the regions that are most and least heavily affected by the implementation by Member States of non-territorial policies (columns 5 and 6).

As can be seen in Table A.3.3, the regions which are mainly targeted by the support policies are on average the regions of relative economic backwardness. What is interesting, is the varying degrees of territorial designation of the support policies (last column). Thus, such policies are less prominent (relative to the other Member States) in countries like Belgium, Finland, and the Netherlands, whereas Spain, Greece, Italy, France, and Germany seem to pay more emphasis on regional support policies, probably due to the extent of regional inequalities as well as their size. A more targeted designation is observed in France, Italy, Spain and the federal states (Belgium and Germany), while policies show lower spatial variation in the smaller countries of the EU.

The first part of Table A.3.4 illustrates the differences and complementarities in targeting that exist between the EU and national cohesion policies. In some countries, for example France, national territorial policies largely prioritise regions that receive relatively little support from the EU. In other countries, e.g., Germany, the two tiers of regional and cohesion policy are much more harmonised in terms of their targeting. If anything, however, larger differences in the targeting of the two tiers of policy exist in the regions which are least supported (columns 2 and


189

4) than in the ones that receive large amounts of support (columns 1 and 3). More detailed analysis (see section on territorial policies) seems to suggest that the existing differences do not necessarily point at functional conflicts between the two tiers of policy, but rather to complementarities and variations in the specific types of support.

Table A.3.3. Regional profiles: designation for main policy support Country Above-average Below-average Levels and within-country

variation Belgium Flanders, Wallonia Brussels Level: low (federal)

Variability: high Finland Mid-Finland, East South, Uusimaa Level: low-to-average

Variability: average France Limousin, Lorraine, Corse,

Auvergne, Midi-Pyrénées, Pays de Loire, Franche-Comté

Aquitaine, Poitou-Charentes, Centre, Rhône-Alpes

Level: average-to-high Variability: average-to-high

Germany Saxony Anhalt, Thuringia, Brandenburg, Berlin, Saxony

Hesse, Baden Württemberg, Hamburg, Schleswig Holstein

Level: average-to-high Variability: high

Greece Dytiki Makedonia, Attiki, Ipeiros, Kriti, Anatoliki Makedonia-Thraki

Sterea Ellada, Notio Aigaio, Ionia Nisia, Peloponnisos, Thessalia

Level: high Variability: average-to-low

Italy Molise, Campania, Calabria, Sicilia, Basilicata, Sardegna

Emilia-Romagna, Trentino, Veneto, Lombardia, Liguria, Aosta, Friuli-Venezia-Giulia, Marche, Toscana

Level: average-to-high Variability: average-to-high

Netherlands Groningen, Friesland, Drenthe Gelderland, Noord-Brabant,Overijsselo

Level: low Variability: low

Portugal Norte, Lisboa y Vale-do-Tejo Alentejo Level: average Variability: low-to-average

Spain Asturias, Murcia, Castilla-LaMancha, Galicia, Andalucía,Extremadura

Madrid, Navarra, Baleares, Cataluña, Rioja, Ceuta y Melilla, Pais Vasco

Level: high Variability: average-to-high

The second part of Table A.3.4, which focuses on the impact of non-territorial national policies, is based largely on subjective evaluations and is thus less informative. In general, national policies with a territorial impact are rarely evaluated in terms of their effects on regional circumstances. It follows that the estimated effects on the regions presented here cannot account for geographical spill-overs related to the effectiveness of national policies for the national economy. Maybe this - as well as cross-country differences in this study of the definitions of national policies with a territorial impact - can partly explain the lack of a general pattern in the results presented in Table A.3.4. Specifically, in some countries, like France and Spain, the more dynamic regions seem to benefit less from national policies. In contrast, in Italy, Germany, and Greece, the situation seems rather mixed, with both advanced and lagging regions benefiting from national policies. It is not clear whether this reflects the selection of policies considered, the unavoidable subjectiveness of the evaluations, or real differences across Member States in the underlying priorities set by economic policy at the national scale.


190

Table A.3.4. Regional profiles: impact of territorial and non-territorial policies by the EU and the Member States

Country EU funds National territorial policies National non-territorial policies Above-average Below-average Above-average Below-average Positive effect Negative effect

Belgium Flanders Brussels N/a N/a - -

Finland North, East Mid-Finland, South,Uusimaa

East, North Mid-Finland, South,Uusimaa

Mid-Finland, South,North, East

-

France Bourgogne, Midi-Pyrénées, Centre, Limousin, Corse

Pays de Loire, Alsace, Haute-Normandie, PACA, Bretagne, Basse-Normandie, Picardie, Rhône-Alpes

Basse-Normandie, Midi-Pyrénées, Nord-Pas-de-Calais, Lorraine, Limousin,Corse

Pays de Loire, Rhône-Alpes,Centre, PACA, Picardie,Bourgogne

Pays de Loire, PACA,Picardie, Haute-Normandie, Franche-Comté, Nord-Pas-de-Calais, Corse

Ile-de-France, Bretagne, Limousin

Germany Saxony Anhalt, Thuringia,Saxony, Brandenburg,Mecklenburg-West.Pomerania

Baden Württemberg,Rhineland Palatinate,Saarland, Hesse, Hamburg

Saxony Anhalt, Thuringia,Saxony, Brandenburg,Berlin

Hesse, Hamburg, SchleswigHolstein, BadenWürttemberg

Bremen, Mecklenburg-West.Pomerania, Thuringia, Saxony,Brandenburg, Berlin

Hesse, Hamburg,Schleswig Holstein,Baden Württemberg,Rhineland Palatinate,Saarland, Bavaria

Greece Ipeiros, Dytiki Makedonia,Ionia Nisia, Voreio Aigaio

Attiki, Kentriki Makedonia,Notio Aigaio, Dytiki Ellada,Peloponnisos

Attiki, Kentriki Makedonia,Notio Aigaio

Voreio Aigaio K. Makedonia, DytikiEllada, Thessalia,Kriti, A. Makedonia-Thraki, Ipeiros

-

Italy N/a N/a N/a N/a Veneto, Lombardia Trentino, Aosta, Friuli,Lazio, Molise, Liguria,Calabria, Basilicata,Toscana, Sardegna

Netherlands Groningen, Friesland,Drenthe, Overijssel

Noord-Holland, Gelderland,Noord-Brabant

Groningen, Drenthe,Friesland

Noord-Holland, Gelderland,Noord-Brabant, Overijssel

-

Noord-Holland, Gelderland, Noord-Brabant, Overijssel

Portugal - Acores - Algrave, Madeira, Alentejo Madeira, Centro, Norte Algrave Spain Castilla y León, Asturias,

Galicia, Castilla-la Mancha,Ceuta y Melilla,Extremadura

Madrid, Baleares, Cataluña,Rioja, Navarra, Pais Vasco

Castilla-La Mancha, Murcia,Galicia, Andalucía,Extremadura

Madrid, Baleares, Cataluña,Navarra, Pais Vasco, Ceutay Melilla, Aragón

Galicia, Andalucía Navarra, Pais Vasco,Ceuta, Rioja, Castilla,Cantabria, Canarias,Asturias, Extremadura

Date post:	20-Jan-2019
Category:	Documents
Upload:	duongxuyen
View:	217 times
Download:	0 times

The Impact of Member State Policies on...

Documents