I.3 Territorial cohesion: towardsamorebalanceddevelopment

The objective of strengthening cohesion specified inArticle 158 of the Treaty is aimed primarily at achiev-ing harmonious development of the Union as a whole.This, indeed, was the rationale for the formulation ofthe European Spatial Development Perspective(ESDP) by the informal Council of Ministers responsi-ble for spatial planning and regional policy inPotsdam in 1999. The ESDP identified three mainaims: a polycentric urban development and a newrelationship between urban and rural areas; equalaccess for all European regions to infrastructure andknow-how and prudent management of the naturaland cultural heritage. The ESDP1 sets out policyguidelines for all Member States, regions and localauthorities as well as for the Commission.

The pursuit of the same objectives, and the generalaim of achieving coherent and sustainable spatial de-velopment, underlies cross-border, transnational andtransregional cooperation, financed by the ERDF.

The enlargement of the EU to the East, which will takeplace over the coming years, following the accessionof Austria, Finland and Sweden a few years ago, willchange the Union significantly, more than doublingits land area in relation to the early 1990s. This raisesnew challenges for territorial cohesion, given the con-tinued importance of reducing regional disparities.

The Union: a very centralised territory

Historically, economic activity, as well as the capitalstock and qualified human resources, have, with afew exceptions, been concentrated in the most cen-tral areas of the Union. While regions on the south-west periphery of the Union have converged in somedegree towards the rest of the EU, including in termsof education levels, this is not yet sufficient – and is

unlikely to be in the medium-term – to undermine thevalidity of the centre-periphery model, which, indeed,is set to be reinforced with the accession of CentralEuropean countries.

Recent studies of the effect of integration on regionalbalance in the EU have emphasised the need foraccompanying policies to prevent a possible widen-ing of disparities between the stronger and weakerareas.2 This conclusion is based on the recognitionthat economic location is characterised by importantexternalities, some positive, some negative, and thatthere is no reason to think that market forces alone willstrike the right balance between positive and nega-tive effects and so result in balanced economic de-velopment across the EU as a whole. While theconcentration of economic activity in the stronger re-gions may lead to greater efficiency of production inthe EU in the short-term, this may be at the expense ofthe longer-term competitiveness of the Union econ-omy insofar as it damages the productive potential ofweaker regions and reduces their capacity to exploittheir comparative advantages. Moreover, the con-centration of both businesses and people in particu-lar regions conflicts with the objective of sustainabledevelopment, not only because of the possible over-crowding and congestion which it causes in these re-gions but also because of the rundown anddepopulation of other areas.

The evidence suggests that although, in the future,three different outcomes from EU integration can beimagined – increased concentration of economic ac-tivity, greater dispersion or little change in the existingpattern – over the past 20-30 years, the spatial patternof activity has remained much the same. Accordingly,economic activity in the Union remains concentratedto a significant extent in a relatively small central area,as indicated above. There is no evidence that the

29

increase in costs in the stronger regions resultingfrom greater congestion and higher wages will, by it-self, correct this imbalance.

To give practical content to the concept of centre-pe-riphery, an index of accessibility has been devel-oped, which measures for each region the timeneeded to reach other regions weighted by their eco-nomic importance. It should be emphasised that thisindex involves a good deal of estimation and that itrepresents the position at the present time rather thanwhat it might be in the future, given the current devel-opment of infrastructure in peripheral regions (partlyfinanced by the Structural Funds) and, perhaps moreimportantly, given the implications for the concept ofaccessibility of the development of the informationsociety. Nevertheless, the results are instructive. Re-gions can be divided into three groups in the terms ofthe index (see Map A.4):

• central regions, for which the accessibility indexis over 50% above the average for the EU plus the12 accession countries, situated in the triangleNorth Yorkshire (UK), Franche-Comté (France),Hamburg (Germany);

• peripheral regions, for which the index is under40% of the average, situated in the north of Eu-rope, in Sweden and Finland; in the north-west, innorthern Scotland and Ireland; in the south, inPortugal, Spain, the Mediterranean islands, thesouthern tip of Italy and Greece, and in the east inthe candidate countries; although the ultra-pe-ripheral regions3 were not included in the study,their accessibility is even less and they have a se-ries of structural handicaps (as mentioned in Arti-cle 299§2 of the Treaty);

• regions in between with an index of between 40%and 150% of the average.

The emerging picture is one of a very high concentra-tion of activities in central regions, which account foronly 14% of the land area but a third of the populationand almost half (47%) of the GDP. Population densityin these regions is 3.7 times higher than in peripheralregions. In all but 11 of the 88 central regions (NUTS 2level) GDP per head in 1998 was above the EU aver-age, while all but 23 of the 111 peripheral regions hada level below the average. Average GDP per head inthe central regions was twice as high as in the periph-eral ones and productivity 2.4 times higher. In 1997,expenditure on research and developmentamounted to 2.1% of GDP in the former as against0.9% in the latter. In 6 of the 7 ultra-peripheral regions,GDP per head was only around half the EU average.

The point on RTD is especially pertinent. The struc-ture of production costs of firms has changed consid-erably in recent years, with the fixed costs of researchand development increasing and costs incurred ontransport declining. Since RTD along with other stra-tegic, high value-added activities, tends to be con-centrated in central regions where the know-how andspecialist infrastructure are located, this is a factorunderlying growing polarisation in the EU and theconcentration of low value-added activities in periph-eral areas.

The transport system is also more developed in cen-tral regions. The density of motorways is four timesgreater than in peripheral ones, while there are also40% more railway lines and twice the length of dou-ble-track lines. There are signs, however, of the rela-tive position changing, especially in areas on theperiphery where the road system is most developedand is continuing to expand, which are tending to be-come important access points, such as Lisbon,Andalucia in Spain and Attiki in Greece.

The sectoral pattern of employment is also very differ-ent in central as opposed to peripheral regions. Al-though the share of employment in industry is the

30

I.3 Territorial cohesion: towards a more balanced development

Table 3 GDP per head in regions grouped according to peripherality, EU27, 1998

Population (% of total in each group) Number of regions (% of total in each group)

GDP p.c. (PPS)above average

GDP p.c. (PPS)below average

GDP p.c. (PPS)above average

GDP p.c. (PPS)below average

Total

Central regions 88.8 11.2 77 11 88

Intermediate regions 70.3 29.7 35 24 59

Peripheral regions 18.1 81.9 23 88 111

Sources: Eurostat, DG Regio

much the same (around 30% of the total), the share ofemployment in agriculture in peripheral regions isseven times larger than in central areas, whereas em-ployment in services is only 53% of the total asagainst 69%. This, of course, reflects underlyingcompetitiveness, which helps explain why the em-ployment rate in peripheral regions is under 59%while in central ones, it is just over 67% (see Table 3and Table A.6).

This concentration of economic activity and popula-tion in such a restricted area of the Union has adverseeffects not only on the peripheral regions but also onthe central ones, where it is responsible for traffic con-gestion and strong pressure on the environment.Whereas transport bottlenecks in peripheral areasare a result of the low standard of infrastructure and alack of connections, in central regions, they arise fromcapacity constraints and excessive traffic (see MapsA.5 and A.6).

A consequence of this congestion and the concentra-tion of economic activity is that toxic emissions in cen-tral areas are 2.3 times greater than in peripheralones4 (see Map A.7).

With the accession of the 12 applicant countries, theUnion will include many more areas where the level ofdevelopment is well below the average. A new eastern,continental periphery will be added to the existingsouthern, maritime one. As a result, economic activitywould tend to be even more regionally concentratedthan in the US, where activity is more evenly distrib-uted, despite its land area being twice as large as anenlarged EU and its population being much smaller(270 million inhabitants, 44% less than in the EU).

Four separate areas of global importance in eco-nomic terms can be distinguished in the US, eachwith over 15 million people and with GDP per headabove the US average in all the individual States in-cluded. These areas together account for 28% of thetotal US land area, 49% of the population and 54% of

national GDP and, accordingly, display a much lowerlevel of concentration than in the EU, though physicalgeography is clearly a contributing factor (Table 4,Graphs A.5, A.6 and Map A.8).

Urban areas

Growth centres for achievingpolycentric development

Connections between urban centres, and betweenthese and rural areas, are a major force for economicdevelopment. In the Union, it is the central areaswhich have the highest level of infrastructure endow-ment and most developed urban networks (see MapA.9).

The studies undertaken for the ESDP divided regionsinto 6 categories: those dominated by a large metro-politan area; densely populated urban regions withpolycentric, or dispersed, economic development;regions with a high density of population in urban ar-eas but containing rural areas and with polycentricdevelopment; rural areas under the influence of ametropolitan area; rural areas with small and mediumsized towns; and remote rural areas.

Polycentric development can occur at two levels:

• at the EU level, with development dispersedacross the Union as a whole in a number of dy-namic growth centres, in contrast to the tendencyfor economic activity to be concentrated in cen-tral regions with other areas being peripheral;

• at a regional level, in areas where there are anumber of urban centres, often interconnected,rather than a single dominant one. In Europe, thetypical examples are the Randstadt in the Nether-lands and the Rhine region in Germany, whichcontrast with the Paris region in France orsouth-east England where development tends tobe focused on Paris and London, respectively.

The central regions of the Union conform very muchto a polycentric model of development, while in pe-ripheral regions, rural areas with small and me-dium-sized towns are much more important(accounting for 47% of the population there). In thelatter, however, the development of urban centres

31

I.3 Territorial cohesion: towards a more balanced development

Table 4 Centres of economic activity inthe EU and the US

% landarea

%population

%GDP

Central regionsin the EU

13.8 33.0 47.2

Developmentcentres in the US

28.1 48.8 53.5

0 100 500 km

SIG16SIG16

Regions dominated by a large metropolis

Polycentric regions with high urban andrural densities

Rural areas under metropolitan influence

Rural areas with small and medium sized towns

Remote rural areas

Polycentric regions with high urbandensities

6 Types régionaux deconfigurations spatiales rurales-urbaines

GDR LIBERGEO - EQUIPE P.A.R.I.S.

Sources : Moriconi - Ebrard, Geopolis, 1994, et Eurostat.

1.3 Territorial cohesion: towards a more balanced development

32

could provide the basis for a more balanced distribu-tion of economic activity across the Union (Map 6).

Over 44% of the population of the EU lives in citieswith over 50,000 inhabitants, but only 22 of these cit-ies have over a million people.5

The most prosperous regions in the Union are virtuallyall urban (see Box) – 54 of the 63 NUTS 3 regions6 withGDP per head of over 50% above the EU average.

Urban areas: significant disparities

Unemployment in EU urban areas taken together aver-aged 10.1% in 1999, slightly above the EU average(9.2%). Long-term and youth unemployment were alsohigher than in the rest of the Union, though the ratesvary considerably between Member States and seemto be affected more by national factors than by factorsgeneral to cities across the EU (see Tables A.8-A.10).

Many European cites have serious problems of pov-erty and social exclusion in particular districts. Ac-cording to the Urban Audit, in cities like Hamburg,Toulouse, Naples, Genoa, Glasgow or Edinburgh,unemployment rates can vary significantly betweendistricts, being up to 10 times higher in the worst af-fected parts than in the least affected. The same istrue of dependency rates.

Rural areas

Over two out of three people in southern Europe andIreland live in rural areas while the figure is under 1 in8 in Belgium, the UK, Germany and Italy (see GraphA.7).

Except in Portugal, the population in rural areas is in-creasing in all Member States, though at differing rates(see Graph A.8 and Table A.11.):

• in Belgium, Germany, Greece and Spain, growthis well above the national average;

• in Italy, the UK and Austria, the rate is similar tothat in other areas;

• in Denmark, France, Ireland, Finland and Swe-den, growth is below that elsewhere and, in Portu-gal, population is declining.

Over the period 1995 to 1999, the growth of employ-ment in rural areas (1.0% a year) in the Union washigher than the overall rate (0.8% a year). The ruralcharacter of a region is, therefore, not an obstacle tojob creation. On the contrary, an attractive naturaland cultural environment, with lower levels of conges-tion, can be an important factor in encouraging busi-ness investment.

At the sectoral level, while rural areas have sufferedlarge scale job losses in agriculture, they have alsoexperienced job gains in industry – except in Ger-many and Austria – and, above all, in services, whichhave more than compensated for these.

33

I.3 Territorial cohesion: towards a more balanced development

Data sources

The two sources used for the analysis are the EU La-bour Force Survey and the Urban Audit.

The LFS enables the data to be broken down intothree kinds of area according to population density:

• urban areas: densely populated, made up of con-tiguous local units, each with a density of over 500inhabitants per square km and with a total popula-tion of at least 50,000;

• intermediate areas: made up of less densely-pop-ulated contiguous local units, each with a densityof over 100 inhabitants per square km and with atotal population of at least 50,000, or of unitsneighbouring a densely-populated area;

• rural areas: made up of sparsely populated localunits located outside urban or intermediate areas.It should be noted that local units with low popula-tion density but completely encircled by an urbanor intermediate area are counted as part of this,while those situated between an urban and inter-mediate one are counted as part of the latter.

The Urban Audit was carried out between 1998 and2000 with the aim of measuring the quality of life intowns and cities in the Union. The pilot phase, whichis the basis of the data cited in the text, covered asmall sample of 58 urban centres, so the resultsneed to be interpreted with care.

The economic and social features of rural areas varysignificantly across the Union as well as within indi-vidual Member States. Three kinds of area can be dis-tinguished in terms of their links with the rest of thenational and international economy and their remote-ness from major centres of activity:

• rural areas which are integrated into the overalleconomy and which are characterised by eco-nomic and population growth. These are oftenclose to urban centres and have, in general,above average income per head. Jobs are pre-dominantly in industry and services rather thanagriculture. In some cases, such areas are at riskof becoming purely residential areas for peopleworking elsewhere, posing a threat to their tradi-tional environment and their social and culturalheritage; others, however, are developingindependently;

• intermediate rural areas, often some distanceaway from urban centres, but with good links tothese and a reasonable level of infrastructure. Inmany such areas, agriculture continues to play asignificant role, though they tend to be experienc-ing increasing economic diversity and growingactivity in, for example, food processing as wellas services. In a number of Member States, largefarms are situated in these areas;

• remote rural areas, usually sparsely populatedand in many cases located in peripheral parts ofthe Union far from urban centres. Their relativeisolation is often due to their topography, such astheir mountainous nature, and they tend to have ahighly dispersed and ageing population, poor in-frastructure, inadequate services, low incomeper head, a relatively unskilled work force, weaklinks with the rest of the economy and high em-ployment in agriculture.

Border regions

Border regions7 cover almost 40% of the EU land areaand account for 25% of the population. They are evenmore important in the candidate countries, account-ing for 66% of the land area and 58% of population.Enlargement will, therefore, lead to a significantgrowth in their prevalence in the Union.

In economic terms, the regions with borders internalto the EU cannot in general be regarded as havingmore difficulties than other regions, in part due to theextent of economic integration in the Union and thesuccess of the INTERREG initiatives. In particular,their level of GDP per head (15% above the averageof the enlarged EU of 27 countries) is similar to that ofnon-border regions (17% above the enlarged EU av-erage – see Map A.10 and Table A.12).

Regions with external borders, however, are in a moredifficult situation, with the notable exception of thosewhich border the candidate countries. While thosewith borders with third countries have a level of GDPof 5% below the (enlarged) EU average, those bor-dering the candidate countries have a level which is15% above the new average of 27 countries. Never-theless, some of these regions might well face tempo-rary difficulties after enlargement.

There are much more significant differences betweenthe regions of the present EU bordering the candi-date countries and the neighbouring ones in the can-didate countries themselves. In the latter, GDP perhead is only 53% of the (enlarged) EU average, iemuch less than half the level in neighbouring regionsin the present EU. Nevertheless, they are still betteroff than regions in the candidate countries with east-ern borders, GDP per head in which averages only37% of that of the enlarged EU.

Areas with specific geographical features

Mountainous areas, coastal and maritime regions, is-lands and archipelagos form an important part of theUnion and are even more significant in some MemberStates. Most of the ultra-peripheral regions are islands.These, however, do not form a distinct geo-morpholog-ical area as such, but are treated as a group of 7 re-gions listed in the Treaty and recognised as having anumber of inherent disadvantages, particularly be-cause of the problem of accessibility caused by theirremoteness from other parts of the Union.

While the regions identified as being entitled to struc-tural assistance from the Structural Funds are definedin terms of administrative and socio-economic crite-ria, the geo-morphological areas are distinguished interms of their physical features. These are not alwayseasy to define and often there is no commonly-ac-cepted definition (urban, rural and so on). Moreover,

34

I.3 Territorial cohesion: towards a more balanced development

the features concerned are not always synonymouswith structural problems.

The three main types of geo-morphological area areconsidered below.

Mountain areas

Mountainous areas represent geographical barriers.Over time, activities concentrated in the valleys whichare natural passages, but today many of these havebecome transport bottlenecks and the growth of trafficof goods and people involves increasing risks to safetyand the environment. Areas such as the Alps, Pyre-nees, Dolomites, the Greek mountains, the Highlandsof Scotland and Fjällen in Sweden cover approximately39% of the EU land area. In many of these areas, eco-nomic activity is concentrated in agriculture – on theland which is usable – tourism and other services. Theothers have very little economic activity at all. Whilesome mountainous areas are economically viable andintegrated into the rest of the EU economy, most haveproblems, as witnessed by the fact that more than 95%of them (in terms of land area) are eligible for assis-tance under Objectives 1 or 2 of the Structural Funds(Map 7 and Tables A.13 and A.14).

Coastal and maritime areas

Coastal areas are defined as those situated on the stripof land around the coasts of the EU, which is of variablewidth depending on geographical features and admin-istrative boundaries. They include many large cities(London, Stockholm, Copenhagen, Helsinki, theHague, Dublin, Lisbon, Barcelona, Marseilles, Rome,Naples and Athens) and cover a significant part of theEU land area. Many of the areas are densely populatedwith a high level of tourist activity, generating significantincome but also substantial environmental pressure,the reconciliation of which poses a serious challenge.Other areas, however, are scarcely populated at all.The growth of maritime traffic involves increasing risks

to safety, the environment and conservation of thecoastline(see Table A.15).

Islands

Islands are particularly important in the four southernMember States, three of which are cohesion coun-tries, though there is also a large number of islands inFrance, the UK and the three Nordic countries, manyof them eligible for Structural Funds support (TablesA.16 and A.17). Indeed, nearly 95% of the populationof EU island regions is eligible for such support underObjectives 1 or 2. In the case of the smaller islands,accessibility is the main problem which makes it diffi-cult to maintain economic activities which are com-petitive and a young work force with a high level ofeducation. Accessibility is an even greater problemfor ultra-peripheral regions. The largest islands aremuch better integrated into the rest of the EU econ-omy, even if many are at present reliant on structuralsupport to catch up with other parts of the Union.

The areas identified above have marked differences interms of their economic and social characteristics. Re-gional policies for furthering their development shouldcontinue to be aimed at strengthening relations be-tween different parts of the Union rather than take theform of isolated measures specific to individual types ofarea. Nevertheless, such policies should include coop-eration programmes between areas of the same type,which are tailored to their particular geographical fea-tures and which can bring additional benefit.

35

I.3 Territorial cohesion: towards a more balanced development

1 In parallel with the development of the ESDP, a programme of studies was launched by the Commission in December 1998, in closecollaboration with the 15 Member States. As a result of this programme, the basis of a common understanding on territorial analysisemerged, an issue which at a Union level had up to then been largely neglected. This part of the report makes use of the results of thestudies undertaken over a period of 10 years.

2 See, in particular, ‘Integration and the regions of Europe: how the right policies can prevent polarisation’, Braunerhjelm et al.3 Canary Islands, Guadeloupe, Martinique, Guyane, Réunion, Açores and Madeira.4 ESDP study programme.5 Approximate figures – source: EUROSTAT-GISCO.6 GDP per head at NUTS 3 level is not always a reliable measure of income because of commuting.7 NUTS 3 regions in the 27 countries eligible for INTERREG III-A or the PHARE-CBC programme.

36

I.3 Territorial cohesion: towards a more balanced development

Objective 1Objective 1

Phasing-out(till 31/12/2005)Phasing-out(till 31/12/2006)

Special programme

Objective 2Objective 2

Objective 2 (partly)

Phasing-out(till 31/12/2005)Phasing-out (partly)(till 31/12/2005)

Regional boundariesNUTS2 boundaries

© MEGRIN for the administrative boundaries

0 100 500 km

7 Structural Funds 2000-2006: Areas eligible under Objectives 1 and 2

Guyane (F)

Guadeloupe

(F)

Martinique

(F)

Réunion

(F)

Canarias (E)

Açores (P)

Madeira

(P)

SIG16SIG16