regions: statistical yearbook 2005 · cronos available. however, to assist comprehen-sion of the...

THEMEGeneral andregional statistics

20

05

ED

ITIO

N

E U R O P E A NC O M M I S S I O N

Regions: Statistical

yearbook 2005Data 1999-2003

PA

NO

RA

MA

O

F

TH

E

EU

RO

PE

AN

U

NI

ON

A great deal of additional information on the European Union is available on the Internet.It can be accessed through the Europa server (http://europa.eu.int).

Luxembourg: Office for Official Publications of the European Communities, 2005

ISBN 92-894-9029-2ISSN 1681-9306

© European Communities, 2005

Europe Direct is a service to help you find answers to your questions about the European Union

Freephone number (*):

00 800 6 7 8 9 10 11(*) Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.

CONTENTS■ INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Regional data in the spotlight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Enlargement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Specialist input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

NUTS 2003 — regions list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

More regional information needed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Regional interest group on the web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Closure date for the yearbook data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

■ POPULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Fertility trends in the 25 EU countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Regional differences in fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Towards an explanation of regional differences in fertility . . . . . . . . . . . . . . . . . . . 18

Regional fertility differences within and between countries . . . . . . . . . . . . . . . . . . . 19

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

■ AGRICULTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Methodological note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Cereal production in Europe’s regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Wheat growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Barley growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Grain maize growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Crop production in Europe’s regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Potato growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Sugar beet growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

Rape growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

■ REGIONAL GROSS DOMESTIC PRODUCT . . . . . . . . . . . . . . . . . . . . . . . . . . 37

What is regional gross domestic product? . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Purchasing power parities and international volume comparisons . . . . . . . . . . . . . 39

Regional GDP in 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Major regional differences within countries too . . . . . . . . . . . . . . . . . . . . . . . . 41

Three-year average for GDP from 2000–02 . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Dynamic development in peripheral regions . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

■ HOUSEHOLD ACCOUNTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Introduction: Measuring wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Private household income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

The measurement unit for regional comparisons . . . . . . . . . . . . . . . . . . . . . . 49

Results for 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

R e g i o n s : S t a t i s t i c a l y e a r b o o k 2 0 0 5 5

Primary income and disposable income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Income and GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Income and direct taxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

■ REGIONAL LABOUR MARKET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Data sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Female employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Self-employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

Unemployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

■ TRANSPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73


Maritime transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Air transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

Road freight transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

■ SCIENCE, TECHNOLOGY AND INNOVATION . . . . . . . . . . . . . . . . . . . . . . . . 83

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85


Research and development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

Human resources in science and technology . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Employment in high-tech manufacturing and knowledge-intensive services . . . . . . . . . . 93

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

■ STRUCTURAL BUSINESS STATISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97


Services concentrated around capitals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Differences in regional wage levels smallest in the Netherlands . . . . . . . . . . . . . . . . . 98

Capital-intensive regions in the EU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Most technology-intensive industries in Germany . . . . . . . . . . . . . . . . . . . . . . . 102

Most specialised regions in EU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

■ HEALTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109


Mortality in EU regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Diseases of the circulatory system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Men and lung cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

R e g i o n s : S t a t i s t i c a l y e a r b o o k 2 0 0 56

Women and ovary cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Healthcare resources in EU regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Hospital discharges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Hospital beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

■ URBAN STATISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

What is the Urban Audit? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Choice of cities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Spatial units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Time line data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Perception survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Dissemination of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Urban liveability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Transport modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

■ EDUCATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131


Participation of 17 year olds in education . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Students in general upper secondary education . . . . . . . . . . . . . . . . . . . . . . . . . 132

Tertiary education students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

Tertiary education attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Lifelong learning participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

■ TOURISM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Methodological notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Capacity (infrastructure) statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Occupancy data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

■ EUROPEAN UNION: NUTS 2 REGIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

■ CANDIDATE COUNTRIES: NUTS 2 REGIONS . . . . . . . . . . . . . . . . . . . . . . . . 151


I N T R O D U C T I O N

Regional data in thespotlightThe decisions on the Structural Funds for2007–13 will probably be taken in 2005. As thesepolitical decisions are based on objective quanti-tative regional data, Eurostat’s offer of a rich setof comparable regional statistics is in the mediaspotlight.

This yearbook highlights many aspects of region-al data and the analyses which can be made withthem. Some fundamental chapters, such as re-gional GDP, regional household accounts, thelabour market, demographic statistics and agri-culture, are a recognised backbone of the book.One chapter is new this year: education, givinginteresting insights into topics such as lifelonglearning participation in the various regions.

As last year for the first time, all regional analysisin this yearbook is based on NUTS 2003. In themeantime, the 10 new Member States have alsobeen integrated formally into the new regionalclassification (in the form of an amendment to theNUTS regulation). The texts of the regulation andthe amendment are available on the CD-ROM —as is the annex, which lists the regions making upthe nomenclature in each country.

EnlargementNo distinction is made in the yearbook betweenthe old Member States, the countries that becameMember States in 2004 and those due to joinaround 2007: wherever data are available for Bul-garia and Romania, these of course also feature inthe maps and commentaries. In the case of Turkeyand Croatia, the situation is rather different. Al-though a regional breakdown has been agreed be-tween these two countries and Eurostat, there arestill too few regional data to justify includingthem in the analyses.

StructureIn each chapter, regional distributions are high-lighted by colour maps and graphs which are thenevaluated by experts in text commentaries. In

keeping with the traditions of the yearbook, an ef-fort has been made to focus on aspects not re-cently covered. The population chapter, for exam-ple, is this year devoted to the fertility rates acrossthe regions where we see a large regional spread.The transport chapter, which reappears this year,focuses on maritime and aviation data.

As last year, the CD-ROM does not contain thedata tables previously specially compiled for theyearbook. With Eurostat’s databases available on-line, free of charge, since 1 October 2004, there isno justification for such tables when users havethe entire wealth of tables in the database New-Cronos available. However, to assist comprehen-sion of the maps, the data series used for the mapsin the yearbook are provided as Excel files on theCD-ROM.

To enable readers to make the fullest possible useof the public database, the CD-ROM again con-tains the latest edition of the reference guide to thedatabase.

Specialist inputThe commentaries in each thematic chapter re-flect the specialist knowledge of Eurostat’s the-matic units (1). By exploiting their experience ofdata at national level, the authors are in a positionto place the regional variation noted in an appro-priate context. The regional statistics team grate-fully acknowledges the contribution made by thefollowing authors, each of whom has had to findthe necessary time within an already overcrowdedschedule:

Chapter Author(s)

1. Population Erik Beekink, Joop de Beer

2. Agriculture Dagmar Binova

3. Regional GDP Andreas Krüger

4. Household accounts Andreas Krüger

5. Regional labour market Michal Mlady

6. Transport Carla Sciullo

7. Science, technology August Götzfried,and innovation Simona Frank and

Håkan Wilén


(1) In the case of the chapters on regional GDP, household ac-counts, regional labour market and urban statistics, theauthors are simultaneously members of the regional statisticsteam and the subject specialists within Eurostat.

8. Structural business statistics Petra Sneijers

9. Health Sabine Gagel

10. Urban statistics Teodóra Brandmüller, Berthold Feldmann

11. Education Birgitta Andrén

12. Tourism Hans Werner Schmidt

NUTS 2003 —regions listIn the maps in this yearbook, the statistics are pre-sented at NUTS 2 level. A map giving the codenumbers of the regions can be found in the sleeveof this publication. At the end of the publicationthere is a list of all the NUTS 2 regions in theEuropean Union, together with a list of the level 2statistical regions in Bulgaria and Romania. Fulldetails of these national regional breakdowns, in-cluding lists of levels 2 and 3 regions and the ap-propriate maps, may be consulted on theRAMON server (2).

More regionalinformation needed?The public NewCronos database contains moreextensive time series (which may go back as far as1970) and more detailed statistics than those givenin this yearbook, such as population, death andbirth by single years of age, detailed results of theCommunity labour force survey, etc. Moreover,there is coverage in NewCronos of a number ofindicators at NUTS 3 level (such as area, popula-tion, births and deaths, gross domestic product,unemployment rates). This is important becausethere are no fewer than eight EU Member States(Cyprus, Denmark, Estonia, Latvia, Lithuania,Luxembourg, Malta and Slovenia) that do nothave a level 2 breakdown.

For more detailed information on the contents ofthe database in NewCronos, please consult theEurostat publication ‘European regional andurban statistics — Reference guide 2005’, a copy

of which is available in PDF format on theaccompanying CD-ROM.

Regional interestgroup on the webEurostat’s regional statistics team maintains apublicly accessible interest group on the web(‘CIRCA site’) with many useful links and docu-ments.

To access it, simply use the URL:

http://forum.europa.eu.int/Public/irc/dsis/regstat/information

Among other resources, you will find:

• a list of all regional coordination officers in theMember States, the candidate countries and theEFTA countries;

• the ‘Regional Gazette’ published at intervals bythe regional team;

• the latest edition of the regional and urban ref-erence guide;

• PowerPoint presentations of Eurostat’s workconcerning regional and urban statistics;

• the regional classification NUTS for the Mem-ber States and the regional classification of thecandidate countries.

Closure date for theyearbook dataThe cut-off date for this issue is 4 May 2005.


IN

TR

OD

UC

TI

ON

(2) See http://europa.eu.int/comm/eurostat/ramon/index.cfm?TargetUrl=DSP_PUB_WELC

http://forum.europa.eu.int/Public/irc/dsis/regstat/information

http://europa.eu.int/comm/eurostat/ramon/index.cfm?TargetUrl=DSP_PUB_WELC

P O P U L A T I O N 1

IntroductionSince the 1960s, fertility has decreased consider-ably in the Member States of the European Unionand in Bulgaria and Romania. However, figuresshow that not all countries experience this declinein the same way, and to the same level. This raisesthe question of the extent to which countriesexperience a common development in fertility dueto common socio-cultural and socioeconomictrends (labelled the ‘second demographic transi-tion’) and the extent to which differences in fertil-ity levels between countries are persistent. Whenlooking for an answer, it is useful to examine dif-ferences in the fertility level between regions. Thischapter compares differences in fertility ratesfrom region to region in the same country withdifferences between countries. If regional differ-ences within countries are relatively small com-pared with differences between countries, thissuggests that country-specific causes are predom-inant in explaining the fertility level. If the oppos-ite is true, this suggests that explanations for dif-ferences in fertility are to be found in factors thatapply transnationally.

The next section of this chapter gives a generaloverview of fertility trends in the 25 MemberStates of the European Union. The following sec-tion examines differences in the fertility rates ofregions at NUTS 2 level. The subsequent sectiondiscusses possible influences on the fertility level

on the basis of available literature. Finally, the ex-tent to which regional differences can be ex-plained by differences between countries is exam-ined, i.e. the extent to which country-specificcauses are likely to affect the level of fertility.

Fertility trends in the25 EU countriesFrom the mid-1960s until the end of the 1990s,the total fertility rate in the EU-25 showed a near-ly constant downward trend (Graph 1.1). The totalfertility rate (TFR) is often used as an indicator for the fertility level, since it makes adjustmentsfor changes in the size and structure of the femalepopulation. The TFR of a given year is the meannumber of children born alive to women who ex-perienced, during their childbearing years, theage-specific fertility rates of that specific year.

In 1964, the average total fertility rate of the EU-25 was 2.72 children per woman. By 1999, therate had decreased to 1.42. During the last coupleof years, the fertility rate for the European Unionseems to have stabilised at around a level of 1.46children per woman.

Not all countries in the Union experienced this de-crease in the same way. Graph 1.2 shows the total


1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Graph 1.1 — Total fertility rates in the EU-25, 1960–2003

Year

Tota

l fer

tilit

y ra

tes

0

0.5

1.0

1.5

2.0

2.5

3.0


PO

PU

LA

TI

ON

1fertility rates by Member State for the years 1972and 2002. In 1972 as well as in 2002, the highestfertility rate occurred in Ireland, where the TFR wasnearly 3.9 and 2 respectively. In 1972, the countryin the EU with the lowest total fertility rate was Fin-land with 1.6. Finland is the only country for whichthe TFR in 2002 exceeded that in 1972. This differ-ence with the general trend is caused by a very lowfertility rate in relative terms in Finland for 1972.

In 2002, the Czech Republic had the lowest fertil-ity rate (1.2). The graph shows that the largest de-crease during this period took place in countrieswhere the TFR in 1972 was high: Ireland, Spain,Portugal, Romania and the Slovak Republic. Arelatively small decrease took place in countrieswhere the TFR was already low in 1972: Ger-

many, Luxembourg, and Sweden. This indicatesthat the total fertility rates in the EU-25 and inBulgaria and Romania have converged. In 1972,the difference between the highest (Ireland) andlowest (Finland) rates equalled 2.7. In 2002, thedifference, then between Ireland and the CzechRepublic, was down to 0.8 (see also Chapter D inthe 2004 edition of ‘Population statistics’). Eventhough there has been a converging downwardtrend in the TFR, there are still considerable dif-ferences in fertility rates between EU countries, asshown by the graph. The overall picture can besummarised by noting that low TFR values aremeasured in central and south European coun-tries, whereas high values are observed in westand north European countries.

CZ

SK

SI

BG

LV

LT

PL

ES

RO

EL

IT

HU

DE

EE

AT

MT

PT

CY

BE

LU

UK

DK

IE

FI

NL

SE

FR

20021972

Graph 1.2 — Total fertility rates by country, 1972 and 2002

2.0 3.01.21.0 4.01.4 1.6 1.8 2.2 2.4 2.6 2.8 3.2 3.4 3.6 3.8

Regional differences infertilityAt regional level, the differences in the level of fer-tility within the EU-25 are even more pronounced

than at national level. Map 1.1 shows the so-called ‘general fertility rates’ for the NUTS 2 re-gions in the EU-25 and in Bulgaria and Romania.A lack of appropriate data for all regions made itnecessary to use general fertility rates (GFR) in-stead of total fertility rates. It is important to notethat the general fertility rate only makes adjust-ments for the total number of women of child-bearing age, but not for the age structure of that


Average general fertility rates

2000–2002 — NUTS 2

> 7.26.1–7.25.3–6.14.6–5.3≤ 4.6Data not available

Data are not available for all countriesfor the years 2000–02

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundaries Cartography: Eurostat — GISCO, April 2005

Map 1.1

group; these rates are calculated by dividing thenumber of births in a given year by the mean totalnumber of women of childbearing age (in thisstudy, we used women in the 20–44 age group). Incontrast to the TFR, the GFR does not take ac-count of age-specific differences in fertility rates.However, the outcomes of this less precise ratewill not change the overall picture of the fertilitydifferences between EU regions. The figures usedhere are based on a calculated average for theyears 2000–02. For some countries, data for 1999had to be used.

The map shows that, during this period, the high-est fertility rates were in North Finland (Pohjois-Suomi), both Irish NUTS 2 regions ‘Border, Mid-land and Western’ and ‘Southern and Eastern’,parts of the Netherlands (Friesland, Drenthe,Overijssel and Flevoland), and France (Nord-Pas-de-Calais, Champagne-Ardenne, Picardie, Haute-Nomandie, Île-de-France, Centre, Basse-Nor-mandie, Bretagne, Pays de la Loire,Franche-Comté and Rhône-Alpes).

Some of the regions with the lowest general fertil-ity rates were in northern Spain (Galicia, Princi-pado de Asturias, Cantabria, País Vasco andCastilla y León) and the eastern part of Germany(Mecklenburg-Vorpommern, Magdeburg,Dessau, Halle, Leipzig, Thüringen, Chemnitz,Dresden, Hamburg and Bremen).

The three regions with the highest general fertilityrates were Ciudad Autónoma de Melilla in Spain(10.6), followed by the Irish region ‘Border, Mid-land and Western’ (10.5) and Flevoland in theNetherlands (10.2). Two of the three regions withthe lowest fertility were also in Spain: Principadode Asturias (4.3) and Galicia (4.8). The third wasthe region of București in Romania (4.7).

Even without sophisticated technical analyses, avisual inspection of the map clearly shows geo-graphical clusters of regions with similar fertilitylevels. Fertility is relatively high in most regions ofIreland, the Netherlands, Finland and France andlow in most regions of Spain, Italy, Greece and thecentral European Member States. More specifi-cally, fertility in northern regions of France tendsto be higher than in southern regions. In Belgium,fertility in the Walloon provinces exceeds that inthe Flemish provinces. In Germany, fertility is rel-atively low in the eastern regions.

Towards anexplanation ofregional differencesin fertilityAs mentioned before, the decline of fertility to lev-els below the so-called replacement level (about2.1 children per woman) that started in the mid-1960s occurred in most European countries. Thissuggests that there may be one general explana-tion for this trend. Van de Kaa and Lesthaeghe in-troduced the concept of the ‘second demographictransition’ in an attempt to provide a frameworkfor explaining changes in family life. These de-mographers claim that ‘the new shifts in demo-graphic patterns result from the interplay of struc-tural, cultural, and technological factors during acomplex process of social change. The welfarestate ensures citizens an income and protects themfrom the vagaries of life. New, highly efficientcontraception has been introduced; restrictionson abortion and sterilisation have in many casesbeen lifted. Significant changes in value systemshave been documented. These ideational transfor-mations accentuate individual autonomy, involvethe rejection of all forms of institutional controlsand authority, and show a rise of expressive val-ues connected with self-fulfilment’. Van de Kaanotes in a more recent essay that, for a while, itlooked like the new transition process would re-main limited to northern and western Europe.However, data for the 1990s show that southernand eastern Europe are increasingly affected. Es-pecially in the 10 new Member States, a period ofrapid demographic change can be observed after1989. Fertility declined sharply. At first, this wasattributed to crisis conditions, but it soon turnedout that these ‘changes in fertility are part of abroader transition in reproductive and family lifemarked by the spread of alternative family forms,non-marital births, postponement of births anddecline in marriage rates which has been gainingground in the west European societies since the1960s’ (Sobotka, 2001).

In their Working Paper for the European Com-mission (3/2004/F/nr 4), Duchêne, Gabadinho,Willems and Wanner study the causes of differ-ences in regional fertility. In an overview of the lit-erature, they conclude that two main types of fac-tors can be used to explain regional differences:


PO

PU

LA

TI

ON

1

the socioeconomic structure of the population (bysocio-occupational class, level of education, na-tionality, etc.) and ‘contextual’ factors linked tothe place of residence (for example, cultural fea-tures, availability of infrastructure, the housingmarket situation). In their conclusion, they pointout that ‘the few studies simultaneously using re-gional data for several countries would tend toshow that the State to which a region belongs is animportant parameter in explaining the level of fer-tility; variations within States are less substantialthan variations between States. In other words, atEuropean level, a large part of the variance of re-gional fertility indicators is “international” ratherthan “intranational”.’

This very concise overview of the aforementionedliterature raises the question of the extent towhich European regions experience a commondevelopment in fertility and the extent to whichdifferences between countries persist. The conceptof the second demographic transition suggeststhat there is one common development in fertilityacross Europe. This seems to be confirmed by theconvergence of fertility rates in the separate EUcountries. However, as noted above, considerabledifferences between countries still exist. The con-clusion of Duchêne et al., namely that regionalvariations in fertility within States are smallerthan between States, suggests that country-specif-ic factors affecting the fertility level play an im-portant part. This would support the hypothesisthat even though there has been a tendency to-wards the convergence of fertility levels acrossEuropean countries, there will not be completeconvergence, and important inter-country differ-ences will persist.

Regional fertilitydifferences withinand betweencountriesGraph 1.3 shows the minimum and maximumvalues of the general fertility rate for each country.This graph indicates the extent to which fertilitydiffers between countries and the extent to whichfertility differs between NUTS 2 regions withincountries. The graph shows that the degree ofvariability differs between countries. Spain shows

by far the biggest differences, as it includes boththe regions with the lowest and the highest levelsof fertility in the European Union. Romania alsoshows relatively high variability of fertility levelsbetween regions. Other countries show less differ-ences in fertility levels between their regions. Nev-ertheless, the ranges of the regional fertility levelsin most countries do show some overlap.

In order to examine the extent to which inter-country differences affect regional differences, aregression model is estimated in which the gener-al fertility rates of all NUTS 2 regions in theEuropean Union are explained using dummy vari-ables for separate countries. These variables areassumed to reflect the economic and cultural dif-ferences between countries that affect the fertilitylevel (in the absence of more detailed informationat NUTS 2 level, only a few explanatory variablescould be used for this model).

Cultural factors refer to norms for the ‘ideal’ fam-ily size. Socioeconomic factors refer to opportuni-ties and restrictions, for example income level,employment and childcare facilities. In additionto factors at national level, economic differenceswithin countries affect the fertility level at region-al level. It appears that in regions with relativelyhigh long-term unemployment, the fertility level issignificantly lower than in other regions. If long-term unemployment as a proportion of total un-employment in region A exceeds that in region Bby 10 percentage points, the GFR in region A is onaverage 0.3 lower than that in region B.

The regression results show that for the regions ofnine countries, the GFR differs significantly fromthe average for the European NUTS 2 regions.The unweighted average GFR for the NUTS 2 re-gions equals 8.4. Taken together, the inter-countrydifferences and the economic conditions (as meas-ured using the long-term unemployment figures)explain 73 % of the variance of the GFRs for theregions of the European Union and the accessioncountries.

For regions in four countries, the general fertilityrate is systematically higher than the EU average:Ireland, France, the Netherlands, and (part of)Belgium. In the two regions of Ireland, the GFR ison average 2.5 higher than the European average.In France, fertility in the north is higher than inthe south. In its northern and southern regions,the GFR is respectively 2.0 and 1.4 higher thanthe average European level. In the Netherlands,the GFR is 1.4 higher than the European average.


In Belgium, the GFR is high in the Walloon re-gions (2.0 higher) but not in the Flemish regions.

The regions of five countries have a low GFR. Inthe regions of Austria, the GFR is 1.2 lower thanaverage. It was mentioned above that the regionsin Spain show a very wide range of values. Never-theless, the fertility level in the regions of Spain isgenerally significantly lower than the Europeanaverage. The average GFR for the Spanish regionsis 1.0 lower than the EU average, indicating thatthe regions with a high level of GFR in Spain areclearly an exception. In the Czech Republic, theGFR is 0.9 lower and in Italy 0.8 lower than theEU average. In Germany, there are systematic dif-ferences between eastern and other Länder. In the

eastern Länder, the GFR is 1.4 lower than the EUaverage, while fertility in other Länder is only 0.3lower.

In the above discussion of the map, the regionswith the highest and lowest levels of fertility wereidentified. As mentioned, some of these regionsare part of clusters of regions with high or low fer-tility rates. However, there are also regions inwhich fertility is considerably lower or higherthan in the other regions of the same country.These regions can be identified by inspecting theresiduals of the regression model. For example, asalready noted, the highest GFR is measured inSpain, in the Ciudad Autónoma de Melilla. ItsGFR equals 10.6, which is considerably higher


PO

PU

LA

TI

ON

1

CZ

SI

LV

SK

EL

ES

LT

BG

IT

AT

DE

PL

RO

EE

HU

PT

CY

UK

BE

SE

MT

LU

FI

DK

NL

FR

IE

Graph 1.3 — Minimum and maximum regional general fertility rates by country,average 2000–02, NUTS 2 (*)

6.04.0 11.05.0 7.0 8.0 9.0 10.0

(*) Data are not available for all countries for the years 2000–02.6.04.0 11.05.0 7.0 8.0 9.0 10.0

than in the other Spanish regions. In the Italianprovince Campania, the GFR equals 7.4, which isnot very high compared with all other Europeanregions, but it is relatively high if one considersthat the GFR in Italy in general is relatively low,and that the proportion of long-term unemploy-ment in this region is very high (74 % of total un-employment). Another region with a relativelyhigh fertility rate is Pohjois-Suomi, the northern-most part of Finland. The GFR there equals 10.1,compared with 8.3 on average in the rest of Fin-land. If we look at regions with a low fertility lev-el, it turns out that in București the GFR is con-siderably lower than in the other regions ofRomania, in Bratislavský kraj considerably lowerthan in the other Slovakian regions and in thePrincipado de Asturias lower than in the otherSpanish regions.

ConclusionSince the 1970s, there has been a general down-ward trend in the fertility level in the countries ofthe European Union and in Bulgaria and Roma-nia. During this process, differences in the level offertility between countries have declined. Thisconverging trend can be explained by the conceptof the ‘second demographic transition’, assuminga common pattern of cultural change acrossEurope. Nevertheless, considerable differences stillcontinue to exist. This is reflected in the consider-able regional differences in the level of fertility.The general fertility rate of most NUTS 2 regionsranges between 4 and 8. There are considerabledifferences in the regional fertility level withincountries, which can only partly be explained by

differences in economic conditions (if long-termunemployment is relatively high, fertility tends tobe low). Regional differences in fertility can to alarge extent be explained by differences betweencountries. This indicates that economic and cul-tural differences between countries have an im-portant effect on the level of fertility. In view ofthese significant differences, one may questionwhether differences in fertility in Europe will dis-appear. It seems more likely that regional differ-ences will persist.

LiteratureDuchêne, J., A. Gabadinho, M. Willems and P.Wanner, ‘Study of low fertility in the regions ofthe European Union: places, periods and causes’.Working Paper population and social condition3/2004/f/nr. 4, Luxembourg, 2004.

Eurostat, ‘Population statistics’, Luxembourg,2004

Lesthaeghe, Ron and Dirk J. van de Kaa, ‘The sec-ond demographic transition revised: theories andexpectations’. In: Population and family in thelow countries, Amsterdam, 1994.

Sobotka, T., ‘Ten years of rapid fertility changes inEuropean post-communist countries: evidenceand interpretation’. Paper presented at the EAPSConference, Helsinki, 1–9 June 2001.

Van de Kaa, Dirk, ‘Demographies in transition:an essay on continuity and discontinuity in valuechanges’. In E. Kotowska and J. Jozwiak (eds.),Population of central and eastern Europe. Chal-lenges and opportunities, Warsaw 2003.


A G R I C U L T U R E 2


IntroductionAgriculture, including forestry and aquaculture,as a major land user plays a key role in deter-mining food and feed safety, as well as the for-mation of the rural landscape. Half of the EU’sland is farmed, highlighting the importance offarming for the EU’s natural environment. Pro-duction of high-quality products demanded bythe market in harmony with the environment isa priority of European agriculture. This year’sedition of regional yearbook focuses on certaintypes of crop production, namely on arablecrops in 2002. The content is divided into twomain parts. The first one looks at cereal pro-duction (wheat, barley and grain maize), thesecond one is wider and consists of a regionalcomparison of potatoes, sugar beet and rapeproduction.

Methodological noteAll data used in this chapter for comparison ofthe regions were the latest which were com-pletely known and available in the NewCronosdatabase in March 2005, and relate to the year2002.

The indicator in Map 2.1 is calculated as the areaunder cereal production as a proportion of theutilised agricultural area (UAA). The indicator inMap 2.5 is calculated as the area under potatoproduction as a proportion of the UAA. Accord-ing to nomenclature for land-use statistics, theUAA includes arable land, land under permanentcrops, permanent grassland, kitchen gardens,crops under glass (and excludes wooded area,other area). The level of shares in % is influencedby the surface of the UAA and its proportion ofthe land area and of the total area (land area +inland waters). In particular, this affects Finland,Sweden, Estonia, Latvia and Slovenia, where theUAA portion of the total area is lower than 25 %.

Comparison of areas, production and yields at na-tional level was done on the basis of agriculturaldatabase (not regional). Some minor differencesbetween national and regional statistics are due tousing different statistical approaches (e.g. oilseedsdon’t include flax and cotton seeds at regionallevel).

Cereal production inEurope’s regionsArable crops include a wide range of annual cropsof primary importance, such as wheat, barley,maize, rape, sugar beet, potatoes, etc. In 2002,they covered around 42.5 % of the EuropeanUnion’s utilised agricultural area, and are foundin all the Member States. Cereal production is oneof the most important outputs of European agri-culture. Cereals are herbaceous plants of thegraminaceous family (except for buckwheat) cul-tivated mainly for their grain. Whole cereals areused primarily for animal feed and human con-sumption. They are also used to produce drinksand industrial products (e.g. starch).

Cereals (including rice) are the largest group ofgrowing crops in the world. In 2002, the EU-25produced nearly 267.6 million tonnes of cerealsand the area under cereals reached 53.2 millionhectares. France, the largest cereal producer in theEU-25, harvested 69.7 million tonnes of cereals,followed by Germany (43.4 million tonnes),Poland (26.9 million tonnes) and the United King-dom (23.0 million tonnes). France, Germany andPoland account for more than half of total pro-duction. The 10 new Member States accountedfor around 20 % of the EU-25 total harvest and29 % of the EU-25 area under cereals.

Cereals are of considerable importance in regionswhere they account for more than 50 % ofutilised agricultural area (Map 2.1). These regionsare found in Balkan countries (Severoiztochen inBulgaria, Sud-Est, Sud, Sud-Vest, București in Ro-mania, Anatoliki Makedonia, Thraki, KentrikiMakedonia and Dytiki Makedonia in Greece), incentral Europe — mainly in Hungary (Közép-Dunántúl, Nyugat-Dunántúl, Dél-Dunántúl, Dél-Alföld), then in Slovakia (Bratislavský, ZápadnéSlovensko), in Poland (Lódzkie, Lubelskie,Wielkopolskie, Dolnośląskie, Opolskie and Ku-jawsko-Pomorskie), in the Czech Republic (Pra-ha, Střední Čechy) and in Germany (Sachsen-An-halt). Coverage of greater than 50 % also exists innorthern Europe (Denmark, Finnish regionsEtelä-Suomi and Länsi-Suomi) as well as in south-ern Europe (Italian region Basilicata). In westernEurope the significant portion of cereal area ofUAA is recorded in the East of England, in BelgianRégion de Bruxelles-Capitale/Brussels Hoofdst-edelijk Gewest (although this region is the small-est in comparison to all European regions) and in


AG

RI

CU

LT

UR

E

2

French regions: Île-de-France, Picardie, Centreand Alsace.

Low representation of cereals in the utilised agri-cultural area is found firstly in southern regions,Alpine regions and the British Isles, where theyoccupy less than 10 % of UAA. In detail, this in-cludes seaside areas in Spain (Galicia, Principa-do de Asturias, Cantabria, Comunidad Valen-ciana and Canarias), in Portugal (Algarve,

Região Autónoma dos Açores, Região Autóno-ma da Madeira), in Italy (Liguria) and in Greece(Ionia Nisia, Peloponnisos, Attiki and Kriti).Alpine regions in Austria (Kärnten, Salzburg,Tirol, Vorarlberg) and in Italy (Valled’Aosta/Vallée d’Aoste, Provincia AutonomaBolzano/Bozen, Provincia Autonoma Trento)and highlands in Belgium (Prov. Luxembourg),in France (Corse, Limousin and overseasdepartment Réunion) and in the United Kingdom

Cereals (including rice)as a % of utilised agricultural area

2002 — NUTS 2> 5035–5020–3510–20≤ 10No cereal productionData not available

DE, UK: NUTS 1

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundariesCartography: Eurostat — GISCO, March 2005

Map 2.1


(North West, Wales, Scotland and Northern Ire-land) prefer grasslands, possibly green fodder, toarable lands for cereal growing. Some regions inthe Netherlands (Friesland, Overijssel, Gelder-land, Utrecht and Noord-Holland) as well as thewhole of Ireland, Bremen urban region in Ger-many and Mellersta Norrland in Sweden, alsooccupy a low proportion of area under cerealscompared to all UAA. Malta is not a producer of

cereals, neither are the French Caribbean regions(Guadeloupe, Martinique).

The most commonly grown cereal crops in theEU-25 are wheat, barley and grain maize, eventhough rice, for instance, predominates in FrenchGuyane. In 2002, area under cereals other thanwheat, barley and grain maize made up only

Wheat (yield 100 kg/ha)

2002 — NUTS 2

> 8060–8040–6020–40≤ 20No or negligible wheat productionData not available

DE, UK: NUTS 1

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundariesCartography: Eurostat — GISCO, April 2005

Map 2.2


AG

RI

CU

LT

UR

E

219.4 % of the total area under cereals in the EU-25 (13.5 % of cereal production).

Wheat growing

Wheat (Map 2.2), including soft wheat, durumwheat and spelt, accounted for 43.9 % of thetotal area under cereals and 46.6 % of the totalEU-25 cereal production in 2002. Wheat produc-tion at 125 million tonnes, of which durum wheatproduction reached 9.9 million tonnes, was 9.3 % higher than in 2001. Average yield in theEU-25 was 53.4 quintal per hectare in 2002. TheEU-15 recorded 57.8 quintal per hectare, whichshows a yield of 19.3 quintal per hectare higherthan in the 10 new Member States. The threelargest producers of wheat in the EU-25, France,Germany and the United Kingdom, accounted formore than 60 % of the total production.

The highest level of yields occurs in north-westernEurope. In the first place, this includes Belgium(Prov. Limburg, Prov. Oost-Vlaanderen, Prov.Vlaams Brabant, Prov. West-Vlaanderen, Prov.Brabant Wallon, Prov. Hainaut, Prov. Liège, Prov.Namur) where the highest value of yields wasreached in Provinces Limburg and Liège —around 88.8 quintal per hectare. Also Germany(Nordrhein-Westfalen, Schleswig-Holstein),Northern France (Île-de-France, Champagne-Ar-denne, Picardie, Haute-Normandie, Nord-Pas-de-Calais), Ireland, the Netherlands (Zuid-Holland,Zeeland, Noord-Brabant, Limburg) and the Unit-ed Kingdom (North East, Yorkshire and TheHumber, East Midlands, Eastern, South East)recorded yields higher than 80 quintal perhectare. High productivity in these regions iscaused by the use of modern technology, artificialand natural fertiliser, well-developed techniqueand last but not least by favourable climatic con-ditions. Similar potential also exists in some re-gions in eastern Europe, but local units are learn-ing how to use these tools effectively, and holdingsare applying for support (subsidies, grants) to in-vest in modern instrumentation and machinery. Itwill be interesting to see whether these regionscapitalise on these possibilities.

The lowest level of yields is recorded in the southof Europe, especially for regions below 45 degreeslatitude. Pyrenean peninsula regions such asAragón and Región de Murcia in Spain, almostthe entire mainland of Portugal and their islands(Iles Balears, Canarias, Região Autónoma daMadeira) are among areas with lowest wheat pro-ductivity, although Spain is one of the six biggest

producers in the EU-25. Basilicata, Sicilia andSardegna fall among the least wheat fertile regionsin Italy. The mountainous environment does notfavour high yields in regions in the Balkan penin-sula (Sud-Est, Sud, Sud-Vest, București, Yugoza-paden, Kentriki Makedonia, Attiki) and in islands(Ionia Nisia, Voreio Aigaio, Notio Ai gaio).

Some urban regions (Bremen, Région de Brux-elles-Capitale/Brussels Hoofdstedelijk Gewest,Berlin, Bratislavský, Hamburg and London)recorded a very low volume of production withregard to small surface areas (in comparison with‘larger’ regions). Alpine regions (Provincia Au-tonoma Trento, Provincia AutonomaBolzano/Bozen, Valle d’Aosta/Vallée d’Aoste, Vo-rarlberg, Tirol, Salzburg) as well as overseas andisland regions (Guadeloupe, Martinique, Guyane,Réunion, Região Autónoma dos Açores, Ca-narias, Região Autónoma da Madeira, Corse,Malta), Nordic regions (Mellersta Norrland,Övre Norrland, Pohjois-Suomi) almost nevergrow any wheat. Similar cases are the coastal re-gions of Principado de Asturias, Cantabria, Lig-uria and Ipeiros.

Barley growing

Barley production in the EU-25 was 56.5 milliontonnes and was harvested from 13.3 millionhectares in 2002. It accounted for 25.0 % of thetotal area under cereals and 21.1 % of the totalEU-25 cereal production in 2002. The EU-25 bar-ley production recorded an average yield of 42.4quintal per hectare: in the EU-15 the average yieldwas up by 3.2 quintal per hectare, whereas in the10 new Member States it was down by 12.2 quin-tal per hectare. The EU-15 made up around 85 %of the harvest and occupied 79 % of the area un-der barley. The three largest producers of barley inthe EU-25, France, Germany and Spain, account-ed for almost 54 % of the total production and50.4 % of the total area.

Map 2.3 shows that the western countries andtheir regions belong to the most productive barleyzones, except for cities such as Berlin, Bremen,Hamburg in Germany and the surroundings ofBrussels and London; bearing in mind that theselargely urban areas have an area under barleylower than 500 hectares. Yields in 14 regions ofNorth France, in eight regions of Belgium, easternUnited Kingdom and Germany (both countries:four regions) exceeded the level of 60 quintal perhectare. Both France and Germany, the twolargest producers, reached harvests of almost


11 million tonnes and their average yields ac-counted for 66.9 and 55.5 quintal per hectare re-spectively. In particular, in southern France,northern Italy, south-western Austria, the majori-ty of Germany, Denmark, western United King-dom and Ireland, yields of between 45 and 60quintal per hectare are typical.

The highest level of barley yields was recorded inthe Dutch region of Utrecht (90 quintal per

hectare), although its area and production is neg-ligible from this point of view. Drier regions inSpain (Principado de Astúrias, Canarias), in Por-tugal (Lisboa, Região Autónoma dos Açores,Região Autónoma da Madeira), in French Corseand in Greek Ionia Nisia are not focused on bar-ley growing. Neither are French overseas depart-ments, nor mountainous areas in Austria and Italy(Tirol, Vorarlberg, Provincia Autonoma

Barley (yield 100 kg/ha)

2002 — NUTS 2

> 6045–6030–45≤ 30No or negligible barley productionData not available

DE, UK: NUTS 1


Map 2.3


AG

RI

CU

LT

UR

E

2Bolzano/Bozen, Provincia Autonoma Trento,Valle d’Aosta/Vallée d’Aoste, Liguria).

In the north, south and east of Europe, the barleyyields level was less than 45 quintal per hectare. Inevaluation of regions, the largest areas under bar-ley were occupied by the Spanish region Castilla yLeón (1.3 million hectares) and the highest level

of production was shown in Denmark (4.1 mil-lion tonnes). Generally, in countries which are bigproducers of beer (per capita: the Czech Republic,the Netherlands, Denmark, Belgium and Ger-many), part of the barley harvest serves to pro-duce malt and processing in the brewing industry.Otherwise, the majority of barley production isused for animal feed (around 60 %).

Grain maize (yield 100 kg/ha)

2002 — NUTS 2

> 9070–9050–70≤ 50No or negligible grain maize productionData not available

DE: NUTS 1

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundaries Cartography: Eurostat — GISCO, March 2005

Map 2.4

Grain maize growing

The EU-25 harvest of grain maize reached 50.3million tonnes in 2002. The former EU-15 pro-duced 40.5 million tonnes and made up almost80.5 %. Grain maize plays an important role inanimal feeding (in particular for pigs and poul-try). It was grown in 16 Member States and occu-pied an area of 6.2 million hectares (in the wholeof the EU-25), while Romania accounted for al-most 2.9 million hectares. Grain maize accountedfor 11.6 % of the total area under cereals and18.8 % of the total EU-25 cereal production in2002. In particular, high yields of this cereal gen-erally contribute to the big proportion in that pro-duction.

According to the Map 2.4, the highest value yields(higher than 90 quintal per hectare) can be foundin Austria, Belgium, Germany, Spain, France,Greece, Italy, the Netherlands and also one regionin the Czech Republic and one in Portugal. To-gether, France (16.4 million tonnes) and Italy(10.6 million tonnes) accounted for half of the to-tal production. The new Member States, exceptfor the Czech Republic and Slovenia, had loweryields than 70 quintal per hectare or did not pro-duce any grain maize at all. A similar situation isnoticeable in regions of Bulgaria and Romania,where their yields were lower than 70 quintal perhectare (except for Yugoiztochen).

Although grain maize is a crop from tropical ar-eas, it is grown in various climatic conditions:Nordic counties, Denmark, Sweden, Finland, theUnited Kingdom, Ireland, Estonia, Latvia and themost northern German region (Schleswig-Hol-stein) do not produce any grain maize in the cold-er climatic zone. Urban areas in Germany (Berlin,Bremen, Hamburg), in the Czech Republic (Pra-ha) and in Belgium (Région de Bruxelles-Capi-tale/Brussels Hoofdstedelijk Gewest) did notgrow grain because of the low amount of arableland and different priorities of its use. Also, is-lands such as Cyprus, Malta, Região Autónomada Madeira, Guadeloupe, Martinique and FrenchGuyane did not grow any grain maize at all.Greek Notio Aigaio and Attiki as well as theAlpine regions Vorarlberg, Valle d’Aosta/Valléed’Aoste and Provincia Autonoma Bolzano/Bozenand Dutch Friesland and Zuid-Holland recordedonly insignificant production of grain maize.

On the whole, it is necessary to say that maize isranged with the principal fodder crops because ofusing green maize and silage.

Crop production inEurope’s regionsThe domain of crop production other than cerealscovers dried pulses, root crops, industrial crops,fodder crops, vegetables and fruit, includinggrapes and olives. From an arable land use pointof view, root crops and oilseeds are the greatestall-European crops. Root crops are field crops,which are very productive and are able to achievehigher farming yields than other crops. Mostly,they produce energy rich substances and productswith low dry matter contents. The importance ofroot crops is attached to the high production abil-ity of organic matter, which assures the energycomponent in human consumption and animalfeedingstuffs. Root crops serve for foods and ani-mal feeds and are a raw material for industrialprocessing. The most important root crops arepotatoes and sugar beet.

Potato growing

The potato is grown primarily for human con-sumption, but it is also used to feed cattle and pro-duce alcohol and potato flour (starch). The pota-to is a tuber, the thickened underground stem ofthe potato plant, especially designed for the stor-age of starch. Starch and moisture content are keypotato characteristics.

Potato growing has been steadily falling in thewhole of Europe in the last 15 years. In 2002,potato production in the EU-25 was 66.7 milliontonnes and was harvested from 2.3 millionhectares (the 10 new Member States occupied 47% of this area) with an average yield of 286.5quintal per hectare. The EU-15 recorded an aver-age yield of 371.4 quintal per hectare in potatoproduction, whereas in the 10 new Member Statesthe average yield reached only 189 quintal perhectare. Yields higher than 400 quintal perhectare were recorded only in regions of the for-mer EU-15, namely in Belgium, Denmark, Ger-many, Spain (La Rioja), France, the Netherlandsand the United Kingdom. These countries includethe five most productive countries of the EU-15(the Netherlands, France, Belgium, Germany andthe United Kingdom) and in 2002 they made uparound 35.7 % of the areas and 52.8 % of theproduction in the EU-25. It is evident that a dif-ferent level of yields in regions of the new Mem-ber States and in regions of the former EU-15 in-fluences the level of the production and market in


the whole of Europe. The level of yields in coun-tries and regions is also determined by the portionof early and other potatoes out of total potato ar-eas and production.

Map 2.5 shows that the biggest areas under pota-toes as a proportion of UAA dominate in regionsof Belgium (Prov. Oost-Vlaanderen, Prov. VlaamsBrabant, Prov. West-Vlaanderen, Prov. Hainaut),

the Netherlands (Groningen, Drenthe, Flevoland,Noord-Holland, Zuid-Holland, Zeeland, Noord-Brabant, Limburg) and Poland (Lódzkie, Ma-zowieckie, Malopolskie, Slaskie, Lubelskie, Pod-karpackie, Swietokrzyskie). If we add SpanishCanarias, French Nord-Pas-de-Calais, PortugueseRegião Autónoma da Madeira (30.9 %) andMalta, we have all the areas with more than a 5 % proportion of the UAA. Map 2.5 points out


AG

RI

CU

LT

UR

E

2Potatoes as a percentage

of utilised agricultural area2002 — NUTS 2

> 5.02.5–5.01.0–2.50.5–1.0≤ 0.5No or negligible potato productionData not available

DE, UK: NUTS 1


Map 2.5

that Poland reached the largest area with a highproportion of UAA. Poland as the biggest pro-ducer of potatoes in the EU-25, occupied 34.5 %of the total area under potatoes (harvested 23.3 % of production) in 2002. Viewed in thelonger term, Poland shows a downward trend inthe area under potatoes, although it constantlyoccupies around 77.6 % (74 % in 2002) of thenew Member State areas and has a predominant

role. Generally, regions in eastern countries pro-duce, process and consume (people, possibly ani-mal) more potatoes per capita than in the formerEU-15 (mainly in west and south European re-gions).

Not all European regions produce potatoes in asignificant volume. In particular, urban regions inBelgium (Région de Bruxelles-Capitale/BrusselsHoofdstedelijk Gewest), the Czech Republic


Sugar beet production(1 kg per capita)2002 — NUTS 2

> 1 000500–1 00010–500≤ 10No sugar beet productionData not available

DE: NUTS 1


Map 2.6

(Praha), Germany (Berlin, Bremen, Hamburg),French overseas departments (Guadeloupe, Mar-tinique, Guyane), Austria (Wien) and also Basili-cata in Italy.

Sugar beet growing

In 2002, sugar beet was the most extensive rootcrop grown in Europe. It is primarily grown as aprocessing/industrial crop (raw material for sugarproduction), and in a small measure it is used forfeeding purposes. The root of sugar beet is used toproduce sugar or alcohol. The leaves and top areused for animal feed or are left in the field. Some-times, the roots are also used for animal feed.

Sugar beet occupied 2.4 million hectares and ac-counted for 48.3 % of the total EU-25 area underroot crops in 2002. Germany, France and Polandeach farmed more than 10 % of area under sug-ar beet (49.7 % in total) in the EU-25. Thelargest areas (more than 80 000 hectares) arefound in German Niedersachsen (117 100hectares), in French Champagne-Ardenne and Pi-cardie (88 100 and 164 000 hectares respectively)and in eastern Britain (93 200 hectares). From thevolume point of view, Germany and France werethe best producers (50.9 % of the EU-15 and42.7 % of the EU-25). The highest production(more than 5 million tonnes) was recorded in Ger-man Bayern (5.3 million tonnes) and Niedersach-sen (6.4 million tonnes) and in French regionsChampagne-Ardenne (6.9 million tonnes), and Pi-cardie (12.4 million tonnes). The highest levels ofaverage yields (more than 800 quintal per hectare)were reached in Spain regions La Rioja (807 quin-tal per hectare), Castilla y León (838 quintal perhectare) and Castilla-la Mancha (836 quintal perhectare) and in French regions Centre (852 quin-tal per hectare) and Alsace (813 quintal perhectare). Western Europe dominated the success-ful growing of sugar beet. The contribution of the10 new Member States was 16.1 % of the EU-25production. This is a similar percentage as a shareof the EU-25 population.

Map 2.6 provides a comparison of sugar beet pro-duction level per capita. It is understandable thatsome European regions produce more than theirinhabitants can use and some of them produceless or almost nothing (Kärnten in Austria, Ré-gion de Bruxelles-Capitale/Brussels Hoofdst-edelijk Gewest in Belgium, Yugozapaden in Bul-garia, Comunidad Foral de Navarra in Spain;Lorraine, Pays de la Loire and Rhône-Alpes inFrance, Sud in Romania). In 2002, more than one

tonne of sugar beet per person was predominant-ly produced in regions of the former EU-15. Thisincludes eastern regions of Austria (Burgenland,Niederösterreich), central Walloon regions (Prov.Brabant Wallon, Hainaut, and Namur), GermanSachsen-Anhalt, Spain Castilla y León, Finnish is-lands (Åland), north France (Champagne-Ar-denne, Picardie, Haute-Normandie and Nord-Pas-de-Calais), north-eastern Greece (AnatolikiMakedonia, Thraki), Italy (Emilia-Romagna andMarche), northern Netherlands (Groningen,Drenthe, Flevoland and Zeeland) and southernSwedish Sydsverige. Only the Polish region of Ku-jawsko-Pomorskie represented the new MemberStates in overproduction per capita.

Rape growing

Seeds and oleaginous fruit are cultivated not onlyfor their richness in fatty matters (intended for hu-man consumption), but also for their richness inprotein (necessary for animal feed). The followingspecies are taken into account in the seeds andoleaginous fruit: colza and the rapeseed namedcolza-rape, sunflower, soya, flax, cotton, ricinus,groundnuts, copra, palm nuts and almonds, otherseeds such as mustard, poppy, hemp, sesame andother fruit not specified elsewhere. Some of theproducts under consideration are exotic and arenot cultivated in the European Union. Therefore,they arrive uniquely as imports in the MemberStates.

In the year 2002, almost 7.3 million hectares ofoilseeds were harvested in the EU-25 and 8.8 mil-lion hectares including areas of Romania and Bul-garia. The biggest areas belong to Germany,France and Romania (46.9 % of the EU-25 +Romania and Bulgaria). Rape was the most im-portant oilseed in the EU-25 (57.5 % of area)and the second position was occupied by sun-flower (29.4 % of the area and around 22 % ofthe production). In Belgium and in the Nether-lands, oil flax predominated in oilseed growing.In Portugal, rape and turnip rape were nevergrown and so sunflower occupied all areas underoilseeds. Sunflower also predominated in Spain,Hungary, Bulgaria and Romania. In Italy, sun-flower made up around half of the area underoilseed, and areas under soya were close behind.Other oilseed growing predominated in Slovenia(pumpkins for oil), Cyprus and Greece (cottonseed).

Oilseed production in the EU-25 reached around17.2 million tonnes (around 19 million tonnes,


AG

RI

CU

LT

UR

E

2

including production of Romania and Bulgaria)and rape and turnip rape production representedaround 68 % of this harvest. With detailed focuson rape production it is possible to find regions(in Map 2.7) which produced more than 50 kg ofrape per capita. Among them are Austrian Bur-genland, five Czech regions (Střední Čechy, Ji-hozápad, Severovýchod, Jihovýchod, StředníMorava), six German regions (Brandenburg,

Mecklenburg-Vorpommern, Sachsen, Sachsen-Anhalt, Schleswig-Holstein, Thüringen), eightFrench regions (Champagne-Ardenne, Picardie,Haute-Normandie, Centre, Bourgogne, Lorraine,Franche-Comté, Poitou-Charentes), four Polishregions (Zachodniopomorskie, Opolskie, Ku-jawsko-Pomorskie, Warminsko-Mazurskie), Zá-padné Slovensko in Slovakia and British EastMidlands. These regions are excellent for rape


Rape production(kg per capita)

2002 — NUTS 2

> 5010–501–10≤ 1No rape productionData not available

DE, UK: NUTS 1


Map 2.7


AG

RI

CU

LT

UR

E

2growing and this is supported by the fact thattheir production was higher than 100 000 tonnes(only Burgenland, Franche-Comté, Warminsko-Mazurskie, Severovýchod and Střední Moravaaccounted for lower production). Regions in theCzech Republic, Germany, Denmark, France,Lithuania, Poland, Slovakia and in the UnitedKingdom represent typical areas for rape produc-tion.

ConclusionBoth climatic and geographical conditions exert agreat influence on land use. Preferences in animaland crop production differ from region to regionin the whole of Europe.

But it is important to point out that quality andintensity of production are not the only factorswhich determine the evolution in the agriculturalsector. Other criteria (rural development, environ-ment, food safety, animal welfare, etc.) will be-come more and more important and will certain-ly influence and change the picture of our presentagriculture.

AG

RI

CU

LT

UR

E

REGIONAL GROSS DOMESTIC PRODUCT3

What is regionalgross domesticproduct?

The economic development of a region is, as arule, expressed in terms of its gross domestic prod-uct (GDP). It is also an indicator frequently usedas a basis for comparisons between regions. Butwhat exactly does it mean? And how can compa-rability be established for regions of different sizeand different currencies?

Regions of differing size achieve different regionalGDP levels. However, a real comparison can onlybe made by indicating the regional GDP of thepopulation for the region in question. This iswhere the distinction drawn between place ofwork and place of residence becomes significant:gross domestic product measures the economicperformance achieved within national or regionalboundaries, regardless of whether this was attrib-utable to resident or non-resident employed per-sons. Reference to GDP per inhabitant is thereforeonly straightforward if all employed persons en-gaged in generating this value are also residents ofthe region in question.

In areas with a high proportion of commuters, re-gional GDP per inhabitant can be extremely high,particularly in such economic centres as Londonor Vienna, Hamburg, Prague or Luxembourg, andrelatively low in the surrounding regions, even ifthese are characterised by high household pur-chasing power or disposable income. RegionalGDP per inhabitant should not, therefore, beequated with regional disposable income.

Regional GDP is calculated in the currency of thecountry in question. In order to make GDP com-parable between countries, it is converted intoeuro using the official average exchange rate forthe given calendar year. However, not all differ-ences in price levels between countries are reflect-ed by exchange rates. In order to equate the cur-rencies, GDP is converted using currencyconversion rates, known as purchasing power par-ities (PPPs), to an artificial common currency,called purchasing power standards (PPS). Thismakes it possible to compare the purchasing pow-er of different national currencies (see box).


Purchasing power parities andinternational volume comparisons

International differences in GDP values, even afterconversion via exchange rates to a common curren-cy, are not due simply to differing volumes of goodsand services. The ‘level of prices’ component is alsoa major contributing factor. Given that exchangerates are determined by many factors influencingdemand and supply in the currency markets (suchas international trade, inflation expectations and in-terest rate differentials), conversion via exchangerates in cross-border comparisons is of limited use.To obtain an exact comparison, it is essential to usespecial conversion rates (spatial deflators) which re-move the effect of price level differences betweencountries. Purchasing power parities (PPPs) aresuch currency conversion rates that convert eco-nomic data expressed in national currencies to anartificial common currency, called purchasing pow-er standards (PPS). PPPs are therefore used to con-vert the GDP and other economic aggregates (e.g.consumption expenditure on certain productgroups) of various countries into comparable vol-umes of expenditure, expressed as purchasing pow-er standards.

With the introduction of the euro, prices cannow, for the first time, be compared directly be-tween countries in the euro-zone. However, theeuro has different purchasing power in the dif-ferent countries of the euro-zone, depending onthe national price level. PPPs must therefore alsocontinue to be used to calculate pure volume ag-gregates in PPS for Member States within theeuro-zone.

In their simplest form, PPPs are a set of price rel-atives, which show the ratio of the prices in na-tional currency of the same good or service in dif-ferent countries (e.g. a loaf of bread costs EUR1.87 in France, EUR 1.68 in Germany, GBP 0.95in the United Kingdom, etc). A basket of compa-rable goods and services is used for price increas-es. These are selected so as to represent the wholerange of goods and services, taking account of theconsumption structures in the various countries.The simple price ratios at product level are aggre-gated to PPPs for product groups, then for overallconsumption and finally for GDP. In order tohave a reference value for the calculation of thePPPs, a country is usually chosen and used as thereference country, and set to 1. For the EuropeanUnion, the selection of a single country as the ref-erence country is inappropriate, so the PPS of theEU is used as an artificial common unit of refer-ence to express the volume of economic aggre-gates for the purpose of spatial comparisons inreal terms.

Unfortunately, for reasons of cost, it will not bepossible in the foreseeable future to calculate re-gional currency conversion rates. If such regionalPPPs were available, the GDP in PPS for


RE

GI

ON

AL

G

RO

SS

D

OM

ES

TI

C

PR

OD

UC

T

3

GDP per inhabitant, in PPS

2002 — NUTS 2

EU-25 = 21 170> 26 00020 000–26 00014 000–20 0008 000–14 000≤ 8 000Data not available


Map 3.1

numerous peripheral or rural regions of the EUwould probably be higher than that calculated us-ing the national PPPs.

The regions may be ranked differently when calcu-lating in PPS instead of euro. For example, in 2002,the German region of Dessau was recorded as havinga per capita GDP of EUR 15 638, putting it wellabove Malta with EUR 10 757. However, with PPS15 499 per capita, Malta ranks above Dessau with itsPPS 14 085 per capita.

In terms of distribution, the use of PPS rather than theeuro has a levelling effect, as regions with a very highper capita GDP also generally have relatively highprice levels. This reduces the range of per capita GDPin NUTS 2 regions in the EU-25 plus Bulgaria andRomania from around EUR 73 000 to around PPS62 000.

Per capita GDP in PPS is the key variable for deter-mining the eligibility of NUTS 2 regions in the frame-work of the European Union’s structural policy.

Regional GDP in2002

Map 3.1 provides an overview of the regional dis-tribution of per capita GDP (in PPS) for the Euro-pean Union, plus Bulgaria and Romania. It rangesfrom PPS 4 337 per capita in north-east Romaniato PPS 66 761 per capita in the UK Inner Londonregion. Brussels (PPS 49 645) and Luxembourg(PPS 45 026) follow in second and third place,with Hamburg (PPS 39 766) and the French capi-tal region Île-de-France (PPS 37 267) in fourthand fifth place.

Prague (Czech Republic), the region with thehighest GDP per inhabitant in the new MemberStates with PPS 32 357 (153 % of the EU-25 av-erage), has already risen to 14th place (2001:15th) among the 268 NUTS 2 regions of the coun-tries examined here (EU-25 plus Bulgaria and Ro-mania). It should be noted, however, that Pragueis an exception. The next regions of those joiningthe EU in 2004 follow some way behind:Bratislavský kraj (Slovakia) is in 46th place(2001: 64th) with PPS 25 351 (120 %), Közép-Magyarország (Hungary) is 127th (2001: 147th)with PPS 20 329 (96 %), Cyprus is 163rd (2001:148th) with PPS 17 558 (82.9 %), Slovenia is187th (2001: 186th) with PPS 15 941 (75.3 %),Malta is 194th (2001: 190th) with PPS 15 499(73.2 %) and Mazowieckie (Poland) 204th(2001: 203rd) with PPS 14 718 (69.5 %). Allother regions of the new Member States have aper capita GDP in PPS of less than two-thirds ofthe EU-25 average.

Of the 268 regions examined here, the per capitaGDP (in PPS) in 2002 was below 75 % of the EU-25 average in 78 of them. This figure is down sig-nificantly on 2001 when the corresponding groupcomprised 85 regions. In Spain and Greece in par-ticular, a number of regions crossed the 75 % ofper capita GDP barrier in 2002. At the other endof the spectrum, 37 regions had a per capita GDPof more than 125 % of the EU-25 average in2002, down from 38 in 2001. The year 2002 didtherefore see further progress in economic con-vergence among the regions of the 27 countriesexamined here.

Major regionaldifferences withincountries tooThere are also substantial differences within thecountries, as Map 3.1 shows. As in 2001, thehighest per capita GDP was more than twice thelowest in 12 of the 19 countries examined here in-corporating NUTS 2 regions. The largest regionaldifferences are in the United Kingdom, wherethere is a factor of 4.3 between the two extremevalues (Inner London: 315 % of the EU-25 aver-age; Cornwall and the Isles of Scilly: 73 %), andin Belgium, with a factor of 3.1 (Région de Brux-elles-Capitale/Brussels Hoofdstedelijk Gewest:235 %; Prov. Hainault: 75 %). As in 2001, halfof the 10 countries with the largest regional dif-ferences are from the older Member States, plusfour from the new Member States and Romania.Comparatively marked regional disparities in percapita GDP therefore emerge in both old and newMember States, although these did widen signifi-cantly in comparison to 2001 in the particularlydynamic new Member States of Slovakia, Hun-gary and the Czech Republic, whereas the situa-tion was more stable in the five EU-15 MemberStates in this group.

Moderate regional disparities in per capita GDP(i.e. factors between the highest and the lowestvalue of less than 2) are, however, almost exclu-sively found in the older Member States. This isparticularly true of Ireland (Southern and Eastern:148 %; Border, Midlands and Western: 92 %)and Sweden (Stockholm: 158 %; Norra Mellan-sverige: 98 %). Bulgaria (Yugozapaden: 42 %;Yuzhen Tsentralen: 23 %) is the only country inthis group that is not one of the EU-15 MemberStates.

In all the new Member States and candidate coun-tries, as well as a number of the EU-15 MemberStates, a substantial share of economic activity isconcentrated in the capital regions. In 14 of the 19countries included here with NUTS 2 regions, thecapital regions are also the regions with the high-est per capita GDP. For example, Map 3.1 clearlyshows the prominent position of the regions ofBrussels, Prague, Madrid, Paris, Lisbon as well asBudapest, Bratislava, London, Sofia andBucharest. There is, moreover, a particular eco-nomic dynamism in the capital regions of the newMember States, as can be seen by the fact that the



RE

GI

ON

AL

G

RO

SS

D

OM

ES

TI

C

PR

OD

UC

T

3

population and GDP of these areas are growingfaster than the national average. By contrast,there is no longer any evidence of such a cleartrend in the EU-15 Member States.

Three-year averagefor GDP from2000–02Map 3.2 gives an overview of the average level ofper capita GDP (in PPS) for the years 2000 to

2002. Three-year averages are of particular im-portance because they are used in deciding whichregions are to receive EU funding under the Struc-tural Funds during the 2007–13 programming pe-riod. The so-called ‘less-developed regions’,whose three-year average for per capita GDP isbelow 75 % of the EU-25 level, are clearly con-centrated in the new Member States and on thesouthern periphery of the EU.

According to the data available in April 2005, 84of the 268 regions examined here lay below thedecisive threshold of 75 % of the EU-25 averagewhich is used to determine eligibility. In additionto the regions in the new Member States, Bulgar-ia and Romania which, with the exception of

BE

CZ

DK

DE

EE

EL

ES

FR

IE

IT

CY

LV

LT

LU

HU

MT

NL

AT

PL

PT

SI

FI

RO

SE

UK

SK

BG

100 150 200 300500 250 325

Bruxelles/BrusselHainaut

PrahaStředni Morava

HamburgDessau

Sterea ElladaDytiki Ellada

Comunidad de MadridExtremadura

Île-de-FranceGuyane

Southern and EasternBorder, Midland and Western

BolzanoCalabria

Közép-MagyarországÉszak-Magyarország

UtrechtFlevoland

WienBurgenland

MazowieckieLubelskie

LisboaNorte

Bratislavsky krajVychodnéSlovensko

Åland/AhvenanmaaItä-Suomi/Östra Finland

StockholmNorra Mellansverige

Inner LondonCornwall and Isles of Scilly

YugozapadenYuzhen

Tsentralen

BucurestiNord-Est

Graph 3.1 — GDP per inhabitant (in PPS) 2002, NUTS 2 level,in % of EU-25 average (EU-25=100)

Average of all areas of the country.

Capital city area of the country.


Prague in the Czech Republic, Bratislavský kraj(Slovakia), Közép-Magyarország (Hungary) andCyprus, all fall below the 75 % threshold, theseregions are mainly found in eastern Germany,Greece, southern Italy, southern Spain, Portugaland, to a lesser extent, in the west of the UnitedKingdom.

There is a second group of regions which is of par-ticular importance in the context of EU structuralpolicy, because their per capita GDP is less than75 % of the EU-15 average, but over 75 % ofthe EU-25 average. These regions are called ‘stat-istical effect regions’ in regional policy discussionsand are therefore shown as a separate category in

GDP per inhabitant (in PPS), in% of EU-25 = 100

(average 2000–02)NUTS 2

> 125

82.2–125

75–82.2

≤ 75


Map 3.2


RE

GI

ON

AL

G

RO

SS

D

OM

ES

TI

C

PR

OD

UC

T

3Map 3.2 (75-82.2 % of the EU-25 average).There are 16 of these in total, most of which arein Spain, Greece and Germany.

Map 3.2 shows not just the economically weakerregions but also the particularly prosperous partsof the EU whose three-year average per capitaGDP (in PPS) is over 125 % of the EU-25 aver-age. These cover 39 NUTS 2 regions distributedacross a whole range of Member States. Contraryto a widely held belief, these particularly prosper-ous regions are by no means all to be found at theheart of the EU, as can be seen from examplessuch as Etelä-Suomi (Finland), north-easternScotland (United Kingdom) or Southern and East-ern (Ireland). It is, however, true that this groupcontains many capital cities, this being the case inparticular for London, Dublin, Brussels, Paris,Stockholm, Helsinki, Prague and Rome.

Dynamicdevelopment inperipheral regionsMap 3.3 shows how much per capita GDPchanged between 1999 and 2002 by comparisonwith the EU-25 average (expressed in percentagepoints of the EU-25 average). Economically dy-namic regions, whose per capita GDP increasedby more than 1 percentage point when comparedwith the average, are shown in orange and red.Less dynamic regions (those with a relative fall ofmore than 1 percentage point in per capita GDPas against the EU-25 average) are shown in yel-low. Figures range from + 24.5 percentage pointsfor Inner London in the United Kingdom to – 11.8percentage points for Hannover in Germany.

The map shows that economic dynamism is wellabove average in the peripheral areas of the EU, inboth the EU-15 countries and in the new MemberStates and candidate countries. Amongst the EU-15 Member States, strong growth is recorded es-pecially in Greece, Spain, southern Portugal, Ire-land, the United Kingdom and Finland. On theother hand, there seems to be confirmation of theworrying trend already revealed by previous data:persistent low growth in many key regions of theEU founding Member States, such as northernand southern Italy. There are currently only a few

Dutch, Belgian and French regions, plus Luxem-bourg, unaffected by this trend.

Of the 10 most dynamic NUTS 2 regions, thereare two each in Greece and the United Kingdom,and one each in the Czech Republic, Ireland, theNetherlands, Hungary, Slovakia and Romania.The fastest growing regions are therefore scat-tered relatively broadly across the 27 countriesexamined here. It should, however, be pointed outthat six of these are capital regions (only two ofwhich are in the EU-15 Member States). Thetrend in recent years has therefore led to a furtherconcentration of economic development in capitalcities. The trend in the Bucharest region (Roma-nia) was particularly dramatic, with per capitaGDP increasing by 17 percentage points againstthe EU-25 average between 1999 and 2002.

Above-average growth has also been recorded inmany other regions of the new Member States andcandidate countries, although it is well short ofthat seen in the capital cities. This is the case, forexample, in Centru and Nord-Vest in Romania (+4.9 and + 4.3 percentage points), Észag-Alföld inHungary (+ 4.1), Stredné Slovensko in Slovakia(+ 3.8) and Severozapaden in Bulgaria (+ 3.7).With the exception of the island States of Malta (–4.4) and Cyprus (– 1.6) which were already rela-tively wealthy, the new Member States with na-tional NUTS 2 or 3 regions also achieved above-average growth. This ranges from + 5.5percentage points in Estonia or + 4.7 in Lithuaniaand Latvia down to + 1.4 in Slovenia. In the EU-15 Member States, particularly strong economicgrowth is recorded not just by Inner London, butalso by Groningen in the Netherlands (+ 13.8percentage points), Southern and Eastern in Ire-land (+ 13.5), Voreio Aigaio in Greece (+ 13.0)and Surrey, East and West Sussex in the UnitedKingdom (+ 11.0).

A rather different picture emerges, however, whenwe look at the 30 regions across all 27 countriescombined where the relative growth rates arestrongest, as 24 of these are in the EU-15 MemberStates and just six in the new Member States andcandidate countries. The main reason for this isongoing dynamic growth in the United Kingdomand Greece, with the United Kingdom accountingfor 12 and Greece for seven of these 30 regions.

Whilst it must then be acknowledged that regionsin the new Member States and candidate coun-tries are relatively poorly represented amongst the30 most dynamic regions, the bottom end of the


ranking presents a more welcome picture: not oneof the regions in the new Member States and can-didate countries is to be found among the 30 re-gions with the lowest growth rates.

Eight of the 10 regions at the foot of this rankingare in Germany, along with one each in Austriaand Italy. The bottom 30 regions include 18 inGermany and six in Italy, providing a clear indi-

cator of the relatively poor growth rates in thesetwo founding members of the EU.

SummaryAll in all, it can be seen that the catching-upprocess is still under way in most of the new

Change of GDP per inhabitant (in PPS)in percentage points of the average EU-25

2002 as compared with 1999 — NUTS 2

EU-25 = 0> +5+1 to +5–1 to +1–5 to –1≤ –5Data not available

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundaries Cartography: Eurostat — GISCO, April 2005

Map 3.3


RE

GI

ON

AL

G

RO

SS

D

OM

ES

TI

C

PR

OD

UC

T

3Member States and candidate countries: of the114 regions that have seen clearly above-averagegrowth rates (> + 1 percentage point above theEU-25 average), 31 are to be found in these coun-tries. On the other hand, only nine of the 103 re-gions with below-average growth rates (< – 1 per-centage point) were from these countries. Takingall the new Member States together, between1999 and 2002 they rose by just under 2 percent-age points to 51.8 % of the EU-25 average.

An analysis of the individual countries shows thatthe dynamics of economic development betweenthe regions in one country can diverge just aswidely as between regions in different countries:between 1999 and 2002, per capita GDP (in PPS)in the most dynamic region of the United King-

dom (excluding London) increased by compari-son to the EU-25 average by 17.8 more percent-age points than in the weakest. For Romania, thecorresponding figure was as high as 22.1 and forthe Czech Republic 15.8 percentage points. At theopposite end of the scale lie Sweden with an inter-regional range of 2.6, Poland with 3.3 and Fin-land with 5.8 percentage points. The strong re-gional discrepancies in the new Member Statesand candidate countries are primarily caused bythe particularly dynamic growth of their capitalcities. On the other hand, there is also consider-able divergence in the distribution of growth ratesin some of the larger — but also the smaller —EU-15 Member States, but in most cases the cap-ital regions exerted no significant influence on thistrend.

H O U S E H O L D A C C O U N T S4

Introduction:Measuring wealthOne of the major aims of regional statistics is un-doubtedly to measure regions’ wealth. This is par-ticularly interesting as a basis for policy measuresto provide support for less well-off regions.

The indicator most frequently used to measure re-gions’ wealth is regional gross domestic product(GDP). GDP is usually expressed in purchasingpower standards (PPS) and per capita to make thedata comparable between regions.

However, regional per capita GDP has some un-desirable features as an indicator of wealth, one ofwhich is that a ‘place-of-work’ figure (the GDPproduced in the region) is divided by a ‘place-of-residence’ figure (the population living in the re-gion). This inconsistency is of relevance whereverthere are commuter flows — i.e. more or fewerpeople working in a region than living in it. Themost obvious example is the ‘Inner London’ re-gion of the United Kingdom, which has by far thehighest per capita GDP. Yet this by no meanstranslates into a correspondingly high income lev-el for the inhabitants of the same region, as thou-sands of commuters travel to London every day towork there but live in the neighbouring regions.Hamburg, Vienna, Luxembourg and Prague areother examples of this phenomenon.

Apart from the commuter flows, other factors canalso cause the regional distribution of actualwealth not to correspond to GDP distribution.These include, for example, income from rent, in-terest or dividends received by the residents of acertain region, but paid by residents in other re-gions. It is therefore useful to compare the region-al GDP with the regional distribution of house-hold income.

Private householdincomeIn market economies with State redistributionmechanisms, a distinction is made between twotypes of private-household income distribution.

The primary distribution of income indicates theincome of private households generated directly

from market transactions, i.e. the purchase andsale of the factors of production and goods. Theseinclude, in particular, the compensation of em-ployees. Private households can also receive prop-erty income, e.g. in the form of interest or rent. Fi-nally, there is also income in the form of anoperating surplus or self-employment income.Any interest or rent payable by the households isrecorded as a negative item. The balance of allthese transactions is termed the primary incomeof private households.

The primary income is the point of departure forthe secondary distribution of income, which de-notes the State redistribution mechanism. All socialbenefits and transfers other than in kind receivedby the households are now added to primary in-come. On the negative side, households must usetheir income to pay taxes on income and wealth,pay their social contributions and effect transfers.The sum remaining after these transactions havebeen carried out, i.e. the balance, is called the dis-posable income of private households.


The measurement unit for regionalcomparisons

When analysing household income, we first needto decide which unit of measurement to use forthe data to ensure that comparisons are mean-ingful.

For the purposes of making comparisons be-tween regions, regional GDP is generally ex-pressed in purchasing power standards (PPS) sothat volume comparisons can be made. Thesame process should therefore be applied to theprivate household income parameters, so thatthese can then be compared with regional GDPand with each other.

However, there is a problem with this. PPS aredesigned to apply to GDP as a whole. The cal-culations use the expenditure approach and PPSare subdivided only on the expenditure side.

In regional accounts, on the other hand, the ex-penditure approach cannot be used, as thiswould require data on regional import and ex-port flows. These data are not available at re-gional level, so regional accounts are only calcu-lated from the output side. This means that thereis no exact correspondence between the incomeparameters and the PPS. PPS only exist for pri-vate consumption.

Eurostat assumes that these conceptual differ-ences are of little importance and converts theincome parameters of private households bymeans of the consumer components of PPS intoPPCS (purchasing power consumption stan-dards).

Results for 2002

It is only in recent years that Eurostat has had datafor these income categories of private households.The data are collected in the regional accounts forNUTS 2 level. Derogations still apply to severalMember States, allowing their data to be submit-ted to Eurostat later than the 24 months after theend of the reference year stipulated in the regula-

tion; other Member States have not always re-spected the deadline laid down in the regulation.

There are still no data available for the followingregions: Provincia Autonoma Bolzano andProvincia Autonoma Trento in Italy, Cyprus, Lux-embourg, Malta, Slovenia and Bulgaria. Valuesfor the EU-25 in this field of the regional accountsconsequently remain unavailable. This chaptertherefore relates to the other 21 Member Statesand Romania.


HO

US

EH

OL

D

AC

CO

UN

TS

4

Primary income of privatehouseholds

per inhabitant, in PPCS2002 — NUTS 2

> 18 00012 000–18 0006 000–12 000≤ 6 000Data not available

IE, NL, PL: 2001FR, AT: 2000UK: 1999


Map 4.1

Primary income anddisposable incomeMap 4.1 gives an overview of primary income inthe NUTS 2 regions of the 22 countries examinedhere. Centres of wealth in southern England,Paris, northern Italy, Vienna, Madrid, Flanders,the western Netherlands, Stockholm and Nor-

drhein-Westfalen, Baden-Württemberg and Bay-ern are clearly evident. There is also a clear north-south divide in Italy and a west-east divide in Ger-many, while the regional distribution is relativelyhomogeneous in France.

In the new Member States, however, householdprimary income lies considerably below the EUaverage. The regions with clearly above-averagelevels of wealth are mainly capital regions,


Disposable income of privatehouseholds

per inhabitant, in PPCS2002 — NUTS 2

> 15 00010 000–15 0005 000–10 000≤ 5 000Data not available

IE, NL, PL: 2001FR, AT: 2000UK: 1999


Map 4.2

particularly Prague, Bratislava, Közép-Mag-yarország (Hungary), Mazowieckie (Poland) andBucurești (Romania). Furthermore, the easternperipheral regions of the new Member States areclearly even further behind the respective nation-al level.

The regional values range from PPCS 2 693 percapita in north-east Romania to PPCS 24 082 inthe Belgian region of Vlaams-Brabant.

A comparison of primary income with disposableincome (Map 4.2) shows the levelling influence ofState intervention. It visibly increases the relativeincome level in several Greek regions, southernItaly, central and northern Spain, the west andnorth of the United Kingdom and in parts of east-ern Germany. State activity moves several regionsin northern and western Germany up to the sameclass as the south-west of the country.

Similar effects can be observed in the new Mem-ber States, particularly in Hungary, Slovakia andcentral and western Poland. However, State inter-vention does not suffice to bring the disposable in-come in the eastern Polish provinces up to a levelthat would be comparable with that of the rest ofthe country.

In spite of State redistribution, most capital re-gions maintain their prominent position with thehighest disposable income for the country in ques-tion, both in the EU-15 and in the new MemberStates and Romania.

The regional values range from PPCS 2 826 percapita in the north-east of Romania to PPCS 18 332 in the Italian region of Emilia-Romagna.State activity lessens the difference between thehighest and the lowest value significantly from afactor of 8.9 to about 6.5.


HO

US

EH

OL

D

AC

CO

UN

TS

4

Vlaams-Brabant

Prov. LuxembourgBE

CZ

DK

DE

EE

EL

ES

FR (2)

IE (1)

IT (4)

LV

LT

HU

NL (1)

AT (2)

PL (1)

PT

FI

RO

SE

UK (3)

SK

2 5000 20 0005 000 7 500 10 000 12 500 15 000 17 500

PrahaSeverozápad

Mecklenburg-Vorpommern

Dytiki MakedoniaPeloponnissos

País VascoExtremadura

Île-de-FranceGuyane

Southern and EasternBorder, Midland and Western

Emilia-RomagnaCalabria

Közép-MagyarországEszak-Alföld

UtrechtGroningen

WienKärnten

MazowieckiePodkarpackie

LisboaNorte

Bratislavsky krajVychodné Slovensko

Åland/AhvenanmaaPohjois-Suomi/Norra Finland

StockholmÖvre Norrland

Inner LondonWest Wales and the Valleys

BucurestiNord-Est

Oberbayern

Graph 4.1 — Disposable income of private households per inhabitant(in PPCS) 2002, NUTS 2

(1) 2001(2) 2000(3) 1999(4) without Bolzano and Trento

Figure 4.1 shows the range of disposable incomeper capita and the names of the regions with thehighest and the lowest value for each country.

Map 4.3 presents the relationship between dis-posable and primary income. This quotient pro-vides an idea of the effects of State activity and ofother transfer payments. First of all, substantialdifferences between the regions of the EU-15 areevident: the disposable income in the capital cities

and other prosperous regions is almost withoutexception below 80 % of primary income. Cor-respondingly high percentages can be observed inthe less economically developed areas, above allon the southern periphery of the EU, in the westof the United Kingdom and in eastern Germany.Low percentages are obvious above all in the cap-ital regions of the new Member States and Roma-nia; the regional redistribution pattern there isthus similar to that in the EU-15.


Disposable income of privatehouseholds

as a % of primary income2002 — NUTS 2

> 10090–10080–90≤ 80Data not available

IE, NL, PL: 2001FR, AT: 2000UK: 1999

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundariesCartography: Eurostat — GISCO, May 2005

Map 4.3

In both the new Member States and the old EU-15, there are a number of regions in which dis-posable income exceeds primary income. For ex-ample, this is the case in eight of the 16 Polishprovinces and in four of the eight Romanian re-gions, and also in eight eastern German, sevenBritish and three Portuguese regions. In Greece,Italy and Hungary, several regions have values ofover 100 %. When interpreting these results,however, consideration should be given to the fact

that not only monetary social benefits from theState but also other transfer payments (e.g. trans-fers from people temporarily working elsewhere,which could play an important role in Poland,Portugal and Romania, for instance) may causedisposable income to exceed primary income.Map 4.3 clearly shows that this is frequently thecase in the less prosperous regions of the countriesin question.


HO

US

EH

OL

D

AC

CO

UN

TS

4

Disposable income of privatehouseholdsas a % of GDP

2002 — NUTS 2> 8070–8060–7050–60≤ 50Data not available

IE, NL, PL: 2001FR, AT: 2000UK: 1999


Map 4.4

Income and GDPCalculating disposable income as a proportion ofGDP gives an idea of the proportion of GDP pro-duced in a region that is actually received as in-come by the population living there.

First of all, the proportion of disposable income incountries with a traditionally high level of Stateactivity can be seen to be relatively low. It is obvi-ous here that the State collects a large amount ofthe GDP in order to distribute it later in the formof monetary transfers or benefits in kind.

In almost all countries, it can be noted that theeconomically prosperous regions must cede someof the GDP that they have generated for the bene-fit of less developed regions, both in the EU-15and in the new Member States and Romania.Map 4.4 shows this phenomenon in, for example,southern Italy, the north-east of Greece, easternGermany, the west and north of the United King-dom and also in eastern and southern Poland.

In most capital regions, disposable income as aproportion of GDP is particularly low. However,care must be taken when interpreting these resultsin more narrowly defined capital regions such asInner London, Prague or București, as cross-re-gional commuters undermine the accuracy of thisquotient, even though the results for more broad-ly defined capital regions such as Madrid, Paris,Etelä-Suomi (Finland) or Mazowieckie (Poland)point in the same direction. For large sections ofthe EU and Romania, Map 4.4 is virtually a mir-ror image of Map 3.1 (gross domestic product percapita) and thus emphasises the considerable in-fluence of the State on the actual regional distrib-ution of income.

Income and directtaxationThe State intervenes in income distribution notonly through monetary social transfers but firstand foremost by taxing income and assets. Thereare characteristic differences between the coun-tries studied here in terms of both the amount andthe regional distribution of these taxes. While inDenmark they represent around 45 % and in Fin-land and Sweden around 25 % of primary in-come, they amount to between 10 and 20 % in

most of the other EU-15 Member States. In gen-eral, the proportion of direct taxation in the newMember States and Romania is considerably low-er than in the EU-15, the lowest levels being about5 % in Romania and almost 7 % in Slovakia.

While a comparison of levels of direct taxation isnot very meaningful due to the typical differencesin the Member States’ fiscal structures, muchmore interesting results can be obtained by ob-serving the development of the rate of direct tax-ation. Map 4.5 shows the trend in direct taxationrates over a longer time period (according to dataavailability between 1995 and 2002). Regions inwhich the tax burden increased are marked in redand orange, and those in which the tax rate fell inyellow and green.

The direct tax burden on private households hasobviously developed in very different ways insome of the regions of the countries examined. Inspite of the sometimes still limited data availabil-ity, it can be concluded that more regions saw anincrease in the tax burden (137) than a decrease(111). This conclusion is confirmed when thestudy is narrowed down to the regions that un-derwent more dramatic changes. In a total of 54regions, an increase of over 1.5 percentage pointscan be observed, whereas only 46 regions experi-enced a decrease of more than – 1.5 percentagepoints.

Moreover, some countries present an entirely uni-form image, while in others the trend varies fromregion to region. For instance, the direct tax bur-den on private households clearly increased inBelgium, Greece, France, Ireland, Italy, Hungaryand Austria. In contrast, it decreased in Denmark,Spain, the Netherlands, Romania, and especiallyin Sweden and Poland. In a third group, particu-larly in Germany, Italy, Portugal, Finland and theUnited Kingdom, an increasing tax burden in eco-nomically strong regions and a decreasing taxburden in economically weak regions can be ob-served. This demonstrates that the generally pro-gressive income tax rates in these countries led toa regional redistribution from the more prosper-ous to the weaker regions.

SummaryThe regional distribution of household incomebroadly resembles the distribution of regional


GDP, but differs from it in a number of NUTS 2regions. This is mainly the result of State activityin the form of monetary social transfers and thelevying of direct taxes, which levels out the dis-parities between regions. In some cases, othertransfer payments and types of income receivedby private households from outside their regioncan also play an important role.

Taken together, State intervention and other ex-ceptional items of income bring the range of avail-able income between the most prosperous and theeconomically weakest regions to a factor of about6.5:1, whereas the two extreme values for region-al per capita GDP differ by a factor of up to 15:1.Information on the income of private householdsas an addition to that on GDP can therefore


HO

US

EH

OL

D

AC

CO

UN

TS

4Change of taxes on income

and wealthin percentage points of disposable income of private househols

NUTS 2> 1.5

0 to +1.5

–1.5 to 0

≤ –1.5

IE, NL: 2001 over 1995; PL: 2001 over 1998HU: 2002 over 2000; FR, AT: 2000 over 1995UK: 1999 over 1995; RO: 2002 over 1998


Map 4.5

provide an important input into a realistic assess-ment of the economic wealth of the regions.Therefore, once a complete record is available, the

income statistics for private households should betaken into account in the decision-making processfor regional policy, alongside statistics on GDP.


R E G I O N A L L A B O U R M A R K E T5

IntroductionRegional labour market statistics play a key rolein the measurement of the economic and socialperformance of European regions. Therefore, weonce again devote a chapter of this yearbook tothis crucial aspect of regional statistics. Eurostatprovides regional labour market information,down to NUTS 3 level, on the economically activepopulation, employment and unemployment, aswell as regional socio-demographic labour forcecharacteristics. All the data can be found on theEurostat website.

As you will see, there is no uniform pattern con-cerning regional employment and unemployment,in spite of the similarities that exist between cer-tain regions of different countries. The rich dataset available at Eurostat allows further detailedanalysis of this interesting field.

Data sourcesDown to NUTS 2 level, the source of the data pro-vided by Eurostat is the EU labour force survey(LFS) with harmonised definitions and methodsfor all EU and candidate countries. The LFS iswidely recognised as a primary tool for studyingthe labour market and provides comparable re-sults. The LFS also represents the main datasource for this chapter.

In the case of two German regions (Branden-burg-Nordost and Brandenburg-Südwest), mi-cro-census data were used, as the LFS data forthem were unavailable. Hereafter, regions men-tioned in this chapter refer to NUTS 2 level re-gions in the EU-25 or the corresponding level 2regions in the candidate countries. Even thoughthe new Member States did not join the EU until1 May 2004, their regions are referred to as EUregions in this chapter describing the regionalEU labour market in 2003. Bulgaria and Roma-nia are referred to here as candidate countries in2003.

There is a break in the time series in 2003 forFrance, Ireland and Luxembourg due to the dif-ferent reference period used for the 2002 and2003 data — first-quarter data (France) and sec-ond-quarter data (Ireland and Luxembourg) areused for 2002, and annual average data for 2003(for all three countries). Similarly, the data on Ro-

mania for 2002 and 2003 are not comparable asthe new weightings from the last census were ap-plied for the 2003 data and the 2002 data havenot yet been recalculated. France, Ireland, Lux-embourg and Romania have therefore not beentaken into consideration when referring to devel-opment between 2002 and 2003.


Definitions

Population covers persons aged 15 and over, liv-ing in private households (population living incollective households are not included). Thiscomprises all persons living in the householdssurveyed during the reference week. This defini-tion also includes persons absent from thehouseholds for short periods (but having re-tained a link with the private household) owingto studies, holidays, illness, business trips, etc.Persons on compulsory military service are notincluded.

Employed persons are all persons aged 15 andover who during the reference week worked atleast one hour for pay or profit or were tem-porarily absent from such work. Employed per-sons comprise employees, self-employed andfamily workers.

Employment rates of the age group 15–64 referto employed persons aged 15–64 as a percentageof the population of the same age group.

Self-employed persons are persons who work intheir own business, farm or professional practicefor the purpose of earning a profit.

Unemployed persons comprise persons aged15–74 who were (all three conditions must befulfilled simultaneously):

1. without work during the reference week;

2. available for work at the time (i.e. were avail-able for paid employment or self-employ-ment before the end of the two weeks fol-lowing the reference week);

3. actively seeking work (i.e. had taken specificsteps in the four-week period ending with thereference week to seek paid employment orself-employment) or who found a job to startwithin a period of, at most, three months.

Unemployment rate represents unemployed per-sons as a percentage of the economically activepopulation (i.e. employed plus unemployed).

More detailed information can be found in themetadata files on the Eurostat website.

EmploymentThe employment rate of the 15–64 age group rep-resents one of the key indicators of the Lisbonstrategy (Lisbon Summit, spring 2000). Targetsfor 2010 were set for a total employment rate of70 % for the 15–64 age group and 60 % forwomen in the same age group. The StockholmEuropean Council added in March 2001 interme-

diate objectives for 2005: 67 % for total employ-ment and 57 % for female employment.

Of the countries that comprise the EU today, onlyeight had an employment rate above 67 % forthe 15–64 age group in 2003. They were Den-mark, the Netherlands, Sweden, the United King-dom, Austria, Portugal, Finland and Cyprus. Inthe case of the first four, the figure was above 70 %. Regional information provides a more


RE

GI

ON

AL

L

AB

OU

R

MA

RK

ET

5

Employment rateof age group 15–64

Percentage2003 — NUTS 2

EU-25 = 62.9

> 7067–7060–6755–60≤ 55Data not available

Sex: TotalAlter: 15–64


Map 5.1

detailed picture of the labour market situation indifferent countries.

In only six EU-25 countries did the majority of re-gions exceed 67 % employment in 2003: Austria(seven out of the nine regions above 67 %), theNetherlands (all regions above 70 % apart fromGroningen in the north-east: 69.0 %), Sweden(three regions above 67 % and the remaining fiveabove 70 %), the United Kingdom (five regions

between 67 and 70 %, 26 above 70 % and onlysix regions below 67 %). This was also the casefor Denmark (75 %) and Cyprus (69 %), eachof which comprises a single NUTS 2 region.

In 88 regions (out of the 254 regions studied) withemployment rate above 67 %:

• The highest growth in total employment (per-sons aged 15 and over) was observed in the


Change in employmentPercentage

2003 as compared with 2002 — NUTS 2

> 2.00.5 to 2.0–0.5 to 0.5–2.0 to –0.5≤ –2.0Data not available

Sex: TotalAge: Total


Map 5.2

northern Dutch region of Flevoland (5.6 % or9 900 employed persons), the two UK regionsof Cumbria (5.7 % or 12 300 employed per-sons) and East Wales (7.7 % or 38 900 em-ployed persons) and Cyprus (3.7 % or 11 800employed persons) (Map 5.2).

• On the other hand, the strongest decline in to-tal employment was recorded in the south-east-ern UK region of Kent (4.9 % or 37 500 em-ployed persons) and the south-western Dutchregion of Zeeland (3.4 % or 6 200 employedpersons).

• Commuters (employed persons working andliving in different regions) as a percentage of allthose in employment exceeded 10 % in 37 re-gions (20 regions in the United Kingdom, sevenin the Netherlands, two in Austria, four in Ger-many and one each in France and the Czech Re-public), while the figure was below 10 % in 42regions.

• The average figure for self-employed persons asa percentage of all those in employment was 12 %.

• The vast majority of employed persons workedin services: over 70 % in most regions.

Employment rates below 60 % (for 77 EU-25 re-gions) were observed in six of 11 Belgian regions,11 of 19 Spanish regions, 11 of 13 Greek regions,11 of 21 Italian regions, eight of 26 French re-gions, all Polish regions, one Czech region(Moravskoslezsko), three of four Slovak regions,four of seven Hungarian regions and in Malta.Characteristics of these regions are as follows.

• Economic activity rate below 67 %, except forregions in Germany, Slovakia, the Czech Re-public and most regions in Greece.

• Services as a share of total employment variedsignificantly from country to country: from42.6 % in the south-eastern Polish region ofPodkarpackie to 91.2 % in the Spanish extra-continental region of Ciudad Autónoma deMelilla.

• Commuters as a percentage of all those in em-ployment exceeded 5 % in only a few cases,but the figure for Belgium was 25 %.

• Most regions showed a downward trend (up to3 %) in the number of self-employed personsas a percentage of all those in employment.

• In spite of a low employment rate there was amarkedly upward trend in total employmentbetween 2002 and 2003 in three Greek regions:

Ionia Nisia in the west (17.2 % or 10 900 em-ployed persons), Sterea Ellada in the centralmainland (14.0 % or 24 800 employed per-sons) and Notio Aigaio in the south-east (11.3% or 11 400 employed persons). The highestdecreases in total employment occurred in twoPolish regions: Opolskie (6.7 % or 23 000 em-ployed persons) in the south and Podlaskie (4.6 % or 20 400 employed persons) in thenorth-east.

In the candidate countries studied, the employ-ment rate of the 15–64 age group in Bulgaria var-ied from 46.4 % in Severozapaden (north-west)to 57.6 % in Yugozapaden (south-west), while inRomania it was between 55.0 % for Centru and61.8 % for Sud-Vest.

In Bulgaria, every region showed an upward trendin the employment rate, most notably Yugoiz-tochen in the south-east (3.1 percentage points inconjunction with a 6.0 % increase in total em-ployment — 15 500 employed persons), thesouth-central region of Yuzhen tsentralen (2.8percentage points with a 5.2 % increase in totalemployment — 34 600 employed persons includ-ing 6 400 self-employed) and Severozapaden inthe north-west (2.7 percentage points and 11.2 % — 14 200 employed persons). However,the economic activity rate, which varied from55.7 % in the north-western region of Severoza-paden to 65.0 % in the south-western region ofYugozapaden, decreased in all but one Bulgarianregion, with the biggest decline occurring inSeverozapaden (4.4 percentage points). The figurefor commuters as a percentage of all those in em-ployment was below 3 % in Bulgaria in 2003.

In Romania, the economic activity rate rangedfrom 59.4 % in Centru to 66.2 % in Sud-Vest in2003. All of the country’s regions, except for thecapital region of București, are distinctive, withvery high employment in agriculture: from25.4 % in Centru to 51.5 % in Nord-Est.

Female employmentAs can be seen in Map 5.3, the best results in 2003 with regard to the Lisbon strategy for fe-male employment (57 % in 2005 and 60 % in2010) occurred in the Netherlands (65.8 %, withevery region above 60 %), the United Kingdom(65.3 %, with only one region below 57 %, four


RE

GI

ON

AL

L

AB

OU

R

MA

RK

ET

5

between 57 and 60 % and 33 above 60 %), Swe-den (71.5 %, with regional figures ranging be-tween 67.1 and 76.1 %), Portugal (61.4 %,with three regions above 60 %, although the ex-tra-continental region of Região Autónoma dosAçores was an exception at 46.1 %), Austria(61.7 %, with every region above 57 % apartfrom Kärnten (56.6 %) in the south and five re-gions above 60 %), Finland (65.7 %, with fourout of five regions exceeding 60 %). The figures

were also good in the majority of German andFrench continental regions, as well as in Cyprus(60.2 %), Denmark (70.5 %), Latvia (57.9 %),Lithuania (58.4 %), Estonia (59.0 %) andSlovenia (57.6 %), each of which comprises asingle NUTS 2 region. On the other hand, Italy(with five regions below 30 %), Greece, Spain,Poland, Hungary and Belgium had the most re-gions where female employment rates were below50 % in 2003.


Female employment rateof age group 15–64


EU-25 = 55.0


Sex: FemalesAge: 15–64


Map 5.3

The female employment rate exceeded 57 % in131 regions in the EU-25. In most of these regionsthe economic activity rate for women aged 15–64was above 65 %, and even above 70 % in Swe-den, the United Kingdom, Denmark and Finland.With regard to the educational level of the femalepopulation aged 25–64, in all of these regionswhere the female employment rate was high —apart from the Finnish region of Åland and fiveregions in Portugal — the percentage of womenwith an intermediate level of education was morethan 40 % of all women in employment. Thebiggest growth in female employment in these re-gions was observed in the western UK regions ofEast Wales (13.1 % or 32 200 employed females)and West Wales and the Valleys (5.4 % or 19 500employed females) and the northern Dutch regionof Flevoland (8.6 % or 7 100 employed females),while the biggest decrease occurred in the centralGerman region of Kassel (5.8 % or 13 300 em-ployed females).

In regions where the female employment rate wasbelow 50 % (72 regions out of 254 studied), theeconomic activity rate for women in 2003 was re-markably low in Italy and Spain (below 50 % inmost regions) and was just over 60 % in only fiveregions (the north-eastern Czech region ofMoravskoslezsko, the north-eastern Greek regionof Anatoliki Makedonia, Thraki, the westernGreek region of Ionia Nisia, the southern Polish re-gion of Małopolskie and the eastern Slovak regionof Východné Slovensko). The percentage of womenaged 25–64 with an intermediate level of educa-tion, in relation to the total female population, wasbelow 40 % in most regions, with the lowest fig-ures occurring in Spain, Greece and Malta and oneregion in both Portugal (extra-continental region ofAçores) and France (Corse in the south-east). InSpain, the figure was below 20 % in every regionapart from Cantabria in the north (21.2 %) andthe two extra-continental regions of CiudadAutónoma de Ceuta (21.5 %) and CiudadAutónoma de Melilla (32.1 %). In Poland, the fig-ure varied from 62.7 to 73.1 %, in Hungary from58.4 to 60.2 %, and in the eastern Slovak regionof Východné Slovensko it stood at 76.8 %.

Among the regions with a low female employ-ment rate, the strongest upward trends in femaleemployment among persons aged 15–64 wererecorded in three Greek regions: Ionia Nisia in thewest (26.8 % or 6 000 employed), the central re-gion of Sterea Ellada (17.4 % or 10 000 em-ployed) and Notio Aigaio in the south-east (14.2 %or 5 000 employed). On the other hand, the

biggest declines in female employment occurred inthe northern Greek region of Dytiki Makedonia(6.9 % or 2 300 employed) and in three Polish re-gions: Dolnośląskie (4.8 % or 20 200 employedfemales) and Opolskie (5.9 % or 8 800 employedfemales), both in the south-west, and Podlaskie inthe north-east (6.4 % or 12 700 employed fe-males).

With regard to the female employment rate in thecandidate countries, the figures in Bulgaria variedfrom 43.0 % in Severoiztochen in the north-east)to 54.8 % in Yugozapaden in the south-west),and in Romania they ranged from 48.4 % in theSud-Vest region to 57.0 % in Centru. Only threeregions in the two countries had an economic ac-tivity rate above 57 %: Yugozapaden in south-west Bulgaria (61.3 %), and Sud-Vest (60.6 %)and Nord-Est (59.7 %) in Romania. In Bulgaria,over 20 % of the female population had highereducation, with the largest figure occurring in Yu-gozapaden in the south-west (34.5 %), while inRomania the rate was below 10 % except in thecapital region of București (21.7 %).

Self-employmentLast year, Eurostat introduced a more detailedbreakdown of employment by economic status toprovide regional information about self-employ-ment making it possible to measure economic dy-namics in regions. When considering the highproportion of self-employed persons among allthose in employment, which can be viewed as agood indicator for the creation of new jobs, theless dynamic performance of self-employment inagriculture due to low productivity must also beconsidered.

In 2003, the percentage of self-employed amongall those in employment in the EU-25 was 14.7 %.The figure was above 20 % in Greece, Italy,Cyprus, Poland and Portugal. In these countries,the proportion of self-employed persons workingin agriculture varied: 32 % in Greece, 10 % inItaly, 14 % in Cyprus, 55 % in Poland and 39 % in Portugal. There was a noticeable declinein self-employment in agriculture in Poland (154 000 self-employed persons, representing 1.1 % of all those in employment) and in Hun-gary (23 500 self-employed, or 0.6 % of all thosein employment). The figure for self-employmentas a percentage of total employment was below


RE

GI

ON

AL

L

AB

OU

R

MA

RK

ET

5

10 % in Denmark, Estonia, Latvia, Luxembourg,Slovenia and Slovakia.

A high proportion of self-employment (more than20 %) could be found in 20 out of 23 Italian re-gions, all 13 Greek regions, six Spanish regions,five out of seven Portuguese regions and one re-gion in the Czech Republic (the capital region ofPraha), the United Kingdom (Cornwall and Islesof Scilly) and the southern Irish region of Border,

Midland and Western (Map 5.4). Among these re-gions, the biggest increases in total employment(above 5 %) occurred in five Greek regions: thecentral region of Sterea Ellada, Notio Aigaio inthe south-east, Kriti in the south, Thessalia in thenorth and Ionia Nisia in the west, which producedthe highest figure of 17.2 %. The biggest reduc-tions were recorded in the north-western Greekregion of Anatoliki Makedonia, Thraki (3.6 %)and in two Polish regions: Lubelskie in the east


Self–employmentshare of total employment


EU-25 = 14.7




Map 5.4

(3.2 %) and Podlaskie in the north-east (4.6 %).Among the regions where the proportion of self-employed was high, there were significant num-bers of people employed in agriculture in almostevery Greek region, with the figures ranging from17.7 % in Voreio Aigaio in the east to 37.3 % inPeloponnisos in the south. Exceptions were theeastern region of Notio Aigaio (8.9 %) and thesouthern region of Attiki (1.0 %). The figureswere also significant in two Polish regions —Lubelskie in the east (37.5 %) and Kujawsko-Po-morskie in the north (17.9 %) — and in the Cen-tro region of Portugal.

In Bulgaria, the proportion of self-employed per-sons varied between 9.2 % in Severozapaden inthe north-west and 14.9 % in the south-central re-gion of Yuzhen tsentralen. In Romania, where 84 % of self-employed persons worked in agricul-ture, the figures exceeded 20 % in every regionapart from the capital region of București (5.6 %).A significant upward trend in self-employment wasobserved in Bulgaria in the south-west region ofYugozapaden, which includes the capital, Sofia.The increase was 24 400 persons, or 30.6 %.

UnemploymentThe unemployment rate for the 25 countries thatnow comprise the EU stood at 9.1 % in 2003,compared with 8.9 % in 2002. In 19 countries,the figure was below 10 %, and the highest un-employment rates were recorded in Slovakia (17.6 %) and Poland (19.6 %). At regional levelacross the continent, the unemployment rate var-ied between 2.0 % in the north-eastern Italian re-gion of Provincia Autonoma Bolzano/Bozen and26.0 % in the south-western Polish region ofDolnośląskie.

Of the 254 regions studied, 69 had an unemploy-ment rate of below 5 % (Map 5.5). Among theseregions:

• The biggest rise in unemployment rates (Map5.6) occurred in Vorarlberg in western Austria(1.5 percentage points). There was an increaseof around 1 percentage point in five Dutch re-gions (the central region of Utrecht, Noord-Holland, Noord-Brabant in the south, Fries-land in the north and Zuid-Holland), the UKregion of north-eastern Scotland and single-re-gion Luxembourg.

• In almost every region in Italy and the Nether-lands the percentage of those with an intermedi-ate level of education among the economicallyactive population was below 50 %. Around orabove 25 % of the economically active popula-tion had a high level of education in every regionin the United Kingdom, the western Belgian re-gion of West-Vlaanderen, the Czech capital re-gion of Praha, the Åland region (29.2 %) of Fin-land, Közép-Magyarország including Budapest(25.7 %) in north-central Hungary, the South-ern and Eastern region (30.8 %) of Ireland andthe single NUTS 2 region of Cyprus (32.3 %).In Italy, the figure varied from 9.1 to 13.7 %.

• The proportion of commuters (persons living andworking in different regions) among all those inemployment varied significantly, with the figureexceeding 10 % in 18 regions in the UnitedKingdom, two regions in Austria, six regions inthe Netherlands and in the western Belgian re-gion of West-Vlaanderen. Figures of below 5 %were recorded in 10 regions in Italy, five in Aus-tria, three in Portugal and two in Hungary, aswell as one region each in Finland (Åland in thesouth-west), Ireland (Southern and Eastern), theCzech Republic (Praha) and Luxembourg, whichcomprises a single NUTS 2 region.

In 2003, the unemployment rate was above 20 %in 19 regions of the EU-25. The biggest increases,of around 1 percentage point, occurred in the UKregions of North Yorkshire, West Wales and theValleys, East Wales, Cheshire and eastern Scot-land. In the regions where unemployment washigh, the proportion of persons with a high levelof education among the economically active pop-ulation varied from 13.5 to 16.5 % in Poland.The figure in two Slovak regions stood at around10 %, while in Germany it exceeded 23 %. Theproportion of commuters among all those in em-ployment was below 5 % in three southern Ital-ian regions and below 8 % in Slovakia, whereasin Germany it varied from 4.7 to 19.9 %.

Of all the countries studied, Bulgaria — whichhad the second highest unemployment rate in2002 (18.2 %) — recorded the greatest improve-ment in 2003, albeit with an increase in econom-ic inactivity. The rate fell by 4.5 percentage pointsto 13.7 %, representing a fall in unemploymentof 160 000 people and an increase in employmentof 93 800 people. The unemployment rate in Ro-mania stood at 7.0 % in 2003.

The situation in the two candidate countries alsovaried significantly at regional level. Whereas the


RE

GI

ON

AL

L

AB

OU

R

MA

RK

ET

5

unemployment rate in every Romanian region wasbelow 10 % (ranging from 5.9 % in the region ofVest to 8.6 % in the capital region of București), inBulgaria it varied between 11.1 % in the south-cen-tral region of Yuzhen tsentralen and 19.4 % in thenorth-eastern region of Severoiztochen.

In spite of higher unemployment, every region inBulgaria showed a positive trend (not taking intoaccount the increase in economic inactivity), espe-

cially in the north-western region of Severoza-paden, the central southern region of Yuzhentsentralen and the south-east region of Yugoiz-tochen.

Every region in Romania, apart from the capitalregion of București, showed very high rates of em-ployment in agriculture, with figures rangingfrom 25.4 % in the Centru region to 51.5 % inNord-Est.


Unemployment ratePercentage

2003 — NUTS 2

EU-25 = 9.1




Map 5.5

ConclusionThis chapter, describing total employment, femaleemployment, self-employment and unemploy-ment in 2003, as well as changes between 2002and 2003, is intended to show the main charac-teristics such as economic activity rates, levels of

education, breakdown of total employment andcommuting, which are all factors determining thelabour market situation in a particular region.The maps and text presented here clearly indicatethat in spite of similarities among countries thereis no uniform pattern concerning the characteris-tics in question for regions with similar levels ofemployment and unemployment.


RE

GI

ON

AL

L

AB

OU

R

MA

RK

ET

5Change in unemployment rate

Percentage points2003 as compared with 2002 — NUTS 2

> 1.50.2 to 1.5–0.2 to 0.2–1.5 to –0.2≤ –1.5Data not available



Map 5.6

T R A N S P O R T 6

IntroductionTransport needs face a steadily raising demand andit is among the main objectives of the Communityto build a truly functioning internal market and toprovide arteries to bring together people and goods.The European Union with 25 members offers astrongly diversified scenario when looking, amongother factors, at infrastructure capacities, access tomain transport links, industrialisation rates, anddisposable income. In this context, planning andmanaging transport development policies, whichcould simultaneously satisfy transport needs, envi-ronmental concern and de-congestion of the mainurban areas, is a challenging task. The main guide-lines to achieve these ambitious objectives have beenoutlined in the White paper published in 2001.Three key issues were underlined: the integration ofthe 10 new Member States, the full incorporation ofenvironmental considerations into Communitytransport policies, focusing on the shifting fromroad transport to other modes of transport such asrail, sea and inland waterways, and the central placethat should be given to the users.

The regional dimension is where the general princi-ples come into action. A substantial part of com-munity funds, taking up to one third of EU budget,has been devoted to developing regional cohesionand reducing regional unbalance. Two specific pro-grammes support the implementation of transportpolicies: the European Regional Development Fund(ERDF) and the Cohesion Funds. The latter con-centrates on completing the missing links in the pri-ority corridors, promoting rail and combined trans-port, developing multi-modal platforms andimproving traffic management.

In this context of strong focus on regional transportpolicies, Eurostat aims at providing support to deci-sion-makers, collecting a broad set of data, buildingindicators, in order to allow implementing soundactions and vigilant monitoring. Currently this col-lection comprises a set of transport indicators atNUTS 2 level on road and railway infrastructures,inland waterways, vehicle stocks and road acci-dents, together with transport flows through ports,airports and on the roads.

Methodological noteEurostat collects, compiles and disseminates vary-ing regional transport indicators. Data on road

and railway infrastructures, inland waterways,vehicle stocks and road accidents are currentlycollected in Member States and candidate coun-tries on a voluntary basis via annual question-naires, while data on maritime and air transportof passengers and goods are directly derived fromthe relative data collections established by legalacts. In addition, journeys made by vehicles werederived from a specific study on road transportdata.

Regional transport indicators are freely dissemi-nated on Eurostat’s database NewCronos underthe ‘transport’ theme and mirrored in the ‘generaland regional statistics’ one.

Data are organised in 19 tables. All indicators,apart from journeys by vehicles, are divided intables including only Member States or only can-didate countries data. Indicators for journeys byvehicles currently cover only regions for the oldMember States, prior to the 2004 enlargement.

With effect from reference year 1999 for oldMember States and 2003 for new Member States,regional data for air and maritime transport offreight and passengers are derived from the ongo-ing data collections, foreseen by the existing legis-lation. Consequently, there has been a series breakwith data prior to those reference years, as themethodology changed. Data according to thisnew methodology are disseminated in specifictables, different from the ones reporting data col-lected in the past using the regional question-naires.

All tables present annual data with the time seriesgoing back to reference year 1978 for transportinfrastructures, air and maritime transport, whilefor road safety data the series start from 1988.

All data in this chapter are presented at NUTS 2level. For Denmark, Cyprus, Estonia, Latvia,Lithuania, Luxembourg, Malta and Slovenia,NUTS 2 level corresponds to the national level.

Due to the nature of transport, a spatial referenceis built into most legal acts dealing with the col-lection of transport flow statistics, which, as men-tioned above, allow to directly derive indicatorson maritime and air transport. Moreover, otherregional transport indicators on transport flowscan be found in the specific domains of the trans-port theme: ‘road transport’, ‘railways transport’and ‘inland waterways transport’. Further infor-mation on transport flows between airports and


ports can be obtained also in the ‘maritime trans-port’ and ‘air transport’ domains.

In order to show the potential of data collected ontransport statistics as an analysis tool for regionalpatterns, this year’s contribution focuses exclu-sively on data on regional transport flows derivedfrom the ongoing maritime, air and road data col-lections based on legal acts. Data described in thefollowing maps have been extracted and aggre-gated directly from the modes’ databases and can-not be found directly from Eurostat’s databaseNewCronos. The objective is to provide an addedvalue to the data already available to the readeron Eurostat’s database NewCronos.

Maritime transportData on maritime transport are currently collect-ed according to the Council Directive 95/64/EC.Data come from national surveys on sea ports.The directive foresees to collect a broad range ofdetailed data for ports handling more than onemillion tonnes and/or more than 200 000 passen-gers per year, while for minor ports only annualaggregated figures are gathered. Consequently,data presented in the following maps may differfrom national totals, as figures for minor ports arenot included. Nevertheless, in order to properlyrepresent the regional distribution of the totalvolume of transport, the contribution of minorports can be considered negligible.

The allocation of ports to the NUTS regions ismade on the basis of the geographical coordi-nates. Data are provided to Eurostat at port leveland then aggregated at NUTS 2 level. In thisprocess the double counting, which was includedin the data previously collected via the regionalquestionnaires, is eliminated. The double count-ing concerns couples of ports that are locatedwithin the same NUTS region and have trafficamong them, in this case the concerned flow isconsidered only once in the total of the region.

The current set of disseminated regional indica-tors for maritime transport comprises embarkedand disembarked passengers and total freightloaded and unloaded, both at NUTS 2 level.

In the following, data on maritime freight trans-port and vessels (ships) are considered. Figures forMember States and available candidates countriesare derived from the data collected according to

the directive, hence comparable methodologieshave been applied. Maps consider data at NUTS2 level.

Map 6.1 presents data for total freight loaded andunloaded in each region, together with the infor-mation on predominant types of cargo in each dir-ection. This map includes data only for coastalareas with major freight ports. Map 6.2 shows thetotal number of vessels entering each region, thetotal amount of gross tonnage corresponding tothose vessels and the predominant type of vessel.The latter has been calculated on the basis of themaximum number of vessels. Due to the sym-metry in the number of vessels entering and leavinga region, only one direction has been considered.

The region of Zuid-Holland, where the port ofRotterdam is located, maintains its absolute lead-ing role as the most important region for maritimetransport, mostly due to the considerable amountof goods unloaded. The volume of goods movedin this region is almost three times the amount ofthe Antwerp region which is the next most im-portant region, followed by the regions of Ham-burg, Provence-Alpe-Côte d’Azur and Haute-Normandie. It is evident from Map 6.1, that thereis a substantial concentration of maritime trans-port, in the area covering the first two mentionedregions, which is not duplicated in any other partof Europe. This contributes to making this area acrucial spot for the whole European transport sys-tem.

Most of European regions unload more than whatthey load, showing the European Union’s relianceon importations. Nevertheless, this trend is over-turned in some regions, notably: the northernSwedish region of Övre Norrland, Highland andIslands, East Scotland and Northumberland in theUnited Kingdom, Estonia, Latvia, Lithuania andPomorskie in Poland. The different pattern ob-served in these regions may be considered as asupporting argument to the presence of naturalresources that are transferred via sea to the finaluser regions, as in the UK’s regions where thereare important volumes of North Sea oil. Quite abalance between volume of goods loaded and un-loaded is reached in the regions of Väst-sverige inSweden, Etelä-Suomi in Finland and Sicilia inItaly. In fact, the latter is characterised by the pres-ence of several refining plants producing and re-distributing oil products by transforming thecrude oil imported from non-European countries,mostly in the Middle East.


TR

AN

SP

OR

T

6


Maritime cargo loaded and unloadedby predominant type of cargo

2003 — NUTS 2

Liquid bulkDry bulkLarge containersRo-ro, mobile self-propelled unitsOther ro-ro, mobile unitsOther cargo, not elsewhere specified


Map 6.1

= loaded

= unloaded

Total freightin millions of tons

The predominance of liquid bulk as a type ofcargo all over Europe is striking. Liquid bulk in-cludes, among different products, crude oil, oilproducts and natural gases. It makes up around40 % of the total volume of goods transported inEurope. Again, the Zuid-Holland region is wherea substantial amount of liquid bulk is unloaded,contributing, with 54 %, to the total goods un-loaded in the region. Following behind, accordingto the quantity of unloaded liquid bulk, are the re-gions of Provence-Alpe-Côte d’Azur, Antwerp,Haute-Normandie, East Riding and North Lin-colnshire, Andalucía and Liguria.

‘Dry bulk cargo’ predominates in both Irish re-gions, Denmark, the south-east region of Roma-nia, Pomorskie in Poland, Puglia, where there is asubstantial amount of steel products unloaded,and Nord-Pas-de-Calais.

‘Ro-ro cargo’, which means goods moved on self-propelled units, i.e. lorries, that can roll-on androll-off the ships, is prevalent in island regions orregions connected to islands. A typical example isshown by the regions of Nord-Pas-de-Calais andKent on the opposite sides of the English Channel.There, the ferry ships couple up with the tunnelunder the channel to produce a virtual bridge forthe road transport between the United Kingdomand continental Europe. Ro-ro cargo is predomi-nant in the Greek regions of Attiki, with the portof Piraeus the connecting point for the CycladicIslands, and Dytiki Ellada, where the port of Pa-tras is facing the islands in the Ionic Sea. These re-gions concentrate the main traffic arriving fromcentral Europe via the Adriatic Italian ports andthe transport of supply provisions to the smallerislands, where tourism is the main economical ac-tivity. Ro-ro type of cargo prevails, also, in thesouth Swedish region of Sydsverige, both for un-loaded and loaded goods. Note that this type ofcargo is strongly linked to the issue of movinggoods from the roads to the sea, using the devel-oping ‘motorways of the sea’, that properly inte-grated in the European road network, should pro-vide costly and timely competitive alternatives toroad transport.

An interesting scenario is offered by the geo-graphical distribution of the regions characterisedby ‘Containers’ as a predominant type of cargo.To some extent, it is possible to recognise themain gateways of Europe, which means the entrypoints for vessels coming from outside Europe,unloading containers that hence will be redistrib-uted all over the Union, again by sea or by other

modes of transport. These are in the north, the re-gions of Hamburg, Zuid-Holland, Antwerp, Bre-men and East Anglia; in the south-west the Span-ish regions of Andalucía, with the port ofAlgeciras, Cataluña where the port of Barcelona islocated and the Comunidad Valenciana; finally, inthe south-east, the Italian regions of Calabria withthe port of Gioia Tauro and Liguria, hosting theport of Genova. Specifically, the regions of Ham-burg, Bremen and Calabria have containers pre-vailing both for loaded and unloaded freight, be-cause of their large hub ports, where almost allgoods unloaded in containers are also redistrib-uted by sea transport.

Denmark outstrips any other European region interms of the number of arriving vessels, with fig-ures one third higher than the next most impor-tant region: the Greek region of Attiki, followedby Sicilia, Campania, Sardegna, Calabria andSydsverige. These regions have in common a highdensity of ferry connections transporting bothgoods and passengers. Specifically, Greek andItalian regions host the major nodes of the net-work linking the islands to the continental part ofthe country; besides, they are major attractionsfor both national and international tourism.

A large number of vessels does not always implyproportionally high values for total gross ton-nage, as this latter depends on the type of vessel.Ferry ships contributing highly to very frequentarrivals do not have very large tonnage. Denmarkis still the region with the highest value for grosstonnage. Nevertheless, the British region of Kent,which is the next most important region for thegross tonnage with 65 % of the Danish figure,contributes to this with only 10 % of the Danishnumber of vessels. A similar argument holds forthe region of Nord-Pas-de-Calais, Zuid-Holland,Schleswig-Holstein in northern Germany, Haute-Normandie and Estonia.

When looking at Maps 6.1 and 6.2 together, it isnot straightforward to mirror the type of pre-dominant cargo with the type of predominant ves-sel, as the type of vessel ‘cargo non-specialised’has an overwhelming weight in terms of numbers.This is in fact a very comprehensive category,whose vessels are used to handle varying types ofcargo. Yet, some interesting findings can be out-lined. All but one Greek region have ‘passengerships’ as a predominant type of vessel, confirmingthe strong network supporting passengers trans-port in the very popular Greek archipelago. In theregions of Hamburg and Bremen, ‘container ship’


TR

AN

SP

OR

T

6


Number of vessels and total gross tonnagearrived by predominant type of vessel

(based on number)2003 — NUTS 2

> 400

200–400

60–200

≤ 60

Data not available

Cargo, non-specialisedPassengersMiscellaneousLiquid bulkDry bulkContainerCargo, specialisedOffshore activitiesUnknown


Map 6.2

Number of vessels in 1 000

vessels

Total gross tonnage

in millions of tons

360

180

60

10

prevails as a predominant type of vessel, in linewith the principal type of cargo. Liquid bulk shipsare the most common in the British regions ofeastern Scotland and Tees Valley and Durham,while dry bulk vessels characterise ships calling atports in Slovenia, Aquitaine and Düsseldorf.

Air transportData on air transport are currently collected ac-cording to Regulation (EC) No 437/2003 of theEuropean Parliament and the Council on statisti-cal returns in respect of the carriage of passengers,freight and mail by air. Data come from nationalsurveys on airports. The regulation foresees tocollect monthly detailed data for airports hand-ling more than 150 000 passengers per year, forairports with less than 150 000 but more than 15 000 passengers only aggregated annual dataare requested, while for minor airports there is nodata provision obligation. Consequently, datapresented in the following maps may differ fromnational totals, as figures for minor airports andfor airports reporting only aggregated data arenot included. Nevertheless, even without data forminor airports the regional distribution can beconsidered representative.

The allocation of airports to the NUTS regions ismade on the basis of the geographical coordi-nates. Data are provided to Eurostat at airportlevel and then aggregated at NUTS 2 level. In thisprocess, the double counting effect of passengerstravelling to/from airports in the same region, ifany, has been eliminated; except for Lithuania,Bulgaria, Cyprus, Slovenia and Poland (only ag-gregated data without information on the part-ners’ airports were provided).

The current disseminated set of regional indica-tors for air transport comprises embarked anddisembarked passengers and total freight and mailloaded and unloaded, both at NUTS 2 level.

In this chapter, data on air transport passengersare considered. Figures for Member States andavailable candidate countries are derived fromdata collected according to the regulation.

Map 6.3 presents data for total passengers em-barked and disembarked in 2003 in each region,together with the change rate from 2002 to 2003.This map includes data only for airports provid-ing detailed data over some thresholds and where

the total passengers in the region are above 50 000 per year. Note that data are not availableif one of the two years has not been provided, asthe change rate cannot be calculated. To properlyread the map, observe that: blue shades indicatedecreases, green shades signify positive increasesup to 10 %, and finally red shades show increas-es over 10 %; besides, the more intense thecolour the higher the number of total passengers.Map 6.4 shows the number of inhabitants divid-ed by the number of airports reporting detaileddata in each region.

The more important regions in terms of air pas-senger transport are Île-de-France, Outer London,Darmstadt, where the Frankfurt airport is locat-ed, Zuid-Holland, with the Amsterdam airport,and Cataluña with three airports close toBarcelona. It is worth noting how the regions con-taining the national capitals are not always asso-ciated with the largest number of passengers. It isquite common that financial and business centresare able to attract more passengers than adminis-trative cities. In addition to the abovementionedregions of Darmstadt and Cataluña, this stillholds true for Milano in Lombardia, where the in-tense economical activities generate a large busi-ness traffic amounting to figures even higher thanthe one observed in Lazio where Rome is located.

A remarkable feature, visible in Map 6.3, is thatregions of all new Member States except Cyprus,observed increases from 2002 to 2003. Even if ab-solute total figures are still constantly lower thanin the regions of old Member States, increases ofmore than 10 % are very frequent. This showsthe desire of these countries, already before theaccession time, to build stronger links with theEU-15, both in terms of economical relationshipsand as tourist attractions.

Following the patterns of the largest increasesfrom 2002 and 2003, it is interesting to examinethe substantial role played by low-cost carriers.Besides being active in contributing to the growthof new Member States traffic, they have influ-enced the more consolidated market of old Mem-ber States. Several regions in the United Kingdom,Spain and Italy benefited from the operations ofthese carriers. The Spanish regions of Cataluña,País Vasco, and Andalucía all saw an increase ofmore than 10 %. The airport of Girona inCataluña, linked to the tourist traffic of CostaBrava, almost tripled the number of passengersfrom 2002 to 2003. The British region of Hamp-shire and Isle of Wight, where the airport of


TR

AN

SP

OR

T

6


Total air passengers in 2003and change rate (%) from 2002 to 2003

NUTS 2

Passengersin millions

No reporting airport (only airports reporting detaileddata, except for CY, LT, SI, PL, BG)Not available or below 50 000 passengers per year

SK: NUTS 1


Map 6.3

Southampton is located, registered a 54 %growth. Similar situations are found in Veneto(IT) and in the German region of Köln. Low-costcarriers are contributing to opening new marketsand revitalising small regional airports. Some-times, this has resulted in a shifting of passengersfrom larger airports to smaller ones. A larger net-work of airports not only increases people’s ac-cessibility to travelling but also boosts localeconomies.

Map 6.4 shows the number of thousand inhabi-tants divided by the number of airports reportingdetailed data in the region. It can be considered asa rough and approximate indicator of potentialpassengers per airport. Naturally it is only an ap-proximation because, first the potentiality is con-sidered only within the borders of the region,while the catchment area for an airport will bevery likely broader or smaller than the region’sborders, secondly because all reporting airports

20

10

1

0> 0 0–10 > 10change rate (%)

are given the same importance. Nevertheless, it isstill possible to outline an interesting finding,which is the potential offered by new MemberStates. It seems as if the present infrastructuresand network are not adequate to support the in-creasing demand and that, bearing in mind thelarge increases shown in Map 6.3, low-cost car-riers are profiting from the opportunities offeredby developing markets.

Road freighttransportRoad freight transport data are collected in theframe of Council Regulation (EC) No 1172/98on statistical returns in respect of the carriage ofgoods by road, which replaced the previous direc-tives. The present regulation provides for the


TR

AN

SP

OR

T

6Number of inhabitants in thousands

for reporting* airport2003 — NUTS 2

> 2 0001 000–2 000500–1 000≤ 500No reporting* airport

(*) Only airports reporting detailed data, except forCY, LT, SI, PL, BG

SK: NUTS 1 and population 2002FR91, FR92, FR93, FR94: population 1989UKM1, UKM2, UKM3, UKM4: population 2000


Map 6.4

transmission of a larger set of variables to Euro-stat, in the form of individual data records on ve-hicles, journeys and goods transport operations.These data are collected via sample surveys ofgoods vehicles in Member States. Starting withreference year 1999, micro-data are transmittedon a quarterly basis, five months after the end ofthe reference period. Each reporting country col-lects data on the activities of road motor vehicles,registered in its country, inside and outside its na-

tional territory, thus, there is no double countingat European level. Data on transport performedby non-European hauliers in Member States’ ter-ritory are not collected. The regulation allows theexclusion from the survey of vehicles with a loadcapacity smaller than 3.5 tonnes.

One major added value brought by the Councilregulation is the description of the regional originand destination of intra-EU road transport.


Number of journeys unloadingper km 2

2003 — NUTS 2

> 1 000500–1 000250–500≤ 250Data not available

UK: NUTS 1995PL: 2004 data, Q4 estimated


Map 6.5

Presently, national transport is reported at NUTS3 level. For international transport, the regulationforesees a transitional period where origins anddestinations can be declared with country codes.The final objective is to have international trans-port also reported at NUTS 3 level.

Four tables with regional road transport data aredisseminated in Eurostat’s database NewCronos,in the road domain under the transport theme.Annual figures for national and total transporttonnes, tonne-kilometres and journeys by regionof loading and region of unloading are shown.

Map 6.5 shows the number of journeys by regionof unloading divided by the dimension of the re-gion in km2. Data for all reporting countries havebeen aggregated by region of unloading. Data arepresented at NUTS 2 level, which can be consid-ered as the best one in order to guarantee spatialdetail and maximum retaining of information forinternational transport. Because internationaltransport is not completely coded at NUTS 3 lev-el, the total in each region may be slightly under-estimated; nevertheless, the regional distributionis properly represented. Intra-regional journeysare included.

When looking at the map, readers should bear inmind that: numbers of journeys have been divid-ed by the size of the region, hence the same im-portance is given to each part of the region whichmay not always be the case, regions which aremostly transit regions may have very low figuresas only unloading journeys are considered, and fi-nally no information is given on where the jour-ney originated.

Map 6.5 shows the busiest areas in terms of en-tering journeys unloading in a region. It confirmsthe importance of the regions in the influence areaof the main ports in northern Europe, in theNetherlands, Belgium and Germany. It outlinestwo main axes of major activities across Europe:from north-east Italy, via Germany up to theNetherlands and hence over to the United King-dom, and an almost parallel one from the SlovakRepublic, via the Czech Republic to northern Ger-many.

Partly because of their very small sizes, regionscontaining capital towns emerged as being very

busy in terms of entering journeys, but on theother hand it has to be considered that urban ar-eas with a highly concentrated population aremore demanding in terms of supplies.

Regions where there is a strong concentration ofindustrial activities are ‘attracting’ many jour-neys: the region of Moravskolezsko in the CzechRepublic and Slaskie in Poland, where major steelindustries are located, Cataluña with its mechan-ics plants, and Veneto which is the beating heartof north-east Italian furniture and textiles pro-duction. The concentration in UK regions of un-loading journeys is conspicuous. The regions onthe ‘spine’ of the country linking the main con-nections in the south to continental Europe andon the centre-west to Ireland are shown as verydynamic in so far as road transport is concerned.

Road transport has a major focus in the regionaltransport policies. Bottlenecks create tremendousproblems in overcrowded areas and alternativesolutions to the predominance of road transportare hot topics. Map 6.5, partially due to absenceof information on countries crossed in transit,outlines a strong unbalance between centralEurope and more peripheral regions, and hencegives a first input on areas where actions have tobe implemented.

ConclusionData shown on the previous five maps representonly a part of a broader set of regional transportstatistics available in Eurostat’s database New-Cronos. The patterns outlined here can be deep-ened with indicators on infrastructures, vehiclesstock and traffic safety. As mentioned, transportpolicies are at the very heart of the process of re-ducing regional inequality and improving region-al cohesion. In the enlarged Europe, economicaland infrastructural disparities are more evidentthan before. One of Eurostat’s long-term objec-tives is to expand the current regional transportindicators in order to provide a better under-standing of the impact of transport policies oneconomic growth, transport needs and the en-vironment.


TR

AN

SP

OR

T

6

SCIENCE, TECHNOLOGY AND INNOVATION7

IntroductionStatistics on science, technology and innovationare related to the Lisbon European Council con-clusions of March 2000. These conclusions saythat ‘the EU should become the most competitiveand dynamic knowledge-based economy in theworld, capable of sustainable economic growthwith more and better jobs and greater social co-hesion.’

These conclusions were complemented at theBarcelona European Summit in 2002, where theEuropean Council remarked that a significantboosting of the overall spending on R & D andinnovation in the EU would be necessary inorder to close the gap between the EU and itsmajor competitors. The objective agreed by EUgovernments at Barcelona was to increase R & Dspending to 3 % of GDP by 2010, with two-thirds of this expenditure to come from the pri-vate sector.

In its contribution to the 2005 European SpringCouncil, the Commission identified ‘knowledgeand innovation for growth’ as one of the threemajor domains of the future Lisbon Action Pro-gramme, a key domain for the future of Europe.Science, technology and innovation are at the coreof these efforts the EU is invited to accomplish onthis subject.

Economic growth is increasingly related to the ca-pacity of an economy to change and to innovate.Considerable efforts should therefore be put intocreating an environment that encourages re-search, development and innovation thus facili-tating the transition to a knowledge economy.Such a policy needs statistical information on sci-ence, technology and innovation, a wide area ofstatistics that covers data on research and devel-opment, patents, high-tech manufacturing andknowledge-intensive services sectors, human re-sources in science and technology and on innova-tion.

The following chapter illustrates the dynamismof regions in providing regional indicators onresearch and development, human resources inscience and technology, high-tech patent appli-cations and employment in high-tech manufactu-ring and in knowledge-intensive services sectors.Also the main flagship indicator ‘R& D intensi-ty’ as determined by the Barcelona EuropeanSummit of 2002 is shown at regional level.

Methodological noteThe data shown in this chapter in maps or tablesis extracted from the theme ‘Science and technol-ogy’ and the sub-domains research and develop-ment, high-tech industry and knowledge-basedservices, European and US patenting systems andhuman resources in science and technology.

Statistics on research and development are col-lected by Eurostat on the basis of the CommissionRegulation (EC) No 753/2004 which determinesthe data set, breakdowns, frequency and trans-mission delays for those statistics. The methodol-ogy for R & D statistics is moreover laid down inthe so-called Frascati Manual (in its version from2002) which is applied on a worldwide level.

The data on Employment in high-tech and medi-um high-tech manufacturing and in knowledge-intensive high-tech and market services are com-piled annually on the basis of the micro-datacollected within the European labour force sur-vey. The high-technology or knowledge-intensiveeconomic sectors are — in general — defined interms of the R & D intensity, calculated as theratio of the R & D expenditure of the respectivesector to its value added.

The data on High-tech patent applications to theEPO are compiled on the basis of micro-data re-ceived from the European Patent Office. Thepatent data reported include the patent applica-tions filed at the European Patent Office (EPO)during the reference year, classified according tothe inventor’s region of residence and to the inter-national patents classification of applications.High-technology patents are compiled in accor-dance with the aggregations of certain groups ofthe international patent classifications which arerelated to high technology.

Finally Statistics on human resources in scienceand technology (HRST) also are compiled annu-ally on the basis of micro-data extracted from theEuropean labour force survey. The methodologi-cal base for these statistics is laid down in theCanberra manual where all the HRST conceptsare laid down.

For more information on methodology, consult alsothe Eurostat webpage under: http://europa.eu.int/comm/eurostat/newcronos/reference/display.do?screen=welcomeref&open=/&product=EU_sci-ence_technology_innovation&depth=2&lan-guage=en


http://europa.eu.int/comm/eurostat/newcronos/reference/display.do?screen=welcomeref&open=/&product=EU_science_technology_innovation&depth=2&language=en

Research anddevelopment

Map 7.1 presents the situation of R & D expen-diture as a percentage of GDP (R & D intensity)in the European regions in 2002. Several clusters

with high R & D intensity can be identified, main-ly in Finland, the United Kingdom, Germany andin southern and eastern France.

One of the EU’s 2010 goals set up by the LisbonSummit is to achieve an R & D intensity of 3 %( = ratio of R & D expenditure to GDP). The datashown on this map identify European regionswhich have already achieved the target ratio of


SC

IE

NC

E,

T

EC

HN

OL

OT

Y

AN

D

IN

NO

VA

TI

ON

7

R&D expenditureas a % of GDP

All sectors — 2002 — NUTS 2EU-25 = 1.9, EU-15 = 2.0 (Eurostat estimates)

> 3.01.9–3.01.0–1.9≤ 1.0Data not available

DE22, DE27, ES63: confidential dataFR9, IE, UK: NUTS 1DE, NL, PT: 2001; IT, LU: 2000; EL, UK: 1999; AT: 1998


Map 7.1


Table 7.1 — Total R & D expenditure in million EUR in the top 3 regions of each country, 2002

Regions by country Mio EUR %EU-25 186 035 s

EU-15 182 488 s

Belgium 5 814 p 100

Czech Republic 959 100Praha 331 34Strední Cechy 247 26Jihovychod 116 12

Denmark 4 634 100

Germany — 2001 52 002 100Oberbayern — 2001 6 989 13Stuttgart — 2001 6 146 12Darmstadt — 2001 3 973 8

Estonia 56 100

Greece — 1999 795 e 100Attiki — 1999 419 p 53Kentriki Makedonia — 1999 126 p 16Kriti — 1999 64 p 8

Spain 7 194 100Comunidad de Madrid 2 278 32Cataluña 1 628 23Andalucia 586 8

France 34 527 100Île-de-France 14 671 42Rhône-Alpes 3 985 12Midi-Pyrénées 2 133 6

Ireland 1 414 100

Italy — 2000 12 460 100Lombardia — 2000 2 793 22Lazio — 2000 2 309 19Piemonte — 2000 1 662 13

Cyprus 34 100

Latvia 42 100

Lithuania 100 100

Luxembourg — 2000 364 100

Hungary 706 100Közép-Magyarország 460 65Dél-Alföld 49 7Észak-Alföld 46 6

Netherlands — 2001 8 090 100Noord-Brabant — 2001 2 011 25Zuid-Holland — 2001 1 572 19Noord-Holland — 2001 1 327 16

Austria — 1998 3 377 100Wien — 1998 1 639 49Steiermark — 1998 596 18Oberösterreich — 1998 392 12

Poland 1 188 100Mazowieckie 517 44Malopolskie 129 11Slaskie 89 7

Portugal — 2001 1 038 e 100Lisboa — 2001 395 e 38Centro (PT) — 2001 338 e 33Norte — 2001 213 e 20

Slovenia 360 100

Slovakia 148 100Bratislavsky kraj 62 42Západné Slovensko 45 31Stredné Slovensko 23 16

Finland 4 830 100Etelä-Suomi/Södra Finland 2 997 62Länsi-Suomi/Västra Finland 1 006 21Pohjois-Suomi/Norra Finland 608 13

Sweden — 2001 10 459 100

United Kingdom — 1999 25 300 100South-East — 1999 6 021 24Eastern — 1999 4 595 18North-West (including Merseyside) — 1999 2 708 11

Bulgaria 81 100Yugozapaden 65 80Severoiztochen 6 7Yuzhen tsentralen 5 6

Romania 184 100Bucuresti 97 53Sud 29 16Nord-Vest 12 7

Malta: no available dataUnited Kingdom: NUTS 1s: Eurostat estimatep: provisional datae: estimated data

3 %. The German regions form strong centres forEuropean R & D, 11 of them have already achievedthis ratio, among them Braunschweig with the high-est overall R & D intensity with 7.1 %.

The remaining regions that exceeded the 3 % levelwere from Finland (three regions out of four),from the United Kingdom and France (two re-gions), from the Netherlands, Austria and theCzech Republic (one region for each of thosecountries). The Czech Republic is the only newMember State which showed one region abovethe 3 % threshold (Střední Čechy).

The EU-25’s average R & D intensity reached1.9 % in 2002. Additional to the abovementioned21 regions exceeding the 3 % threshold, therewere another 23 regions exceeding the EU averageof 1.9 %, but not yet reaching the Lisbon goal of3 %. Most of the regions came again from Ger-many (eight), four from France, three from theNetherlands and the United Kingdom and onefrom Austria, Italy, Spain and Denmark (the latterregion is also the whole country at NUTS 2 level).A number of countries show a high level of R &D intensity in the capital regions, such as Lazio inItaly and Comunidad de Madrid in Spain.

All other European regions stayed below the EUaverage. The lowest R & D intensities are identi-fied in many of the Greek, Spanish and Por-tuguese regions as well in the regions of the newMember States. Bulgaria shows the same patternas some of the EU Member States and the regionaround the capital city scores with the highest R& D intensity (1.0 % in Yugozapaden).

Table 7.1 delineates the regional R & D activitywithin a country by showing the top three regionsfor every country in millions of euro and as per-centage of national R & D expenditure.

For all sectors, very high levels of R & D concen-tration were observed in Greece, Hungary, Portu-gal, Slovakia and Bulgaria, where the top three re-gions accounted for more than 77 % of thenational R & D expenditure. This pattern is evenmore extreme in Bulgaria with 80 % of the R &D expenditure observed in the capital region Yu-gozapaden. The R & D expenditure was alsorather concentrated in one single region in Greece,Hungary, Finland and Romania with more than a50 % share in each of those countries.

The breakdown of the R & D expenditure by sec-tors of performance shows different patterns de-pending on the institutional sectors concerned. In

the government sector (GOV) the R & D expen-diture is often predominant in one region. This isoften different in the higher education sector(HES) where the institutions causing R & D ex-penditure are often less concentrated in certain re-gions.

Human resources inscience andtechnologyHuman resources in science and technology(HRST) comprise all people with either tertiaryeducation or all people employed in professionswhere such an education is normally required. Forthe age group 25–64 years the average percentageof the economically active population that wereclassified as HRST in 2003 reached 39.8 % forthe EU-25 and 41.2 % for the EU-15.

Map 7.2 shows a clear pattern among countriesthat the region where the capital city is locatedalso has the highest concentration of HRST. Thisis true for regions such as Praha (55.7 %) in theCzech Republic, Attiki (40.5 %) in Greece, Île-de-France (56.8 %), Southern and Eastern Ireland(44.0 %), Közép-Magyarország (45.4 %) in Hun-gary, Wien (44.4 %) in Austria, Lisboa (30.5 %)in Portugal, Bratislavský (53.0 %) in Slovakia,Etelä-Suomi (55.0 %) in Finland, Stockholm(62.1 %) in Sweden, Yugozapaden (47.3 %) inBulgaria and București (46.3 %) in Romania.

For a few countries, regions outside the capitalcity have a nearly equal or, in some cases, evenhigher concentration of HRST. In Belgium, threeregions are concerned: Prov. Brabant Wallon(72.7 %), Vlaams Brabant (54.7 %) and Prov.Liège (51.2 %) where more than half of the ac-tive population are classified as HRST, in addi-tion to the Région de Bruxelles-Capitale/BrusselsHoofdstedelijk Gewest (57.2 %) that containsBrussels.

This is also the case for five regions in Germanyoutside Berlin (56.0 %): Oberbayern (53.0 %),Darmstadt (52.4 %), Hamburg (51.7 %), Köln(50.4 %) and Karlsruhe (50.3 %), as well as fortwo regions in the Netherlands outside Noord-Holland (56.8 %): Utrecht (60.9 %) and Zuid-Holland (51.3 %), one region in Spain outside


SC

IE

NC

E,

T

EC

HN

OL

OT

Y

AN

D

IN

NO

VA

TI

ON

7

Comunidad de Madrid (50.9 %): Pais Vasco(53.7 %), and one region in the United Kingdomoutside Inner London (61.0 %): Berkshire, Bucksand Oxfordshire (51.0 %).

For Poland, one region, Slaskie (31.3 %), outsideMazowieckie (35.7 %) that contains Warsaw hasa higher concentration of HRST than 30 %. ForItaly, the five regions with the highest values areLazio (38.9 %) containing Rome, Liguria

(37.6 %), Umbria (35.0 %), Lombardia (33.8 %),Friuli-Venezia Giulia (33.7 %).

For some countries, the NUTS 2 level is equal tothe national level with the following shares of hu-man resources in science and technology: Den-mark (51.0 %), Estonia (45.9 %), Cyprus(43.8 %), Latvia (34.8 %), Lithuania (35.4 %),Luxembourg (39.1 %), Malta (29.3 %) andSlovenia (37.6 %).


Human resources in science and technologyas a % of active population

2003 — NUTS 2EU-25 = 39.8, EU-15 = 41.2 (Eurostat estimates)


Data for Brandenburg – Südwest (DE 42) is includedin Brandenburg – Nordost (DE 41)NL: 2002


Map 7.2

PatentsPatents reflect part of countries’ or regions’ in-ventive activity. Patents also show the country’s orregion’s capacity to exploit knowledge and trans-late it into potential economic gains. Therefore,patent statistics and indicators are widely ac-knowledged as output indicators linked to R & D

and innovation and used to assess the inventiveperformance of the country or regions.

Map 7.3 shows patent applications per million in-habitants from different regions filed at the EPOduring 2003, including both direct applicationsand applications filed under the Patents Coopera-tion Treaty (PCT) that designate the EPO. The re-gional distribution of patent applications is as-


SC

IE

NC

E,

T

EC

HN

OL

OT

Y

AN

D

IN

NO

VA

TI

ON

7

Patent applications filed to the EPOper million inhabitants

2002 — NUTS 2EU-25 = 133.6, EU-15 = 148.5 (Eurostat estimates)


CZ, HU, PL, SK, BG, RO: National level


Map 7.3

signed according to the inventor’s place of resi-dence. If one application has more than one in-ventor, the application is divided equally amongall of them and subsequently among their regions,thus avoiding double counting.

The average number of patent applications permillion inhabitants for 2002 is estimated at 133.6for the EU-25 and 148.5 for the EU-15. These es-timates are built on provisional data from EPO.

For the time being, patent statistics are only avail-able for the EU-15 at NUTS 2 level, although forsix of the 10 new Member States, Estonia, Latvia,Lithuania, Slovenia, Malta and Cyprus the na-tional level corresponds to the NUTS 2 level.

The most outstanding cluster of high productivepatenting regions is found in south-western Ger-many. This cluster also reaches into the neigh-bouring regions of eastern France, Austria and to


Employment in high-tech and mediumhigh-tech manufacturing

(as a % of total employment)2003 — NUTS 2

EU-25 = 6.60, EU-15 = 7.10 (Eurostat estimates for 2003)


The data from NL are from 2002

Statistiscal data: Eurostat database: REGIO© EuroGeographics, for the administrative boundariesCartography: Eurostat — GISCO, March 2005

Map 7.4

some extent into northern Italy. Another smallercluster can be found around the region of Noord-Brabant in the Netherlands that has the highestpatenting figure of all regions, 1 084 patents permillion inhabitants. This cluster reaches from Bel-gium in the west, Luxemburg in the south into astripe of more eastern regions.

In the United Kingdom, a band of regions withcomparable high patenting activity goes from the

region of East Anglia in the east to reach the Bris-tol Channel in the west. A relatively high figure isalso found in north-eastern Scotland but the quitelow population figures for this region makes itless important in absolute patent output.

Also in the Scandinavian Member States there is astripe of highly active regions in patenting, begin-ning with Denmark in the south-west goingthrough southern Sweden up to the Stockholm


SC

IE

NC

E,

T

EC

HN

OL

OT

Y

AN

D

IN

NO

VA

TI

ON

7

Employment in knowledge-intensivehigh-tech and market services(as a % of total employment)

2003 — NUTS 2EU-25 = 10.40, EU-15 = 11.50 (Eurostat estimates for 2003)


The data from NL are from 2002


Map 7.5

region. Southern European regions and the newMember States in general show rather low ratiosof patent applications per million inhabitants atthis stage.

Employment in high-techmanufacturing andknowledge-intensiveservicesIn 2003, the employment in high-tech and medi-um high-tech manufacturing sectors shows an av-erage rate of 6.6 % of total employment in theEU-25 or 7.1 % in the EU-15 respectively. Manyof the regions with the highest shares of employ-ment are located in Germany, Italy, France, Hun-gary or the Czech Republic.

These regions concentrate their economic activi-ties on aerospace, pharmaceuticals, computers,officer machinery, electrical machinery and elec-tronics, motor vehicles, chemicals or other relatedactivities. Amongst new Member States, it is par-ticularly the Czech Republic and Hungary whichshow high shares of employment related to high-tech and medium high-tech industries, also basedon the industrial base already existing there be-fore accession.

Under average shares of this high value added,creating employment is often found outside na-tional economic clusters, characterised by a lowerlevel of industrialisation as such or by a differentstructure of economic activities, sometimes moreoriented towards services. Examples are certainsouthern regions of Spain, Italy or also certain re-gions in Scandinavian countries.

The employment in knowledge-intensive high-tech and market services shows much less concen-tration across the EU. These services cover postand telecommunication, computer and relatedservices, R & D, water and air transport and acertain selection of business services, with an av-erage share in the total employment for the EU-25of 10.4 % and for the EU-15 of 11.5 %.

An over average share of persons are employed inthose services in many regions in southern France,northern Italy or northern Spain. Also Belgium,the Netherlands, the southern regions of theUnited Kingdom, Denmark or a number of Scan-dinavian regions are characterised by a strongerpresence of those services activities.

Compared to the regional distribution shown forthe high-tech and medium high-tech manufactur-ing industries, the knowledge-intensive high-techand market services often show a different re-gional pattern. This means that industrial clustersin certain manufacturing industries sometimes donot coincide with the regional distribution of theknowledge-intensive services which they use. Thispattern can be found in Germany, France, theNetherlands or the United Kingdom.

ConclusionStatistics on science, technology and innovationoffer already a considerable choice of regionaldata across the various domains shown. A largerchallenge in the years to come will be to producealso regional data on innovation. A first broaderattempt in this direction is made with the fourthCommunity innovation survey (based on theCommission Regulation (EC) No 1450/2004) forwhich a number of countries will compile region-al results.


S T R U C T U R A L B U S I N E S S S TAT I S T I C S8

IntroductionWhat effects are the European Union’s commer-cial and regional policies having on the industrialstructure of the regions? How is employment inindustry changing in the regions? What are thewage and investment rates in any particular re-gion or sector of activity? A detailed analysis ofthe structure of the European economy by sectorcan only be made at regional level. Regional struc-tural business statistics (SBS) can provide the datafor this kind of analysis. Regional SBS are collect-ed under Council Regulation (EC) No 58/97 con-cerning structural business statistics. Regionalbusiness statistics are compiled using informationavailable from businesses themselves. The datacover all the EU Member States, including the 10countries that joined on 1 May 2004, andBulgaria and Romania.

Data are collected on the number of local units,wages and salaries, the number of persons em-ployed and investments in tangible assets. The lat-ter variable is collected on an optional basis,which results in more limited data availabilitythan for the other variables. Data are collected atNUTS 2 level, no further detailed breakdowns areavailable. Activity is broken down at the level ofdivisions in the NACE classification.

Maps 8.1 to 8.7 are based on the structural busi-ness statistics available in the NewCronos data-base under the theme ‘Industry, trade and services— horizontal view, structural business statistics— regional statistics’. This includes three tables:regional data (according to NUTS 2003), region-al data (according to NUTS 1995) and multian-nual regional statistics. The revision of the NUTSclassification made it necessary to split the annu-al regional statistics into two tables. The samedata can alternatively be found under the theme‘General and regional statistics — regions —structural business statistics’.

Maps 8.1, 8.2 and 8.3 give a general overview ofthe economies in the regions at NUTS 2 level bycomparing the relative importance of employmentin services and trade with employment in industryand construction, by giving an overview of the av-erage wages per person employed in the total busi-ness economy and comparing the investment ratein industry. In a second part, the specialisations ofthe different regions will be investigated: firstly, byexamining the share of high and medium-hightechnology in total manufacturing (Map 8.4) and

high-technology services in total services (Map8.5). Secondly, an indicator measuring the degreeof specialisation of regions in divisions of manu-facturing (Map 8.6) and services (Map 8.7) is pre-sented in order to give an indication of where themost specialised regions in the European Unionand the candidate countries are located.

Methodological noteRegional structural business statistics are collect-ed on the basis of Council Regulation (EC)No 58/97 concerning structural business statis-tics. As the SBS regulation came into force start-ing with the reference year 1995, data are mostlyavailable from that reference year onwards. How-ever, 1995–98 was a transition period in the im-plementation of the regulation, during which thenational statistical institutes adapted to a systemcomplying with the regulation. Availability istherefore better from 1999 onwards (this is thefirst year after the end of the transitional period).Quality also improved with time: e.g. for the ref-erence year 1999, Belgian data for the first timecovered the local units of all businesses ratherthan only those of businesses with more than 20employees. Similarly, from reference year 2000onwards, German data have covered all localunits, whereas for previous years only the localunits of businesses with more than 20 employeeswere considered in the regional statistics. Forsome old Member States, data are already avail-able from reference year 1985 onwards. As thenew Member States joined the European Uniononly recently, the time series for their regions areshorter, although they do cover several years be-fore their accession.

As SBS data are collected under a regulation, andas definitions of the characteristics are included ina Commission Regulation ((EC) No 2700/98), thecontent of the statistics should be sufficiently har-monised and comparable between countries andregions. The population covered by the SBS regu-lation is the whole of the market economy, exceptfor agriculture and fishing. The population moreor less covers the secondary and tertiary marketsector, corresponding to NACE Rev.1.1 SectionsC to K. Regional data for Section J are for the timebeing only collected on a voluntary basis. As dataare not available for a sufficient number of coun-tries, Section J is not taken into account in thispublication.


The regional data collected under the SBS regula-tion are the number of local units, number of per-sons employed, wages and salaries and invest-ments in tangibles. ‘Number of persons employed’refers to the persons (paid or unpaid) working ina local unit and those working outside the unitwhile remaining part of it and being paid by it.Wages and salaries means all sums in cash andbenefits in kind paid to persons who are countedas employees, including home workers, in returnfor their labour during the accounting year, irre-spective of whether they are paid by the hour, byoutput or at piece rates, or whether they are paidregularly or not. With regard to investment in tan-gibles, account is taken of investments made dur-ing the reference period in all kinds of tangiblegoods, i.e. all those purchased from third partiesor produced for own account (i.e. capitalised pro-duction of tangible capital goods), that have auseful life of more than one year.

Data are collected by the national statistical insti-tutes from the businesses or local units and are ag-gregated by region and by NACE activity divi-sion. Eurostat processes the data — mainly bychecking the plausibility of the developments overprevious reference years — and aggregates themto subsection and section level of the NACE ac-tivity classification. SBS data are also collected atnational level. As these data concern the statisticalunit ‘enterprise’, whereas the regional statisticscover the statistical unit ‘local unit’, differencescan be noted when aggregating regional data tonational level for a certain activity and comparingthe results with the national statistics. In fact, ‘en-terprises’ and local units are classified in certainactivities according to their principal activity. Asan ‘enterprise’ can consist of several local units, itis possible for the principal activity of the localunit to differ from that of the ‘enterprise’ to whichit belongs. Hence, it is possible to note apparentinconsistencies between national and regionalstructural business statistics. It should, however,be noted that in some countries the activity codeassigned is based on the principal activity of the‘enterprise’ in question.

Services concentratedaround capitalsInner London is the region of the European Unionin which the services and trade sector is relatively

the most important employer. As financial ser-vices and non-market services are not taken intoaccount in this analysis, the services sector will inreality still be more important in this region. Ingeneral, the regions that are most active in ser-vices and trade are situated around the capitals ofthe countries, where trade and business servicesplay an important role. This is the case in all theold and new Member States and the candidatecountries, except for Germany, Bulgaria and Por-tugal. In Germany and Bulgaria, the most ser-vices-oriented regions are situated around har-bours, which seems logical as in these regions notonly are water transport services important, butthere is also considerable trade activity. In Portu-gal, the Algarve region, a well-developed touristarea, is the country’s most active region in ser-vices. Map 8.1 also shows that, in general, a rela-tively important services sector can be observed inthe tourist areas of the Mediterranean. Differ-ences in the level of importance of the services andtrade sector as employers as compared with in-dustry and construction can still be seen betweenold and new Member States. It is clear from thestatistics that in the old Member States servicesare relatively more important than in the newMember States.

Differences inregional wage levelssmallest in theNetherlandsIn Map 8.2, the wages and salaries per employeeare shown. From the map it is clear that the aver-age level of wages and salaries is still higher in theold Member States than in the new MemberStates and the candidate countries. This differenceis amplified because wages and salaries are calcu-lated in euro at an average annual nominal ex-change rate, not taking into account purchasingpower parities. If purchasing power parities weretaken into account, the gap between the oldMember States, on the one hand, and the newMember States and the candidate countries, onthe other, would certainly shrink.

When looking at the figures in more detail, it canbe noted that, in general, average wages andsalaries are highest in the regions around the


ST

RU

CT

UR

AL

B

US

IN

ES

S

ST

AT

IS

TI

CS

8

capital of the country. When comparing the aver-age wage level between the regions of a givencountry, the data for the Netherlands indicate thatthe differences are smallest in that country. InSweden, the differences between regions are alsofairly small. The largest difference between the re-gions with the highest and lowest level of wages isin the United Kingdom. The data for Spain alsoindicate that wage levels can vary considerablybetween regions in that country too.

Capital-intensiveregions in the EUMap 8.3 shows the rate of investment in industry,i.e. the physical investments in relation to em-ployment. It illustrates the increase in capital as-sociated with each person employed in industry inthe regions. Since this rate is likely to fluctuate


Relative importance of employment in services and trade

compared to industry and construction2002 — NUTS 2


Population covered: NACE Rev. 1.1 Sections C+D+E (Industry), F (Construction), G (Trade), H+I+K (Services)IE: excluding Section E; DE: 2002: Sections C, D, F, H, I, K; 2001: Sections E and GMT: 2001; excluding Section E; UK: 2000: Sections C, D, E, F, H, I, K; 2001: Section GBE, LV: 2001; CZ: 2001; NUTS 1; EL: 2000


Map 8.1

markedly from one year to the next, the capital-intensity of a given region cannot necessarily beinferred from the fact that investment may havebeen high in 2002. Investment flows would haveto be looked at over several years to enable capi-tal stock figures to be calculated. The figuresshould therefore be seen as an illustration of theavailability of regional structural business statis-tics.

The relatively low rate of investment per capitathroughout the regions of the new Member Statesis yet again accentuated by the fact that the ex-change rates used do not take into account thepurchasing power parity. In other words, it is like-ly that the cost of investment in the 10 new Mem-ber States is lower, so if it were assessed in realterms the figure would be closer to that in theother Member States. A comparison of the data is


ST

RU

CT

UR

AL

B

US

IN

ES

S

ST

AT

IS

TI

CS

8Wages and salaries per person

employed in total business economy(excluding financial services)

2002 — NUTS 2> 23 03518 612–23 03512 984–18 6128 224–12 984≤ 8 224Data not available

Population covered: NACE Rev. 1.1 Sections C to I and KIE: excluding Section E; DE: 2002: Sections C, D, F, H, I, K; 2001: Sections E and GMT: 2001; excluding Section E; UK: 2000: Sections C, D, E, F, H, I, K; 2001: Section GBE, LV: 2001; CZ: 2001; NUTS 1; EL: 2000


Map 8.2

quite difficult given that the reference year cov-ered is not the same for all countries and invest-ments depend on business trends. Bearing in mindthese limitations, the data indicate that the high-est investment rates among the regions for whichdata are available were registered in north-easternScotland (UK, especially influenced by a high in-

vestment rate in the electricity, gas and water-sup-ply industry), Cumbria (UK, manufacture of coke,refined petroleum products and nuclear fuel),Province de Luxembourg (BE, manufacture ofpulp, paper and paperboard), East Riding andNorthern Lincolnshire (UK, manufacture of othertransport equipment).


Investment rate in total industry

2002 — NUTS 2

> 11 4659 118–11 4657 571–9 1185 052–7 571≤ 5 052Data not available

Population covered: NACE Rev. 1.1 Sections C, D and EIE: excluding Section E; DE: 2002: Sections C and D; 2001: Section EMT: 2001; excluding Section EBE, LV: 2001; CZ: 2001; NUTS 1; EL, UK: 2000


Map 8.3

Most technology-intensive industriesin GermanyMap 8.4 shows the share of the high- and medi-um-high technology in total manufacturing.High-technology sectors are ‘Manufacture of

pharmaceuticals, medicinal chemicals and botan-ical products’ (NACE Rev.1.1 24.4), ‘Manufac-ture of office machinery and computers’ (NACERev.1.1 30), ‘Manufacture of radio, television andcommunication equipment and apparatus’(NACE Rev.1.1 32), ‘Manufacture of medical,precision and optical instruments, watches andclocks’ (NACE Rev.1.1 33), ‘Manufacture of air-craft and spacecraft’ (NACE Rev.1.1 35.3). Medi-um-high-technology industries are: ‘Manufacture


ST

RU

CT

UR

AL

B

US

IN

ES

S

ST

AT

IS

TI

CS

8

Share of high technology and mediumhigh-technology

in total manufacturing (% of employment)2002 — NUTS 2


Population covered: NACE Rev. 1.1 Divisions 24, 29, 30, 31, 32, 33, 34 and 35BE, LV, MT: 2001; CZ: 2001; NUTS 1; EL, UK: 2000


Map 8.4

of chemicals and chemical products’ excluding‘Manufacture of pharmaceuticals, medicinalchemicals and botanical products’ (NACERev.1.1 24 excluding 24.4), ‘Manufacture of ma-chinery and equipment’ (NACE Rev.1.1 29),‘Manufacture of electrical machinery and appara-tus n.e.c.’ (NACE Rev.1.1 31), ‘Manufacture ofmotor-vehicles, trailer and semi-trailers’ (NACERev.1.1 34), ‘Manufacture of railway andtramway locomotives and rolling stock’ (NACE

Rev.1.1 35.2), ‘Manufacture of motorcycles andbicycles’ (NACE Rev.1.1 35.4) and ‘Manufactureof other transport equipment’ (NACE Rev.1.135.5). As the regional SBS data are collected onlyat the level of NACE divisions, the figures forbuilding and repairing ships and boats’ (NACERev1.1 35.1) — which is considered to be a medi-um-low-technology industry — are also included.The figures calculated for this publication should,however, be a relatively good approximation of


Share of high technology servicesin total services

(% of employment)2002 — NUTS 2


Population covered: NACE Rev. 1.1 Divisions 64, 72 and 73BE, LV, MT: 2001; CZ: 2001; NUTS 1; EL, UK: 2000


Map 8.5

the importance of the high- and medium-high-technology manufacturing activities. The share ofthese activities in total manufacturing (Section Dof NACE Rev.1.1) has been calculated.

It can be concluded from a detailed evaluation ofthe figures that the most technology-intensive in-dustry can be found in Germany. The 10 regionswith the largest share of high- and medium-high-technology industries are in Germany: Oberbay-

ern, Stuttgart, Braunschweig, Rheinhessen-Pfalz,Darmstadt, Karslruhe, Mittelfranken, Unter-franken, Hamburg and Bremen.

Table 8.5 shows the share of high-technology ser-vices in total services. The following are consideredto be high-technology services: ‘Post and telecom-munications’ (NACE Rev.1.1 64), ‘Computer andrelated activities’ (NACE Rev.1.1 72) and ‘Re-search and development’ (NACE Rev.1.1 73).


ST

RU

CT

UR

AL

B

US

IN

ES

S

ST

AT

IS

TI

CS

8

Degree of specialisation of regionsin divisions of manufacturing

2002 — NUTS 2


Population covered: NACE Rev. 1.1 Section DBE, LV, MT: 2001; CZ: 2001; NUTS 1; EL, UK: 2000


Map 8.6

Total services are calculated as the sum of SectionsH, I and K of NACE Rev.1.1. In the EuropeanUnion, the regions with the largest share of high-technology services in total services are StrednéSlovensko (SK), Münster (DE), Mazowieckie(PL), Berkshire, Bucks and Oxfordshire (UK), Île-

de-France (FR), Stockholm (SE), Comunidad deMadrid (ES), Etelä-Suomi (FI), Sydsverige (SE)and Mellersta Norrland (SE). In Romania, a num-ber of regions also register large shares of high-technology services.


Degree of specialisation of regionsin divisions of services

2002 — NUTS 2


Population covered: NACE Rev. 1.1 Sections H, I and KBE, LV, MT: 2001; CZ: 2001; NUTS 1; EL, UK: 2000


Map 8.7

Most specialisedregions in EUThe regional SBS also allow the calculation of in-dicators on the specialisation of regions. The indi-cator shown in Maps 8.6 and 8.7 measures thedegree of variability of each region within themanufacturing and services sector respectively.This is done by calculating the average deviationof the share of each division of activity in totalemployment in manufacturing or services. Thegreater the deviation, the more specialised the re-gion is. Map 8.6 shows that regions with higherdegrees of specialisation can be found in almostall Member States of the European Union. How-ever, of the 10 most specialised regions in manu-facturing, six are in Greece: Ionia Nisia, VoreioAigaio, Notia Aigaio, Dytiki Makedonia, Ipeirosand Kriti. Two are in Spain: Ciudad Autónomo deCeuta and Ciudad Autónomo de Melilla. The lasttwo are in Portugal (Região Autónomo dosAçores and Algarve). In these regions, the fourmost important industries represent over 50 % ofemployment in manufacturing. It should be not-ed, however, that in these regions the manufactur-ing sector is relatively small. Some of these regionsare rural areas in which agriculture is still impor-tant and in which the ‘Manufacture of food prod-ucts and beverages’ is relatively important. Amore general conclusion can be drawn from thedata: most rural areas of the European Union are(logically) dependent on a limited number of man-ufacturing branches. High degrees of specialisa-tion in certain branches of manufacturing can alsobe found in regions where the services sector isrelatively important.

Map 8.7 shows how specialisation in divisions ofthe services sector is spread over the EuropeanUnion, Romania and Bulgaria, The most spe-cialised regions in certain divisions of services areCentro (PT), Cornwall and Isles of Scilly (UK),

Extremadura (ES), Alentejo (PT), Castilla-LaMancha (ES), Åland (FI), La Rioja (ES), High-lands and Islands (UK), Castilla y Léon (ES) andProvincia Autonoma Bolzano-Bozen (IT). Thefour Spanish regions are most specialised in otherbusiness activities, but hotels and restaurants arealso contributing considerably to employment.The two regions of the United Kingdom and theItalian and Portuguese regions are specialised inthe hotels and restaurants division of the servicessector. In the Finnish region of Åland, the water-transport branch is the most important employerin services.

ConclusionThe regional structural business statistics offerusers who are interested in regional sectoral dataa detailed, harmonised overview of economic ac-tivity by sector in the regions. Those looking forgreater detail can use the full database, of whichthe seven maps presented in this publication onlygive a summary.

In particular, they can compare per capita wagescosts from one region of Europe to another andtheir development, or observe the relative special-isation of the various regions in different sectorsof the economy. As time series are available,changes in the specialisation pattern in the differ-ent regions can be studied. On the basis of re-gional SBS, the regions in which a country’s flag-ship industry is concentrated can be identified.

Regional structural business statistics can also beused in combination with other statistics. For ex-ample, on the basis of regional GDP, data regionscan be identified where economic growth is lag-ging. Regional SBS can then serve to find explan-ations for this trend; it could be linked to a par-ticular activity in which the region specialises.


ST

RU

CT

UR

AL

B

US

IN

ES

S

ST

AT

IS

TI

CS

8

H E A L T H 9

IntroductionSocio-health regions are defined in very differentways from one regional, provincial or local gov-ernment to another or from one Member State toanother. As regional governments have becomemore important in Europe, the role of the regionsas units for the political and administrative man-agement of health issues has also developed. Forexample, in Spain, where regional governmentshave acquired a great deal of autonomy, one prac-tical effect is that they manage the entire healthbudget. The situation is very similar in Belgium.Since 1996, France’s healthcare reform, intro-duced to put healthcare planning on a regionalfooting, has allowed hospitals to be responsiblefor allocating the budget. Healthcare manage-ment is also being drastically reorganised in theUnited Kingdom, with NHS trusts having varyinglevels of responsibility. In other Member Statessuch as the Netherlands and Sweden, the munici-palities are responsible for healthcare.

Hence the difficulty with statistics on health andon medical/health/hospital services at regionallevel stems from the fact that local-governmentboundaries, and thus the regional breakdownwhich is of interest to health authorities in theMember States, do not always coincide with theNUTS and problems may therefore arise withcross-referencing to compare regional statistics.

Currently, two different types of health informa-tion are available at regional level, mostly forNUTS 2 level. Firstly there are data on mortalityby underlying cause, where the illnesses or dis-eases in question are defined according to an in-ternational classification and where data are col-lected using comparable methods. This chapterhighlights two of the main causes of mortality inEurope — diseases of the circulatory system andcancer — and its regional distribution. It alsolooks in detail at lung cancer for men and at neo-plasm of the ovary for women. The second type ofdata available at regional level concerns health-care resources, and this is used here to examine inparticular the regional distribution of hospital dis-charges and hospital beds.

Methodological noteCauses of death (COD) statistics are based on in-formation derived from the medical death certifi-

cate. COD statistics record the underlying causeof death, i.e. ‘the disease or injury which initiatedthe train of morbid events leading directly todeath, or the circumstances of the accident or vio-lence which produced the fatal injury’. This def-inition has been adopted by the World Health As-sembly.

In addition to absolute numbers, crude deathrates and standardised death rates are providedfor COD, at national and regional levels. Region-al level data are provided in the form of three yearaverages. The crude death rate describes mortalityin relation to the total population. Expressed per 100 000 inhabitants, it is calculated as thenumber of deaths recorded in the population for agiven period divided by the population in thesame period and then multiplied by 100 000.Crude death rates are calculated for five-year agegroups. At this level of detail, comparisons be-tween countries and regions are meaningful. Thecrude death rate for the total population (allages), however, is a weighted average of the age-specific mortality rates. The weighting factor isthe age distribution of the population whose mor-tality is being observed. Thus, the populationstructure strongly influences this indicator forbroad age classes. In a relatively ‘old’ population,there will be more deaths than in a ‘young’ one be-cause mortality is higher in higher age groups. Forcomparisons, the age effect can be taken into ac-count by using a standard population. The stan-dardised death rate (SDR) is a weighted averageof age-specific mortality rates. The weighting fac-tor is the age distribution of a standard referencepopulation. The standard reference populationused is the ‘standard European population’ as de-fined by the World Health Organisation (WHO).Standardised death rates are calculated for the agegroup 0–64 (‘premature death’) and for the totalof ages. Causes of death are classified by the 65causes of the ‘European shortlist’ of causes ofdeath. This shortlist is based on the InternationalStatistical Classification of Diseases and RelatedHealth Problems (ICD), a classification developedand maintained by the WHO.

Eurostat collects regional-level statistics onhealthcare staff (numbers of doctors and of otherprofessions; these are not shown in this publica-tion but are available in the NewCronos data-base) and numbers of hospital beds. Regionaldata on hospital discharges have recently becomeavailable, though not yet for all countries. In ad-dition to absolute numbers, density rates are pro-vided for healthcare statistics. Density rates are


used to describe the availability of these resourcesor the frequency of services rendered, expressedper 100 000 inhabitants. They are calculated bydividing the absolute number of healthcare re-sources available or services rendered in a givenperiod by the respective population in the sameperiod and then multiplied by 100 000.

Data on hospital beds should refer to availablehospital beds (occupied or unoccupied) which areimmediately available for use if required by a pa-tient on admission. Bed-counts include only bedsused for full in-patient accommodation. The con-cept of number of beds should correspond asclosely as possible to the resources actually avail-able for the specific type of in-patient care forwhich they are intended. This means fully-staffedand equipped beds excluding provisional bedsand beds for accompanying persons. However,the data reported to Eurostat on the number ofbeds usually refer to an annual average of beds inuse during the year of reporting or according toconcepts of registration or budgetary or plannedapproval. There are still some concerns with re-gards to the comparability of the data, and thedata should therefore be treated with caution dueto the different concepts of ‘hospital’ and ‘hospi-tal bed’ in the EU countries. The figures for ’totalin-patient care beds’ refer to all beds (except cotsfor healthy infants) in general, university and spe-cialised hospitals, mental hospitals, institutionsfor the psychologically impaired, nursing homesand others. Beds in hospitals available for nursingday care, medical children’s homes, nurseries forvery small children under medical supervision andinstitutions for persons with sensorial handicapsare not necessarily included.

A discharge from a hospital or another healthcarefacility occurs at any time when a patient leavesbecause of medically authorised discharge, trans-fer, departure against medical advice, or death.The number of discharges is the most commonlyused measure of the utilisation of hospital ser-vices, in preference to admissions. This is becauseit is at the time of discharge that information isgathered for hospital abstracts for in-patient care.

Mortality in EU regionsMortality patterns differ significantly accordingto age and sex, and also vary considerably be-tween regions. Many factors determine mortality

patterns — intrinsic factors such as age and sex,extrinsic factors such as biological or social col-lective factors, living or working conditions, andindividual factors such as lifestyle, smoking, alco-hol consumption, driving behaviour, and sexualbehaviour.

As a general rule, mortality is higher among menthan women in all age groups. Although there aresigns that the mortality gap is narrowing in someMember States, the difference nevertheless war-rants looking at women and men separately.

Looking at the overall mortality in the EU-25 in2001, diseases of the circulatory system accountfor 42 % of all deaths and are thus the majorcause (46 % for women and 38 % for men). Thesepathologies affect the population at advancedages — over 80 % of deaths due to cardiovascu-lar diseases occur among people aged 70 yearsand older. Malignant neoplasms, i.e. cancer, fol-low as the second most frequent cause, account-ing for 25 % of all deaths in the EU-25 (or 22 %for women and 29 % for men). Malignant neo-plasms mostly affect elderly people, as almost60 % of all deaths due to cancer involve personsaged 70 years and older, but a quarter of all suchdeaths occur at ages 45 to 64.

The following two maps look at these major causesin comparing the differences in mortality betweenwomen and men according to region.

Diseases of thecirculatory systemMale/female mortality ratios compare the differ-ences in mortality between women and men. Theyare calculated by dividing the age-standardiseddeath rate (SDR) for men in a given region and fora specific cause by the corresponding SDR forwomen (for SDR see also above in ‘methodologi-cal note’). A value higher than 1 indicates excessmale mortality, while a value lower than 1 meansexcess female mortality.

Looking at the SDRs for all ages, the male/femalemortality ratios for diseases of the circulatory sys-tem show a male excess mortality in all regionsbut the variation within the EU-25 is relativelysmall, ranging from 1.1 in the Greek region Pelo-ponnisos to 1.8 in Basse-Normandie (France).


HE

AL

TH

9

However, looking at premature mortality due todiseases of the circulatory system, i.e. SDRs forthe ages 0 to 64, then considerably higher maleexcess mortality can be found throughout Europe.The regions with the lowest male excess mortalitybefore the age of 65 already report values ofaround 2.0, and values higher than 4.0 arereached in four European regions: ComunidadForal de Navarra (Spain), Åland and Pohjois-Suo-mi (Finland), and Centru (Romania).

The regions with the highest male/female ratios arefound in France and Spain, in Finland and the Balticcountries, and also in Poland and Romania. On theother hand, in a number of countries almost all re-gions show quite moderate male excess mortality; thisis the case in the Netherlands, the United Kingdomand Sweden, and also in Bulgaria. In a third group ofcountries, to which Austria, Belgium and the CzechRepublic belong, regions with both relatively high andvery low male excess mortality can be found.


Circulatory diseasesMale/female ratio

Age standardised mortality aged 0–64(1999–2001) — NUTS 2


BE: 1994–96DK: 1997–99EL, FR, UK: 1998–2000PL: 2000–02


Map 9.1

CancerAs already mentioned above, cancer is the secondmost frequent cause of death in the EU, and ma-lignant neoplasms are more frequent for men thanfor women. Male excess mortality can be ob-served in all regions throughout Europe, rangingfrom 1.3 in Denmark, in several regions in Swe-den and the United Kingdom, and in Severoza-

paden (Bulgaria) to the maximum value of 2.7which is reached in Cantabria (Spain).

The regions with lowest male excess mortality areconcentrated in the north — Sweden and Den-mark — as well as throughout Ireland and theUnited Kingdom, while high male excess mortali-ty for cancer can be found in the west and in theeast of the European Union. A male excess mor-tality over 2.2 is reported by more than half of the


HE

AL

TH

9

Malignant neoplasm Male/female ratio

Age standardised mortality all ages(1999–2001) — NUTS 2


BE: 1994–96DK: 1997–99EL, FR, UK: 1998–2000PL: 2000–02


Map 9.2

Spanish and French regions. Similarly highmale/female ratios are also seen in Portugal(Região Autónoma dos Açores: 2.4), in Greece(Dytiki Ellada: 2.4; Ipeiros: 2.3) and in Bulgaria(Severen tsentralen and Severoiztochen: 2.2).

The Netherlands together with the western regionsof Germany and parts of Austria form a coherentarea in the middle of Europe with moderatemale/female ratios. Noteworthy is a pattern that

can be found in Germany, in the Czech Republic,in Hungary and in Romania where the capital re-gions (Berlin, Prague, Kozep-Magyarorszag whichincludes Budapest, and Bucharest) show remark-ably lower excess male mortality for cancer thando the surrounding regions.


Malignant neoplasm of larynx and trachea/bronchus/lung

Age standardised mortality in males all ages(1999–2001) — NUTS 2


BE: 1994–96DK: 1997–99EL, FR, UK: 1998–2000PL: 2000–02


Map 9.3

Men and lung cancerRespiratory cancers, i.e. malignant neoplasm oflarynx and trachea/bronchus/lung are commonlyknown as ‘smokers’ cancers’, being mainly causedby smoking. Exposure to carcinogenic dusts andsubstances such as asbestos is another cause ofrespiratory cancers.

For women, only about 11 % of all cancer-relat-ed deaths are due to malignant neoplasm of lar-ynx and trachea/bronchus/lung in the EU-25.However, for men, respiratory cancers are by farthe most frequent cancer-related cause of death,accounting for almost 30 % of all male deaths inthe EU which occur due to cancer. Almost a thirdof the men who died due to respiratory cancers in2001 were aged 45 to 64.

In the EU, the age-standardised death rate for menfor respiratory cancers is 74 (per 100 000 stan-dard population). At national level, SDRs vary be-tween 32 in Sweden and 128 in Hungary. At re-gional level, male SDRs due to malignantneoplasm of larynx and trachea/bronchus/lungrange from values below 30 in a number ofSwedish regions up to values above 150 in Hun-gary (Eszak-Alfold: 155) and Poland (Zachod-niopomorskie: 182).

Regions with a particularly low male mortalitydue to respiratory cancers can be found in theNordic countries, in the southern regions of theUnited Kingdom, in Austria and in the south ofGermany, and also in a few regions in Portugal,Italy and Greece. High mortality due to malignantneoplasm of larynx and trachea/bronchus/lung isconcentrated in the east European regions, in anarea covering the north of France, Belgium andthe Netherlands, the northern part of the UnitedKingdom and some parts of Spain, Italy andGreece.

Women and ovarycancerMalignant neoplasms related to the reproductivesystem accounted for about 28 % of all femaledeaths due to cancer in the EU-25 in 2001.Amongst the cancers of the reproductive system,breast cancer is the most frequent (17 % of allcancer-related deaths of women), followed by ma-

lignant neoplasms of the ovary (just over 5 % ofall cancer-related deaths).

The reasons for malignant neoplasms of the ovaryare still unknown. However, it is assumed that ge-netic disposition influences the likelihood of thistype of cancer. Ovulation also seems to be a fac-tor: research has indicated that cancer of theovary is more frequently found in women whohad never been pregnant or had never taken med-ical treatment to suppress ovulation.

The female standardised death rate for neoplasmof the ovary is 8.5 (per 100 000 standard popula-tion). The highest values are observed in Denmark(13.7) and Lithuania (12.7), while Portugal (5.3)and Greece (5.5) report the lowest SDRs.

The regional pattern of deaths due to this cancershows a clear north-south divide. Almost all re-gions with low female mortality due to neoplasm ofthe ovary can be found in the south — in Portugaland Spain, in Italy and Greece, and in Romania andBulgaria. Almost all regions in these countries re-port SDRs below 8, and more than half of the re-gions in these countries show SDRs below 6 (per100 000 standard population). In the middle ofEurope, a belt of regions with SDRs between 8 and10 stretches from the north-east of France toPoland. Exceptions are found in the Netherlandsand Belgium — here regions with higher and lowermortality can both be found — and the Czech Re-public where all regions except the capital regionPrague report SDRs higher than 10.

Pockets of comparatively high mortality due tocancer of the ovary are found in Ireland (Border,Midland and Western: 12.6), the United Kingdom(Devon: 12.1), Belgium (Prov. West-Vlaanderen:12.3; Prov. Liège: 12.0; Prov. Luxembourg (B):14.3; and Prov. Namur: 12.8), Finland (Åland:18.8) the Czech Republic (Jihozapad: 12.5; Stred-ni Morava: 12.0) and Poland (Lubuskie: 15.3).

Healthcare resourcesin EU regionsHospital discharges

Hospitalisation statistics give a broad picture ofthe healthcare treatment of the population, andalso of general health. Some 16 239 persons per100 000 population were discharged from


HE

AL

TH

9

hospitals in the EU-25 in 2002. However, even be-tween countries, there is a wide range for this in-dicator, from just above 6 400 in Malta to over 30000 in Austria. These differences may partly re-flect the differences in the organisation of health-care services.

Regional data for hospital discharges becameavailable only relatively recently, and not all coun-

tries are yet in the position to provide hospital dis-charges data at sub-national level. Yet, the rangeat regional level is even broader, from just about 6 000 persons discharged in Sterea Ellada to al-most 36 000 in Vienna. Within countries, it is of-ten capital regions or relatively small regions in-cluding a big city which have high discharge rates:Vienna (35 568) in Austria, Bremen (26 825), theSaarland (23 532) and Hamburg (21 220) in


Malignant neoplasm of ovary Age standardised mortality

in females all ages(1999–2001) — NUTS 2


BE: 1994–96DK: 1997–99EL, FR, UK: 1998–2000PL: 2000–02


Map 9.4

Germany, Athens (19 799) in Greece. However,this is not very surprising since hospitals tend tobe concentrated in cities and agglomerations.While the hospitals are located in the cities, theircatchment area is much wider, and people livingin the neighbouring regions may also use thehealthcare facilities offered in the cities.

Hospital beds

For many years, the number of hospital beds hasdecreased continuously in the EU. For the EU-25,it decreased by about 20 % between 1990 and2002. The reduction in the average length of stayin hospitals and the growing financial constraints

which led to the rationalisation of healthcare ser-vices could be mentioned as reasons for this de-crease. An increasing demand for healthcare of el-derly people, most often suffering from chronicdisability or illness which can be met by a transferof hospitals beds into beds in nursing and resi-dential care facilities might also provide anexplanation for the reduction of hospital beds.

Sweden, Spain, Portugal and the United Kingdom,with under 400 beds per 100 000 inhabitants,have the lowest number of hospital beds per 100 000 population in the EU-25. The highest fig-ures are reported for the Czech Republic (1 107)and Ireland (994). Accordingly, the regions with a


HE

AL

TH

9

BE (2)

CZ (3)

DK (2)

DE (1)

EE

EL (1)

ES (3)

FR (2)

IE

IT

LV

LT

HU (2)

NL

AT

RO

UK

BG

CY

LU

MT

SI

SK

FI

SE

PL

PT

BremenBaden-Württemberg

AttikiSterea Ellada

Islas BalearesLa Rioja

AbruzzoPiemonte

WienBurgenland

Vychodné SlovenskoZápadné Slovensko

Itä-Suomi/Östra Finland

Åland/Ahvenenmaa

Northern IrelandWales

YugozapadenZeveroiztochen

5 0000 40 00010 000 15 000 20 000 25 000 30 000 35 000

Graph 9.1 — Hospital discharges — Rate per 100 000 inhabitants, 2002 — NUTS 2

(1) 1999 data.

(2) 2000 data.

(3) 2001 data.

low density of hospital beds are found in thesouth (mainly in Spain and Greece, but also inItaly and Portugal), and also in the United King-dom, the Netherlands and Sweden. At the sametime, a belt of regions with more than 600 hospi-tal beds per 100 000 inhabitants stretches fromFrance via Germany, Poland, the Czech Republic,Austria, Slovakia and Hungary up to Romaniaand parts of Bulgaria.

The density of hospital beds also varies substan-tially within countries. In Finland, Hungary andBulgaria, the region with the highest density ofhospital beds exceeds the region with the lowestdensity only by a factor less than 1.5. In contrast,in France and Portugal, the density of hospitalbeds in the region with the highest value is aboutthree times higher than in the region with the low-est value. In France, the Limousin reports 1 132


Hospital bedsRate per 100 000 inhabitants

2002 — NUTS 2


DK, IE, LU: 1999BE, EE, EL, SE, UK: 2000DE, ES, IT, CY, LV, LT, HU, MT, PT, SK: 2001


Map 9.5

hospital beds per 100 000 inhabitants comparedto 372 on the Réunion island, and in Portugal 714hospital beds per 100 000 inhabitants are report-ed for Madeira compared to 235 in the Algarveregion. In the Netherlands, with 635 hospital bedsper 100 000 inhabitants in Drenthe compared toonly 164 in Flevoland, the density is almost fourtimes higher in Drenthe than in Flevoland. Thebiggest difference between the density of hospitalbeds in the region with the highest value com-pared to the region with the lowest value can beseen in Greece, where 666 hospital beds per 100 000 inhabitants are reported for Athens com-pared to no more than 155 in Sterea Ellada, i.e.the density in Athens is 4.3 times higher than inSterea Ellada. Among the possible reasons whichcould explain these regional disparities are: (1) theeffect of cities or agglomerations and their widercatchment area, i.e. the hospital services providedby cities are also used by residents of the neigh-bouring regions; (2) in regions which attract manytourists or persons in retirement there may be bet-ter facilities for healthcare; (3) in some regionswhich have experienced a considerable popula-tion decrease in recent decades (rural areas), hos-pital capacities may have been maintained be-cause of the distance to healthcare facilities inother regions.

ConclusionThe currently available regional indicators forhealth already provide a good insight into simi-larities and particularities that exist throughoutEurope. However, while analysing the data it hasto be kept in mind that the observed differencesare also influenced by the organisation of health-care systems and by socio-cultural factors. Exam-ples for the latter are the reporting of particularcauses of death such as suicide or alcohol-relateddeaths and their link to culturally determined con-sumption patterns. Healthcare resources are in-fluenced by the organisation of the systems at na-tional and regional levels, and in the medium termfigures on healthcare capacities should be com-plemented by information on their effectiveness.

The main focus of Eurostat’s work in the area ofhealth statistics lies on the further improvement ofthe quality and comparability of the data, and onthe further extension of the regional coverage.


HE

AL

TH

9

U R B A N S T A T I S T I C S10

What is the UrbanAudit?Six years ago, the Commission conducted a tenta-tive data collection of comparable indicators inEuropean cities. The purpose of this so-called‘Urban Audit’ was to test the feasibility of collect-ing comparable measurements of the quality oflife in European cities. Over the entire EU (EU-15at the time), around 480 variables were collectedfor the 58 largest cities — although London andParis were omitted since they were considered toodifficult to cope with in a test phase.

After the completion of the Urban Audit, theCommission decided that there was a clear needto continue and improve this approach of collect-ing comparable information on urban develop-ments. The results of the pilot phase were evalu-ated thoroughly, involving statistical experts fromcity organisations and Eurostat experts for a num-ber of specific fields. This evaluation led to severalconclusions concerning the list of variables col-lected, the list of participating cities, and the spa-tial dimension.

The new data collection for Urban Audit tookplace between 2003 (for EU-15 cities) and2004/05 (for the new Member States). The char-acteristics were as follows:

Variables

Some 336 variables were defined for this exercise,covering most aspects of urban life, e.g. demo-graphy, housing, health, crime, the labour market,income disparity, local administration, education-al qualifications, the environment, climate, travelpatterns, information society and cultural infra-structure. The reference year was 2001.

The Member States were asked to send all datathat were already available in the national statis-tical system and data for all variables that, whilenot currently available, could nevertheless be esti-mated with reasonable accuracy. This approachleft a third group of variables — those that wereneither available nor possible to estimate. Afterthorough reflection, it was decided that a freshsurvey for these data would be too costly. Thefinal response rates for the various variables weretherefore quite heterogeneous.

From the 336 variables, about 270 derived indi-cators were calculated by Eurostat.

Choice of cities

In the Urban Audit pilot phase, it was decided toexclude London and Paris. These two cities were,however, part of the Urban Audit 2003/05 datacollection.

In addition, there was a specific focus on medium-sized cities (50 000 to 250 000 inhabitants),which were not well covered in the pilot phase,even though a large proportion of the EU popula-tion lives in such medium-sized cities. Detailed in-formation on the various aspects of the quality oflife in these cities was considered to be valuablefor the development of European urban policy.

All in all, 237 cities of the European Union (EU-25) and 21 cities from Bulgaria and Romaniatook part in the Urban Audit 2003/05 project.Around 21 % of the 457 million EU inhabitantslive in the participating cities.

Spatial units

As in the pilot phase, there were three levels ofspatial unit for which observations were collected.The first of these is the ‘central’ or ‘core city’, i.e.the administrative unit, for which a rich data setis generally available. Secondly, the larger urbanzone (LUZ) was used in order to capture infor-mation that covers the ‘hinterland’ of the city.Finally, intra-urban discrepancies were taken intoaccount by gathering data for sub-city districts.

Time line data

Last year, Eurostat launched the collection of ‘his-toric’ data, i.e. the collection of data for 1991 and1996. Only a reduced number of 80 variables wasrequired. This data collection allows the calcula-tion of growth rates, which make it possible toanalyse evolution over time.

Perception survey

In January 2004, a parallel perception survey wasconducted in 31 cities of the old Member States.In randomised telephone interviews, citizens wereasked about their perception of various aspects ofthe quality of life within ‘their’ city.

Dissemination of results

The Urban Audit data set including all variables,indicators and methodological information isavailable in NewCronos, the publicly accessibledatabase of Eurostat. The calculated indicators arealso published on the Urban Audit website


(www.urbanaudit.org). This site allows the data tobe examined in different ways. Along with otherfeatures, the tools on the site help to compare andrank cities according to a particular indicator or toobtain an overall profile of the city. The results ofthe Urban Audit data were also published last yearin a book entitled: ‘Urban Audit 2004: key indica-tors on living conditions in European cities’.

Urban liveabilityUrban Europe seems to be enormously heteroge-neous if we look at economic structures, socialcompositions, cultural traditions or environmen-tal characteristics. Despite their diversity, how-ever, cities and towns across Europe face several


UR

BA

N

ST

AT

IS

TI

CS

10

Cinema seats per 1 000 residents(in the larger urban zone)

High

Medium

Low


Map 10.1

common challenges such as social cohesion, eco-nomic prosperity, sustainable development, etc.

The theme of urban liveability was chosen fromthese challenges to demonstrate the various as-pects of the Urban Audit data set, such as therange of spatial units applied or the different datasources used. The following sections are primarilyintended to raise awareness and trigger interestfor urban statistics and to encourage readers toconsult the information available in the New-Cronos database themselves.

CultureCultural development is crucial for the vitality ofcities. One of the characteristics describing thecultural aspect of urban liveability is the numberof cinema seats in a city. Let us look first at the ab-solute numbers — according to the terminologyof the Urban Audit database, absolute values aredefined as variables. Among the Urban Auditcities, London (UK) has the highest number of

cinema seats — more than 100 000 — followedby Paris (FR), Berlin (DE) and Rome (IT). How-ever, these data taken alone could be misleading,since London (UK) has the highest number of in-habitants as well, followed by Berlin (DE),Madrid (ES), Rome (IT) and Paris (FR). In orderto create comparable measurements, indicatorswere derived from the variables. In the case pre-sented, the number of cinema seats per 1 000 res-idents is calculated. Based on the assumption thatcinemas attract audiences beyond the borders ofthe core city where they are situated, the popula-tion of the larger urban zone was taken into ac-count for calculating the indicator. Map 10.1shows the results: the average number of cinemaseats per 1 000 residents tends to be higher inFrench, Spanish, Portuguese and Austrian cities,whereas the figure is rather low in Italy andPoland. Map 10.1 also illustrates the distributionof the cities taking part in Urban Audit, since94 % of them are displayed. (Cities where no datawere available are not marked on the map.)

The existence of cinemas or more complex cultur-al infrastructures merely provides the possibility ofimproving the attractiveness and competitiveness


Stoc

khol

mKøb

enha

vn

Hels

inki

– H

elsin

gfor

sPa

risRot

terd

amW

ienM

ünch

enAm

sterd

amLe

ipzig

Gla

sgow

Antw

erpe

nBe

rlin

Man

ches

ter

Brux

elles

– B

russ

elD

ortm

und

Torin

o

Luxe

mbo

urg

Lond

onD

ublin

Renne

sBa

rcelo

naRom

aLi

ège

Mar

seill

eAth

inai

Nap

oli

Mad

ridLi

sboa

Irakl

eioBr

aga

Mal

aga

0 %

10 %

20 %

30 %

40 %

50 %

60 %

70 %

100 %

80 %

90 %

Graph 10.1 — Perception of quality and quantity of cultural facilities

of cities. Perception indicators based on the per-ception survey results help to reflect the personalimpressions of citizens. Graph 10.1 shows the sat-isfaction index concerning cultural facilities suchas concert halls, theatres, museums and libraries.The index was calculated in two steps: first, a sim-plified index was obtained by combining the‘rather’ or ‘very satisfied’ on the one hand and the‘rather unsatisfied’ or ‘not at all satisfied’ on theother, and the difference between satisfaction anddissatisfaction was divided by the number of re-spondents. Secondly, the index was standardisedat a value between 0 and 100 by multiplying theresulting figure by 50 and then adding 50. Thehigher the index value, the greater the level of sat-isfaction in the city. Values below 50 — which donot appear on Graph 10.1 — would suggest thatmost respondents were dissatisfied. It is notewor-thy that, with the exception of Paris (FR), the fivetop-ranking cities were ‘Capitals of Culture’ in thepast 10 years: Copenhagen (DK) held the title in1996, Stockholm (SE) in 1998, Helsinki (FI) in2000 and Rotterdam (NL) in 2001.

EnvironmentThe characteristics of the urban environment areanother defining factor of urban liveability. Be-sides the indicators measured at national and re-gional levels, urban-sensitive variables were alsoidentified and collected for the environment dur-ing the Urban Audit. The variable of green spaceprovision is one of these. Let us first look at thesatisfaction index displayed in Graph 10.2, whichshows considerable differences between the citiesexamined. Whilst the citizens of Helsinki (FI),Stockholm (SE), Munich (DE) and Rennes (FR)seem to be very satisfied, the majority of the re-spondents from Napoli (IT), Athens (GR) andIrakleio (GR) are dissatisfied. Obviously, the geo-graphic location and climate of the city have a ma-jor influence on the green space provision.

The highest level of green space provision percapita can be found in Frankfurt (Oder) (DE) andDarmstadt (DE). At the bottom end of the scale —


UR

BA

N

ST

AT

IS

TI

CS

10

0 %

10 %

20 %

30 %

40 %

50 %

60 %

70 %

100 %

80 %

90 %

Hels

inki

–

Hels

ingf

ors

Stoc

khol

mM

ünch

enRen

nes

Luxe

mbo

urg

Køben

havn

Leip

zig

Brux

elles

– B

russ

elG

lasg

owW

ienPa

risD

ortm

und

Torin

oLo

ndon

Amste

rdam

Rotte

rdam

Roma

Dub

linM

anch

este

rAnt

wer

pen

Berli

nBr

aga

Mar

seill

eM

adrid

Barc

elona

Lièg

eM

alag

aLi

sboa

Nap

oli

Athin

aiIra

kleio

Graph 10.2 — Perception of green space provision


0

100

200

300

400

500

600

700

Dre

sden

Han

nove

rRot

terd

amAm

sterd

amD

üsse

ldor

fBe

rlin

Mad

rid

Brux

elles

– B

russ

elM

ilano

Catan

ia

Fran

kfur

t (O

der)

Schw

erin

Func

hal

Trier

Misk

olc

Tilb

urg

Caglia

riG

ent

Pont

o D

elgad

aAv

eiro

Irakl

eio V

olos

Wies

bade

nGraph 10.3 — Green space to which public has access per capita

Large cities Medium-sized cities

Aarhus

München

Amsterdam

Luxembourg

Frankfurt am Main

Den Haag

Berlin

Oporto

Braga

0 1 0001 10 100

Graph 10.4 — Green space to which public has access per capita — sub-citydistricts and cities

Average

similarly to the ranking established on the basis ofthe satisfaction index for the green space provi-sion — are the cities of Greece. The reasonsbehind the low values are evidently the same asthose mentioned earlier.

It is also interesting to analyse the differences be-tween large cities (over 250 000 inhabitants) andmedium-sized cities (50 000 to 250 000 inhabi-tants), as is possible with the Urban Audit results.

As expected, the average green space to which thepublic has access per capita in medium-sized citiesis significantly higher than the average for largecities. Graph 10.3 also shows the distinctive val-ues for each city that were taken into accountwhen calculating the averages indicated by thehorizontal lines.

Given the fact that data were also collected inUrban Audit for sub-city districts — albeit for a


UR

BA

N

ST

AT

IS

TI

CS

10

Length of public transport networkper capita

High

Medium

Low


Map 10.2

very limited number of variables — the spread ofthe indicator value can also be examined withinindividual cities. Such an analysis makes it possi-ble to portray a detailed picture of green spaceprovision in the cities of Europe. Graph 10.4 il-lustrates a range of sub-city districts, from thelowest to the highest levels of publicly accessiblegreen space per capita for selected cities. Thewider the range, the greater the disparities withinthe city. As we can see on the chart, the values be-hind the averages can vary considerably (averagesare indicated by the vertical line). Apparently,medium-sized cities — like Luxembourg (LU) orBraga (PT) — are characterised by a narrowerspread. However, there are also exceptions suchas the large city of Frankfurt am Main (DE),which has a rather narrow spread. It would be be-yond the scope of this publication to show thischart for all cities. Readers are invited once moreto download these data from NewCronos if theywish to conduct a comprehensive study.

Transport modes

Urban areas are the locus of multiple forms of un-desirable traffic impacts. Traffic congestions andover-reliance on cars can reduce city efficiencyand personal well-being and increase pollution.Therefore, it is worthwhile to examine the avail-ability of public transport in cities as another fea-ture of urban liveability. Map 10.2 displays a cityclassification using the length of the public trans-port network per capita as an indicator value.Copenhagen (DK) has the longest public trans-port network per capita, followed by other Scan-dinavian and Baltic cities. An extensive publictransport network is also available in Portugal,Luxembourg, Northern Ireland and in most citiesof Italy and Slovenia. Polish cities, on the otherhand, could be generally characterised by lowvalues for this indicator.


0 % 20 % 40 % 60 % 80 % 100 %

Graph 10.5 — Transport modes of the journeys to work — city and the largerurban zone

Other

Car, motorcycle Public transport

Foot, bicycle

Budapest (city)

Praha (city)

København (city)

Tallinn (city)

Barcelona (city)

Madrid (city)

Stockholm (city)

Lisboa (city)

Amsterdam (city)

Dublin (city)

Den Haag (city)

Berlin (city)

Hamburg (city)

Köln (city)

Luxembourg (city)

Graz (city)(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

(LUZ)

LUZ: Larger urban zone

In order to evaluate the usage of public transport,the modes of transport for journeys to work wereanalysed. Budapest (HU), Miskolc (HU), Ostrava(CZ) and Brno (CZ) have the highest proportionof working population using rail, metro, bus ortram for daily commuting. Cycling as the normalmeans of commuting was most characteristic ofDutch, Danish and Swedish cities. In Groningen(NL), Enschede (NL) and Umeå (SE), more than30 % of residents cycle to work. Walking is typi-cal to Spanish and Portuguese cities: more than25 % of workers go to work on foot in Logroño(ES), Oviedo (ES), Vitoria/Gasteiz (ES) and Braga(PT). Driving a car to work is most common in theUnited Kingdom. Almost 90 % of jobholderstravel by car in Wrexham (UK), Stevenage (UK),Worcester (UK) and Gravesham (UK) for in-stance.

Graph 10.5 compares the transport modes forjourneys to work in the core city and in the largerurban zone (LUZ) for selected cities. As we cansee from the chart, the proportion of journeys towork by car is consistently higher in the largerurban zone (LUZ) than in the core cities. The largestdifference between the two proportions was regis-tered in Dublin (IE), Barcelona (ES) and Copen-

hagen (DK). However, not all urban develop-ments fit this pattern: in the case of Graz (AT) andBudapest (HU), the number of car drivers isslightly higher in the core city. As expected, theproportion of workers commuting by publictransport is lower in the larger urban zone (withthe exception of Budapest): the difference rangesfrom 4 % in Graz to 22 % in Barcelona.

OutlookLast year, Eurostat decided to make the UrbanAudit data collection part of its core business. Sofar, the project has been fully financed by theDirectorate-General for Regional Policy of theEuropean Commission. In future, it will be co-fi-nanced by Eurostat and the Directorate-Generalfor Regional Policy. The next data collectionround is planned for 2006. It will also include anew perception survey, this time covering all 25EU Member States. The first results of this datacollection will be available in 2007. Preparationsare ongoing.


UR

BA

N

ST

AT

IS

TI

CS

10

E D U C A T I O N11

IntroductionEducation, vocational training and lifelong learningplay a vital role in the economic and social strategyof Europe. The Lisbon objectives can be attainedonly with efficient use of resources, quality im-provements in the education and training systemsand the implementation of a coherent lifelong learn-ing strategy at national level. To secure educationand lifelong learning opportunities in all its regionsand to all its inhabitants, wherever they live, is onecornerstone in the national strategies towardsachieving this goal. Eurostat’s regional statistics oneducation enrolment, education attainment and life-long learning participation makes it possible tomeasure regional inequalities and monitor regionslagging behind or already reaching the objectives.

Eurostat has been collecting, processing and pub-lishing data on education and lifelong learningparticipation, broken down by region, since 1991.Comparable data on education enrolment are,however, available mainly since 1998 (when theinternational education classification was revised)while data on education attainment and lifelonglearning participation are available since 1999.

The NewCronos databank now contains infor-mation on education on

— total number of enrolments by education leveland sex,

— total number of enrolments, all educationlevels, by age and sex,

— indicators on enrolments related to numbers inthe population.

Data are available for the old Member States since1998 and for the 10 new Member States and Roma-nia since 2000 or 2001 and for Bulgaria since 2002.

The NewCronos database contains informationon education attainment of the population and onlifelong learning participation since 1999. Dataare available for all Member States and for Ro-mania, Bulgaria and Norway.

Methodological noteIn the following, cartographic representation is atNUTS 2 level, except for Germany and the UnitedKingdom, in the education enrolment indicators,where data on enrolments in education are availableat NUTS 1 level only. In the Netherlands and in

Greece and Portugal, data on enrolments by age arenot available at regional level. The indicator partic-ipation rate of 17 year olds in education includes forthese countries only the national figure. No region-al enrolment data are available at all for Greece.National data are shown also in the indicators ongeneral and pre-vocational ISCED 3 enrolmentsand on tertiary education enrolment.

As the structure of education systems varies wide-ly between countries, a framework to collect andreport data on educational programmes with asimilar level of education content is a prerequisitefor intercomparability. The international classifi-cation of education, ISCED, is the basis for datacollection on education. ISCED-97, which is thecurrent ISCED, distinguishes between seven edu-cation levels, from ISCED 0, pre-primary educa-tion, to ISCED 6, second stage of tertiary educa-tion leading to an advanced researchqualification. The full description of ISCED-97 isavailable on the Unesco Institute of Statistics web-site, address: http://www.uis.unesco.org/ev.php?ID=3813_201&ID2=DO_TOPIC

The statistics on enrolments in education includeenrolments in all regular education programmesand in all adult education with subject content sim-ilar to regular education programmes or leading tosimilar qualifications as corresponding regular pro-grammes. All special education is included. Appren-ticeship programmes are included but not entirelywork-based education and training for which noformal education authority has the oversight.

Statistics on education attainment of the populationand on participation in lifelong learning are basedon the Community labour force survey (LFS),which is a quarterly sample survey. The indicatorsrefer to the LFS spring survey 2002 (education at-tainment) and 2003 (participation in lifelong learn-ing). The education attainment is reported accord-ing to ISCED-97. Participation in lifelong learningincludes all kinds of participation in education andtraining during the four weeks prior to the survey.

Participation of 17year olds ineducationAt the age of 17, most young people in the Euro-pean Union are in education, mostly in the upper


http://www.uis.unesco.org/ev.php?ID=3813_201&ID2=DO_TOPIC

secondary level. On average in the EU, 86 % ofthe age group is in education.

The age at which education starts varies betweencountries as does the ending age of secondary ed-ucation. At the age of 17, it is possible to have fin-ished secondary education in some countrieswhile in other countries, at 17 years of age, onemay have just started the upper secondary level.Compulsory education, as well as the age whencompulsory education ends, also varies betweencountries. In most countries, compulsory educa-tion ends at the age of 15 or 16, which is typical-ly the end of lower secondary education. In Bel-gium and Germany, ending age for compulsoryeducation is 18, in the Netherlands it is 17. At theage of 17, it is possible to have completed ISCEDlevel 3C programmes in several countries such asthe Czech Republic, Estonia, Spain, Hungary, theNetherlands, Austria, Slovakia, the United King-dom and Bulgaria. ISCED 3C programmes arethose which do not provide access to any tertiaryeducation and which are often of shorter durationthan three years. In Hungary and Austria, ISCED3A and 3B programmes of three years duration,giving access to tertiary education, may be com-pleted at the age of 17.

Even if compulsory education ends before uppersecondary completion, students continue educa-tion after compulsory school age in most coun-tries. To obtain at least upper secondary attain-ment is necessary for the labour market and sociallife. In 2003, the Council of Ministers set bench-marks for the improvement of education andtraining systems in Europe up to 2010. One of thebenchmarks is that by 2010, at least 85 % of 22year olds in the European Union should havecompleted upper secondary education. To followthe participation rates, also at regional level, it isimportant in order to detect if regions are at riskof lagging behind.

Education is to a great extent embedded in na-tional policy. The regulations on compulsory edu-cation and the programmes available describedabove show this fact. Even so, Map 11.1 showscertain regional variations in the participationrate of 17 year olds in education, even if the na-tional patterns are obvious.

Provincia Autonoma Bolzano-Bozen in northernItaly has the lowest participation of 17 year oldsin education at 59.4 %. Also in Malta, Greece andin several Romanian regions, the participationrate is low.

In Bratislavský in Slovakia, in several Belgium re-gions, especially Région de Bruxelles-Capitale/Brussels Hoofdstedelijk Gewest, in Saar-land and Nordrhein-Westfalen in Germany and inPraha in the Czech Republic, the participationrate exceeds 100 %, in Praha it is 154 %. Thesedark red spots on the map are surrounded by yel-low areas. This means that education facilities aregathered in the capital or big cities and that youngpeople living around commute there for their up-per secondary education. In the education statis-tics, the students are counted in the region wherethey attend school, not in the region where theyare resident and are counted in the populationstatistics.

Students in generalupper secondaryeducationMap 11.2. shows the percentage of students inISCED level 3, (upper secondary education) en-rolled in general or pre-vocational ISCED level 3programmes. The map shows, in an even moreobvious way than in Map 11.1, the differences inthe national education systems. The regional vari-ations are small in most countries. Only in Bel-gium, the Netherlands and in the United Kingdomdo the regional differences show three of the fivecolours of the map. In Belgium, 19.7 % of the stu-dents in Limburg are enrolled in general or pre-vocational streams while in Brabant Wallon thefigure is 64.4 %. In the Netherlands, 23.6 % ofthe students in Groningen are in general or pre-vocational programmes, in Flevoland the percent-age is 46.2 %. In the United Kingdom, 50.2 % ofthe students in Northern Ireland are in general-pre-vocational programmes, in the North-East re-gion the percentage is 22.1.

Even in countries where the regional differencesare relatively small, the highest percentages of stu-dents in general and pre-vocational programmesare most often found in the regions of the capital.This is the case, for example, in Germany (Berlin,47.5 %), in France (Ile-de-France, 49.9 %), inSpain (Comunidad de Madrid, 69.8 %), in Aus-tria (Wien, 36.6 %) and in Poland (Mazowieckieregion, 51.9 %). This is, however, not the case inPortugal, where the highest proportion of general


ED

UC

AT

IO

N

11

and pre-vocational enrolments is in RegiãoAutónoma da Madeira (87.5 %).

Vocational education is particularly strong in theCzech Republic, Slovakia, Austria, and in someregions of Belgium, the Netherlands and the Unit-ed Kingdom. In the Czech Republic, 75 % ormore of students in all regions are enrolled in vo-cational education at ISCED level 3.

Tertiary educationstudents

Map 11.3 shows the number of students in ter-tiary education (ISCED 5-6) as a proportion of allpupils and students in pre-primary, primary,


Participation rate of 17 year olds in education(% of students 17 years old

of the population of 17 year olds)2003 — NUTS 2

EU-25 = 86.3

> 10092.5–10085–92.575–85≤ 75Data not available

BE: Independent private institutions are not includedData refer to 2001

EL, NL, PT: National data as regional data are missing


Map 11.1

secondary and tertiary education (ISCED 0-6) inthe region.

The indicator is based on data on where the stu-dents are studying, not on where they come fromor live. Regions that have universities and othertertiary education institutes, often the big cities,tend therefore to have high percentages, as pupilsand students in lower levels of education most

often attend school close to where they live, whilestudents often travel or move for tertiary educa-tion purposes. The indicator does not in the firstplace show uneven higher education participationbut rather uneven location of higher education in-stitutions over the regions.

Other factors also have to be taken into accountwhen interpreting this indicator. Of particular


ED

UC

AT

IO

N

11Students in general upper secondary education(Number of students in ISCED 3 general and pre-vocational

education as a % of all ISCED 3 students)2003 — NUTS 2

EU-25 = 44.4%> 7560–7545–6030–45≤ 30Data not available

BE: Independent private institutions are not includedUK: ISCED 3 vocational programmes include ISCED 4Pre-vocational programmes are included in vocational


Map 11.2

relevance are demographic variations between coun-tries. In some countries, the age-groups corres-ponding to compulsory education are small rela-tive to the age-groups corresponding to typicalages for tertiary studies. In other countries, the sit-uation is the opposite. Also, the structures of theeducational systems, such as the variable durationof compulsory education and of tertiary educa-tion, affect the indicator. In spite of these limita-

tions, the indicator gives a rough picture of theproportions of tertiary education in countries andthe concentration or spreading of tertiary educa-tion institutions over regions.

On average, the percentage of tertiary educationstudents to all pupils and students is 16.2 % in theEuropean Union. The percentage is highest in theregions of București (36.5 %), Wien (33.0 %),


Tertiary education students (ISCED 5 and 6)as a % of all pupils and students in pre-primary, primary,

secondary and tertiary education (ISCED 0-6) in the region2003 — NUTS 1 and NUTS 2

EU-25 = 86.3%

> 10092.5–10085–92.575–85≤ 75Data not available

BE: Independent private institutions are not includedDE, SI: ISCED 6 is missing


Map 11.3

Bratislavský (31.2 %), Praha (31.2 %), Ma-zowieckie (30.6 %) and Közép-Magyarország(27.4 %), which are capital regions in Bulgaria,Austria, Slovakia, the Czech Republic, Polandand Hungary. These countries also have the re-gions with the lowest proportions of tertiary stu-dents. The lowest percentage is 0.2 % inSeverozapaden in Bulgaria. In Střední Čechy inthe Czech Republic the percentage is 1.5, inFlevoland in the Netherlands 1.6, in Provincia

Autonoma Trento and in Liguria in Italy 1.8 and2.2 respectively, in Niederösterreich and Vorarl-berg in Austria 2.3 in both regions. Also, in Dren-the in the Netherlands, in Reunion in France, inBurgenland in Austria, in Severozápad in theCzech Republic, in the country of Luxembourg aswell as in the Prov. Luxembourg of Belgium thepercentages are low, below 5 %. Most of these re-gions have little, if any, tertiary education infra-structure.


ED

UC

AT

IO

N

11

Education attainment% of 25–64 year olds in the population

with tertiary education2002 — NUTS 2



Map 11.4

Tertiary educationattainmentThe proportion of the population aged 25 to 64years in the regions which have attained the ter-tiary level of education is shown in Map 11.4. Thepattern on this map tends to be similar to the pat-tern in Map 11.3. In most countries, the highestproportions of tertiary attainments are found inthe same regions as the tertiary education stu-dents, that is, where the tertiary education institu-tions are located.

The regional variations are, however, in manycountries relatively small. The variation tends tobe mostly between countries. The national educa-tion systems and the national education policieshave an impact not only on enrolments and pro-vision of general and vocational education butalso on education attainment, which is a conse-quence of many years’ education policy.

The region with the highest education attainmentin Europe is Inner London. In Inner London, everysecond inhabitant between 25 and 64 years hasattained the tertiary level of education, 48.2 %. InBrabant Wallon in Belgium, the percentage is41.2. In the regions of Berkshire, Bucks and Ox-fordshire; Outer London; East and West Sussex;north eastern Scotland and eastern Scotland in the United Kingdom, in Région de Bruxelles-Cap-itale/Brussels Hoofdstedelijk Gewest in Belgium,in Etelä-Suomi in Finland, in Stockholm in Swe-den and in Île-de-France in France the percentagesare all 35 % or above.

The lowest proportion of tertiary education at-tainment in the population is in Região Autóno-ma da Madeira in Portugal, 5.1 %. Also the re-gions Região Autónoma dos Açores, Algarve,Norte and Centro in Portugal have percentagesbelow 8 %. The same applies to the regions Nord-Est, Sud and Centru in Romania, Severozápad inthe Czech Republic, Valle d’Aosta/Vallée d’Aosteand Provincia Autonoma Bolzano-Bozen in Italyand Sterea Ellada in Greece.

The regions with the lowest tertiary education at-tainment levels are also the regions with the low-est participation rates in education among 17 yearolds. These regions have also low proportions oftertiary students. While upper secondary generaleducation, within the countries, is often morecommon in regions with high proportions of ter-tiary students, this is not always the case. Região

Autónoma da Madeira, with the lowest percent-age of tertiary education attainment, has mainlygeneral education at the upper secondary level.

Lifelong learningparticipationLifelong learning refers to participation in anykind of education or training; formal, informal ornon-formal; at the workplace, in the formal edu-cation system or elsewhere during the four weekspreceding the survey. The data are collectedthrough the labour force survey but refer to all ed-ucation or vocational training whether or not rel-evant to current or future employment.

The formal education systems are most often reg-ulated at national level and affected by nationalpolicies. As Map 11.5 also shows, participation inlifelong learning is to a great extent nationallyprofiled. In fact, the regional variations are thesmallest in this indicator compared to the indica-tors on education in regions shown previously inthis chapter.

Participation in lifelong learning is high in all re-gions in Finland, Sweden, the United Kingdomand the Netherlands. The highest percentage is inÖvre Norrland in Sweden, 33.6 %. In practicallyall regions in Finland, Sweden, the United King-dom and the Netherlands, the percentages areabove 15 %, in Sweden the lifelong learning per-centages are all close to or above 30 %.

The participation rates are low in all regions inGreece, Bulgaria and Romania, lowest inSeverozapaden, Severen tsentralen and Yuzhentsentralen in Bulgaria, in Sud-Vest, Sud and Cen-tru in Romania and in Dytiki Makedonia, Pelo-ponnisos and Voreio Aigaio in Greece, below1 %.

Within countries, the highest participation ratesin lifelong learning are often found in the capitalregions. These regions are also most often thosehaving the highest education attainment levels. In the Czech Republic, Praha has the highest per-centage of lifelong learning participation, 9.8 %.In Germany, the highest percentage is in Berlin,9.9 %; in Hungary in the capital region Közép-Magyarország, 6.5 %; and in Poland in the capi-tal region Mazowieckie, 5.9 %.


This is, however, not at all always the case. The re-gion with the highest participation rate, ÖvreNorrland, is the most rural part of Sweden. InFrance, the highest participation in lifelong learn-ing is in Alsace, 8.7 %. In Italy, Sardegna has thehighest percentage, 6.1 %, in the NetherlandsUtrecht, 17.8 % and in Austria Salzburg, 10.1 %.

ConclusionThe above examples are intended merely to high-light a few of the many possible ways of analysingeducation and lifelong learning in the regions ofthe EU and does not constitute a detailed analysis.We hope, however, that they will encourage read-ers to probe deeper into the NewCronos databankand to make many further interesting discoveries.


ED

UC

AT

IO

N

11Lifelong learning participation

(% of 25–64 year olds participating in education ortraining during the four weeks preceding the survey)

2002 — NUTS 2



Map 11.5

T O U R I S M12

IntroductionTourism is an important economic activity in theEuropean Union. It encompasses a wide variety ofproducts and destinations and many differentstakeholders are involved — both public and pri-vate — with very decentralised areas of compe-tence, often at regional and local levels.

Tourism offers an ideal means of contributing to theachievement of a number of major EU objectives,such as economic growth, employment, sustainabledevelopment and economic and social cohesion.

Europe, with the greatest diversity and density oftourist attractions, is the most visited tourist re-gion in the world. European Community tourismis largely domestic. Over 80 % of the recordedtourism activity is attributed to its own citizens.

Following EU enlargement on 1 May 2004, thevarious maps now also show data for the newMember States.

Eurostat has been collecting data on tourism since1994, covering three aspects: capacity, occupancyand demand. At regional level, only data oncapacity and occupancy are collected. Capacity datarefer to the accommodation infrastructure fortourists in the region concerned. Occupancy datarefer to the number of overnight stays in rentedaccommodation in a particular region. Demanddata differentiate between domestic and out-bound tourism: outbound tourism refers to resi-dents of one country travelling to another.

Methodological notesAlthough throughout this section, mainly for rea-sons of map clarity, the regional level adopted forthe analyses is that of NUTS 2, Eurostat’s New-Cronos databank in fact contains extensive dataat NUTS 3 level.

Capacity (infrastructure)statisticsMap 12.1 provides information on the number ofbedplaces, taking account of the region’s residentpopulation.

The map highlights regions with a high accommo-dation density due to the high number of bedplacesavailable (e.g. Islas Baleares (ES), Bolzano (IT) andCorse (FR)). In contrast, other regions have a highaccommodation density due to their small popula-tions (e.g. Åland (FIN), the Highlands (UK) andÖvre Norrland (SE)). Several classic destinationsfor package-holiday flights, such as Islas Baleares inSpain and the Algarve in Portugal, have a very highaccommodation capacity per head of resident pop-ulation. Cyprus, which has a hotel capacity similarto that of the Algarve, can also be counted as oneof these traditional EU destinations.

The region of Tyrol in Austria provides a typicalexample of how tourism can continue all yearround.

Many holidaymakers do not, of course, fly totheir destination, especially on shorter breaks,which are becoming increasingly popular. A num-ber of regions with an extensive hotel infrastruc-ture lie within comfortable driving range of majorconcentrations of urban population. Examples in-clude, in the United Kingdom, West Wales and theValleys, Dorset and Somerset, and the Black For-est region in Germany. Central Sweden is also anattractive destination for short holiday breaks.

Turning specifically to the total number of bed-places, Map 12.2 gives a clear overview of thenumber of beds available in hotels and similar es-tablishments as a proportion of the total bed-places in each region. Apart from hotels and sim-ilar establishments, holiday homes, campsites andother facilities such as youth hostels, tourist resi-dences, etc. are also counted as tourist accommo-dation. It is interesting to note from this map thatthe concentration of hotel beds in urban areas andaround the respective capitals is higher than inother areas. This is most evident in France, wherein Paris over 75 % of the total number of bed-places is accounted for by hotels. This is also thecase in Berlin and the Rhine/Main area aroundFrankfurt in Germany, and in Greater London inthe United Kingdom. It can thus be concludedthat tourist accommodation density varies widelybetween regions, even within the same country.

The number of hotel beds as a proportion of thetotal number of bedplaces is also high in other re-gions, particularly in Scotland, parts of Englandand Greece. In these areas, this can be explainedby the type of accommodation that prevails in thecountry or region in question. In rural regions, forexample in many parts of Belgium and the


Netherlands, the west and south-west of France,Denmark, most parts of Sweden and also inPoland (with the exception of the area aroundWarsaw), the proportion of total bedplaces ac-counted for by beds in hotels and similar estab-lishments can be less than 25 % and does not ex-ceed 40 %. A third group of regions comprisingthe Baltic States (Estonia, Lithuania and Latvia)and Finland can be situated between these two ex-tremes: in these countries, the bed capacity in

hotels and similar establishments represents 40 to75 % of total bedplaces.

Occupancy dataWhile tourist infrastructure figures such as thoseused for Maps 12.1 and 12.2 provide an


TO

UR

IS

M

12Number of bedplaces

(bedplaces per 1 000 inhabitants)2003 — NUTS 2


Population data: 2002EL, FR91, FR92, FR93, FR94: 2002RO: 2001

Statistical data: Eurostat database: REGIO© EuroGeographics, for the administrative boundariesCartography: Eurostat — GISCO, February 2005

Map 12.1

indication of the accommodation capacity avail-able in a specific region, it is still important toknow the extent to which this capacity is actuallyused. An occupancy measurement is thereforenecessary. The NewCronos database containsdata on accommodation and the number ofovernight stays at NUTS 2 level for the years 1994to 2004. These figures are broken down furtheraccording to residents and non-residents. Non-

residents are people living in a country other thanthat in which the region is situated.

Since the indicator here shows the percentage oftotal overnight stays, it is possible to ascertain theproportion of foreign tourists and thus the attrac-tiveness of a region for international tourism.

However, the percentage of overnight stays ac-counted for by foreign tourists also depends, of


Number of bedplaces in hotels and similar establishments

as a proportion of total bedplaces2003 — NUTS 2


EL, FR91, FR92, FR93, FR94: 2002RO: 2001


Map 12.2

course, on the size of a country. This percentagewill always be higher in smaller countries than inlarger ones.

The highest percentage of overnight stays by non-residents can be found in the Austrian Länder ofVorarlberg and Tyrol, Estonia, Cyprus, Luxem-bourg and the Flemish part of Belgium. Thisshows how heavily some of these countries, suchas Austria and Cyprus, depend on foreign visi-

tors. In Germany and Great Britain and also largeparts of Spain and Italy (with the exception of thecoastal regions), on the other hand, there is lessdependence on foreign tourists. Domestictourism plays a predominant role in these bigcountries.

A very different picture emerges if one examinesthe variations in overnight stays in hotels and sim-ilar establishments between 2002 and 2003,


TO

UR

IS

M

12

Nights spent in hotels and similar establishmentsvariation 2003/2002 (%) — NUTS 2

NUTS 2

> 62 to 6–1 to 2–4 to –1≤ –4Data not available

DEB1, DEB2, DEB3: 2003/1999EL, ES64: 2002/2001UK: 2002/2001; UKK4: 2002/1999SK: 2003/2001RO: 2001/2000


Map 12.3

particularly for the new Member States. In Esto-nia and Latvia, on the Baltic coast of Poland, inHungary and in Bulgaria, the number ofovernight stays rose by up to 6 % — and some-times even more — in comparison with the previ-ous period. Although these percentage changesare influenced by the reference values used for thecomparison (a relatively low reference value givesa very high percentage change), they neverthelessreflect a change in travel behaviour. This may well

be due to economic pressure forcing people tochoose holiday destinations closer to home. How-ever, political or other reasons for preferring newdestinations over traditional holiday countriessuch as Spain and Italy may also play a role. It isalso evident that the enlargement process andwith it the attraction of new holiday destinationsin the new Member States already had an effect ontourism in 2003. This trend needs to be moni-tored in future years.


Nights spent in hotels and campsites by non–residents

(proportion of total nights)2003 — NUTS 2


EL, UK: 2002RO: 2001


Map 12.4

The list of the 20 most important tourism regionsin the EU-25 does not include any regions in thenew Member States. Traditionally, south Euro-pean regions predominate in the list. However,established holiday regions such as Tyrol (Aus-tria), Oberbayern (Germany) and Gelderland (theNetherlands) also feature. As far as accommoda-tion type is concerned, there is no obvious patternthat applies to all 20 regions. The accommodationstructure depends on the individual region, al-though hotels and campsites are the most com-monly found types of accommodation.

ConclusionIn recent years, European tourism and related in-dustries have undergone major changes. The datacollected by the Member States and published byEurostat show that tourism is gaining in impor-tance for the European regions. The main factorthat encourages regions to increase their attrac-tiveness is the trend towards more frequent andshorter holidays. The examples given above will,it is hoped, encourage readers to make evengreater use of the regional data on Europeantourism.


TO

UR

IS

M

12

FR — Rhône-Alpes (370 199)IT — Trentino-Alto Adige (367 003)

ES — Andalucia (334 409)FR — Bretagne (319 794)

AT — Tirol (264 746)DE — Oberbayern (232 491)

UK — London (188 139)UK — Dorset and Somerset (173 580)

NL — Gelderland (166 514)EL — Notio Aigaio (152 914)

IT — Sicilia (133 564)PT — Algarve (130 693)

PT — Lisboa e vale do Tejo (127 115)

0 % 20 % 40 % 60 % 80 % 100 %

ES — Cataluña (648 382)FR — Provence-Alpes-Côte d'Azur (431 767)

ES — Islas Baleares (414 396)IT — Toscana (405 259)

FR — Languedoc-Roussillon (392 513)ES — Canarias (377 635)

IT — Veneto (652 721)

Graph 12.1 — Top 20 EU-25 tourist regions — Distribution of bed placesby type of accommodation, 2003 — NUTS 2

Hotels Campsites

Dwellings Other


0

50 000

100 000

150 000

200 000

250 000

300 000

350 000

BE DK DE EL ES FR IE IT LU NL AT PT FI SE UK BGCYCZ EE HULT MTLV PL ROSI SK

Graph 12.2 — Inbound and domestic tourism in 2003 — Nights spent in hotels andcampsites by residents and non-residents

Res./1 000 Non-res./1 000


EUROPEAN UNION: NUTS 2 regionsBE10 Région de Bruxelles-

Capitale/BrusselsHoofdstedelijkGewest

BE21 Prov. AntwerpenBE22 Prov. Limburg (BE)BE23 Prov. Oost-

VlaanderenBE24 Prov. Vlaams-BrabantBE25 Prov. West-

VlaanderenBE31 Prov. Brabant WallonBE32 Prov. HainautBE33 Prov. LiègeBE34 Prov. Luxembourg

(BE)BE35 Prov. NamurCZ01 PrahaCZ02 Střední ČechyCZ03 JihozápadCZ04 SeverozápadCZ05 SeverovýchodCZ06 JihovýchodCZ07 Střední MoravaCZ08 MoravskoslezskoDK00 DanmarkDE11 StuttgartDE12 KarlsruheDE13 FreiburgDE14 TübingenDE21 OberbayernDE22 NiederbayernDE23 OberpfalzDE24 OberfrankenDE25 MittelfrankenDE26 UnterfrankenDE27 SchwabenDE30 BerlinDE41 Brandenburg —

NordostDE42 Brandenburg —

SüdwestDE50 BremenDE60 HamburgDE71 DarmstadtDE72 GießenDE73 KasselDE80 Mecklenburg-

VorpommernDE91 BraunschweigDE92 HannoverDE93 LüneburgDE94 Weser-EmsDEA1 DüsseldorfDEA2 KölnDEA3 Münster

DEA4 DetmoldDEA5 ArnsbergDEB1 KoblenzDEB2 TrierDEB3 Rheinhessen-PfalzDEC0 SaarlandDED1 ChemnitzDED2 DresdenDED3 LeipzigDEE1 DessauDEE2 HalleDEE3 MagdeburgDEF0 Schleswig-HolsteinDEG0 ThüringenEE00 EestiGR11 Anatoliki Makedonia,

ThrakiGR12 Kentriki MakedoniaGR13 Dytiki MakedoniaGR14 ThessaliaGR21 IpeirosGR22 Ionia NisiaGR23 Dytiki ElladaGR24 Sterea ElladaGR25 PeloponnisosGR30 AttikiGR41 Voreio AigaioGR42 Notio AigaioGR43 KritiES11 GaliciaES12 Principado de AsturiasES13 CantabriaES21 País VascoES22 Comunidad Foral de

NavarraES23 La RiojaES24 AragónES30 Comunidad de

MadridES41 Castilla y LeónES42 Castilla-La ManchaES43 ExtremaduraES51 CataluñaES52 Comunidad

ValencianaES53 Illes BalearsES61 AndalucíaES62 Región de MurciaES63 Ciudad Autónoma de

CeutaES64 Ciudad Autónoma de

MelillaES70 CanariasFR10 Île-de-FranceFR21 Champagne-ArdenneFR22 Picardie

FR23 Haute-NormandieFR24 CentreFR25 Basse-NormandieFR26 BourgogneFR30 Nord - Pas-de-CalaisFR41 LorraineFR42 AlsaceFR43 Franche-ComtéFR51 Pays de la LoireFR52 BretagneFR53 Poitou-CharentesFR61 AquitaineFR62 Midi-PyrénéesFR63 LimousinFR71 Rhône-AlpesFR72 AuvergneFR81 Languedoc-RoussillonFR82 Provence-Alpes-Côte

d’AzurFR83 CorseFR91 GuadeloupeFR92 MartiniqueFR93 GuyaneFR94 RéunionIE01 Border, Midland and

WesternIE02 Southern and EasternITC1 PiemonteITC2 Valle d’Aosta/Vallée

d’AosteITC3 LiguriaITC4 LombardiaITD1 Provincia Autonoma

Bolzano/BozenITD2 Provincia Autonoma

TrentoITD3 VenetoITD4 Friuli-Venezia GiuliaITD5 Emilia-RomagnaITE1 ToscanaITE2 UmbriaITE3 MarcheITE4 LazioITF1 AbruzzoITF2 MoliseITF3 CampaniaITF4 PugliaITF5 BasilicataITF6 CalabriaITG1 SiciliaITG2 SardegnaCY00 Kypros/KıbrısLV00 LatvijaLT00 LietuvaLU00 Luxembourg (Grand-

Duché)

AN

NE

X


HU10 Közép-MagyarországHU21 Közép-DunántúlHU22 Nyugat-DunántúlHU23 Dél-DunántúlHU31 Észak-MagyarországHU32 Észak-AlföldHU33 Dél-AlföldMT00 MaltaNL11 GroningenNL12 FrieslandNL13 DrentheNL21 OverijsselNL22 GelderlandNL23 FlevolandNL31 UtrechtNL32 Noord-HollandNL33 Zuid-HollandNL34 ZeelandNL41 Noord-BrabantNL42 Limburg (NL)AT11 BurgenlandAT12 NiederösterreichAT13 WienAT21 KärntenAT22 SteiermarkAT31 OberösterreichAT32 SalzburgAT33 TirolAT34 VorarlbergPL11 ŁódzkiePL12 MazowieckiePL21 MałopolskiePL22 ŚląskiePL31 LubelskiePL32 PodkarpackiePL33 ŚwiętokrzyskiePL34 PodlaskiePL41 WielkopolskiePL42 ZachodniopomorskiePL43 LubuskiePL51 DolnośląskiePL52 OpolskiePL61 Kujawsko-Pomorskie

PL62 Warmińsko-Mazurskie

PL63 PomorskiePT11 NortePT15 AlgarvePT16 Centro (PT)PT17 LisboaPT18 AlentejoPT20 Região Autónoma dos

AçoresPT30 Região Autónoma da

MadeiraSI00 SlovenijaSK01 Bratislavský krajSK02 Západné SlovenskoSK03 Stredné SlovenskoSK04 Východné SlovenskoFI13 Itä-SuomiFI18 Etelä-SuomiFI19 Länsi-SuomiFI1A Pohjois-SuomiFI20 ÅlandSE01 StockholmSE02 Östra MellansverigeSE04 SydsverigeSE06 Norra MellansverigeSE07 Mellersta NorrlandSE08 Övre NorrlandSE09 Småland med öarnaSE0A VästsverigeUKC1 Tees Valley and

DurhamUKC2 Northumberland and

Tyne and WearUKD1 CumbriaUKD2 CheshireUKD3 Greater ManchesterUKD4 LancashireUKD5 MerseysideUKE1 East Riding and

North LincolnshireUKE2 North YorkshireUKE3 South YorkshireUKE4 West Yorkshire

UKF1 Derbyshire andNottinghamshire

UKF2 Leicestershire,Rutland andNorthamptonshire

UKF3 LincolnshireUKG1 Herefordshire,

Worcestershire andWarwickshire

UKG2 Shropshire andStaffordshire

UKG3 West MidlandsUKH1 East AngliaUKH2 Bedfordshire and

HertfordshireUKH3 EssexUKI1 Inner LondonUKI2 Outer LondonUKJ1 Berkshire,

Buckinghamshire andOxfordshire

UKJ2 Surrey, East and WestSussex

UKJ3 Hampshire and Isle ofWight

UKJ4 KentUKK1 Gloucestershire,

Wiltshire and NorthSomerset

UKK2 Dorset and SomersetUKK3 Cornwall and Isles of

ScillyUKK4 DevonUKL1 West Wales and the

ValleysUKL2 East WalesUKM1 North Eastern

ScotlandUKM2 Eastern ScotlandUKM3 South Western

ScotlandUKM4 Highlands and IslandsUKN0 Northern Ireland


CANDIDATE COUNTRIES: NUTS 2 regions

BG11 SeverozapadenBG12 Severen tsentralenBG13 SeveroiztochenBG21 YugozapadenBG22 Yuzhen tsentralenBG23 YugoiztochenRO01 Nord-EstRO02 Sud-EstRO03 SudRO04 Sud-VestRO05 VestRO06 Nord-VestRO07 CentruRO08 București

regions: statistical yearbook 2005 · cronos available. however, to assist comprehen-sion of the...

Documents