Nielsen, Bue: Employment and economic activities in the EC.
Geografisk Tidsskrift 84: 62-69. Copenhagen January 1984.

Regional disparities in unemployment and economic activities in the EC in the late I97o'es. Various concepts of unemployment are discussed. The performance of regions and countries using different concepts are illustrated.

Bue Nielsen, M.Se. Geographical Institute, University of Copenhagen,
Haraldsgade 68, DK-2100 Copenhagen 0.

Keywords: Economic geography, European communities, regional
development, unemployment.

International comparisons of the level of unemployment are difficult. Unemployment statistics in different countries, as they are plotted by the labour exchanges, generally give a picture of unemployment more in alignment with the unemployment policy of the country than a true measure of need for and real lack of employment. This is true also for the harmonized statistics on registered unemployed used in the EC.

LABOUR FORCE SURVEYS

EUROSTAT's distributions of population by categories of activity, published in the so-called Labour Force Surveys, give, however, a uniform material for a more detailed description of unemployment (EUROSTAT, 1977, 1981). Some results from an investigation basedon these data will be presented in the following.

The publications from the Labour Force Surveys distribute the population into different groups with varying contacts to the labour market: persons with a main occupation, unemployed (in the strict sense), persons with occasional employment, seeking employment, housewives, children under 14 years of age, pupils, pensioners and other non-active persons. In this paper the total population, children and pensioners excluded, is called working population, and the paper studies the division of the working population in occupied and non-occupied.

The reasons for including pupils and housewives in the working population might be discussed. The housewives are, however, well-known as a labour reserve. A change in their number is an important indicator of the state of the

labour market, even if this does not apply to their absolutenumbers. The latest years' development in Denmark, where prolonged school attendance has substituted nonexistingjobs .on the labour market - and therefore is not a sign of growing need for education - justifies the inclusionof this group in the working population, even if demographicfactors also are at play here.

The Labour Force Survey is based on a sample of the population. This requires the use of rather large regions to avoid too great unreliability in the compared data. In this paper the 105 regions in the EC used by EUROSTAT are therefore aggregated to 50, centered around 5 millions inhabitants. habitants.Calculations of reliability are presented in Nielsen

Unemployment

The development in the main groups of population from
1975 to 1979 for the total of EC is shown in the tables 1
and 2.

The data in these tables show that the number of occupied persons in the EC increased during the years 1975-79. growth did not keep pace with the growth in the working population, however. Thus the number of nonoccupied persons increased in both absolute and relative numbers. The two sexes indicate rather different performances, as the number of occupied men only shows a very modest development, whereas the employment of women increased relatively and absolutely.

The regional distribution of the non-occupied population is mapped as figure 1. National levels of nonoccupation vary enormously. There are low rates of activity in the Netherlands and in southern Italy. Also northern Italy, Mediterranean France, Ireland, Belgium, northernmost France and the western part of Central Germany give values below the average for the EC.

This distribution of the regions does not indicate the distribution of total unemployment, however. The.social factors behind the equally low rates of activity in the Netherlands and in southern Italy are very different, the composition of the non-occupied population differs and the social problems are thus very different in nature and depth.

The lower rate of activity in the western part of Germany and northernmost France compared to that of the neighbouring areas in the same countries results from the well-known structural problems for coal mining etc., from which the population of those areas suffers.

The non-occupied groups of the working population 1975 and 1979 are shown in table 3. The number of unemployed categorized as such grew 700.000 during these 4 years. But the number of all non-occupied rose with not less than 4'/2 million in the same period. This growth occured in all groups, the housewives excluded. The difficult situation on the labour market clearly results in many different reactions from those without jobs. The strict concept of unemployment used by the EC, and in the national registrations as well, is completely insufficient to grasp this social reality.

large countries alone, we observe a definite fall in the United Kingdom (with Ireland) and a marked rise in France and Italy. In terms of total non-occupied however, United Kingdom has performed badly, whereas France is best placed among the countries in the EC. The picture is reversed! It is necessary to evaluate all groups of nonoccupied if one is to form a realistic opinion of the character and dimension of unemployment.

The regional distribution of the changes in the proportion
of non-occupied persons is mapped in figure 2. Data
for the countries are given in table 6. Denmark, France

This fact is underlined by the geographical distribution of the unemployed and the other non-occupied groups. Changes in the different groups of non-occupied persons in the countries of the EC in the years 1975-79 are shown in table 4. If we look at registered unemployment in the and the Netherlands have performed better than the average of the EC, and the proportion of occupied has increased. In the other end, especially the United Kingdom lost jobs.

The regional performance is complex, particularly in

Germany. But it can be observed, that the new areas of industrialization in western France and central Italy figure among those few regions with a growing share of occupied persons, whereas most regions in United Kingdom, Irelandand southern Italy have witnessed a clear deterioration.The same has happened in traditional problemareassuch as Wallonia and the Ruhr, but also in Lower Saxony and southern Germany.

Further hints about the changes in the structure of employment can be obtained by dividing the population into men and women. The regional distribution of the proportion of the female working population which is occupied is mapped in figure 3. The areas weith total employment expansion are all characterized by a rapidly growing proportion of women in work.

This occurrence is in most of these regions rather precarious, however. All regions with a large growth in the employment of women (western and south-western France, central Italy and the southern part of the Netherlands) are characterized by industries demanding cheap, stable and unskilled labour. The clothing and shoe industries dominate, together with modern component production in electronics and other metal manufacture, cf. Lipietz (1977), Arcangeli et al. (1980) and Läpple & van Hoogstraten (1977). A great deal of the employment is occasional, or outside the regular labour market and thus with long working-days, dangerous environment in the factories and lack of social security.

Comparisons of respective development in men's and women's employment is shown in figure 4. Many regions with growing proportions of female employment simultaneouslywitnessed a decrease in male employment. The

new possibilities for employment in the decentralised
manufacturing and office functions have not benefited the
men, that are rendered superfluous. Instead the demand
has sought other groups on the labour market, often
housewives from rural communities.

Occupation

The material in the Labour Force Surveys will be used to dig a bit further into the regional changes in the distribution of economic activities. The years 1973 and 1979 will be compared in order to get the full impact of the economic crisis in the data (EUROSTAT 1975, 1981).

The Labour Force Surveys give figures for persons with a main occupation, falling in 11 activity groups. The classification of activity groups is shown in table 5. We use the aggregated regions mentioned above in order to reduce the unreliability of the samples. The 1973 Labour Force Survey did not include Denmark and the Republic of Ireland. These countries are therefore omitted from the analysis. The data from 1973 used here has been corrected in several ways by taking the following points into consideration: There have been some changes in the delimitation of regions between 1973 and 1979, especially in the northern part of England, and in Lower Saxony. For some regions, mainly in France, new estimates of the total population in 1973 have been published. Finally the sum of the employment in the 11 activities in each region has been »grossed-up« to comply with the figures for total employment; this concerned especially the United Kingdom and the Netherlands. In total, these correction procedures only involved small modifications.

Shift-share analysis

The data from the Labour Force Surveys 1973 and 1979 distributed over 11 activities and 49 regions are used in a shift-share analysis. Keeble et al. (1981) made an analogue analysis on the basis of similar material on an aggregated regional level. Their conclusions do not need to be repeated. In order to reach more detailed results we will analyse some of the activity groups in the much finer regional division of the EC used here.

The type of shift-share analysis used is more thoroughly
described in Nielsen & Sørensen (1984). We shall only
give the mathematical equations.

The change in the employment in each of the 11 activity
groups in each of the 49 regions is divided into 3 components,
called general, structural and specific:

The development of the total employment from 1973 to 1979 is mapped in figure 5. The growth has been above the average in Italy, France and the Netherlands. Generally the growth rates are high in regions adjoining the traditional centres, as in southwestern UK, central France, the peripheral Netherlands and North East Italy, but the growth rates are high in the most of southern Italy as well. On the other hand employment has declined in Germany, where the traditional manufacturing areas in west and south are hard hit. The same applies to other regions with traditional industries, such as Wallonia and Nord-Pas de Calais.

The structural strength of the regions in 1973 is measured by relating the structural component in the shift-share analysis to the total 1973 employment. The regional distribution of this measure is shown in figure 6. Regions with a favourable industry-mix - using this classification of activities - generally were Great Britain, Wallonia, partly Germany and the capitals of the EC. Activities with low general rates of growth dominated Italy, France and southern Germany.

The industry-mix in the regions 1973 show only very modest signs of connection with the actual development of their employment the following years. 13 of the 19 structurally strong regions mapped in figure 6 had growth

rates well below the average of the EC. The favourable structure of activities was more than offset by a negative specific change in the employment in the regions. On the other hand, 14 of the 22 structural weak regions experiencedgrowth rates above average, caused by great specific gains. These regions include the whole of Italy, except Lazio.

The regional distribution of the specific component for all activities is mapped in figure 7. The specific gains show a regional pattern much in accordance with total growth. Of the areas with a growth clear above the average only He de France and western Netherlands have negative specific components. The growth can thus be related to the structure of activities only in these two regions.

The specific gains in employment which Italy and other peripheral regions in the EC witnessed in the years 1973-79 very unequally distributed among the activities, however. The manufacturing activities show this very clearly. The specific redistribution in metal manufacture (NACE 3) and other manufacturing industries (NACE 4) are mapped in figure 8 and 9 respectively. These activities should be seen as the most important productive ones in a long-sighted development context. The regional pattern for the manufacturing activities is very much different from the pattern for all activities taken together. The Ital-

ian regions all lose, just like the traditional crisis-area of Wallonia. The large specific redistribution of employment to Italy, and particular southern Italy, seems to be either markedly unproductive as consumer service and administration, or associated with an old-fashioned agriculture and speculative building activities, none of which guarantee future growth.

The regional distribution of specific gains in manufacturing is somewhat complex, but seems to favour the southern parts of UK, parts of central Germany and a good section of France. The dispersion of manufacturing employment from the old centres does not reach the more peripheral parts of the EC in the years of crisis.

CONCLUDING REMARKS

In the institutions of the EC much is said about the necessity of regional equalization in the EC as a precondition for further integration. The materials presented in this paper concerning the patterns of unemployment and the most important productive activities do not show tendencies to such an equalization. They rather corroborate that the development leads to bigger disparities as stated by so many observers during the latest years.

Resumé

I artiklen benyttes data fra EUROSTAT's stikprøveundersøgelser
af arbejdsstyrken til at belyse arbejdsløsheden og erhvervsændringerne
i EF i slutningen af 1970'erne.

EUROSTAT opdeler befolkningen i en række kategorier, afhængigt hængigtaf arbejdssituationen: Beskæftigede, arbejdsløse, med tilfældigt arbejde, arbejdssøgende, husmødre m/k, børn under 14 år, elever, pensionister samt andre ikke-beskæftigede. Befolkningen, fraregnet børn og pensionister, kaldes i denne artikel arbejdsbefolkning. Artiklen undersøger ændringerne i arbejdsbefolkningens fordeling på beskæftigede og ikke-beskæftigede.

Tabel 1-3 viser udviklingen for EF som helhed. Kvindernes aktivitetsrate steg, mens mændenes faldt kraftigt. Mens antallet af arbejdsløse, der er registreret som sådan, voksede med 700.000 i årene 1975-79, steg det samlede antal ikke-beskæftigede 4'/2 million. Det snævre arbejdsløshedsbegreb, som almindeligvis bruges, er helt utilstrækkeligt til at indfange denne sociale realitet. Dette understreges af tabel 4, der viser, at indtrykket af arbejdsmarkedet i f.eks. Storbritannien og Frankrig er helt afhængigt af det benyttede arbejdsløshedsbegreb.

Områder med vækst i beskæftigelsen (se figur 2 og 3) er præget af vækst i kvindebeskæftigelsen i erhverv med billig, stabil og ufaglært arbejdskraft som f.eks. fremstilling af beklædning, fodtøj og elektroniske komponenter. Disse regioner har et kønsopdelt arbejdsmarked; de fyrede mænd går arbejdsløse, mens nye lag af kvinder indrulleres i proletariatet.

Beskæftigelsesudviklingen i årene 1973-79 undersøges ved hjælp af en regional omfordelingsanalyse med 11 erhverv fordelt på 49 regioner. Der mangler data for Danmark og Irland. For erhvervene taget under ét er der en specifik omfordeling navnlig til Italien (se figur 7), men denne omfordeling skete ikke i industrierhvervene(figur 8 og 9). Den specifikke omfordeling til Italienhar navnlig fundet sted i de uproduktive erhverv. Spredningenaf

genafindustribesk'æftigelsen fra de gamle centre er under krisen
ikke nået ud til de perifere dele af EF.

Der tales i EFs organer meget om, at en formindskelse af de regionale uligheder er en nødvendig betingelse for fortsat integration. Resultaterne i denne artikel viser imidlertid, at der foregår en regional udvikling, der leder til stadig større forskelle i EF.

References

Arcangeli, F., C. Borzaga & S. Goglio (1980): Patterns of peripheral
development in Italian regions, 1964-77. Papers, Reg. Sc.
Ass., vol. 44, p. 19-34. Urbana.

EUROSTAT (1975): Regional Statistics, 1973-74. Population,
Employment, Living Standards. Luxembourg.

EUROSTAT (1977): Regional Statistics, 1975. Population, Employment,
Living Standards. Luxembourg.

EUR OST AT (1978): Labour Force Sample Survey. Methods and
definitions. Luxembourg.

EUROSTAT( 1981): Labour Force Sample Survey 1979. Regional
Results: Luxembourg.

Keeble, D., P. Owens & C. Thompson (1981): The Influence of
Peripheral and Central Locations on the Relative Development
of Regions. Final Report. University of Cambridge.

Läpple, D. & P. van Hoogstraten (1977): Opmerkingen over de ruimtelijke struktuur van de kapitalistiese ontwikkeling - het vorbeeld van de regionale ontwikkeling in Nederland. ZONE, No. 5, p. 5-45, Amsterdam.

Lipietz, A. (1977): Le capital et son espace. Paris.

Nielsen, B. (1982): Udviklingen i arbejdsløshedens regionale fordeling
i EF 1975-79. Nordisk symposium for kritisk samfunnsgeografi.

Nielsen, B. & P. M. Sørensen (1984): Changes in regional employment
in Denmark in the seventies. Geogr. tidsskr., Copenhagen.