Nielsen, Bue and Sørensen, Peder Michael: Changes in regional
employment in Denmark in the seventies. Geografisk Tidsskrift
84, 69-73. Copenhagen, January 1984.

The paper deals with regional development in employment in Denmark in the 1970'5. It uses shift-share analysis to isolate elements and activities responsible for the regional dispersion experienced. In that context the paper discusses the interrelationship between manufacturing industries and public services.

Bue Nielsen, Senior Lecturer, and Peder Michael Sørensen, M.
Sc. and Research Fellow. Geographical Institute, Haraldsgade 68,
DK-2100 Copenhagen 0.

Keywords: Regional development, occupations, manufacturing,
public services.

During the 1970's Denmark witnessed a reversal of the trend towards a centralization of population and employment which had characterized previous decades. The purpose of the paper is to deliver a short description of the changes in occupational structure involved in this decentralization.

DATA

In this analysis we use data from the Employment Surveys for the years 1972 and 1978.' Changes in the methods of registration² and in the classification of occupation s³ in the surveys since then prevent any meaningful investigation into the changes of employment in later years.

The Employment Surveys are based on samples of people aged 15-74 and record the development of the labour force. An assessment of the development of the employment based on these data might thus be disturbed by differing developments in the regional distribution of unemployment. It can be shown, however, that for the period studied this is not the case. The development described in this paper concerning the labour force can without further qualifications be regarded as valid for the development of the employment as well.

The classification of activities used here is more detailed than that published by Danmarks Statistik. This might, dependent on the size of the sample, influence the results for minor industries in small regions (e.g. Bornholm) for which some reservation thus should be taken.

GENERAL DEVELOPMENT

The total labour force in Denmark showed a moderate growth of 9% in the years 1972-78. Table 1 gives the figures for the regional growth rates to each side of this average.

The labour force grew in all counties, Central Copenhagen excluded, brought about by the recruitment of the labour market. The growth in northeastern Zealand, however, was notably below the average, whereas the other counties, except Fyn and Storstrøm, showed growth rates well above.

The growth rates in the 14 activity groups are shown in
table 2. The structure of activities changed markedly in
the period. Agriculture and Other Services witnessed a

t

decline of 8% whereas the labour force in public activities, e.g. Administration, Education, and Social & Health Services, and in some private ones, e.g. Professional Services and Financing, all grew between 15-40%.

The proportion of the labour force in the different activities in northeastern Zealand is shown in table 3. Generally in 1972 northeastern Zealand had proportions above the average in those activities showing high growth rates 1972-78, proportions below the average in declining activities. Taking the structure of activities as a basis for evaluating the growth in northeastern Zealand would lead to the conclusion that the region's labour force would grow faster than the average. As mentioned, the opposite is the case. In the following, we will use a shift-share analysis to qualify this contradiction.

Shift-share analysis

Shift-share analysis is used in different versions. In this paper we divide the growth in an activity in a county in the period 1972-78 into three components, called general, structural, and specific.

The general growth is the growth in the labour force in a county's activity which is proportional to the total growth in the labour force in Denmark 1972-78. The structural growth is that part of the growth in an activity which contributes to the total growth in a county, and results from a greater growth rate in that activity than that of the total labour force in Denmark.

The specific growth is the growth in the labour force in an activity caused by changing regional distribution of the labour force in that activity. The mathematical equations are:

gu, sty, and sp,j are general, structural and specific growth in activity /in region j 1972-78, respectively, tyi and t_ij2 are employment in activity i in region j in 1972 and 1978, respectively, T_n and T_i2 are employment in activity /in Denmark in 1972 and 1978, respectively. T! and T₂ are the total employment in Denmark 1972 and 1978, respectively. We are not going to examine and critisize the various methodological possibilities in the shift-share analysis .⁴ It can be shown that the greatest problems in the use of the method are associated with the classification of activities .⁵ This problem is all the more crucial when activity groups - as here - are not theoretically defined, but determined by official statistics. However, for our purpose - a description based on over-average occurrences of growth activities and solely this - the method is fully justified.

RESULTS

The main results of the shift-share analysis are shown in table 4. The figures for northeastern Zealand show that the growth of this area stems from the activity structure, whereas within each activity a regional redistribution away from northeastern Zealand has occurred.

In contrast, the other counties show: an activity structure dominated by slowly growing activities, but the specific component is positive. There is only one exception: the county of Århus, where the structure of activities itself yields a small positive contribution to the growth. It can be remarked too that the county of Storstrøm only receives a very modest share of the regional redistribution from northeastern Zealand.

The counties of Ribe and Ringkøbing have extremely high specific gains. (That this does not apply to Viborg is due to the development in agriculture). The specific gains are especially marked in Manufacturing and Public Services, and in the county of Ribe, in Transport as well.

Differences in dispersion

The general pattern of dispersion from northeastern Zealand to the rest of the country, especially Jutland, applies to most of the 14 activity groups. But some deviations deserve attention. The dispersion from northeastern Zealand was intense in Manufacturing, but was noticeable in other activities too, private as well as public ones, cf. table 3. On the other hand, the group of Other Services concentrates to northeastern Zealand, even to the Central Copenhagen area. But generally it is not the case that the white-collar occupations remain in Copenhagen. Professional Services are, in fact, second only to Manufacturing in declining percentage in northeastern Zealand.

Disparities between peripheral counties

As mentioned the county of Storstrøm deviates from the
peripheral counties of Jutland by showing a falling percentageof

centageofthe labour force, even though the growth markedlyexceeded that of northeastern Zealand. This fact is not caused by a different structure of activities in Storstrøm. On the contrary, Storstrøm received only a tiny fraction of the specific redistribution of employment from northeastern Zealand. The development in Manufacturing,where the redistribution did not reach the county, is mainly responsible for this. But the public sectoradded to this deviating behaviour as well. On the other hand, the county received reasonable shares of activities,such as Financing and Professional Services, cf. table 5. The small gains of the peripheral counties in Jutlandin those private activities might indicate the existenceof some hesitation concerning too remote locations in these activities where some contact to a metropolis is necessary.

Explanations

In Denmark and in other countries as well much effort has been made to develop theories concerning regional development in the 1970's.⁶ The theories have mainly dealt with the development of the manufacturing industries including aspects of location.⁷

Also migration data have been used as basis for a theory formation.⁸ This development of theory has too often focused on single parameters and thus hindered a more total understanding of the regional development process. Our stance is that the connection between manufacturing industries and the public sector is central for an understanding of the regional development as described in this paper. These activities show the most marked regional redistributions and seem to be interconnected in the pattern of development.

A marked dispersion of manufacturing industries from Copenhagen to the peripheral counties took place already in the 1960's.⁹ This tendency continued in the 1970's as shown in this paper. This growth of manufacturing employment (excluding handicraft) in the periphery did in the 1970's occur especially in the smaller towns, as shown in fig. 1. With a single exception, growth in manufacturing Peripheral countries are shown in grey. Size of column indicates industrial employment by municipalities with biggest town: 1. less than 5,000 inhab. 2. between 5-10,000 inhab. 3. between 10-20,000 4. between 20-100,000 inhab. 5. more than 100,000 inhab. 6. the whole county.

Source: Danmarks Statistik, Industristatistikken.

in the peripheral counties is characterized by gains in the smaller towns compensating losses in the larger cities. In the more central counties the growth in the small towns did not counterbalance the decline in the cities. Furthermore, it can be shown that only few of the towns in the peripheral counties experience any noteworthy growth in manufacturing.¹⁰ Most of them actually stagnated or declined in manufacturing employment. The growth was localized to a few towns.

The regional redistribution of employment in manufacturingis influenced by a long range of factors such as regional disparities in wages, trade-union attitudes, qualificationsof the labour force, regional subsidies ground rent, environmental regulations and so forth. There are also regional differences in the amount of labour power that can be recruited from agriculture and other smallscaleproduction. Former self-employed and assisting housewives are generally an extremely stable and loyal labour force and constitute the hard core of the labour force in most firms in the small towns, a fact that deserves

more attention in the discussion of regional development. And too, the dissolution of local ties in sales and purchaseshas facilitated the relocation of industry from the larger markets.

The regional redistribution of manufacturing employment has been determined by a marked rise in the level of public service." This concerns possibilities for care of children and retired people - a function of the family badly concordant to wage labour; recreation, such as sport facilities and libraries; better general education and health services. Unlike ten years ago, regional disparities in these functions are now only minor. This development has led to an extraordinary growth in the employment in the public sector. The growth has been tremendous in the peripheral counties, where the dispersion of industry created a large demand for public service. The development in the regional distribution of public employment can thus by and large be explained by the already existing redistribution of manufacturing employment.

CONCLUSIONS

This paper has shown that the labour force in most regions outside northeastern Zealand in the 1970's grew at a higher rate than the average rate for Denmark. This growth cannot be explained by the structure of activities in single regions. Instead specific regional redistribution within the activities away from northeastern Zealand was responsible.

Manufacturing, where employment generally stagnated, showed large redistributions from northeastern Zealand to the 5 peripheral counties in Jutland, as did the public sector. The explanation of the new regional pattern of manufacturing is to be found in changed conditions of production and location, while the regional redistribution in the public activities is connected to the development in manufacturing industry.

Resumé

I Danmark er der igennem 1970'erne sket en fortsat stigning i
arbejdsstyrken. Arbejdsstyrkens stigning har været undergennemsnitlig
i Nordøstsjælland mens resten af landet - med undtagelse
af Storstrøms amt og Fyns amt - har haft en overgennemsnitlig

Erhvervsstrukturen har i samme periode undergået markante
ændringer. Den offentlige sektor og erhvervsgruppen Pengeinstitutter
& Forsikring er vokset meget kraftigt, mens der har været
tilbagegang for Landbrug, Anden Service og Detailhandel.
Selv om væksterhvervene var overrepræsenteret i Nordøstsjælland
vises det ved hjælp af en shift/share analyse, at der gennem
1970'erne er sket en kraftig omfordeling til de øvrige regioner.
Det er specielt de jyske udkantsamter, der har fået størst
andel i denne omfordeling, mens Storstrøms amt ikke har fået
andel heri. Udviklingstendenserne i de danske udkantsregioner
ican derfor ikke siges at være entydige.
Den regionale omfordeling har først og fremmest fundet sted
inden for Fremstillingsvirksomhed og de offentlige erhvervsgrupper,
mens de private væksterhverv kun i mindre udstrækning
omfordeles. Således sker der en koncentration af erhvervsgruppen
Anden Service til Nordøstsjælland.
Grundlaget for industriens og dermed Fremstillingsvirksomheds
regionale omfordeling søges forklaret ud fra følgende forhold.
Regionale forskelle eksisterer stadig inden for en række områder,
f.eks. er muligheden for at få en stabil og loyal arbejdskraft
større i udkantsamterne end i Nordøstsjælland på grund af den
stadige frisættelse af arbejdskraft fra småproduktionen.
For at frigøre arbejdskraft til lønarbejde og dermed muliggøre
en industrialisering af udkantsamterne, har det været nødvendigt
med en statslig udbygning af en række servicefunktioner (børnepasning,
ældreforsorg, efteruddannelse etc.). Den offentlige sektors
regionale omfordeling skal derfor ses i sammenhæng med
udviklingen inden for erhvervsgruppen Fremstillingsvirksomhed.

---

7. e.g. Peter Maskeli: Industriens regionale omlokalisering 1970-80, Copenhagen 1982. Peter Hartoft-Nielsen: Den regionale erhvervsstruktur- og beskæftigelsesudvikling, Lavindkomstkommissionens arbejdsnotat 8, 1980. Chresten Storgaard: Teknologi og regional brancheudvikling, Copen-1981.

8. e.g. P. O. Pedersen: Vandringerne og den regionale udvikling, Esbjerg 1983.

9. Ruth H. Jensen: Changes in the Geographical Distrubution of Danish Industry, Geografisk Tidsskrift, vol. 70, Copenhagen 1971, and R. Guttesen et al: Regional Development in Denmark, with special reference to the postwar period, Geo- Tidsskrift, vol. 75, 1976.

10. Peder M. Sørensen: Den teknologiske udviklings betydning for den regionale industriudvikling, in: Teknik och utveckling, ERU-rapport nr. 18, Stockholm 1982.

Notes

1. cf. Statistiske Efterretninger, 1973, nr. 61 and 1980, nr. A4, Danmarks Statistik.

2. Statistiske Efterretninger, 1983, nr. 7, Arbejdsmarked, yield further details.

3. The Employment Surveys use ISIC 58 until 1978; since then they have used ISIC 68 as a base for classification.

4. cf. Nielsen & Sørensen: (forthcoming) where several methods of shift-share are discussed.

5. cf. Vipond & Forward: A critique of Shift and Share Analyses of Australian Urban and Regional Growth. Australian Geographer, vol. 14, 1979, Sydney, D. 228-236.

6. e.g. Anders Skonhoft: Noen momenter om tilpasningen i regionale arbejdsmarkeder, in: Nordisk Symposium for Kritisk Samfunnsgeografi, Finland 1982, discussing tendencies towards a new regional development in Norway. Doreen Massey: In what Sense a Regional Problem? Regional Studies vol. 13, 1979. Roger Andersson et al: Mot en ny regional industristruktur? Industrisysselsättningens utveckling i Sve- riges kommuner 1974-1980, Forskningsrapport nr. 80, Uppsala