39
Population Recent periurban growth in the Ile-de-France: Forms and causes (Population,2,1997) B. Baccaïni Citer ce document / Cite this document : Baccaïni B. Recent periurban growth in the Ile-de-France: Forms and causes (Population,2,1997). In: Population, an English selection, 10année, n°2, 1998. pp. 349-384. http://www.persee.fr/doc/pop_0032-4663_1998_hos_10_2_6833 Document généré le 17/10/2015

Recent Periurban Growth in the Ile-De-France

Embed Size (px)

DESCRIPTION

Creșterea populației în regiunea periurbană a Parisului

Citation preview

Population

Recent periurban growth in the Ile-de-France: Forms and causes(Population,2,1997)B. Baccaïni

Citer ce document / Cite this document :

Baccaïni B. Recent periurban growth in the Ile-de-France: Forms and causes (Population,2,1997). In: Population, an

English selection, 10ᵉ année, n°2, 1998. pp. 349-384.

http://www.persee.fr/doc/pop_0032-4663_1998_hos_10_2_6833

Document généré le 17/10/2015

AbstractBaccaini (Brigitte). - Recent periurban growth in the Ile de France: forms and causes During thepast several decades, settlement in the Ile de France has, in common with the majority of largeurban areas, experienced a rapid decongestion called suburbanization. After highlighting thespecific situation of suburban development in the Ile de France, which has served as a point oftransition between urban areas and the countryside, and the special features of the populationwho lived there in 1990 (over-representation of families with children and underprivileged socialgroups with housing primarily belonging to owner- occupiers) this paper deals with recentmigration of the suburban population in order to explain the causes and types of demographicgrowth in this type of environment. Decongestion of the Paris urban area is responsible for nearlythree quarters of new suburban dwellers, and involves populations with very different socio-demographic characteristics from those who had migrated from the provinces and those who hadlived in the He de France for a longer period of time. The occupational distribution of recent in-migrants to the suburbs also varied significantly in different geographical areas and selectioneffects have reinforced the existing specificity of various sectors.

RésuméBaccaini (Brigitte). - Modalités et causes de la croissance récente des communes périurbai- nesd'Ile-de-France L'Île-de-France connaît depuis plusieurs décennies, comme la plupart des grandesrégions urbaines, un desserrement rapide de l'habitat connu sous le nom de « périurbanisation ».Après avoir mis en évidence les spécificités du milieu périurbain francilien - transition entrel'espace urbain et le monde rural - et de la population qui y vit en 1990 (surreprésentation desfamilles avec enfants et des catégories sociales défavorisées, place prépondérante des logementsindividuels occupés par leur propriétaire), cet article s'intéresse aux migrations récentes despériurbains afin de saisir les causes et les modalités de la croissance démographique de cetespace. Responsable pour presque les trois quarts des arrivées dans le milieu périurbain, ledesserrement de l'agglomération parisienne touche des populations dont les caractéristiquessocio- démographiques diffèrent sensiblement de celles des migrants venus de province ou decelles de la population installée depuis longtemps dans cet espace. La structuresocioprofessionnelle des nouveaux arrivants du périurbain varie également fortement d'un secteurgéographique à l'autre, ces effets de sélection ayant pour conséquence de renforcer lesspécificités socioprofessionnelles existantes des divers secteurs. Un second article montrera lesrelations qui existent entre ces trajectoires migratoires récentes et la mobilité domicile-travail desactifs du périurbain.

ResumenBaccaini (Brigitte). - Modalidades y causas del crecimiento reciente de los municipios periurbanosde Île-de-France Desde hace varias décadas en Île-de-France se observa, como en la mayoria delas grandes regiones urbanas, una expansion del habitat conocida bajp el nombre deperiurbanización. Des- pués de describir las especificidades del medio urbano de Île-de-France -transición entre el espa- cio urbano y el mundo rural - y de su población en 1990 - sobre-representación de familias con hijos y de categorias sociales desfavorecidas, preponderancia deviviendas individuales ocupadas por su propietario -, el présente articulo estudia las migracionesrecientes de la población periur- bana. El objetivo de tal estudio es entender las causas y lasmodalidades del crecimiento demo- gráfico de este espacio. La expansion de la aglomeraciónurbana de Paris explica casi las très cuartas partes de las llegadas al medio periurbano. Lascaracterísticas socio-demográficas de estas poblaciones difieren sensiblemente de lascaracterísticas de los migrantes llegados de provincias о de las aplicables a poblacionesinstaladas desde hace tiempo en tal espacio. La estructura socio-profesional de los re- ciénllegados al medio periurbano varia también fuertemente de un sector geográfico a otro. Estosefectos de selección refuerzan las especificidades socio-profesionales existentes en los diversossectores. En un segundo articulo se mostrarán las relaciones existentes entre estas trayectoriasmi- gratorias y la movilidad domicilio-trabajo de los activos del medio periurbano.

RECENT PERIURBAN GROWTH IN

THE ILE-DE-FRANCE:

forms and causes*

The power of towns and cities to attract population is a long-established phenomenon, but one whose form has changed in the last twenty years. The population of France's 'old' urban centres has stopped increasing and urban growth is now concentrated in communes which until very recently were rural. Moreover, urban centres exercise their attraction as it were from a distance on the communes that are still rural, by means of the employment opportunities available in the urbanized areas. The communes in the metropolitan periphery or 'peri- urban' of Paris have had an annual growth rate of 2.5% in recent years, compared with 0.5% for the population of France as a whole. In this article Brigitte Baccaíni* presents a detailed analysis of the process of growth of the economically active periurban population of the Ile-de-France. It is based on recent census data, and divides the periurban into two concentric zones and four geographical sectors. The examination of the individual characteristics of the economically active population, and the geographical origin of the newcomers, reveals the existence of an important phenomenon: the increasing social segregation of the urban and periurban populations.

The economically active population in the Ile-de-France is characterized by relatively high levels of residential mobility. A large majority (57.3%) changed residence between 1982 and 1990 (compared with 53.3% for metropolitan France)(1). However, most of this mobility is intraregional: less than 15% of the economically active population lived outside the region in 1982 (10.5% in another region of metropolitan France, 3.5% in another country, and 0.6% in France's overseas departments and territories).

These movements are responsible for a gradual redistribution of population within the Ile-de-France, with the trend being towards a spatial de-

* This article is the first part of a study of geographical mobility among the economically active population of the Ile-de-France. A second article, published in French in Population, 2, 1997, deals with home-workplace mobility and the relations between residential mobility and commuter journeys. The present article is thus concerned only with the economically active population, which alone is involved in commuting.

Translated by Godfrey I. ROGERS. * INED. (1) It must be remembered that census data under-estimate mobility by not registering

the multiple migrations which may have occurred between two censuses.

Population: An English Selection, 10(2), 1998, 349-384

350 B.BACCAINI

concentration. In 1962, one-third of this region's population ('Franciliens') lived in Paris itself, 40% in the 'petite couronne' or inner suburban ring, and 25% in the 'grande couronne' or outer suburban ring. By 1990, however, Paris accounted for only 20% and the inner suburban ring for 37%, and it was the outer suburban ring which had the largest proportion (42%). So while natural surplus is responsible for more than 50% of population growth occurring in the Paris urban area comprising the communes of Paris and the inner suburbs, it is migratory surplus which is responsible for over 50% (even over 80% in the communes furthest from the centre) of the growth occurring in the communes situated outside the Paris urban area.

This strong growth of the Paris metropolitan periphery is in part linked to the dynamism of the new towns, where population growth, although not as high as that observed in the 1970s, exceeded 5% a year between 1982 and 1990 (Atlas des Franciliens, 1991, vol. 1).

Compared with its scale during the three previous decades, the process of population déconcentration from Paris slowed down in the period 1982-90 - with a smaller migratory deficit in the central zone, and a smaller migratory surplus in the periurban communes. But despite this "relative - and selective - slowing down of the shift to the periphery" (Berger, 1996), the movement of 'periurbanization' or population decentralization has nonetheless continued and involves zones that are ever more distant from the capital.

The periurbanization process affects different groups in the population to different degrees and in different ways. The residential choices of individuals and households vary considerably depending on life-cycle stage, family situation, social position and financial resources.

In addition, these selection effects tend to operate in different ways in different geographical sectors of the region, leading to a reinforcement or weakening of socio-spatial contrasts.

The socio-spatial organization of the Ile-de-France has been the object of many descriptions, but these often present methodological shortcomings (Berger, 1992). In addition, most of these studies deal only with the Paris urban area, so the processes responsible for residential segregation of socio- demographic categories in the periurban zone are less well known. As Mar- tine Berger (1992, p. 16) has observed, these processes are "the product of social relations in which mentalities, lifestyle practices and identity construction all have a role, alongside the selective logic of the housing market, the mechanisms of housing finance and the system of social housing allocation".

So while not denying the fundamental role of housing supply type in explaining the spatial distribution of the different socio-demographic groups, it is important to consider other, sometimes contradictory, processes. Place of residence cannot be defined solely by reference to the structure of the housing stock, but must include other characteristics that influence the housing choices of individuals: number and quality of

RECENT PERIURB AN GROWTH IN THE ILE-DE-FRANCE 3 5 1

amenities, quality of environment. Differences in the value of accessibility and situation are in fact at the origin of the sharp socio-demographic contrasts often observed between adjacent communes (Berger, 1992).

An in-depth understanding of the changing pattern of periurban settlement necessarily requires an examination of the various migration and residential behaviours according to age, household structures and socio- economic category.

After a description of the main characteristics and socio-demographic particularities of the economically active population of the Ile-de-France periurban in 1990 (slightly fewer than 650,000 individuals), we analyze the residential mobility of this population between 1982 and 1990. The object is to identify the individual characteristics and housing preferences which are determinant in the choice of a periurban location, while also taking into account individuals' residence histories. In this way it is hoped to construct a clearer picture of the population that is affected by the phenomenon of population decentralization. The results from these analyses are used in a subsequent article to identify the impact of these residential choices on the home-workplace journeys of the economically active inhabitants of the periurban.

A widely-used definition of periurban in France is any rural commune situated in a Zone de Peuplement Industriel et Urbain (ZPIU). For the purposes of this article, however, we have adopted a broader definition of what is periurban, by including the whole of the Ile-de-France region outside of the Paris urban area (Paris intra muros plus suburbs). Under this definition the term periurban is also applied to the population of the secondary urban centres of the Ile-de-France. These are towns which fall very largely in the Paris zone of influence and present many of the characteristics that are specific to the periurban milieu in the narrow definition of the term (rural communes) (Boyer, 1992)(2). The periurban zone thus defined comprises a total of 902 communes (rural communes, secondary urban centres and some parts of the new towns).

It must be noted that confining the analysis to the Ile-de-France does have the disadvantage of excluding some highly urbanized sectors, notably on the northern limit of the region, which could be treated as belonging to the zone under the direct influence of Paris.

Our source for this study is the 1990 census. These are the only data to allow analysis at a very detailed geographical level, despite their limitations for the study of spatial mobility (they supply no information about the multiple migrations which may have occurred between 1982 and 1990,

(2) A new zoning system has recently been introduced by INSEE. Urban areas are defined using a process of iteration, from around an urban pole (an urban unit on whose territory at least 5,000 people are employed and which is not directly dependent on a larger neighbouring pole), to the communes from which at least 40% of the active population work at the urban pole. In each urban area, the communes or urban entitites which are not part of the urban pole form the periurban ring. Under this zoning scheme the Ile-de-France contains seven urban areas (Paris, Melun, Meaux, Fontainebleau, Nemours, Coulommiers and Provins).

352 B.BACCAINI

nor about the place of work in 1982, and supply individual characteristics only for 1990)(3).

I. - The periurban environment and population

Characteristics of the periurban For the purposes of our analysis the milieu in Ile-de-France Paris periurban has been divided into

two concentric zones: a first or inner ring that extends 50.5 km from the centre of Paris, followed by a second or outer ring extending to the edges of the region(4). Each of these two zones has been divided into four geographical sectors (north-east, north-west, south-east, and south-west), based on the geographical coordinates of the communes (Figure 1). It has often been demonstrated that the phenomena of socio-spatial segregation in the Ile-de-France, as in most metropolitan areas, are subject to concentric and sectorial divisions simultaneously.

The first or inner ring contains 71% of the economically active population of the periurban, compared with 29% in the second ring. Within this outer ring, the south-east sector, which is much larger than the others, is the only one to contain a large proportion of the active periurban population (21%).

The second or outer ring is significantly more rural: the rate of urbanization of the active population (which is 70.6% in the periurban region as a whole) is 75.4% in the inner ring as against 58.9% in the outer ring. It may also be noted that within the inner ring, the east is more urban than the west (more than 80% of the active population of the eastern sector of the inner ring live in an urban context).

During the period 1982-90, population growth was significantly higher in the periurban zone than in Paris and the suburbs: 2.4% a year in the first periurban ring and 1.6% in the secqnd ring, as against 0.8% in the suburbs and -0.1% in Paris intra muros. The main source of this strong growth in the periurban zone was migration (+1.7% in the first ring and + 1.3% in the second ring), whereas the natural increase was slightly lower than in the suburbs (where the population is younger).

In addition, the attraction exercised by the periurban zone occurs mainly to the advantage of the small communes: the smaller the population of a commune in 1982, the higher its population growth between 1982 and 1990. But although the rural areas of the Ile-de-France are particularly dynamic, their importance in the region as a whole is slight (4% of all Ile-de-France inhabi-

(3^ We use the quarter sample file, at the individual level. The file contains 2,794,997 observations (of which 1,336,241 economically active individuals), for which the various characteristics collected in the census are known. The file of economically active periurban inhabitants contains 160,424 observations (representing therefore just under 650,000 individuals). These files can of course be used to construct new files or tables at the level of communes, départements, etc.

(4) 50.5 km is the average distance from the periurban communes to Paris.

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 353

W periurban

2 ' - -NW

NW / suburbs periurban 1

Paris

NE periurban 1

NE f \ NE suburbs L гл -, > periurban 2

Figure 1. - The division of the Ile-de-France into 13 geographical sectors

tants and 30% of the periurban inhabitants in an area which nonetheless represents 60% of the region's total surface (Atlas des Franciliens, vol. 1, 1991).

One of the main geographical characteristics of the periurban zone is its low levels of population density. The average population density of communes in the Ile-de-France is 1,270 habitants per km2, but this ranges from over 22,000 in Paris to 3,800 in the suburbs and falls to 245 in the first periurban ring and to 110 in the second'5'.

There is also a considerable variation in population density within a zone from one sector to another (Figure 2 and Table 1). In the suburbs, the north-eastern sector has a relatively high population density, whereas the south-western sector is less densely populated. Overall, population densities are higher in the eastern suburbs than in the western suburbs. The same east-west contrast is found in the periurban zones: in both periurban rings, the eastern sectors are more densely populated than the western sectors.

<5) These figures are not the average densities for the different zones (which for the Ile-de-France as a whole was 887 habitants per km2 in 1990) but the averages of the communal densities.

354 B.BACCAINI

ha<<kiiiJ s ifv-gbr^^Bf CII>ssthan45 ^^f-j^j^"

^^P"

■■92-250 L. JV ■Ц250-1,614 Л ^^

^^B more than 1,614 1г%—ь—,^^рХ/

mj — -

*f

The dark lines are to distinguish the different sectors of the region: - Paris - suburbs: south-west, south-east, north-west, north east

INED 050 97

>

X

M ,{ € >

- first periurban ring (< 50km from Paris): south-west, south-east, north-west, north east - second periurban ring (> 50km from Paris): south-west, south-east, north-west, north east Source: Census 1990

Figure 2. - Population densities in the communes of the Ile-de-France in 1990

The work by Clark (1951) established that the intra-urban distribution of population in most cities follows the 'law of intra-urban density gradient', which expresses a negative exponential relationship between population density and distance from the city centre. In the case of the Ile-de-France region, this relationship is better described by a negative power function with the form D = da.b (where a is estimated equal to -2.54), where D is the density of the communes (outside Paris) and d their distance from the centre of Paris(6).

(6) The fall in density with distance from the centre is faster (than is predicted by Clark's 'law') due to the fact that we are considering not a city but an entire region, in which the densities at the periphery are sometimes very low (less than 50 inhabitants per km2). It must also be noted that an opposite relation is observed inside Paris, with the outer districts (arrondissements) having higher population densities than those of the centre.

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 355

Table 1 . - Average communal densities in the Ile-de-France

Zones of residence

Paris Suburbs

north-east north-west south-east south-west

1st periurban ring north-east north-west south-east south-west

2nd periurban ring north-east north-west south-east south-west

Total periurban Ile-de-France

Number of inhabitants per km2 Average 22 381.0 3 808.9 4 824.1 3 630.9 3 885.7 3 337.0

244.7 284.1 196.7 292.9 209.2 114.2 108.7 99.0

132.5 54.4

183.9 1 270.9

Standard deviation -

4 175.9 4 291.9 4 275.0 4 250.1 3 899.0

463.6 484.7 294.6 607.1 413.7 276.3 145.6 90.3

343.7 46.3

393.9 2 885.9

Source: Census 1990.

If instead of the total population of communes we consider only the active population at the place of residence, parameter a (estimated equal to -2.66) is found to be slightly higher in absolute value: the density of the active population falls faster with distance from Paris than does that of the inactive population. In other words, the active population is more strongly concentrated at the centre of the region than is the inactive population.

This fall in density with distance from Paris is even faster when we consider the active population at the place of work (parameter a estimated equal to -3.0), reflecting the fact that employment is more highly concentrated at the centre of the region than are the places of residence.

The structure of the housing stock in the metropolitan periphery differs from that of the Paris urban area: owner-occupied individual dwellings account for over 60% of the stock in the periurban zone (as against 18% in Paris intra muros and the suburbs). By contrast, apartment dwellings, both in rental and owner-occupation (though excluding furnished accommodation) are very under-represented in the periurban zone (less than 12% of the stock, compared with 50% in the Paris urban area). The same is true, to a lesser degree, of the stock of social housing (HLM) (less than 15% in the periurban zone as against 22% in the city and suburbs).

In all the major cities of the industrialized countries, strong correlations are observed between the age of the housing stock and that of the

356 B.BACCAINI

inhabitants, as well as between household structures and the size (and even occupancy status) of housing. The highly concentric structure of the housing stock in the Ile-de-France and the strong over-representation in the periurban zone of large housing units, particularly in the form of individual dwellings, are thus determinant in explaining the demographic structure of the periurban population.

A particular socio- demographic structure

The economically active population of the periurban zone differs from that of the Ile-de- France region as a whole by its

over-representation of the under-25s and of the 35-50 age group; by contrast, members of the 25-35 age group and the over-55s are under-represented (Figures 3a and 3b).

20 15 Percentage

10 15 20 Percentage

Figure За. - Age structure of active population resident in the Ile-de-France periurban zone in 1990

20 15 10 5 0 0 5 10 15 20 Percentage Percentage

Figure 3b. - Age structure of active population resident in the Ile-de-France region in 1990

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 357

More than half of the active population of the periurban zone are members of couples, with a working partner and children (55.6% in the first periurban ring and 50.4% in the second ring), whereas this type of household accounts for barely more than a quarter in Paris intra muros. Above all it is the greater presence of children which distinguishes the periurban households: 72.7% and 69.7% of the economically active population living in the first and second periurban rings respectively have children, compared with 59.8% for the Ile-de-France as a whole. The choice of a peripheral residential location is in fact closely linked to the life-cycle stage of the family. The stock of individual dwellings of recent construction available for home-buyers that is characteristic of the metropolitan periphery is well-suited for young families with children, and who wish to move to more spacious housing.

Manual workers are over-represented in the periurban zones, particularly in the outermost ring (28% of the economically active here, compared with 19.2% in the Ile-de-France as a whole and just 12.1% in Paris intra muros) (Table 2). Middle-level professions and clerical or sales employees are relatively more likely to be found in the first periurban ring than in the second. The higher social categories (managers(7) and members of the liberal professions) are under-represented throughout the periurban zone (accounting for barely 11% of the economically active population in the second periurban ring and 14.5% in the first periurban ring, compared with over 30% in Paris intra muros and 16.8% in the suburbs).

Table 2. - Socio-occupational structure of the economically active population in Ile-de-France, by zone of residence

Socio-occupational category

Farmers Craftsmen, tradesmen 'Cadres' Liberal professions Middle-level profess. Clerical, sales Service sector Skilled manual Unskilled manual Unemployed

Paris

0.1 5.3

27.4 3.0

19.4 17.0 6.1 7.4 4.7 9.7

Suburbs

0.2 4.4

15.7 1.1

21.4 24.0 3.8

13.4 7.1 8.8

Periurban 1

1.5 5.3

13.5 0.9

22.3 23.4 3.1

14.7 7.9 7.3

Periurban 2

3.8 6.9 9.7 1.0

18.2 20.1 3.6

17.0 11.0 8.7

Total Ile-de-France

0.4 4.7

17.8 1.5

21.0 22.4 4.2

12.4 6.8 8.9

Source: Census 1990, quarter sample.

It must be noted that this continued strong concentration of the most affluent social categories in the centre of the region and the recent nature of their migration to the periurban zones are both fairly specific to the

<7' Used for convenience to translate 'cadres', the French socio-occupational category covering senior civil servants, senior managerial staff and the higher intellectual professions.

358 B.BACCAINI

Paris region. In many European urban regions (the London region and the Dutch Randstad, for example), the proportion of the highest social categories increases with distance from the centre to the periphery (Berger 1992).

The over-representation of the least privileged social categories in the second ring of the Paris periurban was observed in the last census and prompted J.-C. Boyer (1992) to consider the possibility of residential segregation by distance, related to the concentric structure that characterizes the map of land and house prices.

The spatial distribution of the social categories within the Ile-de-France region also varies strongly by geographical sector. Irrespective of the zone considered (suburbs, first or second periurban ring), the south-west sector contains a significantly higher proportion of managers and members of the professions than any of the other sectors, whereas the north-east sector has a high proportion of manual workers (Aldeghi, Tabard, 1990). The periurban region would thus seem to possess the same sectorial structure as the Paris urban area, with the characteristics of the sectors of the agglomeration tending to be continued into the adjacent peripheral zones. Nicole Tabard (1993) has also shown that a geographical division exists within the region between two sections of the higher socio-occupational categories. This takes the form of a central location for the creative activities, specialist services and administrative functions, and a more peripheral location for the technical and research activities.

Socio-occupational categories As noted earlier, the main charac- and access to housing teristic of the housing stock in the

riurban zone compared with Paris and the suburbs is the large proportion of individual dwellings in owner-occupation (63% of the economically active population in the periurban zone are owners of an individual dwelling). However, access to the different segments of the housing stock is very unequal, depending on social category. As is to be expected, owner-occupation of an individual dwelling is much more common at the top of the social hierarchy (77% of the managers and professionals in the periurban) than among the less privileged groups (only 44% of unskilled manual workers). The housing survey conducted in 1988 found that household income in the Ile-de-France, excluding Paris, was 26% higher for households in individual dwellings than for those in apartment dwellings (for France as a whole, this disparity is only 7%). When household income is broken down by housing occupancy status, that of home-buyers is 60% higher than that of tenants. Owner-occupiers who already own their homes, who form a minority in the periurban zone, have lower average incomes than those who are in the process of buying their homes {Atlas des Franciliens, vol. 2, 1992).

A question to be examined concerns the degree to which the fact of living in the periurban zone rather than in the Paris urban area influences

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 359

the probability for the different socio-occupational categories of living in one type of housing rather than another.

To do this we have calculated for each socio-occupational category a logistic distance (LD) between the probability of living in a particular type of housing depending on whether living in the periurban zone or in the Paris urban area. For example, if we denote pi the probability for a manager of being the owner-occupier of an individual dwelling in the periurban, and p2 the same probability in the Paris urban area, LD will be:

LD(p],p2) = \ogit(pl) -\ogit(p2) = log f -^—1- log [lp 4 )

The variations in this distance can be examined on a linear scale and compared according to the different socio-occupational categories and the various types of housing (Table 3).

The disparity between residents of the periurban and the Paris urban area (measured in terms of logistic distance) are greater for the higher social categories than for the intermediate or lower socio-occupational categories, and this is true for almost all types of housing. This means that the housing structure is more sharply contrasted between the periurban and the Paris urban area for managers and professionals than for lower categories.

It is for the managers and professionals that the fact of living in the periurban increases the most (in terms of logistic distance) the likelihood of being the owner-occupier of an individual dwelling. It is also for these higher social categories (as well as for craftsmen/tradesmen) that residence in the periurban zone produces the largest fall in the likelihood of being the owner or tenant of an apartment.

For the higher socio-occupational categories, residence in the periurban zone is usually the result of choosing a particular quality of life (and in particular an individual dwelling) which is unobtainable in the city and suburbs. For the members of these categories, moving to live a long way from Paris is only worthwhile if it makes possible a different life-style, and because of their financial resources they are usually able to satisfy their desire to purchase individual dwellings. The enthusiasm for individual home ownership has several explanations: psychological (sign of social status) and environmental (garden, space), but also financial, given that loan repayments, though higher than rent, have the advantage of being concentrated in the period of working life (a large majority of home-buyers are aged 30-50), which means that they will enjoy free housing (apart from taxes) in retirement.

The choice of a periurban location is less closely linked to home- buying for the lower socio-occupational categories. Although this place of residence can indeed reflect a deliberate choice of a quality of life, it is probably often imposed by financial considerations. In order to obtain housing that is large enough and yet still affordable (even a rented apartment)

360 B.BACCAINI

О Z О I

level и

я ■a 3 о X 1 ад eh о

-a с } с •- я

Soci я X ел с с г н с с Е-* -осе. Г> с О эо- (U с б - О

3 С Е я о. я

"я тз > ind tment я D. Я "я ТЗ '> ТЭ С nent я О. Я

, , ТЗ

<^> СМ ГО го о __ го m о о ^' см го ~~ о ГО г- CN с eriurba eu ТЗ 2 Crafts, t - о о NO го О — . о го о г^ см [^ in 00 ^) см см

го — о m О 1 NO о 1 СМ о о — 1 о ON о 1 00 о

ooQ

in ГО ГО О ЧО СМ г^ о NO vO 00 ■^- oo чО r- e eriurba cu 'Cadres - ^r 00 о О ro О _; ГО ' — 1 о о 1 in о 1 о о -н I

ro in CM о ГО ^ ГО CM gglom

< о ON о 1 о , 1 Û

— см го О см — in о см ON in (^ г~ с eriurba eu 'о Си Liberal — о о см т* 1П о ro _ о >/-) го 00 CN 00 го ГО 00 ■51- см gglom < го о CN о 1 о 1 чО о о — 1 in о о ^-< 1 о . ! Q

о о^ см in Tt О ON О — 1 vd m uo ON ГО 00 ON с eriurba eu к Middle- - d о in о —

00 ~~

NO О CM — CN ON „_, CN OO ro CN о о Tř О 1 О 1 го О о 1 о 00 О 1 ON

_o "ад _ aoQ

о чО "~> m О ON 1 ' ГО — ON чО T)- ■О о 4L) С eriurba

eu я Clerica - о о ON Tt — >n <— ) го (^ о NO _)- CN о о u-> О 1 О 1 ГО о NO о 1

on m — ro nO ■ — ' OO о о 1 00 о

E "ад _ >-|

о

см On г^ о -(- NO ГО — го ON 1^ 00 о го ^ V) С eriurba Си о И Service - ON 17.

ON CN ГО см NO о ON см

го см CN 1 — ' „ CN — gglom ГО о NO о 1 CN о 1 о NO о 1 NO о о 1 ON о Q _J

о ГО чО о го On го — ON чО чО СМ ^ J — с eriurba Си ual с я Е Skilled Q о о — см т^

ГО о о чО О 1 ГО о 1

00 СМ о чО ГО см in CN CN ' — ' ON gglom о чО О 1 ч~> О t~- о 1 оо О О ■-I

о <^ чО т* 00 О CN ON (N 00 — чО 00 s о с^ см >п го чО (^ го с^ о чО см

ЧО ON 00 •/О СМ ON ГО ^ř с eriurba Си Я fci ТЗ Unskill см ON ГО — gglom <. ' — 1 о о i см о 1 Tt О NO о 1 in о чО о о Q 1-1

irea jrban г я си т> С я с я 3 per een ebetw с distž у .S2 LD: log 'q. Я ter я 3 СТ

о sus с

и Source:

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 36 1

it is necessary to move away from the city centre and the near suburbs. Residential decentralization thus tends to be unavoidable for the least well- off households seeking to improve their housing conditions.

It is also possible that home ownership for these less favoured social categories tends to occur by stages, at the end of a residence itinerary within the periurban region. If this hypothesis were true we would expect to observe a narrowing of the differences as regards home-ownership between privileged and less privileged social categories in the course of the life-cycle of individuals. The census data do not allow us to work in a longitudinal perspective, so we can only compare the proportions of individual home-owners in the various social categories by age groups. Consequently it is not possible to distinguish age effects from cohort effects.

However, we have calculated the disparity in terms of logistic distance between the probability of individual home-ownership for manual workers and the higher socio-occupational categories (managers and professionals) for three age groups (30-39, 40-49 and over-49). This distance between the social categories is found to increase with age, going from 0.28 between 30 and 39 years, to 0.49 between 40 and 49, and to 0.54 for the over-49s. The contrasts between managers and manual workers as regards ownership of an individual dwelling are thus greater for the older cohorts, aged over 40 in 1990, than for the younger ones, aged under 40 in 1990, which therefore tends to undermine our hypothesis about a 'staged' access to owner-occupation of individual dwelling in the lower social categories.

By contrast, the opposite phenomenon is observed as regards ownership of an apartment dwelling (a housing type which is relatively rare in the periurban): the disparity between managers and manual workers is reduced between 30 years and over-50, with a logistic distance equal to 0.19 for the over-49s. This is consistent with the observation of M. Berger (1990) that the higher social categories are relatively uninterested by the stock of privately-owned apartment dwellings, whereas it is with this section of the housing stock, which is less prestigious and less sought-after than other types of dwelling, that the middle and working classes are able to achieve home-ownership.

II. - The residential mobility of the economically active population of the periurban

General mobility characteristics Members of the economically ac- of the periurban population tive population living in the

urban in 1990 were less mobile in the period 1982-90 than their counterparts in the Paris urban area: 55.7% of the former had changed address during that period, compared with 58% and 57.4% of those living in Paris and the suburbs, respectively. Among the periurban population which had changed address between 1982 and

362 B.BACCAINI

1990, 45% had moved within the periurban zone, while 55% had come from outside the periurban zone (as a result of déconcentration of the Paris urban area or as arrivals from the provinces or abroad).

Of the new arrivals in the periurban zone (economically active individuals who lived outside this zone in 1982), 72% came from the Paris agglomeration (63% from a suburban commune and 9% from Paris intra muros). Slightly more than a quarter of those moving into the periurban zone had previously lived in another region of metropolitan France in 1982 (11% in a nearby region that was part of the Paris basin ZEAT(8); 15% in a more distant region) while only 1% had come from outside metropolitan France. In all, between 1982 and 1990 more than 136,000 economically active individuals left the Paris urban area and moved to a periurban commune in the Ile-de-France. These individuals represent 22% of the total active population of the Paris periurban zone.

Slightly more than half (55%) of those moving to the periurban zone from the Paris urban area were actually born in Paris or the suburbs. The remaining 45% had had more complicated itineraries, having been born outside the Ile-de-France (27% in the provinces and 18% abroad) but moved to the Paris urban area prior to 1982, before moving out to a periurban commune between 1982 and 1990. In all, more than 37,000 individuals have experienced a migratory itinerary which took them from the provinces to the Paris urban area (probably for reasons of higher education or first employment) and then into the periurban region between 1982 and 1990. More complicated itineraries, involving successive places of residence, cannot be identified using census data. Such analysis is only possible with material from event history surveys.

Much less common are residence histories in the form: 'birth in the Paris urban area => move to the provinces before 1982 => arrival in the periurban between 1982 and 1990'. Barely 20% of those moving to the periurban from the provinces between 1982 and 1990 were born in the Paris urban area, representing fewer than 10,000 economically active individuals: most of the former provincials of 1982 had actually been born in the provinces.

The strong over-representation of migration flows from the Paris urban area among the new arrivals in the periurban zone is more pronounced in the rural part of this zone than in its urban part: 79% of the new arrivals in rural communes of the periurban (coming from a commune outside the periurban) come from the Paris urban area, compared with 69% of those moving to a secondary town of the region. These secondary urban centres have consequently received a relatively large proportion of migrants who originate in the provinces (29% as against 19% of incomers to rural communes). The two migration flows into the Paris periurban zone (decentral-

(8) The Paris basin ZEAT (Zone d'étude et d'aménagement du territoire) comprises the following six regions: Champagne-Ardennes, Picardie, Haute-Normandie, Basse-Normandie, Centre and Bourgogne.

RECENT PERIURB AN GROWTH IN THE ILE-DE-FRANCE 363

ization from the Paris urban area, and arrivals from the provinces) thus differ significantly, with the rural part of the periurban zone exercising a stronger attraction on migrants from Paris than on those from the provinces. We shall see that this contrast is related to different choices as regards housing types.

The economically active population of the periurban who already lived in the periurban zone in 1982 have relatively low levels of mobility (36% have changed address): once resident in the periurban, subsequent moves are rare. However, this mobility within the periurban was slightly higher for those living in 1990 in the second periurban ring (40.4% of the active population of this zone had changed address at least once between 1982 and 1990) than among those living in the first periurban ring (33.4%). This difference is attributable to a déconcentration of the population actually within the periurban zone (9% of the active population who lived in the second periurban ring had been residents of the first periurban ring in 1982).

Among those living in the periurban for more than eight years, the proportion who had been born in the provinces is exactly the same as among those moving to the periurban since 1982 from the Paris urban area (27%).

This does not mean that centripetal movements from the periurban to the Paris urban area can be overlooked, however. Between 1982 and 1990, economically active individuals moving from a periurban commune to the Paris urban area numbered slightly more than 60,000: of these, 25% were born in the provinces and 12% were foreign-born. However, the balance (approximately 80,000) is still clearly in favour of the periurban ring.

The active population which moved from the periurban to the Paris urban area between 1982 and 1990 had a very specific age structure compared with the population already living there in 1982 and with the population which moved in the opposite direction, that is from the urban area to the periurban. Migrants from the periurban to the Paris urban area include an over-representation of young people: the under-30s account for 44.6% of these migrants, as against 21.4% of the non-migrants of the urban area and 20% of those migrating from the urban area to the periurban.

The migration flow from the periurban to the urban area also contains more women than the non-migrant population of the urban area or the flow from the agglomeration to the periurban.

The average distance from the centre of Paris to the place of residence of periurban inhabitants was 44 km in 1990. For those who already lived in the Ile-de-France in 1982, the residential migration(s) they made between these two dates took them further from the centre of Paris: their average distance from the centre of Paris in 1982 had been 38km. Those who had lived outside the region in 1982 lived on average 286 km from the capital.

Centrifugal intraregional moves, to communes further from the centre of Paris, accounted for 62% of the migrations made between 1982 and

364 B.BACCAINI

1990 by members of the economically active population living in the peri- urban in 1990.

Socio-occupational structure It is worth recalling that the data of the periurban active population concerning individuals (occupa-

and type of migratory itinerary tion, family status, etc.) are those relative to the place of residence

in 1990. They cannot be used to reconstruct the 'individual itineraries' that include the situation in 1982, for the characteristics of the migrant populations are established a posteriori. This can be problematic given that migration is frequently associated with family and professional changes. What we thus aim to identify here are not the 'determinants' of migration but the specific current characteristics of the individuals who have changed dwelling between 1982 and 1990 (and who are living at present in the periurban zone) compared with those who have not changed dwelling.

The different socio-demographic characteristics of individuals (sex, age, socio-occupational category, household type) are inter-related. By using the method of logistic regression it is possible to identify the contrasting patterns of mobility between the different socio-occupational categories while allowing for the relations between these characteristics.

We begin by examining the probability for the active population of the periurban of changing dwelling between 1982 and 1990 (Table 4), considering men and women separately, whose spatial behaviour is very different (lower mobility of women, in particular).

A clear and significant contrast exists between two main groups of population (for the same age, household type and place of residence): the higher social categories (and in particular the members of the professions) have frequently moved to their present dwelling since 1982, whereas the middle classes and, especially, manual workers have been more stable in their housing. This pattern has to be qualified depending on sex: male clerical or sales employees have a level of mobility comparable to managers, that is, relatively high, whereas their female equivalents tend to be in the categories of relatively low mobility (though less so than female manual workers).

These results can be illustrated with some examples. Consider a male, aged 35, member of the professions, married to a woman who works but without children, and who in 1990 lived in the south-east of the first periurban ring: his probability of having changed dwelling since 1982 is 0.90. For a woman presenting identical characteristics the probability of having changed dwelling will be lower (0.87). For a male manual worker, of the same age, same household type and same place of residence, this probability is 0.86 and in the case of a woman 0.81.

In these conditions it is interesting to examine whether the differences in mobility levels observed between the different socio-occupational ca-

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 365

Table 4. - Results of logistic regressions on the question: "have changed dwelling between 1982 and 1990" (active population living in periurban zone in 1990)

Constant SOC (overall effect) Craftsmen, tradesmen 'Cadres', company managers Liberal professions Middle-level professions Clerical, sales Service sector Skilled manual Unskilled manual Age group (overall effect) less than 20 20-24 25-29 30-39 40-49 50-59 60 and over Household type (overall effect) Single person Lone parent Couple (two-earner) no children Couple (single-earner) no children Couple (two-earner) with children Couple (single-earner) with children Place of residence in 1990 (overall effect) 1 st periurban ring N-W 1 st periurban ring N-E 1st periurban ring S-W 1st periurban ring S-E 2nd periurban ring N-W 2nd periurban ring N-E 2nd periurban ring S-W 2nd periurban ring S-E

Men

Parameter

1.5 *** -0.2 0.0 0.2

-0.2 0.0 0.0

-0.2 -0.3 *** -1.3 -1.1

0.0 0.0

-1.5 -2.5 -2.7 #** 0.6

-1.4 0.6 0.2 0.0

-0.1 #** -0.2 -0.2 -0.1

0.0 -0.2

0.1 -0.2 -0.1

Marginal effect (%)

81.7

-2.5 0.2 2.2

-3.1 ref

-0.1 -3.2 -4.6

-26.9 -21.7

0.0 ref

-32.3 -55.3 -59.2

7.2 -28.8

6.9 2.9

ref -1.6

-3.2 -2.5 -1.3 ref -2.7

1.1 -4.0 -0.9

- ** **#

_ **# ***

*** *** _

*#* *#* ***

*** *** *** ***

*##

*** **# *##

*** - ** **

Women

Parameter

1.0 *** 0.1 0.2 0.3 0.1 0.0 0.0

-0.2 -0.1 *** -0.6 -0.3

0.8 0.0

-1.5 -2.4 -2.7 *## 0.8 0.1 0.6 0.1 0.0

-1.4 *** -0.2 -0.2 -0.1

0.0 -0.2

0.0 -0.3 -0.1

Marginal effect (%)

73.8

1.8 3.0 5.2 1.3

ref 0.2

-3.1 -2.4

-13.4 -7.3 12.2 ref

-35.3 -52.8 -57.8

12.0 2.7

10.0 1.9

ref -32.3

-3.3 -3.5 -2.4 ref

-4.3 -0.8 -5.2 -2.8

** *** **# ***

_ *** ***

*** *** **

*** *** **#

*** *** *** -

***

#** #** ***

*** -

*** ***

The logit model of the probabilities of having changed dwelling between 1982 and 1990 is of the kind: P = 1/1 + exp(-fs) where fs = constant + (31 (SOC) + (32(age) + (33(household type) + p4(place of residence). The table reads: all things being equal, a skilled manual worker is less likely to have changed dwelling since 1982: this probability is 3.2 points below that of the reference individual. *** : significant at 1% level ** : significant at 5% level

* : significant at 10% level - : not significant Source: Census 1 990, quarter sample.

tegories in the periurban are the same in the other parts of the region, and, which amounts to the same thing, if the mobility differences between

366 B.BACCAINI

the population of the periurban and of other parts of the region are the same for all the socio-occupational categories. To answer this we have calculated for each socio-occupational category a logistic distance (LD) between pi (% of migrants in the periurban zone) and p2 (% of migrants in another part of the region). For any given category, this distance measures the 'ratio of the likelihoods' of having moved house between 1982 and 1990 depending on whether residence is in the periurban or in another zone (Table 5).

Managers and individuals in middle-level professions who lived in a periurban commune in 1990 had a lower mobility than individuals in the same categories who lived in the Paris urban area (and in particular in Paris intra muros). Clerical, sales and service sector workers living in the periurban are also less mobile than their counterparts in suburban communes but in contrast are more likely to have changed dwelling than those who lived in Paris in 1990. Manual workers, professionals and, to a lesser extent, craftsmen/tradesmen resident in the periurban zone in 1990 have a higher level of mobility than individuals in the same categories living in the Paris urban area (and in particular than those living in Paris intra muros).

These different levels of mobility correspond in fact to different types of mobility depending on the socio-occupational categories. For the individuals who had migrated since 1982 and who lived in the periurban zone in 1990, we have tried to identify the relationship between their different characteristics and the location of their place of residence in 1982 (and hence on the nature of their itinerary between the two dates) (Table 6). Unlike the procedure when longitudinal data are available, when it is possible to identify the effects of different characteristics on the direction of a migration which can occur from a given point in time /, here we try to show how these characteristics can be used to distinguish, within the population of the periurban zone, different types of 'itineraries' which have occurred.

All other things being equal as regards their other characteristics (age, household type, place of residence in 1990), migrants belonging to the higher socio-occupational categories (professionals, managers) and craftsmen/tradesmen are far more likely than individuals lower down the social hierarchy to have moved out from Paris. The former groups are also characterized by their high probability of coming from another region (indicative of the strong attraction the Paris region exercises on highly qualified individuals). Individuals in the middle-level professions and clerical/sales employees have often migrated from the suburbs, whereas the mobility of the lower social categories (manual and service sector workers) occurs mainly within the periurban. These are social categories who have been resident in this zone for long periods, and whose mobility within it is probably related to efforts to improve their housing conditions or, as will be seen later, to obtain social housing.

The socio-occupational structure of the periurban population varies greatly depending on geographical origin (Table 7). Those coming from

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE

367

eu о с я disi о '■S '5b о ^л С Я ьЬ Е 'о

я и, я с X з я X 3 u Он rbs 3 X an/Su

X 3 Per SUB, 1 3 tu Он Я X 3 Он я <0 S3 с 3 ел irbs xi з СЛ ел Я Он

SOC

о 8 о ГО о о чО 00 — 00 ■** чО £! с i о о 1 чО О 1 — н ^> !П о о ЧО ЧО оо On

чО С adesmt ~, с eu ел Craft in <N en чО res

1 'Cad S о S о о о г- in о о s о 1 s о 1

1П 1П

m ^- ON in m

о чО in r- (N in a Libei On in ON ^^ чО sions profes; > Л Л S 8 о о о 1 го О о ГО ON 1П ON ON On in ON in ел

I

Cleri (N О О 1 in о о 1 ГО о о о го m ГО ^> in о чО in о с с о о о о ГО чО 1П ГО in C3N in оо о in in

с u ел eu и Servi

я 3 с 1 •о

1 Skill S о (N О О in о Tf <N in — in in in On nual я E ■a «i ill.

1 Unsk 2 1 2 о 1 S о 1 t\ >o о in Total

CU "EL E ce Й 3 er

ON ел 3 С tu и ?■> Soun

368 B.BACCAINI

в ti u ter H я я

e-Fr. •о le- -а uts о II 'So 'С и cd •е CU II с igi О X! С/3 II О ris я и с ад

Or

с р 1 и 5Ž н 1 с и н 1 с и с и р Wo и

_ 1 чО _ 1 о I чО о 1 _ 1 fN О 1 _ СО 1 Os <N 1 Constant

* # * * * * * * * * * * SOC (overall effect) * * о * * # r- o " -H o 1 * * * CO о * * fN О 1 1 _ о 1 # * * чО О * * ■«fr О Craftsmen, tradesmen * * * со о * * # О * * * 00 о 1 * * * О 1 1 О о * * * __ о 1 * * * — * * # О и 'Cadres', company manag * * со О * * * о * * * in о 1 * * * со О 1 * fN О 1 1 _^ О 1 * # # — . * * # со _ Liberal professions * * * tN О ♦ * * СО о * * * СО о 1 1 о о 1

о о о * * * ■«fr о 1 о о Middle-level professions ref о о <и

о о о о о о <+ч JJ о о 'ч- <и

о о а

о о о о Clerical, sales

i о 1 _н о 1 о о * * * о 1 * СО о 1 1 о о * * * ЧО о Service sector * # * tN О * * * fN О * * * СО о * * ЧО О

* * со О 1 * * * см о 1 1 tN О 1 # * * о 1 Skilled manual

i о о * * # о 1 * * * о * * * о * * Os о 1 * * * •л о 1 # * * 00 о 1 * <N о 1 Unskilled manual

* * * * * * * * * * * * *

* <N О 1 1 о 1 ^_ о 1 _ о 1 о 1 1 _^ о 1 * о 1 * ГО о 1

Age group

(overall effecl less than 20 # * * о 1

* * * со О 1 # * * чо О * * *

* * * см о i * * * со О 1 * * * СО о * * * со со О •х- * * Os О 1 * * * О 1 * * «о — - i * * со — 1 20-24 о * * * г- о 1 * * * о 1 * * * чО о 1 * * * Г- о 1 25-29

ref о

о о о Ч-| о о 'ч- *"■ о о о о ч- о о о a о о ч- 14 о о 30-39

i о о

о 1 * * # со О * * * tN О * * * _^ о 1 * о 1 * # * со О 1 * о 1 40-49

i о

о 1 о о * * * о * * * 1Г) о 1

о 1 * * * fN о 1 * tN о 1 * fN о 1 50-59

i о 1 о * * * о * * * >п о 1 fN О 1 * fN О 1 1 со О * * * чо О 60 and over

* * * * * * * * * * * * ffect си Household type (overall * * о i * * * о 1 # * * со О 1

о о * * * о — 1 * * * ОО О 1 1

О * * fN О Single person * * * Os О 1 * со О 1 * * # >п о * * ■<fr о * * # о 1 * * *

* * * о i * * * fN О 1 1 О о * о со О 1 * * * СО СО о 1 1 о 1 1 о о Lone parent о 1 * fN О * о

Id Couple (two-earner) no ch

i <N О 1 * О 1 1 _^ о * tN о 1

* * # о 1 * * * со О 1 * ■4- о 1 CN о с d ldi j=

ref

о о Й о о Ч-ч О о (1 «)

О о <и

о о «4-1 о о •ч- и о о Л)

О о п «í ■а 'г, f )

Couple (single-earner) no Couple (two-earner) with

1 о * * * __ о 1 1 о о * # fN о 1 1 __ о 1 * * * __ о 1 1 см о * о ren ■а hil о Couple (single-earner) wit

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE

369

ТЗ <U JBUI ■с S В Parai

e-Fr.

43 le- u T3 3 о II 'Бр о с -e 3 (U II с 'ад Ori iburbs 3 on II с 'ад Ô ris II с rigi

О

с u E 1 E 1 с <u с <o Wom с u с E 1 с

■jf * * ■ft * * * * ■ft * * * * # * * * * * * * ffect) CU "3 CU О о Hni и 73 resi о ti я Su * * # CN * # * * * * ЧО d i # * * d i * * * 0.6 * d

l d i о d N-W M с ban 3 'C D * * _ d i * _ d l * * * — d i * * * _ d i * * * 0.1 * * — ' d

* * * d * * d N-E # * d * * * О * * * d i * * # го О 1 * * * 0.2 * * * сч о 1 о i о о S-W ref о о Ч-ч Vh О С ч- о d ч-< О О

ч- U 0.0 ч- о d <ú

о о ref о d S-E ад ад at с

с с

>- ban 3 'С О 1- ban 3 'С D —

i- ban 3 'С

. о * # * — * * * * * * о * * d i * * * 0.4 * * * d l CM d i * * d 1 Ž с с rbai u a T3 с fN # * On — * * * О * * * * * * Т;|- d * 0.5 * * * d

i * d * * d N-E Dt С С rbai О ТЗ с CN * * * oi * * * (N 1 О d d i i d i l о d l

d i _ d % su с С rbai <u * * * d * * * 00 d * * * d * ЧО d * * * 0.9 i * * * 00 d i i о d i о d S-E ад с с rbai ni. u D. (X

T3 с (N ТЗ с (N

~~ ce ifican с ад !" * * u >■

i te ce lific ад * *

с ce CJ iignifi о 1 ~Б Cl) ^™ О te G ce liflC с '(Л *

cl ce t; ее 3 о* 3 с и Li a о

370 B.BACCAINI

the Paris urban area have a socio-occupational structure very similar to that of the urban area as a whole, though with a slight over-representation of the middle-level professions, clerical or sales employees and skilled workers; managers, however, are slightly under-represented among migrants to the periurban relative to their importance in the population of the Paris urban area. As F. Beaucire (1992) has established for the earlier periods, the successive waves of arrivals in the periurban zone tend to "give the periurban ring the same social structure as the urban area"(9).

Table 7. - Socio-occupational structure of active population of the periurban in 1990 according to place of residence in 1982

SOC in 1990

Farmers Craftsmen, tradesmen 'Cadres' Liberal professions Middle-level profess. Clerical, sales Service sector Skilled manual Unskilled manual

Total

Not moving (1)

3.1 6.5

11.8 0.9

21.4 23.8 3.7

17.7 10.9

100.0

Immigrants from Paris urban area

0.4 6.1

16.0 1.5

26.7 26.4 2.7

14.8 5.5

100.0

'lace of residence in 1990 Periurban

Immigrants from Paris

basin 2.0 5.5

13.3 0.7

24.8 24.9 3.6

16.3 8.8

100.0 (1) Individuals already resident in the periurban in 1982. Source: Census 1990, quarter sample

Immigrants other region

1.0 3.0

21.8 0.9

27.9 24.9 3.5

11.0 6.0

100.0

Immigrants external

0.5 2.4 9.2 1.2

16.5 34.7 8.0

13.8 13.6

100.0

Paris urban area

0.1 5.1

20.3 1.8

23.1 24.5 4.8

13.2 7.1

100.0

The socio-occupational structure of migrants to the Paris periurban zone from elsewhere in France varies significantly depending on whether they have come from a region close to the Ile-de-France (in the Paris basin ZEAT) or from a more distant region. Manual workers and craftsmen/tradesmen are disproportionately present in the flow of arrivals from the Paris basin, while the flows from the other regions contain more managers and individuals in middle-level professions. The attraction of the Ile-de-France periurban zone for the economically active in the provinces is thus different depending on distance, with short-distance migrations involving mainly manual workers, while highly qualified individuals predominate in long-distance migrations.

i9) The socio-professional structure of the 1968-75 wave of immigration by 'Parisians' to the metropolitan periphery differed significantly from that of the urban area, with a strong over-representation of manual workers but an under-representation of clerical and sales employees, middle-level professions and managers (Beaucire, 1992).

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 37 1

Let us now examine in greater detail the flows between the periurban zone and the Paris urban area (Table 8).

The balance of the flows is very strongly in favour of the periurban zone for all socio-occupational categories. An examination of the socio- occupational structure of this surplus shows that in quantitative terms the middle-level professions and clerical or sales employees have a major role in the decentralization of population from the Paris urban area (these categories are also proportionately more present in the surplus than in the economically active population of the two zones).

The qualitative effects of these migration flows between the Paris urban area and the periurban zone are relatively more complex. The flows that run from the urban area to the periurban have a selective effect in favour of four socio-occupational categories: managers, professionals, middle- level professions and clerical/sales employees. For these groups the index of differential immigration is positive, indicating that the arrivals of these categories in the periurban was large relative to the periurban population which did not move. In contrast, farmers, unskilled workers, service sector workers and craftsmen/tradesmen are relatively under-represented among arrivals in the periurban from the Paris urban area. Migrations in the opposite direction, from the periurban to the Paris urban area operate mainly in favour of managers, the middle-level professions and clerical/sales employees. For these three socio-occupational categories, as well as for service sector workers, the index of differential emigration is above or equal to the index of differential immigration, which is also reflected in a low 'effectiveness' of the migration flows for these categories. In contrast, for craftsmen/tradesmen and professionals, more than 50% of the total flow (sum of the urban area-periurban flow and the periurban-urban area flow) had a real 'effectiveness'.

The socio-occupational structure of the new arrivals in the periurban (individuals living outside the periurban in 1982, either in the Paris urban area or outside the Ile-de-France) varies clearly from one geographical sector to another. So as to identify the socio-occupation selection effects of the migrations towards the periurban, we have defined a communal-level indicator of the social distribution of arrivals in the periurban as the ratio of the number of arrivals in a 'low' social category (clerical, sales and service sector workers) to those in a 'high' social category (managers, members of the professions).

This indicator has an average value of 8.6 but is above three in half the communes of the periurban zone, signifying that the new arrivals from the low social categories are three times more numerous than those of the higher social categories. A relatively strong selection operating in favour of managers and members of the professions (an indicator less than unity) is observed in two main sectors (Figure 4): first, a group of communes in the west-south-west, second, a group of communes centred on Fontainebleau, in the south of the region. These communes, which are already char-

372 B.BACCAINI

area Hon s -S a

а ■^

sa. ^

=1 ban

'Hon -S •2

а il

LOf ntial ition 1

к p jy i? "c fcë 'S w

"3

<U

— с

ч- .S .2 x 1

s! ^

Ind diffe immi (

.M ^ ^^ 8 ^

^

w w

otal

ice %oft balar

о я "я И

*2 с

с я

d E M

g 3 щ £ "~

Й W

û< ся

С Í-

Я С xi а. 2 х

•£ я

III -

Е <ё -с ^ Z

1990 с

on

о CN •о CN 42

On CN О | 00 о 1 го CN U0 in о О 1 — О 1 го CN m

40 00 — CN ГО ^+ —

о чО 00 чО *-* О irmers

u- ГО >п 00 CN CN чО Г*- [ — о 00 го 00 esmen -о я raftsmen, tr;

и Ci ^ř d — о ON ^« со — CN r- On OO ЧО О О ЧО CN О CN CN ■О ЧО ГО anagers В Cadres' , со. y 00 о d CN О

го CN >n ON

g ro _ °o CN On CN CN d --. о о ( — со СО чО CN ЧО г~ CN оо ""

CN 00 ON

CN ON 00 _ CN ГО ЧО CN О '35

beral profe; «л U0

CN 00 С*") ГО rj- CN On ofess. CL iddle-level 2 CN •4- CN d -H о ■^- *n со CN ■^- CN ON !2 00 00 ГО ГО ГО oo 00 >n 00 in lerical, sale:

U 00 \)-' 42 О d CN О 1 ЧО CN ON CN m _- чО О CN 00 CN tJ" ro 25 oo ON ;rvice secto CN "П 42

d 1 о о CN ^f in К o< d i ro о 1 r^ CN О

ЧО ЧО

oo чО ■^" -m ~" О CN 00 о 00 ~^ 00 ЧО "я cilled manu

oo on g CN ON CN CN in On о ci Cî 00 к ^o о 100. 42 •о \}" On 42 CN CN !o 00 42 42 ЧО 42

^4- ^}- 00 ЧО CN "я 3 с nskilled ma P CN ^ 42 42 263 "g

(Sx/S)]

it, ^

0П on ' I

§ s II

TJ ей о Jo я с défi .22 x, this index и О on я i

i и Т) Ix = 0 then о II '•Б с u •£ on on II Mu

II '•a с 1 о II X on 0П X El

on ■ Ш II T3 ? О fol я ТЗ S с défi M x, this index (j on я

i и •o Ex = 0 ther

Яз

:hen de = 0 Ex/E = Sx/: Яз

и •o с D .g О II X 00 4-1

>am] m

1990, quarti 3 >мдге:

Censi

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 373

acterized by a strong presence of 'higher' social categories, are experiencing a process of 'embourgeoisement' as a result of migration. By contrast, in the communes of the north-eastern periurban, migrants in the lower social categories far outnumber those of the higher social categories, thereby reinforcing the existing social structure of this sector. However, the map shows that within the different sectors there is a complex imbrication between communes that tend to attract the higher social categories, and communes which on the contrary tend to attract the lower social categories.

number of in-migrants (manual + clerical/sales + service sector) indicator i : number of in-migrants (managers + professionals) Source- census 1990

Figure 4. - Index of social distribution of arrivals in the Paris periurban between 1982 and 1990

Migration into the periurban communes in the period 1982-90 thus had the effect of reinforcing the existing socio-spatial contrasts, with the predominantly working-class and white-collar areas (north and east) attracting mainly these categories, whereas the more 'bourgeois' areas (chiefly in the south-west) attracted proportionately more managers and professionals. This is consistent with the findings of I. Aldéghi and N. Tabard (1990) who showed that for the period 1975-82 increasing social segregation was,

374 B.BACCAINI

in statistical terms, the most significant social transformation affecting the Ile-de-France. This finding has been confirmed for the period 1982-90 (Chenu, Tabard, 1993).

Throughout the north-east of the periurban zone, half of all communes received more than four times more migrants belonging to the lower social categories than those from the higher social categories. This is especially marked in the most peripheral part of the periurban ring (where the ratio between the two groups of in-migrants exceeds sixteen in a quarter of the communes). By contrast, in the south-western sector of the first periurban ring the number of newcomers from the higher social categories exceeds that from the lower social categories in more than a third of communes.

The relative importance of newcomers from the 'lower' social categories compared with migrants from the 'higher' social categories also tends to increase with distance from Paris (there is a positive correlation, significant at the 1% level, between the distance of the commune from the centre of Paris and the indicator of the social distribution of the migrants).

Behaviour by age and household type

As is usually observed, levels of mobility among the economically active population of the periurban increase up to the age of 25-29

years, then decrease. The geographical origins of migrants to the metropolitan periphery vary considerably according to their age (Table 9) The 20-29 age group is particularly important among migrants from outside the Ile-de-France region (with the 20-24 age group being more likely to come from a nearby region, within the Paris basin, whereas the 25-29 age group tend to come from more distant regions). This age group thus makes an important contribution to the migration flow from outside the Ile-de- France: economically active members of the 20-29 age group account for 25% of migrants to the periurban zone but 41% of those coming from a different region. Those involved in the déconcentration of the Paris urban

Table 9. - Place of previous residence (in 1982) of the economically active population migrating to the periurban between 1982 and 1990, by age

Age in 1990

15-19 20-24 25-29 30-39 40-49 50-59 60-69 Total

Place of residence in 1982

Paris

7.6 3.8 6.8

10.6 10.4 12.0 22.7 9.5

Suburbs

66.2 46.8 48.0 67.8 68.6 66.9 68.3 62.9

Paris basin region 13.2 21.5 17.8 8.8 8.8 9.9 7.8

11.4

Other region 12.1 24.8 25.5 12.0 11.5 10.6 10.9 15.1

Outside France

1.0 3.2 1.8 0.8 0.7 0.6 0.3 1.1

Total

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Source: Census 1990, quarter sample.

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 375

area are, by contrast, predominantly of the over-30s (and under-20s), being drawn mainly from the suburbs between 30 and 59 years but from Paris intra muros when older.

All things being equal regarding their characteristics in 1990, migrants in the 30-39 age group still have a high probability of coming from the Paris urban area (Table 6). By contrast, the characteristics of migrants from the provinces or from abroad are significantly modified, in relation to the crude results. For example, including the 1990 place of residence has the effect of considerably reducing the probability for young people of having come from a region outside the Ile-de-France. Newcomers to the Ile-de-France whose place of residence is in the periurban have a high probability of living in the second periurban ring, and, in addition, relatively more members of the active population under-30 live in this second ring than those aged 40-49. The effects of geographical origin, choice of place of residence in the periurban, and of age, are thus extensively imbricated. For example, when the residence sector in 1990 is not included in the model, the probability of having come from the provinces is significantly higher between the ages of 20 and 30 than at higher ages.

The age structure of the migrants to the periurban varies from one geographical sector to another, although less clearly than the socio occupational structure. As before, we have constructed a communal-level indicator of the distribution by age of migrants to the periurban as the ratio between the number of economically active migrants under-35 and the number over-35: the higher the value of this indicator, the more that particular commune has exercised a disproportionate attraction on the young.

The indicator has an average value of 1.8 but is below 0.9 in half of the periurban communes. The east of the periurban (except for the communes of the south-east closest to Paris) attracts a mainly young population, whereas migrants aged over-35 outnumber the younger migrants in many communes of the western periurban, with especially low values of the indicator in a group of communes adjacent to the Paris urban area (Figure 5).

Within the periurban ring, the further from Paris, the stronger the selection of migrations operates in favour of young people under 35 years (positive correlation, significant at the 1% level, between distance from Paris and the indicator of age distribution of the migrants). This relationship is especially strong in the south of the region.

Analysis of the effects of family situation on spatial decisions is problematic for several reasons, given that we know this situation only for 1990. For example, an individual who was living in a couple in 1990 could have been living alone when he or she migrated. Similarly a couple who had children in 1990 might not have had any at the time of the migration.

Individuals living alone in 1990, men or women, have a high probability of having changed dwelling since 1982, all other things being equal regarding age (Table 4). High levels of mobility are also observed in the double-income families with no children. This may indicate mobility by

376 B.BACCAINI

indicator i : number of in-migrants < 35 years number of in-migrants > 35 years Source, census 1990

Figure 5. - Index of age distribution of migrants to the Paris periurban between 1982 and 1990

the couple which was already formed when the migration occurred, or mobility of two single individuals, who formed a couple after or at the time of the migration. By contrast, couples in which only one member is economically active but who have children have a low probability of having changed dwelling between 1982 and 1990.

All other things being equal, migrants from the suburbs are often couples with children (and in particular two-earner couples). Lone parent families are noteworthy for their high probability of having migrated within the periurban zone (Table 6). The high proportion of individuals from the provinces among the lone individuals (13.4% of these individuals, as against 8.2% of the economically active population of the periurban) is no longer apparent when the other characteristics are included, given the important effects of age and geographical sector of residence.

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 377

Migration itineraries and housing choice

The decision to live in the metropolitan periphery is closely linked to housing strategies. In particular, the demand for private

ownership of individual dwellings is easier to satisfy in the periphery than in Paris and the suburbs. However, the housing choices of households also vary according to their migration histories.

Economically active individuals occupying rented accommodation in 1990 have experienced higher residential mobility than home-owners. This contrast is not specific to the metropolitan periphery, since the fact of becoming a home-owner (which may be a reason for mobility) is in general responsible for residential stability (Table 10). For comparable housing occupancy status, levels of mobility are also higher for those living in apartments compared with individual dwellings, except in the social housing sector (HLM) where residents of apartment blocks (the large majority in this part of the housing stock) had lower mobility than those living in individual dwellings.

Table 10. - Types of housing of periurban residents in 1990, by their place of residence in 1982

Owner-occ. individual Owner-occ. apartment Tenant individual Tenant apartment Social: individual Social: apartment Tenant furnished ace. Housed at no charge Total

migrant

73.1 3.3 4.8 3.1 0.7

10.6 0.2 4.3

100.0 Source: Census 1990, quarter sample

Migrant intra-

periurban

44.9 4.4 9.3

10.7 1.9

22.7 0.6 5.6

100.0

Migrant Paris urban area 75.0 2.9 6.2 4.8 1.0 6.2 0.4 3.5

100.0

Migrant Paris basin

32.6 3.6

16.6 16.1 1.8

18.4 1.4 9.6

100.0

Migrant distant region

31.5 3.4

13.6 18.7 1.9

18.9 1.6

10.5 100.0

Migrant from

abroad

37.2 2.2 5.1

11.9 2.2

33.2 2.0 6.3

100.0

Total

63.1 3.5 7.0 6.5 1.1

13.4 0.5 4.9

100.0

The economically active individuals who have changed residence within the periurban zone since 1982 are relatively less likely to be owner-occupiers in 1990. By contrast, they are over-represented in the social housing (HLM) sector, probably because such housing is easier to obtain for individuals who have lived for a long time in the same geographical sector. It is also observed that the lower social categories are over-represented among these long-term residents of the periurban, and it is these categories which have priority for the allocation of this type of housing (manual, clerical, sales and service sector workers account for 70% of the occupants of apartment blocks in the social housing sector in the Paris periurban).

By contrast, a majority of those who have moved into the metropolitan periphery from Paris and the suburbs live in privately-owned individual dwellings (75% of them), therefore supporting the hypothesis that decisions

378 B.BACCAINI

to move out of the Paris urban area are related to a desire to purchase this type of housing.

A large proportion of the migrants who come from outside the Ile- de-France go into housing in the private rental sector. Migrants from the neighbouring regions are more likely to go into rented individual dwellings while those from regions further away tend to go into rented apartment dwellings. Migrants from outside the Ile-de-France are also often housed at no charge (by a member of their family or in housing provided by an employer). These housing preferences account for the over-representation of migrants from the provinces in the secondary urban centres of the peri- urban zone, whereas for opposite reasons the dominant preference of migrants from the Paris urban area for individual dwellings explains their frequent choice of residential location in a rural commune.

The different segments of the housing stock can thus be seen to correspond to highly specific 'migration profiles'.

However, although a link exists between the type of housing occupied in 1990 and the migration itinerary, housing type is also related to an individual's socio-occupational category. It is worth examining the effects on the migration itinerary of the interaction between socio-occupational category and the choice of a particular type of housing.

We have proceeded by making a two-by-two comparison of four types of migration itineraries: first, between non-migrants and migrants inside the periurban zone (two groups already resident in the periurban in 1982); second, between those who came from Paris and those who came from the provinces (two groups of migrants to the periurban zone). For this we use a logistic scale in order to compare proportions that vary over a wide range.

Begin with the case of economically active individuals who already lived in the periurban zone in 1982 (Table lia). The probability that they will have changed place of residence (within the periurban) between 1982 and 1990 is more influenced by the type of housing occupied in 1990 (and thus on their housing preferences) than by their socio-occupational category: the logistic contrasts produced by housing type are actually greater than those associated with socio-occupational category. Whatever their socio-occupational category, owners of individual dwellings are always the least likely to have migrated. The fact of being an owner-occupier, particularly of an individual dwelling, is responsible for greater residential stability and thus a lower probability of being a migrant. The highest probabilities of having changed dwelling within the periurban zone since 1982 are, depending on socio-occupational category, among the tenants of an apartment (professionals, clerical/sales, service sector, skilled workers), the tenants of furnished rented accommodation (unskilled workers, managers) and the tenants of individual dwellings in the social housing sector (craftsmen/tradesmen).

The greatest variation in the probability of having migrated by the type of housing occupied in 1990 is for the higher and middle social ca-

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 379

Table 1 la. - Economically active individuals living in the periurban in 1982 and 1990. Comparison of non migrants and migrants by SOC and housing type in 1990

soc

Craftsmen, tradesmen 'Cadres'

Liberal professions Middle-level profess. Clerical, sales Service sector Skilled manual Unskilled manual

LDmax(2)

Comparison of migrants and non-migrants on a logistic scale (1)

Owner-occ. individual

-0.52 -0.61 -0.46 -0.42 -0.42 -0.48 -0.35 -0.42

0.26

Owner-occ. apartment

-0.31 -0.24 -0.40 -0.05 -0.06 -0.03 -0.12 -0.20

0.37

Tenant individ.

-0.02 0.02

-0.03 0.21 0.14

-0.11 0.09 0.00

0.32 (1) logit (p) = log (p/(l-p)) with p = % of migrants.

Tenant apartmt.

0.05 0.20 0.46 0.41 0.46 0.12 0.24 0.08

0.41

(2) LD max = logistic distance between the two extreme values. Source: Census 1990, quarter sample.

Social: individ.

0.24 0.26 -

0.29 0.29 0.03 0.16 0.04

0.26

Social: apartmt.

0.01 0.15 - 0.16 0.17

-0.02 0.05 0.00

0.18

Tenant

ace. 0.06 0.56 - 0.26 0.31

-0.04 0.10 0.21

0.60

Housed

charge -0.34 -0.19

-0.08 -0.05 -0.21 -0.18 -0.15

0.28

LDmax (2)

0.76 1.16 0.91 0.83 0.88 0.60 0.59 0.63

Table lib.- Housing types and socio-occupational categories in the periurban ZONE: COMPARISON OF MIGRANTS FROM THE PARIS URBAN AREA AND MIGRANTS FROM

THE PROVINCES

SOC

Craftsmen, tradesmen 'Cadres'

Liberal professions Middle-level profess. Clerical, sales Service sector Skilled manual Unskilled manual

Logistic distance between former Parisians and former provincials ( 1 ) Owner-

occ. individual

0.5 0.7 0.4 0.9 0.9 0.8 0.8 0.8

Owner- occ.

apartment -0.1 -0.2 -0.3 -0.1

0.0 -0.1

0.0 0.4

Tenant individual

0.0 -0.5 -0.3 -0.5 -0.4 -0.3 -0.4 -0.4

Tenant apartment

-0.4 -0.7 -0.3 -0.8 -0.6 -0.5 -0.7 -0.4

Social: individual

-0.5 -0.4

-0.3 -0.2 -0.4 -0.3 -0.2

Social: apartment

0.0 -0.7 -0.5 -0.6 -0.6 -0.4 -0.6 -0.4

Tenant furnished

ace. -0.2 -0.6 -1.0 -0.7 -0.7 -0.5 -0.7 -0.6

Housed at no charge

-0.3 -0.4 -0.3 -0.5 -0.7 -0.4 -0.4 -0.4

(1) let plx = % of owners of individual dwellings among the former Parisians of SOC x p2x = % of owners of individual dwellings among the former provincials of SOCx

then LD = logit(plx) - Iogit(p2x) = log(plx/(100-plx)) - Iog(p2x/(100 - p2x)). Source. Census 1990, quarter sample.

380 B.BACCAINI

tegories. For example, a manager who owns an individual dwelling is far less likely to have migrated than an unskilled worker living in the same type of housing in 1990. On the other hand, a manager who is the tenant of an apartment has a higher probability of having migrated than an unskilled worker living in the same type of housing.

Now let us consider those moving into the periurban zone, either from the Paris urban area or from a region in the provinces. The relationship between migration histories and type of housing occupied in 1990 must now be considered from the opposite point of view. Whereas earlier it could be assumed that the choice of a particular type of housing influenced the propensity to migrate within the periurban, now it has to be assumed that the geographical origin of individuals (Paris urban area or provinces) influences the choice of type of housing. Consequently, for each socio-professional category and for each type of housing we have calculated a logistic distance between the probability of having obtained this type of housing depending on whether one has migrated from Paris or from the provinces (Table lib).

Whatever their socio-occupational category, migrants from the Paris urban area are more likely than migrants from the provinces to be owners of an individual dwelling: buying a home, especially an individual dwelling, is often a key aim of individuals when moving out of Paris or the suburbs. In addition, the origin (place of birth) of these migrants to the periurban from the Paris urban area between 1982 and 1990 is not associated with any difference in the proportions who are owners of individual dwellings: for all the socio-occupational categories combined, 75% own or are in the process of buying an individual dwelling, regardless of whether they were born in the provinces (and had thus already migrated to the Paris urban area before moving to the periurban) or in the urban area.

The difference between former Parisians and former provincials ('former'

referring here to the situation in 1982, and not to their place of birth) is greater for the lower and middle socio-occupational categories than for managers, professionals and craftsmen/tradesmen. In other words, between former Parisians and former provincials the probability of owning an individual dwelling is more contrasted among manual workers, clerical or sales employees and the middle-level professions, than among the higher categories. Less than 30% of the members of the lower socio-occupational categories can aspire to this section of the housing stock when they migrate from the provinces, compared with more than 45% of managers and professionals.

Conclusion

The economically active population of the Paris periurban zone forms a highly specific sub-population in the Ile-de-France, as is clear from its socio-demographic structures and spatial behaviour.

RECENT PERIURB AN GROWTH IN THE ILE-DE-FRANCE 381

A majority in this population is formed by young adults (30-50 years) among which families with children are over-represented. The types of housing available in the periurban zone match the needs of this category of households. Overall, the periurban zone is socially less privileged than the Paris urban area, though it contains marked geographical disparities which tend to reproduce the sectorial pattern observed in Paris and the suburbs.

The structure of the housing stock in the periurban zone makes it especially attractive for individuals wishing to purchase individual dwellings. This is especially clear for the higher socio-occupational categories (managers, professionals), for whom a periurban location seems to be the most frequently directly linked to the choice of this type of housing. In the lower categories, where financial resources are more limited, purchase of individual dwellings is less frequent (even when it is an ambition), though the periurban zone does make possible an improvement in housing conditions (surface area, quality of surroundings) and sometimes own-home purchase (perhaps in apartment blocks), which would be hard to achieve in the centre of the urban area.

The rapid growth of the communes in the periurban zone results mainly from the decentralization of population from the Paris region, and owes much less to migration from the provinces or from abroad. Examining the birth place of individuals does, however, reveal the existence of itineraries that are often more complex, detailed analysis of which is only possible using event history material. For example, almost half of the migrations from the Paris urban area to the periurban between 1982 and 1990 were made by individuals who had been born in the provinces and who had thus merely had a temporary stay in Paris or the suburbs before moving to live in a periurban commune. Migrations in the other direction, from the periurban to the Paris urban area between 1982 and 1990, involved only half as many economically active individuals, with moreover a socio- demographic structure that was very different, closer to that of migrants from the provinces into the Paris urban area (young people, living alone).

Detailed analysis of the migratory itineraries and the characteristics of the different types of migrants (or non-migrants) living in the periurban enables us to identify a number of principal types of residential strategies. Purchase of an individual dwelling emerges as the main aim for individuals in the higher social categories who move out of the Paris urban area. Consequently, once resident in the periurban zone these individuals have low mobility. Individuals in the lower social categories are less often involved in home purchase and more likely to live in apartment dwellings (whether or not in the social housing sector), and as a result tend to experience longer and more complex itineraries within the periurban zone.

Migration from Paris to the periurban zone reinforces the existing socio-spatial inequalities. As a result of these movements, the higher social categories increase their presence in the sectors where they are already

382 B.BACCAINI

strongly represented, whereas the lower social categories tend to move to the areas where they already predominate.

Locational decisions are always the result of trade-offs between the advantages and disadvantages associated with residence in a particular place. The advantages of the periurban zone are a greater chance of home-ownership and of living in a larger apartment or in an individual dwelling, as well as offering a better quality of life in terms of natural surroundings. On the other hand, a peripheral location has the disadvantage of placing economically active individuals further from the main areas of employment which, although subject to some decentralization, remain strongly concentrated in the Paris urban area.

Several studies (Baccaïni, 1996a; Berger, 1996) have established the relatively low importance that individuals attach to the place of work when making residential decisions, the balancing of home-work locations usually being achieved after the residential migration. The main reasons for moving within the periurban zone are the desire for more spacious and more comfortable housing, and the possibility of owner occupation and an individual dwelling, whereas the factor of proximity to work counts for little, particularly for the households in which two earners is increasingly the norm(l0).

Brigitte Baccaïni

REFERENCES

Aldeghi I., Tabard N., (1990), Transformation socio-professionnelle des communes d'Ile-de- France entre 1975 et 1982, CREDOC, Collection des rapports, n° 80, 70 p.

Atlas des Franciliens, (1991-1992), tome 1 : «Population et logements»; tome 2: «Âges, emplois, modes de vie », IAURIF-INSEE.

Baccaïni В., (1996а), « L'évolution récente des navettes en Île-de-France», L'Espace géographique, 1, pp. 37-52.

Baccaïni В., (1996b), « Les trajets domicile-travail en Île-de-France : les contrastes entre catégories socioprofessionnelles», Économie et Statistique, n° 294-295, pp. 109-126.

Baccaïni В., (1997), « Commuting and residential strategies in the Île-de-France. Individual behaviour and spatial constraints», Environment and Planning A, vol. 29, pp. 1801- 1829.

Beaucire F., (1992), «Migrations résidentielles et ségrégation sociale dans la couronne périur- baine d'Île-de-France entre 1968 et 1982», Villes en parallèle, n° 19, pp. 47-73.

Benoît Ph., Benoît J.-M., Bellanger F., Marloff В., (1993), Paris 1995, le grand desserrement, éd. Romillat, 301 p.

Berger M., Guillon M.^Rhein C., (1989), « Évolution socio-démographique et parc de logements», dans L'Île-de-France en mouvement, collection Reclus modes d'emploi, n° 16, pp. 256-279.

Berger M., (1990), «.Les périurbains d'Île-de-France : stratégies de localisation», in : Stratégies résidentielles, Congrès et Colloques n° 2, INED, pp. 369-383.

Berger M., (1991a), «L'urbanité des périurbains d'Île-de-France; de la diversité à la ségrégation » Les Annales de la Recherche Urbaine, n° 50.

Berger M., (1991b), «Périurbains et ex-urbains d'Île-de-France», Géographie ̂ Sociale, n° 11. Berger M., (1992a), «Division sociale de l'espace et parc de logements en Île-de-France»,

in : L'Ile de France et la recherche urbaine, Datar-Strates, pp. 11-43.

(l0) The second part of this study examines the consequences of these residential choices on the daily journeys of the economically active populations and explores the relations which exist between residential itineraries and commuting.

RECENT PERIURBAN GROWTH IN THE ILE-DE-FRANCE 383

Berger M., (1992b), «.Le rôle du logement dans les stratégies de localisation : l'exemple des périurbains d'Île-de-France», Villes en parallèle, n° 19, pp. 105-112.

Berger M., Saint-Gérand T., (1996), « Périurbanisation et métropolisation en Île-de-France dans les années 1980: les ménages et les aménageurs», Xe colloque national de démographie, Bordeaux.

Boyer J.-C., (1988), «Péri-urbanisation et migrations de population active en Île-de-France», Espace, Populations, Sociétés, n° 3, pp. 495-500.

Boyer J.-C, (1992), «Des espaces périurbains en évolution», Villes en parallèle, n° 19, pp. 11-27. Brun j, Fagnani J., (1994), «Lifestyles and locational choices. Trade-offs and compromises :

a case-study of middle-class couples living in the Île-de-France region », Urban Studies, vol 31, 6, pp. 921-934.

Camstra R , ( 1 994), Household relocation and commuting distance in a gender perspective, Amsterdam : Department of Planning and Demography, University of Amsterdam, 20 p.

Camstra R., (1996), «Commuting and gender in a lifestyle perspective», Urban Studies, vol 33, 2, pp. 283-300.

Cervero R., (1989), « Jobs-housing balancing and regional mobility », Journal of the American Planning Association, 55, pp. 136-150.

Chenu A., Tabard N., (1993), «Les transformations socioprofessionnelles du territoire français, 1982-1990», Population, 6, pp. 1735-1770.

Clark W.A.V., Kuijpers-Linde M., (1994), «Commuting in restructuring urban region», Urban Studies, vol 31, 3, pp. 465-483.

DREIF, (1995), Les déplacements des Franciliens en 1991-1992. Enquête globale de transport, 70 p.

Fortin S., Le Jeannic T., (1993), «Le parcours quotidien des couples biactifs », INSEE-/?e- gards sur l'Île-de-France, n° 22, pp. 9-12. л

Fradin J.-R., (1989), «Urbanisme et transports en Île-de-France» in: L'Île-de-France en mouvement, collection Reclus modes d'emploi, n° 16, pp. 344-354.

Gordon P., Wong H.L., (1985), «The cost of urban sprawl : some new evidences», Environment and Planning A, 17, pp. 661-666.

Gordon P., Richardson H.W., Wong H.L., (1986), «The distribution of population and employment in a polycentric city : the case of Los Angeles », Environment and Planning A, 18, pp. 161-173.

Gordon P., Kumar A., Richardson H.W., (1989), «Gender differences in metropolitan travel behaviour», Regional Studies, vol 23, 6, pp. 499-510.

Jacquot A., Rajaonarison D., (1993), «D'un recensement à l'autre : la redistribution géographique des emplois entre 1975 et 1990», Économie et Statistique, n° 270, pp. 23-35.

Lefranc Ch., Tabard N., (1996), Géographie des structures familiales, INSEE, Document de travail n° F9613, 168 p.

Neveu A., Zembri P., (1989), «Migrations alternantes : des comportements sociaux différenciés dans un espace polarisé», Données sociales Ile-de-France, pp. 146-150.

Poulain M., (1981), «Contribution à l'analyse d'une matrice de migrations internes», Recherches Démographiques, 3, Louvain-la-Neuve.

Ronsac J.-J., (1989), «Géographie des déséquilibres entre Г habitat-emploi : des surprises», Données sociales Ile-de-France, pp. 134-137.

TABARD N., (1989), «Voisinage social en Île-de-France», dans : Données Sociales Île-de- France, INSEE, pp. 73-82.

Tabard N., (1993), «Des quartiers pauvres aux banlieues aisées : une représentation sociale du territoire », Économie et Statistique, n° 270, pp. 5-22.

Zax J.S., (1991), «The substitution between moves and quits», The Economic Journal, 101, pp. 1510-1521.

384 B.BACCAINI

Baccaini (Brigitte). - Recent periurban growth in the Ile de France: forms and causes During the past several decades, settlement in the Ile de France has, in common with

the majority of large urban areas, experienced a rapid decongestion called suburbanization. After highlighting the specific situation of suburban development in the Ile de France, which has served as a point of transition between urban areas and the countryside, and the special features of the population who lived there in 1990 (over-representation of families with children and underprivileged social groups with housing primarily belonging to owner- occupiers) this paper deals with recent migration of the suburban population in order to explain the causes and types of demographic growth in this type of environment.

Decongestion of the Paris urban area is responsible for nearly three quarters of new suburban dwellers, and involves populations with very different socio-demographic characteristics from those who had migrated from the provinces and those who had lived in the He de France for a longer period of time. The occupational distribution of recent in-migrants to the suburbs also varied significantly in different geographical areas and selection effects have reinforced the existing specificity of various sectors.

Baccaini (Brigitte). - Modalités et causes de la croissance récente des communes périurbai- nes d'Ile-de-France L'Île-de-France connaît depuis plusieurs décennies, comme la plupart des grandes

régions urbaines, un desserrement rapide de l'habitat connu sous le nom de « périurbanisation ». Après avoir mis en évidence les spécificités du milieu périurbain francilien - transition entre l'espace urbain et le monde rural - et de la population qui y vit en 1990 (surreprésentation des familles avec enfants et des catégories sociales défavorisées, place prépondérante des logements individuels occupés par leur propriétaire), cet article s'intéresse aux migrations récentes des périurbains afin de saisir les causes et les modalités de la croissance démographique de cet espace.

Responsable pour presque les trois quarts des arrivées dans le milieu périurbain, le desserrement de l'agglomération parisienne touche des populations dont les caractéristiques socio- démographiques diffèrent sensiblement de celles des migrants venus de province ou de celles de la population installée depuis longtemps dans cet espace. La structure socioprofessionnelle des nouveaux arrivants du périurbain varie également fortement d'un secteur géographique à l'autre, ces effets de sélection ayant pour conséquence de renforcer les spécificités socioprofessionnelles existantes des divers secteurs.

Un second article montrera les relations qui existent entre ces trajectoires migratoires récentes et la mobilité domicile-travail des actifs du périurbain.

Baccaini (Brigitte). - Modalidades y causas del crecimiento reciente de los municipios periurbanos de Île-de-France Desde hace varias décadas en Île-de-France se observa, como en la mayoria de las grandes

regiones urbanas, una expansion del habitat conocida bajp el nombre de periurbanización. Des- pués de describir las especificidades del medio urbano de Île-de-France - transición entre el espa- cio urbano y el mundo rural - y de su población en 1990 - sobre-representación de familias con hijos y de categorias sociales desfavorecidas, preponderancia de viviendas individuales ocupadas por su propietario -, el présente articulo estudia las migraciones recientes de la población periur- bana. El objetivo de tal estudio es entender las causas y las modalidades del crecimiento demo- gráfico de este espacio.

La expansion de la aglomeración urbana de Paris explica casi las très cuartas partes de las llegadas al medio periurbano. Las características socio-demográficas de estas poblaciones difieren sensiblemente de las características de los migrantes llegados de provincias о de las aplicables a poblaciones instaladas desde hace tiempo en tal espacio. La estructura socio-profesional de los re- cién llegados al medio periurbano varia también fuertemente de un sector geográfico a otro. Estos efectos de selección refuerzan las especificidades socio-profesionales existentes en los diversos sectores.

En un segundo articulo se mostrarán las relaciones existentes entre estas trayectorias mi- gratorias y la movilidad domicilio-trabajo de los activos del medio periurbano.

Brigitte Baccaini, Institut national d'études démographiques, 133 boulevard Davout, 75980 Paris Cedex 20, France, tél. (33) 01 56 06 21 53, fax (33) 01 56 06 21 99, e-mail : [email protected]