The working class, minorities and housing in Paris, the rise of fragmentations

Transcription

1 GeoJournal 46: 51 62, Kluwer Academic Publishers. Printed in the Netherlands. 51 The working class, minorities and housing in Paris, the rise of fragmentations Catherine Rhein CNRS Université de Paris 1, U.M.R. LADYSS, No. 7533, 191 Rue Saint-Jacques, Paris, France Received 29 December 1997; accepted in revised form 12 July 1998 Key words: housing conditions, minorities, Paris, suburbs, residential segregation, tenurial polarisation, working class Abstract Which places are left to minorities in the housing and labour markets in metropolitan Paris? Over the last two decades, job structures have evolved dramatically, eroding the formerly prominent Paris working-class, shaking the social and political roots of its identity. These indicators would lead one to diagnose a growing fragmentation of metropolitan society. In the political debate, the burning issue of the crisis of suburbs has replaced outdated debate of the class struggle. This paper aims first to re-examine the interactions between social status and national origins, then analyses the housing conditions of different social strata, and finally demonstrates the pre-eminence of social status over national origins in the pattern of residential segregation. Over the last two decades, recurrent riots in housing estates have challenged French social policy. These events have been depicted in common opinion, and hence in political discourse, as the result of the over-representation of foreigners in social housing. But to what extent is there such over-representation of foreign households in social housing, when differences in social status are considered? We will first consider the positions of minorities within social strata, then the relationships between social strata and the housing market, and finally the positions of the minorities within the working class and the housing stock. Social strata and minorities Let us first investigate the jobs held by foreign workers in the Paris labour market and within each social stratum, especially within the working class. In metropolitan Paris 1,job structures have all the characteristics of a global city : the share of skilled service jobs is larger than the share of blue collar, industrial, workers. However there is no social polarisation as such within the labour market, contrary to Sassen s thesis (1991). Unskilled jobs, in particular manual jobs, have fallen from 39% to 22% between 1954 and 1990, while highly skilled jobs grew from 9% to 20% in Paris, just as they did in London (Hamnett, 1996; Rhein, 1996). Such fragmentation processes can be 1 In this paper metropolitan Paris, or Greater Paris, is defined as the core of the urban area as defined by the Census Bureau, INSEE. It includes the City of Paris, as the set of 20 arrondissements, and the three inner départements of Hauts-de-Seine, Seine-Saint-Denis and Val-de-Marne. The Ile-de-France, as the political and administrative region, includes four other départements. The population over 15 years living in metropolitan Paris is 4.87 millions, or 62% of the population living in the Ile-de-France region and 11.6 % of the French population. examined through an analysis of the position of foreigners in the labour market and in social strata, as well as of the school-age population living in households headed by foreigners. Social strata and social categories In France since the 1950s the position of workers within job structures has been depicted through changes in socioprofessional categories. These categories, in particular those of heads of household, are still taken as a good index or proxy for the position of these households within class structures. According to Desrosières and Thévenot, this vision is different from the Anglo-Saxon one, where the social milieu has been less strongly associated with the profession and was a more complex notion, within which income, housing and way of life are implied (Desrosières and Thévenot, 1988). The métier (profession) has had an enduring, pervasive effect within the French labour movement. The role of the State as a mediator between management 2 and workers unions is prominent in the elaboration of these social categories, through the negotiations of the collective conventions, conducted every year at the national level. To each social category, in particular the cadres (executives, professionals) and ouvriers (blue collar workers), wages, rights and social advantages are linked. In other words, the social categories which have been integrated within the census system since 1954 are an institutionalised version of these overall social relationships of production. 2 The French word, derived from Latin, is patrons for bosses, yielding patronat for the group of bosses as a highly structured sociopolitical group; the Conseil National du Patronat Français (CNPF) is the official union of the patronat.

2 52 The category ouvrier 3 should be considered not merely as a social status, but as a direct outcome of the institutionalised version of these class relations, as they have been built over the past century through conventions collectives, which still exist 4, and in the survival of a communist party still in charge of almost half of the largest municipalities around the City of Paris. The Red Belt, whose consolidation occurred in the 1950s, has played and is still playing an important role in the reproduction of the Parisian working class and in the building of working-class spaces around Paris, in particular through social housing (see Section 2). In this paper, we will focus on blue collar workers, as most foreigners belong to this social category in the labour market and to this social stratum. Indeed, in 1990, 31% of blue collar workers were foreigners in metropolitan Paris, against 12% in France as a whole (Table 2). Among unskilled white collar workers ( employés ), the share of foreigners was much lower than among blue collar workers, even in metropolitan Paris, where foreigners accounted for 11% of the total category in 1990, against 4% in France. In fact women accounted for 78% of this category and most are French 5. Indeed social categories as such are far more than variables in a strictly technical sense, but are real social constructs, as fully demonstrated by Desrosières and Thévenot (1988). Minorities, immigrants and foreigners Minorities, immigrés and foreign population are also social constructs as are social categories (catégories socioprofessionnelles), but of a different nature and meaning. Social categories have been built through specific processes of socio-political regulation, while categories related to national origin are still evolving, their meanings being much less clearly determined. The meanings of categories related to national origins minorities, immigrés and foreign population are all the more complex, for some of those minorities, namely Algerians, are from a former French colony where natives were not admitted to full French citizenship during the colonisation period ( ). Here we have chosen to use the term minority exclusively as a synonym for foreign communities, since such a category is not relevant in French civil law, as it is in the United States. In French censuses, the quest for national origin is not as developed as in America. There is no question about foreign origins or foreign ancestry, even for the first generation. Individuals with foreign origins can only be identified as such in as much as they are still living with their (foreign) parents. Special surveys are required to provide information similar to that used by Lieberson in his classic work on blacks and white immigrants over a century in the United States (Lieberson, 1980). Such lack of reliable information 3 Translated here as blue collar or manual workers. 4 As demonstrated by the recurrent strikes led by truck drivers, in October, during the annual bargaining process. 5 Table POP5, INSEE, 1990 census. does not help the development of knowledge about the social integration of foreigners into French society. In this paper, an attempt has been made to deal with the school-age population (5 19 years old), still living in households whose heads are foreigners 6. Three different populations will be examined: workers, heads of household and school-age population, thus providing a wide analytical scope. Foreigners in the metropolitan labour market and in social strata In the Ile-de-France region, the proportion of foreigners in the labour force is much greater than in France as a whole, since 45% of foreigners working in France live and work in this region; in the 1920s they were only 6% of the total (Guillon and Rhein, 1985). Other large foreign and working-class communities have existed, and still exist, elsewhere in the country, mainly in eastern and northern France, in large industrial and mining regions. Deindustrialisation is one of the major factors explaining the growing concentration of foreigners in the Paris region, since industrial jobs have almost disappeared in the other regions. In the Paris region, other types of unskilled jobs were substituted for former industrial/manual jobs, still within the ouvriers category, but not necessarily in industrial activities. Economically active foreigners are much more frequently blue collar workers than are French workers (Table 1). In 1990, 18% of the French active population, but 52% of the active foreigners, were blue collar workers in the Ile-de-France region. In the metropolis, in particular in the City of Paris, contrasts are much sharper, if one considers the employment structures as a whole. One active foreigner in two, and two North Africans in three, are blue collar workers (Table 1). The picture is even more striking if we aggregate the two unskilled categories, namely unskilled white- and blue collar workers, who are those experiencing the highest rates of unemployment, the lowest job security and the lowest wages. This group as a whole involves 49% of the French labour force, but almost 80% of active foreigners living in metropolitan Paris. On the other hand, 31% of the French labour force living in the City of Paris belong to the most skilled, best-paid category of executive/professional workers, against 17% in the inner suburban ring and 21% in the Ile-de-France region as a whole. It should be borne in mind that 38% of the highly skilled individuals working in France are concentrated in this region, where the share of total national employment reached only 23% in 1990 (Rhein, 1996). There is no evidence of social polarisation in employment structures, but investigation of job structures, controlling by citizenship, exhibits some evidence of social polarisation in metropolitan Paris. These results contrast with those provided by Peach for metropolitan London (Peach, 6 This analysis was possible thanks to special tables drawn, by the author, from the original 1990 census files bought by the LASMAS-CNRS, at the Centre National Universitaire Sud de Calcul (Ministère de l Education Nationale, de l Enseignement Supérieur et de la Recherche) at Montpellier.

3 53 Table 1. Socio-economic status by citizenship within the Paris region, active population, percentages City of Paris Inner ring Ile-de-France region French Foreign French Foreign French Foreign total Non salaried Executive, Professional Skilled white collar Unskilled white collar Blue collar Total Sources: Table NAT1, 1990 census, INSEE. 1997). Indeed, in London, contrasts are greater between Indians and Bangladeshis or West Indians, than between Indians and whites, since the distribution of Indians in job structures is quite similar to the distribution of whites. In metropolitan Paris, the structures of the foreign population are different from those in London. In Paris, highly skilled foreigners from highly developed countries such as western Europe, North America and Japan accounted for only 10% of active foreigners in 1990, while people from the Maghreb and southern Europe accounted for 67% of the foreign labour force. The most important community was of the Portuguese (22%), followed by migrants from Algeria (17%) and Morocco (11%). These three countries accounted for 50% of the foreigners living in metropolitan Paris in This is in large part a result of a policy developed by the car industry in the 1960s to hire cheap labour force from less developed countries, within a contractual framework controlled by the public authorities (Tripier, 1990). Not surprisingly, active foreigners were mostly concentrated in low-skilled jobs in 1990 (Table 2). It can be argued that discrimination based upon national origins can explain such an unbalanced distribution. For most public jobs, French citizenship is required. Indeed, public jobs account for 15% of total jobs, but their share is higher among white collar sectors, skilled (34%) as well as unskilled (31%), slightly lower among the highest skilled jobs (23%), but only 10% among blue collar jobs. So this argument provides only a small part of the explanation ; instead differences in skills and training between French and non-french workers must be put forward as important factors. This strong polarisation had obvious effects upon the social representation of foreigners as immigrés and blue collar workers. Migrants are mostly in blue collar occupations and, as Tables 1 and 2 help to demonstrate, the gap is quite large between French and non-french workers and it has been growing over the last two decades. The increase in foreigners among blue collar workers has challenged the former ideological construct of the unified working class, while the image of the immigré has replaced the former ideologically-built representation of the unionised industrial worker. Foreign executives and professionals, university teachers and researchers, artists and writers are generally not primarily categorised as foreign, but as highly skilled executives or professionals. Social distance between French and non- French workers is, in such cases, much smaller than between French and non-french manual workers. The category immigrés designates those foreign unskilled workers who are located at the greatest possible distance from French highly skilled workers and heads of household. We can now document recent changes within the foreign population, whether those who are economically active or not. In the 1960s, most migrant workers were males living alone or in special hostels. In the middle of the 1970s, the regulation of foreign migrants was changed, due to rising unemployment rates: immigration flows became far more controlled and reduced in volume. Public policies dealing with family recomposition therefore assumed new and prominent roles. As a result, the foreign-born population became younger and the sex ratio changed as the proportion of spouses, and more generally of women, grew significantly along with the proportion of children born in France. This set of demographic changes was among the major factors in the transformation of the non-french population. Simultaneously, mainly in the inner suburbs, the Frenchborn population was ageing, as most of the suburbs were built and populated in the 1960s and entered this ageing process at the end of the 1980s. This differential evolution resulted in contrasts among the school-age population, these contrasts being much stronger than within the total household population (Table 3). Simultaneously, educational policy also changed. An important reform at high school level was progressively implemented during the 1970s, broadening education through the introduction of a single curriculum offered to all teenagers, whatever their social origins (Rhein, 1997). Problems arose quickly because the challenge of doing so was more difficult than expected, due to unanticipated contrasts in social origins between teenagers, when controlled by national origins (Table 4). Half of the school-age population from foreign origins is also from working-class origins, but this rises to almost 2 out of 3 of those from Maghrebin and southern European origins, while this is the case for only one out of five youngsters of French origins. These increased differences and contrasts have two important implications. First, they played an important role in the crisis of the school system itself, as youngsters from foreign origins had to face two combined that is multiplied rather than added sets of problems, namely those dealing with social ori-

4 54 Table 2. Proportion of foreigners in different socio-economic status groups Economically active Paris Inner Ring Ile-de- France population France Non-salaried Executive, professionals Skilled white collar Unskilled white collar Blue collar Total Sources: Table NA T1, 1990 census, INSEE Table 3. Total and school-age population according to citizenship of heads of households, percentages Metropolitan Paris Total population School-age population French Italy, Spain, Portugal Maghreb Other countries Total foreigners Total Sources: censuses, INSEE, tables computed by the author, population living in households gins and those relating to foreign origins, language and the cultural gap. Second and consequently, these differences played an unexpected role in the reproduction and reinforcement of the spatial pattern of residential segregation, since catchment areas were introduced along with the high-school reform. Indeed a high degree of residential segregation had been registered in 1954, when census data were available for the first time at the local level 7. Research confirmed a decrease in the degree of residential segregation based upon social status over the period, then an increase over the period. However it will not be argued that these socio-demographic changes in the school-age population are the key or the only factor in this evolution. Rather, it is argued here that the pre-existing spatial pattern of the housing stock, combined with these socio-demographic changes, formed the preconditions of such changes. We shall provide some clues about the housing stock and its spatial structures. The share of foreigners has increased significantly among manual workers. The process of growing fragmentation within a declining working class has turned out to be more relevant than the process of polarisation. Such fragmentation has had significant implications for demographic structures, as well as for consumption patterns and housing. New boundaries arose between upper and lower strata, as well as among working class strata, based on questions of citizenship. 7 Local means at the commune level. The commune is the basic local authority in French territorial administration. Housing and social strata Let us examine here what is often the hidden dimension of residential segregation, namely the spatial structure of the housing stock and its relations to social strata and patterns of residential segregation. In this section we will use a rather schematic definition of dwellings, which is adequate for our purpose, since we need to consider first the social contents of different housing stocks, then the housing conditions of the social strata. This issue has not attracted much attention in France, unlike in British urban geography. In particular it did not fuel a debate over gentrification similar to that concerning the London case, between Hamnett and Randolph (1988) and Forrest and Murie (1986), nor stimulate research about interrelations between housing tenure and social segregation (Murie and Musterd, 1996). Some questions may not have equivalent meanings in France, such as the basic issue of housing classes, set forth by Rex and Moore in the 1970s, at least on theoretical grounds, for reasons provided in the preceding section. Social relations of production are fundamentally defined through positions in employment structures, that is mainly in the sphere of production, as expressed through the categorisation process in the French census. Five housing stocks Because of the development of housing over the last century, and the operation of rules and laws about tenures, a sociotenurial polarisation has grown over the last four decades

5 55 Figure 1. in French cities, in particular in metropolitan Paris 8.Inthe 1950s, such polarisation was indeed rather weak, due to the fact that three out of four households were renting their dwelling, while social housing accounted for no more than 5% of the housing stock and 7% of the rental stock. The situation in 1990 is in sharp contrast; within the same territorial limits, rental stock accounts for 57% only of the total housing stock, while social (rented) housing accounts for one-third of the total rented stock and for 20% of the total housing stock. At the same time, the share of owneroccupied apartments and houses rose from 14% to 34% in Due to a lack of relevant tables, it is unfortunately impossible to document the evolution within the social rental sector since the 1950s. Over the period, housing stock did not increase significantly in metropolitan Paris (Table 5). However changes in housing tenure followed the expected routes, except in the case of social housing: namely, a sustained increase of owner-occupancy, a slight decrease in the free- 8 The following figures are provided for the 20 arrondissements for the City of Paris and the 80 communes belonging to the former département de la Seine; this provides an accurate picture of metropolitan Paris back to the 1950s. Sources are tables published in the 1954 census of population, INSEE. rental segment and a strong decrease of the marginal stock of dwellings. As shown in Table 5, there has been a net increase of 16% in the social housing stock, mostly in peripheral suburbs, contrary to the evolution everywhere else in western Europe. The increase in owner-occupancy is an overall tendency, well on its way elsewhere in Europe, especially in Great Britain and in the Netherlands. In French cities, home ownership rose between the two world wars, as one-family dwellings were for the first time available to affluent strata of the working class and to the lower middle class. But in central cities, apartments were still mostly rented, since ownership was, prior to the 1938 law, restricted to whole apartment buildings. For the first time, this law allowed coownership of apartment buildings. However two other laws were needed, in 1965 and 1966, to organise co-ownership and the maintenance and management of co-owned apartment buildings (Haumont et al., 1971). The Malraux Law protecting historic districts, which was passed in 1962, and these laws on co-ownership have to be considered as fundamental landmarks promoting the revival of urban cores in France. Simultaneously the extension of mortgage lending

6 56 Table 4. Social origins of total and school-age populations controlled by citizenship of heads of household, percentages 1990 Total population 5 19 years old population Citizenship of heads of French Southern Maghreb Other Total French Southern Maghreb Other Total household Europe Europe Non salaried Professional, managerial Skilled white collar Unskilled white collar Blue collar Others Total Sources: original census files, INSEE, 1990, table computed by the author. Table 5. Dwellings according to types of housing and tenure in metropolitan Paris Types of dwelling 1982 % 1990 % Evolution One-family dwellings % % + 7% Multi-family Social housing % % +16% dwellings Owner-occupied % % +14% Free rental stock % % 5% Others (furnished rooms, hotels) % % 20% Total % % +3% Sources: census, INSEE, 1982 and 1990, central agglomeration of Paris for prospective apartment owners opened up this housing segment and contributed to its re-evaluation on the market. In the City of Paris, home ownership was made accessible to low-income households, since most of them were former renters, entitled to buy their apartment at special low prices. Most of these households were or are composed of retired people. Over the last two decades, in metropolitan Paris, retired households have been the fastest growing social category, in terms of income and of number. This is one of the most striking, though little acknowledged, impacts of the welfare system 9 and this process has been especially strong in the housing market (see also Rhein, 1998a). In the inner suburbs, home ownership is mostly to be found in a stock of multi-family dwellings built over the period (Map 1) and in the remnants of a stock of one-family dwellings built before the second world War (Map 3). The stock of free rented apartments is highly concentrated in the City of Paris (Map 2), where it still accounts for 46% of the total housing stock, against 25% in the inner suburbs. This housing stock is quite heterogeneous, in terms of prices, size and equipment. However two-thirds of the apartments have only one or two rooms. This small mean size weakens the operation of gentrification processes in central Paris and is one of the differences with London s housing stock. One-family dwellings account for 13% of the total housing stock. For obvious reasons, their share in central Paris 9 The French welfare system includes the Social Security system (health and retirement insurances ) and family allowances; in most professions, there are complementary benefits, provided by a mutual system. is minimal (Map 3). This stock accounts for more than one in four dwellings in the inner suburbs and this stock is still growing at the eastern periphery of metropolitan Paris while, over the last three decades, it has been slowly eroded in the suburbs closer to the City of Paris. The dominant tenure is owner occupancy (85% in 1990). For this reason housing type, rather than housing tenure, was adopted in this paper, assuming that housing type was more important than tenure status for further analyses. In metropolitan Paris, social housing is much newer than in London, and its share is much lower than in most cities of the Netherlands and the United Kingdom (Murie and Musterd, 1996). Indeed, in the City of Paris, social housing only accounts for 12% of the total housing stock of which one-third was built in the 1930s, as a result of a housing policy launched by the Conseil Général de la Seine,inorder to provide jobs in a weakened building industry during the economic crisis. However, in the inner suburbs, two-thirds of the stock was built during the period. In the City of Paris, half of the social housing stock is concentrated in the 13ème, 19ème and 20ème arrondissements, asparts of urban renewal schemes launched during the 1970s. The largest shares of social housing are to be found in northern and southern industrial and working-class peripheral suburbs (Map 4), although this spatial pattern does not coincide exactly with the working-class suburbs as a whole. Finally the marginal stock is quite heterogeneous: furnished rooms, hotels and rooms for caretakers and domestic servants still accounted for 14% of total stock in 1954 in the former département of the Seine, but for 9% only in This decreasing tendency has been well documented

7 57 Figure 2. (Merlin, 1983) as the erosion of a de facto social housing stock. Part of this stock was linked to domestic services, reappearing at the end of the 1990s; another part was linked to the role of Paris, as the heart of a national and international migratory system, which offered housing over short periods for young migrants trying to enter the metropolitan housing and labour markets. The spatial pattern of this stock approximates to that for free-rented apartments (Map 2). The housing stocks and their social contents The debate over tenurial polarisation has not been developed so far in France, partly due to lack of relevant temporal data and partly due to a lack of interest in housing other than social housing. Alternatively, tenurial polarisation can be fully understood as the outcome of the functioning of the whole housing system. Not surprisingly, members of the upper strata (nonsalaried, executive/professional) are the most likely to be found in owner-occupied apartments or houses, with household heads who are economically active or retired. However,

8 58 Figure 3. households whose heads are executive/professional workers, active or not, are more frequent in apartments: 46%, compared with 32% in houses, while houses are both the residence and workplace of non-salaried workers and also part of their capital (Table 6). At the opposite extreme, 47% of the social rented dwellings are occupied by active unskilled white collar and blue collar households, rising to 65% when active and retired household heads are considered, against 44% among total households. The free rented stock is the most accessible, in particular to young workers, migrant or not, affluent or not. These households are younger than those living in social housing and owner-occupied dwellings, since the share of retired households is lower than elsewhere (19% against 25% among total households). The second important characteristic of this stock is its relative social diversity. In this stock, the social profile is the closest to the metropolitan profile, although the share of executive/professional households is higher than in the total, due to the concentration of large rented apartments in the so-called Beaux Quartiers of the western arrondissements of the city of Paris. Lastly, the marginal stock is quite heterogeneous, but it houses members of the least affluent strata blue col-

9 59 Table 6. The social content of housing stocks in metropolitan Paris, percentages One-family Social Owner- Free Others Total dwellings housing occupied rented flats (furnished, flats hotels) Non-salaried, craftsmen Executive, professionals Skilled white-collar Unskilled white-collar Blue collar Retired Execut./Profess Retired white/blue collar Non-active Total Sources: special tabulations, original census files, INSEE 1990 lar and unskilled white collar workers as well as non-active individuals. Sociotenurial polarisation seems to exist, even if it is not as marked as expected. In fact, controlling by number of rooms within these five stocks would provide a more accurate picture of socio-tenurial polarisation, in particular within the rental stock and within the City of Paris, and would also reveal a clear-cut geography of stocks of large and small apartments, a geography which already existed in 1921 (Rhein, 1998b). Social strata and their housing conditions We reach now the last step in this second part, the reconstitution of the housing conditions of social strata at the metropolitan scale (Table 7). This step is necessary, given the different importance of each of the five housing stocks. It is also necessary to provide this broad though schematic picture, since the third part of this paper will only focus on the blue collar households and on differences in their housing conditions according to national origins. Not surprisingly, most affluent households are home owners. 50% of the non-salaried heads are home owners of houses or apartments; amongst executive and professional heads of household, 57% of the retired, but only 46% of those in employment are owner-occupiers. An age effect accounts for this difference and is also to be found among the least affluent strata. In 1990, the retired generations were composed of those workers born before 1925, who benefited the most from the peak of the welfare state between 1950 and They also represent those generations left behind in the central city, since young baby-boomer families were moved out of the central city to the inner ring as a result of the huge suburbanisation process. Consequently they are the most likely to have been renters with enough capital to be able to buy the dwellings they had been living in for decades. This age effect is most clear-cut among the least affluent, unskilled white or blue collar household heads: 43% of retired heads, but only 22% of active heads are home owners. Since it was initiated under economic conditions different from those prevailing over the last twenty years, such a trend is probably not a sustainable one and the picture will be different twenty years from now. Blue collar households are four times more likely than executive/professional households to be living in social housing. However this contrast cannot lead to a diagnosis of acute sociotenurial polarisation either. Minorities and housing So far, sociotenurial polarisation could not be diagnosed as such, which leaves open the question of tenure as being one of the factors in residential segregation in metropolitan Paris. Let us now go further into the role of housing, more precisely of social housing, in the processes contributing to the spatial concentration of foreign communities, namely as one of the major factors in ethnic segregation. Such a relationship is often taken for granted, at least at the political level. This relationship seems to be widespread in those European countries in which social housing accounts for a significant share of the total housing stock, but different meanings are assigned to it. In Amsterdam, social housing accounts for 42% of the total housing stock and segregation indices are quite low. However, as Musterd and Deurloo demonstrate, the causal relation between these two facts is much more complex than had been expected (Musterd and Deurloo, 1997). In metropolitan Paris, as the share of social housing is much lower, this would imply that segregation indices would be much higher than in Amsterdam. Employing arguments different from those used by Musterd and Deurloo, we will demonstrate that this is not the case either. Actually the notion of segregation recently took a very specific meaning in France, since the set of public interventions on social housing called politique de la ville (city policy) is a highly territorialized one. To put it more bluntly, the urban crisis supposedly takes place in large housing estates 10, because such estates are easily identifiable as a territory, since there is coincidence between their ownership and management, maintenance and urban form. As it would be more 10 Housing estates can be grands ensembles, if the number of dwellings reaches But most housing estates are smaller and are called cités.

10 60 Table 7. Housing conditions of social strata in metropolitan Paris 1990 Non sala- Executive Skilled Unskilled Blue Retired Retired Non Total ried Profes- white white collar Executive white or blue active craftsmen sional collar collar Profess. collar One family dwgs Social housing Owner-occ.flats Free rented flats Others Total Sources: special tabulations, original census files, INSEE 1990 Table 8. Proportions of foreigners among blue collar heads by types of dwellings Types of dwelling City of Paris Inner ring Metropolitan Paris One-family dwellings Social housing Owner occupied flats Free rented flats Others (marginal stock) Total Sources: INSEE original census files, 1990 difficult to define precisely sets of buildings concentrating so many deprived households in a more traditional urban fabric, most public interventions are concentrated on housing estates, even though their social, ethnic and economic structures were quite heterogeneous, as demonstrated by Champion and Marpsat (1996). In the remaining part of this paper, we will deal with this question, concentrating on the housing conditions of blue collar households. In the first place, taking account of their social status, foreign households are no more likely than French households to be living in social housing. As demonstrated in the first part of this paper, the majority of foreign heads belong to the blue collar stratum. In 1990, households whose heads were blue collar workers, accounted for 26% of the total households living in social housing. This is made up of 19% of French and 7% of foreigners. So French blue collar households are twice as represented in social housing as in the housing stock as a whole, where they made up 10% of all households. In other words, when social status is controlled for, it becomes clear that it is French households that are concentrated in social housing, not foreign households. This assertion needs verification at the metropolitan level, through considering the relative importance of foreign headed households among all blue collar households in the five housing stocks (Table 8 and Map 5). When social status is held constant, it is precisely in social housing that the share of foreign households is the lowest. It is the highest in the marginal stock, which is the least spacious and least comfortable (Table 8; Maps 6 and 8). In other words, foreign blue collar households are concentrated in two specific parts of the housing stock, which are both residual stocks, namely the marginal stock and the most decaying part of the rented stock (Map 7). Not only is there no evidence to support the opinion (not the hypothesis) that foreign households are concentrated in social housing, but Table 8 demonstrates that the situation is quite the reverse (see also Guillon and Chauviré, 1991; Rhein, 1998a). To what extent is there a residualisation process at work within social housing in metropolitan Paris? At the metropolitan level, the answer is positive. Over the period, the share of blue collar heads among total households decreased from 20% to 16%, and from 30% to 26% for those living in social housing. However the share of blue collar households living in social housing increased from 28% in 1982 to 34% in This increase did not result only from an increasing share of foreigners among blue collar heads. Indeed, the share of households living in social housing increased from 20% to 26% among foreign blue collar households, and from 32% to 38% among French blue collar households. So the residualisation hypothesis turns out to be valid for both groups, namely for French and foreign blue collar households. It might even be concluded that foreign households are somewhat lagging behind. Undoubtedly, there are concentrations of foreign blue collar households in distant communes located in the northeastern part of the City of Paris and on the eastern periphery of metropolitan Paris (Map 6). Social housing located in these parts of the agglomeration is generally considered as relegation stock by most French blue collar households. Working class spaces and concentrations of foreigners Let us now consider a second set of hypotheses and data concerning the residential segregation of households and of the school-age population, while holding social status constant and controlling by citizenship. In fact foreign blue collar households live in the worst dwellings, those that are the smallest and the least comfortable. To some extent, their role is similar to the one they have been playing on the labour market for decades: they replace French blue collar households in this housing stock and this is one of the reasons why they are still more concentrated within the City of Paris than are French blue collar households.

11 61 Schematically, blue collar spaces are located in the eastern arrondissements in the City of Paris, and in the northeastern and south-eastern inner suburbs, most of which belong to the Red Belt 11, where the share of social housing may be quite high (Map 4). The proportion of blue collar households in 1954 is, in this analysis, quite a good proxy for the definition of such working class spaces; the correlation coefficient between proportions of blue collar households in 1954 and of social housing in 1990 reaches only +0.65, while it reaches between blue collar heads in 1990 and social housing, and even for French blue collar heads, but only for foreign blue collar heads and social housing. The last check is the correlation between the share of foreigners among blue collar heads and social housing: the value of the coefficient ( 0.21) leads to the conclusion that the relationship is not significant and confirms the diagnosis advanced at the metropolitan level (Table 8). When controlling by social status, social housing does not play any significant role in the concentration of foreign blue collar households, since these households in 1990 still did not have access to this housing stock in the same proportion as French households. Even if this notion so far does not have any juridical meaning in French law, it should be considered that foreign households are discriminated in social housing. The mythical correlation between young Maghrebins and social housing The impact of such changes upon residential segregation has to be checked for the school-age population. In this last set of analyses, housing conditions have not been reconstituted for this population. However their social and national origins are known, as presented in the first part of this paper. In 1990, social housing only accounted for 23% of the share of foreigners among youngsters (15% in 1982). The fit 12 is even worse for youngsters whose parents were Italian, Spanish or Portuguese (6% in 1990; 3% in 1982), as these communities have settled since the 1960s in the eastern and southern suburbs. However, the model fits a little better to youngsters from Maghrebin origins (42% in 1982, 47% in 1990). But a further check by social origins dissolves the strength of the last relationship. Social housing accounts for 61% of the share of youngsters of French blue collar origins (52% in 1982), but only for 21% of the share of youngsters of Maghrebin blue collar origins. This leads to a final set of checks, based upon the use of indices of dissimilarity. In 1990, the index of dissimilarity between youngsters of blue collar and upper strata origins reached 42 (41 in 1982), whatever their national origins. But the index between youngsters from French and foreign ori- 11 See figure No. 1, p. 441 in Rhein, 1998a, and maps Nos. 9 and 10, Rhein, 1998b. 12 In this paragraph, results are based upon a linear regression model. The percentages are the proportion of variance taken by the model. Social housing is the independent variable. gins is much lower: 20 in 1990 (16 in 1982), whatever their social origins. When we control for national origins, the index of dissimilarity remains fairly high between youngsters from French blue collar and from French executive/professional households: 43 in 1990 (42 in 1982). The value of this index is slightly higher than the index between executive/professional heads and blue collar heads (39), and between the total population whose heads are executive/professional on the one hand, and blue collar on the other (40). The last, though most important, piece of evidence is as follows. The dissimilarity index between youngsters from French blue collar and from Maghrebin blue collar origins is only 07. In other words, social origins play the dominant role in the overall pattern of segregation, much larger than that relating to national origins (see also Rhein, 1997). Conclusions In metropolitan Paris in the 1990s, places left to minorities are located mostly at the bottom of the labour and housing markets. As Noiriel argued, this has always been the way integration worked in France (Noiriel, 1988); however the French melting pot seems to be far less efficient today than in earlier times. In this respect, minorities are quite different in Paris and in London, as far as skills and social status are considered. In Paris, half of foreign heads are blue collar workers, 80% belong to the least affluent social strata and their situation evolved in such a way that interaction between citizenship and social status has never been as strong as it is today. However the so-called crise des banlieues (crisis of the suburbs) is much too often attributed to foreigners overrepresentation in social housing. As demonstrated in this paper, although foreign households have gained access to social housing over the last two decades, their share is still lower than the share of French households, social status being held constant. Foreign households entered social housing as blue collar households, where they are in no case overrepresented, but generally still underrepresented. Contrary to what happened in metropolitan areas in the U.S.A., the strength of former class struggles deeply influenced the urban fabric of metropolitan Paris, through political and sociospatial polarisation, to such an extent that the actual pattern of residential segregation fundamentally remains a social one. But as far as processes of social integration are considered, students from both working-class and non-french backgrounds cumulate handicaps within the highly selective and demanding French public school system and set new challenges for reform. References Champion J. B. & Marpsat M., 1996: La diversité des quartiers prioritaires: un défi pour la politique de la ville, Economie et Statistique, ,

12 62 Desrosières A. & Thévenot L., 1988: Les catégories socio-professionnelles, Editions La Découverte, Paris. Forrest R. & Murie A., 1986: Marginalization and subsidised individualism. Int. J. Urban Regional Res. 10, Guillon M. & Rhein C., 1985: Travail et emploi en région parisienne, Espace-Populations-Sociétés 2: Guillon M. & Chauviré Y., 1991: Les étrangers dans l agglomération parisienne, unpublished research report, Université Paris I-Direction Régionale de l Equipement de l Ile-de-France, Paris. Hamnett C., 1996: Les changements socio-économiques à Londres, croissance des catégories tertiaires qualifiées ou polarisation? Sociétés Contemporaines 22/23: Hamnett C. & Randolph W., 1988: Cities, housing and profits. Hutchinson, London. Haumont A., Haumont N. & Raymond H., 1971: La copropriété. Editions du Centre de Recherche sur l Urbanisme, Paris. Lieberson S., 1980: A piece of the pie, Blacks and white immigrants since University of California Press, Berkeley. Merlin P., 1983: Pour une véritable priorité au logement social à Paris, La Documentation Française, Paris. Murie A. & Musterd S., 1996: Social segregation, housing tenure and social change in Dutch cities in the late 1980s, Urban Studies 33(3): Musterd S. & Deurloo, R., 1997: Ethnic segregation and the role of public housing in Amsterdam, Tijdschrift voor Economische en Sociale Geografie 88(2): Noiriel G., 1988: Le creuset français. Le Seuil, Paris. Peach C., 1997: Pluralist and assimilationist models of ethnic minority settlement in London 1991, Tijdschrift voor Economische en Sociale Geografie 88(2): Rhein C., 1996: Social segmentation and spatial polarization in Greater Paris, In: O Loughlin J. & Friedrichs J. (eds.), Social polarization in post-industrial metropolises, pp , Walter de Gruyter, New-York. Rhein C., 1997: De l anamorphose en démographie, polarisation sociale et flux scolaires dans la métropole parisienne, Ann. Rech. Urbaine 75: Rhein C., 1998a: Globalization, social change and minorities in metropolitan Paris: the emergence of new class patterns, Urban Studies 35(3): Rhein C., 1998b: Ségrégation résidentielle et parc de logements en banlieue parisienne ( ), In: Girault, J. (ed.), Les ouvriers en banlieue, XIXème et XXème siècles, Paris, Editions de l Atelier. Sassen S., 1991: The global city: New York, London, Tokyo, Princeton University Press, New-York. Tripier M., 1990: L immigration dans la classe ouvrière en France, CIEMI- L Harmattan, Paris.