The Persistence of Intergenerational Inequalities linked to Immigration: Labour Market Outcomes for Immigrants and their Descendants in France
Pages 645 to 682
Cite this article
- MEURS, Dominique,
- PAILHÉ, Ariane
- and SIMON, Patrick,
- Meurs, Dominique.,
- et al.
- Meurs, D.,
- Pailhé, A.
- and Simon, P.
https://doi.org/10.3917/popu.605.0763
Cite this article
- Meurs, D.,
- Pailhé, A.
- and Simon, P.
- Meurs, Dominique.,
- et al.
- MEURS, Dominique,
- PAILHÉ, Ariane
- and SIMON, Patrick,
https://doi.org/10.3917/popu.605.0763
Notes
-
[*]
Université Paris II, Ermes.
-
[**]
Institut national d’études démographiques.
Translated by Godfrey Rogers -
[1]
Each country produces its own categories that are superimposed on the classification by “generation”. The “visible minority” category groups together all individuals whose “skin is not white”. In practical terms, the different “visible minority” categories correspond to regions of origin, such as Asia, Africa, Middle East, or to “cultural” subgroups such as “Arab”. On the classification of origins, see Simon (1997 and 2004).
-
[2]
See in particular Reitz and Banerjee (2005).
-
[3]
With the Generation 1998 survey we can analyse the first three years of working life of young people who completed their schooling in 1998.
-
[4]
Among 18-40 year-olds, 10.2% report being school or university students or unpaid trainees. The proportion of 18-40 year-olds in education is highest among the immigrants who entered after age 10 (17%), as it includes immigrants who entered the country to pursue their education. The proportion is lowest among the children of mixed couples (5%) and stands at around 9% for the other three categories.
-
[5]
Calculation of unemployment rates (ratio of the number of unemployed to the number of economically active) gives much higher results for female immigrants (around 30% for G1 and 26% for G1.5) than for males. This is because the denominator for unemployed women is smaller due to the large proportion of immigrant women who are not economically active.
-
[6]
The fact that in the EHF survey the descendants of immigrants of North African origin cannot be distinguished from the children of repatriated settlers may well lead to an under-estimation of the unemployment rate for this population. The figure given, although high, is a minimum indicator (see Appendix).
-
[7]
The economically inactive were excluded from the analysis as the numbers involved are too small.
-
[8]
A logistic regression restricted to economically active women gives results very similar to the multinomial model presented here regarding the relative risk of being unemployed for the different generations.
-
[9]
Summary statistics for these variables by generation and by sex are given in Appendix Table 1.
-
[10]
A quadratic variable was introduced for age to see if the risk of unemployment falls faster as age increases.
-
[11]
In the absence of information on educational level of the parents, the socio-occupational category was used. We constructed a composite variable with seven modalities: both parents higher-level occupations/one parent higher-level occupation/one or both parents farmer or self-employed/father clerical worker or middle-level occupation, mother manual or clerical worker or middle-level occupation/father manual worker, mother clerical or manual worker or activity not given/father manual worker, mother inactive/both parents inactive or other activity.
-
[12]
In these estimates by country of origin, no distinction is made between immigrants arriving after or before age 10, or between second generations of mixed couple parentage and the others, since the numbers would have been too small.
-
[13]
The procedures for acquiring French nationality vary with the situation of the foreign person. According to the country of birth (France, or a country formerly administered by France) or the legal tie with a French national (spouse), the procedure followed can involve declaration, with minimum waiting periods for obtaining French nationality, or reversion. For persons who are foreign and are not covered by any particular procedure, acquisition is by naturalization. This cannot occur until after five years’ residence in France, though this waiting period can be reduced to two years or even waived in certain circumstances.
-
[14]
The average age at entry to France in this group is 22.6 years; half of the sample was aged 18–27 at the time of immigration.
-
[15]
Results not reported here. They are available from the authors on request.
-
[16]
The same result is observed by Canaméro et al. (2000) for the chances of finding a job after a period of unemployment: possessing French nationality does little to make up for the disadvantage associated with North African origins. However, Fougère and Safi (2005), using longitudinal data from the Permanent Demographic Sample (EDP) and estimating simultaneously the probability of acquiring French nationality and of being in work, show that naturalization has a positive impact on labour market access for immigrants, particularly for males from sub-Saharan Africa and Morocco and for females from Turkey and North Africa.
-
[17]
Non-wage employment is slightly more prevalent among male immigrants who entered after age 10 (13% of jobs held). This might suggest that this status is an alternative to wage earning as a mode of economic integration for immigrants who entered when adults. However, a logistic regression (not reported) indicates that this apparent over-representation of immigrants in non-wage employment is no longer observed once the main individual characteristics (age, qualification level, parental occupation) are introduced.
-
[18]
The dependent variable is the fact of having an unstable job (fixed-term contract, temporary work, subsidized contract, paid work placement, apprenticeship contract). Besides the generation with respect to immigration, the independent variables are qualification, age (and age squared), living in a couple, having children under age 3, residence in the Paris region, working in the civil service. The economic sector, the socio-occupational category (four headings) and that of the parents are also included.
-
[19]
The logistic regressions are not reported here; they are available from the authors on request.
-
[20]
Even though only part of the population studied (aged 18-40 in 1999) is potentially affected directly by the age limit. Note that the age limits were abolished in 2005 except for certain examinations to occupy positions classed as active service (police, prison staff, nurses, etc.) or requiring more than two years of post-examination training.
-
[21]
Candidates must have three years of higher education to sit the main examinations for category A status, and have the baccalauréat (high-school diploma) to sit the examinations for category B status.
-
[22]
90% of jobs in the state civil service are filled though competitive examination. Candidates are selected on the basis of anonymous written examinations and assessment interviews. In the local government civil service, 50% of recruitment is by competitive examination.
-
[23]
The presence of French nationals born abroad in the civil service is explained by the large numbers of expatriates who belonged to the colonial administration and by the policy of redeploying returnees in the civil service. It is estimated that one-third of returnees from Algeria were civil servants when they arrived in metropolitan France (Baillet, 1975).
-
[24]
The significance of the other coefficients does not vary.
-
[25]
The FQP survey that they use contains a question on parental nationality.
-
[26]
Simon (2003) speaks of “integration through reproduction” in reference to the descendants of Portuguese immigrants.
-
[27]
Calculation made on the same database.
-
[28]
Because the number of types of occupation in the civil service is different from that in the private sector, the indices of segregation for two sectors cannot be compared directly. However, the range of occupations in the civil service is smaller than in the private sector, which should make for a lower index of segregation than in the private sector. We observe the opposite, signifying that segregation is greater in the civil service.
-
[29]
This and the other percentages are based on the EHF survey data (Borrel and Simon, 2005, table 5A, p. 435). The estimate is based on an backward projection of the characteristics of foreign-born French nationals belonging to the age groups corresponding to “potential parents” of immigrants’ descendants. Adjustments were made to take account of mortality and differences in fertility between immigrants and returnees. Comparison of these estimates with FQP survey data that includes the information on parental nationality at birth confirms their robustness.
Working with the same source, M. Tribalat (2004) adopted a different logic to distinguishbetween descendants of returnees and descendants of immigrants, by using the languagetransmitted from parents to their children as a marker. The proportion of descendants of returneesis broadly the same whichever method is used.30. Unfortunately, the EHF survey did not collect this information. -
[30]
Unfortunately, the EHF survey did not collect this information.
-
[31]
This is true of the Canadian and Australian censuses, the Netherlands population registers and US surveys (Jensen and Chitose, 1996).
1The fact that immigrants tend to occupy menial and unstable jobs might be explained by characteristics acquired prior to migration, such as a low educational level for example. But what about their children who are socialized and educated in the host country? In this article, Dominique Meurs, Ariane Pailhé and Patrick Simon assess the occupational integration of the children of immigrants in France, the so-called "second generation". Using indicators such as access to employment, occupational status, access to jobs in the civil service or other occupations, and comparing the children of migrants with first-generation immigrants and with natives whose parents were born in France, they show that the "second generation" is still severely disadvantaged on the labour market, even though present in the civil service and in a wider range of economic sectors. Distinguishing by nationalities of origin reveals the prevalence of discrimination against descendants of immigrants of sub-Saharan African, North African and Turkish origin.
2Research into the situation of immigrants in countries of immigration has, for a number of years, focused increasingly on the life trajectories of their descendants. This interest is first of all a response to a generalized demographic process. The great waves of immigration that occurred in Europe between the end of the Second World War and the 1970s, and in the United States after the country was reopened to migration in 1965, have led to the formation of “second generations” composed of immigrants’ children who were born and brought up in the receiving countries (Crul and Vermeulen, 2003; Portes and Zhou, 1993). The second reason for the emergence of this research question is that outcomes for these second generations occupy a strategic place in the integration models developed in the countries of immigration. The main theories proposed for studying the processes of integration – i.e. full and unrestricted participation in the host society – assume that these processes unfold over several generations. Extending analysis to members of the second generation enables us to observe the different modes of integration of immigrants and their descendants and the capacity of the host societies to integrate these populations. The central question in this context concerns the social trajectories followed by migrants’ descendants. Will they experience upward mobility in relation to their parents – most of whom have limited formal education and a low social status in the host countries – or will the relative social positions be reproduced across the generations?
3The US social science literature is divided over the interpretation of current developments. The downward assimilation hypothesis is expounded by Portes, Rumbault and Zhou in several of their theoretical works (Portes, 1996; Zhou, 2001) and seems to be borne out by results from the survey conducted by the first two authors on the children of immigrants in two US agglomerations (Miami/Fort Lauderdale and San Diego) (Portes and Rumbault, 2001). Their theory of segmented assimilation, i.e. occurring in socially differentiated ways – entry into the mainstream for a fraction of immigrant descendants and into the underclass for a large proportion – is contradicted by Alba, Farley and Nee who, using data from the Current Population Survey for the years 1998-2000, conclude that assimilation continues to have beneficial effects in terms of upward mobility (Farley and Alba, 2002; Alba and Nee, 2003). In a nuanced and more critical perspective on assimilation theories, Perlmann and Waldinger (1997) emphasize the strong similarity between the integration processes affecting post-1965 immigrants and their children, and those experienced by immigrants in the great immigration waves of the late-nineteenth and early-twentieth centuries – integration within a system of class-based inequality but where mobility is a reality, one that all the evidence suggests will be repeated for the “new second generations”. Finally, Kasinitz, Mollenkopf and Waters (2002) continue this critical approach by introducing a dual dimension: the role of institutions, and the combined effects of belonging to the second generation and to racial minorities. The results from their New York survey on immigrants’ descendants show that incorporation of second generations into the city’s organized minorities does not have exclusively negative effects and can, on the contrary, give access to resources that facilitate social mobility.
4Economists, for their part, have sought to formalize the benefits for immigrants themselves and to identify the parameters that facilitate or hinder mobility in successive generations (Borjas, 1999, 1993; Chiswick, 1978; Solon, 1999). Language skills and the creation of informational and social networks in the host country are the main factors for convergence between the social positions of immigrants and native-born populations, while levels of educational attainment and the characteristics of family background, and possibly of neighbourhood or community likewise, appear to be determinant for second generations. These explanations based on personal characteristics are not the only approach to the analysis of disparities, however, and other studies have focused on the labour market players, revealing how discriminatory practices, judgments and systems affect the chances of finding a job, and on job characteristics and income differentials by ethnic and racial origin (for a recent review, see Blank et al., 2004).
5Applications of these models to immigration countries other than the United States remain limited, however, though more numerous national analyses would provide the starting point for a theoretical generalization based on experiences not drawn exclusively from the history of the United States. Boyd (2002) studied the case of Canada and found that the models of downward mobility and integration into the underclass are not validated for second generations belonging to visible minorities [1]. These results are tenuous, however, given the small sample size used. The inclusion in the Canadian 2001 census of a question on parents’ country of birth and its use in studies currently in progress should provide valuable insights [2]. This theme has so far received relatively little attention in the UK literature, which has focused rather on “ethnic minorities” without clearly distinguishing “generations” within the groups formed. The reference studies of the team headed by T. Modood, set up to analyse the PSI (Policy Study Institute) national survey of ethnic minorities conducted in 1994, make only very marginal reference to the position of second generations and do not use them as a significant category for analysing inequality in Great Britain (Modood et al., 1997). Lastly, the comparative research projects being developed in Europe have already supplied a number of answers that go beyond the diversity of the historical, political and social contexts of the individual societies, as can be seen from the two special issues of the journals International Migration Review (2003) and Journal of Population Economics (2003).
6The problematic emerged relatively recently in France (Simon, 2003) but research has developed rapidly since the late 1990s. The lack of statistics on the situation of immigrants’ descendants – a category until recently not identified in the large-scale socioeconomic surveys, notably the census – accounts in part for the small amount of quantitative research. The exception to this paucity of sources is the Mobilité géographique et insertion sociale (geographical mobility and social integration, MGIS) survey conducted by INED in conjunction with INSEE in 1992, which yielded an extremely rich research base on immigrants and their descendants (Tribalat et al., 1996). However, this survey only covered immigrants from certain countries of origin and did not provide an exhaustive picture of the labour market situation of migrants and their descendants. Research activity has increased through use of databases hitherto rarely employed in this area – like the Echantillon démographique permanent (Permanent Demographic Sample, EDP) (Richard, 2004) and the Formation et qualification professionnelle (Education and Vocational Qualification, FQP) surveys (Silberman and Fourrier, 1999) – or new sources that specifically identify the descendants of immigrants, of which the most important are the Etude de l’histoire familiale (Study of Family History, EHF) survey (1999), which will be used in this article, the Generation 1998 survey by the CEREQ (2001) [3] and the Histoire de Vie (Life History) survey (2003).
7This question of intergenerational mobility must also be set in the context of the far-reaching transformation of the labour market since the 1970s. Conditions of entry into the labour market have become more difficult, particularly for the younger generations, causing a slowdown in social mobility and even, for certain groups, a dynamic of downward mobility and exclusion (Chauvel, 1998). Particularly affected by this transformation are immigrants’ children who combine several negative factors. The great majority are from manual-worker backgrounds and must adapt to the new economic conditions without the relational and cognitive resources for labour market success passed down by the previous generation. In addition, a proportion of them face discrimination whose impact is becoming all the more decisive at a time when labour market opportunities are severely restricted and when the qualification gap with respect to the native population is tending to close. Conversely, other factors may have a positive effect on the prospects for social mobility of immigrants’ offspring. Their educational level is higher than that of their parents. In fact, if the comparison is made between the generations of immigrant origin, some form of upward mobility is likely to be observed. What are the relative disparities between the second generation and the “natives” (see Box)? Do they carry over from one generation to the next, indicating continued occupation of menial positions, or is there convergence from the second generation onwards? And if disparities persist, does the reason lie with inadequate education and family resources, or with discriminatory treatment, or a combination of both?
8This article uses data from the EHF survey combined with data from the 1999 census. The survey questions included the parental country of birth, thus making it possible to create the “descendants of immigrants” category and to compare the labour market status of the different generations of immigrant origin. By combining nationality at birth, and country of birth of the individuals and of their parents, we formed three categories of “generations”, two of which are split into subgroups, one by age at entry into France, the other by the mixed origins of the parents (see Box). The total number of survey respondents was 380,000, so relatively large numbers of individuals could be included in each of our “generations” and analyses performed by exact origin. Our study is concerned with the labour market status of the second generations relative to immigrants of comparable age, and not with the labour market position of immigrants’ children relative to their own parents. To minimize the effects of age structure on comparisons between our “generations”, the analysis is restricted to the population aged 18-40, which corresponds to 80% of the second generations of North African, Sub-Saharan African, Turkish, Portuguese and Asian origin. Also excluded from our study are school or university students and unpaid trainees, since they are not available to take up wage employment as their main activity [4]. The final sample contains 129,366 individuals.
Comparing generations
The total population is divided into five categories according to the generation with respect to immigration. The first two categories comprise immigrants by age on arrival in France, while the following three comprise people born in France with immigrant parents or otherwise. The numbers shown correspond to the number of observations in the 1999 EHF (Study of Family History) survey (population aged 18-40).
G1: immigrants who arrived after age 10 (n = 5,775);
G1.5: immigrants who arrived at age 10 or below (n = 2,472);
G2: people born in France whose parents were both born abroad (n = 7,286);
G2 mixed: people born in France one of whose parents was born abroad and the other in France (n = 9,614);
Natives: people born in France and whose parents were both born in France (n = 104,219).
The choice of age 10 (in completed years) to differentiate immigrants is based on the reasoning that schooling in France and early immersion in French society facilitate acquisition of French language skills and qualifications, with implications for labour market integration. This cut-off point and the term “generation 1.5” are widely accepted in the literature on the “second generation”.
9In the first section we examine labour market participation and vulnerability to unemployment in the different generations and by origins. The second section examines employment status, first the prevalence of insecure jobs, and second, the inequalities in access to civil service employment. The differences between types of occupation by generation and by country of origin are presented in the third section.
I – Labour market participation and vulnerability to unemployment
10The immigrant population observed today was largely formed during the period of labour migration and is structurally tied to the labour market. This initial tie – which A. Sayad highlighted thus: “What is an immigrant? An immigrant is essentially labour power, and labour power that is non-permanent, temporary, in transit” (Sayad, 1979, p. 7) – underwent a profound transformation in the 1980s and 1990s, a period marked by the end of immigration by low-skilled workers and by the continuation of immigration for family reunion that had started a few years earlier. Thus the proportion of immigrants in the economically active population, which stood at 8.8% in 1999, has been relatively stable over the last 25 years (Boëldieu and Borel, 2000). Of these economically active immigrants, around one quarter entered France at age 10 or below and make up the “generation 1.5”. The second generation represents 11.7% of the economically active population.
11In addition, industrial restructuring affected the economic sectors where immigrant workers were concentrated, and especially the low-skilled positions they occupied. In all sectors of industry, immigrant workers were progressively replaced by French workers, the latter often having higher skill levels, and foreigners were also more frequently made redundant, thereby magnifying the effects of restructuring. Between 1975 and 1982, the textile and construction industries lost 6.9% of the jobs held by French nationals and 24.3% of those held by foreigners (Merckling, 1998).
12Industrial restructuring and the accompanying rise in mass unemployment coincided with the increasing feminization of the immigrant population and the entry of these women into the labour force. Between 1982 and 1999, the number of jobs held by male immigrants declined by 11%, while among female immigrants it grew by 49% (Boëldieu and Borel, 2001). The increase in female labour force participation is reflected in the proportion of women in the economically active immigrant population, which rose from 28.3% in 1982 to 43.5% in 2005. Female participation rates have climbed steadily, from 33% in 1982 to 42% in 1995 and 46% in 2005.
1 – A disjunction between joining the labour force and being in work
13Levels of activity among immigrant men, high for structural reasons in the early phases of immigration, started falling when it became more difficult to redeploy redundant workers. The phenomenon of labour market withdrawal among male immigrants is more clearly observed with longitudinal data, as has been shown using the Life History survey (Houseaux and Tavan, 2005). With cross-sectional data of the kind used here, it is not surprising to find that almost all men under age 40 are in the labour market (Table 1). However, becoming economically active does not signify obtaining a job, and among immigrant men, whether they arrived in France before or after age 10, the proportion who are unemployed is nearly double that of the native population. Moreover, these difficulties in finding jobs are not specific to immigrants, and the levels are comparable for the second generations. This absence of change suggests that the potential gain from early socialization (G1.5) or birth in France does not translate into improved chances of finding a job. The “mixed generation” (one parent born in France, the other born abroad, see Appendix 2) tends to occupy a position closer to that of natives, initiating the real break between the “generations”. Rooth and Ekberg (2000) identify the same pattern in Sweden and attribute it to an input of “Swedish-specific” social capital particular to mixed couples.
Labour market status by sex and generation with respect to immigration (%)
Labour market status by sex and generation with respect to immigration (%)
14Women are more likely than men to be economically inactive and their labour market status varies more by generation. Note that these observations concern women under 40, i.e. age groups for which variations in activity levels by sex have grown much narrower. The proportion who are inactive drops sharply between women who immigrated after age 10 (31%) and those who arrived before this age (16%), and in the second generations, activity levels are similar to those of native women, with around 12% of inactive women in these three groups. Though activity levels converge, the same is not true for overexposure to unemployment, which follows the same gradient as for the male population. Among 18-40 year-olds, the proportion of unemployed women is close to that observed for men in the case of immigrants and of the second generation where both parents are immigrants, and is around 3 percentage points higher for the mixed generation and for native females [5]. These disparities in labour market activity and in levels of unemployment are reflected in large differences in access to employment. Just under half of immigrant women who arrived in France after age 10 have a job, against nearly three quarters of the native female population.
15This general picture must be broken down to take account of ethnic origin (Tables 2 and 3). Among men, the proportion who are inactive does not vary significantly. Among women, it is considerably higher for female immigrants – except those from Italy, Spain and Portugal – than for native French women. The frequency of unemployment varies by a factor of three for men and women alike. Those most vulnerable are from North Africa, sub-Saharan Africa and Turkey. The contrasts by origin are not quite so sharp for the second generations, but the general pattern is stable, with disparities in a proportion of one to two. Three different types of “employment situation” can be distinguished:
- a strong labour market position characterized by relatively small gender differences and few differences between the generations with respect to immigration. Immigrants and second generations from southern Europe (Portugal, Italy, Spain) are in this position and are not differentiated from the natives;
- over-exposure to unemployment combined with high levels of female inactivity among immigrants that carries over to the second generations, although with a lower frequency of unemployment and smaller male-female differentials. This second type corresponds to immigrants and “second generations” of North African origin;
- greater difficulties in obtaining work, with large gender differentials and positions that are reproduced to some extent in the second generation. This corresponds to the groups of Turkish origin and, for males, of Southeast Asian origin.
Labour market status of immigrants by country of origin (%)
Labour market status of immigrants by country of origin (%)
Labour market status of second generations by parents’ country of origin (%)
Labour market status of second generations by parents’ country of origin (%)
16Here we see contrasting models of labour market participation characterized by divisions that tend to be reproduced across generations. In particular, the persistent over-exposure to unemployment for people of North African [6], Turkish and Southeast Asian origin argues against the hypothesis of intergenerational mobility resulting from education and socialization in France.
2 – Inequalities persist even after controlling for structural effects
17The disparities in the risks of unemployment observed between the generations of immigrant origin arise partly from differences in individual and structural characteristics (age, educational attainment, neighbourhood, and so on). We also know that immigrants form a relatively heterogeneous group in respect of origin, and that the specific history of each migration stream, its social and demographic characteristics, and even the migrants’ own ambitions explain in part the socioeconomic differences observed. The econometric estimates presented here evaluate the degree to which over-exposure to unemployment for immigrants and second generations relative to the native population persists after controlling for structural effects.
18For men, a logistic regression was used to estimate the relative risk ratio of being unemployed for all economically active individuals (employed or unemployed) [7] net of the effect of the individual characteristics observed. For women, a multinomial logit model was preferred so as to compare the difficulties in finding work across the different generations but also the differences in labour market participation. The model estimates the relative risk of being unemployed [8] and the probability of being economically inactive relative to that of being active and in work. In both types of regression, the variables introduced to control for the structural effects were [9]: educational level; age and age squared [10]; unemployment rate in the region of residence of the individual observed; conjugal status (in a couple or not) and number of children under age 3; a variable summarizing the socio-occupational category (SOC) of both parents [11], assuming that it captures in part the family’s social capital (the influence of the family network in labour market access and, for women, the potential influence of the mother’s activity on the choice of activity).
19Of these variables, educational level is one of the most influential factors conditioning employment prospects. Levels and types of qualification are strongly differentiated between the generations of immigrant origin and may thus provide a partial explanation for the differences in unemployment rates. Broadly speaking, qualification levels for immigrants who arrived after age 10 exhibit a bipolar distribution, corresponding to a large proportion with no qualifications but also a large proportion with higher education qualifications. The presence of students who come to France to complete their higher education contributes to raising the average qualification level of immigrants. Educational qualifications among immigrants who arrived before age 10 are roughly the same as for the second generations, with an over-representation of intermediate vocational qualifications for women. On the other hand, the distribution by qualification of children from mixed couples is marked by an over-representation of degree-level education for men as well as for women. In general, the sex inequalities observed for immigrants – characterized by a very low educational level for women – are reversed in the following generations, the change-around occurring between short vocational courses, taken by men, and higher education courses, taken by greater numbers of women (Meurs et al., 2005).
20The full results of the regressions are reported in Appendix table 2. The coefficients of the structural variables have the expected effects: age and qualifications reduce the risk of unemployment, residence in a region of high unemployment increases it. As regards the variables of interest (Table 4), the relative risk of being unemployed rather than active and in work for male immigrants arriving after age 10 is more than double that of native French people, and remains well above 1 for all the other generations, including for the children of immigrants in mixed couples. Similar proportions are observed for women, but with smaller disparities relative to the native French. Female labour market participation also continues to be differentiated, irrespective of the generation considered, with a relative risk of inactivity that is 2.5 times higher for the first generation who immigrated after age 10 than for native French people, and significantly higher, though at much lower levels, for the other generations.
Labour market status by generation with respect to immigration (odds ratios)
Labour market status by generation with respect to immigration (odds ratios)
21The descriptive statistics presented earlier show large variations in the proportions unemployed by ethnic origin. Accordingly, we introduced into the models the country of origin for immigrants, and for second generations, that of their parents (Table 5) [12].
Labour market status by country of origin of the different generations (odds ratios)
Labour market status by country of origin of the different generations (odds ratios)
22Immigrants of either sex from Algeria, Morocco, sub-Saharan Africa and Turkey are three to four times more likely to be unemployed than the native population. Those from Italy, Spain and Portugal, for their part, are indistinguishable from the reference group. These findings are consistent with those of E. Maurin (1991) obtained using the Labour Force survey and the Career survey of 1989. At that time he concluded that foreign workers faced a “spiral of insecurity” and calculated that the risk of unemployment for North Africans was 79% higher than for the French reference population, while for Portuguese it was 49% lower. Nearly twenty years later the situation has not improved, as our results and recent similar studies make clear. Using the 2002 Labour Force survey on a larger range of ages, M. Domingues Dos Santos finds that the probability of being out of work is three times higher for a North African man than for his French counterpart, while it is one-third lower for a Portuguese man with equivalent characteristics (Domingues Dos Santos, 2005).
23To a lesser degree but for the same ethnic origin, the higher risk of unemployment revealed for immigrants is also observed for the second generations. It is between one-and-a-half and two times higher for them than for the native French. This risk remains significant for children of mixed couple backgrounds, even though in theory they are better equipped for labour market success. However, the disadvantage associated with immigrant parentage does not concern all origins, and second generations of southern European origin do not differ significantly from native French. The children of immigrants from North or sub-Saharan Africa, in common with their parents and despite their educational capital and familiarity with French social norms, face higher risks of unemployment than native French. These results parallel those obtained earlier by Richard (2004) using the Permanent Demographic Sample from the 1990 census, by Silberman and Fournier (1999) from the Entry into Active Life (EVA-Entrée dans la vie active) survey conducted by the CEREQ in 1993, by Canaméro et al. (2000) from the Trajectories of Jobseekers (Trajectoires des demandeurs d’emploi) survey conducted by the DARES in 1995 and, more recently, those of Dupray and Moullet (2004), Frickey et al. (2004) and Silberman and Fournier (2006) based on the Generation 1998 survey by the CEREQ. Silberman and Fournier (1999) show that young people of North African origin follow markedly less positive trajectories than other young people of immigrant origin, and less so again than those with parents born in France. In the four years after the end of their schooling they experience more periods of unemployment and benefit less than others from employment support measures. This situation persists after controlling for qualification level and is due to lesser mobilization of social capital (family members and contacts in the labour market) and a poorer match between qualifications and the labour market. Frickey et al. show that young people of North African origin wait longer to find their first job than young native French people (4 months against 3.3 months), are more often in subsidized jobs and spend longer out of work during the first three years after leaving school. These disparities persist after controlling for levels of qualification. Using an original survey on a cohort of new clients at the French national employment agency (ANPE – Agence nationale pour l’emploi), Canaméro et al. establish that the relative probability of finding another job after a period of unemployment is 15 percentage points lower for French nationals of non-European descent than for French natives, i.e. lower even than for non-European foreign nationals. Lastly, Dupray and Moullet explain the discrepancy between the probabilities of having a job for young people of French and North African origin. Demographic factors (having children under age 3, having a partner) have positive effects only for young people of French ancestry, while variables relative to employment experience have more contrasted effects. Finally, the authors estimate a disparity of 16 percentage points between the two groups in the probability of getting a job, two thirds of this being attributable to “differences of individual attributes in favour of young people of French ancestry” and the remaining one-third to a “difference of treatment”, i.e. discrimination.
24Our analyses do not fully bear out the hypothesis of additional discrimination against men on account of representations and behavioural norms judged negatively by employers. The positive image of young immigrant women presented by the media and the public authorities does not seem to be powerful enough to overcome the “double discrimination” based on origins and gender, notably for second generation women of Turkish descent. Moreover, the “risk” of inactivity for second-generation women is lower than for immigrants but is nonetheless still higher than for native women. At particular risk of inactivity are the daughters of North African and Turkish immigrants. Frickey et al. report an over-exposure to unemployment among young women of North African origin. Dupray and Moullet, for their part, find that mention of a work placement in the curriculum vitae does not have any beneficial effect for young women of North African origin whereas for all others it does.
3 – Factors influencing immigrants’ employment prospects
25Among the determinants affecting immigrants’ employment prospects, the literature has identified length of residence in France and acquisition or not of French nationality as potentially significant factors. Having French nationality can in fact make it easier to find work and to get access to a wider range of jobs (particularly those in the civil service – see below – and some independent professions). Table 6 indicates the proportion of immigrants with French nationality in our sample, according to whether they arrived in France before or after age 10. Nearly a quarter of immigrants who arrived after age 10, and half of those who arrived at younger ages, have acquired French nationality, with wide variations by country of origin. Proportionally the most numerous to acquire French nationality are immigrants from Southeast Asia and sub-Saharan Africa. The absence of any prospect of return migration for the former group and the possibility of reversion to French nationality for former citizens of the French colonial empire, partly explain these high levels. At the other extreme, Turkish and Portuguese nationals are least likely to acquire French nationality. This pattern was already observed after the 1990 Census and the MGIS survey of 1992. Since most countries have requirements for minimum duration of residence before naturalization can be applied for [13], acquiring the nationality of the country of residence is a lengthy process, and the procedure itself is relatively protracted. Besides, immigrants take this step only after abandoning their original plans to return to their home country after many years of residence in France. The fact that immigrants who arrived before age 10 are more likely to acquire French nationality than those who arrived later is explained partly by a longer average length of residence, but also by a weaker attachment to the country of origin and, in some cases, by a collective effect of the parents’ naturalization which applies equally to them.
Proportion of immigrants with French nationality by country of origin (%)
Proportion of immigrants with French nationality by country of origin (%)
26Duration of residence is also a potentially positive factor for “economic assimilation”, since it enables individuals to develop their language skills and to create social networks (Piché et al., 2002). Its impact on occupation and earnings has been studied by a large and predominantly Anglo-Saxon body of research (for a review of the literature, see Borjas, 1999; Lalande and Topel, 1997). In his early work, Chiswick (1978) found that after thirteen years of residence in the United States, the earnings gap between immigrants and natives tends to disappear. However, this estimate was based on cross-sectional data, which may distort the estimates owing to the compositional changes between cohorts of immigrants (see Borjas, 1985 and 1995). Furthermore, similar work carried out for other countries suggests that duration of residence has less effect on earnings growth for immigrants (Chiswick and Miller, 1995, for Australia; Baker and Benjamin, 1994, for Canada). A possible explanation is that the advantage procured by duration of residence is smaller when immigrants have characteristics similar to those of natives: quite simply they have fewer new skills (linguistic or occupational) to acquire during their stay. The “duration of residence” effect should therefore be differentiated according to the origin of immigrants.
27Among the immigrants aged 18-40 recorded in the 1999 French census, the oldest flows are those from southern Europe (Spain, Italy), then Portugal, followed by Southeast Asia. The more recent arrivals are from sub-Saharan African and “other countries”, the majority of them European (European Union members or not). The flow of immigrants from the North African countries has been the most constant over time. Restricting analysis to immigrants who arrived after age 10 brings out the more recent character of immigration from other European countries and from sub-Saharan Africa. The durations of residence are relatively similar irrespective of origin, however, due to the 40-year upper age limit of the population studied here.
28To what extent can the labour market disadvantage for immigrants from countries outside southern Europe be explained by differences related to length of time since immigration and/or to the proportion acquiring French nationality? To test the probability of unemployment for immigrants, two additional control variables were introduced into the previous specification: having French nationality and duration of residence. Analysis is restricted to the immigrants who entered France after age 10 [14], for whom duration of residence is the most pertinent variable in the analysis of employment prospects [15].
29According to our results, possessing French nationality has no significant effect on the risk of unemployment [16]. A longer duration of residence is associated with a lower risk of unemployment for men and of inactivity for women. Introducing into the regression an interaction between “short duration of residence” and “origin” to test for any differences in the duration of residence effect by origin shows that neither variable has a significant effect. In other words, duration of residence does indeed have an effect on reducing the risk of unemployment, but this effect does not vary with origin.
30All in all, introducing these two variables does not reduce the differences in the unemployment risk between groups according to origin. Immigrants of non-European origin continue to have a risk of unemployment distinctly and significantly higher than that of immigrants of European origin, a finding that holds for men and women alike.
II – Status in employment
31Employment prospects are not the only factor of labour market differentiation between immigrants, second generations and the native French. Job characteristics, and in particular employment status, can also be a major source of disparities. In the population of under-40s considered here, by far the most common status, irrespective of the generation observed, is that of wage earner, with proportions in excess of 90% for men and 95% for women [17]. In a context where job insecurity is on the increase, are the chances of finding stable work the same for immigrants and their offspring as for natives? The public sector is a major source of employment in France, particularly for women. Civil service positions are filled through impartial examinations, which in theory guarantee recruitment without discrimination on grounds of sex, or social or ethnic origins. To what extent are civil service jobs really accessible to immigrants and second generations?
1 – More insecure wage employment
32Insecure or unstable jobs refer here to the full range of subsidized employment, temporary employment, fixed-term work contracts, paid work placements and apprenticeships. Under this definition, among wage earners, 20% of men and nearly 22% of women aged 18-40 are in insecure jobs. The proportion of wage earners not in stable jobs varies across the generations (Table 7). Native French people are least exposed to insecure employment (19% for men, 21% for women). The most likely to be in insecure employment are immigrants who entered before age 10 and the second generations (around 25%). Male immigrants who entered France before age 10 are more often in temporary work, the second generations, particularly women, are slightly more often in subsidized jobs, while immigrants who entered after age 10 are more often on fixed-term contracts.
Proportion of wage earners with unstable jobs by different generations with respect to immigration (%)
Proportion of wage earners with unstable jobs by different generations with respect to immigration (%)
33Unstable employment is much more prevalent among immigrants of non-European origin than among those from southern Europe. Over one-third of male wage earners among sub-Saharan African immigrants, and over a quarter of male immigrants from North Africa and Turkey are in insecure jobs, compared with 19% of native French people and 13% of Portuguese immigrants. Female immigrants from these countries of origin are equally vulnerable: 40% of female immigrants from Morocco and Tunisia, 35% of those from Turkey, 34% from Sub-Saharan Africa and 31% of female wage earners from Algeria hold insecure jobs, compared with 21% for natives (and 14% for female Portuguese immigrants). The situation is scarcely better for the second generations, since over a quarter of male and female wage earners with immigrant parents from North or sub-Saharan Africa are in unstable jobs. The second generation of Turkish origin is particularly exposed to job insecurity (36% of women and 41% of men, which is higher than among the immigrant generation). Similarly, the second generations of Portuguese origin much more frequently occupy unstable jobs than the immigrant generation (25% of the second generation versus 13% of immigrants for men, and 26% versus 14% for women).
34In theory, insecure employment is associated with lower skill levels and with particular sectors of activity that have more frequent recourse to temporary work and subsidized employment schemes. The over-exposure of immigrants and particularly their offspring to unstable jobs could therefore be explained by their personal characteristics (qualifications, line of work). As in the previous section, we estimated a logistic regression for the relative risk of being in an insecure job (compared with the other types of wage employment). We used the same specification, and included the sector of activity and socio-occupational category as control variables [18]. After controlling for these structural factors, the probability of having an unstable job is still three times higher for male immigrants than for natives, and twice as high for female immigrants (Table 8). This increased risk of being in an insecure job is also observed for the second generations. Having mixed-couple parentage is associated with a reduced risk of job insecurity, for men in particular.
Unstable employment by generation with respect to immigration (odds ratio)
Unstable employment by generation with respect to immigration (odds ratio)
35Both male and female immigrants from sub-Saharan Africa, North Africa, Turkey and Southeast Asia, (with the notable exception of female immigrants from Turkey), are at greater risk than natives of being in an unstable job (Table 9). This is also the case for the sons of immigrants from Algeria and for the daughters of immigrants from North Africa and sub-Saharan Africa. Subsidized jobs thus appear decisive for getting the offspring of North African immigrants into the labour market. They remain so for those with higher education qualifications, as observed by Frickey et al. (2004) using the CEREQ survey. Finally, contrasting effects are obtained when variables relative to French nationality and duration of residence in France are introduced into the equation estimated on the immigrant population [19]. For male and female immigrants, acquisition of French nationality has no significant effect on the probability of having a stable job. On the other hand, the risk of employment insecurity declines with duration of residence in France, which seems consistent with a process of economic assimilation and progressively better knowledge of the labour market and of the skills that facilitate access to stable employment.
Unstable employment by ethnic origin of the different generations (odds ratio)
Unstable employment by ethnic origin of the different generations (odds ratio)
2 – Becoming a civil servant
36Faced with these difficulties in the labour market, immigrants, and above all their children, may seek a form of refuge against job discrimination in the public sector. The impartial character of its examination-based mode of recruitment could be expected to reduce the risk of discrimination on grounds of skin colour or ethnic origins. However, it is known that the propensity to enter the examinations is not randomly distributed and that the civil service tends to attract people with similar socio-cultural profiles (Calvès, 2005; Peres, 2004). So is the civil service open or closed to the descendants of immigrants?
37Access to the status of civil servant is regulated by relatively stringent conditions, of which at least two automatically weigh against immigrants: applicants are required to be French citizens or – since very recently – nationals of EU member states, and the age limit for taking the examinations is between 35 and 45 years. These two restrictions are cumulative, since not all immigrants become French nationals, while for those that are, acquisition of French nationality often occurs at relatively high ages, thereby effectively limiting the possibilities of entry into the civil service [20] (CERC-association, 1996). Logically, therefore, immigrants of both sexes, whether or not they went to primary school in France, are found to be under-represented in the three branches of civil service employment (state, local government, hospital services). For those with French nationality, under-representation among wage earners in the public sector (all statuses combined) ceases for men and becomes less acute for women. However, the immigrants with French nationality who enter the civil service do so with an insecure status, on fixed term contracts and in subsidized jobs (contrats emploi solidarité (CES), contrats emploi consolidé (CEC), emplois jeunes). Thus, at ages 18-40, only 4% of male immigrants with French nationality are fully tenured civil servants, against 11% of native French men, and 6% of female immigrants against 18% of native women (Table 10). Immigrants, particularly men who entered France after age 10, are much more likely to be working on fixed-term contracts: 39% of male immigrants with French nationality in civil service employment are on a fixed-term contract against 13% of native men, and 24% against 13% for women (Table 11). Note also that a large proportion of female immigrants who entered after age 10 are in subsidized jobs.
Proportion of civil service employees in the different generations with respect to immigration (%)
Proportion of civil service employees in the different generations with respect to immigration (%)
Status of civil service employees in the different generations with respect to immigration (%)
Status of civil service employees in the different generations with respect to immigration (%)
38The rules relating to the entrance examinations and the effects of the nationality requirement go some way to explaining the difficulties that immigrants experience in acquiring the status of tenured civil servant. But what is the situation for their children, who have French nationality at least from the age of majority? Although their general level of civil service employment is close to that observed for native French, children of immigrants are also less often tenured civil servants and are more frequently in subsidized jobs or on fixed-term contracts (Table 11). Here too, the second generation of mixed parentage stands out for having a high level of civil service employment, though the proportion in untenured positions is higher than among natives.
39To take account of structural effects, we used a logit model to estimate for the different generations the relative probability of being a civil service employee, first among all wage earners, then among those with French nationality, and finally the relative probability of being a civil servant among all wage earners with French nationality. In addition to the generation with respect to immigration, we controlled for age, qualification, marital status, number of children under 3, region of residence, socio-occupational category and parents’ socio-occupational category. All other things being equal, the relative probability of working in the civil service is significantly lower for immigrants than for natives (Table 12). Moreover, immigrants and the children of immigrants are less likely to have tenured civil servant status, irrespective of their nationality and qualification level [21]. The only group with the same chance as natives of obtaining a civil service post are second-generation men with mixed parentage.
Civil service employee or tenured civil servant status by generation with respect to immigration (odds ratio)
Civil service employee or tenured civil servant status by generation with respect to immigration (odds ratio)
40How can these differences be explained? Immigrants – especially those who arrived as adults – may be disadvantaged by the examination-based recruitment that is the norm in the civil service [22], by the fact that most of the jobs on offer are skilled positions for which the minimum requirement is three years in higher education, and by the importance of formal criteria particular to the French educational system (notably a written essay on a topic of general knowledge). This selection method can also be an obstacle for the second generations, who have grown up in a minority cultural environment that affords limited contact with the “common cultural background”. Assimilating this “common cultural background” is a prerequisite for the examinations, and whereas for native French people it is acquired spontaneously in everyday life, for the children of immigrants it often demands a specific effort of learning. Indeed, a study of the participation of “young people of immigrant origin” in the Category A (senior grade) recruitment exams showed that they are less well-informed about civil service careers and modes of entry, receive fewer encouragements to take the examinations and spend less time preparing for them (Peres, 2004). It is too early, however, to conclude that the recruitment exams are a discriminatory filter, since we would need to have information on the origins of all candidates and to compare their respective success rates in the different examinations.
41Access to civil service employment varies with ethnic origin. For all immigrant wage earners with French nationality, both men and women, the probability of being civil servants is lower than for natives, but it is particularly low for those from Southeast Asia, sub-Saharan Africa, Portugal, Italy and Spain (for men). The Italian and Spanish second generations, like their parents’ generation, are less likely than natives to be civil servants. The children of immigrants from Algeria, by contrast, have the same likelihood as natives of being civil servants, for men and women alike.
42Interpreting that last result is complicated by the heterogeneity of the group and by the “statistical noise” linked to the presence of children of repatriated settlers from Algeria who cannot be distinguished from the children of immigrants in our source (see Appendix 2). This problem also arises for the descendants of immigrants from Morocco, Southeast Asia and sub-Saharan Africa. French nationals born abroad are in fact over-represented in the civil service, particularly those from Algeria, among whom civil service employees represent 44% of wage earners, compared with a French average of 30% [23]. The hypothesis of intergenerational transmission of civil servant status was studied by Audier (2000) and has been verified by Fougère and Pouget (2004) in their study on the social origins of civil servants. Hence it is probable that descendants of repatriated settlers are also over-represented in the civil service and that this tends to artificially increase the proportion of immigrant second generations in this sector. It has been estimated that around half of all French-born people with at least one parent of Algerian origin are descendants of repatriated French settlers (Borrel and Simon, 2005). They make up an even larger proportion in the second generation with mixed parentage. When the latter are excluded, the relative probability of being a civil servant is significantly below 1 for second-generation men of Algerian origin (it stands at 0.68) [24]. Using data which identifies the children of returnees [25], Fougère and Pouget (2004) show that women with at least one parent of North African origin are over-represented among category C civil servants and under-represented among those of category A. Second-generation men of North African origin occupy category A jobs in broadly similar proportions to natives but are disproportionately present in untenured positions.
III – Occupational segregation becomes less marked over the generations
43If immigrants and their descendants suffer discrimination in access to stable work, is there also horizontal segregation, i.e. a differential distribution of occupations by origin? It is known that immigrants and their descendants remain largely confined to low-skill jobs and are over-represented in the economic sectors most exposed to fluctuating economic conditions (Tanay and Audirac, 2000; Meurs et al., 2005). The historical legacy of the conditions that prompted recourse to immigrant labour can be seen in the occupations held today. Hired to fill low-skill jobs during the decades of post-war prosperity, the immigrants have not followed the upward career paths of their French colleagues and are for the most part still in manual jobs. In their study of immigrants’ occupational histories, J.-L. Dayan, A. Echardour and M. Glaude (1996) noted that after 20-25 years of activity, nearly 75% of immigrant workers are still manual workers, whereas the proportion drops to 30% for men born in France. The study also shows that Spaniards have better-than-average mobility prospects, whereas Turks and Moroccans suffer from occupational immobility.
44A segmented labour market and an origin-specific specialization of immigrants in certain occupations are characteristics of multicultural societies, and these phenomena are still largely unstudied in France. The distribution of immigrants by occupation reveals fairly predictable patterns of polarization. For example, immigrant men are over-represented among building workers (9.5% against 3.4% for natives), domestic helpers (4.2% against 2.2%), directors and managers (3.7% against 2.2%), and health professionals. On the other hand, they are under-represented among legislators, senior officials and managers, and among farmers. Immigrant women, meanwhile, are over-represented among cleaning and other assistants, cleaners and launderers (12.2% against 5.9% for natives), messengers, porters, doorkeepers and related workers (2.3% against 0.3%) and among personal care and related workers (12.3% against 10.8%). On the other hand, they are under-represented among office clerks and nursing and midwifery professionals.
45The occupational distribution of the second generations is much closer to that of natives, with again a few particularities. Men are slightly over-represented among building workers, in certain craft and related trades (metal moulders, welders, sheet-metal workers, structural-metal preparers, and related trades workers, electrical and electronic equipment mechanics and fitters), as well as among computing professionals. They are slightly under-represented among legislators, senior officials and managers, agricultural workers, food processing and related trades workers, and assemblers. For their part, women of the second generation are over-represented among finance and sales associate professionals, among domestic and related helpers, and among cleaners and launderers. Like their parents’ generation, they are under-represented among personal care and related workers, office clerks, agricultural workers, food and related products machine operators.
46The analysis by occupations conducted by Lainé and Okba (2005) using the Generation 1998 survey confirms a fairly strong occupational specialization among the descendants of immigrants at the start of their working life, associated with a redistribution of occupations between the generations. The intergenerational discontinuity is sharper for young women “who prefer administrative occupations in the service sector” and for young men of North African origin who enter “the service sector at the expense of the industrial sector” (2005, p. 83). The social services sector has an important role here, pointing up the importance of youth employment schemes (emplois jeunes) in the trajectories of descendants of North African immigrants, as mentioned earlier. By contrast, a degree of continuity is observed for young men of southern European origin who tend to remain in the building and industrial sectors [26]. The hypothesis of “occupational niches” in which networks of social relations play an active role is substantiated by the preference shown by young people of southern European origin for the building trades, independently of any vocational training received. Similarly, the specialization of young people of North African origin in the social services sector is not simply a consequence of their initial educational choices (Lainé and Okba, 2005).
47As a summary measure of the degree of occupational segregation (i.e. the concentration of immigrants and non-immigrants in particular jobs) we use Duncan’s index of dissimilarity. This index is calculated in the following manner:
49where N is the total number of occupations considered,
50fit the proportion of workers from origin f employed in occupation i in year t,
51mit the proportion of workers from origin m employed in occupation i in year t.
52We computed the index using structure level 3 of the International Standard Classification of Occupations (ISCO) available in the census, corresponding to 106 occupations. The closer the index is to 100, the stronger the segregation, the closer it is to 0, the more the occupational distribution of the two groups is similar. This index is interpreted as the percentage of people who would have to change occupation to make the distribution between each group identical with that of the population as a whole.
53In comparison with occupational segregation by sex (the segregation index between men and women stands at 55%) [27], segregation by origin appears weaker, at around 20% for male immigrants (Table 13). However, the mean indicator for all immigrants combined is distorted by the diversity of employment situations within the immigrant group. Calculating the index by origin reveals greater occupational segregation: between one-third and half of immigrants from each origin would have to change occupation in order to obtain the same occupational distribution as the population in general. Compensation effects may operate when we go from populations by country of origin to immigrants as a whole, and heterogeneity may thus be reduced.
Index of occupational segregation by immigrants’ country of birth (G1 + G1.5) (%)
Index of occupational segregation by immigrants’ country of birth (G1 + G1.5) (%)
54For men, segregation appears especially marked for immigrants of Turkish and Asian origin. For their Moroccan- or Tunisian-born counterparts, the occupational distribution is more like that of the non-immigrant population. Occupational segregation between immigrants and non-immigrants appears weaker for women than for men, except for female immigrants of Portuguese and, most of all, sub-Saharan African origin. Once again, immigrants of Tunisian and Moroccan origin are less specialized by occupation, though the same also applies to those from Italy and Spain.
55Occupational segregation is found to decline across the generations with respect to immigration and the aggregate figure for the children of immigrants falls to around 9%, signifying a diversification of occupations in the second generations (Table 14). Occupational segregation lessens for those of European and North African descent, men and women alike. On the other hand, segregation remains high among the descendants of African and Asian immigrants, particularly for men. It has even increased between immigrants and the second generation for those of Turkish origin.
Index of occupational segregation of second generations (G2 +G2 mixed) by parents’ country of birth (%)
Index of occupational segregation of second generations (G2 +G2 mixed) by parents’ country of birth (%)
56When the index is calculated between immigrants or between second generations and they are compared pairwise by country of origin, the disparities appear more marked than when compared with the natives (Table 15). The indices are particularly high when the comparison is between immigrants, with the maximum obtained between Turkish and Portuguese immigrants (the index reaches 45% for the men, 30% for the women). Apparently, Turkish immigrants of both sexes work in occupations in which neither natives nor Portuguese immigrants are employed. And Portuguese immigrants do not have the same occupations as Moroccan or Algerian immigrants, thus confirming their high degree of specialization in niche sectors. The narrowing of differences in the second generations is also observed between origin groups, except for the divide between Turks and Portuguese, which even increases slightly in the next generation. Note that for women, segregation is weaker between the different immigrant origin groups – except for those of Turkish origin – than in relation to the natives.
Index of occupational segregation by ethnic origin for immigrants and second generations (%)
Index of occupational segregation by ethnic origin for immigrants and second generations (%)
57Finally, the index of dissimilarity between natives and immigrants was computed for civil service employees, for tenured civil servants only, and for private sector wage earners. Occupational segregation between immigrants and natives appears to be stronger in the civil service than in the private sector [28], particularly for men, and is especially high for tenured civil servants. Male and female immigrants who enter the civil service have a limited number of occupations: they are over-represented among secondary education teaching professionals, health professionals, domestic helpers, cleaners and launderers, social work associate professionals and office clerks, and in the case of men are distinctly under-represented among senior government officials.
Conclusion
58The situation of the descendants of immigrants is a central issue of political and social debate in multicultural societies. In an unprecedented movement of convergence, social science scholarship in North America and in Europe is giving close attention to the trajectories followed by the second generations. This interest shares the perspective of the classic works on intergenerational mobility, the innovation here being that the categories studied relate not to social class but to immigrant origins. With the new quantitative data sources now available, the labour market positions of the second generations can be examined in detail and compared with those of immigrants and natives. We made this comparison using the EHF survey, first by looking at levels of labour force participation, and particularly the risk of unemployment, and then by analysing the characteristics of labour market participation from the viewpoint of employment status and type of occupation.
59Despite the limitations specific to this source (which does not distinguish between descendants of immigrants and descendants of French nationals born abroad), our results show clearly that the types of employment occupied have changed substantially from one generation to the next, but equally that the second generations continue to face serious difficulties in accessing the labour market. Considerable heterogeneity is observed, however, between different ethnic groups. Descendants of immigrants from southern Europe are relatively unaffected by unemployment, whereas second generations of North African, sub-Saharan African and Turkish origin are extremely vulnerable. This greater exposure to unemployment is not determined exclusively by personal characteristics such as initial educational levels, social background, age, and family composition. All other things being equal, the discrepancies are considerable for non-European immigrants, and remain large for second generations of North African, Turkish and sub-Saharan African origin. In addition to unemployment, they experience greater job insecurity and rely heavy on subsidized employment schemes. On the other hand, the marked occupational segregation observed for immigrants declines at the next generation, indicating a process of occupational assimilation. A degree of occupational specialization is nonetheless observed for second generations, notably in the building trade and small industrial sector for the offspring of Portuguese immigrants, and in social services for the descendants of immigrants from the Maghreb countries.
60Several studies, cited in this article, have produced similar results that highlight the over-exposure of descendants of North African immigrants to unemployment, their over-representation in insecure jobs and even their downward assimilation. After controlling for the active variables, all conclude that immigrant origin has a specific effect on labour market integration. We can quote here the terse conclusion from the DARES study on unemployment among foreign-origin workers: “an inherited non-European origin constitutes a disadvantage in the labour market which is not removed by acquiring French nationality” (Canaméro et al., 2000, p. 7). The disadvantage associated with an “inherited” origin is reflected in a wide range of constraints that affect not just the chances of finding a job but the characteristics of the job held. It attests to the existence of discrimination whose widespread and persistent character pertains more to the system as a whole than to individual measures by employers. This discrimination weighs primarily on the trajectories of immigrants from the Mahgreb countries, Africa and Turkey, but also on those of their children. Silberman and Fournier (2006) reach the same conclusion, arguing that improvement in the overall employment situation is not, in itself, enough to solve the difficulties related to discrimination.
61In a context of labour market discrimination, the civil service could function as a refuge or even as an exemplary model (Calvès, 2005). Our findings and those of Fougère and Pouget (2004) show that this is not the case, and that although “young people of immigrant origin, and in particular those of North African origin are strongly drawn to civil service employment” (Calvès, 2005, p. 26), their access to civil service jobs is still characterized by inequality and low-level entry. This point deserves attention, since the civil service not only accounts for a large proportion of employment opportunities in France but its accessibility to the descendants of immigrants, and most particularly those of North African or sub-Saharan African origin, is a strategic component of anti-discrimination policies (Versini, 2004). It remains to be seen if this discrimination and its consequences on access to employment and the careers of immigrants’ children will prove enduring or transitory. A longitudinal follow-up of the occupational trajectories of immigrants’ descendants will provide an answer to this question.
Methodology used for descendants of repatriated settlers and children of “mixed couples”
62Construction of the “second generation” category using the EHF survey raises two major methodological problems. The first relates directly to the survey design, the second involves an ambiguity specific to the category itself.
Descendants of repatriated settlers
63Members of the second generation are defined by their place of birth and by their immigrant origins, characterized by the fact of having at least one immigrant parent. Yet a non-negligible proportion of people born in foreign countries are French by birth. Among the people born abroad, foreign nationals can be distinguished from French nationals using information on the nationality of their parents at birth. This information – which even when it is collected is not always reported correctly – does not feature in the EHF survey. Hence it is not possible to distinguish between descendants of French nationals born abroad and descendants of immigrants from the same country.
64Although historically the French have had rather low levels of international mobility, colonization produced large movements of settlement migration. For this reason the proportion of “repatriated” French nationals (i.e. returned former settlers) is particularly large among those born in Algeria, but it is also significant in the populations originating from other territories formerly in the French sphere of influence.
65Because French repatriated settlers from Algeria cannot be distinguished from “French Muslims” who subsequently became Algerian nationals and are thus defined as immigrants, the second generation of Algerian origin includes both descendants of immigrants and descendants of returnees. According to the estimate by Borrel and Simon (2005), descendants of returnees make up 55.8% of the second generation of Algerian origin [29]. Similarly, the second generations of Tunisian, Moroccan, African or Asian origin all include sizeable proportions of descendants of repatriated settlers (47%, 35%, 39% and 44%, respectively). What is the effect of the confusion between descendants of immigrants and descendants of returnees? In their comparison of returnees from Algeria and Algerian immigrants, Alba and Silberman (2002) show that the former have a higher educational level and are much less likely than the latter to be manual workers. They then show that the two groups are so clearly differentiated that by using a small number of descriptive variables [30] the origin group can be predicted correctly in 94% of cases. In conclusion, the descendants of repatriated settlers form a group that is socially distinct from the descendants of immigrants, and the fact that they cannot be distinguished in our data leads to an underestimation of the negative effects of inadequate social capital among immigrants.
Descendants of mixed couples
66The criterion selected to characterize the second generation is having at least one parent born abroad. This definition – which is common to studies and official statistics [31] on the second generation – results in children with two immigrant parents being treated in the same way as those with one immigrant parent and one non-immigrant parent. Yet the characteristics of the parental couple exert an influence on the educational and occupational trajectories followed by second generations. According to the EHF survey, 58% of descendants of immigrant were from mixed-couple backgrounds (Borrel and Simon, 2005, pp. 437-8). This proportion varies considerably depending on origins, ranging from three-quarters for descendants of Southeast Asians to a quarter for the descendants of Turkish and Portuguese immigrants.
67The socio-demographic parameters of mixed couples differ in many ways from those of couples where both partners are immigrants. The immigrant member of the couple entered France on average at a younger age, comes from a higher social category, is more likely to have a qualification higher than a basic vocational diploma (CAP) and to be employed in an intermediate or higher-level occupation. In addition to these factors, the non-immigrant parent possesses a cultural and social capital. Whether with regard to contact networks and knowledge of the unstated rules that govern the social world and particularly the labour market, or with regard to the visibility of non-native status (through physical appearance or family name), mixed couples escape some of the disadvantages to which immigrant families are subjected, or perceive themselves to be subjected. The phenomenon of invisibilization is quite complex to measure and is not the automatic consequence of birth into a “mixed” family. When the father is of immigrant origin, transmission of the surname will be a marker as powerful as if both parents were of the same origin. But it can be assumed that mixed couples are on average more richly endowed with human capital (in the sense of qualifications and specific linguistic and cultural skills) than immigrant couples, on condition that mixed status itself is a socially accepted norm. The benefit of being born into a mixed family would be nil if these families were heavily stigmatized, such that the children were prevented from finding a place in either reference community.
Characteristics of the sample (weighted averages)
Characteristics of the sample (weighted averages)
Factors influencing labour market status (odds ratios)
Factors influencing labour market status (odds ratios)
REFERENCES
- Alba Richard, Nee Victor, 2003, Remaking the American Mainstream: Assimilation and Contemporary Immigration, Cambridge, Harvard University Press, 359 p.
- Alba Richard, Silberman Roxane, 2002, “Decolonization immigrations and the social origins of the second generation: The case of North Africans in France”, International Migration Review, 36(4), pp. 1169-1193.
- Audier Florence, 2000, “La transmission du statut dans la fonction publique”, Économie et Statistique, 337-338, pp. 121-133.
- Baker Michael, Benjamin Dwayne, 1994, “The performance of immigrants in the Canadian labor market”, Journal of Labor Economics, 12, pp. 369-405.
- Baillet Pierre, 1975, “L’intégration des rapatriés d’Algérie en France”, Population, 30(2), pp. 303-314.
- Blank Rebecca M., Dabady Marilyn, Citro Constance F. (eds.), 2004, Measuring Racial Discrimination, Washington D.C., National Academies Press, 317 p.
- Boëldieu Julien, Borrel Catherine, 2000, “La proportion d’immigrés est stable depuis 25 ans”, Insee Première, 748, 4 p.
- Boëldieu Julien, Borrel Catherine, 2001, “De plus en plus de femmes immigrées sur le marché du travail”, Insee Première, 791, July, 4 p.
- Borjas George, 1985, “Assimilation, changes in cohort quality and the earnings of immigrants”, Journal of Labor Economics, 3(4), pp. 463-489.
- Borjas George, 1993, “The Intergenerational mobility of immigrants”, Journal of Labor Economics, 11(1), pp. 113-135.
- Borjas George, 1995, “Assimilation and changes in cohort quality revisited: What happened to immigrants earnings in the 1980s?”, Journal of Labor Economics, 13(2), pp. 201-245.
- Borjas George, 1999, “The economic analysis of immigration”, in Ashenfelter O. and Card D. (eds.), Handbook of Labor Economics, Vol. 3., Elsevier Sciences.
- Borrel Catherine, Simon Patrick, 2005, “Les origines des Français”, in Lefèvre Cécile and Filhon Alexandra (eds.), Histoires de famille, histoires familiales, Cahier de l’Ined 156, Paris, Ined-Puf, pp. 425-441.
- Boyd Monica, 2002, “Educational attainment of immigrants offspring: Success or segmented assimilation?”, International Migration Review, 36(4), pp.1037-1060.
- Calvès Gwénaële, 2005, Renouvellement démographique de la fonction publique de l’État : vers une intégration prioritaire des Français issus de l’immigration ?, Paris, La documentation française, 76 p.
- Canaméro Cécile, Canceill Geneviève, Cloarec Nathalie, 2000, “Chômeurs étrangers et chômeurs d’origine étrangère”, Premières Synthèses, Dares, 46-2.
- Cerc-association, 1999, “Immigration, emploi et chômage”, Les dossiers de Cerc-association, no. 3.
- Chauvel Louis, 1998, Le destin des générations. Structures sociales et cohortes en France au XXe siècle, Paris, Puf, 301 p.
- Chiswick Barry, 1978, “The effect of Americanization on the earnings of foreign-born men”, Journal of Political Economy, 86(5), pp. 897-921.
- Chiswick Barry, Miller Paul W., 1995, “The endogeneity between language and earnings: International analysis”, Journal of Labor Economics, 13(2), pp. 245-287.
- Crul Maurice, Vermeulen Hans, 2003, “The second generation in Europe”, International Migration Review, 37(4), pp. 965-986.
- Dayan Jean-Louis, Échardour Annick, Glaude Michel, 1996, “Le parcours professionnel des immigrés en France : une analyse longitudinale”, Économie et Statistique, 299, pp. 107-128.
- Domingues Dos Santos Manon, 2005, “Travailleurs maghrébins et portugais en France : le poids de l’origine”, Revue économique, 56(2), pp. 447-464.
- Dupray Arnaud, Moullet Stéphanie, 2004, “Quelles discriminations à l’encontre des jeunes d’origine maghrébine à l’entrée du marché du travail en France ?”, in Lahcen Achy, Jérôme de Henau, Nacer El Kadiri (eds.), Marché du travail et genre : Maghreb-Europe, Proceedings of the international conference in Rabat, 11-12 April 2003 organized by Mage, Dulbea and Insea, Bruxelles, Ed. du Dulbea, 413 p.
- Farley Reynolds, Alba Richard, 2002, “The new second generation in the United States”, International Migration Review, 36(3), pp. 669-701.
- Fougère Denis, Pouget Julien, 2004, “L’emploi public s’est-il diversifié ? Sexe, niveau d’étude, origine sociale et origine nationale des salariés de la fonction publique”, Addendum to a report on La diversité dans la fonction publique submitted by Madame Dominique Versini to the Minister for the Civil Service and State Reform, Paris, La Documentation française, 60 p.
- Fougère Denis, Safi Mirna, 2005, “L’acquisition de la nationalité française : quels effets sur l’accès à l’emploi des immigrés ?”, in France, Portrait social, édition 2005-2006, Paris, Insee, pp. 163-184.
- Frickey Alain, Murdoch Jake, Primon Jean-Luc, 2004, “Les débuts dans la vie active des jeunes après des études supérieures”, Notes Emploi-Formation, no. 9, Céreq, 37 p.
- Houseaux Frédérique, Tavan Chloé, 2005, “Quels liens aujourd’hui entre l’emploi et l’intégration pour les populations issues de l’immigration ?”, Revue économique, 56(2), pp. 423-446.
- International Migration Review, 2003, The Future of the Second Generation: The Integration of Migrant Youth in Six European Countries, special issue, 37(4).
- Jensen Leif, Chitose Yoshimi, 1996, “Today’s second generation : Evidence from the 1990 census”, in Portes Alejandro (ed.), The New Second Generation, New York, Russell Sage Foundation, pp. 82-107.
- Journal Of Population Economics, 2003, Symposium on “Second Generation Immigrants and the Transition to Ethnic Minorities”, 16(4).
- Kasinitz Philip, Mollenkopf John, Waters Mary, 2002, “Becoming American/becoming New-Yorkers: Immigrant incorporation in a majority minority city”, International Migration Review, 36(140), pp. 1020-1036.
- Lainé Frédéric, Okba Mahrez, 2005, “Jeunes de parents immigrés : de l’école au métier”, Travail et Emploi, 103, pp. 79-93.
- Lalonde Robert J., Topel Robert H., 1997, "Economic impact of international migration and the economic performance of migrants," in Mark R. Rosenzweig and Oded Stark, Handbook of Population and Family Economics, Handbooks in Economics, vol. 14, Amsterdam/New York/Oxford, Elsevier Science/North-Holland, pp. 799-850.
- Maurin Éric, 1991, “Les étrangers : une main-d’œuvre à part ?”, Économie et Statistique, 242, pp. 39-50.
- Merckling Odile, 1998, Immigration et marché du travail. Le développement de la flexibilité en France, Paris, L’Harmattan, 408 p.
- Meurs Dominique, Pailhé Ariane, Simon Patrick, 2005, “Immigrés et enfants d’immigrés sur le marché du travail : une affaire de génération ?”, in Lefèvre Cécile, Filhon Alexandra (eds.), Histoires de familles, histoires familiales, Cahier de l’Ined no. 156, Paris, Ined-Puf, pp. 461-482.
- Modood Tariq, Berthoud Richard, Lakey Jane, Nazroo James, Smith Patten, Virdee Satnam, Beishon Sharon, 1997, Ethnic Minorities in Britain: Diversity and Disadvantage, London, Policy Institute, 420 p.
- Peres Hubert, 2004, “L’accès aux fonctions publiques des jeunes d’origine immigrée”, Migrations Études, 22.
- Perlmann Joel, Waldinger Roger, 1997, “Second generation decline? Immigrant children past and present – A reconsideration”, International Migration Review, 31(4), pp. 893-922.
- Piché Victor, Renaud Jean, Gingras Lucie, 2002, “Economic integration of new immigrations in the Montreal labor market: A longitudinal approach”, Population-E, 57(1), pp. 57-82.
- Portes Alejandro (ed.), 1996, The New Second Generation, New York, Russell Sage Foundation, 246 p.
- Portes Alejandro, Rumbault Ruben, 2001, Legacies: The Story of the Immigrant Second Generation, Berkeley, University of California Press and New York, Russell Sage Foundation, 406 p.
- Portes Alejandro, Zhou Min, 1993, “The new second generation: segmented assimilation and its variants”, Annals of the American Academy of Political and Social Science, 530, pp. 74-96.
- Reitz Jeffrey, Banerjee Rupa, 2005, “Racial diversity, inequality and the cohesion of Canadian society: Research findings and policy implications”, paper presented at the conference on Diversity and Canada’s Future: Recognition, Accommodation and Shared Citizenship, Institute for Research and Public Policy, Montreal, October 2005.
- Richard Jean-Luc, 2004, Partir ou rester ? Destinées des jeunes issus de l’immigration étrangère en France, Paris, Puf, 258 p.
- Rooth Dan-Olof, Ekberg Jan, 2003, “Unemployment and earnings for second generation immigrants in Sweden. Ethnic background and parent composition”, Journal of Population Economics, 16, pp. 787-814.
- Sayad Abdelmalek, 1979, “Qu’est-ce qu’un immigré ?”, Peuples méditerranéens, 7, pp. 3-23.
- Silberman Roxane, Fournier Irène, 1999, “Les enfants d’immigrés sur le marché du travail. Les mécanismes d’une discrimination sélective”, Formation Emploi, no. 65.
- Silberman Roxane, Fournier Irène, 2006, “Les secondes générations sur le marché du travail en France : une pénalité ethnique ancrée dans le temps. Contribution à la théorie de l’assimilation segmentée”, Revue française de sociologie, 47(2), pp. 243-292.
- Simon Patrick, 1997, “La statistique des origines. Race et ethnicité dans les recensements aux États-Unis, Canada et Grande-Bretagne”, Sociétés contemporaines, n° 26, p. 11-44.
- Simon Patrick, 2003, “France and the unknown second generation”, International Migration Review, 37(4), pp. 1091-1119.
- Simon Patrick, 2004, Comparative Study on the Collection of Data to Measure the Extent and Impact of Discrimination within the United States, Canada, Australia, Great Britain and the Netherlands, Report to the European Commission, DG Employment and Social Affairs, Fundamental Rights and Anti-Discrimination, Luxembourg, 102 p., available online at: http://europa.eu.int/comm/employment_social/fundamental_rights/pdf/pubst/compstud04_en.pdf
- Solon Gary, 1999, “Intergenerational mobility in the labor market”, in Ashenfelter O. and Card D. (eds.), Handbook of Labor Economics, Vol. 3A, Amsterdam, Elsevier Sciences, pp. 1761-1800.
- Tanay Alice, Audirac Pierre-Alain, 2000, “Les salariés étrangers : loin des secteurs porteurs et des positions valorisées”, Premières synthèses, Dares, 46-1, November.
- Tribalat Michèle, 2004, “An estimation of the foreign-origin populations of France in 1999”, Population-E, 59(1), p. 49-80.
- Tribalat Michèle, Simon Patrick, Riandey Benoît, 1996, De l’immigration à l’assimilation. Enquête sur les populations étrangères en France, Paris, Ined/La Découverte, 302 p.
- Versini Dominique, 2004, Rapport sur la diversité dans la fonction publique, report presented to Mr Renaud Dutreil, Minister of the Civil Service and State Reform, Paris, La Documentation française, 107 p.
- Zhou Min, 2001, “Progress, decline, stagnation? The new second generation comes of age”, in Waldinger R. (ed.), Strangers at the Gates. New Immigrants in Urban America, Berkeley, University of California Press, pp. 272-307.
This article is available in open access under our model Subscribe To Open.
Uploaded: 08/01/2007
https://doi.org/10.3917/popu.605.0763