Article body


For over twenty-five years, public debate in France over “insecurity” has revolved around a few central themes, particularly the supposed relationship between crime and immigration. Discussion of this alleged relationship focuses on two groups: immigrants, particularly those who who have entered French territory illegally, and “young people of immigrant origin”, particularly those whose parents are black and come from Maghreb and North Africa. Most of those in this second group are French citizens. For some time, the relationship between crime and immigration has been seen mainly through an ideological and moralistic lens. Those on the right and the extreme right accuse, those on left or in other anti-racist movements respond, in a rigidly divided discourse. However, during the 1990s, this division decreased a bit (Mucchielli, 2002: 30-32). At the turn of the 1980s and 1990s, the “Islamic headscarf affair,”[3] the appearance of “urban riots,” and the context of the (first) Gulf war led to increased fear of the Arab world and of the increasing presence of Islam in France (Cesari, 1997). Subsequently, changing policies in the socialist party and in Jospin’s government (1997-2002) over questions of “security”, particularly in relation to juvenile delinquency, greatly reduced the left’s traditional criticism of responses to such issues and allowed more open discussion of the culpability of immigrants and young people with immigrant parents. Symbolic of this change, individuals recognized for their fight against racism, such as the president of SOS Racisme (Le Figaro, 25 April 2002) and Father Christian Delorme (Le Monde, 3 December 2001), an important actor in collective actions by young people of North African origin early in the 1980s, publicly called for ”breaking the taboo”, acknowledging a “hyper-delinquency” among young people with immigrant parents.

These changes would have been positive if they had led to intellectual debate that allowed argumentation and contextualizing demonstration to take precedence over emotion and common sense. That was not the case. After a brief reminder of the dominant discourse in public debate and various attempts to criticize the links between delinquency and immigration, I discuss how this topic is understood and what sociology might contribute to the discussion.

A new model: the “city mafia spiral”

The picture of new forms of delinquency provided by a number of contributors to the public debate is essentially this: we are dealing with youth who lack socialization, education, moral and social references, and whose parents have “resigned” or become indifferent. Worse, many of these young people are drug addicts and, both to meet their own habits and to make a living, turn to trafficking in drugs, often by becoming part of criminal armed groups, creating an underground economy and terrorizing their cities and surrounding areas. Car fires are used to make stolen cars disappear and riots are not an expression of feelings of injustice but a way to keep the police at a distance in order to protect dealers. The eldest initiate those younger into this criminal system, gradually developing a “mafia system” organized around a few “crime families” who, little by little, gain control over an entire neighbourhood and use younger members to protect them from the police[4]. This description is not, however, applicable to just anyone but only to North Africans. There are thus “cultural” reasons for this phenomenon: “considered relatively safe according to the traditions and culture of the North African population and further legitimized by its positive economic impact, the hashish traffic structures the schedules and supports the consumption capacity of the district. As a contributor to social peace, it draws the veil of mafia silence over the neighbourhood” (Bousquet, 1998: 119[5]). However the idea that consumption (and hence trafficking) of drugs is legitimized by the “culture of the North African population” is incorrect. In reality, only one North African country – Morocco – produces hashish (Observatoire géopolitique des drogues, 1996: 113-115). The problem therefore is not North Africa in general but Morocco in particular, and even then not its culture but its colonial history and the current status of its economy[6].

Another example: on 28 March 2000, one of the contributors to the debate on crime was cited in the Senate Commission on Decentralization as having said said that “the major taboo in matters of urban crime concerns the perpetrators’ origins and that he had been part of a general investigation that found that out of 3,000 perpetrators of urban violence, only approximately fifty had a ‘Gallic’ family name.” (Raufer, 2000). During the same period, Claude Goasgen, Paris deputy for the French Republican Party, used his Internet site for posts denouncing the “hidden reality” of immigration as well as explanations of the links between delinquency and immigration: “these links are clearly mechanical: a population that is young, male, and uprooted is going to be more prone to crime than old ladies living in a village – a four-year-old child could understand this.” The problem is that a large majority of the young people referred to here were born in France and their only experience of life has been in France. They are in no way “uprooted,” as their parents may sometimes have been, particularly when they came from a rural milieu[7]. This error makes sense, however, within a system of thought that, as part of the understanding of “uprootedness,” attributes to the “stranger” an irreducible otherness and an intrinsic (ontological, according to Sayad (1997)) potential for crime.[8]

Faced with these charges, and without assuming that any one of the elements in the broader argument will turn out to be correct, I now turn to certain aspects of contemporary knowledge, starting with the topic of immigrant crime.

1. Overly high levels of crime among immigrants?

The limited information available in police data

The most common source of statistical information on crime is the police; the information they provide is the least incomplete administrative data available[9]. However these statistics provide little of interest to a sociologist as they usually contain information about only three aspects of the demographics of those charged (that is, those people identified by the police as perpetrators and who appear in official statements sent to the public persecutor’s department): sex, status (minor or adult), and nationality (French or foreigner). They do not provide specific information on crimes commited by immigrants but only on those committed by individuals with a foreign nationality. Within this framework, both the interest and the limits of police information are easily recognized (Tournier and Robert, 1990; Robert et al., 1994). The idea is to provide, year after year, an overview of the results of police force service activity. The limit is thus contained in the statement: the data is intended to capture the activities of the police, not that of criminals. The two are obviously not unrelated, and the former captures an aspect of the latter, but only an aspect. Police statistics sift through the reality of crime and in order to judge the relevance of the data, we must understand the logic behind the sifting[10].

Police force statistics are gathered in two very different ways: from what victims tell the police and from what police officers discover themselves (either by catching the culprit in the act or as the result of medium- or long-term investigations by police). Nothing in the investigations following reports by victims makes it possible to determine whether the nationality (or even the skin colour) of the perpetrator influenced their decisions about whether or not to notify the police about offenses that affected them or those close to them. But this is not the case for the police. “Identity checks” are a classic example. But what does this say? That the police, as individuals, are racist? This is frequently suggested by those who display physical signs of African, particularly North African, origin, whether or not they are guilty of a crime. The problem is, however, much more complex and, in some senses, much more significant. Two police specialists have summarized this so well that I prefer to let them speak rather than paraphrase.

Researchers who have looked closely at police practises, in France and abroad, have concluded that there are widespread ideas about racism that create a norm that is difficult for an ordinary policeman to escape and, to an even larger extent, to oppose. The normative nature of police racism is an element of police culture, distinct from environmental racism or racism relating to the police officer’s social background, and is not ideologically or dogmatically constructed. … police racism is essentially reactive. Individuals do not join the police because they are racist, they become racist through a process of professional socialization. The habit of judging individuals by their alleged ethnic features is acquired on the job, over the course of professional socialization. … racist representations have an operational character in that they make it possible to differentiate individuals. In practice, by guiding police observations, they participate in a process of creative prediction. In a way, they are tools, part of a collection of practical knowledge that provides the background – the references – for police work. Ethnic attributes play a functional role in the same way as age or sex, in the sense that street police rely, above all, on an understanding of normality believed to be appropriate to a given population, space, and moment. Any discrepancy between these three parameters triggers police suspicion and can lead to an intervention.

Lévy and Zauberman, 1998: 293-294; see also Monjardet, 1996

This also applies to those of African appearance and, in another context, to gypsies (Zauberman, 1998). The problem then is evaluating the effect of police professional practices. In the absence of systematic local studies – the difficulty of carrying out such research can easily be imagined – such an evaluation cannot be exact. However, there are a number of elements that can be used to understand the mechanisms involved in selective sifting by police officers.

While the police recorded approximately four million offences each year throughout the 1990s, only a quarter of these were investigated and led to charges. Data published by the police provide information about the number of cases investigated, making it possible to see the difference between initial arrests and cases that are further investigated (table 1); “by moving from one to the other, one kind of offenses shrinks considerably, that of theft and destruction, which nevertheless constitutes the largest proportion of complaints from individuals. On the other hand, cases whose discovery depends mainly on the police initiative occupy a relatively disproportionate place among the accused” (Robert et al., 1994: 60). In terms of this study, these observations are important for two reasons.


Comparison of offenses and arrests in 2000

Comparison of offenses and arrests in 2000

TABLE 1 (continuation)

Comparison of offenses and arrests in 2000

Note: the rate of successful investigation is sometimes larger than 100%, perhaps because of cases being postponed from one year to another and/or being counted by more than one police force. This usually arises only in cases involving drugs.

* Pickpockets; ** Burglary involving a motor vehicle.

Source: Interior Ministry

-> See the list of tables

On the one hand, if the number of those charged is mainly affected by police initiatives, recognizing that, for small crimes, the police usually give priority to immediate observations over investigation (Lévy, 1987) makes it clear that police bias can play a large role, leading to an artificial overrepresentation of strangers among those arrested. This is particularly true in cases of theft and assault committed on public roads.

On the other hand, if the rate of solution varies according to the nature of the offence, are foreigners overrepresented in arrests for offences that are more likely to be solved? It turns out this is precisely the case for the majority of offences (with the exception of pickpocketing which is usually not investigated by the police but by department store security forces). In ascending order, these are procuring (225 foreigners arrested, 78% of whom were charged), rape (703 foreigners arrested, 75% charged), drug trafficking (1,992 foreigners arrested, 117% charged), carrying or firing a prohibited weapon (2,025 foreigneers arrested, 100% charged), violence and threats toward authorities (3,724 foreigners arrested, 97% charged), shoplifting (10,737 foreigners arrested, 87% charged) and, finally, attacks and nonfatal injuries (11,747 foreigners arrested, 72% charged).

This broad statistical analysis can be clarified by more precise studies, both qualitative or quantitative. For example, a study of drug trafficking across the French-Spanish border highlights the extent to which the high level of involvement of young white men from well-off backgrounds remains unacknowledged by police, who focus on the resellers who are, for the most part, members of the gypsy community (Missaoui and Tarrius, 1999). More generally, investigations of marijuana use among middle and high school students found that those who use it come from all levels of society, often from the upper classes (Choquet and Ledoux, 1994). However, most of those arrested for simple cannabis use are young, come from working-class backgrounds, and are undoubtedly of African origin (Barré and Godefroy, 1999), clearly demonstrating the effect of social and “ethnic sifting” by the police. Sifting can also be observed in the actions of private security officers. Ocqueteau and Pottier (1995: 138) looked at the way security operates in large supermarkets and found that when a shoplifter is caught, the chances of being reported to the police (as opposed to an in-store resolution) are much higher for foreigners.

For all of these reasons, police data must be analyzed very carefully. Let us then look in detail at the only statistical information available from this source – the number of foreigners among those arrested by the police (table 2) – before attempting to interpret it.

The overall influence and general characteristics of foreigners arrested by the police

Calculating the overall influence of foreigners in delinquency and criminality requires new calculations. In the 1999 census, France had 60.2 million residents (58.5 in cities), 3.3 million of whom were foreigners and 2.4 million of whom were French by naturalization. In the metropolitan area, the breakdown was 90.4% French by birth, 4% French by naturalization, and 5.6% foreigners (43.5% of whom were of African nationality, more precisely, 35% from northwest Africa) (Lebon, 2000). We now have the basis for our comparison – 5.6% foreigners. But a comparison with what? Certainly not with the overall number of crimes and infractions listed in police statistics. We must first eliminate “administrative offenses” in which foreigners are by definition strongly overrepresented. These are 1) violation of the law on foreigners, 2) offences related to false identity or other administrative documents, and 3) clandestine work (see also the comments of Desdevises, 1998: 270-273). When these offenses are removed, the number of foreigners arrested in 2000 was around 89,000, a little less than 11% of arrests (See table 3) This percentage has remained remarkably stable for over a quarter of a century, between 12 and 13%[11]. It is almost twice the percentage of foreigners living in metropolitan areas (5.6%). However, as Robert et al. (1994: 67) remind us, this comparison is misleading, first because police data includes all foreigners, whether French residents, tourists, seasonal workers, those on the French border, or undocumented immigrants and second because the foreign population living in France does not have the same demographic structure (in terms of sex and age) or the same social structure (particularly in relation to socioprofessional categories) as the population as a whole. Administrators in the statistical services of the Interior Ministry acknowledge these elements as sources of error in their annual data collection:

the figures relative to the foreign population in France … are established based on valid residency permits and take into account only foreigners resident in France. Foreigners travelling in France on visas or other tourist permits are therefore excluded from this data collection. However, their status as foreigners is taken into account in [police] reports, creating a problem in the calculation of crime rates. There is at present no way to determine the level of this problem

Interior Minister, 1996: 110


The role of foreigners in arrests in 2000*

The role of foreigners in arrests in 2000*

TABLE 2 (continuation)

The role of foreigners in arrests in 2000*

* The numbers and calculations exclude offences considered to be “administrative” (see explanations in the text)

Source: Interior Ministry

-> See the list of tables


Percentage of foreigners arrested between 1975 and 2000 for each category

Percentage of foreigners arrested between 1975 and 2000 for each category

TABLE 3 (continuation)

Percentage of foreigners arrested between 1975 and 2000 for each category

Note: there are some disparities in data due to changes in the way the Interior Ministry presentated data between 1985 and 1990. The percentage of foreigners in burglaries is high before 1990 because the data does not distinguish burglaries from other “entries to the home based on deception.” The same applies to car theft, which was not distinguished from theft of two-wheeled vehicles. Drug trafficking was not distinguished from drug use or from drug use and resale, categories that were introduced many years after the 1970 law on the criminalization of narcotic use. Other categories of infractions were not followed before1990 so no data is available.

Source: Interior Ministry; calculations CESDIP

-> See the list of tables

This creates a major difficulty, discussed in more detail later but important to recognize from the outset because, given the current state of available data, it cannot be corrected.

In looking at the overall proportion of foreigners arrested, it is important to recognize that this percentage conceals large disparities. Foreigners are slightly overrepresented among the perpetrators of car thefts, arson, destruction and damage to public or private property, and domestic abuse and are much more frequently accused of burglary, car theft, fraud, fakes and counterfeiting, and sexual violence other than rape. The overrepresentation becomes more apparent for offences such as violence and insulting authorities, carrying or possessing prohibited weapons, rape, theft with non-gun violence, or even homicide. It reaches its peak with drug trafficking, small thefts, and procuring. In considering the most important elements in this list of offences, it is important to look at both the offence categories in which overrepresentation is the most important (the fourth column of table 2) and the categories in which the number of those involved is the highest (second and third columns of table 2). Without forgetting the other question, let us look first at the number of those involved. Six main observations stand out (shown in italic in table 2). In descending order of importance:

  1. around one out of every eight foreigners arrested is arrested on a charge of shoplifting;

  2. around one in seven foreigners arrested is charged with aggravated assault (i.e., fighting);

  3. a little more than a quarter of the total number of those arrested for drug trafficking are foreigners (we will consider the specific problems this creates later);

  4. nearly half of those arrested for pickpocketing are foreigners;

  5. foreigners are clearly overrepresented in arrests for homicide and attempted homicide, even if there are only a small number of such cases;

  6. foreigners are clearly overrepresented in arrests for procuring, although there are only a very small number of cases reported.

Does the selective logic used in police work restrict sociological interpretation of the data? This is a delicate question. It is clear that this selective logic is directly responsible for the overrepresentation of foreigners in the police category of “violence and insults toward figures of authority” (that is, toward police officers). It is, in fact, the most direct consequence of the police practise of racial profiling. The same findings also apply to foreigners who are caught in the act in minor thefts (particularly shoplifting, discussed earlier). The research discussed above on the consumption and trafficking of drugs suggests that this area should be approached cautiously, particularly as police statistics do not specify the country of origin of the accused, nor whether those arrested are residing lawfully or unlawfully in France. Many of those investigated for trafficking are questioned by customs officers at borders or in airports and are often not French residents. Confusion is also created by using the same category for cases of both international trafficking and local resale, activities that are often linked but are clearly different[12]. Research based on cases in Bobigny, Nanterre, and Lille clearly indicates that most foreigners arrested for drug trafficking were either illegal immigrants, largely from African countries and living in precarious conditions at the very bottom of the trafficking hierarchy, or foreign nationals from countries that export cannabis (notably Morocco), who often work in networks that include those from several nationalities but whose dominant positions are often held by native-born French citizens (Duprez and Kokoreff, 2000; Duprez et al., 2001).

Foreigners residing in France: the criminal poor?

Some police data is so large that it must capture some of the actual situation regarding crime in France. The key factor then is how to interpret what is being provided, which seems to me to be fairly easy. Looking at the numbers (90,000 arrests), the higher crime rates for foreigners resident in France seem to reflect the plight of the downtrodden: low-level offenders involved in shoplifting and pickpocketing, small-scale drug suppliers, those involved in assault, and occasionally murderers. This interpretation is supported by research on the citizenship of at least a hundred murders who appeared before the criminal courts in the 1990s in a neighbourhood on the periphery of Paris (Mucchielli, 2004). In looking at a crime where the effect of “ethnic sorting” is likely to have little, if any, effect, it clear that foreigners are overrepresented – in this sample the perentage of foreign murderers was twice that of the percentage of those of similar backgrounds in the region studied. Those of North African and Portuguese origin were the most numerous, which corresponds to their demographic importance in the area. It is equally clear that these foreign criminals come almost exclusively from unstable working-class environments, as demonstrated by their employment status (over half are inactive or unemployed, others are essentially labourers), its nature (precarious or very precarious) and their place of residence (often the most dilipidated large buildings in the region).

In other areas, crimes commited by foreigners is essentially what the police refer to as “serious crime” (trafficking, procuring, large-scale fraud) and is tied to criminal organizations whose leaders do not necessarily reside in France. Without more information, this is all we can say.

It seems clear, then, that the overrepresentation of foreigners living in France among those arrested is strongly linked to their personal circumstances. In France, the majority of nationals from African countries—and their children, whether or not they have French citizenship—live in very precarious socio-economic conditions. A large percentage of them come from working-class backgrounds: in the 1990 census, foreigners from North African countries are listed as labourers in 66.5% of cases, employed in 15.8% of cases, and unemployed and have never worked in 4.2% of cases (INSEE, 1994). Combining these three categories covers 86.5% of the population, to which can be added the 5.2% who are craftspeople and merchants (employed in restaurants or grocery stores, etc.), the lucky few. Foreign workers thus belong to the working class in around 90% of cases (compared to 60 to 65% of the French population). Foreigners are also overrepresented in fields that require the fewest qualifications and the most arduous labour[13]. They are clearly overrepresented among “acting”[14] and fixed-term contract workers[15]. Finally, they are much more likely to be unemployed. In the 1992 INSEE survey on “Employment,” the rate of unemployment for French citizens was 9.5%, for foreigners as a whole it was 18.6%, but for North Africans it was 29.6% for the group as a whole and 50.6% for those aged 15-24 years old (INSEE, 1994). The same survey in March 2000 showed that these percentages have not changed. The rate of unemployment for foreigners (20%) is double the rate of unemployed for the whole work force (10%), triple (30%) if nationals of the European Union are excluded. Recent quantitative research has shown that there is a significant correlation between thefts and the rate of unemployment of youth below 25 years of age without diplomas (Lagrange, 2001a).

The INSEE (1997) revenue survey found that while 7% of all households in France live below the poverty line, this percentage rises to 25% for households headed by someone of Algerian, Moroccan, or Tunisian nationality (Hourriez et al., 2001). And the situation continues to deteriorate: in one region of Paris (Île-de-France), where around 12% of all foreigners live, in 1978 they comprised 18% of the poorest households. In 1996, this had risen to 32% (Observatoire national de la pauvreté et de l’exclusion sociale, 2002: 80-81).

As might be expected, the poor are concentrated in suburbs affected by urban policies (Castellan et al., 1992). In 1992, slightly more than 500 neighbourhoods were part of a city contract that applied to around 3 million residents whose main demographic characteristics were an overrepresentation of foreigners (18% or three times more than in the metropolitan area overall), of youth under 20 years of age (33% compared with 26% for the entire metropolitan area), and of large families (6 people or more in the household; 7.5% compared with 3.2% in the metropolitan area as a whole). In these areas, those under 15 years of age made up 21.6% of the foreign population (the same age group was 7% of the metropolitan population as a whole.) The “sensitive urban areas” created by the city’s 1996 stimulus package now contain more than 4.5 million people with similar characteristics and proportions with the notable exception that the rate of youth unemployment clearly increased during the 1990s (Le Toqueux and Moreau, 2002).

Ultimately, the high levels of crime among foreigners residing in France seems to be a modern version of the high levels of crime traditionally found in the most precarious social environments of industrial societies, traces of which can be found in literature and the media since the 19th century. The list of problems tied to such precarious living conditions is well-known: families frequently destabilized by the father’s unemployment or by work accidents; generally precarious living conditions that block access to numerous resources, leading to feelings of frustration among children; numerous families living in small spaces, leading to the street being used as a place to play or spend time (which creates opportunities for both high levels of social interaction and problems with parental surveillance); parents with low levels of education, making it difficult for them to help their children succeed in school [16]; and a higher incidence of psychological disorders and alcoholism among parents. For certain categories of foreigners, these classic social problems are accompanied by those linked to specific stigmas. To understand these stigmas, we need to look at the situation of young men with an African immigrant background – that is, those born in France to foreign parents.

2. High levels of crime among “youth from immigrant backgrounds”?

In public debate in France today it is not foreigners but young French men and women, particularly those with a North African background, or even Africa more generally, whose parents or grandparents were foreigners, who are treated with suspicion, often due to the assumption that they are dangerous. The French statistical system does not make it possible to evaluate the situation of those who are French but whose parents from other countries with the same precision as the situation of foreigners in general[17].

Survey data

In a self-report survey of delinquency conducted in 1999 with a sample of 2,288 participants between 13 and 19 years of age, Sebastian Roché (2001) found a clear over-representation of delinquency among those with two foreign parents and a slight overrepresentation of North Africans compared to other foreigners in this group. He notes that “North African youth often display a number of factors associated with delinquency: a lower level of parental supervision, more time spent in the HLM park outside the city centre, low levels of parental income and education, higher rates of school absenteeism” (p. 221). He emphasizes the problems created by bad relations between North African youth and the police, without clarifying the effect these problems have on levels of delinquency. The more important problem with this study, however, is its quantitative and global approach, which obscures the effects of local contexts.

Research by Hugues Lagrange (2001b), carried out at the end of the 1990s, is more interesting on these contexts. Lagrange systematically analyzed the surnames of groups of delinquents identified by the police, first in a provincial town like Amiens and then in the county of Mantois (Mantes-La- Jolie, Mantes-La-Ville, and Les Mureaux), a zone said to be “very affected” by high levels of Parisian redevelopment, typically involving those with economic and social handicaps and including a large number of foreigners or those born to immigrant parents. (These are what Dubet and Lapeyronnie (1992) call “exile neighbourhoods”). His results are clear. In Amiens, delinquency is not over represented among those born to African immigrant parents; their behaviour is similar to that of others from poor families (some of whom are aggressive and thieves). This supports my hypothesis. But the situation is different in Mantois, where there is a higher level of delinquency among youth born to African immigrant parents (those from sub-Saharan Africa as well as North Africa)[18]. Lagrande concludes that the delinquent or rebellious behaviour of youth from an immigrant background varies greatly depending on the local context. He does not ask whether police discrimination also varies according to local context, despite the readily apparent hypothesis that tension between youth and police is more likely in poor areas, particularly those with a greater concentration of foreigners or those with foreign parents who are often involved in the consumption of or resale and/or trafficking in cannabis. Research on police violence shows that young men, either foreign or of immigrant background, are strongly overrepresentated among victims of such violence, particularly in cases related to the consumption and/or resale of drugs (Jobard, 2002:210-211). The attitude of the local police forces undoubtedly depends on their relationships with other local institutional actors and with elected locals (who may pressure them to move in conflicting directions), with the prefecture, and, finally, on the personalities and experiences of local police.

The question thus shifts toward research on what characterizes local contexts, particularly (or at least primarily) as they affect youth with immigrant backgrounds. At this point in his analysis, H. Lagrange invokes “violent collectives” and their “affirmation of identity,” noting the importance of Islam in these neighbourhoods in the sense that it gives youth an opportunity to regain their dignity. The explanation, however, deserves further investigation, both with regard to how it works and the nature of the criminal behaviour.

Psychosociological elements in delinquency among youth from an African immigrant background

Culturalist explanations involve reductionism, something the social sciences have tried to distance themselves from for some time, both in France and in the United States. We have seen how this works with the theme of uprooting. Another way of explaining “youth from an immigrant background” involves foregrounding their “dual culture,” inevitably presented as a source of contradiction and conflict and thus of psychological disturbance and deviance. As a generalization, this idea is obviously incorrect, first because it relies on an idea of perfect cultural homogeneity that makes no sense for very old immigrant nation such as France, second because it reduces the construction of social personality to a cultural problem, ignoring the effect of socioeconomic integration. What such explanations do get right is that the question of cultural origins and relationships with the parents’ culture must inevitably be dealt with by “second generation” immigrants. But how? Past research, such as recent sociological investigations, highlights that in France, French youth born to foreign parents (whether during the waves of Italian and Spanish immigration early in the 20th century or the post-war waves of immigration from North African) have always had to deal, simultaneously, with a dual process: on the one hand, very strong pressure to conform to the norms of the host country, which can lead to early rejection of the culture of their parents, which is seen as so inferior or backward that it can lead to feelings of shame, and, on the other hand, a contrasting movement toward rejecting the stigma and publicly acknowledging their specificity. Is this a contradiction? In his remarkable work, Noiriel (1998:211-245) leaves this question unanswered. To me, it seems that there is no contradiction between these two attitudes, which are due to the same source: the fact that the culture of the country of origin is stigmatized in the host country. All these supposedly insurmountable contradictions linked to “dual culture” seem to me to arise from a single, central problem: how to deal with stigma and the devalorization of identity that it has led to since childhood (Vinsonneau, 1996[19]; Esterle-Hedibel, 1999) and, inseparable from this, the defensive reactions and strategies of self-valorization of the self that it, to a greater or lesser degree, engenders (Malewska, 1982; 1991; Camilleri et al., 1990). But “greater or lesser” with regard to what? It seems to me that management of stigma is tied primarily to three elements: primo, the success – or not – of socioeconomic integration, an integration whose crucial phase is the educational pathway taken; secondo, family determinants, which are family history (including the nature of the migration project and the discourse surrounding it), parents’ professional success, the attitude of parents toward school, and more broadly, the values of the host country; tertio, context, in particular whether the environment is more or less stigmatizing, the level of development at which racism is experienced and its intensity, and the influence of peer groups.

Contexts favourable to the development of over-representation of delinquency among those from immigrant backgrounds

To the extent that sociological analysis has been able to break with cultural reductionism and look instead at the central element that distinguishes the way identity is constructed among youth from an African immigrant background as opposed to that of other French youth, the focus shifts toward research into the environments that can lead to an over-representation of youth from immigrant backgrounds in criminal activities.

In contemporary French society, this environment is the crowded, run-down suburbs of large cities (sometimes the neighbourhoods around large cities), where foreigners and those of immigrant background often make up the majority of residents, the rate of unemployment is particularly high, the proximity to attractive municipalities further increases frustration, and the stigmatization of place (its negative reputation) further increases stigmatization of the population residing in it and contributes to discrimination by certain institutions and in the job market (Duprez, 1997). It seems to me that the determining element here is less the environment itself, captured at a given moment, than the sentiment shared by those who live in these neighbourhoods, who see them as an inevitability, a trap that one is unlikely to excape. From the end of the 1980s, for economic reasons and perhaps especially for political reasons (the failure of collective mobilization between 1982 and 1986), the effect of the temporal factor on the trajectory of those in the working class and first-generation immigrants has become inverted (Mucchielli, 2002: 107 and onwards)[20]. As Khosrokhavar (1997: 186) notes, “for the first time, a generation does not believe in the utopia of social ascent in its time.” Coupled with stigmatization in the public space and in relationships with institutions, this idea of a destitute social destiny produces intense feelings of frustration, exclusion, and collectively experienced contempt. It is in this sense that I suggest we look at feelings of collectivevictimization as a way to analyze the mentality of youth who often see themselves as living in ghettos, and to point out the way in which these collective representations solidify to such an extent that they come to be seen as the result of a conspiracy: the result of a plot by an unfair and racist society (Mucchielli, 1999b).

The behaviour of juvenile delinquents is certainly found in some strong arguments in support of such behaviour as a way of dealing with feelings of victimization[21]. But exactly which delinquent behaviours are we referring to? The feeling of living in a ghetto does not, in itself, seem to encourage violent acts such as murder or rape. In contrast, isolation in micro-localities can lead to increased violence between youth living in peripheral cities or neighbourhoods. In fact, the large increase in attacks and nonfatal injuries in police statistics since the end of the 1980s, corroborated by surveys on victimization (Robert et al., 1999), undoubtedly reflects this intensification in confrontations between juveniles, often in small groups. The greatest rise in crime among minors throughout the 1990s involved theft (notably car theft, which, as we know, can be undertaken for various reasons, including seeking revenge against another group (Esterle-Hedibel, 1996)); fighting (which again refers to aspects of local conflict); the consumption of and trafficking in drugs; destruction of and damage to public goods; and “attacks and violence” directed at law enforcement officers (Aubusson de Cavarlay, 1997). And, again, while this institutional data is in no way an exact measure of the level and evolution of such behaviour, it does confirm which social relations are particularly important to public institutions. This leads me to highlight two other probable causes of the localized over-representation of youth from immigrant backgrounds among those who commit crimes. The first is economic. The feeling that any possibility of economic and social integration is blocked is a powerful factor in investment in a classic underground economy (for example, stealing, fencing, and reselling car parts) or its more recent format (the systems for trafficking and resale of cannabis that have developed since the 1980s). The second is anti-institutional – “violence against institutions” – and refers to various forms of humiliation and violence directed toward objects and people symbolically tied to institutions. This affects many different stakeholders (police, public transport officials, schools and public buildings, sometimes firemen), but one has a decisive role, owing to the frequency and nature of its relationships with the youth of the neighbourhoods in question: the police. “Tensions with police are a major part of the urban experience for youth in poor neighbourhoods,” observes Michel Kokoreff (2004: 144). In fact, many studies clearly indicate that relations between young people of African descent and police in these sensitive neighbourhoods operate according to cycles of provocation, replies, reprisals, etc. on the part of both sides (for a summary see Esterle-Hedibel, 2002). It is in such contexts that “riots” sometimes occur, solidifying the collective understandings on both sides and their “essential claim of dignity and of justice” (Lapeyronnie, 1993: 263)[22]. As Bachmann and Le Guennec (1996: 355) explain, beyond the incidents that precipitate a riot, “who are the rioters fighting against? Against a faceless enemy. Against those who, on a daily basis, ignore them, condemning them to social inexistence and to a future with no possibiities. Their daily environment is woven of mistrust and hostility: their future is obliterated. No allies. No escape . . .. These two strong emotions, of being thwarted and of being held in contempt, are always at the heart of suburban rage.”


It is not an exaggeration to say that, in the ordinary universe of social representations[23], North African youth (“Blacks” or second generation “Beurs”) constitute a “type” of young delinquents and that the social housing neighbourhoods in which they are concentrated are seen as dangerous (Rey, 1996; Boucher, 2001). Such stereotypes have always been directed at the stranger. At the end of the 19th century, for example, Italians and Belgians were the objects of intense xenophobia and numerous acts of collective violence (notably involving the working class). The historian Gérard Noiriel (1988: 249ff) points out that three economic crises in modern France (at the end of the 19th century, in the 1930s, and in the middle of the 1970s) led to waves of xenophobia. Such problems also affected a number of its European neighbours[24]. Given this, we cannot say whether today’s stereotypes are more powerful than those that followed the wave of immigration initiated by French companies in the 1950s. They are, however, specific in the sense that they originate in France’s colonial past and the dominance (military, political, and economic) it has imposed on other civilizations, with the resulting feelings of individual and collective superiority that always accompany such a history. In this sense, devalorizing stereotypes are a source of additional difficulties for a population that is the result of a wave of immigrants that came largely from the unqualified working-class and which has found itself “trapped” by economic crises at the very moment that it became established in France, notably through family reunification. In capturing both institutional practices and the representations such actors have of themselves, these stereotypes resemble the self-fulfilling prophecies aptly proposed by Robert Merton (1965:140-161).

However, even if the personality development of every young person with this profile must include constructing ways to manage stigma, the resulting construct does not inevitably lead to criminal activities and a fortiori to a criminal career[25]. These activities and the careers they result in can be observed (in relation to other characteristics of the population) only in certain local contexts where the systems of segregation and discrimination have accumulated and become entrenched, transmitted between generations. For the rest, the determining factors in the delinquency of foreigners residing in France and of French citizens born to foreign parents are the result of familial and educational problems that are not particular to these populations but, on the contrary, are comparable to problems that have already arisen as a result of the rural exodus or among other foreign and working-class groups in times of economic crisis. It also seems that the general social mechanisms discussed in this article can be found in a large number of other Western nations[26].