Keywords

The line (presumed or real) between immigration and criminality has for some time been a special field of analysis in social research . Notable examples include the pioneering studies of members of the Chicago school of sociology, such as Frederic Thrasher (1927), John Landesco (1929), Clifford Shaw and Henry McKay (1942), and in more recent research in a European environment by Martin Killias (1997), Pierre Tournier (1997), Uberto Gatti (2004), Ernesto Savona and Andrea Di Nicola (1997).

While this research has in large part endeavored to determine whether or not immigrants were more involved than natives in criminal acts, in this essay we propose a markedly different point of view aimed at showing the paradoxes and aporias implicit in the analysis of the question of “immigration and criminality” in Italy . As we shall see, this theme contains two contrasting currents of research originating from 20th century sociological traditions; one is oriented in a positivistic sense, and the other comes from social constructivism (Collins 1988). Despite the existence of identical empirical results on the subject from the statistical point of view, the two “schools” of research arrive at diametrically opposed conclusions regarding the involvement of foreigners in criminality .

Based on a critical survey of the most important studies, the first part of this article outlines the different approaches referred to previously from the perspective of applied research; the second shows the limits and aporias existing in both these.Footnote 1

A Word from Statistics

The earliest research statistics systematically analyzing the involvement of foreigners in Italy in criminality were cited by Marzio Barbagli in his book Immigrazione e criminalità in Italia [Immigration and Criminality in Italy]; further revisions and new chapters were added in subsequent versions of this book (Barbagli 1998, 2002, 2008). In reference to the 2008 edition, which outstanding elements emerge from his research? First of all, statistical analyses of a broad range of crimes between 1988 and 2007 show foreigners (both legal and illegal) more involved than Italians in crime, or, more precisely, crimes of a violent and predatory nature since, as the author clearly states, no study was done of “white collar” crime, the exclusive prerogative of natives. Thus “in the past twenty years, the percentage of foreigners among those arrested and sentenced has increased greatly for all crimes. [This analysis focused on 19 predatory and/or violent crimes.] For some crimes this has doubled, for others tripled, and for some increased as much as sixfold” (Barbagli 2008, 53). The percentage of foreigners out of a total arrested for murder increased from 6 % in 1988 to 24 % in 2007; the same values are shown for attempted murder (from 5 to 32 %); assault and battery (from 5 to 29 %) and theft (from 14 to 49 %). The same progress is shown for other crimes, such as robbery (from 6 to 33 %), sexual violence (from 9 to 40 %), and “production, trafficking and sale of narcotics” (from 3 to 30 %).

The second important element regards the percentage of illegal foreigners responsible for crimes. Based on data from the Interior Ministry, the period between 1988 and 2006 shows a particularly high percentage of illegals among those foreigners accused for a broad range of crimes. As an example, during the last year examined, 2006, 74 % of foreigners accused of murder were illegals, 62 % of those accused of assault and battery, and 80 % of those accused of theft; the percentage of illegals runs from a minimum of 61 % for brawling to a maximum of 88 % for pickpocketing and petty theft (Barbagli 2008).

In comparing foreign and Italian offenders, the author concludes by stating “the data we have available leave no doubt regarding the fact that foreigners present in our country commit a quantity of crimes disproportionate to their number. From 1.4 % of the Italian population in 1990, they [the foreign population] grew to 5 % in 2007. But … in 2007 they made up between 25 and 68 % of reports, depending on the crime. Just as certain is that those committing these crimes are mostly those with no residence permits. … Therefore, if legal immigrants today commit crimes more frequently than natives (at least in certain age groups), illegals show many times higher levels of criminality than either of these groups” (ibid., 104–106) .

Let us look at the reference to “age groups” in the last statement. Here Barbagli is referring to the separation by age group among those foreigners accused and not, as some passages in the book might lead us to think, of separation by age group in comparing foreigners and natives. As he himself recognizes, gender and age are closely linked to the probability of committing crimes. As regards age, this probability increases in the age range between 14 and 20, and subsequently decreases as age advances. Foreigners in Italy actually show a different individual composition both in terms of gender (depending on the community, either balanced or unbalanced in favor of men or women) as well as age compared to natives. The appropriate comparison of crime levels between foreigners and Italians should therefore take into account these variables in the two spheres of reference (Ferraris 2008).

This aspect has however been considered in recent research by Gian Carlo Blangiardo (2010). His analysis initially focuses on reports for violent crimes (robbery and battery) of known foreign and Italian offenders in the three-year period between 2004 and 2006, standardizing the comparison between the two based on gender and age (the latter in two age groups of 18 to 29, and over 30). This comparison shows foreigners with a higher level of violent crime than natives, in the range of five to six per thousand for the former and one per thousand for the latter. A more detailed look according to a geographical partition by macro-regions shows Central Italy with the highest levels of criminality of foreigners compared to natives: 7.4 per 1,000 foreigners (legal and illegal) as compared to 1.2 for natives .

These differences become even more pronounced when the crime of theft is taken into consideration, with values in Central Italy equal to 11.9 known perpetrators per 1,000 foreigners compared to 0.8 Italians. Finally, based on categorization by national origin, the highest levels of violent crime are by immigrants from: Serbia and Montenegro (with an average provincial percentage of perpetrators at 18.4 per each 1,000 foreigners of the same nationality in Italy ); Tunisia (17.5), Morocco (11.1), Romania (5.9) and Albania (4.6); and the highest levels for theft go to Serbia and Montenegro (72.2), Romania (18.8), Tunisia (8.9), Morocco (8.1) and Albania (5.0) (Blangiardo 2010).

These results seem to show the greater criminal involvement by foreigners compared to Italians, though within the limits of the crimes examined by Blangiardo. However, some methodological observations carried out by Franco Pittau, coordinator of the research team at the Dossier Statistico Immigrazione of Caritas/Migrantes, lead us to believe this may not be the case.

In the words of Pittau: “We know that the comparison between these two populations is at times the principal if not the only reason for interest regarding this subject [immigration and criminality in Italy], but we should also understand that no reliable basis for comparison exists because, while complete data are known about Italians (above all, the number of resident citizens), the reference population of foreigners is instead difficult to determine, since this includes different categories” (Pittau 2010, 121) . In fact, the different typologies of people who compose the foreign component make overall quantification of this entity very difficult. These typologies include: resident foreign citizens; foreigners possessing a resident permit but not yet registered as residents; foreigners authorized for entrance in Italy but not included in the “immigrant” category, such as holders of visas for reasons of medical treatment, visits, or business; illegal foreigners; and finally foreign tourists.

This last group arrives in Italy, as Pittau states, “for the most part exempt from visa requirements (overall more than 25 million a year) and, although present only for brief periods, are not immune to the possibility of violating the law. … Given this is the state of things, it is impossible to establish the level of crime by legal foreigners, much less those by illegal foreigners. If a drug courier with a regular tourist visa is intercepted by the police; if a tourist with or without a visa steals something from a supermarket and is reported; … in these and other cases, the crime will always be recorded improperly as committed by an ‘immigrant’ and we cannot know exactly to which category this will be assigned” (ibid., 121–122).

Any statistical datum, either a value taken as is by the researcher or, on the contrary, extrapolated independently according to personal conceptual categories, is clearly the result of a process of analysis. Thus, the methodological observations of Pittau on the criteria for attribution of crimes by foreigners as performed by the Interior Ministry (distinguishing, as it seems, only between legal and illegal immigrants without considering the variety of categories of foreigners present in Italy ) leave room for a level of distortion in crime statistics (either crimes reported independently by law enforcement or reported by the population) that is not insignificant. On the one hand, this method of calculation overestimates the charges against foreigners since the numerator increases inaccurately, while on the other it is impossible to precisely define the size of the denominator (the reference population of foreigners) if these calculations only include the number of resident foreigners, those waiting for registration by the records office, and finally an estimate of illegals. In other words, those charged may also include foreigners in transit, yet this category is not taken into consideration in the denominator.

Though these observations draw from an approach based on statistical data analysis, other schools of research tend instead to show the basic non-reliability of criminal statistics. Let us examine why this is.

Social Stigmatization

Official statistics on crime fail to take into account the hidden number; these limits met with empirical confirmation only in the second half of the 20th century, when new reporting tools were perfected such as self-report studies and victimization surveys.Footnote 2 These new investigation tools made it possible to confirm that statistics on criminality diverged from the much greater number of crimes actually committed .

The limits of official statistics can be partially overcome by a comparison between known foreign and Italian perpetrators of crimes, in such as way as to exclude all those crimes labeled as by an “unknown offender,” though this would include most of them. Nevertheless, this comparison only apparently solves the problem. The moment we introduce a new aspect such as the discretionary power of the population and law enforcement in reporting crimes, the problem of the reliability of official statistics returns once again (Kitsuse and Cicourel 1963; Chapman 1971). Here are some “hypothetical” examples to help better understand this.

The owner of a supermarket discovers a person in the act of taking some goods from inside his sales area. He decides to call law enforcement and report the thief caught in the act. Later, when the trial begins, the owner will be called on to testify and report the events. As standard procedure, he will be called as witness (and plaintiff) the morning of the hearing as part of a list of other witnesses, each of them called based on proceedings scheduled for that morning. If everything goes as planned, he will be able to testify that same morning, otherwise, and not so improbably, his testimony will be postponed till the next hearing, since the proceedings scheduled for that day, for the same judge, have run long. Therefore, the supermarket owner will have to count on spending a half day, more likely two or maybe more, to give his testimony in the penal proceedings in which he is involved against his will. In light of all this, the next time our owner will clearly consider whether or not he should report the theft. The next time a thief has been identified, he may be heavily reprimanded, perhaps threatened with report to the police, but with full knowledge the affair will end then and there. Alternatively, the owner could adopt a discretionary attitude based on the person discovered stealing. If, for example, the thief is an elderly person from the neighborhood and someone he knows, he will limit himself to scolding him and recovering the merchandise; on the other hand, if he is an underage gypsy or immigrant, the owner will call law enforcement and begin the entire judicial process.

The second example regards the conduct of law enforcement in a mid-to-large Italian city. A neighborhood in the city center, having in short time become an area of street prostitution, has seen the birth of a citizens committee organized for the purpose of combating urban decay in the area where they live. Committee leaders send letters to newspapers to attract the attention of city politicians and public opinion to the decay, which, in their words, is no longer tolerable. Thanks to support from a few politicians from the local opposition and the contribution of several “friendly” newspapers, the mobilization of the committee achieves its initial results when the mayor personally calls the chief of police to request more law enforcement presence in that part of the city.

These hypothetical situations can be theoretically classified within the cognitive framework of the “tautology of fear” as outlined by Alessandro Dal Lago (1999). He expresses a definition of fear from the sociological standpoint. This “does not mean that specific elements of danger (physical or social) are distributed within a social environment, but rather that there is a collective legitimization and interpretation of more or less arbitrary evidence of danger as indisputable proof of a threat to the stability or existence of a society. From this viewpoint, foreigners (individually or as a group) are the most susceptible to being considered dangerous” (ibid., 9). Therefore, “the simple declaration of an alarm (in this case ‘the invasion of criminal immigrants’) demonstrates the reality that this itself reports” (ibid., 19). The author reiterates William Thomas’s classic theorem of the self-fulfilling prophecy: “if men define situations as real, they are real in their consequences.” Thus, even when a social actor incorrectly defines a situation with respect to reality, he will behave in a manner consistent with the accepted definition, creating the conditions for the consequences originating from that interpretation to actually take place.

From this standpoint, the social actors that concur, each in their own way, in the social construction of the frame “immigrant equals criminal” are of various types. These are in order: organized segments of civil society, like the above-mentioned citizens’ committees mobilized in the public arena in order to demand greater security where they live; “political entrepreneurs” or in other words those political representatives that instrumentally support requests for law and order to achieve an electoral consensus; the mass media, which emphasizes the social alarm posed by the presence of foreigners, if for no other reason than to increase newspaper sales or television viewership; and finally, the experts who give scientific dignity to common sense evaluations. Thus, within the pre-established frame, “the facts from news reports on crimes, which foreigners may or may not be responsible for, are nothing but empirical proof of a truth taken for granted in information from the mass media” (ibid., 12).

The above-mentioned viewpoint calls attention to the procedural and relational nature of social events, qualifying them as a process of social construction involving the different actors in play. More specifically, this approach recalls a tradition of research, in the field of the sociology of deviance, dating back to the theory of labeling. According to this, the causes provoking deviance can be found in the process of stigmatization imposed on socially disadvantaged minorities and groups (Becker 1997). Edwin Lemert explained this labeling process in two phases, differentiating between primary and secondary “deviation.” For the deviant, the former brings marginal consequences that are easily overcome. Instead, the latter occurs with the entry onto the field of the stigmatization process, establishing in the subject a reorganization of his identity in the direction of deviance (in the case examined here, the immigrants subject to the tautological mechanism of fear). The deviant falls into the role assigned to him and uses this as a tool for defense, attack or adaptation in the face of the labeling processes he has experienced (Lemert 1981).

Limits and Aporias in Social Research

As can be seen, we are dealing with two noticeably divergent interpretations and methodologies regarding causal factors at the root of the criminality of immigrants in Italy . The first, in reference to a research framework oriented toward a neo-positivist approach, starts from the statistical analysis of data. Once established that foreigners commit more crimes than natives, this works to identify the basic factors of this difference. Following Barbagli’s analysis, this involves a theory of either the conflict between cultures, social control, or relative privation; the latter, in his opinion, is the most convincing of the three perspectives (Barbagli 2008).

The second gives priority to qualitative research tools, and is instead interested in revealing (and deconstructing) the process of stigmatization experienced by immigrants in Italy, tracing their criminal involvement to a hostile and arbitrary institutional and social context. Exasperated by their constant subjection to law enforcement checks and searches, they are thus provoked into violating the rules, refusing to trust a State perceived as “absolutely unjust and underserving of their loyalty” (Palidda 1999, 86).

Even when scholars who are part of this research tradition rely on criminal statistics (something that rarely occurs for the reasons mentioned previously), they would probably trace the greater involvement of foreigners in criminality to the discretional and arbitrary nature of the checks and controls applied by law enforcement . Therefore, faced with identical evidences, corresponding to the higher statistical incidence of foreigners in criminality, we are given two explanations that are diametrically opposed depending on the type of research perspective. In fact, we could say that where statistics show the greatest involvement of foreigners in crime, the existence could be posed of a “crystallization” of the labeling process faced by them.

All this calls attention to the possibility of establishing a correspondence only within the same research approach between the “context of discovery” and the “context of justification,” between empirical observation of the phenomenon and the explanation of the same phenomenon. Put another way, we are dealing with a theoretical and interpretative “short circuit”—a sort of impasse—regarding the (possible) link between immigration and criminality . In fact, it would be difficult for those who adopt official statistics (reports) as a starting point when examining the crime levels of foreigners to take into account the possible discretional nature of law enforcement and the population in reporting crimes. If they were to do this, the criminal phenomenon (the “context of discovery”) to be analyzed would be seriously compromised internally by an intervening variable capable of invalidating the procedures of constructing the statistical data, or in other words the starting point from which the research proceeds. On the contrary, those who follow the theory of labeling presume that the “social reaction,” put into action according to various modalities, gives rise to the criminality of foreigners. In fact, according to this line of research, without a stigmatization process, we would not likely have what society defines as a “criminal individual.” This formulation thus reveals an explanatory circularity in which the research leads to conclusions that can be easily corroborated by the original presuppositions.

Considering what has been said till now, we can finally ask ourselves what correspondence can be shown, on the one hand, between the empirical formula furnished so far on the theme of “immigration and criminality” in Italy and, on the other, both of the two research perspectives.

The theory of labeling leaves several basic questions unanswered. First of all, this does not tell us why the process of discrimination is not translated, ceteris paribus, into greater criminal involvement for all the different foreign communities present in Italy. Unless law enforcement can be shown to give different treatment to different nationalities, it is difficult to explain why, for example, Chinese, Senegalese, and Nigerians, although in principal experiencing similar sorts of rejection from Italian society, fail to show similar values in terms of their involvement in criminality . Faced with this division of immigrants into nationalities of origin, the exponents of social constructivism would respond, and in principal correctly, by observing that the variable of “nationality” is probably one of the most spurious and artificial aspects. If anything, rather than taking this as a starting point, national origin should be the end point of an explanatory reasoning aimed at telling us how this has been categorized and constructed (Dal Lago 2006). The same observations are advanced by those who recall the fallacies of so-called “methodological nationalism,” based on studies founded on nationality as a pre-established variable, and who consider this an unquestionable fact once and for all (Beck 2003; Wimmer and Schiller 2003).

However, in spite of all the theoretical and epistemological limits that may derive from “methodological nationalism,” analysis by nationality can help us establish the notable differences in the incidence of crime between the different foreign communities. For example, in examining the data gathered by Blangiardo, why do Serbs and Montenegrins show an average percentage for theft of 72.2 per 1,000 foreigners (legal and illegal from the same nationality), while Filipinos and Chinese have a level of 0.8? These differences raise doubts, also based on the validity of critical observations advanced by Pittau, regarding the limits relative to the distribution of charges against “immigrants” by the Interior Ministry. Put another way, the notable differences by nationality can hardly be “explained” by referring, on the one hand, to an improper system of categorizing reports of different crimes by the Interior Ministry or, on the other, the process of discrimination experienced by immigrants in Italian society. This leads to the consideration that other aspects, such as for example the cultural universe of each foreign community and models of socialization into the values of the host society, may play a significant role in the involvement of foreigners in criminality.

In other ways, to presuppose, as Barbagli seems to hypothesize, the existence of different characteristics for illegal compared to legal immigrants (e.g., a greater propensity for risk and illegality of the former compared to the latter) fails to give adequate importance to the fact that between 1986 and 2002 five general amnesties were approved that involved around 1.5 million people. We could also add the hidden amnesties that in large part legalized ex post facto those who found themselves already in Italy and could show they had a work contract. Among these, the 2006 amnesty involved 520,000 people, and the one in 2009 gave legalization to 300,000 “domestics and caregivers” (Cnel 2008; Pittau 2010). Thus, in the past few decades, we have witnessed a continuous shift in the positions of foreigners from illegal to legal conditions and, most likely, as a result of the economic crisis and current regulations on immigration, from a legal to an illegal status.Footnote 3 In this sense, although illegal immigrants constitute the major percentage of foreigners involved in crimes, this group cannot be considered as a segment separate from the totality of immigrants present in Italy. If anything, we could hypothesize with some plausibility that it was their state of illegality that fed their criminality, putting them in contact with illegal opportunities that at times may be a functional alternative to institutionally sanctioned paths of social mobility (Merton 2000; Bell 1964; Becucci and Garosi 2008) .Footnote 4

To paraphrase Franco Cassano, one way of not bowing down in advance to the majesty of science is to make clear the problematic aspects fueling the scientific debate on the issue of “immigration and criminality” (Cassano 1989). This is precisely what we have attempted to do here, yet without adopting any perspective, as Benedetto Croce supported in his time, that would consider social research a pseudoscience.