Migration,not migrants,is the problem: Delinquency among migrants and non-migrants in Switzerland and ex-Yugoslavia

Abstract

The nexus between migration and crime has been studied over nearly a century across many countries from all continents. Research has concentrated on comparisons of migrants (or their offspring) with natives. Comparisons between migrants and comparable samples from their countries of origin have not been undertaken so far, however, because data were usually limited to the host country. The International Self-Reported Delinquency Study (ISRD-3, Enzmann et al., 2018) allows this gap to be overcome. In Switzerland, with its large immigrant minority – one student in two has roots in a foreign country – migrants of different backgrounds can be compared with native Swiss and with students who attend schools in ex-Yugoslavia where many migrants are from. We compare data on self-reported offences and victimization in the family collected through interviews with some 4000 juveniles in Switzerland and more than 6000 students of the same age in four countries of ex-Yugoslavia (Bosnia-Herzegovina, Serbia, Kosovo, Macedonia). Native-born youths in Switzerland report fewer offences than their immigrant peers, including those from ex-Yugoslavia. Although differences between students from ex-Yugoslavia and other foreign countries are relatively small, juveniles in ex-Yugoslavia report far lower offending rates than immigrants of the same age in Switzerland. Their rates are similar or lower than among native Swiss students. Further, rates of physical punishment and maltreatment are higher among immigrants than among non-migrants in Switzerland and in ex-Yugoslavia. We conclude that cultural background is unrelated to delinquency and parental punishment, but the experience of migrating goes along with violence within the family and self-reported offending. Differences exist between various family constellations, students born or with at least one parent born in Switzerland committing fewer offences and experiencing less parental violence.

Keywords

Juvenile delinquency migration migrants parental punishment self-report delinquency

Introduction

In traditional immigration countries, the nexus between migration and crime had started to interest criminologists as early as the beginning of the 20th century, both in the United States (Sellin, 1938) and in Latin America (Aebi, 2005). In Europe, aside from some isolated early examples (such as Exner, 1939), interest in this issue started only once migration from Southern to Northern Europe reached major proportions during the 1960s and 1970s (Tonry, 1997). Despite the differences in migration policies, the origin of migrants and the reasons for migration, the experiences and debates were similar across nations (Tonry, 1997). To the extent that authors observed higher offending rates among migrants than natives, explanations centred on cultural factors (Sellin, 1938), social disadvantages (Ross, 1937) or concentration of migrants in problematic neighbourhoods (Shaw and McKay, 1942). Interestingly, however, most studies from Europe found that, even according to official records, crime among migrants was lower or at least not higher than among native males of similar age, a pattern that changed more or less across the continent after 1975 (Killias, 1997). The change around 1975 in Europe suggests that migration may not go along with higher crime rates in all circumstances. Probably, much depends on who migrates from what background to which countries and with what plans and perspectives of integration. The unusually high offending rates among migrants from the UK in Argentina at the beginning of the 20th century (Aebi, 2005) further illustrate that migrants may represent a more or less favourably selected group from their home countries, depending on the historical context of migration.

Within Europe, Switzerland has, alongside tiny states such as Luxembourg or Liechtenstein, an unusually high proportion of residents born abroad. According to official records, Switzerland’s foreign population in 2016 reached 25 percent of the permanent resident population (FSO, 2018: 7). Citizens of four former Yugoslavian countries (Serbia, Macedonia, Bosnia-Herzegovina and Kosovo) make up 14 percent of Switzerland’s total population (FSO, 2018: 7), Kosovars being the largest group among them (BFS, 2017; FSO, 2018: 7; Sharani et al., 2010). Together, nationals from these four countries constitute the largest group of foreign residents. These four ex-Yugoslavian countries are also included in ISRD-3. Owing to ongoing high rates of naturalization, the ‘real’ share of residents with an immigrant background is likely to be close to 40 percent, however, particularly among younger age groups. In the sample interviewed for the ISRD-3 in 2013 (and used in the present study), 49 percent reported having immigrated or having at least one parent born abroad. This high rate of immigration has led to some social tensions, and immigration is among the most controversial political issues (Linder, 2011). National police statistics show that one criminal code offence in two is committed by a foreign national, of whom 41 percent are non-residents (Polizeiliche Kriminalstatistik, 2018: 27).

The situation of juveniles growing up in a foreign country received attention first in the USA, when, in her Yale doctoral dissertation under the direction of Thorsten Sellin (and largely abstracted in Sellin, 1938), Dorothy Krall coined the term ‘second generation’. In contrast to the first generation, the so-called second generation is seen as more prone to offending owing to confusing cultural norms, that is, a sort of a conflict between their parents’ imported cultural norms and those prevailing in the host society (Sellin, 1938). So far, the question has remained undecided whether children of migrants are (if at all) more delinquent because (1) they share more pro-violent (‘macho’) attitudes learned from their parents’ cultural norms, or because (2) difficulties resulting from migration itself (and all it entails for children growing up abroad) are the causes of their (eventually) higher criminal involvement. With data from ISRD-2, a first test allowed us to see, a few years ago, that juveniles living in Bosnia-Herzegovina tend to offend less than students in Switzerland (Killias et al., 2010). The present study allows, for the first time to our knowledge, a comparison of young migrants in the host country – originating both from ex-Yugoslavia or from other parts of the world – with native (Swiss) juveniles, as well as with young people growing up in their parents’ country of origin. Thanks to reasonably large samples in all the countries involved, and owing to the unusually high proportion of foreign residents in Switzerland, both from former Yugoslavia and from other parts of the world, statistically meaningful analyses are possible comparing all these sub-samples. In sum, Switzerland offers – within Europe – a particularly useful field of analysis of the question of whether migrant youths are (if at all) more delinquent because of cultural factors related to their origin, or whether circumstances due to their living conditions in the host country are the key factors.

In the ISRD of 2006, as well as in other surveys among Swiss students, juveniles of foreign origin were found to admit to more offences than their peers without an immigration background (for example, Killias, 1997, 2009; Killias et al., 2010; Vazsonyi and Killias, 2001; Vazsonyi et al., 2018). Most authors explained this phenomenon through problems of integration within West European societies, or by cultural or religious traditions (for example, Bui and Farrington, 2016; Helbling, 2008; Killias et al., 2010; Kohler, 2012).

In this article, we will first describe the design of the study.¹ Next, we will correlate offending rates with commonly used independent variables of social integration to assess whether, across sub-samples (that is, native and immigrant juveniles in Switzerland, and young people in ex-Yugoslavia), (1) delinquency rates differ, (2) the explanatory power of these variables differs, and (3) migration relates similarly to offending, physical punishment and maltreatment by parents.

Beyond descriptive statistics and bivariate analyses, we will test, through binary logistic regressions, whether the association between violent, property and minor offences, on the one hand, and classic independent variables, including gender and the experience of migration, parental control, having delinquent friends, religious affiliation, and weak self-control, on the other hand, holds across sub-samples.

Methodology

Design of the study

In the context of the Third International Self-Reported Delinquency Study (ISRD-3, Enzmann et al. 2018), juveniles enrolled in secondary schools, that is, in grades 7 to 9 and aged typically between 12 and 16 years, were interviewed in 40 countries. Among these countries was Switzerland, with a national probability sample, and four countries of ex-Yugoslavia, namely Bosnia-Herzegovina, Kosovo, Macedonia and Serbia, with samples drawn from schools in at least two major cities. In the four Balkan countries, the following cities were included: (1) in Bosnia-Herzegovina, Bihać, Mostar, Odžak, Sarajevo, Tuzla and Zenica; (2) in Kosovo, Pristina and Prizren, including some suburbs; (3) in Macedonia, Skopje and Kumanovo, including some suburbs; and (4) in Serbia, Belgrade and Novii Sad. For the present purposes, the samples from the four ex-Yugoslavian countries have been collapsed into a combined sample of 6269 students. The Swiss sample consists of 4158 students in all, of whom 2096 were born in Switzerland from two parents born in the country (that is, the ‘native’ Swiss sample), and 1979 were born abroad or had at least one foreign-born parent; 83 respondents did not indicate the birthplace of their parents or provided invalid answers. Among the students with a background of migration, 441 had their origins in former Yugoslavia.²

In the four Balkan countries, high schools were selected randomly from a list of schools provided by the Department of Education of the various countries, provinces or cities. Within each school, classes of grades 7 to 9 were selected randomly.³ In the target classes, all students present on the day of the survey who were willing to participate filled out the questionnaire. Participation among schools, classes and students was high. The schools’ response rate ranged (in the overall assessment) from 75 percent to 96 percent. Details can be found in Table 3 in the Appendix.

In Switzerland, a list of all eligible high schools (with grades 7 to 9) was obtained through the national authority in charge of coordinating the PISA surveys.⁴ In all, 132 schools (or 75 percent of all those contacted) participated in the Swiss national survey. Among the students enrolled in all participating classes, 92 percent actually responded to the online questionnaire.

Interview methods

The questionnaires were in computer-based form and interviews took place, according to local conditions, online or offline. In Switzerland, online interviews were available in all classrooms.⁵ Controlled experiments had previously confirmed that outcomes do not significantly differ between electronic and paper-and-pencil questionnaires (Lucia et al., 2007). During the interviews, teachers supervised and, if necessary, assisted their students – two previously conducted randomized controlled trials in Finland (Kivivuori et al., 2013) and in Switzerland (Walser and Killias, 2012) had shown that results do not differ across the mode of supervision (by teachers or by researchers).

In the ex-Yugoslavian countries, respondents answered the questionnaire offline using the FluidSurveys program.⁶ For this purpose, the research staff brought netbooks or, in Bosnia-Herzegovina, tablets to the classroom for the interview sessions. Research assistants were present during interviews.

The questionnaire

Central to ISRD-3 is a common questionnaire. A Steering Committee under the direction of Ineke Marshall developed it for all countries participating in this project (Enzmann et al., 2018). The core questionnaire comprises 11 modules covering delinquency, victimization and substance use. For the present purposes, questions on offending have been grouped into three categories (violent, property and minor offences). We also used the questions that measure ‘minor physical punishment by parents’ and ‘parental maltreatment’.

The questionnaire also includes many demographic and other independent variables derived from prevailing criminological perspectives. Also available is information on the personal history of migration, that is, the country of birth of the child and of both parents, and on religious affiliation, including the importance of religion in everyday life.

Variables used in the present analysis

To assess the role of migration and cultural identity in the development of delinquency, we will take into account several variables or indexes. We will use three indices of delinquency: namely violent, property and minor offences. ‘Violent’ offences is a dichotomized index of robbery and/or assault, committed at least once over the previous 12 months. ‘Property offences’ is an index of five offences committed at least once during the previous 12 months, namely burglary, motorbike/car theft, bicycle theft, breaking into a car and theft of personal belongings. ‘Minor’ offences stands for having committed, at least once over the previous 12 months, vandalism, graffiti, shoplifting, carrying weapons, group fights and cruelty towards animals.

Given the international character of ISRD-3, it has been necessary to measure migration in a way that is independent of national legislation about citizenship. Therefore, the questionnaire includes items on the birthplace of both the respondent and his/her parents. This offers the additional advantage of taking the family history of migration into account. For the present study, we have grouped our respondents into the following five categories:

(1) The respondent and both parents were born in Switzerland. This is the native-born sub-sample.

(2) The respondent and his/her father were born in Switzerland, but the mother was born abroad.

(3) Child and mother were born in Switzerland, but the father was born abroad.

(4) The respondent was born in Switzerland, but both parents arrived from abroad.

(5) The respondent and both parents were born abroad.

Beyond migration, for the present study we have used, among the many independent variables collected in the ISRD-3 instrument, variables that, from a theoretical point of view, might shape the association between migration and offending, namely parental control, delinquent peers, self-control, religious affiliation and physical punishment by parents. We will briefly explain how these variables were defined and recoded.

Concerning parental control over the child, we collapsed the following variables into an index:⁷ Do parents know (1) where their children usually are when they are out, (2) whom they are with and (3) what they are doing when they go out. We consider parental control to be weak if respondents answered that, for at least two out of these three items, their parents ‘never/seldom’ knew about their whereabouts. On the other hand, we rate parental control as strong if respondents answered that their parents ‘sometimes/often/always’ know about their whereabouts.

The questionnaire included questions about having friends who have consumed drugs and/or committed shoplifting, burglary, robbery and assault.⁸ We consider respondents with at least one among their peers having committed three or more different offences out of a list of five as ‘having delinquent friends’.⁹ According to this way of indexing, respondents with no friends having committed more than two offences are considered as having no delinquent friends. The cutting point of three offences avoids excessively high estimates.

Our self-control index is a shortened version of the scale by Grasmick et al. (1993).¹⁰ For the present purposes, self-control is ‘weak’ if respondents agree (fully or somewhat) with at least three out of nine statements about ‘low’ self-control, and all other respondents are considered as having ‘strong’ self-control. The cutting point of three out of nine items is in line with a reference for professionals working with adolescents provided by the American Psychological Association (APA, 2002). Empirically, it has the advantage of dividing the respondents into sub-samples of nearly equal size.

Religious affiliation is by essence a categorical variable. Religion has often been found to be an important aspect of social control (for example, Cretacci, 2003; Ellwood, 1918; McCullough and Willoughby, 2009). We will assess to what extent, in the context of other cultural variables, religion (and especially Islam) shapes social adaptation and delinquency.

Finally, we will also consider, as a form of parental control, minor physical punishment and maltreatment by parents. These variables have received increasing attention as risk factors for later delinquency (Averdijk et al., 2012; Farrington and Malvaso, 2019; Hoeve et al., 2009). Minor physical punishment includes, in the questionnaire, hitting, slapping or shoving (including also times when this was punishment for something a juvenile committed), whereas maltreatment covers all serious forms of physical aggression by parents, namely hitting with an object, punching or kicking forcefully, or beating up (including also times when this was punishment for something a juvenile committed).

Data analysis

In this article, we use, beyond descriptive and bivariate statistics, binary logistic regression analysis to assess whether origin is associated with delinquency (violent, property and minor offences) and whether or not there are differences between several constellations of migration of parents and children. Further, we will assess whether different forms of delinquency (that is, violent, property and minor offences) are associated in a similar or a different manner with cultural background. We will also assess whether dominant criminological theories, such as parental control, networks of delinquent friends and self-control, go along with delinquency in a similar way across cultures. Throughout the following tables and analyses, weighted data have been used for descriptive results and non-weighted data for the multivariate analysis.¹¹ For this analysis, we used the software package SPSS (Version 23.0) for statistical analysis.

Results

Delinquency and parental punishment by experience of migration

Figure 1 gives an overview of how migration relates to delinquency and parental punishment styles. The figure compares native Swiss juveniles (with both parents born in Switzerland), juveniles attending schools in ex-Yugoslavia and migrants (that is, students born abroad or with at least one parent born abroad) in Swiss schools. Among the migrants, we compare those with roots in former Yugoslavia and those from other backgrounds.¹² Although minor offences and minor physical punishments by parents are relatively common across sub-samples, serious forms of offending and victimization by parents seem to follow a clear pattern. For property and violent offences, as well as for parental maltreatment, young migrants have significantly higher rates than those without experience of migration, among both Swiss students and those living in ex-Yugoslavia.

Figure 1.

Offending and experience of physical punishment among youths in Switzerland and in ex-Yugoslavia, by experience of migration (percent).

One should keep in mind that property as well as violent offences include relatively serious offences only. Violent offences include robbery (mugging) as well as assault, the latter being defined, in the common ISRD-3 instrument, as bodily injury that, according to the respondent’s assessment, was likely to require some medical attention. Property offences include breaking into houses or cars, vehicle theft and theft of personal belongings, such as sports equipment or valuables. Shoplifting, which is fairly common among teenagers across all Western countries, is included among minor offences, where differences are small. The same is true with respect to milder physical punishment, although Swiss juveniles seem to experience this the least. On the other hand, parental maltreatment, which includes hitting with an object, punching, kicking forcefully or beating up, is far more common among juveniles whose families have experienced migration. Interestingly, and across all categories of offending and victimization by parents, juveniles of Swiss descent or living in ex-Yugoslavia show significantly lower rates than respondents with experience of migration. Further, children with roots in ex-Yugoslavia do not necessarily fair worse compared with those who arrived from other countries, or whose parents did. We will return to the issue of the cross-cultural validity of questionnaires such as those used in the ISRD-3. We would like to point out, however, that the distribution of serious problems including victimization does not seem to follow common-sense expectations. Most observers might be surprised to learn that growing up in ex-Yugoslavia goes along with fewer problems than living as a young migrant in Switzerland, and this independently of the origin. Further, although one study observed higher rates of parental violence among Hispanic migrants in the USA (for example, Antunes and Ahlin, 2018), the fact that migrant parents seem to abuse their children more often than their relatives in the countries of origin has not received any attention so far.

Patterns that are even more surprising appear when we look at different family histories of migration (Figure 2). Slightly more than half of respondents in Switzerland were born and have parents who were both born in this country (the ‘Swiss native’ sample). Among students with roots in a foreign country, 73 percent were born in Switzerland and 27 percent abroad. The ISRD-3 questionnaire allows us to assess the family history of migration sufficiently in detail¹³ to warrant further analyses with respect to delinquency and parental punishment styles. Since the focus is on the impact of different family histories of migration, the analysis is limited to respondents living in Switzerland.

Figure 2.

Offending and experience of physical punishment among youths in Switzerland, by family history of migration (percent).

Self-reported robbery and/or assault are, compared with the native Swiss sample, about six times as common among youths who arrived from a different country with their foreign-born parents. Among juveniles born in the country, rates are much more similar to the Swiss sample, despite some differences according to foreign origin of either the father and/or the mother. Minor and property offences, as well as the experience of minor physical punishment by parents, are similarly frequent among the several sub-samples, although the native Swiss sample has the lowest rates. Parental maltreatment, however, shows a similar pattern to self-reported violent offences, with foreign-born children of foreign-born parents reporting this kind of abuse more than four times more often than native Swiss juveniles. Overall, juveniles of foreign background but born in Switzerland report lower rates of parental abuse as well as of offending than do foreign-born respondents. Interestingly, and with the exception of minor offences, juveniles with a foreign-born father seem to experience more problems than those whose mother had arrived from abroad.

In sum, the personal and family history of migration obviously matters. Children born in Switzerland clearly have better odds of successfully integrating into the host society than those who were born abroad and arrived in the country later. It is not entirely clear why the father’s origin seems to be more important than the mother’s background. It might be that the father’s social and, presumably, professional background matters more for the family’s successful integration. The remarkable parallel between our respondents’ self-reported delinquent behaviour and their experience of parental punishment suggests that migration might be a stressful experience not only for juveniles but also for their parents. Physical punishment, as seen in Figure 1, is more than an ‘imported’ style of bringing up children; it is probably the result of the stress that immigrant parents might experience more often than other families. We will come back to this issue in the further analyses.

Delinquency by migration and social control

In Table 1, delinquency and parental punishment are analysed, for every sub-sample used in Figure 1, by the influence of parental control, self-control, peer influences, gender and religion. All these variables relate in some way to social control, in the sense that girls are more controlled than boys, that juveniles with delinquent peers experience less social control by conventional others, and religion – especially Islam – implies strict behavioural control in many aspects of everyday life. These variables may mediate the effects of migration in many relevant ways. The data presented in Table 1 indeed suggest this to be the case.

Table 1.

Self-reported delinquency, by different variables of social control (percent).

	Violent offences				Property offences				Minor offences
	Students with both parents born in Switzerland	Students in CH with at least one parent born abroad	Students in CH with at least one parent born in ex-Yugoslavia	Students in ex-Yugoslavian countries	Students with both parents born in Switzerland	Students in CH with at least one parent born abroad	Students in CH with at least one parent born in Ex-Yugoslavia	Students in Ex-Yugoslavian countries	Students with both parents born in Switzerland	Students in CH with at least one parent born abroad	Students in CH with at least one parent born in Ex-Yugoslavia	Students in Ex-Yugoslavian countries
	Percent^a	Percent^a	percent^a	Percent^c	Percent^a	Percent^a	Percent^a	Percent^c	Percent^a	Percent^a	Percent^a	Percent^c
Gender
Female	1.1	2.1	0.5	0.5	4.8	9.8	5.3	2.2	15.7	20.3	15.5	18.6
Male	3.1	8.7	10.5	2.7	12.0	23.2	33.0	4.2	35.7	39.5	41.5	27.2
N ^b	2078/2032	1891/1897	406/420	6049	2080/2033	1900/1902	408/422	6038	2072/2023	1880/1888	404/419	5974
p ⩽	.001/.000	.000/.000	.000/.002	.000	.000/.000	.000/.000	.000/.000	2.2	.000/.000	.000/.000	.000/.000	.000
Having delinquent friends
No^d	1.2	3.5	1.5	1.0	5.9	10.8	11.2	2.8	19.8	22.1	19.0	22.7
Yes^e	8.6	14.4	24.4	18.5	26.6	41.8	59.0	20.6	68.2	64.9	74.0	67.0
N ^b	2074/2030	1890/1896	406/420	5532	2079/2032	1898/1901	409/422	5522	2070/2021	1878/1887	404/419	5459
p ⩽	.000/.000	.000/.000	.000/.000	.000	.000/.000	.000/.000	.000/.000	.000	.000/.000	.000/.000	.000/.000	.000
Parental control
Strong^f	1.6	4.6	4.9	1.4	6.6	15.5	19.6	3.0	23.2	28.4	28.1	21.7
Weak^g	11.1	20.2	31.6	5.3	40.0	35.2	31.6	7.1	64.4	57.5	57.9	42.1
N ^b	2075/2029	1886/1893	406/419	6040	2076/2030	1897/1898	407/421	6029	2070/2020	1875/1884	403/418	5968
p ⩽	.000/.000	.000/.000	.000/.000	.000	.000/.000	.000/.000	.203/.042	.000	.000/.000	.000/.000	.005/.030	.000
Self-control
Strong self-control^h	0.3	0.7	1.4	0.4	2.4	6.2	5.4	1.4	12.6	12.3	8.8	13.4
Weak self-controlⁱ	3.9	8.4	9.3	2.6	13.9	23.7	29.6	4.9	38.0	41.4	41.5	31.0
N ^b	2017/1981	1818/1835	395/410	5993	2019/1982	1828/1839	396/411	5986	2013/1973	1812/1829	396/410	5932
p ⩽	.000/.000	.000/.000	.002/.001	.000	.000/.000	.000/.000	.000/.000	.000	.000/.000	.000/.000	.000/.000	.000
Religious affiliation
Islam	n/a	7.4	5.9	1.4	n/a	21.4	22.7	2.0	n/a	31.2	31.9	19.0
Atheist, any other religion	1.9	4.9	6.0	1.8	8.1	15.9	16.9	4.7	25.1	30.2	26.2	27.6
Total^b	2051/2015	1870/1882	402/419	6038	2054/2016	1879/1887	404/421	6027	2047/2006	1859/1873	399/418	5965
p ⩽	.000/.000	.053/.146	.969/.795	.294	.219/.543	.013/.058	.152/.100	.000	.098/.618	.713/.842	.220/.080	.000

Notes:

Weighted data.

Weighted/non-weighted data.

Non-weighted data.

No = having no friends who have committed more than two out of five offences.

Yes = having friends who have committed at least three out of five offences.

Strong parental control = parents know whereabouts ‘sometimes/always’ in at least two out of three respects.

Weak parental control = parents know whereabouts ‘seldom/never’ in at least two out of three respects.

Strong self-control = all other respondents who have not been considered as having ‘weak’ self-control.

Weak self-control = at least three out of nine statements were agreed.

Not surprisingly, boys commit more offences of all categories and in all sub-samples, with a few noteworthy nuances, however. For example, the gender gap is larger among juveniles from ex-Yugoslavia living in Switzerland than in any other group – including students living in ex-Yugoslavia, suggesting that problems of integration might be more salient among male migrants than among girls. Next, the influence of delinquent peers is consistent and strong across all types of offences and sub-samples. The same is true with respect to parental and self-control¹⁴, although these variables seem to be somewhat less important among migrants and juveniles living in ex-Yugoslavia, perhaps due to less variance in control exercised by parents and families in these sub-samples. Beyond these associations, the findings confirm the low delinquency rates in former Yugoslavia, particularly with respect to property offences (presumably due to fewer consumer goods). Further, having friends who committed at least three out of five different types of offences is far less common in ex-Yugoslavia. Regarding religion, it turns out that young Moslems are moderately more involved in violent and property offences among migrant students in Switzerland¹⁵, but not in former Yugoslavia, confirming that the effect of religion depends much on the context (Killias and Lukash, 2017). Minor offences are more common and less influenced by whatever variables.

Multivariate analysis

In Table 2, the impact of family constellations of migration on delinquency and parental punishment styles is analysed in the context of the same variables of social control used in Table 1. As the reference category, we use the sub-sample of respondents born abroad from two foreign-born parents. Since the focus is on different constellations of migration, the analysis is limited, as in Figure 2, to respondents interviewed in Switzerland.

Table 2.

Self-reported delinquency and parental punishment, by family constellation and different variables of social control.

		Violent offences		Property offences		Minor offences		Minor physical punishment by parents		Parental maltreatment
		N	R ²	N	R ²	N	R ²	N	R ²	N	R ²
		3705	.251	3711	.231	3691	.257	3743	.089	3746	.154
		Sig.	OR	Sig.	OR	Sig.	OR	Sig.	OR	Sig.	OR
Family constellation	Child born abroad, both parents born abroad^a	.00		.01		.11		.00		.00
	Child born in CH, both parents born in CH	.00	−2.7	.06	−1.4	.31	−1.1	.00	−1.5	.00	−3.3
	Child born in CH, father in CH, mother abroad	.01	−3.1	.60	−0.9	.34	−0.8	.93	−1.0	.19	−1.5
	Child born in CH, mother in CH, father abroad	.05	−2.0	.35	−0.8	.76	−0.9	.56	−1.1	.24	−1.4
	Child born in CH, both parents born abroad	.00	−2.7	.66	−1.1	.20	−1.2	.46	−0.9	.61	−1.1
	Weak parental control^b	.00	2.9	.00	2.7	.00	2.4	.00	2.1	.00	3.0
	Having delinquent friends^c	.00	5.0	.00	4.3	.00	5.2	.00	2.5	.00	3.5
	Weak self-control^d	.00	5.1	.00	3.5	.00	3.6	.00	1.8	.00	2.2
	Religious affiliation^e	.02	1.9	.01	1.6	.39	1.1	.55	1.1	.06	1.5
	Gender (1 = male)	.00	3.3	.00	2.1	.00	1.8	.01	0.8	.09	0.8
	Constant	.00	0.0	.00	0.0	.00	0.1	.00	0.2	.00	0.0

Notes: R² = Nagelkerke R Square; OR = Exp(B), calculated in the software package for statistical analysis of SPSS (Version 23.0).

Reference category.

Weak parental control (1 = at least two out of three statements were answered as ‘never’/’seldom’).

Having delinquent friends (1 = having friends who have committed at least three out of five offences).

Weak self-control (1 = at least three out of nine statements were agreed).

Religious affiliation (1 = Islam, 0 = Christian, atheist and any other religion).

The results show, in the upper part of Table 2, that violent offences, compared with foreign-born juveniles from two foreign-born parents, are significantly less common among all the other sub-samples, that is, among respondents born in Switzerland, independently of their parents’ origin. With respect to property offences, no significant differences appear between the reference group and the other sub-samples, although the difference from the Swiss native sample is near the .05 limit (p < .06). The rate of minor offences does not significantly differ across sub-samples, as also the visual impression in Figure 1 suggests. Physical punishments and maltreatment by parents differ significantly between the reference group and the Swiss native sub-sample, but not compared with respondents born in Switzerland with other backgrounds of migration.

The lower part of Table 2 shows the weight of the independent variables used in this analysis. For all three categories of offences, but also for parental punishment including maltreatment, the selected variables are significant, the strongest ones being weak self-control and having delinquent peers. Only religion does not significantly relate to minor offences and parental punishment styles. Visual inspection of Figure 2 indeed suggests the same conclusion. The associations between parental punishment and weak parental and self-control, as well as having delinquent friends, suggest that children who escape parental control, who are poorly self-controlled and who have delinquent friends have relationships that are more conflictual with their parents. Finally, girls seem to experience physical punishment by parents more often than boys, although the difference is not quite significant with respect to maltreatment.

Discussion

Many studies have assessed the extent of delinquency and other problem behaviours among migrants and natives. Whenever migrants seemed to have higher crime rates than the native-born population, it was not clear how to explain the gap. Sociologists focused on the discrimination or strain (owing to unfavourable socioeconomic conditions) that migrants experience in the host country, whereas cultural theorists argued that migrants were socialized in a different and allegedly more violent culture. A further possibility, rarely envisaged, would be that individuals who leave their home country just might not be average people but have a more problematic profile before they arrive in the host country. The present study addresses some of these conflicting views. Thanks to an unusual design, with the inclusion of samples in the former home countries of many migrants as well, we can assess delinquency and victimization not just for juveniles of Swiss and migrant backgrounds, but also for young people of the same age who attend school in the home countries of many migrants living in Switzerland.

As it turns out, juveniles born of Swiss parents have lower delinquency rates than those who were born in a foreign country or whose parents were. Indeed, roughly one in two students in Switzerland of the age group included in this study have some background in migration, illustrating the extent of migration to Switzerland over the last two generations. Interestingly, juveniles who were born in former Yugoslavia, or whose parents were, do not differ markedly from immigrant juveniles from other backgrounds. On the other hand, students of the same age attending schools in four of the former Yugoslav countries (Serbia, Macedonia, Bosnia-Herzegovina and Kosovo) had remarkably low rates of delinquency. Thus, our data do not support the cultural conflict perspective, nor do they support the view that violence among juveniles from former Yugoslavia might be the result of their exposure to armed conflicts during the civil wars in the Balkans.

The data do, however, support the view that migration itself may be a difficult experience for juveniles and their families. As the data on physical punishment (including maltreatment) by parents illustrate, physical abuse by parents is more prevalent among immigrant parents – whatever their origin – than among parents who were born in Switzerland, or among parents living in former Yugoslavia. As it seems, it may be hard to bring up children in a foreign country where support from the extended family is not available in critical situations.

Several alternative explanations are not in line with the data. For example, there is little room to assume that differential validity of self-report instruments across cultures might have affected the findings. If disclosure of unfavourable information was indeed discouraged in the Balkans, it is hard to see why migrants – particularly those from the same region – would admit more self-reported offences than would juveniles from native Swiss families. If the low rate of self-reported property offences in former Yugoslavia is in line with limited opportunities, that is, the lesser availability of many goods, it is not feasible to explain the findings regarding violence in this way. The match with the findings on parental maltreatment further supports the view that migrants indeed experience problem behaviour and violence within the family more than native Swiss juveniles or those living in ex-Yugoslavia. Further, the difference between migrants from former Yugoslavia and those from any other regions of the world is very small. This finding points, once more, to the possibility that the process of migration itself and not the cultural background matters in connection with delinquency and problem behaviour.

In the ISRD-3 and in the present study, much attention has been devoted to the control of independent variables that might interact with migration. As expected, gender, parental control, self-control and having delinquent friends all contribute to the explanation of violence, property and minor offences across all sub-samples. Interestingly, having delinquent friends is more important than self-control in most instances, but the difference is more marked among juveniles in former Yugoslavia than among those in Switzerland. Further, the family history of migration is also important. Children born in Switzerland have, in all comparisons, lower rates of delinquency than those who were born abroad. Children with at least one Swiss-born parent also have lower rates of delinquency and physical abuse than those with two parents from abroad. This suggests that gradual integration, through birth in the host country and the origin of at least one parent, widely mitigates problems related to migration. Finally, young Muslims have higher rates of delinquency in Switzerland but lower rates in former Yugoslavia (Killias and Lukash, 2017), suggesting that religion might interact with culture.

Summing up, criminological theories seem well able to explain delinquency in different settings and sub-groups of a society. However, the total amount of variance explained remains relatively modest and never exceeds 26 percent. In other words, despite some progress in the explanation of intercultural differences, many facets of youth crime continue to remain rather obscure.

Some caveats should also be kept in mind. The ISRD-3 instrument does not include direct measures of parental education, family income and other aspects of socioeconomic status (SES), mostly because such items did not work well in preceding waves of the ISRD across nations. Indirect measures of SES, such as unemployment of parents, welfare dependency or well-being of the family relative to others, do not contribute significantly to the explanation of delinquency in the multivariate models, presumably owing to the low prevalence of such forms of strain even among the most disadvantaged groups of migrant respondents (with, for example, an unemployment rate that does not exceed 4.5 percent even among the most disadvantaged groups).

Further, the ISRD instrument does not include items on experienced discrimination in everyday life. It cannot be ruled out, therefore, that some forms of more subtle discrimination provoke offending as a reaction. However, the data used here refer to self-reports rather than official statistics and should, therefore, not be biased owing to police discrimination. The European and, especially, the Swiss situation may differ in many important respects from the US experience. The conclusions should, therefore, be generalized to other contexts with extreme caution. The data used here cannot answer whether migrants are different from those who do not leave their countries. However, the migrants among our respondents came to Switzerland as children or were even born there. In other words, if selection related to delinquent tendencies were taking place whenever a person decides to leave his/her country, it would have operated at the level of our respondents’ parents. Further, the lower delinquency scores among respondents born in the host country, as well as among those with at least one parent of Swiss descent, suggests that successful integration is far from impossible. The challenge will be, for now and in the future, how to make this possible on a larger scale.

Conclusion

Several results may come as a surprise to our readers. They were challenging also for us, although an earlier paper (Killias et al., 2010) pointed to the possibility that juveniles in former Yugoslavia might be far more law-abiding compared with young migrants from that area living in Switzerland. With far larger samples and no longer limited to just one ex-Yugoslav country (Bosnia-Herzegovina), the present study allows conclusions that are more valid and more robust. What, however, should be the conclusion? If we know that the process of migration itself rather than cultural background is the problem, policy-makers can focus even more on promoting the integration of young migrants into the host society. In other words, the problem is not where people come from but what kind of opportunities and difficulties they encounter in their new environment. This may be helpful as a starting point for new policies, although concrete answers to the challenges of living alongside migrants in daily life remain to be developed.

Footnotes

Appendix

Table 4.

Independent variables (Switzerland and in ex-Yugoslavia), frequencies and in percent.

	Switzerland (CH)	Swiss students with both parents born in Switzerland	Students in CH with at least one parent born abroad	Students in CH with at least one parent born in ex-Yugoslavia	Ex-Yugoslavian countries
	Percent^a	Percent^a	Percent^a	Percent^a	Percent^c
Gender
Female	51.3	53.9	48.3	45.1	50.1
Male	48.7	46.1	51.7	54.9	49.9
Total^b	4155/4156	2124/2096	1950/1977	423/441	6266
Family constellation
Child born in CH, both parents born in CH	52.8
Child born in CH, father in CH, mother abroad	8.6
Child born in CH, mother in CH, father abroad	9.0
Child born in CH, both parents born abroad	18.2
Child born abroad, both parents born abroad	11.4
Total^b	4001/3988
Having delinquent friends
No^d	85.0	89.3	80.7	80.9	96.3
Yes^e	15.0	10.7	19.3	19.1	3.7
Total^b	4123/4092	2105/2063	1935/1947	420/435	5637
Parental control
Strong^f	94.7	95.8	93.7	95.4	94.4
Weak^g	5.3	4.2	6.3	4.6	5.6
Total^b	4146/4141	2119/2088	1946/1972	421/439	6254
Self-control
Strong self-control^h	44.0	50.0	38.2	37.7	46.4
Weak self-controlⁱ	56.0	50.0	61.8	62.3	53.6
Total^b	3962/3945	2036/2001	1847/1864	401/417	6054
Religious affiliation
Islam	9.1	n/a^k	18.9	58.8	54.9
Atheist; any religion	90.9	99.8	81.1	41.2	45.1
Total^b	4106/4119	2099/2079	1930/1962	419/440	6254

Notes:

Weighted data.

Weighted/non-weighted data.

Non-weighted data.

No = having no friends who have committed more than two out of five offences.

Yes = having friends who have committed at least three out of five offences.

Strong parental control = parents know whereabouts ‘sometimes/always’ in at least two out of three respects.

Weak parental control = parents know whereabouts ‘seldom/never’ in at least two out of three respects.

Strong self-control = all other respondents who have not been considered as having ‘weak’ self-control.

Weak self-control = at least three out of nine statements were agreed.

During previous 12 months.

The number of cases is too low in this group (N = 4/6).

Acknowledgements

Many people helped, at one time or another over recent years, to realize this research project. In Switzerland, our first thank goes to Dr Kathrin Weder (Swiss National Science Foundation), Mr S. Sommer (Jacobs Foundation), Dr M. Peek (Swiss Federal Office of Migration), Direction of the Conference of Education Departments, namely former cantonal Minister I. Chassot, Secretary General H. Ambühl and Dr T. Mattmann-Arnold, and Cantonal Departments of education.

In Bosnia-Herzegovina, we thank Dr Almir Maljevic (Associate Professor of Criminal Law, University of Sarajevo, Faculty of Criminal Justice Sciences)

In Kosovo, we thank Mensut Ademi (PhD candidate, University of Pristina; Executive Director of Kosovo Association of Criminal Justice/Criminology)

In Macedonia, we thank Dr Natasha Jovanova (Assistant Professor), Professor Dr O. Bacanovich (Dean), J. Ivanovski (Assistant Professor) and A. Nedev (Teaching Assistant), all from the Faculty of Security-Skopje, University St. Kliment Ohridski – Bitola

In Serbia, we thank Professor Dr Vesna Nikolic-Ristanovic and Ljiljana Stevkovic (PhD candidate), both at the Faculty of Special Education and Rehabilitation, University of Belgrade.

We further acknowledge the kind and generous technical support that we received, throughout this project, from other members of the ISRD Steering Committee and in particular from Professor Dr I. Haen Marshall, Professor Dr D. Enzmann, Professor Dr M. Steketee and Professor Dr J. Kivivuori

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was supported by the Swiss National Science Foundation (grants N 100015_138401/1 and 10001A_162816), by Jacobs Foundation (grant 2012-1026) and by the Swiss Federal Office of Migration.

Notes

References

Aebi

(2005). Immigration et délinquance: Le mythe du conflit de cultures. In Queloz

, et al. (eds) Délinquance des jeunes et justice des mineurs. Les défis des migrations et de la pluralité ethnique. Bern: Stämpfli/Brussels: Bruylant, 97–123.

Antunes

MJL

Ahlin

(2018) Minority and immigrant youth exposure to community violence: The differential effects of family management and peers. Journal of Interpersonal Violence . First published online, 3 February, doi.org/10.1177/0886260518755491.

APA (American Psychological Association) (2002) Developing Adolescents: A Reference for Professionals. Washington, DC. URL (accessed 1 February 2019): http://www.apa.org/pi/families/resources/develop.pdf.

Averdijk

Malti

Eisner

Ribeaud

(2012) Parental separation and child aggressive and internalizing behavior: An event history calendar analysis. Child Psychiatry Human Development 43: 184–200.

BFS (Bundesamt für Statistik) (2017) Die Bevölkerung der Schweiz 2016. 348-1600. URL (accessed 1 February 2019): https://www.bfs.admin.ch/bfsstatic/dam/assets/3902098/master.

Bui

Farrington

(2016) First-generation immigrants feel socially excluded and have greater pro-violence attitudes than the native population in England and Wales. Journal of Aggression, Conflict and Peace Research 8(1): 46–60.

Cretacci

(2003) Religion and social control: An application of a modified social bond on violence. Criminal Justice Review 28(2): 254–277.

Ellwood

(1918) Religion and social control. The Scientific Monthly, 7(4): 335–348.

Enzmann

Kivivuori

Marshall

Steketee

Hough

Killias

(2018) A Global Perspective on Young People as Offenders and Victims: First Results from the ISRD3 Study. Cham, Switzerland: Springer International Publishing.

10.

Exner

(1939). Kriminalbiologie in ihren Grundzügen. Hamburg: Hanseatische Verlagsanstalt.

11.

Farrington

Malvaso

(2019) Physical punishment and offending in two successive generations of males. In: Bryce

Robinson

Petherick

(eds) Child Abuse and Neglect: Forensic Issues in Evidence, Impact, and Management. London: Elsevier (in press).

12.

FSO (Federal Statistical Office) (2018. Statistical Data on Switzerland 2018. URL (accessed 1 February 2019): https://www.bfs.admin.ch/bfsstatic/dam/assets/4522209/master.

13.

Grasmick

Tittle

Bursik

Jr Arneklev

(1993) Testing the core empirical implications of Gottfredson and Hirschi’s General Theory of Crime. Journal of Research in Crime and Delinquency 30(1): 5–29.

14.

Helbling

(2008) Islamophobia in Switzerland: A new phenomenon or a new name for xenophobia? Paper prepared for Annual Conference of the Midwest Political Science Association (MPSA) Chicago, 3–6 April 2008. URL (accessed 1 February 2019): https://www.unil.ch/files/live/sites/issr/files/shared/8._Telechargement/Colloque_de_Recherche/Helbling_2010_Islamophobia.pdf.

15.

Hoeve

Dubas

Eichelsheim

van der Laan

Smeenk

Gerris

JRM

(2009) The relationship between parenting and delinquency: A meta-analysis. Journal of Abnormal Child Psychology 37: 749–775.

16.

Killias

(1997) Immigrants, crime, and criminal justice in Switzerland. In: Tonry

(ed.) Ethnicity, Crime, and Immigration: Comparative and Cross-National Perspectives. Chicago: University of Chicago Press, 375–406.

17.

Killias

(2009) Paradise lost? New trends in crime and migration in Switzerland. In: McDonald

(ed.) Immigration, Crime and Justice. Bingley: Emerald, 33–45.

18.

Killias

Lukash

(2017) Does religion make a difference? Juvenile delinquency in Switzerland and four Balkan countries. In: Bijleveld

van der Laan

(eds) Liber Amicorum Gerben Bruinsma. The Hague: Boom criminologie, 193–200.

19.

Killias

Maljevic

Lucia

(2010) Imported violence? Juvenile delinquency among Balkan youths in Switzerland and in Bosnia-Herzegovina. European Journal on Criminal Policy and Research 16(3): 183–189.

20.

Kivivuori

Salmi

Walser

(2013) Supervision mode effects in computerized delinquency surveys at school: Finnish replication of a Swiss experiment. Journal of Experimental Criminology 9(1): 91–107.

21.

Kohler

(2012) Cultural integration in Switzerland. In: Algan

Bisin

Manning

Verdie

(eds) Cultural Integration of Immigrants in Europe. r. Oxford: Oxford University Press, 210–259.

22.

Linder

(2011). Multicultural Switzerland and the challenge of immigration. Journal of Minority Studies 5: 201–225.

23.

Lucia

Herrmann

Killias

(2007) How important are interview methods and questionnaire designs in research on self-reported juvenile delinquency? An experimental comparison of Internet vs. paper-and-pencil questionnaires and different definitions of the reference period. Journal of Experimental Criminology 3(1): 39–64.

24.

McCullough

Willoughby

BLB

(2009) Religion, self-regulation, and self-control: Associations, explanations, and implications. Psychological Bulletin 135(1): 69–93.

25.

Polizeiliche Kriminalstatistik (2018) Jahresbericht 2017 der polizeilich registrierten Straftaten. Sektion Kriminalität und Strafrecht. Herausgeber - Bundesamt für Statistik (BFS). Neuchâtel. URL: https://www.bfs.admin.ch/bfsstatic/dam/assets/4822913/master.

26.

Ross

(1937) Crime and the native born sons of European immigrants. Journal of Criminal Law and Criminology 28(2): 202–209.

27.

Sellin

(1938) Culture Conflict and Crime. New York: Social Science Research Council.

28.

Serajzadeh

(2001–2002) Islam and crime: The moral community of Muslims. Journal of Arabic and Islamic Studies 4: 111–131. URL (accessed 1 February 2019): https://www.lancaster.ac.uk/jais/volume/docs/vol4/4_111-131serajzade1.pdf.

29.

Sharani

Efionayi-Mäder

Hammer

Pecoraro

Soland

Tsaka

Wyssmüller

(2010) Die kosovarische Bevölkerung in der Schweiz. Herausgeber: Bundesamt für Migration (BFM). URL (accessed 1 February 2019): https://www.sem.admin.ch/dam/data/sem/publiservice/publikationen/diaspora/diasporastudie-kosovo-d.pdf.

30.

Shaw

McKay

(1942) Juvenile Delinquency and Urban Areas. Chicago: University of Chicago Press.

31.

Tonry

(ed.) (1997) Ethnicity, Crime, and Immigration: Comparative and Cross- National Perspectives. Chicago: University of Chicago Press (Crime and Justice 21).

32.

Vazsonyi

Killias

(2001) Immigration and crime among youths in Switzerland. Criminal Justice and Behavior 28(3): 329–366.

33.

Vazsonyi

Javakhishvili

Ksinan

(2018) Routine activities and adolescent deviance across 28 cultures. Journal of Criminal Justice. 57: 56–66.

34.

Walser

Killias

(2012) Who should supervise students during self-report interviews? A controlled experiment on response behaviour in online questionnaires. Journal of Experimental Criminology 8(1): 17–28.