Are Children of Immigrants Born in Sweden More Law-Abiding Than Immigrants? A Reconsideration

Abstract

Previous Swedish studies have found that second-generation immigrants participate in offending less than first-generation immigrants, which is a striking finding from an international perspective. The aim of this study is to critically investigate the idea that second-generation immigrants have lower rates of registered offending than first-generation immigrants. The population of the present study consists of all 4.4 million persons who were registered as living in Sweden at the end of 1996. These people were followed for 5 years. During this time, 280,000 persons were registered for 1.5 million offenses. As far as the participation in overall offending is concerned the previous results were reproduced. However, when focusing on the offending rates in different offense types, the first-generation immigrants have the highest risks of being registered for violent crimes and second-generation immigrants with no Swedish-born parents have the highest risks for property crimes. The pattern found is further discussed in relation to the risk of second-generation immigrants with no Swedish-born parent to develop a criminal career and thereby frames the problem of criminality in culture differences, immigration policy, and so on.

Keywords

immigration and crime criminological theories race class and gender articulation longitudinal crime trends bias in the criminal justice system race and public opinion treatment by the police

Introduction

In public and political debate, immigrants are often presented as being more crime-prone than the indigenous population. The question about the linkage between immigration and crime is an old criminological question that classical criminologist such as Shaw and McKay, Sellin, and Sutherland have all posed (Ajzenstadt & Shapira, 2012; Bersani, 2012; Martinez & Lee, 2000). But, it is still an important area of research and a global issue (Gabbidon, 2010; Haen Marshall, 1997; Martinez & Lee, 2000; Tonry, 1997).

From classical criminological theories addressing differential opportunity structures, criminal subcultures, and social disorganization, it also seems plausible to predict higher crime rates among immigrants (Reid, Weiss, Adelman, & Jaret, 2005; Martinez & Lee, 2000). In the United States, the link between immigration and crime has a long research tradition dating back to the 1920s and even earlier (Bersani, 2012; Bursik, 2009; Martinez & Lee, 2000; Mears, 2002; Tonry, 1997). A lot of empirical research has accumulated throughout the years in which different kinds of data with varying quality on both the individual level and the macro level have been used and analyzed (Bersani, 2012; Mears, 2002). It does not come as a surprise, that different studies have arrived at contradictory conclusions. However, the bulk of the empirical studies done in the United States over the last century have, contrary to public opinion and theoretical assumptions, found that the immigrants engage less in crime than the native-born (Martinez & Lee, 2000; Reid et al., 2005; Wortley, 2009). From various macro-level studies on the immigration–crime nexus, it has been concluded that a high concentration of the foreign-born in a metropolitan or a suburb area does not lead to an increase in the crime rate, but stabilizes or reduces the rates. In some cases, a concentration of the foreign-born might be vitalizing for the whole neighborhood and thereby suppress the crime rate (Bersani, 2012; Reid et al., 2005; Sampson, 2008). Hence, both macro-level and individual-level studies have found lower crime involvement among foreign-born immigrants than among native-born Americans.

Research on the immigrant–crime nexus has also attracted many criminologists in some of the western European countries, but this research has a much shorter history than in the United States (Killias, 2011; Tonry, 1997). Immigration to western European countries dates back to the years after World War II. Until the beginning of the 1950s, most European countries with some exceptions (e.g., Switzerland) used to be ethnically homogenous. The first wave of immigrants who arrived in the 1950s and 1960s were labor force immigrants (guest workers) and their families. From the 1980s onward, refugee immigrants were received (as asylum seekers) in successive waves. During a few decades, the number of immigrants has increased and thereby changed the ethnic composition radically. A great deal of the criminological studies on immigrants’ criminal involvement in western European countries is based on official crime statistics. Studies have usually revealed that immigrants tend to have higher crime rates than the indigenous population. However, as pointed out by Killias (2011), it has not always been so. Before 1970, the crime statistics rather showed that the immigrants, the majority being labor force immigrants, had lower crime rates than the indigenous population (see also Junger-Tas for the Netherlands, Albrecht for Germany, and Killias for Switzerland in Tonry, 1997). At that time, the immigrants’ motivation to migrate to another country was to find new and better opportunities for oneself and one’s family, or even a part of a career. In a critical analysis of the official police statistics for suspected persons in Germany, Geissler (2008) draws the conclusion that the nucleus of the German immigrants, the “migrant workers” (Arbeitsmigranten¹), for example, the immigrants living permanently in the country and being in the work force, are at least as law-abiding as the native Germans. After having controlled for some relevant background factors, the migrant workers turned out to be even more law-abiding than the natives. Their lower criminality might even have contributed to suppressing the crime trend in the population.

Offending among the offspring of immigrants born in the new country, the second-generation immigrants, is another issue that researchers of the immigrant–crime nexus in the United States and western Europe have called much attention to. In the United States, sociologists of the early years in the 20th century noticed that the children of the immigrants tend to be more involved in criminal activities than their foreign-born counterparts (Bersani, 2012; Bursik, 2009). The research on this issue, which was revitalized in the 1960s in the United States, has since then repeatedly found empirical evidence that whereas the first-generation immigrants have lower crime rates than the native population, the second generation often reach a crime level in the vicinity of their naive-born counterparts or even higher (Bersani, 2012). Similar conclusions for second-generation immigrants have been drawn from the research results reported by western European criminologists (Killias, 2011; Tonry, 1997). To explain the higher crime rates among the second-generation immigrants, researchers have referred to theories of alienation, blocked opportunities, social disorganization, relative deprivation, deviant subcultures, and so on (Tonry, 1997). Similarly, Killias (2011) called attention to the fact that social research has found that immigrants, more often than natives, live in disadvantaged neighborhoods where the children have limited opportunities for a successful school career and the risks are greater to associate with delinquent friends.

Sweden as an Immigrant Country

Sweden has gradually developed to a multiethnic country during the last 60 years. Among the 9 million civic registered inhabitants in Sweden, every eighth person was born in a foreign country. Compared to other European countries, the influx of foreign-born in the population is large. During and shortly after the Second World War, immigrants were refugees from the neighboring Nordic and Baltic countries. In the 1950s and 1960s, Swedish industry rapidly expanded, resulting in a shortage of labor force. Laborers, therefore, were welcomed from other countries such as West Germany, Austria, Italy, Greece, Spain, Turkey, and former Yugoslavia. By the late 1960s, the labor market was saturated and immigration was highly restricted. An open labor market within the Nordic countries still remained, resulting in continuous immigration from Finland in particular. Between 1970 and the early 1990s, the influx of refugees and their families from countries outside of Europe increased. In the early 1990s, the number of foreign-born from countries outside Europe passed the number born in a Nordic country other than Sweden. After Sweden joined the European Union in 1995, there has been a larger influx of labor force immigrants from the other European countries. The immigrants today come from more than 200 countries, the dominating countries being Finland, the former Yugoslavia, Iran, Bosnia–Herzegovina, Iraq, Poland, Norway, Denmark, Germany, and Turkey. Furthermore, the proportion of Swedish-born persons whose parents were born in a country outside Europe has gradually increased and will continue to do so in the future (Statistics Sweden, 2003).

Swedish Research on Immigrant Crime

The Swedish research on the immigrant–crime nexus dates back to the 1970s and was based mainly on official data on police-registered offending and convictions. The early studies defined immigrants on the basis of citizenship, with the consequence that the naturalized foreign-born were statistically treated as native Swedes. The definition also made it impossible for the researchers on the immigrant–crime nexus to single out second-generation immigrants and to discuss how Sweden might differ from other countries regarding the second-generation immigrants’ offending pattern (Kardell, 2011; in press). The first studies on offending among second-generation immigrants versus first-generation immigrants and indigenous Swedes were presented in the mid-1990s. The first study by Ahlberg (1996) was a statistical analysis of the criminal offenses reported to the police during a 5-year period concerning the whole Swedish population between 15 and 55 years of age. In contrast to the findings in the United States and other western European countries, Ahlberg found lower crime involvement among the second-generation immigrants than among the first-generation immigrants. The indigenous Swedes had the lowest involvement. This striking finding that second-generation immigrants engage less in criminal activities than the first generation has been reproduced in subsequent Swedish research for both convicted persons and crime suspected persons (Kardell, 2011; in press; Swedish National Council for Crime Prevention, 2005b; von Hofer, Sarnecki, & Tham, 1997). Ahlberg’s results received international attention and were further commented upon by Tonry (1997) and later also by Hiatt (2007). Tonry (1997) draws the conclusion from the Swedish results that a country’s settlement policies might have a suppressing influence on the crime rates in the next immigrant generation. This interpretation has fallen into good ground among the domestic criminal justice authorities in Sweden and has been presented as being a success for the country’s welfare system.

In a more recent comparison between convicted persons in the Nordic countries of Denmark, Norway, and Sweden, where the studies in the three countries were designed in practically the same way, the previous results for Sweden were not reproduced (Kardell & Carlsson, 2009). The striking finding was that the second generation had the highest conviction rate and the indigenous population the lowest rate. After controlling for age, second-generation immigrants had about the same conviction rate as the indigenous population. These recent results seem closer to the research results on the issue reported in other western European countries and the United States. The higher rates among the second-generation immigrants reported in the recent studies may indicate that a change has taken place in the Swedish society where integration into Swedish society is now more difficult.

Aim and Scope of the Study

The aim of this study is to critically address the conclusion that second-generation immigrants in Sweden are less involved in crime than first-generation immigrants. Is it possible to state from the reported findings that the Swedish-born offspring of immigrants are more law-abiding than immigrants? In this article, results from the study of the Swedish National Council for Crime Prevention (2005b) will be analyzed in greater depth. The reported research findings are remarkable even in a Swedish context, when one takes into account that social research in Sweden has concluded that immigrants and their families often live in more disadvantaged neighborhoods and their life conditions have changed to the worse during the economic crisis in 1990s (Kommittén Välfärdsbokslut & Palme, 2002).

Many studies from other countries on offending among the second-generation immigrants have been restricted to self-reported information on criminal activities and immigrant background from youth living in urban neighborhoods (Bersani, 2012). In the present study, the entire civil-registered people and their police-registered suspects of crime are considered. All of those who are unregistered and their suspected offenses are excluded; this includes asylum seekers, tourists, illegal immigrants, visitors who have purposely entered the country to commit crime, and so on. The population in the study was set to comprise all persons who were civilly registered in Sweden on December 31, 1996, and were born between the years of 1945 and 1981. The number of persons totaled almost 4.4 million between 15 and 51 years of age. The registered offending among the population members was followed up for the 5-year period of 1997–2001. Information about the population members such as country of birth, parents’ country of birth, citizenship, educational level, earned income, income from social allowances, and so on, and other allowances was obtained from the Register Over the Total Population and the LOUISE Database, which is a longitudinal database with information on education, income, and employment for the whole population in Sweden (Statistics Sweden, 2002).

Registered Offending

The data for registered offending were obtained from the Register of Persons Reasonably Suspected of an Offence, which means that the police have opened a file for the case. In order to cover the whole spectrum of criminal offenses specific types of crime are created from the crime headings of the Swedish Penal Code. This makes it possible to detect eventual different offending profiles between different populations. We use the categories of all offences (all offenses taken together), violent offences (completed and attempted murder, manslaughter and assault resulting in death, assaults not resulting in death, unlawful threat, rape and other sexual offenses, robbery, violence and threats to public servants, possession of weapons), property crimes (such as crimes inflicting damage, car theft, theft out of or from motor vehicles, burglary into dwellings, etc., theft from shops and department stores), fraud, and so on, crimes against the Law of Road Traffic Offences Act (comprising only the offenses that may impose imprisonment) and crimes against the Narcotics Drugs (Punishment) Act. Almost 280,000 persons in the population were found to be registered for at least one offense against the Penal Code and against other penal offenses. The number of offenses for which the suspected were registered during the 5-year period totaled almost 1.5 million. During the same period, the total number of offenses reported to the police reached almost 6 million.²

It should be mentioned that of all the police-registered offenses during the period of 1997–2001, the estimation is that 7% were for persons who were not civilly registered in Sweden (Swedish National Council for Crime Prevention, 2005b). This estimation has more than doubled since the previous study from the Swedish National Council for Crime Prevention (Ahlberg, 1996). The breakdown of the Communist regimes in Eastern Europe, the free movement within the Schengen zone, and expanded traffic networks that made it easier to reach Sweden from Eastern and Continental Europe can partly explain the increase. Also, the number of asylum seekers in the country had increased in the later period.

Immigrant and Swedish Background

We create subpopulations based on immigrant and indigenous background. The information on the individual’s country of birth and their parents’ country of birth are obtained from the official population register. Individuals born in Sweden are disaggregated into three categories: those with both parents born in Sweden (native population), those with one Swedish-born and one foreign-born parent (second-generation immigrants with one Swedish-born parent), and those with two foreign-born parents (second-generation immigrants without a Swedish-born parent). The foreign-born individuals are treated as a fourth category (first-generation immigrants). Persons born and raised in Sweden in a family where one of the parents was also born and raised in Sweden may integrate easier into the Swedish society than persons with two foreign-born parents. Ramakrishnan (2004) suggested the concept of the “two-and-a-half generation” to designate those with a foreign-born and a native-born parent.

Control variables

Three background variables, often pointed out in the criminological literature as correlating with offending, are considered here: gender, age, and education level. The proportion of males and females in the population is 51% for males versus 49% for females. Among the Swedish-born categories (the native population and the second-generation immigrants), the proportions are about the same as in the total population. However, among the immigrants the proportion of females is slightly higher than for males, 49.5% for males versus 50.5% for females. Age is measured at the starting point of the observation period in 1997. The population members in the study concerning the suspicion data were between 15 and 51 years of age at the beginning of 1997. The median age is 33.6 years for the whole population. The second-generation immigrants without Swedish-born parents is the youngest category with a median age of 28.5 years. Twenty-two percent of this category is less than 21 years of age and 19% is older than 40 years. Immigrants have the highest median age of 35.2 years. Not quite 10% of this category is younger than 21 years of age and almost a third are older than 40 years of age.

Education level can be seen as an indicator of both socioeconomic status and a risk or protective factor. Education level is measured during the year 2000. About 10% of the immigrants have no information about formal education, whereas the percentage missing in the other groups is considerably lower. A scale of education level ranging from 0 to 5 is used, where lack of information is given the score of (0), 9 years or less of primary school is set to the score of (1), 2 years of secondary education or a corresponding vocational education is given the score of (2), 3 years of secondary education is scored as (3), and the score of (4) is reserved for a postsecondary education of 2 years. The highest score of (5) indicates postsecondary education of at least 3 years (including a PhD degree). In all categories, it is most common to have an education of the second level (attended secondary education for 2 years). About 16% of the indigenous Swedes have the highest education level compared to 14% among the immigrants and almost 12% among the second-generation immigrants without a Swedish-born parent (which is the category with a high proportion of young persons). In the total population, the median educational level is 2.6. The indigenous Swedes have the highest median score of 2.7 and the immigrants the lowest score of 2.3. In all four categories, there is a tendency that the younger age categories (under 25 years of age) have a higher educational level than the older categories (40 years of age or older). Females have a higher median educational level than males, the median score being about 2.8 for females and 2.5 for males. Females have a higher educational level than males in all four categories.

Method

We present a number of descriptive statistics, separated by sex and foreign background, to address our research question. When comparing offending rates between the different categories of Swedish and immigrant background, we sometimes refer to adjusted rates. In that case, we have recalculated the immigrants’ crime rates as to reflect the age, education level, and gender composition in the native population using the standard population method to ensure that any detected differences are not based on these factors. The obtained differences between the categories of Swedish and foreign backgrounds reveal the differences in the whole Swedish population. Therefore, no tests of statistical significance between the compared categories will be performed.

Following interrelated statistical measures for the level of offending will be referred to in the presentation of the results: the prevalence rate, the incidence rate, the relative risk, the relative incidence ratio, and the average number of offences among the suspected persons.

The prevalence rate is based on the number of persons per 1,000 in the population or a population category who have been registered for at least one offense during the 5-year period studied. In statistical terms, it is the probability or the risk that the members of a category will be suspected of an offence. The incidence rate is based on the number of offences per 1,000 registered among the members of a population category during the time period. The rate shows the average number of offenses the members of a category have been suspected of and gives an idea about the offending activity of a category. The relative risk, or ratio, of being suspected of an offense for a nonindigenous category is obtained by dividing a nonindigenous category’s prevalence rate with the corresponding rate of the indigenous category. The prevalence rate for indigenous Swedes is used as point of reference to inform how many times the rate for the nonindigenous categories exceed the rate for the indigenous category. A ratio close to 1 indicates that the nonindigenous category has about the same probability of being registered as a suspect, as had the indigenous Swedes. A ratio greater than 1 shows how many times the risk of being suspected of an offense the category in question exceeds the risk of the indigenous Swedes. A ratio below 1 indicates that the indigenous Swedes have a higher risk than the nonindigenous category of being registered.

The relative incidence ratio is a measure showing how much the crime activity of the nonindigenous categories deviate from the activity of the indigenous Swedes and is obtained by dividing the incidence rate of the former by the rate of the latter. A relative incidence ratio being close to 1 for a nonindigenous category indicates that the offending activity is on about the same level as for the indigenous group. A ratio being greater than 1 for the nonindigenous category tells how many times higher the offending activity is when related to the activity of the indigenous category. A ratio below 1 indicates that the offending activity is higher in the indigenous category than in the nonindigenous one.

The average number of registered offenses among the suspected persons (sometimes referred to as lambda in the literature of the field) tells how frequently the suspected persons in a population category on the average have been registered for an offense during the 5-year period. It is a measure of the suspected persons’ offending activity.

Limitations of Registered Offending

The use of police-registered offenses is also problematic from a methodological viewpoint (Geissler, 2008; Killias, 2011; Sommer, 2012). First, not all police reported offenses have a suspect. In the study presented below, the majority (about 80%) of the offenses registered have no official suspect. Furthermore, not all crimes committed in the society are detected, not all detected crimes are reported to the police, and not all detected crimes reported to the police lead to a conviction. Thus, the offenses being analyzed in this study are only a limited part of the offenses committed. The “dark figure” of crime and clearance rates vary between different types of reported offenses (National Council for Crime Prevention, 2002). This limitation of the material is important to have in mind when interpreting the results. Furthermore, one might have objections, and rightly so, against choosing persons suspected of offending instead of persons found guilty of the offenses they have been charged for. Kardell (2006) has shown that the picture of the offending rates among immigrants and their offspring is slightly overestimated when using data on suspected persons rather than on persons found guilty for an offense. This implies that there is an overreporting of immigrants and that their case is dropped at a later stage in the criminal justice system. They may occur because the police develop investigation strategies restricting the search for offenses and potential offenders to for example, specific neighborhood areas or persons with specific appearances (Pettersson, 2005). This disparate registration of suspected persons with an immigrant background can be seen as the consequence of so-called structural discrimination of foreign-born (Kardell, 2006).

Results

Overall Offending

Figure 1 displays unadjusted and adjusted prevalence rates for overall offending in terms of per 1,000 of the members in the respective categories of Swedish and immigrant background.

The prevalence rates for overall offending confirm the results of previous Swedish studies that second-generation immigrants have been registered as a suspect to a lesser extent than have the first generation. Both the unadjusted and the adjusted rates show the highest rates for the first-generation immigrants, followed the second-generation immigrants without Swedish-born parents, second-generation immigrants with one Swedish-born parent, and the indigenous Swedes. The risk of being suspected is 2.4 times higher for the first-generation immigrants than for the indigenous Swedes. For the second-generation immigrants without Swedish-born parents, the relative risk is 1.7 times higher and for the second-generation immigrants with one Swedish-born parent the risk is 1.4 times higher.

Figure 2 shows the unadjusted and adjusted the average number of offenses registered among the different categories multiplied by 1,000, that is, the incidence rates.

The incidence rates of overall offending give a slightly different picture than the prevalence rates. The unadjusted rates reveal that the second-generation immigrants without Swedish-born parents have been suspected for a greater number of offenses than have any other category. When controlling for gender, age, and educational level, the rates are reduced, in particular for the second-generation immigrants without Swedish-born parents. Their rate drops 25%, from 700 offenses per 1,000 persons to almost 540 offenses per 1,000 persons. The adjusted rate is practically on the same level as the corresponding rate for the first-generation immigrants. The second-generation immigrants without Swedish-born parents consist of relatively many young people who generally also tend to be crime-prone which might partly explain the drop when the age distribution is adjusted to the distribution of the indigenous Swedes. The relative incidence ratios for the first-generation immigrants and the second-generation immigrants without Swedish-born parents reveal that they on the average have been registered for an offense twice as many times, as have the indigenous category. The relative incidence ratio for the second-generation immigrants with one Swedish-born parent is 1.6.

When turning the attention to the average number of registered offenses among the suspected persons in the different categories (displayed in Figure 3), one can see that the first-generation immigrants have the lowest number of offenses and the second-generation immigrants without Swedish-born parents have the highest. The adjusted number for the first-generation immigrants is about 4.5 offenses compared to 6.5 offenses for the second-generation immigrants without Swedish-born parents. However, the corresponding number among the second-generation immigrants with one Swedish-born parent is 6.2 offenses, which is close to the figure for those with two foreign-born parents. The number of offenses among the suspected indigenous Swedes is 5.4. Both categories of second-generation immigrants, who appear in the register of suspected persons, appear more frequently than the first-generation immigrants and the indigenous Swedes. The first-generation immigrants appear less frequently in register than any other category.

Different Types of Offenses

In this section, the relative risks for being suspected of different types of offenses among immigrant background categories will be investigated as well as the categories’ relative crime activity in the different crime types. Tables 1 and 2 display the adjusted relative risks of being registered as a suspect in the different types of offenses for males and females, respectively. The prevalence rates for the indigenous Swedes are reference values (shown in the last column of the tables) and the relative risks for the categories with immigrant background give information about how many times their rates exceed the rates for the indigenous Swedes. The highest relative risk in each offense type is underscored.

Table 1.

Relative Risk of Being Registered as a Suspect of Crime in Different Offense Types Among the Immigrant Categories—Males.

	Relative risk			Prevalence per 1,000
	First- generation immigrants	Second-generation immigrants/without Swedish- born parent	Second- generation immigrants/with one Swedish-born parent	Indigenous
Number of persons in the population (N)	284,621	72,698	208,532	1,677,503
Completed and attempted murder and other lethal violence	4.1	3.0	1.9	0.6
Assault (not resulting in death)	3.0	1.9	1.5	23.4
Assault against children under 15 years of age	3.9	1.7	1.4	1.6
Assault against women 15 years of age or over	3.4	2.0	1.5	9.8
Assault against men 15 years of age or older	2.9	2.0	1.5	14.3
Unlawful threat	3.7	2.1	1.6	12.6
Crimes against liberty and peace other than unlawful threat	2.6	1.9	1.5	9.5
Rape/gross rape, completed and attempts	5.1	2.0	1.5	0.9
Sexual offenses other than rapes	2.7	1.5	1.3	2.1
Crimes against public activity	2.9	2.3	1.6	7.6
Crimes against law on possession of weapons/knives	2.1	2.4	1.8	5.9
Robbery/gross robbery (without and with fire arms)	4.0	3.6	2.1	1.5
Crimes inflicting damage	1.5	1.7	1.5	10.8
Car theft	1.5	2.1	1.7	5.0
Theft out of or from motor vehicle	1.3	1.9	1.6	4.0
Burglary involving dwellings	1.6	2.1	1.6	1.7
Theft from shops, department stores, etc.	3.5	2.1	1.6	8.5
Fraud, embezzlement, and other breaches of trust	2.5	1.8	1.5	14.8
Crimes of falsification	2.8	1.8	1.6	2.3
Drunken driving, gross drunken driving	1.4	1.6	1.3	14.1
Other crimes against the Road Traffic Offences Act	2.7	1.9	1.5	12.1
Crimes against the Narcotics Drugs (Punishment) Act	2.2	2.3	1.8	8.2

Note. The prevalence rates multiplied by 1,000 for the indigenous Swedes are displayed in the right column. The highest relative value for each offense type is underscored. Adjusted for age and educational level.

Table 2.

Relative Risk of Being Registered as a Suspect of Crime in Different Offense Types Among the Immigrant Categories—Females.

	Relative risk			Prevalence per 1,000
	First-generation immigrants	Second- generation immigrants/without Swedish-born parent	Second- generation immigrants/with one Swedish-born parent	Indigenous
Number of persons in the population (N)	290,160	68,881	198,444	1,596,353
Completed and attempted murder and other lethal violence	–	–	–	–
Assault (not resulting in death)	2.9	1.9	1.5	3.6
Assault against children under 15 years of age	4.6	1.6	1.3	0.5
Assault against women 15 years of age or over	2.7	1.9	1.5	1.7
Assault against men 15 years of age or older	2.8	2.1	1.7	1.5
Unlawful threat	3.5	2.0	1.5	1.6
Crimes against liberty other than unlawful threat	2.4	1.7	1.5	2.0
Rape/gross rape, completed and attempts	–	–	–	–
Sexual offenses other than rape	–	–	–	–
Crimes against public activity	2.1	2.0	1.6	1.0
Crimes against law on possession of weapons/knives	1.4	2.8	1.7	0.7
Robbery/gross robbery (without and with fire arms)	–	–	–	–
Crimes inflicting damage	1.8	2.0	1.6	1.2
Car theft	0.6	1.6	1.6	0.7
Theft out of or from motor vehicle	0.6	1.4	1.6	0.5
Burglary involving dwellings	0.8	1.6	1.7	0.3
Theft from shops, department stores, etc.	3.8	1.8	1.4	5.4
Fraud, embezzlement, and other breaches of trust	1.9	1.6	1.4	4.7
Crimes of falsification	1.6	1.6	1.4	1.0
Drunken driving, gross drunken driving	1.0	1.5	1.4	1.9
Other crimes against the Road Traffic Offences Act	1.9	1.8	1.5	1.7
Crimes against the Narcotics Drugs (Punishment) Act	1.0	1.5	1.5	1.9

The prevalence and incidence rates for various types of offenses reveal that the first-generation immigrants have higher rates than any other group in many types of offenses considered in the tables, but not in all types. For first-generation immigrant males, the relative risk of being registered is most accentuated for violent offenses such as lethal violence, assaults, rape, and robbery. For first-generation immigrants, the risk to be registered for rape is 5 times higher than that for the indigenous Swedes. For murder and assaults leading to death and for robbery, the risk is 4 times higher. The risk of being registered for various types of assault is between 3 and 4 times higher for the first-generation immigrant than for the indigenous Swedes. First-generation immigrant females have their highest rates in assault against children and thefts from shops and department stores, where the risk to be registered is 4.5 and 3.8 times higher than that for the indigenous females. Females have seldom been registered as suspects of the most serious violent offenses and the number of suspects is too low for calculating prevalence and incidence rates. Both first-generation immigrant males and females have relatively high prevalence and incidence rates of thefts from shops and department stores. The risk of being registered for this type of offense is about 3.5 times higher for first-generation immigrants than for the indigenous Swedes.

There are some types of offenses where the second-generation immigrants without Swedish-born parents have the highest risks of being registered. This is the case for crimes inflicting damage, car theft, theft out of and from motor vehicles, and burglary into dwelling, which are offenses usually associated with juvenile delinquency and everyday criminality. The risks are about twice as high for the second-generation immigrants without Swedish-born parents. Males in this category also have a higher risk than those in the other categories of being suspected for offending the law of possessing weapons and the law prohibiting carrying knives, a fact that could also be related to the high risk of being suspected for robbery. The risk of being registered for drunken driving and crimes against the narcotics drug act is also relatively high for this group.

Even if females generally have lower probabilities of being registered than males, their relative risks show a pattern resembling the pattern of males. Second-generation immigrants without Swedish-born parents females have the highest risk of being registered for crime against the law of possessing weapons and the law prohibiting carrying knives, crimes inflicting damage, thefts out of and from motor vehicles, and drunken driving.

The conclusion that the second-generation immigrants without Swedish-born parents have high rates on offenses associated with juvenile offending and everyday crimes are confirmed and even further underscored by the incidence rates. Among females this pattern even comes out more distinctly with the incidence rates than with the prevalence rates.

The incidence rates (Tables A1 and A2 in Appendix) for different types of offenses and the categories of Swedish and immigrant background show about the same pattern, as do the prevalence rates (Tables 1 and 2).

Discussion

Overall Offending

First-generation immigrants have the highest offender rate (participation rate), at the same time as they (the suspected offenders) have the lowest average number of offenses. This indicates that there are quite a number among the suspected offenders with only one single registered offense, that is, are low-rate offenders.

Second-generation immigrants have lower offender rates than the first-generation immigrants, at the same time they have the highest average number of suspected offenses. This indicates that quite a number among the second-generation immigrants have been suspected of several offenses, that is, are high-rate offenders. This pattern is slightly more pronounced for the second generation with two foreign-born parents.

Indigenous Swedes have the lowest offender rate, but the suspected offenders have higher average number of registered offenses than first-generation immigrants but a lower average number than second-generation immigrants. Apparently, there are more high-rate offenders among the suspected native Swedes than among the first-generation immigrants.

As long as the prevalent rates in the different subpopulations are considered, the results from previous Swedish studies are reproduced and thereby differ from the research findings reported in other countries. But when crime frequencies are taken into account first-generation immigrants have the lowest crime activity and the second generation the highest activity which is similar to the patterns found in the United States. Furthermore, the higher proportion of low-rate offenders among the suspected persons in the first-generation immigrants can be a consequence of an underlying “structural discrimination” (Kardell, 2006).

Different Offending Profiles

The offending profile for first-generation immigrant males is characterized by higher involvement in lethal violence, assaults not leading to death (e.g., assault against children and women), rape, and robbery, which are all serious violent crimes. For females, assault of children and unlawful threat dominate the profile. These are all offenses of violence and occur in interaction situations with other known and unknown persons. The first-generation immigrants, both males and females, have also a relatively high involvement in thefts from shops and department stores, that is, an unlawful way of getting access to goods like food and clothes, but also smaller luxury goods found in department stores and boutiques.

The recurring finding in the United States that the first-generation immigrants are less likely to be involved in crime than the native-born Americans, runs counter to the expectations from classical criminological theories on social strain, differential opportunity structures, cultural conflicts, social disorganization, and so on (Mears, 2002). Apparently, a high proportion of the immigrants in the United States are motivated people who are eager to get ahead in the new country. Tonry (1997, p. 21) describes them as “self-selected economic migrants” in contrast to the high proportion of refugee immigrants received generously in Western Europe, who, to begin with, are more motivated to find a tentative shelter from war and terror in their home countries. The mentioned criminological theories seem, however, quite useful to give some theoretical insight into our Swedish research findings. The first-generation immigrants’ higher involvement in violent crimes can partly be explained by cultural differences in how one is used to handle stress and conflicts in day-to-day situations. On the other hand, the emphasis on social equality in the Swedish public and political debate might be provocative for a foreign-born from a patriarchal society. The strong emphasis on the equality between men and women might even be food for conflict within an immigrant family. The experience of discrimination and reoccurring blocked opportunities for instance to get a better paid job, or to get a job at all, is frustrating for an immigrant and might turn to hate and revenge or even to expressive violence (Wortley, 2009). The higher level of reported petit thefts among the first-generation immigrants can partly be explained by a poor economy and poverty leading to a temptation to use illegitimate means in order to get access to material goods.

The offending profile for the second-generation immigrants without Swedish-born parents is determined by their relatively high involvement in car thefts, thefts out of and from cars and other motor vehicles, burglary into dwellings and houses, and damage of other people’s property. Their involvement in robbery is also comparably high as well as in unlawfully carrying weapons and knives. Weapons and knives can be used as tools to shock and threaten a victim in a robbery or just carried for self-protection in case of a threatening situation. Drunk driving and drug-related offences are also to some extent crimes that belong to this group’s offending profile. This profile is mainly applicable to males, whereas the females’ profile is not that pronounced except for fraud and embezzlement being particularly frequent. The offending profile for the second-generation immigrants with two foreign-born parents is mainly crimes against property in which some offenders might use threats and violence to break a victim’s resistance. These crimes are often associated with juvenile delinquency.

The classical theories of social disorganization, social strain, and subcultures of deviant youths are often used to explain causes of juvenile offending (Geissler, 2008; Killias, 2011; Mears, 2002; Tonry, 1997; Wortley, 2009). Second-generation immigrants quite often grow up in urban neighborhoods characterized by various social problems and poverty. Maybe this socialization environment is more common among those with two foreign-born parents than among their peers with one Swedish-born parent. Unavoidably, immigrants’ children internalize the host country’s mainstream values and want the same high material standards as their middle class native peers already have. The immigrant parents cannot always provide their children with a mainstream standard of living. The economic barriers may cause immigrant children to obtain attractive, high-status goods through illegitimate means. In particular, the second-generation immigrants with both parents born abroad seem to have an enhanced risk of becoming a property crime offender.

The Risk of Becoming a Chronic Offender

Among the second-generation immigrants without Swedish-born parents who were suspected of an offense comparably many are high-rate offenders. Some of these people may become chronic offenders. The offending profile found for this immigrant group may be illuminated by Swedish studies on strategic offenses. Svensson (2002) has found that when vehicle theft is the principle offense in the first conviction of a young person, there is a relatively high risk that he or she will follow “the career path of a chronic offender” (Svensson, 2002, p. 401). Also robbery/aggravated robbery as the first conviction indicates a fairly high risk of developing a chronic criminal career. In a more recent similar study, these results were confirmed, but with a slightly stronger emphasis on robbery and aggravated robbery and violent crimes (National Council for Crime Prevention, 2011). These so-called strategic offences, that is, offenses appearing early in a criminal career that predict a serious criminal career (Wikström, 1995), show up in the offending profile for the second-generation immigrants without Swedish-born parents. It seems plausible to assume that this group of second-generation immigrants also has a higher risk than the other categories for being convicted for these kinds of offenses as the first conviction and will thus end up in a long-lasting criminal career. This is by no means to say that chronic offenders mainly are recruited from this small group, they develop in all of the groups of Swedish and immigrant background. Simply, the offending profile of the second-generation immigrants with two foreign-born parents contain strategic offences, which may mean that they are “more likely to cluster in a persistent offending group” (Bersani, 2012, p. 8).

Further Disaggregation of the Immigrant Concept

The immigrants are not a homogenous group and should not be treated as a single category in the research of the immigrant–crime nexus. A reoccurring issue in the literature of the field is how to decide who is an immigrant and how to disaggregate this wide concept in different subcategories in a meaningful way. The use of broad categories conceals important differences in subgroups (Martinez & Lee, 2000). To divide the immigrants into generations is the most common way of disaggregating. Western European criminologists have made the distinction between labor force immigrants and refugee immigrants where the former have revealed lower crime rates than the indigenous population and the latter higher rates. A recent concept that has turned up in the literature of the field is the one-and-a-half generation immigrants denoting foreign-born persons who came to the host country before the age of 12 years (Wortley, 2009). Swedish studies have revealed that first-generation immigrants from South East Asia and from advanced industrial democracies like the United States, Canada, Australia, and New Zealand as well as western European countries have crime rates just above the level of the indigenous Swedes. In contrast, immigrants from politically instable countries in North Africa and West Asia have considerably higher crime rates than native Swedes (Ahlberg, 1996; Martens, 1997). There is also a clear tendency of first-generation immigrants from countries with high prevalence in Swedish crime statistics to have Swedish-born offspring with high crime involvement as well. Correspondingly, first-generation immigrants from countries with low crime involvement also have offspring with low involvement—the correlation being as high as r = .80 (Martens, 1997, p. 229). Hence, as a group the resident immigrants in a multicultural society is complex in nature and needs more solid and insightful clarification in the future immigrant–crime nexus research. According to Bursik (2009), this is an issue where immigration research rather has regressed. Tonry (1997) concluded that the reasons for a category of immigrants to leave the home country might be predictive of their subsequent behavior in the new country. Also, the efforts of the receiving countries to meet the needs of the newcomers and facilitate their integration into the society to become active and self-supporting citizens are important aspects that might lower the risks for the immigrants’ crime involvement.

Immigrant Background as an Explanatory Factor

It should be pointed out that the vast majority of the population members in the presented study have never been suspected of an offense during the 5-year observation period. Most people, no matter which category of Swedish or immigrant background they belong, are not connected with the offenses registered by the police.

This leads to the more general question of how important ethnic background is as an explanatory factor for offending. When using social background factors such as gender, age, educational level, and income level together with ethnic background in multivariate analyses, ethnic background explains less of the variance in registered offending than do the other factors (not reported here). Similar conclusions have been drawn in a more extensive study by Hällsten, Sarnecki, and Szulkin (2011). This implies that ethnic background alone has a limited explanatory power. In a study of self-reported offending among pupils in the ninth grade in Sweden, Ring (1999) concluded that ethnic background had a low explanatory power. This conclusion has been drawn by Ring in various follow-up studies done at the National Council for Crime Prevention Sweden (2005a). It is known from the reanalyses of Glueck and Glueck’s (1950) material by Sampson and Laub (1993) and from some Swedish studies on juvenile offending (e.g., Martens, 1992; Ring, 1999; Svensson, 2004) that structural background variables generally have little direct effect on offending rates during the life course. Instead, the influence of these factors is indirect and mediated through sociopsychological factors such as attachment to family, school, delinquent peers, and the labor force. To arrive at a better understanding of the eventual differences in offending among various categories in the society, one has to focus upon the unequal distribution of social and material resources. This revives the issues discussed in the classic sociological literature on social class stratification, among others, pointing to stratification of social classes by ethnic categories.

Conclusion

The overall offending rates have revealed that there are more first-generation immigrants suspected for a crime than any other group, but relatively many of them are low-rate offenders, that is, are involved in only one offense. There are even more high-rate offenders among indigenous Swedes than among the first-generation immigrants. Furthermore, the higher proportion of high-rate offenders among the second-generation immigrants indicates that they run a higher risk to become chronic offenders—in particular those with two foreign-born parents. When both prevalence and incidence of offending are used to measure crime involvement the relationship between crime involvement and immigrant background turns out to be more complicated than previous Swedish research has realized.

To return to the initial question of this article asking if the offspring of immigrants born in Sweden are more law-abiding than the immigrants, it cannot be answered with just yes or no. It is not a question which of the two population categories is more law-abiding; it is more a question in what types of crime they are more engaged in. Hence, to the extent the first- and second-generation immigrants are involved in criminal activities they are engaged in different types of crime. Even after standardization for age, gender, and educational level, the offending profiles of the two subpopulations seem to reflect criminal activities among offenders in different phases of the life cycle.

Footnotes

Appendix

Table A1.

Relative Incidence Ratio for Different Offense Types Among the Immigrant Categories—Males.

	Relative incidence ratio			Number of offenses per 1,000
	First- generation immigrants	Second- generation immigrants/without Swedish-born parent	Second- generation immigrants/with one Swedish-born parent	Indigenous
Number of persons in the population (N)	284,621	72,698	208,532	1,677,503
Completed and attempted murder and other lethal violence	4.0	3.0	1.9	0.8
Assault (not resulting in death)	3.0	2.0	1.5	47.1
Assault against children under 15 years of age	4.1	1.4	1.3	2.4
Assault against women 15 years of age or over	3.3	2.0	1.6	18.2
Assault against men 15 years of age or older	2.8	2.1	1.6	25.3
Unlawful threat	3.6	2.1	1.5	26.0
Other crimes against liberty and peace	2.5	1.8	1.5	18.8
Rape/gross rape, completed and attempts	5.4	1.9	2.5	1.3
Sexual offenses other than rapes	2.3	1.4	1.1	4.3
Crimes against public activity	2.9	2.4	1.7	17.8
Crimes against law on possession of weapons/knives	1.8	2.5	1.8	13.2
Robbery/gross robbery (without and with fire arms)	4.6	4.0	2.2	2.4
Crimes inflicting damage	1.4	1.8	1.6	21.3
Car theft	1.3	2.0	1.6	16.4
Theft out of or from motor vehicle	1.2	1.7	1.6	14.0
Burglary involving dwellings	1.7	2.2	1.6	3.2
Theft from shops, department stores, etc.	3.2	2.3	2.2	26.5
Fraud, embezzlement, and other breaches of trust	1.9	1.7	1.5	46.3
Crimes of falsification	2.6	1.9	1.7	4.5
Drunken driving, gross drunken driving	1.4	1.6	1.4	23.2
Other crimes against the Road Traffic Offences Act	1.7	2.2	1.7	65.2
Crimes against the Narcotics Drugs (Punishment) Act	1.9	2.2	1.8	30.7

Note. The right column displays the incidence rates multiplied by 1,000 for the indigenous Swedes. The highest relative incidence ratios for each offense type are underscored. Adjusted for age and educational level.

Table A2.

Relative Incidence Ratio for Different Offense Types Among the Immigrant Categories—Females.

	Relative incidence ratio			Number of offenses per 1,000
	First-generation immigrants	Second- generation immigrants/with no Swedish-born parent	Second- generation immigrants/ with one Swedish-born parent	Indigenous
Number of persons in the population (N)	290,160	68,881	198,444	1,596,353
Completed and attempted murder and other lethal violence	–	–	–	–
Assault (not resulting in death)	2.9	1.9	1.5	5.5
Assault against children under 15 years of age	4.8	1.7	1.4	0.7
Assault against women 15 years of age or over	2.6	1.9	1.5	2.5
Assault against men 15 years of age or older	2.7	2.0	1.7	2.3
Unlawful threat	3.5	2.0	1.5	2.6
Other crimes against liberty and peace	2.3	1.9	1.4	3.7
Rape/gross rape, completed and attempts	–	–	–	–
Sexual offenses other than rape	–	–	–	–
Crimes against public activity	1.9	1.9	1.6	2.0
Crimes against law on possession of weapons/knives	1.2	2.5	1.7	1.2
Robbery/gross robbery (without and with fire arms)	–	–	–	–
Crimes inflicting damage	1.7	1.8	1.7	1.9
Car theft	0.5	1.9	1.5	1.8
Theft out of or from motor vehicle	0.5	1.6	1.4	1.3
Burglary involving dwellings	0.8	1.4	1.6	0.5
Theft from shops, department stores, etc.	3.3	2.0	1.5	13.2
Fraud, embezzlement, and other breaches of trust	1.6	3.5	1.5	12.7
Crimes of falsification	1.2	1.3	1.4	2.1
Drunken driving, gross drunken driving	0.9	1.5	1.5	2.6
Other crimes against the Road Traffic Offences Act	0.9	2.0	1.8	8.1
Crimes against the Narcotics Drugs (Punishment) Act	0.9	1.5	1.4	6.4

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Ahlberg

(1996). Offending among immigrants and the children of immigrants. A statistical analysis (National Council for Crime Prevention 1996:5). Stockholm, Sweden: Fritzes (Report in Swedish).

Ajzenstadt

Shapira

(2012). The socio-legal construction of otherness under a neo-liberal regime: The case of foreign workers in the Israel criminal courts. British Journal of Criminology, 52, 685–704.

Bersani

(2012). An examination of first and second generation immigrant offending trajectories. Justice Quarterly, 1–29. Advanced online publication. doi:10.1080/07418825.2012.659200

Bursik

(2009). The Dead Sea scrolls and criminological knowledge: 2008 Presidential address to the American society of criminology. Criminology, 47, 5–16.

Gabbidon

(2010). Race, ethnicity, crime and justice: An international dilemma. London, England: Sage.

Geissler

(2008). Der “kriminelle ausländer”—vorurteil oder realität? Zum stereotyp des “kriminellen ausländers.” Überblick, 14, 3–9.

Glueck

(1950). Unraveling juvenile delinquency. New York, NY: The Commonwealth Fund.

Haen Marshall

(Ed.). (1997). Minorities, migrants, and crime. Diversity and similarity across Europe and the United States. London, England: Sage.

Hällsten

Sarnecki

Szulkin

(2011). Crime as a price for inequality? The delinquency gap between children of immigrants and children of native Swedes (Working Paper No. 2011:1). Stockholm, Sweden: SULCIS/Stockholms universitet.

10.

Hiatt

K. D.

(2007). Immigrant danger? Immigration and increased crime in Europe. Berkeley: Institute for Research on Labor and Employment, University of California. Retrieved from http://escholarship.org/uc/item/3gt4s8w0

11.

Kardell

(2006). Discrimination of persons with foreign background in the criminal justice system—A quantitative analysis. In Sarnecki

(Ed.), Is the justice just? Ten perspectives on discrimination of ethnic and religious minorities within the criminal justice system. Official Report from the Swedish Government SOU 2006:30. Stockholm, Sweden: Fritzes (Report in Swedish).

12.

Kardell

(2011). Foreign background and registered criminality. Over-representation in the Swedish official statistics (Licentiatuppsats). Stockholm, Sweden: Department of criminology/Stockholm University (Report in Swedish).

13.

Kardell

(in press). Foreign background and registered criminality in Sweden. An analysis of the research concerning the over-representation of individuals with a non-Swedish background in registered crime in Sweden. Journal of Ethnicity in Criminal Justice.

14.

Kardell

Carlsson

K.-M.

(2009). Conviction of immigrants and immigrants offspring in the Nordic countries. Nordisk tidskrift for kriminalvidenskap, 3, 237–261 (Report in Swedish).

15.

Killias

(2011). Immigration and crime: The European experience. EU-US Immigration Systems 2011/19. San Domenico di Fiesole (FI), Italy: Robert Schuman Centre for Advanced Studies, European University Institute.

16.

Kommittén

Välfärdsbokslut

Palme

(2002). Welfare in Sweden: The balance sheet for the 1990s. Stockholm. Sweden: Fritzes.

17.

Martinez

Lee

(2000). On immigration and crime. Criminal Justice, 1, 485–524.

18.

Martens

(1992). Family, environment and delinquency (National Council for Crime Prevention 1992:1). Stockholm, Sweden: Fritzes (Report in Swedish).

19.

Martens

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

(Ed.), Ethnicity, crime, and immigration: Comparative and cross-national perspectives (pp. 375–405). Chicago, IL: University of Chicago Press.

20.

Mears

(2002). Immigration and crime: What’s the connection? Federal Sentencing Reporter, 5, 284–288.

21.

National Council for Crime Prevention. (2002). Crime clearance and efficiency. An analysis of the factors affecting trends in the clear-up rate. (National Council for Crime Prevention 2002:4). Stockholm, Sweden: Fritzes (Report in Swedish).

22.

National Council for Crime Prevention. (2005a). Theft, drugs and violence among ninth grade youth. Results from five self-report surveys (National Council for Crime Prevention 2005:4). Stockholm, Sweden: Fritzes (Report in Swedish).

23.

National Council for Crime Prevention. (2005b). Crime among Swedes and immigrants (National Council for Crime Prevention 2005:17). Stockholm, Sweden: Fritzes (Report in Swedish).

24.

National Council for Crime Prevention. (2011). Strategic crimes among youths in the 00’s (National Council for Crime Prevention 2005:17). Stockholm, Sweden: Fritzes (Report in Swedish).

25.

Pettersson

(2005). Are persons with a foreign background discriminated in the criminal justice system? In Diesen

Lernerstedt

Lindholm

Pettersson

(Eds.), Equality before the law. Stockholm, Sweden: Natur och Kultur (Report in Swedish).

26.

Ramakrishnan

S. K.

(2004). Second-generation immigrants? The “2.5 generation” in the United States. Social Science Quarterly, 85, 380–399.

27.

Reid

Weiss

Adelman

Jaret

(2005). The immigration-crime relationship: Evidence across US metropolitan areas. Social Science Research, 34, 757–780.

28.

Ring

(1999). Home and school, peers and offending. Stockholm, Sweden: Stockholm University/Department of Criminology (Report in Swedish).

29.

Sampson

(2008). Rethinking crime and immigration. Context, 7, 28–33.

30.

Sampson

Laub

(1993). Crime in the making: Pathways and turning points through life. Cambridge MA: Harvard University Press.

31.

Sommer

(2012). “Ausländerkriminalität”—statistische daten und soziale wirklichkeit. Bonn, Germany: Bundeszentrale für politische Bildung.

32.

Statistics Sweden. (2002). A longitudinal database regarding education, income and employment (LOUISE) 1990-1999. Background facts for the statistics on labor market and education (Statistics Sweden 2002:2). Örebro, Sweden: Statistics Sweden (Report in Swedish).

33.

Statistics Sweden. (2003). The future population of Sweden 2003-2020. Swedish and foreign background (Statistics Sweden 2003:5). Örebro, Sweden: Statistics Sweden (Report in Swedish).

34.

Svensson

(2002). Strategic offences in the criminal career context. British Journal of Criminology, 42, 395–411.

35.

Svensson

(2004). Social control and socialisation. The role of morality as a social mechanism in adolescent deviant behaviour. Stockholm, Sweden: Almqvist & Wiksell International.

36.

Tonry

(1997). Ethnicity, crime, and immigration. In Tonry

(Ed.), Ethnicity, crime, and immigration: Comparative and cross-national perspectives (pp. 375–405). Chicago, IL: University of Chicago Press.

37.

von Hofer

Sarnecki

Tham

(1997). Minorities, crime, and criminal justice in Sweden. In Haen Marshall

(Ed.), Minorities, migrants, and crime. Diversity and similarity across Europe and the United States (pp. 151–174). London, England: Sage.

38.

Wikström

P-O.

(1995). Self-control, temptations, frictions and punishment: An integrated approach to crime prevention. In Wikström

P. O.

Clarke

McCord

(Eds.), Integrating crime prevention strategies: Propensity and opportunity (National Council for Crime Prevention 1995:5). Stockholm, Sweden: Fritzes.

39.

Wortley

(2009). Introduction. The immigrant-crime connection: Competing theoretical perspectives. International Migration & Integration, 10, 349–358.