Two of a kind? A comparative multicohort study of juvenile violence in Finland and Sweden

Abstract

Youth violence has decreased in many Western countries, including Finland and Sweden, during the 2000s. A worrying deviation from this decline concerns lethal youth violence in the two countries, in particular Sweden, which has experienced a clear increase. In this study, we analyze the extent to which this divergence mirrors a concentration of violence whereby fewer and fewer young people are subject to the criminal justice system, but where those who are, are increasingly high-frequent and/or severe offenders. We employ full population longitudinal register data on Finnish and Swedish birth cohorts born between 1986 and 1998 who are followed with conviction data for violent crimes between ages 15 and 20. We disaggregate the trends across these cohorts into criminal career parameters, break down violent crime into specific violent categories, and scrutinize the trends by sociodemographic background variables that are typically associated with violent crime: sex, foreign background, family income, and urban residence. Our findings show that the overall decline in juvenile violent convictions is primarily driven by a declining proportion of violent offenders. This decline is fairly uniform across the examined background variables in both countries. At the same time, this study gives a less optimistic picture of the development within the convicted population of juvenile violent offenders. Convicted juvenile violent offenders born in the late 1990s offend with the same frequency, have a similar debut age of convicted violent crime, but are recorded for more severe offenses as their same-age counterparts born just a decade earlier. Longitudinal data on multiple birth cohorts are needed to understand whether we are seeing more lasting changes in criminal careers, and how these developments look like in comparative terms.

Keywords

Comparative research longitudinal multicohort violent crime

Introduction

Juvenile violence has declined since the early 1990s across many Western nations (Estrada, 2001; Farrell et al., 2014; Tonry, 2014), and this trend seems to have largely continued into the 21st century (Pickett et al., 2013). The Nordic countries, including Finland and Sweden, are no exception to this overall decline in juvenile violence. Surveys conducted on nationally representative samples of pupils indicate that the proportion of youth perpetrating physical violence has decreased markedly since the late 1990s until the late 2010s in Sweden (Svensson and Oberwittler, 2021) and in Finland (Näsi et al., 2024), particularly during the 2010s. Official statistics on violent convictions and hospital data on violent victimization in the two countries show a similar decline (Estrada, 2022; Haapakangas, 2020). The decline in violence has occurred during a time period that may be described as a unitary trend in Western societies toward less risky behaviors, including a decline in alcohol consumption and unstructured face-to-face interactions (Ball et al., 2023). Recent empirical evidence points out that such broader changes in routine activities are important explanations for the decline in juvenile violence (Baumer et al., 2021; Näsi et al., 2024; Osgood, 2023; Svensson and Oberwittler, 2021).

Despite the broader tendencies toward less juvenile violence, there is variation between historical time periods, geographical contexts, crime types, and demographic groups that may be informative and, to some extent, also question the generality of the decline of violent crime (Aebi and Linde, 2012; Baumer et al., 2018). For one thing, violence is a heterogeneous construct that often includes a broad repertoire of aggressive behavior, including the most serious forms of physical violence such as homicides, much more common ‘street’ violence such as assaults, instrumental forms of violence such as robberies, and also non-physical violence such as unlawful threats. To be sure, lethal violence makes up a tiny fraction of all violent crime, which is perpetrated by a small proportion of the offender population. Still, homicide has typically been viewed as ‘the tip of the iceberg’, which reflects underlying levels and trends of crime in general, and non-lethal violent crime in particular (Van Breen et al., 2023). Finland and Sweden are examples of countries that, from a long-term historical perspective, have shown a strong association between lethal and non-lethal forms of violence (von Hofer, 2011; Vuorela, 2018).

However, as concluded by van Breen et al. (2023), the association between lethal and non-lethal violence is dependent on the social and economic context and varies with time periods and locations. Considering the recent two decades in Finland and Sweden, the co-development between lethal and non-lethal violence has become increasingly diffused, especially when focusing on juvenile offenders and victims (Näsi et al., 2024; Sivertsson et al., 2022). Figure 1 shows that the rates (offenders per 100,000) of convictions at ages 15–20 for assault have decreased markedly in both Finland and Sweden, in Sweden by roughly half and in Finland even more. In contrast, rates of homicide/manslaughter convictions initially show a stable trend followed by a marked increase during the 2010s in Sweden, while Finland shows a relatively stable trend over the entire period, albeit with large fluctuations between years as the number of convictions is small. These trends in the respective countries mirror those observed in other data sources for juvenile violent crime, such as self-reported assaults among school pupils (e.g. Näsi et al., 2024; Svensson and Oberwittler, 2021) and juvenile lethal violence, for example, as reported in the European Homicide Monitor and national cause-of-death statistics (Hradilova Selin et al., 2024; Suonpää et al., 2024).

Figure 1.

(a, b) Number of offenders (ages 15–20) convicted of assault (including aggravated) and homicide/manslaughter (including attempts) per 100,000 residents in Finland and Sweden, 2000 (2005)–2020.

As emphasized by a body of criminal career research (Blumstein et al., 1988; DeLisi and Piquero, 2011), aggregate crime trends are a function of both the proportion of the full cohort who commit crime (i.e. participation) and the frequency of crime in the offender population. There is increasing evidence that the historical decline in juvenile crime rates since the 1990s has mainly been caused by a reduction of participation, whereas frequencies, or recidivism tendencies, have remained relatively stable (Andersen et al., 2016; Berg et al., 2016; McAra and McVie, 2018; McCarthy, 2021; Payne and Piquero, 2020; Pitkänen et al., 2022; Sivertsson et al., 2021). The divergence between aggregate developments of lethal and non-lethal juvenile violence, as seen in Finland and Sweden over the first two decades of the 2000s, may therefore mirror a ‘concentration effect’ whereby fewer and fewer young people are subject to the criminal justice system, but where those who are, are increasingly comprised of high-frequent and/or severe offenders in marginalized groups of individuals (McAra and McVie, 2018).

At the same time, there may be distinct differences between Finland and Sweden in how such a concentration is distributed with regard to different types of violent crime and between different sociodemographic groups in society. For instance, it is by now well established that the increase in lethal juvenile violence in Sweden is driven by criminal milieu homicide, with more or less organized criminal networks competing over drug markets, and where the use of firearms is a common element (Gerell et al., 2021; Granath, 2024; Hradilova Selin et al., 2024; Sturup et al., 2019; Suonpää et al., 2024). This group of serious violent offenders has been shown to resemble career criminals in that they tend to commit a multitude of crime and exhibit multiple risk factors for persistent antisocial behavior, such as poor school performance and widespread networks of criminally active peers (Granath, 2024). In contrast, in Finland the basic structure of lethal violence—involving conflicts between friends and acquaintances under alcohol intoxication—has largely remained the same (Suonpää et al., 2024). In that sense, criminal milieu homicide, in addition to having caused an increase in the total homicide rate in Sweden, also marks a qualitative change in juvenile lethal violence that distinguishes Sweden from the rest of the Nordic region (Hradilova Selin et al., 2024; Suonpää et al., 2024). However, while this shift in lethal juvenile violence in Sweden is well documented, it is highly unclear whether it also captures a change between different forms of non-lethal violence (e.g. between robberies and assaults), and, relatedly, to what extent it marks distinct differences between Finland and Sweden in terms of concentration of general youth violence.

In this exploratory study, we analyze individual-level longitudinal register data on complete Finnish and Swedish birth cohorts born between 1986 and 1998 who are followed for convictions for violent crimes from the minimum age of criminal responsibility, age 15, until age 20 (i.e. a period of 6 years per individual). The employment of longitudinal data on multiple successive birth cohorts makes it possible to analyze the development of criminal career parameters, such as participation and individual crime frequency, over time. Furthermore, we break down violent crime into specific categories that are defined in a similar manner between the two jurisdictions, such as assault, robbery, and homicide/manslaughter. We also scrutinize the development of violent crime in the respective countries by sociodemographic background variables that are typically associated with violent crime: sex, foreign background, family income, and urban residence. Do convicted violent crime across birth cohorts of juveniles differ between Finland and Sweden with regard to:

The mean number of violent crimes per cohort member (i.e. incidence)

The proportion of juvenile violent offenders (i.e. participation)

The composition of the population of juvenile violent offenders in terms of:

their frequency of offending?

the age at which they were first convicted for a violent offense?

the types of violent crimes they are convicted of?

Sociodemographic gradients to the above developments (1–3)?

A comparison between Finland and Sweden with regard to criminal career parameters and types of violent crime across sociodemographic groups provides a deepened understanding of the overall decline of juvenile violent crime over the first two decades of the 2000s and the extent to which this is a general or diversified trend. In addition to substantive research questions, our study also has a methodological aim. We attempt to conduct a harmonized comparison of violence trends in two countries using data on officially recorded convictions. While the benefits of going beyond aggregate statistics in comparative criminology are obvious, coming up with comparable data from different countries is difficult, and criminologists are rightly cautious in interpreting the results of such studies (Aebi et al., 2021). Our approach is based on utilization of full population register-linkage data that has been increasingly used in criminological research (Lyngstad and Skardhamar, 2011), but not yet much in comparative research. We use similar sample inclusion criteria in both countries and match criminal codes as closely as possible to facilitate equivalence of measurement.

Contextual similarities and differences between Finland and Sweden

Finland and Sweden resemble each other in many respects and often represent the cluster of Nordic welfare states in international comparisons (Kangas and Kvist, 2018). Shared features are low levels of income inequality, generous and comprehensive welfare state provisions, and high employment rates. Crime trends in the Nordic countries have largely moved in tandem since the 1950s (von Hofer et al., 2012; Vuorela, 2018). The Nordic countries are also associated with leniency and humane approaches toward punishment (Lappi-Seppälä and Tonry, 2011), and they have a long tradition of compiling population statistics that rely on high-quality registers and linkages between different types of registers with reliable personal identification numbers (Lyngstad and Skardhamar, 2011). This makes individual-level comparative studies across the Nordic countries feasible, although the register systems and measures are not identical.

There are, however, some notable differences between Finland and Sweden in terms of demographic and socioeconomic development patterns that are important for this comparative study, in particular with regard to residents with foreign origins and residential patterns/urbanity. It is well established that persons with a foreign background are overrepresented in registered violent crime when compared to natives in the Nordic countries (Bäckman et al., 2021; Beckley et al., 2017; Skardhamar et al., 2014). The proportion of the population with foreign background may therefore be an important demographic component in comparative crime trend studies. In a Nordic context, Sweden has had a significantly more liberal migration policy. This is reflected as a difference in the proportion of residents who were born abroad or born in the country with two foreign-born parents between Finland and Sweden. According to this definition, just above a fourth of the Swedish population in 2020 had a foreign background, while the corresponding percentage in Finland was only 8%. These demographic differences, in particular the size and character of migration, have in the public debate been linked to alleged increases in crime and to a worrisome situation that distinguishes Sweden from the rest of the Nordic countries (Bäckman et al., 2021).

The nature of immigration also differs between Finland and Sweden. Until the 1970s, migration to Sweden was dominated by labor and family reunification migration from other Nordic countries and southern Europe, and since the 1990s, Sweden has experienced a clear increase in refugee migration from non-European countries (Beckley et al., 2017). Finland, in contrast, has a much shorter history of large-scale immigration, with numbers of foreign-born residents starting to increase markedly only at the start of the 1990s, and where the most common reasons for immigration at present are work, family, and study (Pesola et al., 2024). Relatedly, the socio-spatial segregation in metropolitan areas is lower in Finland than in Sweden. Finnish cities, in particular, do not have highly ethnically segregated areas such as Sweden, which is in part connected to the lower proportion of foreign background in the Finnish population. For instance, Finland's capital city, Helsinki, has internationally low levels of segregation, with lower levels than all its Nordic metropolitan counterparts, including Stockholm (Tunström et al., 2016).

Finally, cross-national comparative research tends to find a statistical association between economic inequality and crime (Kim et al., 2022). At the individual level, this has partly been explained by relative deprivation, that is, processes of resentment resulting from being economically disadvantaged compared to others, which, in turn, may result in various negative consequences, including crime (Itskovich, 2024). It is well known that income inequality is low in Finland and Sweden from an international perspective. However, it is in this regard worth noting that income inequality, for example, as measured by the Gini coefficient for household disposable income, has increased in both countries since the early 1990s (Pareliussen et al., 2018). Importantly, while the level of income inequality has been relatively stable in Finland during the 2000s until the mid-2010s, it has continued to increase in Sweden over this time period (Pareliussen et al., 2018).

Data and methods

In this study, we examine between-cohort changes in violent offending that have led to criminal convictions in Finland and Sweden. We use population registers to select individuals born between 1986 and 1998 and registered as residents at age 15 (i.e. the minimum age of criminal responsibility) in Sweden and Finland. Data on criminal convictions, urban residence, foreign background, family income, and mortality were then linked to all individuals in the respective populations using personal identifiers. All individuals were followed over a period of 6 years from age 15 up to and including age 20. In total, the data included 820,959 individuals residing in Finland and 1,502,376 individuals residing in Sweden (for population sizes by demographic subgroup, see Table S1 in the supplementary file). The linking of various registers and the de-identification of the data have been carried out by Statistics Finland and Statistics Sweden, respectively. The microdata have been stored on Statistics Finland's system for making microdata available for online research, FIONA, and the Swedish equivalent called MONA, administered by Statistics Sweden. We aggregated the data by birth cohort in the respective systems and were then able to export the data and generate comparative statistics in this format.

Conviction data on violent crime

We use conviction data to operationalize violent crime. Conviction records constitute an important source in Nordic criminological register-based research (Lyngstad and Skardhamar, 2011), and may in broad terms be said to capture offenders who have committed several offenses, and/or serious offenses, and/or traditional person- and property-oriented offenses (Kyvsgaard, 2002).

There are many similarities in the respective criminal justice systems of Finland and Sweden. The minimum age of criminal responsibility is 15, and young offenders who are 15 or older are dealt with in adult courts, although under policies calling for mitigated punishments (for a more in-depth description of the criminal justice system in the Nordic countries, see Lappi-Seppälä and Tonry, 2011). The conviction registers in the respective countries are population-based and event-based that contain every conviction since 1973 in Sweden and 1977 in Finland. A conviction may include several crimes. Each crime, in turn, can refer to different sections of the criminal code or to special criminal law statutes, and they may also have been committed on different days. Our analyses are based on all crimes in the convictions and the time of offense, instead of the time of the conviction decision.

A statistical difference is that the Swedish conviction register includes all types of convictions, including waivers of prosecution and summary sanction orders, whereas the Finnish conviction register includes only court decisions and summary sanctions, but not waivers of prosecution. In Sweden, waivers of prosecution are typical sanctions if a prison sentence is not the likely outcome on the basis of the penal code, and/or when the offender is young and has no recorded criminal history. Generally, this would mean a larger underestimation of crime in the Finnish data relative to the Swedish. In order to increase comparability, we therefore excluded all waivers of prosecution in the Swedish conviction data.¹ It should also be noted that all court convictions relate to first instance (i.e. district court judgments) and that we were not able to measure changes in judgments in the event of any appeals (e.g. acquittals).

We defined an omnibus measure of violent crime and also separated violent crime into 10 categories (for penal codes, see Table S2 in the supplementary file): Lethal (homicide, manslaughter, attempted homicide, infanticide), Aggravated robbery, Robbery, Aggravated assault, Assault, Violence against a public servant, Threat, Unlawful possession/distribution of knife or equivalent (including stiletto, knuckles), Unlawful possession/distribution of firearm, and Other violent crime (e.g. causing another's death, kidnapping, unlawful coercion, unlawful intrusion, stalking, rioting, inciting rebellion).

Background variables

Besides birth cohort, we measure four background variables: sex, foreign background, residence, and family income.

Sex was available in the total population register and is based on a legal definition divided into male and female.

Foreign background was generated as a three-category variable: (1) Native was defined as an individual born in Finland/Sweden with at least one Finnish/Swedish-born parent; (2) first-generation migrant was defined as an individual who was born abroad; and (3) second-generation migrant was defined as an individual born in Finland/Sweden to foreign-born parents. Country of birth was available in the total population register, allowing us to separate between individuals born in Finland/Sweden and abroad. In order to separate between natives and descendants, we linked the multi-generational register to the study population, thus allowing us to identify the country of birth for the parents. In cases where there was only information for one parent among the Swedish/Finnish-born, this parent's country of birth was used in creating the categorization.

Residence was a three-category variable based on the municipality where the individual resided when he/she was 15 years old: (1) Urban included the main metropolitan municipalities which were Helsinki (including Espoo and Vantaa), Turku, and Tampere in Finland and Stockholm, Gothenburg, and Malmoe in Sweden; (2) Mixed included individuals residing in a municipality that were not classified as either urban or rural; and (3) Rural where more than 50% of the population lives in areas defined as rural by the Swedish Agency for Economic and Regional Growth in Sweden, whereas in Finland these include municipalities where < 60% of the population lives in the largest town that is smaller than 15,000 inhabitants (or 60–90% live in a town with maximum of 4000 inhabitants.

Family income was based on a measure of disposable household income when the individuals were 16 and 17 years old. This is an indicator of socioeconomic family background. We took the average disposable household income during these 2 years and generated a three-category variable divided into the lowest 20%, the mid-60%, and the top 20%. This variable was calculated separately for each birth cohort and should therefore be seen as a relative, rather than an absolute, measure of economic conditions in the family.

Analytical approach

Longitudinal data on multiple successive birth cohorts make it possible to study how crime is distributed between (inter-cohort) and within (intra-cohort) birth cohorts (Bäckman et al., 2014; Sivertsson et al., 2021; von Hofer, 2014). Figure 2 illustrates the data structure whereby all individuals born between 1986 and 1998 who resided in the country at age 15 were followed prospectively between ages 15 and 20, that is, a period of 6 years.² We aggregated these individual-level data to the birth cohort level for all analyses.

Figure 2.

Data coverage of birth cohorts born 1986–1998 followed from age 15 through 20.

We base our trend analysis on four main measures as defined in the criminal career literature (Blumstein et al., 1988): incidence, participation, frequency, and age of onset. Incidence and participation relate to the full birth cohort, whereas frequency and age of onset relate to the subpopulation of convicted violent offenders in each birth cohort. Incidence is calculated as the sum of all convicted violent crimes in a given birth cohort divided by the number of individuals in that cohort. Participation distinguishes offenders from non-offenders and is calculated as a percentage by taking the number of individuals who were convicted by the birth cohort size. Frequency is a property of the offending population which reflects the degree of individual criminal activity by active offenders. Frequency is calculated as an average by taking the sum of convicted violent crimes in a given cohort divided by the number of offenders convicted for violent crime in that cohort. Finally, we calculated the mean age at first violent offense by subtracting the birth date from the date of the first convicted violent offense.³

To reiterate, the concentration effect posits that fewer and fewer young people are subject to the criminal justice system, and that those who are, are increasingly comprised of high-frequent and/or severe offenders. Ideally, this would be expressed as decreasing rates of participation (i.e. across complete birth cohorts) and a simultaneously increasing frequency rate and decreasing age of onset (i.e. within the subpopulation of convicted violent offenders across birth cohorts). Furthermore, the frequency rate would comprise more serious types of violent crime. In our data, this could, for example, be expressed as an increase in lethal violence and/or a shift toward aggravated forms of assault and robbery (as compared to the normal degree of these crime types).

In the analyses of specific crime types and demographic subgroups, we collapsed the birth cohorts into three groups (1986–1989, 1990–1994, 1995–1998). This was needed to facilitate the interpretation of change across the cohorts, as the breakdown concerns unusual types of crime and, in some cases, small groups in the population. In this regard, we also calculated absolute and relative change, comparing the youngest (born 1995–1998) to the oldest (born 1986–1989) cohort. Absolute change was calculated as a percentage (participation) or mean (frequency and age of onset) difference. Relative change was calculated as a ratio and expressed as a percentage change.

Finally, we want to highlight that our analyses are based on the total population of registered residents in the respective country (e.g. as compared to a random sample of residents). All changes should, in that sense, be considered to be free from sampling error. Therefore, we have not conducted any inferential tests. We also want to emphasize that our focus is on the overall trends over the study period and not on annual fluctuations.

Results

In total, the Finnish data included 820,959 residents, 65,956 convicted violent crimes, and 30,246 convicted violent offenders. The corresponding numbers in the Swedish data were 1,502,376 residents, 116,650 convicted violent crimes, and 51,707 convicted violent offenders. Measured over the age period 15–20, 3.7% of residents in Finland were convicted of a violent crime (participation), and they were convicted of an average of 0.08 violent crimes (incidence). The corresponding figures among residents in Sweden were 3.4% and 0.08 violent crimes.

Figure 3 shows the country- and birth cohort-specific developments of four violent crime outcomes: incidence, participation, frequency, and mean age at first violent offense. The key observation from the figure is that the two countries appear very similar in terms of both level and development of the outcomes. Incidence and participation have both decreased, especially in the youngest birth cohorts. Frequency, on the other hand, has remained stable, which shows that the declining incidence is driven by declining participation. Age of onset has also remained stable and is on average 8 months lower in Sweden than in Finland. Hence, the convicted violent population has decreased over the studied birth cohorts, but violent offenders are similar in terms of frequency and age of onset regardless of birth cohort. These patterns are remarkably similar in Finland and Sweden.

Figure 3.

(a–d) Incidence, participation, frequency, and mean age of first convicted violent offense by birth cohort and country.

In Table 1, we take a closer look at developments of participation by type of violent crime. While the relative contribution of different offense types to the overall participation of violence differs somewhat between countries, assault offenses are the main driver of overall trends in both countries. Participation in assaults has declined almost identically (in relative terms, ∼40%) in both countries. Interestingly, in Sweden participation in violence has declined in all crime types except lethal violence, whereas in Finland lethal violence has decreased, while robberies (especially aggravated robberies) and assaults against public officials have increased in the younger cohorts. Hence, the main similarity between the two countries is that fewer and fewer youth are convicted for assault crimes. There is some degree of divergence between the developments of types of violent crimes between the countries. Given that lethal violence, and to some extent robbery, are to be considered more serious forms of violence, this is an indication for a change in participation toward more serious forms of violence that is masked by the omnibus measure and the decline of assault.

Table 1.

Percentages of convicted offenders by type of violent crime, grouped birth cohort, and country.

(a) Finland	Grouped birth cohorts			Absolute % change (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–98	Absolute % change (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
Homicide/manslaughter	0.036	0.033	0.033	−0.003	−9
Aggravated robbery	0.028	0.028	0.053	0 . 025	89
Robbery	0.208	0.219	0.233	0.025	12
Aggravated assault	0.181	0.143	0.121	−0.060	−33
Assault	2.756	2.324	1.664	−1.092	−40
Violence against a public servant	0.712	0.791	0.767	0.055	8
Threat	0.246	0.282	0.252	0.006	2
Possession/distribution of knife	1.073	0.877	0.672	−0.401	−37
Possession/distribution of firearm	0.197	0.124	0.167	−0.030	−15
Other violent crime	0.570	0.514	0.417	−0.153	−27
Total violent crime	4.247	3.738	2.961	−1.286	−30
N	250,326	329,064	241,569

(b) Sweden	Grouped birth cohorts			Absolute % change (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–1998	Absolute % change (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
Homicide/manslaughter	0.021	0.020	0.023	0.003	13
Aggravated robbery	0.020	0.017	0.015	−0.005	−25
Robbery	0.361	0.361	0.312	−0.049	−14
Aggravated assault	0.254	0.202	0.135	−0.119	−47
Assault	2.281	2.110	1.328	−0.952	−42
Violence against a public servant	0.629	0.591	0.511	−0.119	−19
Threat	0.651	0.627	0.479	−0.172	−26
Possession/distribution of knife	0.648	0.601	0.621	−0.028	−4
Possession/distribution of firearm	0.267	0.204	0.201	−0.066	−25
Other violent crime	0.802	0.780	0.563	−0.239	−30
Total violent crime	3.851	3.614	2.716	−1.135	−29
N	463,593	627,534	411,249

Note: Individual items do not sum up to ‘total violent’, as the same offender can commit multiple offense types. Absolute and relative increases in participation are bolded.

In Table 2, we examine the extent to which the total frequency measure, as illustrated in Figure 3(c), masks any degree of heterogeneity between types of violent crime. As with participation, the frequency of assaults has declined in the two countries, in relative terms by 18% in Finland and 25% in Sweden. The decline also concerns aggravated assaults, although it is modest in Finland (only 4%). This decline is the reason why the omnibus frequency measure is relatively stable over the studied cohorts, as the frequency measures for all other types of violent crime have increased, the only exceptions being knife and gun possession/distribution in Finland. Hence, this indicates a marked qualitative shift in terms of the types of violent crimes in convicted youth born during the mid- to late 90s to that of convicted youth born only a decade earlier. Again, since homicide/manslaughter, and arguably also robbery, are in general more serious forms of violence than assaults, this is an indication that the subpopulations of convicted offenders have become increasingly serious.

Table 2.

Mean frequency in convicted violent offenders by type of violent crime, grouped birth cohort, and country.

(a) Finland	Grouped birth cohorts			Mean difference (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–1998	Mean difference (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
Homicide/manslaughter	0.009	0.010	0.014	0 . 005	50
Aggravated robbery	0.008	0.008	0.021	0.013	163
Robbery	0.074	0.087	0.124	0.050	68
Aggravated assault	0.049	0.045	0.047	−0.002	−4
Assault	1.145	1.071	0.941	−0.204	−18
Violence against a public servant	0.222	0.267	0.341	0.119	54
Threat	0.071	0.099	0.108	0.037	52
Possession/distribution of knife	0.394	0.323	0.335	−0.059	−15
Possession/distribution of firearm	0.055	0.038	0.071	0.016	29
Other violent crime	0.167	0.177	0.169	0.002	1
Total violent crime	2.196	2.133	2.181	−0.015	−1
N	10,645	12,377	7224

(b) Sweden	Grouped birth cohorts			Mean difference (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–1998	Mean difference (1995–1998/1986–1989)	Relative % change (1995–1998/1986–1989)
Homicide/manslaughter	0.006	0.006	0.010	0.005	78
Aggravated robbery	0.005	0.005	0.006	0.001	13
Robbery	0.157	0.168	0.194	0.037	24
Aggravated assault	0.079	0.065	0.057	−0.023	−28
Assault	0.949	0.912	0.708	−0.240	−25
Violence against a public servant	0.285	0.291	0.327	0.041	15
Threat	0.248	0.251	0.255	0.007	3
Possession/distribution of knife	0.209	0.204	0.293	0.084	40
Possession/distribution of firearm	0.077	0.061	0.083	0.007	9
Other violent crime	0.271	0.287	0.283	0.012	4
Total violent crime	2.286	2.251	2.217	−0.069	−3
N	17,855	22,682	11,170

Note: Mean frequency = count in offense type/total number of violent offenders. Absolute and relative increases in frequency are bolded.

Table 3 provides a disaggregation of participation of all violent offenses by background characteristics of youth. Overall, the percentages of convicted violent offenders have decreased across all sociodemographic categories, albeit with some differences in degree. In both countries, the decrease over time has been stronger among men than women. With regard to immigrant background, it seems that in Sweden the decrease has been slightly less for second-generation migrants than for natives and first-generation migrants (see also Bäckman et al., 2021). The differences by residence are also fairly small. In both countries, the decline is larger in urban youth than in youth residing in semi-urban municipalities. A country difference is that the decline in youth residing in rural areas is larger in Sweden than in Finland. In both countries, the decrease in violence has been stronger among children from wealthy families, but the decrease has been more even in Finland. Throughout the follow-up, the differences between opposite ends of family income distribution are somewhat larger in Sweden.

Table 3.

Participation (%) by demographic background, grouped birth cohort, and country.

	(a) Finland			(b) Sweden			Absolute % change (1995–1998/1986–1989)		Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–1998	1986–1989	1990–1994	1995–1998	FIN	SWE	FIN	SWE
Male	6.822	6.006	4.709	6.339	5.891	4.411	−2.114	−1.928	−31	−30
Female	1.556	1.414	1.181	1.224	1.215	0.914	−0.376	−0.310	−24	−25
Native	4.082	3.618	2.840	3.237	3.053	2.108	−1.242	−1.129	−30	−35
First generation	10.618	9.267	6.899	8.769	7.527	5.703	−3.718	−3.066	−35	−35
Second generation	7.407	5.952	5.426	6.828	6.646	4.927	−1.981	−1.901	−27	−28
Urban	4.876	4.258	3.698	4.847	4.554	3.612	−1.178	−1.235	−24	−25
Mixed	4.272	3.685	2.863	3.864	3.578	2.586	−1.409	−1.278	−33	−33
Rural	3.635	3.415	2.476	3.299	3.191	2.492	−1.160	−0.807	−32	−24
Low family income	7.225	6.580	5.190	7.302	7.147	5.740	−2.035	−1.562	−28	−21
Middle family income	3.910	3.374	2.726	3.446	3.140	2.278	−1.184	−1.169	−30	−34
High family income	2.264	2.080	1.528	1.635	1.526	1.030	−0.737	−0.605	−33	−37

Finally, in Table 4, we turn to the population of convicted violent offenders and examine to what extent there are sociodemographic gradients in terms of the frequency of violent crime and the age of first violent offense in the respective countries. As expected, the development of these measures over time across birth cohorts is considerably smaller compared to participation and, overall, indicative of stability rather than change. In Sweden, the most visible changes over time are that female frequency has increased somewhat (by 7%) and that the frequency among first-generation immigrants has decreased somewhat (by 12%). In Finland, the main difference is observed for second-generation immigrants, among whom the frequency of violent crime has increased 14%, whereas among first-generation immigrants it has decreased by 9%. The mean age of first violent offense has largely remained stable across the compared birth cohorts, although a slight change toward earlier ages, with a few months at the most, may be noted. It should also be noted that for the most part, the differences in frequency and age of first violent crime between demographic subgroups are lesser than those observed for participation; larger differences are observed mainly by sex and family income. Overall, the results suggest that the two countries have much in common in cohort-specific trends in recorded violent crime.

Table 4.

Mean frequency and age of first violent offense by demographic background, grouped birth cohort, and country.

	(a) Finland			(b) Sweden			Mean difference (1995–1998/1986–1989)		Relative % change (1995–1998/1986–1989)
	1986–1989	1990–1994	1995–1998	1986–1989	1990–1994	1995–1998	FIN	SWE	FIN	SWE
Mean frequency
Male	2.324	2.260	2.339	2.391	2.336	2.293	0.015	−0.098	1	−4
Female	1.635	1.639	1.675	1.712	1.819	1.829	0.040	0.117	2	7
Native	2.166	2.122	2.186	2.152	2.140	2.081	0.020	−0.071	1	−3
First generation	2.713	2.472	2.465	2.714	2.550	2.401	−0.248	−0.314	−9	−12
Second generation	2.150	2.526	2.457	2.542	2.542	2.495	0.307	−0.048	14	−2
Urban	2.533	2.366	2.613	2.517	2.544	2.423	0.080	−0.095	3	−4
Mixed	2.139	2.145	2.118	2.292	2.213	2.229	−0.021	−0.062	−1	−3
Rural	1.960	1.836	1.771	2.091	2.131	1.989	−0.189	−0.103	−10	−5
Low family income	2.527	2.524	2.504	2.624	2.561	2.521	−0.023	−0.102	−1	−4
Middle family income	2.062	1.963	2.078	2.119	2.083	2.037	0.016	−0.083	1	−4
High family income	1.809	1.820	1.837	1.829	1.836	1.708	0.028	−0.122	2	−7
Age of first violent offense
Male	18.1	18.1	18.0	17.6	17.4	17.5	−0.118	−0.097	−1	−1
Female	17.9	18.1	18.0	17.6	17.4	17.4	0.040	−0.191	0	−1
Native	18.1	18.1	18.0	17.7	17.5	17.5	−0.092	−0.141	−1	−1
First generation	17.8	17.8	17.9	17.2	17.2	17.4	0.105	0.149	1	1
Second generation	17.8	17.7	17.6	17.3	17.1	17.2	−0.213	−0.062	−1	0
Urban	18.0	17.9	17.7	17.4	17.2	17.3	−0.310	−0.052	−2	0
Mixed	18.1	18.1	18.1	17.6	17.4	17.5	−0.004	−0.095	0	−1
Rural	18.2	18.3	18.3	17.7	17.5	17.5	0.091	−0.185	1	−1
Low family income	17.9	18.0	18.0	17.4	17.2	17.3	0.038	−0.048	0	0
Middle family income	18.2	18.2	18.0	17.7	17.5	17.5	−0.162	−0.152	−1	−1
High family income	18.3	18.3	18.2	17.8	17.6	17.8	−0.103	−0.027	−1	0

Note: Absolute and relative increases in frequency and decreases in age of first violent offense are bolded.

Discussion

The aim of this study was to examine between-cohort changes in participation and frequency of violent offending that has led to criminal convictions in Finland and Sweden. Using full population register data, we included birth cohorts 1986–1998 followed between ages 15 and 20 and used similar sample inclusion criteria and similar definitions of violent crime to make the comparison as valid as possible. Our results extend prior crime trends research by providing a nuance by types of violent offenses and sociodemographic background and by a comparison between two Nordic countries which differ in population composition concerning some important sociodemographic predictors for violent crime.

The overall picture concerning between-cohort developments in the two countries is one of remarkable similarity. To begin with, participation and incidence have declined markedly over time in both countries. This decline is largely driven by the single largest violent category, assault offenses, in both Finland and Sweden. This finding should be seen against the strong historical link between alcohol consumption and criminalized violence in both countries (Vuorela, 2018), and the fact that alcohol consumption in the youth population has decreased with a magnitude that is fairly similar to the decline in assaults. More specifically, heavy episode drinking among youth shows a marked, and remarkably similar, decrease from the late 1990s until the mid-2010s—from 49% to 25% in Finland and 43% to 25% in Sweden (Raitasalo et al., 2021). This decline in alcohol consumption is part of broader changes in the ways in which young people socialize and spend their free time, and also involve, among other things, declines in unstructured face-to-face interactions (Ball et al., 2023), which have been shown to be highly correlated with the decline in self-reported violence among school pupils (Baumer et al., 2021; Svensson and Oberwittler, 2021). Since assaults comprise the single most common violent offense type in conviction registers, these overall changes in routine activities among youth would also be important explanations for the decline in total violence. Our findings are consistent with prior research in this respect, but further show that this downward trend is fairly uniform across sex, foreign background, urbanicity of residence, and family income in both Finland and Sweden. These findings underline the generality of the decline in the proportion of youth who engage in physical violence during the first two decades of the 2000s.⁴

However, the results also show deviations from the overall downward trend in juvenile violence in at least two respects, and these patterns are, again, highly similar between the two countries. First, the frequency of violent crimes and the age of first convicted violent crime have remained relatively stable throughout the observation periods. These results mirror prior multicohort studies (Andersen et al., 2016; Berg et al., 2016) in showing that overall declining trends in juvenile crime, in this case, violent crime, tend to be primarily driven by a decreasing population of juvenile violent offenders and not the changes in frequency or recidivism of violent crime among offenders. Second, the relatively stable overall frequency masks a shift in the repertoire of violent offenses in the population of juvenile convicted offenders born in the mid- to late-1990s, compared to their same-age counterparts born only a decade earlier—fewer assaults but more of most other types of violent offenses, including homicide/manslaughter and robbery. Also, while a downward trend is seen in participation of most offense types, participation increases for homicide/manslaughter in Sweden and robberies, especially aggravated robberies, in Finland. In that sense, the shift in the types of violent crimes that juvenile offenders are convicted for during the 2000s in Finland and Sweden is one toward increasing severity.

Our approach of using longitudinal multicohort data aligns with recent developments in criminological life-course research. This line of research suggests that adolescents and young adults are strongly shaped by the historical context in which they grow up (Sampson and Smith, 2021). The marked changes in routine activities among youth during the 2000s, which may largely be described as a unitary trend in Western societies toward less risky behaviors and unstructured face-to-face interactions (Ball et al., 2023), may well capture those structural changes that explain the overall decrease in juvenile physical violence. However, the divergence we point to in terms of participation and frequency, and toward an increasingly hard core of juvenile violent offenders, complicate the notion of a general decline in violent offending. This development would better be described in terms of a polarization (Andersen et al., 2016; Sivertsson et al., 2021) or concentration (McAra and McVie, 2018), whereby those juveniles who come into contact with the criminal justice system are increasingly comprised of frequent and serious offenders and, presumably, marginalized groups of individuals with little opportunity to adapt to conventional society.

However, our subgroup analyses disaggregating participation and frequency of violent crime by sex, foreign background, residential area, and family income showed little indications for divergence in total violent crime across these particular sociodemographic subgroups. Participation declined markedly across all subgroups, and while the frequency and age of onset measures captured some degree of variation, there were no distinct deviations. That being said, given that assault was such a decisive type of crime for the overall development of participation and frequency, there may well be important nuances regarding other types of violence between different subgroups of the population that could vary between the two countries. For example, whereas lethal violence in Finland has declined and continued to be related to conflicts between friends and acquaintances under alcohol intoxication, in Sweden it has increased and also shifted toward conflicts in a criminal milieu where young men from so-called vulnerable areas, often located in the suburban areas of the main metropolitan areas, are both the perpetrators and victims (Suonpää et al., 2024). Given the degree of segregation in terms of foreign background and socioeconomic conditions of these areas, we would expect that these background variables capture some degree of divergence for the development of juvenile lethal violence, at least in Sweden.

Despite the similarities between the two countries, there are also some differences that ought to be highlighted. First, age at first violence offense leading into a conviction is slightly higher in Finland than in Sweden. The similarities in crime patterns can be considered surprising, as prior evidence (Aebi et al., 2021) shows that convictions for both homicides and assaults have been much more common in Finland on the population level. These findings underline the importance of age-graded analyses in comparing national rates of crime. Second, while the differences between countries are small, there are some indications of larger differences in relative declines by some background variables in Sweden. Family income, in particular, showed a stronger gradient in the development of participation in violent convictions in Sweden, where the magnitude of the decline in the low-income category was lower than in the mid- and high-income categories. This means that the share of violent offenders that is recruited from the least well-off families has increased across the studied birth cohorts in Sweden but not in Finland. This development should be noted because Sweden has at the same time had an increase in economic inequality that exceeds the OECD average and the rest of the Nordic region during the 2000s (see Pareliussen et al., 2018). The extent to which these two developments (violent crime and relative deprivation) are connected is an avenue for future comparative research involving Finland and Sweden.

Limitations

Conviction data may be considered a back-end measure of criminal offending, which captures only a small percentage of all acts of violence (Andersen and Skardhamar, 2017). The assumption is that conviction data typically measure to a greater extent more serious and repeated crime as compared to self-reported or police-recorded measures of crime (Kyvsgaard, 2002). However, as we have emphasized, we do not claim to make strong inference on the magnitude of violence in Finland and Sweden, but focus on the temporal development of juvenile violence and the similarities/differences between the two countries in this regard. Therefore, the main potential caveats instead concern judicial and/or administrative changes during the studied time period (von Hofer et al., 2012). Below, we discuss this issue in relation to our study design and findings.

Changes in the development of the clearance rate is a central concern when using official crime records for the purpose of measuring criminal behavior (e.g. McAra and McVie, 2018). The clearance rate for violent crimes has decreased in the two countries since the early 2000s, slightly so in Finland and more markedly in Sweden. However, by following each person over several years from the age of criminal responsibility up to and including age 20, we increase the chance of identifying those offenders who commit more than a few crimes, as the risk to be detected and ultimately convicted increases with the length of the observation period. Common problems with official statistics for the purpose of describing the actual development of crime, such as detection risk and identification, can therefore be assumed to be less with the longitudinal multicohort approach than with traditional annual reports, especially with regard to participation (von Hofer, 2014). Our measure of crime frequency may, however, be more sensitive to a declining clearance rate, thus likely leading to an increasing underestimation of crime frequency over time. If this is the case, it would strengthen rather than undermine our conclusions since the factual divergence between the respective developments of participation and frequency would be even more pronounced than what we show.

The drop in the clearances in Sweden is largely driven by assaults, which have close to halved over the time period. This drop must be seen against the increase in reported assault offenses that has been seen in the Nordic countries since the 1990s (Tonry, 2014; von Hofer, 2014). This development is, in turn, probably linked to an increasing sensitivity toward violence, which in turn has increased the propensity to report minor violence (Estrada, 2001, 2006; Kivivuori, 2014). Again, the decline in violent convictions aligns well with the development of other indicators of juvenile prevalence of violence, such as self-reported violence in school pupils and victimization as measured by hospital inpatient data. The disparity between the two countries (i.e. the more marked decrease in assault clearances in Sweden) is probably partly linked to differences in the counting procedure and police discretion. Reported offenses in Sweden, in contrast to Finland, are registered at the time they are reported; reported acts that are later proven to be non-offenses are not removed, and all offenses listed on the same police report or committed on the same occasion appear as separate offenses (von Hofer et al., 2012: 25).

Finally, the decline in homicide clearances in Sweden must be seen against the changing nature of lethal violence over the time period. To reiterate, firearm-related homicide among first and foremost young men has increased to the degree that total lethal violence in this demographic group has increased (Sturup et al., 2019). During this time period, there has been a considerable decrease in the clearance rate for firearm-perpetrated homicides, but a relatively stable development of clearances for non-firearm-related homicide (Granath and Sturup, 2018). Given the fact that this study concerns the age group for which firearm-related homicide has increased in Sweden, the increase in convictions that we show for lethal violence (both participation and frequency) in Sweden is most probably somewhat underestimated. However, this would strengthen rather than invalidate our conclusion of an increasingly severe subpopulation of juvenile violent offenders, although the differences between Finland and Sweden in this regard may be larger than as implied by our measures. Taken together, we believe that our results give a reliable picture of between-cohort developments of violent crime in Finland and Sweden.

Conclusion

This study indicates that the prevalence of juvenile violent offending has decreased markedly across cohorts of adolescents that grew up during the first two decades of the 2000s in both Finland and Sweden. Assault stands out as the main driver of the overall decline. Moreover, this decline is not concentrated in certain sociodemographic groups but instead appears fairly uniform across the examined background variables in both countries. At the same time, this study gives a less optimistic picture about the development across birth cohorts of offenders. The population of convicted juvenile violent offenders born in the late 1990s offend with the same frequency, have a similar debut age of convicted violent crime, and are recorded for more severe offenses, as their same-age counterparts born just a decade earlier. The central question that these patterns raise is why violent crime in Finland and Sweden has become concentrated in an increasingly smaller group of serious violent offenders. Our results, alongside a growing number of longitudinal multicohort studies (e.g. Andersen et al., 2016; Sivertsson et al., 2021), indicate that the explanations put forward for the decline in youth crime (e.g. changes in routine activities among young people) concern debut crimes and, to a lesser extent, frequency/recidivism in the small group that is registered for crime, in this case, violent crime, at a young age. This could be indicative of a selection mechanism where (presumably) broad layers of low-risk youth refrain from committing street crime while a small group of (presumably) high-risk youth are less susceptible to the proposed mechanisms of social change. A more in-depth analysis of risk factors and how they potentially moderate the relationship between birth cohort and violent crime is needed. The increasing severity among young convicted violent offenders also calls for an increasing need to bridge the gap between criminal career research, which uses a broader definition of criminal offending, and homicide research, which focuses on the most serious cases only. Longitudinal data on multiple birth cohorts are needed to understand whether we are seeing more lasting changes in criminal careers, and how these developments look in comparative terms.

Supplemental Material

sj-docx-1-euc-10.1177_14773708251335533 - Supplemental material for Two of a kind? A comparative multicohort study of juvenile violence in Finland and Sweden

Supplemental material, sj-docx-1-euc-10.1177_14773708251335533 for Two of a kind? A comparative multicohort study of juvenile violence in Finland and Sweden by Fredrik Sivertsson, Mikko Aaltonen, Olof Bäckman, Pekka Martikainen, Felipe Estrada, Joonas Pitkänen, Anders Nilsson and Karoliina Suonpää in European Journal of Criminology

Footnotes

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Aaltonen's research contribution was funded by the Strategic Research Council (SRC) established within the Research Council of Finland (grant no. 352543-352574). Bäckman’s, Estrada’s, Nilsson’s, and Sivertsson’s research contributions were partially funded by the Swedish Research Council (grant no. 2022-05510). Martikainen’s research contribution was supported by the European Research Council under the European Union’s Horizon 2020 Research and Innovation Program (grant agreement no. 101019329), the Strategic Research Council (SRC) within the Academy of Finland grants for ACElife (no. 352543-352572) and LIFECON (no. 345219), and grants to the Max Planck—University of Helsinki Center from the Jane and Aatos Erkko Foundation (no. 210046), the Max Planck Society (no. 5714240218), the University of Helsinki (no. 77204227), and the Cities of Helsinki, Vantaa, and Espoo. The study does not necessarily reflect the Commission’s views and in no way anticipates the Commission’s future policy in this area. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Suonpää’s research contribution was partially funded by the Strategic Research Council (SRC) established within the Research Council of Finland (grant nos. 352600 and 352601).

ORCID iDs

Fredrik Sivertsson

Mikko Aaltonen

Olof Bäckman

Pekka Martikainen

Joonas Pitkänen

Anders Nilsson

Karoliina Suonpää

Supplemental material

Supplemental material for this article is available online.

Notes

References

Aebi

Caneppele

Harrendorf

, et al. (2021). European Sourcebook of Crime and Criminal Justice Statistics–2021. Götingen, Germany: Universitätsverlag Götingen.

Aebi

Linde

(2012) Crime trends in Western Europe according to official statistics from 1990 to 2007. In: van Dijk

Tseloni

Farrell

(eds) The International Crime Drop: New Directions in Research. London: Palgrave Macmillan UK, 37–75.

Andersen

Anker

AST

Andersen

(2016) A formal decomposition of declining youth crime in Denmark. Demographic Research 35: 1303–1316.

Andersen

Skardhamar

(2017) Pick a number: Mapping recidivism measures and their consequences. Crime & Delinquency 63(5): 613–635.

Bäckman

Estrada

Nilsson

, et al. (2014) The life course of young male and female offenders: Stability or change between different birth cohorts? The British Journal of Criminology 54(3): 393–410.

Bäckman

Estrada

Nilsson

, et al. (2021) Criminal convictions and immigrant background 1973–2017 in Sweden–have differences increased or declined? Nordic Journal of Criminology 22(2): 149–168.

Ball

Grucza

Livingston

, et al. (2023) The great decline in adolescent risk behaviours: Unitary trend, separate trends, or cascade? Social Science & Medicine 317: 115616.

Baumer

Cundiff

Luo

(2021) The contemporary transformation of American youth: An analysis of change in the prevalence of delinquency, 1991–2015. Criminology: An Interdisciplinary Journal 59(1): 109–136.

Baumer

Vélez

Rosenfeld

(2018) Bringing crime trends back into criminology: A critical assessment of the literature and a blueprint for future inquiry. Annual Review of Criminology 1(1): 39–61.

10.

Beckley

Kardell

Sarnecki

(2017) Immigration and crime in Sweden. In: Pickering

Ham

(eds) The Routledge Handbook on Crime and International Migration. London: Routledge, 41–54.

11.

Berg

Baumer

Rosenfeld

, et al. (2016) Dissecting the prevalence and incidence of offending during the crime drop of the 1990s. Journal of Quantitative Criminology 32(3): 377–396.

12.

Blumstein

Cohen

Farrington

(1988) Criminal career research: Its value for criminology. Criminology: An Interdisciplinary Journal 26(1): 1–35.

13.

DeLisi

Piquero

(2011) New frontiers in criminal careers research, 2000–2011: A state-of-the-art review. Journal of Criminal Justice 39(4): 289–301.

14.

Estrada

(2001) Juvenile violence as a social problem. Trends, media attention and societal response. British Journal of Criminology 41(4): 639–655.

15.

Estrada

(2006) Trends in violence in Scandinavia according to different indicators: An exemplification of the value of Swedish hospital data. British Journal of Criminology 46(3): 486–504.

16.

Estrada

(2022) Ungdomsbrottslighetens utveckling, omfattning och struktur. In: Den svenska ungdomsbrottsligheten. (Femte upplagan). Lund: Studentlitteratur, 35–78.

17.

Estrada

Bäckman

Nilsson

(2016) The darker side of equality? The declining gender gap in crime: Historical trends and an enhanced analysis of staggered birth cohorts. The British Journal of Criminology 56(6): 1272–1290..

18.

Farrell

Tilley

Tseloni

(2014) Why the crime drop? Crime and Justice 43(1): 421–490.

19.

Gerell

Sturup

Magnusson

, et al. (2021) Open drug markets, vulnerable neighbourhoods and gun violence in two Swedish cities. Journal of Policing, Intelligence and Counter Terrorism 16(3): 223–244.

20.

Granath

(2024) Firearm and Criminal Milieu Homicide in Sweden. Journal of Contemporary Criminal Justice 40(3): 493–510.

21.

Granath

Sturup

(2018) Homicide clearance in Sweden 1990–2013 with special reference to firearm-perpetrated homicides. Journal of Scandinavian Studies in Criminology and Crime Prevention 19(1): 98–112.

22.

Haapakangas

(2020) Nuorten rikollisuus on laskussa - mutta pieni joukko nuorista tekee yhä enemmän ja vakavampia rikoksiaTieto and Trendit. https://www2.tilastokeskus.fi:443/tietotrendit/artikkelit/2020/nuorten-rikollisuus-on-laskussa-mutta-pieni-joukko-nuorista-tekee-yha-enemman-ja-vakavampia-rikoksia/ (accessed 7.2.2023)

23.

Hradilova Selin

Krüsselmann

Suonpää

, et al. (2024) Trends in firearm homicide in 23 European countries—Is Sweden an outlier? Nordic Journal of Criminology 25(1): 1–24.

24.

Itskovich

(2024) Economic inequality, relative deprivation, and crime: An idividual-level eExamination. Justice Quarterly: 1–22. https://doi.org/10.1080/07418825.2024.2435859

25.

Kangas

Kvist

(2018) Nordic welfare states. In: Greve

(ed) Routledge Handbook of the Welfare State. Abingdon, UK: Routledge, 124–136.

26.

Kim

Seo

Hong

(2022) A systematic review and meta-analysis of income inequality and crime in Europe: Do places matter?. European Journal on Criminal Policy and Research 28(4).

27.

Kivivuori

(2014) Understanding trends in personal violence: Does cultural sensitivity matter? Crime and Justice 43(1): 289–340.

28.

Kyvsgaard

(2002) The Criminal Career: The Danish Longitudinal Study. New York, NY: Cambridge University Press.

29.

Lappi-Seppälä

Tonry

(2011) Crime, criminal justice, and criminology in the Nordic countries. Crime and Justice 40: 1–32.

30.

Lyngstad

Skardhamar

(2011) Nordic register data and their untapped potential for criminological knowledge. Crime and Justice 40(1): 613–645.

31.

McAra

McVie

(2018) Transformations in youth crime and justice across Europe: Evidencing the case for diversion. In: Goldson

(ed) Juvenile Justice in Europe. London: Routledge, 74–103.

32.

McCarthy

(2021) How universal is the youth crime drop? Disentangling recent trends in youth offending through a socio-economic lens. Victims & Offenders 16(6): 796–818.

33.

Näsi

Aaltonen

Kaakinen

, et al. (2024) Explaining long-term trajectories in youth delinquency: Evidence from survey and register data. Crime & Delinquency. https://doi.org/10.1177/00111287241295982 .

34.

Osgood

(2023) Delinquency, unstructured socializing, and social change: The rise and fall of a teen culture of independence. Criminology: An Interdisciplinary Journal 61(4): 681–704.

35.

Pareliussen

Hermansen

André

, et al. (2018) Income inequality in the Nordics from an OECD perspective. Nordic Economic Policy Review 2018: 17–57.

36.

Payne

Piquero

(2020) Developmental Criminology and the Crime Decline: A Comparative Analysis of the Criminal Careers of Two New South Wales Birth Cohorts. Cambridge University Press. https://doi.org/10.1017/9781108882149.

37.

Pesola

Sarvimäki

Virkola

(2024) Employment of People Immigrating to Finland for Different Reasons. Helsinki: Ministry of Economic Affairs and Employment of Finland. Available at: https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/165475/TEM_2024_9.pdf?sequence=1&isAllowed=y .

38.

Pickett

Molcho

Elgar

, et al. (2013) Trends and socioeconomic correlates of adolescent physical fighting in 30 countries. Pediatrics 131(1): e18–e26.

39.

Pitkänen

Aaltonen

Suonpää

, et al. (2022) Nuorisorikollisuuden esiintyvyyden ja kasautumisen muutokset syntymäkohorteissa 1986–2000. Yhteiskuntapolitiikka 87(5–6), 475–490.

40.

Raitasalo

Kraus

Bye

, et al. (2021) Similar countries, similar factors? Studying the decline of heavy episodic drinking in adolescents in Finland, Norway and Sweden. Addiction 116(1): 62–71.

41.

Sampson

Smith

(2021) Rethinking criminal propensity and character: Cohort inequalities and the power of social change. Crime and Justice 50(1): 13–76.

42.

Sivertsson

Nilsson

Bäckman

(2021) Participation and frequency in criminal convictions across 25 successive birth cohorts: Collectivity, polarization, or convergence? Justice Quarterly 38(6): 995–1018.

43.

Sivertsson

Nilsson

Estrada

(2022) Färre men värre? En kohortanalys av lagförd våldsbrottslighet bland unga män födda 1985–1999 [Fewer but worse? A cohort analysis of convictions among young men born 1985–1999]. In: Rostami

Sarnecki

(eds) Det Svenska Tillståndet [The Swedish Condition]. Lund: Studentlitteratur, 97–123.

44.

Skardhamar

Aaltonen

Lehti

(2014) Immigrant crime in Norway and Finland. Journal of Scandinavian Studies in Criminology and Crime Prevention 15(2): 107–127.

45.

Sturup

Rostami

Mondani

, et al. (2019) Increased gun violence among young males in Sweden: A descriptive national survey and international comparison. European Journal on Criminal Policy and Research 25(4): 365–378.

46.

Suonpää

Kivivuori

Aarten

, et al. (2024) Homicide drop in seven European countries: General or specific across countries and crime types? European Journal of Criminology 21(1): 3–30.

47.

Svensson

Oberwittler

(2021) Changing routine activities and the decline of youth crime: A repeated cross-sectional analysis of self-reported delinquency in Sweden, 1999–2017. Criminology 59(2): 351–386.

48.

Tonry

(2014) Why crime rates are falling throughout the Western world. Crime and Justice 43(1): 1–63.

49.

Tunström

Anderson

Perjo

(2016) Segregated cities and planning for social sustainability—A Nordic perspective. Available at: https://norden.diva-portal.org/smash/get/diva2:1128924/FULLTEXT01.pdf.

50.

Van Breen

Devarakonda

Liem

(2023) Can homicide serve as an indicator of non-lethal crime? A systematic literature review. International Criminology 3(2): 99–115.

51.

von Hofer

(2011) Punishment and crime in Scandinavia, 1750–2008. Crime and Justice 40(1): 33–107.

52.

von Hofer

(2014) Crime and reactions to crime in 34 Swedish birth cohorts: From historical descriptions to forecasting the future. Journal of Scandinavian Studies in Criminology and Crime Prevention 15(2): 167–181.

53.

von Hofer

Lappi-Seppälä

Westfelt

(2012) Nordic Criminal Statistics 1950–2010: Summary of a Report, 8th revised ed. Stockholm: Stockholm University. Available at: https://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-76420 .

54.

Vuorela

(2018) The historical criminal statistics of Finland 1842–2015—A systematic comparison to Sweden. International Journal of Comparative and Applied Criminal Justice 42(2–3): 95–117. https://doi.org/10.1080/01924036.2017.1295395

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB