Juvenile Crime Trends in Post-War Europe

JUVENILE CRIME TRENDS IN POST-WAR EUROPE 23

European Journal on Criminal Policy and Research 7: 23–42, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

FELIPE ESTRADA

JUVENILE CRIME TRENDS IN POST-WAR EUROPE*

ABSTRACT. In the literature, two models – routine activity and social control – aremost often used in attempts to account for a continuous upward trend in the number ofjuvenile offenders during the post-war period. In Sweden, contrary to what we mightexpect given these models, the number of juvenile offenders has been stable, and mayeven have decreased, over the last 25 years. This article will present an analysis of juvenilecrime trends in West European countries during the post-war period (1950–1995). Asensible way to begin a comparative study is to take advantage of the analyses alreadycarried out by researchers in the relevant countries. Besides the official crime statisticsthis study also uses alternative statistics, that is, self-report studies and victim surveys.An obvious advantage with these surveys is that they are independent of the relevantcountry’s judicial system and official statistics. The availability of data played an importantpart in the choice of countries to be included. In addition, contacts were established withresearchers and research centres in most countries covered by the study.

KEY WORDS: control theory, crime statistics, juvenile delinquency, lifestyle, routineactivities theory

INTRODUCTION

Comparative analyses of crime trends show that the level of crime hasincreased in practically all West European countries during the post-warperiod (Gurr 1978; Wilson and Herrnstein 1985; Smith 1995; Westfelt1998). It is also well established that the number of juveniles committingoffences and being registered in the official crime statistics is much greaterin the 1990s than it was in the 1950s. One of the most common explanationsfor this crime trend is the change in routine activities (Cohen and Felson1979). It is Cohen and Felson’s contention that the availability of suitabletargets and the opportunities for crime have increased during the post-warperiod. This development is seen as the result of changes in the structureof legitimate activities, brought about by the modernisation of the westernworld. Another major explanation stems from control theory. As we know,

*Many thanks to Dave Shannon for his translation.

FELIPE ESTRADA24

control theory seeks to explain variations between individuals in thepropensity to commit crime and can thus perhaps complement the routineactivities model, which has been criticised for taking the motivatedoffender for granted. Although control theory is primarily concernedwith conditions at the individual level, when societal crime trends are tobe explained, it is not uncommon to operationalise as structural variablesthose factors which are assumed to affect the delinquency of individuals.Family variables, such as the percentage of the population that is divorcedand the percentage of households headed by women, are seen as powerfulpredictors of crime rates (for examples of this kind of aggregated use ofcontrol theory’s ‘child rearing’ variables see Wilson and Herrnstein 1985;Friday 1992; Smith 1995).

In the literature, the two models mentioned above are most often usedin attempts to account for a continuous upward trend in the number ofjuvenile offenders during the post-war period (e.g. Friday 1992; Smith1995). In Sweden, contrary to what we might expect given these models,the number of juvenile offenders has been stable, and may even havedecreased, over the last 25 years (Estrada 1995; Ward 1998; this point isdiscussed in more detail below). If the situation is not unique to Sweden,it may have important consequences both for the kind of explanationsapplied to juvenile crime trends in post-war Europe and for the formula-tion of criminal justice policies (where attempts to combat juvenile crimeare a central concern).

This article will therefore present an analysis of juvenile crime trendsin West European countries during the post-war period (1950–1995). Twomodels can be formulated to describe these trends:

– the linear upward trend model (the usual description, e.g. Smith 1995);– an initial increase followed by a levelling-off (as in Sweden).

Since these models differ only in respect of the second half of the periodunder study, the following analysis focuses chiefly on those trends thathave characterised the last three decades. The question to be answered iswhich of these two models best describes the trends in juvenile crimefrom a West European perspective.

On the whole, crime trends among juveniles are dominated by theftand criminal damage. In more recent years, however, violence amongyoung people has increasingly become a focus for the interest of thepublic, the media and criminologists alike (Estrada 1997). Since violentcrimes make up a very small part of the total of juvenile offences, changesin the level of violent offending are easily lost in descriptions ofunderlying crime trends. A more specific analysis of the trends in juvenile


violence will therefore be conducted. This analysis is limited to the periodafter 1980.

METHOD AND MATERIAL

There are a number of problems associated with the study of juvenilecrime over a period of several decades. The task is not made easier by theadditional cross-national comparative element. Official crime statisticsare the most accessible source of data, but as we know their use isproblematical. Our knowledge of longitudinal changes in reportingbehaviour, as well as judicial and policing practises, is all too oftenunsystematic and incomplete (for a more extensive discussion see e.g.Maguire 1994; Coleman and Moynihan 1996). It is therefore wise to nourisha healthy scepticism in the face of claims that crime statistics actually describe‘real’ crime trends. In order to use official statistics as a means of interpretingcrime trends, one must be in a position to control for the other factors which,together with criminal activity, affect their content.

A sensible way to begin a comparative study is to take advantage of theanalyses already carried out by researchers in the relevant countries. It isafter all reasonable to assume that these people know a good deal aboutthe factors which affect their domestic crime statistics. Besides theseanalyses this study also uses alternative statistics, that is, self-report studiesand victim surveys. An obvious advantage with these surveys is that theyare independent of the relevant country’s judicial system and officialstatistics. Such surveys are few in number, however, and comparisons overtime present problems.

This study covers the following countries: Austria, Denmark, Englandand Wales (hereafter referred to simply as ‘England’) Finland, (West)Germany,1 the Netherlands, Norway, Scotland, Sweden and Switzerland.The availability of data played an important part in the choice of countriesto be included. A search for relevant literature was carried out in a numberof data bases and there were countries, such as those in southern Europe, forwhich insufficient material could be found.2 In addition, contacts were

1The description of Germany for the years 1950–1995 refers exclusively to that partof the country which was known as West-Germany up until 1989.

2The following electronic data bases were used: Sociofile, Criminal Justice Abstractsand National Criminal Justice Reference Service. These data bases were trawled usingthe following keywords: ((Youth or Juvenile) and (trends or crime trends) or crime rates)),for the years 1980–1996. Abstracts on Criminology and Penology was available only in

FELIPE ESTRADA26

established with researchers and research centres in most countries coveredby the study. Data achieved this way will be referred to below as PI(personal information).3

The validity of the present study depends upon whether its conclusionsare grounded in adequate descriptions of crime trends in the various countries.It is vital that these descriptions are both sufficiently representative andsufficiently reliable.

With regard to representativeness, the issue is the degree to which thestudies referred to contain all of the relevant information available in thecountries in question. Researchers often have differing opinions. How arewe to know that the researchers whose studies form the basis for the presentproject are those whose work best represents the available research in thesecountries? The honest answer is that we cannot be certain. Insofar asresearchers have similar chances of getting their analyses published inscientific journals, however, the review of data bases ought to produce areasonably undistorted sample. One might still object that analyses ofnational crime trends are often not intended for an international audience,and that they are therefore rarely published in international scientificjournals. The collection of data from specific experts in the variouscountries can in this sense be seen as complementary. The contacts withexperts from the Nordic countries, Germany, England and Scotland werefairly extensive, and the material from these countries can be regarded asreasonably representative. In the case of the Netherlands, material wascollected from a wide variety of sources which ought to contribute to ahigh level of representativeness. The representativeness of the materialcollected for Austria and Switzerland is more doubtful.

The question of reliability has to do with the quality of the studies used;not an easy thing to estimate. One approach is to examine the data onwhich the studies are based. The literature concerning crime trends hastaught us that analyses based on a single statistical indicator are oftenfairly unreliable, particularly if the statistics used relate to convicted

book-format and here the search was limited to the sections Juvenile Delinquency,Comparative Analysis, Crime Measurement, Crime Pattern and Time Series Researchfor the years 1980–1996.

3The following researchers have been consulted: Austria: Christian Grafl (Universityof Vienna); Denmark: Britta Kyvsgaard (Copenhagen University); England: David Smith(University of Edinburgh); Finland: Kauko Aromaa and Janne Kivivouri (NationalResearch Institute of Legal Policy); Germany: Christian Pfeiffer (KriminologischesForschungsinstitut, Hannover); the Netherlands: Josine Junger-Tas (University of Leiden);Norway: Sturla Falck (NOVA, Oslo); Scotland: Joe Curran (Scottish Office).


offenders. In short, any attempt to produce an ideal description ofinternational trends in juvenile crime should take the following factors intoaccount.

1. The analysis should be based not only on official crime statistics, butshould also utilise other statistics that are less affected by changes inthe criminal justice system or in the methods used to produce the officialstatistics. Repeated self-report studies would constitute the bestalternative but sadly such studies exist only in Denmark, the Netherlandsand Sweden. There are victim surveys available for a larger number ofcountries but these present problems partly because they do not alwaysinclude a separate presentation of juvenile victimisation, and partlybecause their relevance for juvenile crime trends lies primarily in thefield of violent crime.

2. Official crime statistics should be used to complement these alternativestatistical sources. Statistics should be used which lie ‘nearer’ to theactual crime event than court convictions. One example would be touse crimes reported to the police to study trends in those specific typesof crime for which juveniles constitute the majority of offenders.

3. Statistics relating to identified offenders should also be analysed. Onceagain it is better to use those statistics that lie ‘closer’ to the crime –statistics relating to suspected rather than convicted offenders.

4. In the literature, the reduction in the clearance rate for property offencesis seen to indicate a reduced risk of detection (e.g. Schneider 1991;Ahlberg 1992). It is therefore wise to analyse this type of crimeseparately, thus producing a control for changes in the detection rate.Presenting trends for more than just one category of offenders (e.g.juveniles and adults) allows for a further control for such changes.Individual differences in detection risk are often small, and thus trendsimilarities over different offender categories offer more support to theidea of a diminishing detection rate than do trends which are dissimilar.

5. Where possible, a presentation of the domestic debate regarding juvenilecrime trends and the reliability of the data contributes to the analysisby making it possible to judge the relative worth of the various indicators.

Validity should thus be seen as less satisfactory for those countries wheredescriptions are based on the analysis of a very few indicators such as‘persons convicted for all offence types’ (Table I). In those cases wherethe description is based on a number of sources including alternativestatistics, and where a discussion of juvenile crime trends is included,validity is much improved. On the basis of these criteria, I estimate the

FEL

IPE E

STR

AD

A28

TABLE I

Collation of validity indicators.

Convicted All Crimes Good Juveniles Suspects Victim Theft Domestic Detection Self-reportpersons crime of represen- and surveys debate rate studies

types violence tativeness adults

Sweden + + + + + + + + + + +Norway + + + + + + + + + + –Denmark + + + + + + + – + – +Netherlands + + + + + + + + – – +Germany + + + + + + – – + + –England + + + + + – + + + – –Finland + + + + + + + + – – –Scotland + + + + – – + – – – –Switzerland + + + – + – – + – – –Austria + + – – – + – – – – –

+ = yes – = no. Explanation of the column headings: Convicted persons – statistics covering convicted persons. All crime types – statistics coveringjuvenile crime trends for all types of crime. Crimes of violence – statistics covering trends for crimes of violence committed by Juveniles. Goodrepresentativeness – see above. Juveniles and adults – separate presentation of crime trends for juveniles and adults. Suspected persons – statisticscovering persons suspected of having committed offences. Victim surveys – victim surveys that allow for comparisons over time. Theft – statisticscovering trends in thefts committed by juveniles. Domestic debate – a documented domestic debate regarding juvenile crime trends. Detection rate– a control for possible changes in the detection rate. Self-report studies – self-report studies that allow for comparisons over time.


validity to be good for Denmark, England, Germany, the Netherlands,Norway and Sweden. For Finland and Scotland the level of validity isjudged to be acceptable, and for Austria and Switzerland the validity israther less satisfactory.

This study is based on data that was produced for other purposes. Forthis reason the indicators used, the periods studied and the age range forthe juveniles included vary to a certain extent between the differentcountries. Since my primary interest is in general trends and not precisecomparisons of crime levels, this is not a significant problem. Thedifferences between countries are acceptable. The official statistics for mostof the countries cover the age group 14–20 and the period 1970–1994.

RESULTS: JUVENILE CRIME TRENDS IN POST-WAR EUROPE

Trends for the separate countries are presented below. In brief, the resultsindicate that the post-war upward trend is broken in Denmark, theNetherlands, Norway, Scotland and Sweden. Austria and Switzerland showa similar pattern but here the data is considered weaker. In Finland, Englandand Germany the upward trend has continued. Juvenile violence appearsto have increased in almost every country during the 1990s.

Western and Central Europe

The material from Austria is insubstantial. There is no data concerningcrime trends prior to the 1970s and likewise no alternative statistics. Sincethe beginning of the 1970s the number of convicted juveniles has fallenby more than half (Grafl 1989; Bogensberger 1992). This is assumed tobe primarily the result of changes in working practice within the criminaljustice system. Statistics in respect of juvenile suspects show that theirnumber has been more or less stable since the beginning of the 1970s(Bogensberger 1992). One fact which lends added credibility to the officialcrime statistics is that the proportion of juveniles suspected of more seriousoffences has shown a continuous decline during this period (Grafl 1989).If it were simply that the Austrian police were giving informal, non-registered cautions more and more often, then the proportion of individualsregistered as suspects in connection with more serious crimes would beexpected to rise rather than fall.

Registered juvenile crime in (West) Germany rises from the beginningof the 1950s (Kaiser 1992; Walter 1996). During the first half of the 1980s,however, the number of convicted juveniles starts to fall sharply (Kaiser

FELIPE ESTRADA30

1992; Jehle 1996) whilst for the most part, statistics relating to juvenilesuspects show an increase for the whole of the period 1954–1994 (Walter1996). The point is often made that the reduction in German clearancerates may indicate a drop in the detection rate, a situation that leads toan underestimation of juvenile crime trends where official crime statisticsform the basis for conjecture (Schneider 1991). Statistics relating tosuspects portray violent juvenile crime as having been stable in (West)Germany during the period 1980–1988 (Pfeiffer et al. 1996). The levelof registered violent juvenile crime rises sharply from 1989 onwards. Theextent to which this increase might be explained by changes in reportingbehaviour is difficult to determine in the absence of German victimsurveys.

Between 1955 and 1980 there is a sharp increase in the number ofjuveniles convicted in the Netherlands (Junger-Tas 1988, 1992), a trendwhich is broken at the start of the 1980s. The number of juvenile suspectsis stable during the period 1980–1992 whilst the number of adult suspectsincreases by about 80% (Junger-Tas 1992, PI; SCP 1995). Juvenile self-report studies also suggest a stable pattern in the crime trend at this time.These studies reveal a stable level of self-reported crime for the period1986–1994 (Junger-Tas 1992, PI). In the same way, victim surveyssuggest that the level of crime increased between 1976 and 1979 but thenremained more or less stable during the period 1980–1992 (Franke 1994;Kester and Junger-Tas 1994). Violent juvenile crime on the other handhas increased dramatically in the Netherlands from the end of the 1980sonwards (Junger-Tas 1996, PI). Part of this increase is assumed to be dueto changes in reporting behaviour since the levels of both fatal violence andthe violence reported in victim surveys have remained more or less stablesince 1980, whilst complaints to the police regarding serious violence andattempted murder have increased sharply (Franke 1994). Self-report studiesindicate nonetheless that the level of violent crime committed by Dutchjuveniles has increased (Kester and Junger-Tas 1994; Junger-Tas 1996, PI).

The data from Switzerland is insubstantial. No statistical data relatingto suspects, nor any alternative statistics have been found. The number ofconvicted juveniles doubled during the period 1955–1970 (Niggli andPfister 1997). This trend is broken at the beginning of the 1970s, and thereare no increases in the level of juvenile crime thereafter (Eisner 1993; Reber1993; Niggli and Pfister 1997). The level of reported crime, and specificallytheft, has not increased since 1982 (Niggli and Pfister 1997). Swiss trendsin juvenile violence have been a matter for debate. Reber (1993) suggeststhat levels did not increase between 1984 and 1990. Eisner (1993) agreesthat the 1980s were a period of stability but feels that this trend was broken


at the start of the 1990s. This position is supported by the findings of Niggliand Pfister (1997), who show that aggregate levels of violent crime haveincreased in Switzerland during the 1990s.

Great Britain

A description of juvenile crime trends in England is made more complicatedby the fact that the procedures whereby crime is recorded have changedduring the period under study (Rutter and Giller 1983; Farrington 1992).At the end of the 1960s the police registration procedures changed so thata larger proportion of police cautions were registered than had previouslybeen the case (Rutter and Giller 1983). In the mid-1980s the police onceagain altered their procedures, this time in the opposite direction (Farrington1992). Comparability over the whole of the period under study is thereforeless than satisfactory. It ought, however, to be possible to describe crimetrends within the following periods: 1950–1968, 1970–1985 and 1989–1995.

The number of juveniles in England either convicted or cautioned risessharply from the mid-1950s (Rutter and Giller 1983). This strong upwardtrend is broken in the mid-1970s (Rutter and Giller 1983). According tothe official crime statistics there is no increase during the period 1974–1979. Between 1979 and 1985, however, the level of registered juvenilecrime climbs once again (Home Office 1980–1996). During the period1985–1995, the statistics indicate a substantial reduction in juvenile crime.In contrast, the levels of crime committed by adults display an upward trendthat is on the whole unbroken throughout this period (Rutter and Giller1983; Home Office 1980–1996).

Farrington (1992) is sceptical about the reduction in juvenile crime. Hesuggests that the apparent decrease is a result of the procedural changesconcerning the registration of crime by the police. During the period 1981–1995, the level of crime increases according to both statistics for crimesreported to the police (+90%), and victim surveys (+80%) (Mirrlees-Blacket al. 1996). Both Farrington (1992) and Smith (1995) feel that this isindicative of a continued increase in juvenile crime. If we accept thisinterpretation, then juvenile crime in England, with the exception of a fewyears during the 1970s, has risen more or less continuously throughout thepost-war period.4

4It is nevertheless possible to find indicators which suggest that the ‘levelling off’ thatcertain researchers identify in the mid-1970s (Rutter and Giller 1983) has continued. It isinteresting, for example, that reported crime only increased by 7% during the years 1985–1989, which Farrington identifies as the period where the statistics relating to juvenile

FELIPE ESTRADA32

The number of juveniles convicted and cautioned for violent offencesis stable during the 1980s, and falls during the 1990s. However, the levelof violent crime as evidenced by both victim surveys and statistics relatingto crimes reported to the police increases conspicuously (Mirrlees-Blacket al. 1996). Violent crimes against young people (aged 16–29) increaseby about 70% between 1981 and 1995 (Mirrlees-Black, PI). If we acceptFarrington’s criticism of the statistics concerning persons convicted orcautioned, it is reasonable to assume that violent juvenile crime in Englandhas increased continuously since 1980.

As in most other West European countries, the level of registered crimein Scotland rises sharply from the mid-1950s (Young 1997). The numberof convicted juveniles increases throughout the 1970s. There is a levellingoff, however, at the beginning of the 1980s (Scottish Office 1979–1995).This trend is confirmed by alternative statistical sources. Victim surveysshow no increases in the level of crime between 1981 and 1992 (Andersonand Leitch 1996). Neither victim surveys nor statistics relating to convictedoffenders indicate an increase in the level of violent crime in Scotland(Scottish Office 1979–1995; Anderson and Leitch 1996).5

The Nordic Countries

In Denmark the post-war upward trend in the number of convicted juvenilesis broken in the mid-1970s (Kyvsgaard 1992, pp. 23, 30). This finding issupported by a study of juveniles’ self-reported criminality for the years1979 and 1989 (Kyvsgaard 1992). The number of convicted adults hascontinued to rise during this period (Kyvsgaard 1992). In Denmark thetrend in juvenile violence shifts from relative stability during the 1980s toan upward trend in the 1990s (Balvig 1995). Danish victim surveys showthat juveniles are being exposed to more violence (Balvig 1995). However,

crime are most unreliable (Farrington 1992; Home Office 1980–1996). Nor is there anupward trend during the 1990s. Between 1991 and 1995 reported crime falls by 3% andthe number of registered juveniles (aged 14–17) by 4% (Home Office 1980–1996). Theexception to this stabilising trend is the sharp rise in reported theft during the period1989–1991 (an increase of 40% in just two years). The portrayal of an unbroken upwardtrend in juvenile crime can thus be called into question.

5It is interesting that England and Scotland display such contrasting crime trends. Thisphenomenon has only recently become a focus for attention within British criminology(Anderson and Leitch 1996; Young 1997). At the beginning of the 1980s the differenceswere felt to be smaller (Smith 1983; Mayhew and Smith 1985). No explanations for thecontrasting trends have as yet been presented.


a study that compares hospital data with police data indicates that theviolence among juveniles has not increased. Nor has there been an increasein the level of fatal violence (Brink et al. 1997). What has increased is thewillingness, especially among the youngest (15–19) to report the case tothe police (Brink et al. 1997).

In Finland crime statistics portray the upward trend in juvenile crimeas more or less continuous from 1950–1990 (Törnudd 1984; Kivivouri,PI). During the 1990s this trend is broken and the number of juvenilesuspects begins to decline instead. The registered crime of adults followsthe same pattern. However, victim surveys covering the period 1980–1996show no increase (Aromaa 1991; Heiskanen and Aromaa 1996). In thematter of juvenile violence, assault statistics shows an increase from 1980.At the same time victim surveys reveal a reduction in the victimisation ofjuveniles for the period 1980–1988 (Aromaa 1991).

In Norway the sharp upward trend in juvenile crime is broken at thebeginning of the 1970s (Christie 1982; Falck 1994). Clausen (1996) hasdemonstrated that it is primarily adults who are the cause of the increasein reported crime for the period 1980–1992. However, crime statisticsreveal a strong increase in the level of violent crime committed by juvenilessince the mid-1980s (Falck 1994; Clausen 1996). The proportion of youngNorwegian males (aged 16–24) who according to victim surveys have beenthe victims of a violent crime remained at a fairly stable level between 1983and 1991 (Olaussen 1995, PI). The figures for 1995, however, representan increase.

Between 1950 and 1972, the number of juveniles convicted of theftin Sweden increases. From 1972 juvenile crime as measured by thesesame statistics falls. This has been the subject of much debate (Estrada1995). Several analyses of the various components of the official crimestatistics show it to be unlikely that the dramatic post-war increase intheft by juveniles should have continued into the 1970s (Estrada 1995;Von Hofer 1997). Self-report studies confirm that the number of juvenileoffenders has not increased since the beginning of the 1970s (Ward 1998).The rise in crime since the beginning of the 1970s is instead due to anincrease in the amount of crime committed by adults. The trends in violentcrime committed by juveniles since 1980 are more difficult to interpret.Distinct increases in assaults by juveniles can be seen in the officialstatistics from about 1987 (Von Hofer 1995; Olsson and Lindström 1995).At the same time juvenile violence leading to the death of the victim hasbeen stable since 1980. Victim surveys and self-report studies also showa stable pattern (Qvarnström and Mårtensson 1996; SCB 1998; Ward1998).

FELIPE ESTRADA34

SUMMARY OF RESULTS

A summary of the results together with references to the most importantsources is presented in Table II. It shows that the underlying trend for thenumber of juvenile offenders follows the ‘levelling off’ hypothesis in amajority of the countries included in the study (and perhaps mostimportantly in four of the six countries where the data was judged to bemost reliable). This means that the post-war period is not characterisedby an ever expanding population of young criminals. The exceptions tothis rule are England, Germany, and perhaps Finland, where post-warupward trend seems to have continued at a more or less undiminished rateduring the entire period.

In several countries trends in juvenile violence deviate from theunderlying crime trends (we have to remember that violence onlyaccounts for a small proportion of the total of juvenile offences). In allcountries the crime statistics display a rise in juvenile violence. InEngland and Finland this rise has been continuous throughout the period1980–1995. The increase in Denmark, Germany, the Netherlands,Norway, Sweden and Switzerland is probably the result of a trend-shiftat the end of the 1980s. Junger-Tas (1996) shows that it is primarilyviolence of a less serious nature which has contributed to the rise injuvenile violence in various European countries. Violence leading to thedeath of the victim has not increased (Junger-Tas 1996). Junger-Taspoints out that it is difficult to quantify the size of the increase since crimestatistics and alternative statistics paint somewhat different pictures. Thepresent study shows that this is especially true for the Scandinaviancountries. In Western and Central Europe, however, both alternativestatistics and crime statistics indicate that the level of violent juvenilecrime has increased during the 1990s. The question of recent trends injuvenile violence certainly merits further research.

The German criminologist Pfeiffer (1997) has recently carried out ananalysis of juvenile crime in Europe. He describes juvenile crime trendsas falling in Denmark, England, Switzerland and Sweden. The trends inGermany and Austria are judged to be stable until the end of the 1980s, atwhich point they begin to rise. Juvenile crime in the Netherlands ispresented as having been stable since 1985. Finland, Norway and Scotlandare not mentioned. There are certain differences between Pfeiffer’sinterpretation and my own but the general picture is much the same. Wherethere are differences, these are due in part to Pfeiffer’s reliance on officialcrime statistics as the only source and in part to the fact that his study coversa shorter period (1980–1995). On the question of trends in violent juvenile


TABLE II

Summary of results. Underlying trends in juvenile crime, trends in juvenile violence, key sources.

Country Underlying trend Key sources Trends in juvenile Key sources1950–1995 violence 1980–1995

Austria Model 2: Grafl 1989 – –Stable since firsthalf of 1970s.

Denmark Model 2: Kyvsgaard Contradictory indicators Balvig 1995Sharp upward 1992, PI Increasing during 1990’s Brink et al. 1997trend to first half according to assault Kyvsgaard 1993of 1970s, followed statistics. Stable accordingby ‘levelling off’. to hospital data.

England Model 1: Farrington 1992 Rising throughout the Mirrlees-BlackMore of less un- Mirrlees-Black period according to both et al. 1996; PIbroken increase et al. 1996 assault statistics andthroughout the Rutter et al. victim surveys.period. 1983

Finland Model 1: Kivivouri PI Contradictory indicators. Aromaa 1991More or less un- Törnudd 1984 Increase according to Heiskanen andbroken increase to assault statistics. Decrease Aromaa 19961990, followed by according to victim surveys. PIreduction duringthe 90’s.

Germany Model 1: Kaiser 1992 Stable between 1984– Pfeiffer et al.More or less un- Pfeiffer et al. 1989. Thereafter sharp 1996broken increase Walter 1996 increase during 1990sduring the entire according to officialperiod. crime statistics.

Netherlands Model 2: Junger-Tas 1992; Stable to end of 1980s. Franke 1994Sharp upward trend PI Thereafter rising during Junger-Tas 1996;to beginning of SCP 1995 1990s according to both PI1980’s, followed by assault and alternative Kester et al. 1994‘levelling off’. statistics.

Norway Model 2: Christie 1982 Contradictory indicators Clausen 1996Sharp upward trend Clausen 1996 during 1980s. Increase Olaussen 1995to first half of 1970s, Falck 1994 during 1990s according tofollowed by both assault statistics and‘levelling off’ victim surveys.

Scotland Model 2: Anderson et al. No increase according to Anderson et al.Upward trend to 1996 victim surveys and assault 1996beginning of 1980s, Scottish Office statistics. Scottish Officefollowed by 1979–1995 1979–1995‘levelling off’. Young 1997

Sweden Model 2: Estrada 1995 Contradictory indicators Von Hofer 1995Sharp upward trend Von Hofer 1985 Increase during 1990s Olsson andto first half of 1970’s, Ward 1998 according to assault Lindström 1995followed by statistics. Stable according‘levelling off’. to self-report data and

victim surveys.Switzerland Model 2: Niggli et al. 1997 No increase during 1980s. Eisner 1993

Sharp upward trend Reber 1993 Possible break in this Niggli et al.to beginning of trend and increase 19971970s, followed by during 1990s. Reber 1993‘levelling off’.

crime, Pfeiffer’s results and my own are on the whole very similar. Weboth find indications of increasing violence among juveniles in the 1990s

FELIPE ESTRADA36

according to crime statistics in several countries.A central question is the extent to which the described trends are real,

or simply the result of amongst other things procedural changes in thecriminal justice system or variations in reporting behaviour (see above).The most obvious answer is of course that both explanations are true toa varying extent. In most of the countries studied here, for example,statistics concerning convicted juveniles indicate a clear reduction overthe last 20 years or so. Those indicators which lie ‘closer’ to the crimeevent, however, and which are thus less sensitive to changes within thecriminal justice system (such as statistics relating to suspects), suggestthat the reductions are not real but are rather the result of ‘system effects’(see e.g. Von Hofer 1985; Grafl 1989; Ahlberg 1992; Farrington 1992;Junger-Tas 1992; Estrada 1995; Clausen 1996; Walter 1996). It isnonetheless important to remember that for most countries these sourcesdo not indicate that the number of juvenile offenders has continued toincrease at an undiminished rate during this period.

This interpretation is reinforced by alternative statistics. In thosecountries where self-report studies are available over time they suggest astable level of juvenile offenders. In addition, with the exception ofEngland, the available victim surveys suggest that the underlying crimetrends have probably levelled off over the past 15 years, and that theincreased crime levels portrayed in the official crime statistics are to someextent the result of changes in reporting behaviour (Aromaa 1991; Kesterand Junger-Tas 1994; Olaussen 1995; SCB 1995; Anderson and Leitch1996; Mirrlees-Black et al. 1996). The data from several countries suggestthat the rise in crime that remains to be explained once changes in reportingbehaviour have been taken into account, really ought to be ascribed to adultsrather than juveniles. Crime trends for adult offenders over the last 20 yearsare essentially different from those for juveniles in Denmark, the Netherlands,Norway, Sweden and Switzerland.

In summary, it is likely that trends in juvenile crime are no longer uniformto all West and North European countries. In many countries the availablematerial suggests that the steep upward trend in the number of juvenileoffenders which characterised the first half of the post-war period, has beenbroken. In some countries, however, the increase seems to have continued.

DISCUSSION

It was pointed out in the introduction that explanations of juvenile crimetrends often try to account for a continuous upward trend. This study


indicates, however, that what ought in fact to be explained is a situationwhere the number of juvenile offenders first increases (circa 1950–1975)and then is stable during the subsequent decades. My intention here is tolimit myself to a brief discussion of how well the explanatory factors thatwere presented in the introduction cope with this empirical pattern.

For the routine activities approach, the structural changes affecting theavailability of crime opportunities in the post-war period are of centralimportance. Here we are dealing with an empirical question that goesbeyond the scope of this article. It is nonetheless reasonable to contendthat the sharp post-war rise in the availability of crime opportunities haslevelled off in parallel with the slowing economic trends that followed theoil crisis of 1973 (Korpi 1992, pp. 55, 108). At the same time, faced withthe sharp post-war crime increases, most western countries have made useof situational crime prevention measures in order to reduce or at least limitthe number of crime opportunities (Garland 1996), and these measures haveprincipally been targeted at typical juvenile offences such as theft andcriminal damage. This description fits the observed slow-down in juvenilecrime rather well. A stabilisation in the availability of opportunities forcrime is moreover primarily assumed to affect those groups of juvenileswho have not yet become ‘persistent offenders’ (Wikström 1995). For thosewho are already embarked upon a ‘criminal career’, restrictions in the numberof opportunities for crime are rather less important (Wikström 1995). Thisline of argument could thus also provide us with a way of making sense ofdifferences in the crime trends displayed by adults and juveniles over thelast 20 years. One might, however, object that economic trends have variedfrom country to country (Korpi 1992). In Norway for example, the oil crisisof 1973 initiated a period of accelerated growth rather than an economic slow-down.6 If we, in fact, have a slow-down in juvenile offending whilst structuralconditions still favour a continued increase in crime, then the changing trendsin these countries demand a different explanation.

The arguments from the control theory present problems in that risesin the levels of the important social phenomena (divorce rate etcetera)take place at a time, around 1965–1975 (Hess 1995), which fits withneither the sharp rise in juvenile crime between 1950 and 1975 nor thesubsequent levelling off that can be seen in several European countries.

6A common objection to the importance of the availability of crime opportunities isthat economic trends in Japan have been similar to those in Western European countrieswithout there having been a corresponding rise in the level of crime (see e.g. Wilson andHerrnstein 1985).

FELIPE ESTRADA38

The explanatory value of these factors is thus unclear. Another argumentoften raised in connection with the control approach is that the level ofsocial control exercised by adults over juveniles has been significantly andcontinuously reduced in post-war societies due to a increasing ‘generationgap’ (see e.g. Wilson and Herrnstein 1985, p. 436; Friday 1992; Smith1995). This segregation of young people from adult society has been helpedalong by such factors as the increasing length of educational careers, a laterand later entrance into the labour market and the growth of a youth focusedpopular culture. To evaluate this argument properly would require anextensive comparative analysis of the changing trends in juvenile socialcontrol in Western Europe. There is room here for no more than a briefdescription of the situation in Sweden.

The modernisation of Sweden led to a massive expansion of popularculture which quickly became focused on young people. As early as the1930s and 1940s the increasing ‘gap’ between youth culture and adultsociety was a matter of much debate. After the war, and especially duringthe 1950s and 1960s, the pace of this modernisation process increaseddramatically. In 1960, the number of students aged 15–19 was five timesthe number in 1940 (Bjurström and Fornäs 1988). Between 1960 and1970, the number of teenagers working in Sweden was cut by very nearlyhalf (Bjurström and Fornäs 1988). If the trend towards a generationspecific youth culture can be said to have started in the 1930s and 1940s,then the youth revolts at the end of the 1960s can be seen as the peak ofthis growing divide between the generations. It seems reasonable tosuggest that the development of a gap between young people and the adultworld has not continued at the same pace since the beginning of the 1970s.The generational divide which has impacted upon the social control ofjuveniles may well have been more or less complete by this point.Couched in these terms, the social control model fits much better withthe apparent slow-down in juvenile crime trends than with the hypothesisof a continuous increase with which it is usually associated (e.g. Wilsonand Herrnstein 1985; Friday 1992; Smith 1995). Indeed, self-reportstudies from Denmark and Sweden indicate that the youth of today aremore and not less disciplined than juveniles in the 1970s (Kyvsgaard1992; Ward 1998).

To sum up, this study suggests that the usual accounts of post-war trendsin juvenile crime do not fit well with the factual development. For themajority of the Western European countries, what needs to be explainedis not a continuous upward trend in the number of juvenile offenders butrather an initial increase followed by a levelling off.


REFERENCES

Ahlberg, J., Myten om en minskande ungdomsbrottslighet [Reductions in juvenile crime:The myth]. Apropå, 1, pp. 4–8, 1992.

Andersson, S. and S. Leitch, Main Findings from the 1993 Scottish Crime Survey.Edinburgh: The Scottish Office Central Research Unit, 1996.

Aromaa, K., Deliberate violence. In: M. Heiskanen et al., Accidents and Violence 1988.Rapport 108, pp. 103–128. Helsingfors: Rättspolitiska Forskningsinstutet, 1991.

Balvig, F., Ungdom oplever mere vold: En oversigt over danske voldsofferundersogelser1970–94 [Juveniles are experiencing more violence]. Kopenhagen: Jurist- ogOkonomfobundets Forlag, 1995.

Bjurström, E. and J. Fornäs, Ungdomskultur i Sverige [Youth culture in Sweden]. In: U.Himmelstrand and G. Svensson, Sverige- vardag och struktur: Sociologer beskriverdet svenska samhället. Stockholm: Norstedts, 1988.

Bogensberger, W., Jugendstrafrecht und Rechtspolitik. Wien: Manz, 1992.Brink, O. et al., Mindre vold og hyppigere politianmeldelse [Less violence and more

reports to the police]. Nordisk tidsskrift for kriminalvidenskab, 84, pp. 103–114, 1997.Christie, N., Hvor tett et samfunn? Andra reviderade upplagan [How tightly knit a

society?]. Oslo: Universitetsforlaget, 1982.Clausen, S.E., Barne- og ungdomskriminalitet i Norge 1980–1992 [Child- and juvenile

crime in Norway 1980–1992]. NIBR-rapport 1996(4). Oslo: Norsk Institutt for by- ogRegionsforskning, 1996.

Cohen, L. and M. Felson, Social change and crime rate trends. American SociologicalReview, 44, pp. 588–608, 1979,

Coleman, C. and J. Moynihan, Understanding Crime Data: Haunted by the Dark Figure.Buckingham: Open University Press, 1996.

Eisner, M., Zu- oder Abnahme der Gewaltkriminalität Jugendlicher: Anmerkungen zumBeitrag von Rolf Reber. Kriminologisches Bulletin, 19, pp. 91–98, 1993.

Estrada, F., Ungdomsbrottslighetens utveckling 1975-1994 [Juvenile Crime Trends 1975–1994]. Rapport 1995(1). Stockholm: Kriminologiska Institutionen, 1995.

Estrada, F., Ungdomsvåld: Upptäckten av ett samhällsproblem: Ungdomsbrottslighet isvensk dagspress 1950–1994 [Juvenile violence: The discovery of a social problem:Juvenile crime in the Swedish daily newspapers 1950–1994]. Sociologisk Forskning,97(4), pp. 51–72, 1997.

Falck, S., Hvor lenge skal barn få lov til å vaere barn? [How long are children to beallowed to remain children?]. Oslo: Barnevernets Utviklingssenter, 1994.

Farrington, D.P., Trends in English juvenile delinquency and their explanation. InternationalJournal of Comparative and Applied Criminal Justice, 16, pp. 151–163, 1992.

Franke, H., Violent crime in the Netherlands: A historical-sociological analysis. Crime,Law and Social Change, 21, pp. 73–100, 1994.

Friday, P.C., Delinquency in Sweden: Current trends and theoretical implications. Inter-national Journal of Comparative and Applied Criminal Justice, 16, pp. 231–246,1992.

Garland, D., The limits of the sovereign state: Strategies of crime control in contemporarysociety. British Journal of Criminology, 36, pp. 445–471, 1996.

Grafl, C., Juvenile delinquency in Austria in the eighties. Paper presented at the VIIthInternational Workshop on Juvenile Criminology, Noordwijkerhout, May 29–31,1989.

FELIPE ESTRADA40

Gurr,T.R., Crime trends in modern democracies since 1945. In: The Cranfield Papers:The Proceedings of the 1978 Cranfield Conference on the Prevention of Crime inEurope. London: Peel Press, 1978.

Heiskanen, M. and K. Aromaa, Victimisation in Finland 1996: Survey Findings from theFinnish Part of the International Victimisation Survey. Forskningsmeddelanden (28).Helsingfors: Rättspolitiska Forskningsinstutet, 1996.

Hess, L. E., Changing family patterns in Western Europe: Opportunity and risk factorsfor adolescent development. In: M. Rutter and D.J. Smith (Eds), Psychosocial Disordersin Young People: Time Trends and their Causes, pp. 104–193. New York: John Wileyand Sons, 1995.

Home Office, Criminal Statistics England and Wales. London: HMSO, 1980–1996.Jehle, J.M., Strafrechtspflege in Deutschland: Fakten und Zahlen. Bonn: Bundesministerium

für Justiz, 1996.Junger-Tas, J., Patterns in delinquent behavior. In: J. Junger-Tas and R. Block (Eds),

Juvenile Delinquency in the Netherlands, pp. 13–40. Amstelveen: Kugler, 1988.Junger-Tas, J., Juvenile delinquency in the Netherlands: Trends and perspectives.

International Journal of Comparative and Applied Criminal Justice, 16, pp. 207–230,1992.

Junger-Tas, J., Youth and violence in Europe. Studies on Crime and Crime Prevention, 5,pp. 31–58, 1996.

Kaiser, G., Juvenile delinquency in the Federal Republic of Germany. InternationalJournal of Comparative and Applied Criminal Justice, 16, pp. 185–206, 1992.

Kester, J.G.C. and J. Junger-Tas, Criminaliteit en strafrechtelijke reactie: Ontwikkelingenen samenhangen, Ministry of Justice, WODC. Arnhem: Gouda Quint, 1994.

Korpi, W., Halkar Sverige efter? Sveriges ekonomiska tillväxt 1820–1990 i jämförandebelysning [Is Sweden getting behind? Sweden’s economical development in acomparative illustration]. Stockholm: Carlssons, 1992.

Kyvsgaard, B., The decline in child and youth criminality: Possible explanations of aninternational trend. In: A. Snare (Ed.), Youth, Crime and Justice. Scandinavian Studiesin Criminology, Vol. 12, pp. 26–41. Oslo: Universitetsförlaget, 1991.

Kyvsgaard, B., Ny ungdom? Om familie, skole, fritid, lovlydighed og kriminalitet [Newyouth? On the family, school, leisure time, living within the law and crime].Kopenhagen: Jurist- og Okonomfobundets Forlag, 1992.

Kyvsgaard, B., Kriminalitetens demografi og forandring [Demography and change incriminality]. In: J. Bay (Ed.), Kriminalistisk instituts årbog 1993, pp. 109–123.Kopenhagen: Kriminalistisk Institut, 1993.

Maguire, M., Crime statistics, patterns and trends: Changing perceptions and theirimplications. In: M. Maguire, M., R. Morgan and R. Reiner (Eds.), The OxfordHandbook of Criminology, pp. 233–294. Oxford: Clarendon Press, 1994.

Mayhew, P. and L. Smith, Crime in England and Wales and Scotland: A British CrimeSurvey comparison. British Journal of Criminology, 25, pp. 148–159, 1985.

Mirrlees-Black, C., P. Mayhew and A. Percy, The 1996 British Crime Survey Englandand Wales. Statistical Bulletin, 96(19), 1996.

Niggli, M.A. and F. Pfister, Paradise lost? Paradise ever? On crime development inSwitzerland. Studies on Crime and Crime Prevention, 6, pp. 73–100, 1997.

Olaussen, L. P., Voldskriminalitetens utvikling de to siste tiårene [Trends in violent crimeover the last twenty years]. Nordisk tidsskrift for kriminalvidenskab, 82, pp. 97–116,1995.


Olsson, M. and P. Lindström, Vad säger kriminalstatistiken? [What do the crime statisticstell us?]. In: Forskningsrådsnämnden, Det obegripliga våldet: Forskare om ungdomaroch våld. Källa 46. Stockholm: FRN, 1995.

Pfeiffer, C., Juvenile Crime and Juvenile Violence in European Countries. Hannover:Kriminologisches Forschungsinstitut Niedersachsen, 1997.

Pfeiffer, C. et al., Steigt die Jugendkriminalität wirklich? Hannover: KriminologischesForschungsinstitut Niedersachsen, 1996.

Qvarnström, G. and R. Mårtensson, Drogvanor och erfarenheter av våld bland årskurs 9elever i Stockholms skolor 1996 [Selfreported drug use and violence among schoolpupils in Stockholm]. ANT-informationen, Stockholms Skolor, 1996.

Reber, R., Evidenz für eine Abnahme der Jugendkriminalität in der Schweiz.Kriminologishes Bulletin, 19, pp. 85–99, 1993.

Rutter, M. and H. Giller, Juvenile Delinquency: Trends and Perspectives. Harmondsworth:Penguin Books, 1983.

SCB (Statistiska Centralbyrån), Offer för vålds- och egendomsbrott 1978–1993 [Victims ofviolent- and property crimes]. Levnadsförhållanden, rapport 88. Stockholm: SCB, 1995.

SCB (Statistiska Centralbyrån), Unpublished data from the victim surveys 1994–1997,1998.

Schneider, H.J., Kriminologishe Aspekte der Kinder- und Jugenddelinquenz. Forensia-Jahrbuch, Vol. 2, pp. 177–199, Berlin/Heidelberg: Springer-Verlag, 1991.

SCP (Social and Cultural Planning Office), Social and Cultural Report 1994 TheNetherlands. Rijswijk: SCP, 1995.

Scottish Office, Criminal Proceedings in Scottish Courts. Edinburgh: Scottish Office,1979–1995.

Smith, L., Criminal Justice Comparisons: The Case of Scotland and England and Wales.Research and Planning Unit (17). London: Home Office, 1983.

Smith, D.J., Youth crime and conduct disorders: Trends, patterns, and causalexplanations. In: M. Rutter and D.J. Smith (Eds), Psychosocial Disorders in YoungPeople: Time Trends and Their Causes, pp. 389–489. New York: John Wiley andSons, 1995.

Törnudd, P., Kriminalitetsutvecklingen i Finland 1950–1982 [Crime trends in Finland1950–1982]. Nordisk tidsskrift for kriminalvidenskab, 71, pp. 149–168, 1984.

Von Hofer, H., Brott och straff i Sverige [Crime and punishment in Sweden]. Urval (18).Stockholm: Statistiska Centralbyrån, 1985.

Von Hofer, H., Violence criminelle et jeunes en Suède: Une perspective á long terme.Les politiques sociales, (3/4), pp. 33–45, 1995.

Von Hofer, H., Kohortstatistik: Ett bra sätt att studera unga lagöverträdare ochungdomsbrottslighet [Cohort statistics: A good way to study young offenders andjuvenile crime]. Stockholm: Kriminologiska Institutionen, Stencil, 1997.

Walter, M., Kriminalpolitik mit der polizeilichen Kriminalstatistik? DVVJ-Journal, 4/96(153), pp. 209–214, 1996.

Ward, M., Barn och brott av vår tid? Självdeklarerad ung domsbrottslighet 1971 och1996 [Youth and crimes of our time? Self reported delinquency 1971 and 1996].Stockholm: Kriminologiska Institutionen, 1998.

Westfelt, L., Utvecklingen av registrerade brott under efterkrigstiden: Sverige i europeiskbelysning [Trends in registered crime during the postwar period: Sweden in the lightof the situation in Europe]. In: H. Von Hofer (Ed.), Svensk brottslighet ochkriminalpolitik i europeisk belysning, pp. 20–40. Lund: Studentlitteratur, 1998.

FELIPE ESTRADA42

Wikström, P.O.H., Self-control, temptations, frictions and punishment: An integratedapproach to crime prevention. In: P.O.H. Wikström, R.V. Clarke and J. McCord (Eds),Integrating Crime Prevention Strategies: Propensity and Opportunity. Brå-rapport,1995(5). Stockholm: Brottsförebyggande Rådet, 1995.

Wilson, J.Q. and R. Herrnstein, Crime and Human Nature. New York: Simon and Schuster,1985.

Young, P., Crime and Criminal Justice in Scotland. Edinburgh: Scottish Office, 1997.

Department of CriminologyStockholm UniversityS-106 91 Stockholm, SwedenE-mail: [email protected]

Documents

Juvenile Crime Trends in Post-War Europe