Self-Reported Crime and Deviance Studies in Europe : Current State of Knowledge and Review of Use [1 ed.] 9789054875581

Citation preview

Self-reported crime.book Page 1 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe Current State of Knowledge and Review of Use

Self-reported crime.book Page 2 Thursday, April 23, 2009 8:02 AM

Self-reported crime.book Page 3 Thursday, April 23, 2009 8:02 AM

Renée Zauberman (ed.)

Self-Reported Crime and Deviance Studies in Europe Current State of Knowledge and Review of Use Contributions from Marcelo F. Aebi, Lina Andersson, Cécile Carra, Giada Cartocchi, Thomas Görgen, Janne Kivivuori, Susan McVie, Lieven Pauwels, Stefaan Pleysier, Susann Rabold, Philippe Robert, Giovanni Battista Traverso, Simona Traverso, Renée Zauberman This edited book is the outcome of a seminar held within a Coordination Action Assessing Deviance, Crime and Prevention in Europe, (CRIMPREV, www.crimprev.eu) funded by the European Commission (contract 028300 under FP6) and coordinated by the Groupe européen de recherches sur les normativités (GERN, Centre national de la recherche scientifique, CNRS).

Self-reported crime.book Page 4 Thursday, April 23, 2009 8:02 AM

Criminological Studies In this series studies are published which contribute significantly to the development of the criminological sciences, through either theoretical elaborations, empirical research or criminal policy discussions. Studies originating from other scientific disciplines can also be considered if they are relevant to the criminological domain. The editorial board consists of academics from the criminological research institutes of the Vrije Universiteit Brussel (VUB) and the Universiteit Gent (UG). All manuscripts are reviewed by at least two referees. Editorial Board: Prof. Dr. J. Christiaens (VUB and UG), president Prof. Dr. T. Decorte (UG) Prof. Dr. C. Eliaerts (VUB) Prof. Dr. S. Gutwirth (VUB) Prof. Dr. P. Hebberecht (UG) Prof. Dr. S. Snacken (VUB and UG)

Cover design: Book design: Style, Hulshout Print: Flin Graphic Group, Oostkamp © 2009 VUBPRESS Brussels University Press VUBPRESS is an imprint of ASP nv (Academic and Scientific Publishers nv) Ravensteingalerij 28 B-1000 Brussels Tel. ++ 32 2 289 26 50 Fax ++ 32 2 289 26 59 E-mail [email protected] www.vubpress.be ISBN 978 90 5487 558 1 NUR 824 Legal Deposit D/2009/11.161/017 All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.

Self-reported crime.book Page 5 Thursday, April 23, 2009 8:02 AM

Contents

1. Introduction Renée Zauberman, Philippe Robert

7

2. Self-Reported Delinquency Surveys in Europe Marcelo F. Aebi

11

3. Self-Report Studies in Belgium and the Netherlands Lieven Pauwels, Stefaan Pleysier

51

4. Self-Reported Delinquency Surveys in Finland Janne Kivivuori

77

5. Self-Reported delinquency Surveys in France Cécile Carra

101

6. Self-Reported Delinquency Studies in Germany Thomas Görgen, Susann Rabold

125

7. Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland Susan McVie

155

8. Self-Reported Delinquency in Italy Simona Traverso, Giada Cartocci, Giovanni Battista Traverso

189

9. Self-Reported Delinquency in Sweden Lina Andersson

221

10. On the Authors

249

Self-reported crime.book Page 6 Thursday, April 23, 2009 8:02 AM

Self-reported crime.book Page 7 Thursday, April 23, 2009 8:02 AM

Introduction Renée Zauberman, Philippe Robert

The European Commission has been funding, within FP6, a Coordination Action called Assessing Deviance, Crime and Prevention in Europe. Within this framework, among six Workpackages, one has been specifically devoted to Methodology and Good Practices.

I.

The Methodology and Good Practices Workpackage

The aim of this Workpackage was to deliver a synthetic review of the most significant knowledge tools in the field of crime and of their use This entailed: 앫 the mapping of the European situation, 앫 the recognition of good as well as bad practices, 앫 opening the way to European comparisons. During the last half-century, new and powerful tools emerged for the knowledge of crime. They have in common to free themselves from the institutional data in which the study of crime had traditionally closeted itself. Not only have these instruments renewed, even if only partially, the academic knowledge of crime, but they have a high potential as decisionmaking aids. Yet, they have been initiated under varying circumstances across Europe and their mastering is unequally distributed, the number of confirmed specialists being quite small. Consequently, their use is not always appropriate and users outside the academic sphere have often only limited knowledge of their potential. This makes room within CRIMPREV for a Workpackage devoted to a review of the most significant operations of these instruments in the main European countries, in the hope of establishing and circulating both a state of knowledge and a catalogue of good practices in the field. The workpackage is under the joint responsibility of an academic centre, the French Centre de recherches sociologiques sur le droit et les institutions pénales (CESDIP), a professor at the Universidad central de Barcelona, a European network of local communities sharing urban safety concern and programmes, the European Forum for urban safety (EFUS), and an Italian 7

Self-reported crime.book Page 8 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

regional organisation coordinating local safety programmes (Città Sicure). This makes this workpackage an interface between the academic and public policies spheres. Four methods have been selected: 앫 general population surveys on victimisation and insecurity, 앫 general population surveys on self-reported crime and deviance, 앫 comparisons between survey-based data and others based on institutional sources, like police statistics, 앫 evaluative research on crime prevention and safety policies. For each of these methods, we have proceeded according to a six-phase protocol: 앫 phase 1: elaboration of a report grid; selection of a general rapporteur in charge of synthesising the information collected; and rapporteurs – a half-dozen per topic – in charge of drawing up the state of knowledge and usage in different countries where the method is sufficiently developed. Of course, given problems of availability of the approached experts, we could not cover them all. We were not aiming at exhaustiveness but rather at presenting a reasonably representative selection of what is being done in the European zone, especially in the principal countries; 앫 phase 2: drafting by each rapporteur of a review pertaining to the country or zone of which she/he is in charge and circulation of these drafts among all rapporteurs; 앫 phase 3: meeting of all rapporteurs, national and general, with the promoters of the workshop in a seminar where the reports are presented and discussed; 앫 phase 4: drafting by the general rapporteur of a synthesis of the reports and discussions; 앫 phase 5: validation of this document by the workpackage promoters and dissemination in the form of a 50-page booklet in English and French, published by the FESU; 앫 phase 6: publication of all the contribution in two volumes, one in English, the second in French, both edited by a member of the steering group of the workpackage. A first volume has already been published in this series in 2008, based on the first workshop of the workpackage: Zauberman R., (Ed.), Victimisation and Insecurity in Europe ; a review of Surveys and their use. This second volume is devoted to self-reported delinquency and deviance surveys.

8

Self-reported crime.book Page 9 Thursday, April 23, 2009 8:02 AM

Introduction

II.

The workshop on self-reported delinquency and deviance surveys

These surveys, invented in the 1940s, had developed in some European countries – notably Scandinavian or British – around the 1970s: Their use has later on decreased, as the growing momentum of victimisation surveys attracted a large amount of academic energy. They enjoy nowadays renewed favour and are henceforth a powerful tool for measuring and studying specific forms of deviance or crime, like juvenile delinquency, school violence or substance use (illegal drugs, tobacco or alcohol). The working party gathered around this topic comprised: 앫 Lieven Pauwels (Universiteit Gent) and Stefaan Pleysier (Expertisecentrum Maatchappelijke Veiligheid KATHO University College associated to the Katholieke U. Leuven) for the Netherlands and Belgium, 앫 Janne Kivivuori (Oikeuspoliittinen tutkimuslaitos, Helsinki) for Finland, 앫 Cécile Carra (Institut universitaire de formation des maîtres du Nord-Pasde-Calais and Centre de recherches sociologiques sur le droit et les institutions pénales – CESDIP) for France, 앫 Thomas Görgen (Deutsche Hochschule der Polizei, Münster) and Susan Rabold (Kriminologisches Forschungsinstitut Niedersachsen) for Germany, 앫 Susan McVie (University of Edinburgh) for Great Britain and Ireland, 앫 Simona Traverso, Giada Cartocchi, Giovanni Battista Traverso (Universita degli studi di Siena) for Italy, 앫 Lina Andersson (Stockholms Universitet) for Sweden, 앫 Marcelo Aebi (Université de Lausanne) as general rapporteur, 앫 Renée Zauberman and Philippe Robert (Centre de recherches sociologiques sur le droit et les institutions pénales – CESDIP), Amadeu Recasens i Brunet (Universidad central de Barcelona), Rossella Selmini (Città sicure, Università degli studi di Macerata), Michel Marcus (European Forum for Urban Safety – EFUS) for the steering committee. The national rapporteurs received an order letter from the workpackage’s steering committee in April 2007 and handed in their reports in December that same year. Then a three-day seminar chaired by Renée Zauberman brought together the steering committee, the general and the national rapporteurs in order to present and discuss the reports and draw upon them for feeding transnational lessons into the general report. Following this meeting, the general rapporteur drafted an intermediate report and asked the national correspondents for possible minor amendments to their papers according to the discussions that had taken place during the seminar. The final versions of the national reports were then handed over to him, allowing him to draft a first version of his final

9

Self-reported crime.book Page 10 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

report, which was discussed with the steering committee of the Workpackage during a meeting held in Bologna on 9 July 2008 and sent for final comments to the national correspondents. This volume includes the synthesis by Marcelo F. Aebi along with the sries of national reports. Beside this English publication, this book will also be published in French in the subseries Deviance et Société of L'Harmattan (Paris). Both publications have been edited by Renée Zauberman who also revised all the translations.

References AEBI M.F., 2009, Self-Reported Delinquency Surveys in Europe/Les enquêtes de délinquance autoreportée en Europe, Paris, Forum Européen pour la Sécurité Urbaine. ZAUBERMAN R., 2009, Dir., Les enquêtes de délinquance et de déviance autoreportées. État des savoirs et des usages, Paris, L'Harmattan.

10

Self-reported crime.book Page 11 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe Marcelo F. Aebi1

This report is based on the national reports, the discussions that took place during the Paris seminar and the Bologna meeting, and the comments of the national correspondents, as well as on bibliographical research conducted by the general rapporteur. As a consequence, it includes also references to Self-Reported Delinquency (hereafter SRD) surveys conducted in countries not represented in the workshop. This report is focused on general SRD studies, but surveys conducted in order to measure specific behaviours such as bullying or drug use are also covered. It includes a discussion of definitional issues, a short synthesis of the national reports, a historical overview of the development of SRD studies, a discussion on methodological issues related to that measure of delinquency, as well as an analysis of the impact of SRD surveys on criminological theories and criminal policies.

I.

Definitional issues

Self-reported delinquency surveys are studies in which people – usually juveniles – are asked to reveal information about their delinquent behaviours. However, the terminology may be misleading for two reasons. First of all, because respondents give information not only about delinquency but also about their life-style in general, their attitudes toward different subjects, their families, their school, their friends, and many other socio-demographic factors. Thus, one could consider that delinquency is the dependent variable in such surveys, and that respondents give also a lot of information about independent variables that are supposed to be related to delinquency. The second reason is that the concept of delinquency may also be misleading. Needless to say that delinquency, as any other concept, is a social construct. However, it must be mentioned that the concept of delinquency used in the criminological literature is a very broad one. In fact, many of the behaviours included in a self-reported survey are not criminal offences in most European countries. Such behaviours include, 1.

Université de Lausanne, École de sciences criminelles, Institut de Criminologie et de droit pénal.

11

Self-reported crime.book Page 12 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

for example, running away from home, skipping school or fare-dodging. In this report the term delinquency is used in that broad sense, and therefore it includes all sorts of antisocial or deviant behaviour, even if these behaviours are not defined as an offence in the criminal law. In that context, it is interesting to point out that in countries that use languages derived from Latin (for example Catalan, French, Italian, Spanish, and Portuguese) the term delinquency has the same root as the word used to describe a criminal offence. For example, in Spanish we find “delincuencia” (delinquency) and “delito” (offence), in French “délinquance” and “délit”, and in Italian “delinquenza” and “delitto”. For that reason, in these countries the word delinquency is immediately related to behaviours forbidden by the criminal law. On the contrary, the Webster’s New Universal Unabridged Dictionary, defines delinquency as “wrongful, illegal, or antisocial behavior”. Thus, in English speaking countries, the common understanding of the concept of delinquency includes a wider series of acts ranging from deviant or antisocial behaviour to criminal offences. This different evolution of the term delinquency − which is derived from the Latin delinquere (to do wrong)2 − might be related to the fact that England developed a system of common law−based on custom or court decision instead of written law − and exported it to its former colonies, while Continental European countries followed typically the civil or Roman law system which is based on the written law and led to the introduction of criminal codes and a specific terminology related to them. In this context, the publication of Dei delitti e delle pene −translated into English as On Crimes and Punishments− by Beccaria in 1764 is particularly important. That book inspired many reforms in criminal justice throughout Europe and had a strong influence on the intellectual leaders of the French revolution. Thus, in 1795 −year IV according to the calendar introduced after the revolution−, the (French) National Convention approved its Code of Crimes and Punishments, whose name (Code des délits et des peines) took up the title of Beccaria’s book. Indeed, it was under that same name that the Criminal Code (Code pénal) was first introduced by Napoleon Bonaparte in 1810. That code is considered as the first criminal code and it inspired most of the European codes approved during the following years. In that way, in Continental Europe, the term delinquency was definitely linked to a violation of the criminal law. Nevertheless, the term self-reported delinquency study or self-reported delinquency survey − coined in the United States of America, where the first 2.

According to the Webster’s New Universal Unabridged Dictionary, the Latin delinquere (to do wrong) led to the Late Latin delinquentia (fault, crime).

12

Self-reported crime.book Page 13 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

surveys took place − was translated literally to the different European Latin languages3, creating thus some confusion about the contents of such a survey. This confusion was aggravated by the fact that some of the behaviours included since the first American4 SRD surveys are considered as (juvenile) status offences in most States of the United States of America. These are offences based on the personal condition (the status) of the offender. This means that the same behaviour is not an offence when it is done by an adult, but it is an offence when it is done by a minor. Status offences include behaviours such as truancy, running away, tobacco smoking or underage consumption of alcohol. As they constitute a sort of offence in the USA, American researchers are right when they state that the behaviours include in their surveys constitute violations of the criminal law, but the situation is completely different in Europe, where such behaviours are not considered as offences.

II.

On the development of SRD surveys in Europe

In this chapter, we present a brief summary of the national reports, an overview of the International self-reported delinquency study (ISRD) and the European School Survey Project on Alcohol and Other Drugs (ESPAD), as well as an outline of self-reported delinquency studies in other European countries.

1.

SRD surveys in Finland5

The fieldwork for first Finnish survey on SRD took place in 1959 and the results were published in the early 1960s. That survey was part of the Nordic Draftee Research Program which included other Scandinavian countries in which similar surveys were conducted in the 1960s. The technique was seldom use from the mid-1970s to the 1990s. However, since then, several surveys took place. In particular, Finland was the only Nordic country to participate in the first ISRD in 1992 and has also participated in the second one in 2006. In both cases, the survey is based 3. 4.

5.

For example, encuesta de delincuencia autorrevelada in Spanish, sondage de délinquance autoreportée in French, and indagine sulla delinquenza autorivelata in Italian. In this report we follow the conventional use of the term American to refer to the United States of America, even if we are fully aware that America is a continent and the United States are only one of the countries of that continent. This chapter is based on Kivivuori (2009, with references).

13

Self-reported crime.book Page 14 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

on a city sample (Helsinki), but in the meantime the country has developed a series of national surveys. Thus, in 1995, Finland launched the Finnish Self-Report Delinquency Study (FSRD) that is conducted periodically in schools with samples of 9th grade students. The latest available results refer to 2004 and the FSRD will be conducted again in 2008. Moreover, since 2000/1, the Finnish School Health Survey includes also some questions on SRD. This survey is a largescale one and it includes results at the municipality level. An analysis of the trends from 1995 to 2004 shows a drastic decrease of property offending (especially of shoplifting), a relative stability of violent offences and an increase in soft drug use. Computer related offences were not included in the FSRD but results from the ISRD-2 suggest that there are relatively common. Taking into account that currently juveniles spend a lot of time in front of their computers, it is possible to imagine a displacement from property offences in public places to computer offences. Apart from that, the country has conducted in 2006 a Young Male Crime Survey (YMCS) based on the same questionnaire that was used in 1962 for the Nordic Draftee Research Program. The sample is composed by young males that attend their pre-military screening and therefore are slightly older than the adolescents included in most European samples. A comparison of the results obtained in 1962 and 2006 shows a decrease in workplace theft, buying or selling stolen goods, and theft from a car; an intriguing stability in shoplifting, and an increase in drunken disturbance in public places, drunken driving and bicycle theft. These trends can be explained by changes in the opportunity structure including a late arrival of juveniles to the working market, an improvement of the economic situation of the country, and an increasing in the availability of alcohol. Finland has also participated in the Mare Balticum Youth Victimisation Survey that took place in 2002/3. The questionnaire used for that survey included a SRD scale. Finally, the country also participates in the ESPAD surveys. In sum, one can say that SRD surveys have been institutionalized in Finland and constitute a usual measure of delinquency. As a consequence, they are playing a role in the development of criminal policies. For example, the Ministry of Justice used self-reported data in its estimation of the crime situation and crime trends in the country. In addition, the results of the FSRD were taken into account by the committees that reformed the law concerning young offenders and for the planning of the Finish national violence reduction program.

14

Self-reported crime.book Page 15 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

2.

SRD surveys in Sweden6

As in the case of Finland, the origins of SRD surveys in Sweden are linked to the Nordic Draftee Research Program. The fieldwork for the first survey was conducted at the end of the 1950s and the first results were published in 1960. A few other surveys were conducted during the 1960s and the early 1970s. The technique was somehow abandoned in the mid-1970s but, since the beginning of the 1990s, the country is conducting regular SRD studies. Thus, in 1995, Sweden conducted a national SRD survey and since 1999 that survey is run every second year. The sample is national and varies from 5,000 to 10,000 juveniles attending the 9th grade. An analysis of the trends from 1995 to 2005 shows a decrease of theft and vandalism, and a relative stability of violent offences and drug use. A comparison of the first and the latest surveys shows a decrease in response rates from almost 100% in the 1960s to 80% in 2006. At the same time, it must be mentioned that Swedish researchers have paid a lot of attention to methodological issues, running reliability and validity tests that included measures of the effects of teacher and researcher supervision, anonymity and non-anonymity of the respondents, test-retest controls, observation of the attitudes of pupils during the filling of the questionnaire, tests of different versions of a questionnaire, and follow-up interviews. Sweden has also participated in the second ISRD with city samples. Apart from that, local or regional SRD surveys are conducted regularly with relatively large samples of high school students. Surveys on drug use among youth are also regularly conducted using the self-report technique. All in all, Sweden is a country where SRD surveys are institutionalised and represent currently a typical measure of crime. As a consequence, results from national surveys are used in the political debate on crime and crime prevention. Their influence on public policies can be seen mainly at the local level.

3.

SRD surveys in the United Kingdom and Ireland7

Since the early 1960s, thirty major SRD studies − collecting data on more than 140,000 individuals − have been conducted in the United Kingdom and Ireland. Apart from that, there have been several local and regional studies. Northern Ireland, England and Wales participated in the first 6. 7.

This chapter is based on Andersson (2009, with references). This chapter is based on McVie (2009, with references).

15

Self-reported crime.book Page 16 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

ISRD, and Ireland, Northern Ireland and Scotland participated in the second one. The majority of the self-reported studies took place in Great Britain − more precisely in England − and probably the most well known is the Cambridge Study on Delinquent Development which spans over a 40 year period (1961-2004). Leaving aside that longitudinal study, two other major crosssectional surveys were conducted in the 1960s in England, Wales and Scotland (1963) and London (1967). The technique was seldom used in the 1970s and the 1980s when surveys were conducted only in Sheffield (1975), England and Wales (1983), and Scotland (1989). Then, since the 1990s, there was a sudden explosion of the number of SRD surveys. Thus, new longitudinal studies such as the Peterborough Adolescent and Young Adult Development Study, the Edinburgh Study of Youth Transitions to Crime, and the Belfast Youth Development Study were launched. At the same time, in the 2000s, SRD surveys were introduced in the Republic of Ireland and institutionalized in England and Wales with the introduction of the Offending Crime and Justice Survey that is a repeated cross-sectional survey with a partial longitudinal panel design and a sample of 12,000 individuals. The same is true for Northern Ireland, where the Northern Ireland Crime and Justice Survey has a repeated cross-sectional design with a sample of 3,000 to 3,500 individuals including (in 2005) persons living in private households, offenders on probation, and offenders in custody. In the United Kingdom, the increase in the number of SRD surveys may be due to the fact that the Central Government − that is the main provider of financial support for such studies − changed its attitude towards crime with campaigns such as “tough on crime, tough on the causes of crime”. The idea was to develop an “evidence-based” approach and, in that context, empirical data was clearly needed. All in all, the number and diversity of SRD surveys in the United Kingdom is impressive. As a consequence, the information provided by the available surveys is a clear source of inspiration for public policies. Thus, the Cambridge Study on Delinquent Development had a strong influence on policy makers and inspired partially the reform of the juvenile justice system putting the accent on the early detection of problematic behaviours and ineffective parenting practices. Apart from that, SRD studies conducted in the United Kingdom have often been used for testing the validity and reliability of this measure of crime, as well as for the development and testing of criminological theories.

16

Self-reported crime.book Page 17 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

4.

SRD surveys in Germany8

In Germany, the first SRD surveys were conducted in the late 1960 s and the early 1970s with local or regional samples. Like in most countries, the vast majority of the German research based on SRD surveys, focuses on adolescent populations. However, since 1980, the German General Social Survey (ALLBUS) is conducted every two years and, from 1990 on, that survey includes four items on SRD. As the survey is based on a national sample of the German population, the respondents are mainly adults. Following the Swiss model, the self-report technique has also been used with adults for the evaluation of the involvement in delinquency of the participants in heroin prescription programs. As the first SRD surveys were conducted in different regions and using different methodologies, their results were not easily comparable. Nevertheless, during the 1990s, the Criminological Research Institute of Lower Saxony (KFN) developed a SRD questionnaire that has been used since then in many German cities. Moreover, in some of these cities, the KFN survey is being used on a regular basis. Germany is also one of the few European countries where longitudinal studies based on SRD surveys are available. Such studies are taking place at the local level in different cities, and some of them include comparisons with official data. Most of these studies started in the 2000s but between 1977 and 1996 a longitudinal survey followed a group of 399 children from 13 to 25 years old. Finally, Germany has also participated in the second ISRD. In sum, Germany has a long tradition of SRD studies but the technique is not institutionalized yet. Surveys are organized at the local or regional level and, although the German Ministry of Interior has recently funded a big survey, not all the German States took part on it. Given the federal organization of the country, it is difficult to foresee if national surveys will be conducted in the future. As it happened in Switzerland, the positive results of the heroin prescription programs, measured through SRD surveys, had a strong influence on the German drug policy. Apart from that, SRD surveys did not have a strong impact on criminal policies yet, although they are regularly quoted in the Periodic Security Reports published by the German government.

8.

This chapter is based on Görgen and Rabold (2009, with references).

17

Self-reported crime.book Page 18 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

5.

SRD surveys in the Netherlands9

SRD surveys were introduced in the Netherlands in the late 1960s and quite a few surveys were conducted in the following years. The country has also participated in the first and the second ISRD. Currently, the Scientific Research and Documentation Center (WODC) conducts systematic SRD surveys studies with representative samples of Dutch adolescents. The survey is called the WODC monitor and it takes place every second year. Apart from that, the Netherlands Institute for Social Research (SCP) and the Duch Institute for Budget (NIBUD) are also financing SRD studies. Since the middle of the 1980s, SRD studies have been used to test different criminological theories including differential association, social disorganization theory, strain theory and social bonding theory. The effects of poverty, peers and neighbourhood were also studied. Apart from that, some researchers have tested the reliability and validity of SRD studies and their use with adult samples. As one can see, SRD surveys are institutionalized in the Netherlands and are having some influence not only in the academic field but also in the development of public policies.

6.

SRD surveys in Belgium10

In Belgium, the first SRD survey was conducted in 1976 with a local sample. In the 1980s two other surveys were conducted in both linguistic areas of the country. The country participated with a city sample (Liège) in the first ISRD and with samples from both linguistic regions in the second one. Since the 1990s no systematic large-scale representative self-reports have been conducted, but there is research usually based on urban samples. Thus, two surveys based on large scale samples have been conducted in the Flemish region and in Brussels in the 2000s. However, the recent creation of a platform on adolescent research by three institutions may lead to a more systematic implementation of SRD studies. In sum, SRD studies are not yet institutionalized in Belgium and, apparently, they are not playing a major role in the development of criminal policies. Nevertheless, they have been used to test and develop criminological theories. 9. This chapter is based on Pauwels, Pleysier (2009, with references). 10. This chapter is based on Pauwels, Pleysier (2009, with references).

18

Self-reported crime.book Page 19 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

7.

SRD surveys in Italy11

Leaving aside a small SRD survey (N=198) conducted in Milano and published in 1980, the history of SRD surveys in Italy is strongly related to the ISRD project and the leadership of Uberto Gatti, professor at the University of Genoa, who has coordinated the participation of the country in the first two waves of that survey. Italy participated in the first ISRD with a sample of three cities (Genoa, Messina, and Siena), and the questionnaire was used again in Siena in the mid 1990s. The country has also participated in the second ISRD with a sample of 15 cities (N=7,278). In 2006, a selfreported survey was also used to measure the involvement in delinquency of foreign youth living in Italy. Research on bullying has been conducted in different cities and regions since the middle of the 1990s using self-reported surveys. In this context, sometimes the ISRD questionnaire was combined with a specific questionnaire on bullying. The self-report technique has also been used to measure drug and alcohol use among juveniles since the 1980s. In sum, SRD surveys are not institutionalized in Italy and their findings are mainly used by the scientific community. As a consequence, they do not play a role in the development of national public policies; nevertheless, there is evidence that they have been used at the local level in the city of Sienna.

8.

SRD surveys in France12

Among the countries included in this overview, France was the latest to introduce SRD studies. The first study of this kind was conducted in 1999 − and published in 2001− with a sample of two cities. The questionnaire was based on the one used for the first ISRD. In 2003, a second survey was conducted using the same methodology in one of the two cities surveyed in 1999. Finally, in 2006, the country participated in the second ISRD. Violence at school has also been measured using the self-report technique since the mid-1990s. In that case, the survey was based on a questionnaire focused on victimisation, but it included also some questions on selfreported violence. The same questionnaire was later used in other countries, providing thus the possibility of performing some cross-national comparisons. A recent research is based on open questions about violence at school which are later recoded by the researcher. 11. This chapter is based on Traverso, Cartocci, Traverso (2009, with references). 12. This chapter is based on Carra (2009, with references).

19

Self-reported crime.book Page 20 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

One can easily see that SRD studies are not at all institutionalized in France. The findings from the very few surveys available are used by the scientific community, but they do not play a role in the development of public policies. The only exceptions are surveys on drug abuse which were quoted in the new law on crime prevention introduced in 2007.

9.

The International Self-Reported Delinquency Study (ISRD)

The ISRD-1 The origins of the ISRD project can be traced back to a meeting of experts that took place in 1988 in the Netherlands and led to the publication of a book on issues related to cross-national research based on the SRD technique (Klein, 1989). Some of the experts that assisted to that meeting decided to launch the project, which started formally in 1990 and was coordinated by the Research and Documentation Centre of the Dutch Ministry of Justice (WODC), headed at that time by Josine Junger-Tas. The first survey (ISRD-1) had three main objectives: to compare prevalence and incidence of delinquent behaviours across countries; to contribute to the explanation of (differences in) delinquent behaviour; and to contribute to the solution of methodological problems related to cross-national research (Junger-Tas, 1994a, 2). It was based on a common questionnaire developed by a steering committee of researchers in criminology that was translated into each national language13. The theoretical framework of the questionnaire was heavily inspired by social bonding theory (Hirschi, 1969) while the questions on delinquency had similarities with the ones used in the National Youth Survey and the Denver Youth Study in the United States (Elliott, Huizinga, 1989)14. Twelve countries participated in the study15. Four of them (England and Wales, the Netherlands, Portugal, and Switzerland) used national random samples; Spain participated with a stratified urban sample; Germany, 13. The original English questionnaire can be found in Junger-Tas, Terlouw, Klein (1994, 387-441) 14. According to Moffitt et al. (1994, 358), 80% of the delinquency items included in both instruments (ISRD-1 and National Youth Survey) overlapped. 15. Our presentation of the ISRD-1 is based on the main publication about it (Junger-Tas, Terlouw, Klein, 1994). The authors of that publication mention thirteen countries because they include New Zealand. However, data from New Zealand refer to people aged 18 and comes from a longitudinal study that did not use the ISRD questionnaire, but the SRD instrument developed by Elliott and Huizinga (1989) for the National Youth Survey and the Denver Youth Study in the United States (Moffitt et al., 1994, 357), that we have mentioned in the previous footnote.

20

Self-reported crime.book Page 21 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

Greece and Northern Ireland participated with city samples (Mannheim, Athens, and Belfast, respectively); Belgium (Liège), Finland (Helsinki), Italy (Genoa, Messina, and Siena) and the United States (Omaha) participated with school city samples. With the exception of the countries that used school samples, participants in the survey were aged 14 to 2116. Data collection took place in 1991 and 1992 and was based on face-to-face interviews with the exception of the countries that used school samples, where the questionnaire was self-administered17. The ISRD-1 database is not freely available for researchers yet. A short synopsis of the preliminary results was presented by Junger-Tas (1994b, 379), but it took more than ten years to have a detailed presentation of the main findings in a publication by Junger-Tas, Haen Marshal, Ribeaud (2003) that are summarized here : 앫 Lifetime prevalence rates are surprisingly similar across countries. However, looking at the nature of the offences, there are clear crossnational disparities. 앫 Property offences are highest in the Northwest-European countries and the USA (Omaha) supporting the idea − based on opportunity theory − that these rates would be highest in the most prosperous countries. 앫 There were high rates of violent offences in the USA, England and Wales, Spain and Finland (Helsinki). 앫 High rates of mainly soft drug use were found in England and Wales, the USA, Northern Ireland (Belfast) and Switzerland. The Netherlands, that apply a tolerant policy on possession and use of soft drug use, occupied a medium position on the scale of drug use rates. 앫 Delinquent behaviour peak between age 14 and 18 across countries. The peak age was 15 in the Anglo-Saxon cluster and 16 in the other clusters. 앫 The peak age vary according to the offence. For example, the peak age of vandalism was clearly lower (between age 14 and 15), while that of drug use was considerably higher. 앫 Age of onset vary between clusters. The lowest age of onset for vandalism was found in Southern Europe, while the lowest age of onset for property offences was found in Northwest Europe. Violent offences and drug use had lowest age of onset in Anglo-Saxon countries. 앫 Serious delinquents tend to start offending at an earlier age than their less seriously-involved counterparts. 16. Belgium applied a mixed approach combining a school sample with a random sample of persons aged 18 to 21. 17. However, Italy used a school sample, but the questionnaire was administered through face-to-face interviews.

21

Self-reported crime.book Page 22 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

앫 Both prevalence and frequency of offending are lower for females than for males. However, gender disparities vary according to offense type and regional country-cluster. Gender disparity is smallest for property offending and drug use while it is largest for violent and serious offences. In addition, the disparity is smaller in Northwest Europe compared to other clusters, which may be related to the high rates of property offences and vandalism found in that cluster. 앫 Family break-up has important effects only as far as it results in father absence. However, father absence is closely related with delinquent behaviour in the Anglo-Saxon cluster and in South Europe, but not in Northwest Europe. 앫 Relationship with parents is not related to overall delinquency, but is related to serious delinquency and to drug use in all clusters. 앫 In all countries there is tighter control on females than on males. Compared to males, the relationship of females with their parents deteriorates somewhat more with age. On the other hand, girls participate considerably more in family outing than do boys. 앫 Truancy and disliking school are related to all types of delinquency in all country clusters. Having to repeat a grade, while not being related to petty delinquency, does appear to be related to violence, serious delinquency and drug use for males, but not for females. 앫 Peer group membership is related to age, enrolment in school, disliking school, parents’ informal control and not participating in family outings. Delinquents typically spend most of their leisure time with the peer group, while non-delinquents spend much more time with the family (Junger-Tas, Haen Marshal, Ribeaud, 2003, 139-142). The ISRD-2 A first meeting of experts interested in participating in the second wave of the ISRD took place during the 2003 Annual Conference of the European Society of Criminology, in Helsinki. A steering committee was established and, during the following months, the research design and the revised questionnaire were established. The questionnaire was then translated into each national language. Data collection took place in 2006 in thirty countries18 − of which twentyfour were European − and a first publication including national chapters is 18. Armenia, Aruba, Austria, Belgium, Bosnia-Herzegovina, Canada, Cyprus, Czech Republic, Denmark, Estonia, Germany, Hungary, Iceland, Ireland, Italy, Lithuania, the Netherlands, the Netherlands Antilles, Northern Ireland, Norway, Poland, Portugal, Russia, Scotland, Slovenia, Surinam, Sweden, Switzerland, the United States of America, and Venezuela.

22

Self-reported crime.book Page 23 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

attended for 2009. In the meantime, some countries have already published their national findings. The ISRD-2 survey is based on school samples of at least 2,000 students − attending 7th to 9th grade − per country. The main sampling design is city based with a minimum of five cities including a metropolitan area, a medium sized city and three small rural towns. Such a design allows multilevel HLM analyses (Raudenbusch, Bryk, 2002) using students as first level units and cities as second level units. Nevertheless, some countries used national random samples drawn from lists of all school classes in the country. These countries oversampled some urban or rural areas in order to make comparisons possible with the rest of the countries. The majority of countries used self-administered paper and pencil questionnaires filled by the students in their classrooms, however some countries (e.g. Switzerland) used computer assisted personal interview (CAPI) or computer assisted self-administered interview (CASI)19. The ISRD-2 questionnaire can be seen as a revised version of the one used for the ISRD-1. This is completely logical, as one of the goals of the project is to achieve comparability between both surveys. However, there are some major modifications. As far as offences are concerned, the questionnaire includes new ones (e.g. computer related offences) and the wording of some questions has been modified. The instrument includes also a short victimisation survey. The theoretical framework keeps variables related to social bonding theory, but the new questionnaire includes a summary of the Grasmick self-control scale (Grasmick & al., 1993) that allows testing the general theory of delinquency − or self-control theory − developed by Gottfredson and Hirschi (1990), as well as questions related to social learning theory and lifestyle variables that allow testing opportunity-based theories. It also asks for information about life events as well as school and neighbourhood context. The survey included also a short questionnaire for the interviewers about the context in which the administration of the survey took place, as well as a questionnaire for the teachers of the classes being surveyed. Most countries used Epidata software to input data from the questionnaires. In that context, the labelling of the variables was standardized, simplifying thus the construction of an international database that will be a major tool for European research in the future. As a final consideration about the ISRD project, it must be mentioned that even if the survey has not been institutionalized at the European level − both the ISRD-1 and the ISRD-2 were financed in each country by local 19. The different techniques for administering SRD surveys will be discussed later, in the chapter on Methods of administration.

23

Self-reported crime.book Page 24 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

sources or eventually with the financial help of another European country20 − the creation of a steering committee and a big consortium of researchers is a guarantee that the ISRD project will continue in the future.

10. The European School Survey Project on Alcohol and Other Drugs (ESPAD) The success of self-reported studies on substance abuse (tobacco, drugs and alcohol) is surely due to the fact that even the most survey sceptic person must admit that official statistics do not provide any valid information on the extent of such use. In that perspective, in the mid 1980s, the Council of Europe created a group of experts that started working on a common questionnaire on substance abuse. Building on the experience of that group, in 1993, the Swedish Council for Information on Alcohol and Other Drugs (CAN) initiated the European School Survey Project on Alcohol and Other Drugs (ESPAD). The first survey took place in 1995 and data have been collected every four years since then. The latest available results refer to the 2003 survey. Results from the 2007-8 survey will only be available by the end of 2008. The ESPAD project has received financial support from the Pompidou Group at the Council of Europe, the Swedish Ministry of Health and Social Affairs and the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). However, data collection in individual countries is funded by national sources. The number of countries taking part in the ESPAD project has increased over the years. There were 23 countries in 1995, 30 in 1999, 35 in 2003 and 43 in 2007-8. Indeed, all the countries represented at the Paris workshop on SRD surveys are participating in it. The survey is based on a common questionnaire that covers tobacco, alcohol, and illicit drugs use. Data are collected through groupadministered questionnaires in schools. Samples are composed by 15-16 year old students, although some countries include also students aged 1718. With only a few exceptions, samples are nationally representative. For researchers, one of the big advantages of the ESPAD project is that the main results are available on a free website (www.espad.org) where it is possible to download reports and generate key results graphs. Summarizing the main results of the 2003 SPADE survey, Hibell & al. (2004, 22) state:

20. For example, Switzerland funded partially the participation in the survey of Armenia and Bosnia-Herzegovina.

24

Self-reported crime.book Page 25 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

앫 The pattern of alcohol consumption reveals that frequent drinking is most prevalent among students in the western parts of Europe, such as the British Isles, the Netherlands, Belgium but also in Austria, the Czech Republic and Malta. Very few students in the northern parts of Europe drink that often. 앫 Beer consumption is most prevalent in Bulgaria, Denmark, the Netherlands and Poland, while wine consumption is most prevalent in typical wine producing countries such as Austria, the Czech Republic, Greece, Italy, Malta and Slovenia. 앫 The consumption of spirits is less uniform, with high prevalence rates in as disparate countries as the Faroe Islands, Greece, Ireland, Isle of Man, Malta and the United Kingdom. 앫 The prevalence of drunkenness seem to be most concentrated to countries in the western parts of Europe, such as Denmark, Ireland, Isle of Man and the United Kingdom. Very few students report frequent drunkenness in Mediterranean countries such as Cyprus, France, Greece, Portugal, Romania and Turkey. 앫 The illicit drug use is dominated by use of marijuana or hashish. Frequent use is mainly reported from countries in the central and western parts of Europe, where more than one third of the students have used it. The high prevalence countries include the Czech Republic, France, Ireland, Isle of Man, Switzerland and the United Kingdom. The low prevalence countries are found in the north as well as the south of Europe (Hibell & al., 2004, 22). As far as the trends between 1995 and 2003 are concerned, Hibell & al. (2004, 128) conclude: 앫 To sum up, the trend development over the 8 years of the ESPAD history is indicative of the fact that smoking remains at about the same level or decreased in a majority of the countries. 앫 With regard to alcohol an unchanged or a somewhat decreasing consumption was observed in the western parts of Europe while increases mainly were found in the eastern parts. 앫 The use of drugs is still dominated by the use of cannabis. The high prevalence countries in 1999 are still at the top in 2003, but a clear increasing tendency can be observed in the eastern parts of Europe. It is also clear that an increasing number of students in many European countries find cannabis easily available (Hibell & al. 2004, 128).

25

Self-reported crime.book Page 26 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

III. Historical overview of SRD surveys After having presented an overview of SRD surveys in different European countries and some European projects based on the self-report technique, this chapter will provide a general overview on the historical development of that measure of crime. In the criminological field, the first use of the self-report technique to measure delinquency can be traced back to the works of Porterfield in the United States in the 1940s. The technique was improved during the 1950s and a major step was taken when Nye and Short introduced the first Guttman delinquency scale constructed on the basis of an SRD survey, in 1957. As we have explained elsewhere (Aebi, 2006), in the context of The American Soldier in World War II project, Guttman (1950) developed a scalogram that allowed combining in a table with N columns and N lines the answers of each person to selected items of a questionnaire. Simplifying things a little bit, these items constitute what is commonly called a Guttman scale. Indeed the selection of the items is the key element in the construction of the scale because they should measure a single dimension and must be classified in a hierarchical order according to the principle that a positive answer to an item implies agreement with items appearing lower in the scale. For example, from a logical point of view, the person who gives an affirmative answer to the question “do you smoke more than 15 cigarettes per day?” should also answer yes to the questions “do you smoke more than 10 cigarettes per day” and “do you smoke more than 5 cigarettes per day”. Thus, knowing the number of positive answers, the researcher can identify the questions that have been answered in an affirmative way. In the field of criminology, a Guttman scale classifies offences from the less serious to the most serious one. However, as the samples are usually composed by juveniles attending high school, it is very difficult to establish such a scale without including minor deviant behaviours at the bottom of the scale. Accordingly, Nye and Short (1957) selected the following seven items – presented here from the less serious to the most serious − from the twenty-one behaviours included in their questionnaire: driving without a license, taking little things (worth less than $2), buying or drinking beer, wine, or liquor, skipping school without a legitimate excuse, having sex relations with a person of the opposite sex, purposely damaging or destroying public or private property, and defying parents’ authority to their face. The researchers were obliged to drop items such as running away from home and taking things of medium value (between $2 and $50) because they were committed by less than 10 percent of the sample. Their scalogram took also into account the frequency of the behaviour according to four categories (never, once or twice, several times, and very often). 26

Self-reported crime.book Page 27 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

Actually, one can see here a nice example of the interrelationship between science and technology. In the 1950s, the analysis of the data took a lot of time and it was almost impossible to conduct sophisticated statistical analysis. The Guttman scale was a technical innovation that allowed a relatively cheap and quick way of improving the quality of the analysis. Indeed, the article by Nye and Short (1957) can be seen as a statement in support of the use of such a scale in criminological research in order to break away from the simple dichotomies between delinquents and nondelinquents. Most researchers followed their advice and, in the same way that since the mid-1990s it is very difficult to find an empirical article that does not include a logistic regression − and in the 2000s it seems that a multilevel HLM analysis is a must for every research − in the 1960s most articles included analysis based on Guttman scales. The problem comes from the fact that the scores in Guttman scales were used to classify respondents in different groups but, many times, one has the impression that the ones classified as heavily involved in delinquency are not really offenders but merely deviant youth. In any case, only a few years later, SRD surveys were introduced in Europe. In Scandinavian countries, the first SRD surveys were part of the Nordic Draftee Research (NDR) Program and the fieldwork took place in Sweden already in 1959 and in Finland in 1962. In the United Kingdom, the technique was adopted in 1961 by the Cambridge Study in Delinquent Development (1961-2004), which was the first European longitudinal study of this kind, and it was also used in cross-sectional studies since 1963. In Germany and in the Netherlands, the first SRD surveys were conducted in the late 1960s, while in Belgium the first survey took place in 1976. In Italy, a small survey was conducted in 1980 but it was only with the participation of the country to the first ISRD in 1992 that the technique was used with larger samples. Finally, in France, the first ISRD was conducted in the late 1990s21. It is possible to analyze the evolution of SRD surveys in Europe using the categories proposed by Kivivuori (2009) for Scandinavian countries. One can thus distinguish three periods: the first one runs from the 1960s to the mid-1970s and corresponds to the moment when the first surveys were conducted in many countries. A second period runs from the mid-1970s to the end of the 1980s and is characterized by the fact that the technique was 21. It is difficult to write the history of SRD surveys in Europe without mentioning the name of Josine Junger-Tas, who conducted the first survey in Belgium in 1976, promoted such surveys very actively in the Netherlands while she was Head of the WODC during the 1980s and was one of the main promoters of both ISRD surveys in 1992 and 2006.

27

Self-reported crime.book Page 28 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

seldom used. The third and current period corresponds to a kind of golden age of such surveys that includes the development of the ISRD project at the European level in 1992 − that allowed countries such as Italy, Portugal, Spain and Switzerland to conduct their first large scale or national surveys − a clear increase in the financing of such projects in the United Kingdom, and their institutionalization in countries such as Finland, Sweden, the Netherlands, England and Wales and Northern Ireland. In order to understand this evolution, one should take into account once more the interrelationship between science and technology. This link is clearer in the natural sciences. For example, the Galilean revolution would have been impossible without the help of the first telescopes (Morin, 1990), and Pasteur would have been unable − as Semmelweiss, a few years before him (Hempel, 1966) − to develop his theories without the microscope. More recently, the development of very powerful telescopes allowed researchers to discover galaxies and planets that could not be seen before and led to major changes in astronomy theories. At the same time, in order to refine their theories, researchers ask for more powerful instruments and help developing them. This leads to a “loop phenomenon” between science and technology (Morin 1990, 60) that can be appreciated when one analyzes the evolution of crime measures and crime theories (Aebi, 1999, 2006; Aebi, Jaquier, 2008). For example, the development of victimization surveys in the 1970s allowed the development of life-style theory (Hindelang, Gottfredson, Garofalo, 1978); but, before that, the development of SRD surveys had also play a major role on the evolution of criminological theory. As we have mentioned before, the first SRD surveys included a lot of problematic behaviours and trivial offences. These items were admitted by the great majority of the respondents, and such a result challenged previous criminological theories. The typical criminal was no longer a young urban male belonging to an ethnic minority. Indeed criminality seemed normal22. Moreover, if everyone was engaged in delinquency, the overrepresentation of some categories of the population found in official statistics could only be due to a differential reaction of the criminal justice system. This interpretation was also supported by a very

22. The hypothesis of the normality of delinquency had already been posed by Durkheim (1988/1895, 160) in the 19th Century. However, Durkheim considered crime as normal in the sense that it is hard to imagine a society without deviants; but, as research has shown, crime is not normal in the sense that anyone commits serious offences.

28

Self-reported crime.book Page 29 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

influent article by Kitsuse and Cicourel (1963) on the limits and biases of official statistics. Such findings played a major role on the development of labelling theory in the 1960s and Marxist criminology in the early 1970s. However, the self-report technique was being constantly improved and very soon it became clear that serious delinquency was not normal. As Kivivuori (2009) states: However, the NDR data harboured findings that could potentially contradict the “crime as normality” interpretation. For example, it was found that police detection likelihood reflected offending intensity (Christie, Andenaes, Skirbekk, 1965). Nils Christie, one of the pioneers of Nordic self-report research, explicitly warned that the crime-as-normality rhetoric, appealing as it was, should not be pressed into absurdity (Christie 1966 [1964], 59). However, he himself kept on using the normality argument (Christie 1975, 73). The stubborn existence of the chronic offender thus haunted the early NDR researchers. At the same time the spirit of the times took an anti-positivist turn. The early NDR design had reflected the influence of empirically oriented American social science. This emphasis lost much of its appeal towards the end of the 1960s as researchers wanted to engage politically and started to criticise quantitative methods (Kivivuori, 2009). The Scandinavian situation described by Kivivuori was probably not very different than the one that could be found in the rest of Western Europe. At the beginning of the 1970s, a vast majority of the new generation of European social researchers was fascinated by the emergency of the paradigm of the social reaction. Thus, they concentrated their efforts on the study of the social reaction to crime and the process of labelling. As a consequence, even if one can find some noteworthy exceptions, research based on SRD surveys was fairly uncommon. Curiously enough, in 1977, an analysis made by Farrington using the SRD data from the Cambridge study was giving some support to labelling theory, even if it was also showing that police activity was not as arbitrary as some researchers thought because juveniles contacted by the police where the ones that were more engaged in delinquency according to self-report measures (Farrington, 1977). The latter result had also been found by Christie, Andenaes, Skirbekk (1965) some years before with a Nordic sample, but it has seldom been quoted by Christie himself. The situation changed drastically at the beginning of the 1990s. Since then, the number of SRD surveys conducted in Europe has increased at an incredible pace. Currently, they are being conducted in all Western European countries and also, through the ISRD-2 project, in some Central

29

Self-reported crime.book Page 30 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

and Eastern European ones. Several explanations can be given to this evolution, but we think that the following three are the main ones : 앫 Improvements in computer technology: Once more we insist here on the link between science and technology. Since the 1990s it has been possible to conduct all the statistical analysis of a big SRD database using a single personal computer. This implies a reduction of the costs of a research and also of the time needed to produce a final report on it. 앫 A change in attitudes towards crime and security in developed nations23: There is currently no doubt that most developed countries became more punitive since the 1990s. Moreover, in many of these countries there are unending debates − leading sometimes to modifications of the Criminal Code − on juvenile delinquency and the ways in which juveniles should be treated by the criminal justice system. The causes of this change are probably twofold: on the one hand, there has been a real increase in delinquency during the second half of the 20th Century in most developed nations (Aebi, 2004, with references); on the other hand, politicians reintroduced the debate on crime and security in their agendas, amplifying thus the phenomenon, and asking for (and funding research on) data on crime and delinquency. The idea was to introduce policy reforms based on an evidence-based approach, although it seems that in many cases it was guided by merely electoral goals. 앫 A change of the main paradigm in criminological research: Thomas Kuhn (1970, 15) had doubts about the existence of paradigms in the social sciences. As a consequence, his concept of scientific revolutions is probably impossible to apply to them. In the natural sciences, a new paradigm replaces the old one completely (e.g. no serious scientist would defend today that the conception of Newton is superior to the one of Einstein). In the social sciences, on the contrary, there is no replacement but addition of paradigms. By the end of the 1960s, the paradigm of social reaction was introduced in the criminological scene, but it never replaced the former paradigms and nowadays is only one paradigm among others. Nevertheless, during the 1970s and early 1980s most European criminologist adhered to the social reaction paradigm and where quite reluctant to the types of empirical research that could be labelled as positivist24. The situation changed in the 1990s 23. An interesting analysis of some aspects of this change in England and the United States has been done by Garland (2001). 24. The depletion in SRD surveys during that period (late 70s et early 80s) may be also have been caused by the concurrent development of the first important European programs of victimisation surveys, in Britain, the Netherlands, Sweden...: the energy invested in this innovations may have been diverted from SRD surveys.

30

Self-reported crime.book Page 31 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

when a new generation of criminologists, trained in the use of personal computers, entered the scene and started to put the accent on alternative explanations of delinquency developed by researches that did not belong to the mainstream in the 1980s and that could be tested through SRD surveys. Thus, in the 1990s research was relatively cheaper than before, in some countries politicians were willing to finance it, and there was a new generation of criminologists ready to conduct it. As a consequence, SRD surveys became an established measure of juvenile delinquency in Europe and therefore it is necessary to discuss their methodology.

IV. Methodological issues Several authors have identified the main methodological problems related to the use of SRD surveys. At the same time, many of the national reports presented before include considerations on that topic. This chapter offers a short summary of such problems based on the national chapters as well as on the reviews of Hindelang, Hirschi, Weiss (1981), Aebi (1999, 2006), Junger-Tas and Haen Marshall (2003), Thornberry, Krohn (2000), and Aebi, Jaquier (2008).

1.

The content of SRD surveys

A typical self-reported delinquency survey usually includes questions on problematic behaviours, property offences, violent offences, substance abuse, fraud and dishonesty. Problematic behaviours These are behaviours that are wrongful or antisocial but that are usually not included in criminal law. Typical examples are truancy, running away from home, tobacco smoking, underage consumption of alcohol, and defying parental authority. The first American surveys included also having sex with a person of the opposite sex as an item that was included in self-reported delinquency scales (see Nye, Short, 1957). As mentioned before, when self-reported delinquency surveys were introduced in the United States of America, most States considered these behaviours as status offences. As a consequence, such behaviours were included in self-reported delinquency surveys and, later, in the delinquency scales constructed on the basis of such surveys. We have

31

Self-reported crime.book Page 32 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

already seen that the delinquency scale proposed by Nye and Short (1957) included mainly problematic behaviour and trivial offences. Later, the delinquency scale developed by Erickson (1972) encompassed eight behaviours including the following four: smoking, buying tobacco, drinking beer, wine or liquors, and buying beer, wine or liquors. Of course, these behaviours are bad for the health of an adolescent, but from a European point of view, they cannot be considered as offences. However, as European self-reported delinquency surveys were inspired by American research, some of these behaviours were − and are still being included − in many European questionnaires. As a consequence, researchers should pay special attention when analyzing the results in order to avoid mixing problematic behaviours with serious delinquent acts. Property offences SRD surveys regularly include questions on theft and many times they introduce a distinction between minor theft (less than a certain amount of money) and serious theft. Burglary is also generally included and there are questions about theft of a car − as well as about theft of motorcycles or bicycles − and theft from a car too. Other property offences frequently included in the survey are graffiti, vandalism and arson. Curiously enough, questions on intellectual property theft are rare. Only in recent years, there has been an interest in this behaviour but it is usually measured in an indirect way through the use of computers to download illegal files. It is a pity that such questions were not included in the first SRD surveys, because one can imagine that the rates of intellectual property theft in the 1970s and 1980s (through the recording of cassettes) and in the 1990s (through the burning of compact discs) were not very different to the ones found in the 2000s (through the downloading of MP3 files). Also, theft of intellectual property through Xerox copies or – more recently − scanning of books and other printed material are regularly excluded from the survey. Violent offences Violent offences typically include robbery − which in many European continental countries is defined as theft with violence and therefore considered a property offence − assault, bullying and use of weapons. The questions on assault have been improved with time in order to avoid including minor incidents such as slapping. Bullying has been included in recent years and therefore it is impossible to compare the current situation with the one that existed in the 1960s and 1970s.

32

Self-reported crime.book Page 33 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

Substance abuse Questions on tobacco, alcohol and drug use were already included in the first SRD surveys. Currently, there is an international specialized survey – the ESPAD project presented previously − on that topic. In European countries that are wine producers, consumption of wine by adolescents during the meals was fairly common until some years ago and it was not seen as problematic. The same was true for beer. Our personal impression, based on research conducted in Southern Spain and informal interviews in Switzerland, is that there has been a displacement from consumption at home − where supervision by adults was assured − to consumption while going out with peers. It also seems that there has been a displacement from consumption of wine to consumption of liquors such as whisky, gin or vodka. Unfortunately, as many European SRD questionnaires were inspired by the American questionnaires that − according to the American laws − considered the use of alcohol by adolescents as an offence, SRD surveys have seldom measured the context in which consumption was taking place and the nuances between the different types of alcohol being used. Fraud and dishonesty SRD surveys frequently include questions on fare-dodging as well as on receiving and handling stolen goods. In countries where checks are a typical way of paying, there are usually questions on the misuse of them. Also, in some cases, one can find questions related to the abuse of credit cards and, more recently, computer related offences has been added in most questionnaires although sometimes their wording is doubtful. For example, in the ISRD-2 questionnaire, it is not clear whether the download that is being measured is legal or illegal. The issue of trivial offences As a rule, some of the behaviours included in SRD surveys are criminal or administrative offences but, as they are not serious, they would probably never lead to a criminal procedure. A typical example of such offence is fare-dodging. Both trivial offences and problematic behaviours − discussed above − can be seen, under certain circumstances, as good predictors of future delinquency. Therefore we are not advocating for the removal of all of them from self-reported delinquency scales. However, as suggested by JungerTas (1989), it is very important to avoid mixing them with serious offences when constructing delinquency scales. The inclusion of such items in

33

Self-reported crime.book Page 34 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

general delinquency scales – especially when lifetime prevalence is studied – usually leads to results that show little differences in the distribution of delinquency by sex, socioeconomic status and other socio-demographic factors. As mentioned before, such indexes led in the 1950s and 1960s to the erroneous idea that delinquency was distributed homogeneously across the population.

2.

Delinquency scales

The first SRD surveys were mainly interested in the lifetime prevalence of different behaviours (Have you ever…?). For example, in the first Scandinavian SRD surveys, measures of incidence/frequency of delinquency were non-existent or rudimentary (Kivivuori, 2009). However, as indicated by McVie (2009), nowadays most SRD surveys include a general prevalence questions (i.e. have you done…?), for the period ‘ever’ and ‘last year’, followed by an incidence/frequency question (i.e. how many times have you done it?). Both ISRD surveys and most current SRD survey include also follow-up questions about the characteristics of the latest delinquent behaviour. However, in many cases these questions − and many others included in the questionnaire − have not been fully exploited in the available publications on SRD surveys. Prevalence and incidence questions are used to construct delinquency scales that include different behaviours. Variety (i.e. number of different offences committed) is also an interesting index that can easily be calculated on the basis of the questions on lifetime and last year prevalence. It is particularly interesting for testing versatility and specialization in delinquency. For example, Aebi (1999, 2006) found that most heavy heroin-addicted adult offenders had committed a large range of offences during their lifespan, but that they tend to specialise themselves in one or two types of offences − small scale drug-trafficking and shoplifting − when shorter periods of time were studied. Their situation can be compared to the one of the smoker that prefers a certain brand but has tasted most of the available ones. Moreover, variety of offending was the best predictor of the risk of being arrested by the police − the greater the range of offence, the greater the risk of being arrested − while incidence was not linearly correlated to that risk. However, frequency is seldom used in publications on SRD surveys. An exception can be found in Junger-Tas, Marshall, Ribaux (2003) that used it as a good index for international comparisons.

34

Self-reported crime.book Page 35 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

3.

Sampling

In recent years, many cross-sectional SRD surveys are being conducted at school with students attending the ninth grade. Sometimes − for example in the ISRD-2 − the sample is enlarged in order to cover 7th, 8th and 9th grade students. Only exceptionally are post-adolescents included in the sample; for example, the ISRD-1 included juveniles 14 to 21 years old. Finally, one can find a few exceptions of surveys conducted with adults using samples of drug users or prisoners or in the context of longitudinal studies. One of the typical critics to school samples is that juveniles that most engaged in deviant behaviours are probably not at school. Also, sometimes special education classes are not included in the sample. As a consequence, SRD surveys tend to underestimate the most serious types of delinquency. Moreover, taking into account that the influence of some factors − such as the family or the school − is likely to diminish with time, limiting samples to the 9th grade may lead to an overestimation of the influence of such factors in future delinquency. For example, a study based on the Swiss ISRD-1 sample showed that family structure had an influence on drug consumption for adolescents aged 14-17 − males from broken homes were more involved in soft drugs use than the ones coming from traditional families − but this influence disappeared for the ones aged 18-21 (Aebi, 1997).

4.

Response rate

Response rates vary widely across time and space, ranging typically from 70 to 90%. In Sweden, a decreased trend was observed in recent years (Andersson, 2009), but in other countries the evolution is not linear. One advantage in some European countries is that parents do not have to authorize explicitly their children to participate in the survey. Typically, parents are informed that, without opposition from their part, their children are going to participate in a survey. The experience of the ISRD shows that, under such circumstances, few parents refuse the participation of their children in the survey.

5.

Method of administration

The first SRD surveys were based on personal interviews. Paper-pencil questionnaires were introduced relatively soon. Sometimes – for example in the ISRD-1 − both techniques are combined, with juveniles answering in 35

Self-reported crime.book Page 36 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

a written form to the screen questions on delinquency (using the sealedenvelope technique). Since the 1990s, CAPI (computer assisted personal interview) and CASI (computer assisted self-administered interview) techniques are being used. In the first case, the interviewer administers the questionnaire, while in the second one the respondent reads the questions on the screen and enters directly the answers on the computer. Switzerland used CAWI (computer assisted web interview) for the ISRD-2. The only difference between CASI and CAWI is that, in the first one, the answers are stocked on the computer used for the interview, while on the second one they are stocked directly on the main database which is accessible through the Web (in a secured webpage only accessible through a password given to the interviewers). Finally, it is also possible to use ACASI (audio computer assisted audio self-interviewing), a technique in which the answers are recorded and the respondent can listen to them through the computer. As far as the validity of the answers is concerned, comparisons of these techniques show only small differences from one to the other. However, the use of ACASI may solve at least partially the problems of illiteracy of some of the respondents even if, when the research is conducted at schools, this issue does not seem a serious threat to validity. At the same time, the use of CAPI, CASI and CAWI improves reliability by reducing the risk of introducing wrongful information in the database while coding the paperpencil questionnaires. It also allows the use of more complex questionnaires − especially as follow-up questions are concerned − because the computer will lead the respondent directly to the next relevant question. This is a major problem in paper-pencil questionnaires, where the respondent has to look by himself/herself for the next question according to his/her answers (e.g. if yes, please go to question No 55). Finally, comparisons of the presence of teachers or researchers during the administration of the questionnaire as well as anonymous versus nonanonymous versions of a questionnaire show that lower prevalence delinquency rates are found when the teacher is present and when anonymity is not assured.

6.

Validity of SRD surveys

Every instrument designed to measure a phenomenon can be considered as a measure of that phenomenon. The validity of a measure can be defined as its capacity to measure efficiently the phenomenon under investigation. In particular, the validity of a crime measure can be defined as its capacity to measure efficiently the crime phenomenon.

36

Self-reported crime.book Page 37 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

There are several classifications of the different types of validity. In the following chapters we will present these types according to the classification that we proposed a few years ago (Aebi 1999, 2006). 앫 Content validity 왌 Face validity (apparent validity) 왌 Logical validity 앫 Pragmatic validity 왌 Predictive validity 왌 Concurrent validity • Known group validity • Correlational validity 앫 Construct validity Before starting the presentation of these types of validity, it is necessary to point out that it is possible to test the validity of the whole survey or the validity of the delinquency scales derived from it. Content validity Content validity refers to question of whether the content of the survey – i.e. the offences included in it − corresponds to its label (is this survey really a self-reported delinquency survey?). As such, it entails examining if the behaviours included in the SRD survey are actually delinquent behaviours (face validity) and if they constitute a representative sample of the universe of delinquent behaviours (logical validity). From that definition it can be seen that there are two subtypes of content validity that we will discuss in the following chapters. Face validity By definition, a self-reported delinquency survey is supposed to measure delinquency. However, as we have explained at the beginning of this paper, some of the behaviours included in SRD surveys are not considered as criminal offences in Continental European countries. As mentioned before, the reason of this apparent paradox is that the concept of delinquency is broader in English speaking countries – i.e. it includes wrongful and antisocial behaviour − than in Continental European countries. The only empirical way of testing face validity is to compare the items included in the survey with the offences included in the criminal laws of the country. American researchers include under such heading the statutory laws that foreseen the status offences mentioned above and usually conclude that the survey meets the requirements of face validity. European researchers usually include such behaviours under the heading of problematic behaviour and conclude that the survey measures 37

Self-reported crime.book Page 38 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

delinquency and other forms of antisocial behaviour. Once more it must be mentioned that, in order to overcome this threat to face validity, the researcher must avoid mixing serious offences with problematic behaviours or trivial offences in a general delinquency scale. Logical validity We have already discussed the content of SRD surveys indicating that typically they include problematic behaviours, property offences, violent offences, substance abuse, fraud and dishonesty. From that list, it is obvious that many criminal offences are not included in the surveys, and therefore one can conclude that the behaviours included in it do not constitute a representative sample of all criminal offences. This implies a challenge to the logical validity of the survey. However, it has generally been considered that most SRD surveys include the most typical offences committed by juveniles and that therefore their logical validity is acceptable. In that context, we have already regretted the fact that intellectual property offences were not included until recent years, and the same can be said about bullying and the context in which alcohol is being consumed. From that point of view, it is possible to affirm that the logical validity of most SRD surveys has been improved in recent years. In SRD surveys conducted with adults, the behaviours included in the survey must be modified. For example, in Switzerland and Germany, SRD surveys were used for measuring delinquency of drug-addicts following heroin prescription programs and the offences included in such surveys were adapted to those related to hard-drug addict life-style. Pragmatic validity Pragmatic validity refers to question of whether an instrument allows establishing the current state (concurrent validity) or the future state (predictive validity) of the concept that is being measured. In the case of SRD surveys, concurrent validity entails examining if the survey measures the involvement in delinquency of the respondents; while predictive validity requires examining if the survey predicts correctly the future involvement in delinquency of them. Indeed, the only way to establish if an instrument measures correctly a phenomenon is to compare its results with the results provided by an instrument whose validity has already been proven. For example, a researcher that develops a new balance can test its measures with the one provided by the balances that are already available. The problem comes from the fact that there is no universally accepted measure of delinquency. 38

Self-reported crime.book Page 39 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

Furthermore, all existing measures of delinquency are considered imperfect. As a consequence, it is possible to compare different measures of delinquency, but it is not clear how to interpret the differences found between these different measures. In that context, the results of SRD surveys have usually been compared with the results of official measures of crime, such as police or court files. Predictive validity The correct way of testing predictive validity implies using an SRD survey at time 1 (t1) and, on the basis of the analysis of the survey, making a prediction about the future implication in delinquency of the respondents to the survey. After some time (t2), a second SRD survey must be conducted with the same sample and a comparison of the predictions with the observations should be performed. In spite of the existence of quite a few SRD longitudinal surveys, predictive validity is never tested that way. Usually, the researchers do not make a prediction. Instead of that, they compare the results at t1 and at t2 in order to establish retrospectively which would have been the best predictive factors at t1. Concurrent validity In order to estimate concurrent validity, one needs to compare the results obtained with two instruments. In that perspective, researchers distinguish two subtypes of concurrent validity: Known group validity and co relational validity. – Known group validity In order to test known group validity, one compares the scores in selfreported delinquency of two or more groups whose implication in delinquency is different according to another crime measure. For example, using official records of crime − such as police or court records − one can compare the SRD scores of the group of persons officially delinquent (known by the police or convicted by a court) with the ones that are officially non-delinquent. The logic of the comparison implies that the groups that are different according to official measures of crime should also be different according to SRD measures. Thus, officially delinquent youth should disclose more offences than officially non-delinquent youth. If that is not the case, there is a problem of validity with one of the measures. The alternative measures used to test the validity of SRD surveys vary widely. Some are close to official measures of crime (contacts with the police, contact with a court, institutionalisation), some use non-official

39

Self-reported crime.book Page 40 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

measures (reports from the teachers, personality tests, class play role assignment, etc.) and some combine both types of measures25. In that context, the typical way of conducting known group validity tests is to compare self-reported police contacts with the police with contacts registered by the police. If respondents do not admit contacts registered by the police, it is clear that they are lying. The results of the studies that tested known group validity show that the group that was supposed to be more involved in delinquency did present higher SRD scores − in incidence and variety scales − than the group that was supposed to be less involved in it. – Correlational validity According to Hindeland, Hirschi and Weiss (1981, ch. 5) correlational validity can be seen as a refined form of known group validity. Both of them compare two measures of crime, but the difference comes from the fact that in known group validity one uses the crime measures as a criterion to assign individuals to different groups, while in correlational validity the goal is to calculate the correlation between both measures. Thus, studies that tested correlational validity have also compared SRD surveys to official and non-official measures of crime. A review of the studies that conducted comparisons of SRD surveys and official measures of crime (Aebi, 2006)26 shows that they have usually found weak positive correlations. In particular, the size of the correlation is heavily influenced by the scale used to measure delinquency. The strongest correlations are found with the widest scales of delinquency in terms of the number of offences included and the period studied. In addition, the correlations are sensitive to the sex, age and ethnic background of the sample, suggesting a differential validity of SRD surveys. In that context the validity of SRD surveys seems stronger for juveniles and for autochtones. The fact that the correlations are positive shows that both measures are measuring the same concept. As this concept was defined as delinquency, it is possible to conclude that both measures are relatively valid as measures of crime. However, as we will see in the following chapter, it is not easy to interpret the size of the correlations found. Some reflexions on the logical problems related to the comparisons of SRD surveys and official measures of crime The history of criminology can be seen as the history of the search for reliable and valid measures of crime. In that context, Hindelang, Hirschi, 25. For a detailed review of these studies, see Aebi (2006, 162-164, with references). 26. For a detailed review of these studies, see Aebi (2006, 169-177, with references).

40

Self-reported crime.book Page 41 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

Weis (1981, 97s.) pointed out that the fact that researchers would turn to official measures of crime as a criterion of validity of SRD surveys was seen as a paradox or at least as an admission of the superiority of such measures by some researchers. Indeed, if official measures of crime were considered as non-valid and SRD survey were developed as an alternative to them, it is not logic at all to use official measures of crime to test the new measure. Trying to find a solution to that situation, Hindelang, Hirschi, Weis (1981, 97s.) pointed out that a good survey should at least identify all those who had been identified by the official measure, because the latter is only considered completely non-valid by those who are unconditional supporters of the SRD surveys. This is precisely the kind of study that is conducted when testing known group validity, and we have seen that SRD surveys usually meet the requirements of such test. In that context, the fact that persons identified as delinquent by the police or the courts show higher scores in SRD than those not identified as such has been interpreted as a corroboration of the validity of SRD surveys. However, it can also be interpreted as a corroboration of the validity of official measures of crime. Indeed, if a person that admits having committed a lot of offences has already been contacted by the police while someone who has not admitted offences has never been contacted by them, one could conclude that such a result shows that the police are doing a relatively good work. The problem with this kind of reasoning is that it is deductive in nature and it starts with an axiom. Indeed, researchers have adapted the classic syllogism “Man is mortal, Socrates is a man, therefore, Socrates is mortal” in this way: “Official measures of crime are relatively valid; SRD surveys match official measures; therefore, SRD surveys are relatively valid”. However, this method − severely criticized by Russell (1961/1912) − can only lead to a valid conclusion when the major premise can be accepted without any restriction (i.e. when it constitutes an axiom), and that is not the case as far as the validity of official measures is concerned. From a logical point of view, the problem cannot be solved because no measure of delinquency is considered as completely valid. From an empirical point of view, the only possibility is to test both measures in both ways. This implies testing whether SRD surveys can differentiate official delinquents, and also whether official measures can differentiate the respondents that defined themselves in the SRD survey as delinquents. The combination of both types of tests is interesting because, on the one hand, if persons identified by the police do not admit offences, one can conclude that SRD surveys are not valid; while, on the other hand, if the police (or courts) do not identify the persons that admit offences in the

41

Self-reported crime.book Page 42 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

SRD survey, one can conclude that official records are not valid. Finally, if the two measures coincide, the validity of both of them is corroborated. In practice, the second type of test − creating groups of delinquents according to the SRD survey and comparing their scores in official records − has seldom been used because usually SRD studies have been conducted mainly with adolescents that are not heavily involved in delinquency and therefore it is difficult to establish a group of delinquents. One European exception is our research with adult hard-drug addicts following a heroin prescription treatment in Switzerland (Aebi, 2006). In that case, both measures were associated − even if SRD surveys allowed identifying more offenders and many more offences than police and court records − and therefore they were considered as relatively valid. Construct validity The degree of correspondence between the empirical results of the research and the theoretical previsions is called construct validity. However, when the results do not follow the theoretical previsions, sometimes it is difficult to know whether the theory is not correct or whether it is correct but the instruments used to test it are not valid (Selltiz, Whrigtsman, Cook, 1977). Of course, the same problem also exists when the theoretical previsions are corroborated by the results. In that case, one can wonder whether the results are not due to the research design or the instruments used. Construct validity has been studied mainly in the field of psychology, in order to test the efficiency of some tests. In the mid-1980s, Huizinga and Elliott (1986) considered that it had seldom been used in delinquency research because it was too complicate to examine within the same study both tests of theory and validity issues. However they concluded that there was some indication of the construct validity of SRD measures because, when correlations were not the ones specified by a given theory, researchers have concluded that the theory was wrong and not that the SRD measure was invalid (Huizinga, Elliott, 1986). Even if in recent years many theories have been tested using SRD studies, it is difficult to establish clearly the construct validity of the surveys because in many cases they do not include questions that would allow testing alternative theories. However, the importance of construct validity should not be underestimated because, as it has been shown in the field of domestic violence, research based on the self-report technique (Straus, Gelles, 1990) leads to completely different conclusions to the one based on victimization surveys.

42

Self-reported crime.book Page 43 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

7.

Reliability of SRD surveys

Reliability refers to the extent to which a measuring instrument – in our case, the SRD survey − would produce identical scores if it were used to make multiple measures of the same object (Huizinga, Elliott 1986). According to Hagan (2005, 298) there are two types of reliability: stability and consistency. Stability implies testing whether, assuming that conditions have not changed, a respondent gives the same answer to the same question on second testing; while consistency implies testing whether the set of item used to measure a phenomenon are highly related and measure the same concept (Hagan, 2005, 298). As with validity, it is possible to test the reliability of the whole survey or the reliability of the delinquency scales constructed on the basis of the survey. Reliability of SRD surveys has been tested using the test-retest method (i.e. administering the survey twice to the same population27), the split-half technique (in which the instrument is divided in two halves and each half is analyzed separately and then compared to the other in order to see whether the scores are similar), and the split-ballot technique (in which the sample is divided randomly in two halves and alternate forms of the instrument are administered to each half before testing whether the scores are comparable).

V.

Self-reported delinquency surveys and criminological theory

SRD surveys are a major tool for the development and testing of criminological theories, in particular of those that try to explain juvenile delinquency. In fact, since the 1960s, most theories in criminology have been developed on the basis of the results of SRD surveys. We have already mentioned the influence that the first SRD studies had on development of the labelling approach. Other paradigmatic examples in the United States are social bonding theory (Hirschi, 1969) and self-control theory (Gottfredson, Hirschi, 1990). In Europe, these surveys played a major role in the development of the integrated cognitive antisocial potential theory (Farrington, 2005), and Situational Action Theory (Wisktröm, 2005). Apart from that, SRD surveys are regularly used for testing existing theories. For example, as we have already mentioned, the ISRD-1 27. It is also possible to test alternate forms of the instrument on the same group. This technique is known as multiple forms (Hagan, 2005, 299).

43

Self-reported crime.book Page 44 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

questionnaire was inspired mainly by control theory, and the second one includes different theoretical approaches. The later approach − a combination of different theories − is used by the majority of the recent studies, including the Edinburg Study of Youth Transitions and Crime, the Belfast Youth Development Study and the KFN surveys in Germany. From a theoretical point of view, the ideal SRD survey should include questions referring to different theories in order to allow comparisons of the explanatory power of each of these theories. However, leaving aside the classical problems related to the operationalisation of the concepts of each theory − that are at the origins of many debates in criminology − it must be mentioned that some theories are harder to test than others. This is typically the case for labelling theory that requires a longitudinal design in order to compare at time t2 the implication in delinquency of persons who had contacts with the criminal justice system and persons who did not have such contact, but controlling that both groups had a similar implication in delinquency at time t1. Situational-based theories create also problems when it comes to measures the opportunities to commit offences. Usually, the questionnaires include questions on the lifestyle of the person that are then used to infer his/her exposition to the risk of committing an offence but, with only a few exceptions, the concrete occasions of committing offences are seldom taken into account.

VI. Self-reported delinquency surveys and criminal policies At a national level, the influence of SRD surveys on criminal policies is clearly related to the influence of this indicator in each country. In countries with a weak tradition of SRD studies, it seems that these studies are not playing a major role on the development of criminal policies or are having only some influence at the local level (e.g. Italy). On the other hand, when this crime measure becomes part of the criminological scene, it is often taken into account for such policies. This is the case mainly in the United Kingdom, where the Cambridge Study in Delinquent Development has inspired some legal reforms. In countries such as Finland, Germany and the Netherlands, SRD surveys are also playing a role in the political debate on crime and crime prevention.

44

Self-reported crime.book Page 45 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

VII. Conclusion SRD surveys have become a standard measure of delinquency in Europe. However, their validity cannot be established on an abstract basis. On the contrary, it is necessary to analyze each survey − paying particular attention to the sampling, the conditions of administration of the survey and the construction of the questionnaire − in order to establish its degree of validity. SRD surveys do not measure the most serious types of offences. However, they provide extremely useful information for minor and less-serious types of offences. Recommended good practices include : 앫 When using school samples, include special education classes and find a way of adding youth non attending school to the sample. 앫 Improving a common questionnaire, such as the ISRD-2, in order to allow comparisons across time and space. 앫 Clearly separate problematic behaviours and trivial offences from the rest of the offences. 앫 Include more serious offences in the questionnaire (e.g. sexual abuse). 앫 Improve measures of socio-demographic variables. 앫 Use CAPI, CASI and CAWI in order to reduce the costs of the survey and the risks of introducing mistakes while entering the information in the database. 앫 Include questions on victimization in order to have a more complete picture of the sample. The national reports summarized in this article show that in countries such as England and Wales, Finland, the Netherlands, Northern Ireland and Sweden, SRD surveys have been institutionalized and national surveys are run on a regular basis that allows the development of time series. In particular, the United Kingdom has a strong tradition of SRD studies and currently an impressive number of studies − including longitudinal ones − are taking place there. Germany is also running longitudinal as well as cross-sectional regional and local surveys on a more or less regular basis. Finally, in Belgium, France, Italy and Ireland, SRD surveys are not institutionalized but surveys are taking place punctually and the four countries have taken part in the second ISRD. The reports also show that the use of SRD studies decreased since the mid1970s to the end of the 1980s and clearly increased since the beginning of the 1990s. At the general European level, the positive experience of the ISRD-1 in 1992 – in which eleven of the twelve participant countries were European − showed that it was possible to develop joint and comparable research and 45

Self-reported crime.book Page 46 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

provided a common questionnaire that was later used in many studies. Finally, the creation of the European Society of Criminology (ESC) in 2000 provided a forum where European criminologists could meet and build up joint projects and clearly helped the development of the ISRD-2, in which 24 of the 30 participating countries are European, and their national correspondents meet regularly at the ESC Annual Conferences. In brief, not only SRD surveys seem to have found their place in European criminology as a major measure of juvenile delinquency, but it seems also clear that their use will probably increase in the future.

References AEBI M. F., 1997, Famille dissociée et criminalité. Le cas suisse, Kriminologisches Bulletin de Criminologie, 23, 1, 53–80. AEBI M.F., 1999, La validité des indicateurs de la criminalité. Les sondages de délinquance autoreportée face aux données de police et du casier judiciaire, Thèse de doctorat en criminologie, Lausanne, Université de Lausanne, Faculté de Droit, Institut de police scientifique et de criminologie. AEBI M.F., 2004, Crime Trends in Western Europe from 1990 to 2000, European Journal on Criminal Policy and Research, 10, 2-3, 163-186. AEBI M.F., 2006, Comment mesurer la délinquance? Paris, Armand Colin. AEBI M.F., JAQUIER, V., 2008, Les sondages de délinquance autoreportée, Origines, fiabilité et validité, Déviance et société, 32, 2, 205-227. ANDERSSON L., 2009, Self-Reported Delinquency Surveys in Sweden, This volume BECCARIA C., 1991, Des délits et des peines, Paris, Garnier-Flammarion (original ed.: Dei delitti e delle pene, 1764). CARRA C., 2009, Self-Reported Delinquency Surveys in France. This volume CHRISTIE N., 1966 [1964], Hvorfor blif unge mennesker forbrytere? in Andenaes J., Bratholm A., Christie N., Eds., Kriminalitet og samfunn, Bergen, J. W. Eides Boktrykkeri. CHRISTIE N., 1975, Hvor tett ett samfunn? Oslo, Christian Ejler’s forlag. CHRISTIE N., ANDENAES J., SKIRBEKK S., 1965, A study of self reported crime, Scandinavian Studies in Criminology, 1, 86-116. DURKHEIM E., 1988, Les règles de la méthode sociologique, Paris, Flammarion, (Original ed.: 1895). ELLIOTT D. S., HUIZINGA D., 1989, Improving self-reported measures of delinquency, in KLEIN M. W., Ed., Cross-national research in selfreported crime and delinquency, Dordrecht [etc.], Kluwer Academic Publishers, 155-186. 46

Self-reported crime.book Page 47 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

ERICKSON M., 1972, The changing relationship between official and self– reported measures of delinquency: An exploratory–predictive study, The journal of Criminal Law, Criminology and Police Science, 63, 3, 388– 395. FARRINGTON D. P., 1977, The effects of public labelling, The British Journal of Criminology, 17, 2, 112-125. FARRINGTON D.P., 2005, The integrated Cognitive Antisocial Potential (ICAP) theory in FARRINGTON D.P., Ed., Integrated Developmental and Life-course theories of offending, New Brunswick, NJ, Transaction, 73-92. GARLAND D., 2001, The culture of control: Crime and social order in contemporary society, Oxford, Oxford University Press. GÖRGEN T., RABOLD S., 2009, Self-Reported Delinquency Surveys in Germany. This volume. GOTTFREDSON M. R., HIRSCHI T., 1990, A General Theory of Crime, Stanford, CA, Stanford University Press. GRASMICK H.G., TITTLE C.R., BURSIK R.J., ARNEKLEV B. J., 1993, Testing the Core Empirical Implications of Gottfredson and Hirschi's General Theory of Crime, Journal of Research in Crime and Delinquency, 30, 1, 529. GUTTMAN L., 1950, The basis for scalogram analysis in STOUFFER S. A., GUTTMAN L., SUCHMAN E. A., LAZARSFELD P. F., STAR S. A., CLAUSEN J. A., Eds., The American Soldier: Studies in Social Psychology in World War II. Vol. 4, Measurement and Prediction, Princeton, NJ, Princeton University Press, 60-90. HAGAN F.E., 2005, Research methods in criminal justice and criminology (7th ed.), Boston [etc.], Allyn and Bacon. HEMPEL C., 1966, Philosophy of Natural Science, Englewood Cliffs, NJ, Prentice Hall. HIBELL B., ANDERSON B., BJARNASON T., AHLSTROM S., BAKALIREVA O, KOKKEVI A., MORGAN M., 2004, The 2003 ESPAD report: Alcohol and other drug use among students in 35 European countries, Stockholm, Swedish Council for Information on Alcohol and Other Drugs / Strasbourg, Council of Europe. HINDELANG M. J., GOTTFREDSON M. R., GAROFALO J., 1978, Victims of personal crime: An empirical foundation for a theory of personal victimization, Cambridge, MA, Ballinger. HINDELANG M. J., HIRSCHI T., WEIS J.G., 1979, Correlates of delinquency: The illusion of discrepancy between self–report and official measures, American Sociological Review, 44, 5, 995–1014. HINDELANG M. J., HIRSCHI T., WEIS J.G., 1981, Measuring delinquency, Beverly Hills, CA, Sage.

47

Self-reported crime.book Page 48 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

HIRSCHI T., 1969, Causes of Delinquency, Berkeley, CA, University of California Press. HUIZINGA D., ELLIOTT D.S., 1986, Reassessing the reliability and validity of self–report delinquency measures, Journal of Quantitative Criminology, 2, 4, 293–327. JUNGER–TAS J., 1989, Self–report delinquency research in Holland with a perspective on international comparison, in M. W. KLEIN (Ed.), Cross– national research in self–reported crime and delinquency, Dordrecht [etc.], Kluwer, 17–41. JUNGER-TAS J., 1994a, The international self-report delinquency study: Some methodological and theoretical issues, in JUNGER-TAS J., TERLOUW G.-J., KLEIN M. W., Eds., Delinquent behavior among young people in the western world: First results of the international self-report delinquency study, Amsterdam [etc.], Kugler, 1-13. JUNGER-TAS J., 1994b, Delinquency in thirteen western countries: Some preliminary conclusions, in JUNGER-TAS J., TERLOUW G.-J., KLEIN M. W., Eds., Delinquent behavior among young people in the western world: First results of the international self-report delinquency study, Amsterdam [etc.], Kugler, 370-380. JUNGER–TAS J., HAEN MARSHALL I., 1999, The self–report methodology in crime research, Crime and Justice: A Review of Research, 25, 291–367. JUNGER–TAS J., HAEN MARSHALL I., RIBEAUD D., 1999, Delinquency in an International Perspective, Monsey NY, Criminal Justice Press / The Hague, Kugler. JUNGER-TAS J., HAEN MARSHALL I., RIBEAUD D., 2003, Delinquency in an International Perspective: The International Self-Reported Delinquency Study (ISRD), The Hague, Kugler. JUNGER-TAS J., TERLOUW G., KLEIN M., 1994, Delinquent Behavior Among Young People in the Western World: First Results of the International SelfReport Delinquency Study, Amsterdam [etc.], Kugler. KIVIVUORI T., 2009, Self-Reported Delinquency Surveys in Finland, This volume. KITSUSE J., CICOUREL A., 1963, A note on the use of official statistics, Social Problems, 11, 2, 131-139. KLEIN M. W., (Ed.), 1989, Cross-national research in self-reported crime and delinquency, Dordrecht / Boston / London, Kluwer Academic. KUHN T. S., 1970, The structure of scientific revolutions (2nd ed.), Chicago, University of Chicago Press. MCVIE S., 2009, Self-Reported Delinquency Surveys in the United Kingdom and Ireland. This volume. MOFFITT T.E. SILVA P.A., LYNAM D.R., HENRY B., 1989, Self-reported delinquency at age 18: New Zealand’s Dunedin multidisciplinary health 48

Self-reported crime.book Page 49 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in Europe

and development study, in JUNGER-TAS J., TERLOUW G.-J., KLEIN M. W., Eds., Delinquent behavior among young people in the western world: First results of the international self-report delinquency study, Amsterdam [etc.], Kugler, 354-369. MORIN E., 1990, Science avec conscience, Paris, Fayard. NYE F. I., SHORT J. F. JR., 1957, Scaling delinquent behavior, American Sociological Review, 22, 2, 326-331. PAUWELS L., PLEYSIER S., 2009, Self-Reported Delinquency Surveys in Belgium and the Netherlands. This volume. PORTERFIELD A.L., 1946, Youth in trouble, Fort Worth, Leo Portisham Foundation. RAUDENBUSH, S.W., BRYK, A.S., 2002, Hierarchical Linear Models: Applications and Data Analysis, Method (2nd. Edition), Newbury Park, CA, Sage. RUSSELL B., 1961, The problems of philosophy, in EGNER E., DENONN L. E. (Eds.), The basic writings of Bertrand Russell, New York, Simon and Schuster (article published originally in 1912). SELLTIZ C., WRIGHTSMAN L.S., COOK S. W., 1977, Les méthodes de recherche en sciences sociales. Montréal, HRW. STRAUS M.A., GELLES R.J. (Eds.), 1990, Physical Violence in American Families: Risk Factors and Adaptations to Violence in 8,145 Families, New Brunswick/London, Transaction. TRAVERSO S., CARTOCCI G., TRAVERSO G.B., 2009, Self-Reported Delinquency Surveys in Italy, this volume. THORNBERRY T.P., KROHN M. D., 2000, The self-report method for measuring delinquency and crime, in D. DUFFEE, (Ed.), Criminal Justice 2000 Volume 4: Measurement and analysis of crime and justice, Washington, DC, 33-83. WIKSTRÖM, P. O., 2005, The Social Origins of Pathways in Crime: Towards a Developmental Ecological Action Theory of Crime Involvement and Its Changes, in FARRINGTON D.P. (Ed.), Integrated Developmental and Life-course theories of offending, New Brunswick, NJ, Transaction, 211246.

49

Self-reported crime.book Page 50 Thursday, April 23, 2009 8:02 AM

Self-reported crime.book Page 51 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands Lieven Pauwels, Stefaan Pleysier

In this contribution a synthetic overview of self-report studies in Belgium and the Netherlands is offered. We will focus on the background, scope, and scientific and technical organisation of the most important self-report studies. From the observation that these kinds of surveys are sometimes lacking in attention towards methodological issues and measurement problems, we will stress research that invested in quality assessment and methodological improvement of the measurement instruments and methodologies used. We will discuss research and scientific assessments that concentrated on delineating good and bad practices, in the past, and for future research to come.

I.

Overview of Self-Report Studies in Belgium

Belgian research concerning youth and delinquency has primarily focused on the social and penal reaction towards delinquent behaviour of minors as well as the treatment of delinquent youth, i.e. the juvenile justice system (Christiaens, De Fraene and Delens-Ravier, 2005). In this paragraph we briefly overview the use of self-report studies from a historical point of view and highlight the most important studies. The focus is on juvenile delinquency, but also substance use is the subject of self-reports. Contrary to crime and victim surveys, which in Belgium are administered by federal authorities, research on adolescent behaviour and opinions in Belgium is far more fragmentized. Therefore it is difficult to generalize findings from several Belgian studies. The first self-report studies in Belgium were conducted by Junger-Tas (1976). She surveyed 402 adolescents (15-18 years old) that lived in a small city, about acts of crime they committed (N=27). Junger-Tas also applied this method for the study of substance use in Brussels. Several years later Vettenburg, Walgrave (1988) surveyed 1,698 adolescents that studied vocational training in Flanders, while Born (1987) conducted similar research in Wallonia.

51

Self-reported crime.book Page 52 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Most classical self-report studies were conducted to monitor substance (ab)use. The most common method was to conduct a classroom based Paper-and Pencil (PAPI) survey. These classical surveys guarantee anonymity and additionally include questions on socio-demographic background variables (age, sex, immigrant background) but also attitudinal questions (social bonds, moral values, parental controls, peer and teacher relationships) and questions about life styles, alcohol, tobacco and the use of illegal drugs. One detailed self-report study of substance abuse was conducted by De Ruyver et al. (1992) and more recently by Vermeiren et al. (2003). One reason of the popularity of the self report method to measure the prevalence of substance use has to do with the fact that it is a victimless crime, which limits the number of alternative measurement instruments. Melis (1994) and Enhus and De Pauw (1996) were among the first to explicitly study life styles of adolescents through self-reports.

1.

Recent developments

During the nineties and from the beginning of the twenty-first century few large-scale self-report studies can be mentioned. In Belgium, unlike the situation in The Netherlands, no systematic large-scale representative selfreports were conducted. However, existing studies show that juvenile delinquency can be looked upon as ‘normal’ behaviour that is a part of the process of growing up. Results from Belgian studies do not differ systematically from international findings (Junger-Tas, Terlouw, Klein, 1994) when prevalence of offending and correlates of offending are taken into account. We will demonstrate this later in this contribution. Belgian self-report studies have primarily focused on the study of undetected, unreported, unresolved crime (the so-called ‘dark number’) and in most cases urban samples were used instead of national samples (e.g. Born, Gavray, 1994; Spaey, 2004). An extensive research project of the KULeuven (Catholic University of Leuven) on Flemish youth included an inquiry (N=4,891) into the offending behaviour of 12 to 18 year old students (Goedseels, Vettenburg, Walgrave, 2000). Despite the known distortions of self-report data, this study reveals interesting information on the dark number (unreported crimes). We briefly comment some key findings of this study. Students were asked eight types of offences or problem behaviour, committed during the past year. Fare dodging was reported mostly (25,5%), followed by theft (23,4%), vandalism (20,7%), use of drugs (17,4%), carrying weapons (12,7%), assault and battery (12,6%), running away from home

52

Self-reported crime.book Page 53 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

(6,5%), and drug dealing (5,7%). With regard to the use and dealing of drugs, in 90% of the cases it involved only soft drugs. About half of the respondents (52%) committed at least one of the offences. One fifth committed only one of the offences, a same proportion two to three acts, 10% reported to have committed more offences. The latter group is often considered as a group at risk to develop a criminal career: this group didn’t only commit more types of offences, but also committed every offence more frequently and the police detected their acts more often. These career delinquents suffered significantly more from depression, suicidal thoughts, and experienced less parental support and control. Their school careers tended to be problematic. These findings correspond rather well with another Belgian study of Walloon juveniles in Liège (14 to 21 years old): the total prevalence of delinquent behaviour during the last year aggregates 56,1%, the total prevalence in the total course of their young lives amounts to 82,5%. Only the drug-related offences are remarkably less prominent among the Walloon youths (8,8% drug use and 1,6% drug dealing during the past year) (Born, Gavray, 1994). Most offences are committed only once (especially ‘assault and battery’ and ‘running away’) or occasionally. Carrying weapons, drug use and dealing are committed more recurrently (Goedseels et al., 2000). Criminal acts classified as serious have a rather low frequency (Born, Gavray, 1994; Pauwels, 2007). Similar to other research, girls are less delinquent, although the differences are less striking for running away, drug use and fare dodging. Regarding age, the well known and documented age-crime curve is found in Belgian self-report studies, delinquency peaks on the age of 15-16 and declines from age 17 (apart from the use of drugs). Vandalism and especially drug use typically take place in group, while running away, assault and battery and drug dealing are committed alone (Goedseels et al., 2000). Differences between Belgian and immigrant juveniles are existing, but small, only for fare dodging, assault and battery and carrying weapons the latter group presents slightly higher percentages. Students with a higher educational level commit significantly less offences. Most crimes (70-90%) are never discovered; vandalism and assault and battery are the most detected by the police. Noteworthy is that the group of offenders of these crimes was also more victimised (Goedseels et al., 2000). Generally – as far as a comparison with other research studies is allowed – Belgian youths nowadays do not seem to be more criminal as their peers in the past or their peers abroad. A third large scale self report study worth mentioning is the representative self report (N=4,496) conducted by Vercaigne et al (2000) in Brussels. This 53

Self-reported crime.book Page 54 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

research project was a cooperation between the Department of social geography (Catholic University of Leuven) and the Department of juvenile criminology (Catholic University of Leuven), ordered by the Belgian Science Policy. The researchers surveyed 4.496 students within the Belgian “high-school” educational level (age 12-18), in 48 different schools in Brussels. The questionnaire focused on several delinquency items, family situation, background variables, but most of all the detection of crime. 48,9% said that they did not commit any of the offences listed in the questionnaire within the last twelve months. Of all respondents that reported an offence in this self-report, only 19% reappeared in official crime statistics.

Shoplifting Carrying weapons Use of drugs Vandalism Graffiti Harrassing people on the streets Hitting someone on purpose Car theft Selling drugs Burglary Theft of purse/wallet

Boys 40.1 33.0 20.4 22.3 17.6 17.6 19.9 10.1 8.0 6.5 6.3

Girls 27.1 8.3 12.1 6.2 9.1 8.5 5.4 1.1 2.2 0.7 1.2

All 33.6 20.7 16.2 14.2 13.3 13.0 12.6 5.6 5.1 3.6 3.7

Source: Vercaigne et al (2000)

Table 1: Percentages having committed some offences (12 months)

Some key findings are that adolescent-limited offences (graffiti, shop lifting, vandalism…) tend to peak at age 16 and then decline. Nonadolescent limited offences (use of drugs, car-theft, burglary) continue to rise also after the age of 16. This category needs to be studied more in depth as these adolescents are continuing a pathway into structural delinquency. This study also revealed that adolescents with immigrant background tend to have somewhat higher frequencies in offending concerning carrying weapons, use of drugs, vandalism, shoplifting and burglary. Street crimes are more likely to be reported by adolescents in lower social class. These offences are also more likely to be committed by adolescents with an immigrant background. In 2002, a representative sample of adolescents (14-18 years) was surveyed to measure antisocial behaviour. The results show that different kinds of problem behaviour are strongly 54

Self-reported crime.book Page 55 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

correlated (De Groofs, Smits, 2006). Recently Pauwels (2008) used a selfreport study (N=2,486) to measure involvement in informal youth groups that use violence. Only very few adolescents (5.9% of the sample) reported to be member of an informal youth group where violence was committed. The results show a strong interaction between immigrant background and sex of respondents, suggesting that male adolescents with immigrant background have the highest risk of becoming involved in such a “street oriented group”.

2.

The use of self-report studies to test theories of delinquency

Self report studies are not only used to estimate prevalence of delinquency, but also to test theories of offending. Belgian self-report studies are very seldom used to test theories of (juvenile) delinquency. Vettenburg and Walgrave (1988) developed the theory of social vulnerability based on selfreport studies. The latter was conducted by order of the Ministry of Education. These authors found empirical evidence for the negative effect of labelling of social vulnerable adolescents. These adolescents tend to have higher chances of getting involved in delinquency because of negative experiences with teachers, police and the criminal justice system in general. More recently, Pauwels (2007) tested an integrated version of social disorganization theory using a large-scale sample of Antwerp young adolescents (age 12-15 on average) in his doctoral dissertation. Pauwels questioned the nature of the ecological correlation between the number of juvenile offenders and poverty at the neighbourhood level. Descriptive results of this self-report study can be found in appendix 1. Key findings are that the neighbourhood level distribution of juvenile offenders is unequal, and has concentrations in poor neighbourhoods, a finding similar to previous Belgian self-reports in metropolitan context (Melis, 1994; Vercaigne et al, 2000). However, as aggregate level analyses are not suitable to test contextual effects of neighbourhood characteristics on juvenile delinquency, the author used multilevel analysis to test whether growing up in poor neighbourhoods increases the frequency of offending. It was shown that the housing market is partially responsible for the concentration of juvenile offenders in poor neighbourhoods. But neighbourhood characteristics cannot explain individual differences in offending. Further Pauwels integrated social disorganization theory with Wikström's situational action theory (Wikström, 2004) and found empirical evidence for the latter theory. Life style risk was the strongest

55

Self-reported crime.book Page 56 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

predictor for delinquent behaviour, but the effect of life styles was dependent on adolescents’ morality.

3.

The JOP-Monitor in Belgium: a new initiative

Background Research on adolescents’ behaviour and attitudes in Belgium is, as we have argued, rather unsystematic, shows a lack of coordination and therefore is characterised by many overlapping issues but also is characterised by empty fields. Policy makers and practitioners experience this lack of uniformity as a problem. To solve this problem a “research platform on adolescent research” (in Dutch, Jeugdonderzoeksplatform (JOP)) has been established as an initiative of the Flemish Ministers of the Interior, Culture, Youth and Civil Service. The JOP monitor is an interdisciplinary collaboration between three research groups, Research Group Tempus Omnia Revelat (Free University Brussels), Department of Social Welfare (University of Ghent), Research Group Juvenile Criminology (Catholic University of.Leuven). The Mission of the JOP consists of (1) collecting existing research on juveniles, analyse this material, synthesise it, and (2) to initialise new research through exchange with the field of practitioners, policy makers and scientific partners to try to collect information on juveniles and children and make these data accessible for all people that show interest. Scope The monitor is a research tool designed to systematically measure experiences, living conditions and activities of adolescents. The aim of the monitor is to gather techniques and item instruments used in earlier research, and promote the continuous use of valid measurement instrument and questionnaires. One central aspect is that the monitor is probing data that are important for adolescents and children. The monitor does not depart from adult concern for children. The first fieldwork was planned in 2005 and first results have recently been published (Vettenburg, Elchardus, Walgrave, 2007a, b).

II.

Overview of Self-Report Studies in the Netherlands

Self report studies have a stronger tradition in the Netherlands, contrary to the situation in Belgium. The Scientific Research and Documentation 56

Self-reported crime.book Page 57 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

Centre (WODC) of the Dutch Ministry of Justice conducts on a recurrent base self-reports studies on representative samples of Dutch adolescents. Apart from the WODC periodical study, there is also a joint effort between the SCP (Netherlands Institute for Social Research) and the NIBUD (Dutch Institute for Budget). As the WODC Monitor is of major importance for knowledge on juvenile offending, we discuss its general methodology and sampling issues. The WODC Monitor has gone through a lot of transformations. We discuss the methodology as it is today. The Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) is participating in several ongoing longitudinal self-report studies. However, the main subject in these studies is lying on testing criminological theories. A thematic group on adolescent development and offending is continuously working within the NSCR on the subject of adolescent offending.

1.

The WODC self-reported delinquency studies

General Methodology The WODC survey takes place every second year, and is ordered by the Ministry of Justice. Sample size is about one thousand respondents. Central questions include frequencies of offending, risk and antisocial behaviour in general, truancy, substance use, bullying, the detection of offences, background variables and attitudinal questions (Kruissink, Essers, 2001, 2004). The items on juvenile offending are asked by the use of a portable computer, to avoid item-nonresponse. Other information is gathered by face-to-face interviews. Questions on offending are asked in a particular order. First, it is asked if an offence “ever” has been committed. After the general question, it is asked more in detail how many times during the past twelve months the offence has been committed, and if the police found out about the behaviour. In the latest edition the self-reported offending-items were revised. Some items have been removed and other items have been introduced. By large the questionnaire remained identical in order to compare results and to detect trends in adolescent offending. One example is the removal of questions on vandalism of public phones, which have disappeared because of the large-scale use of mobile phones. In the latest version attempts were made to create equilibrium between violent offending and property offending. In appendix 3 the questions used to evaluate adolescent offending between 1990-2001 can be found.

57

Self-reported crime.book Page 58 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Sample and response In recent years, a lot of attention is paid to proportionally sample ethnic minorities. This is necessary to have a better estimation of the prevalence rate by subgroups at the national level. Efforts were made to obtain a better representation of “traditional” ethnic minorities in the sample, such as Turkish, Moroccan, Antillean and Surinam adolescents. To select Dutch adolescents, a telephone-based screening took place. First the researchers contacted households at random and asked if an adolescent of the target population was present and if the adolescent was prepared to participate in the survey. If that was the case, social and demographic background variables were taken into account to have a representative sample. Next, an introductory letter was send to the respondents. This letter was also addressed to the parents to inform them on the survey. Afterwards, the interviewer made an appointment. To approach immigrant adolescents, screening by telephone was not seen as appropriate. Households with immigrant background dispose less frequently of a telephone. Therefore direct contact was taken at the home address of the respondents. The introductory letter was given to the parents at that moment.

2.

The use of self-report studies to test theories of offending

In the Netherlands, several PhD studies have been published the last decade where self-reports are used to test criminological theories. We highlight some major tests. Bruinsma (1985) tested the theory of differential association and find some empirical evidence for this theory. Interestingly, Bruinsma found differences between boys and girls and discovered some contradictions with Hirschi’s theory of the social bond, that were in favour of differential association theory. Rovers (1997) looked for possible ghetto effects of neighbour poverty on adolescent offending. Similar to the earlier mentioned Belgian study (Pauwels, 2007), Rovers used an integrated theory (integration between social disorganization theory and strain theory) and found only empirical evidence for the effects of his revised strain theory. Similar to Pauwels, Rovers did not find evidence for ghetto effects of neighbourhood structural conditions on adolescent offending. Rovers (1997) paid a lot of attention to method bias in his PhD study. Weerman (1998) used self-report data to test aspects of social bonding theory and offending. Weerman (1998) found empirical evidence for the fact that bonding variables (social control, parent-child relation, attachment) could explain both individual differences and

58

Self-reported crime.book Page 59 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

changes in offending. Weerman (2001) has also paid a lot of attention to the role of peers in offending.

III. International survey participation of Belgium and the Netherlands Just as it is the case in Belgium, the Netherlands have been participating in two sweeps of the ISRD (Junger-Tas et al., 1994). The ISRD, based on a common set of questions for respondents, was started in 1990 with 12 countries participating. These countries included the Netherlands, Portugal, Spain, England and Wales, Switzerland, Belgium, Finland, Germany, Northern Ireland, Greece, Italy, New Zealand, and the state of Nebraska in the United States. One general aim of the ISRD-project is to generalize findings of the restrictions of official data, but also to gather information about differences in prevalence of offending at the country level. One problem in the first sweep is that there existed rather big sampling differences. Sometimes national samples were used but in many cases city samples were used, thus impeding generalization at the country level and cross-national comparison. During the 7th edition of the annual conference of the European Society of Criminology (Bologna, 2007) some results of the second sweep were presented.

Netherlands England and Wales Portugal Switzerland Spain Mannheim Belfast Liege (Belgium) Athens Nebraska 3 Italian cities Helsinki

Property Violence Drugs offences 29.5 29.3 15.3 16.0 15.8 25.3 21.4 29.5 11.3 33.5 29.1 20.9 20.1 34.5 15.4 20.7 21.7 7.0 25.5 23.8 19.9 27.3 29.9 8.2 34.9 51.8 9.1 36.9 34.9 17.3 16.7 14 6.2 28.6 34.7 13.2

Source: Newman (1999, 16)

Table 2: Percentage of respondents that that reported at least once in the past twelve months an offence (1992)

59

Self-reported crime.book Page 60 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Table 2 shows the percentage of respondents that reported at least one of the mentioned offences. Percentages do not differ dramatically from each other. Percentages of violence are rather high. One partial explanation is that violence against property and persons are included in the definition. Researchers of the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) have been participating in an international group (‘the Eurogang Network’) applying self reports to study ’problematic youth groups’ within the European and international context (Decker, Weerman, 2005; Haaland, 2000; Weerman, 2000; Esbensen, Weerman, 2005; Klein, Weerman, Thornberry, 2006,). Covariates of adolescent offending in Belgian and Dutch studies are very similar to covariates of adolescent offending found in the ISRD. In appendix 2, we added the correlation matrix of self reported delinquency from Pauwels’ (2007) study.

IV. Method issues in Belgian and Dutch self-report studies 1.

Method bias: an overview

The growing importance of large-scale general surveys, not the least within politics and the media, has undoubtedly increased the awareness and concerns about the quality, validity and reliability of survey data. In an international context, concern of data quality in self-report studies has led to a substantial body of research showing the weaknesses and strong points of the self report method (Hindelang, Hirschi,Weiss, 1981; JungerTas, Haen-Marshall, 1999; Thornbery, Krohn, 2000). Despite the international focus on method issues, the topic gained less attention in Belgium, compared to the Netherlands (Van Kerckvoorde, 1995; Ponsaers, Pauwels, 2003). Researchers in the Netherlands have by and large good experience with method issues, and self-report studies are especially used to test theories of crime causation. Decent overviews of the possibilities, method issues and problems and prospects are available from Junger-Tas, Haen-Marshall (1999). The authors pay attention to most common problems in self-reports, such as sampling options, participation and response rate concerns, and validity problems related to respondent characteristics, criminal involvement, and memory effects. Other central issues include instrument construction, conceptualization of the dependent variable, administration of the instrument, and reliability. According to Junger-Tas and Haen Marshall (1999) the self-report method has improved greatly over the past fifty years. Many of its problems and limitations have been addressed. Although the self-report method does 60

Self-reported crime.book Page 61 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

not replace other measures or methods, it has become a valuable tool for measuring criminal involvement and testing theory. Especially in the Netherlands issues of reliability and validity have been studies in detail as shall be shown later. In this contribution we shall evaluate the method issues following Groves’ (1989) typology of potential survey error i.e. sampling error, coverage error, nonresponse error and measurement error. Sampling error results from using a sample instead of surveying the entire population. A coverage error results from the failure to give all the units of the population on which information should be gathered a nonzero chance to be included in the sampling frame. Nonresponse error results from the fact that not all the units of the gross sample actually participate in the survey. Finally, different processes related to the instrument, the interviewer, the respondent, and the data collection mode contribute to measurement error (Braun, 2003, 137-138). Sampling problems School-based (multi stage) sampling is the most common used methodology in Belgian self-report studies. Large-scale surveys, such as the interdisciplinary scientific research measuring living conditions and perspectives of Belgian adolescents use multi stage sampling (Goedseels, Vettenburg, Walgrave, 2000). In that case, schools are sampled in a first stage, and in a second stage pupils are sampled by additional criteria to capture the diversity in educational groups. It is fair to state that consequences of sampling error and coverage problems in Belgian studies never have been a main issue of research. The problem is recognized in selfreport studies, but the methodological topic has never been prioritised. The PAPI-method is by large the most used method to collect data. Other methods include personal interviews, but as these are more expensive, they are less used. In the Netherlands household sampling is used in the WODC Monitor. To correct for school-based sampling multilevel models can be used, to deal with the clustered data structure. While this methodology has become common within other fields, such as sociology, multilevel studies on school based self report studies are extremely rare in the Belgian context. One example can be found in the self-report study conducted by De Fraine et al (2005) and Pauwels (2007). In the Netherlands multilevel studies on self-report data have been conducted by Rovers (1997) and Bruinsma (1992).

61

Self-reported crime.book Page 62 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Non coverage and nonresponse error Non-observational error is error due to the problem of non-coverage of respondents (e.g. students that go to schools outside the reach of the sampling area in case of school based sampling), non-contact of respondents (e.g. the absent students, truants and drop-outs in school based delinquency surveys) and noncooperation or the unwillingness of respondents to participate in the survey (Stoop, 2005). According to many scholars this kind of error is generally seen as the largest problem of error in survey research (Assael, Keon, 1982; Marker, 2002). Unit nonresponse has repeatedly been reported as one of the most serious validity problems that threaten conclusions drawn from self-reported delinquency studies (Cernkovich, Giordano, Pugh, 1985; Junger, 1989; Van Kerckvoorde, 1995; Ponsaers, Pauwels, 2003; Shannon, 2006, 78-100). Effects of nonobservational error on results of self-report studies in Belgium have up to now never been conducted, but the problem has been studied in the Netherlands (Junger, 1989). Measurement error Measurement error is caused by a difference between the true score of a respondent and the score that in reality is assigned to the respondent. Some important causes of measurement issues are discussed below with regard to Belgian self-report delinquency studies. This can be due to a variety of reasons concerning the measurement instrument or mode used, the interviewer, the respondent, etc.

2.

Method bias: specific issues

Item non-response as measurement error Missing data for a specific question or item in a questionnaire is referred to as item nonresponse. It occurs when a respondent fails to provide an answer to a required item (Huisman, 1999). Item nonresponse differs from other sources of error and should also be clearly differentiated from unit nonresponse error, which is a form of non-observational error (Stoop, 2005). In the classroom PAPI survey, item nonresponse is also supposed to occur as a consequence of “bystanders” who may influence the response behaviour of students, but little research has been conducted in that field (Oberwittler, 2004). This effect can be avoided using envelopes (Bjarnason, 1995). It is obvious that item nonresponse should always be avoided (1) as the effective sample size decreases, which is especially a problem in 62

Self-reported crime.book Page 63 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

multivariate analyses, and (2) as survey estimates (e.g. means or percentages) might be biased if respondents who do not answer certain questions differ from those that do answer all questions. Item nonresponse on delinquency items is to some extent related to being male, immigrant background, school delay and survey disinterest (Pauwels, 2007). In one recently finished comparative study Pauwels, Svensson (2008) paid attention to the problem of item nonresponse on both questions on offending and on questions used to construct latent variables, representing etiological variables in criminological research. Two “PAPI” surveys of young adolescents in the cities of Antwerp (Belgium) and Halmstad (Sweden) were examined on the occurrence of item nonresponse. Item nonresponse rates on separate scales did not come close to the threshold value of 5% of missing data, the threshold above which scholars suggest the use of statistical imputation methods, i.e. methods to estimate missing data from other data (Rubin, 1987). Item nonrespons is however a potential problem when multivariate analyses are considered. A loss of data goes up to 10% when several variables are used in multivariate models. Although some significant bivariate relations exist between background variables, survey disinterest, and key theoretical variables, the key finding in this comparative study is that the use of multivariate logistic regression of item nonresponse on the mentioned background variables does not lead to correct predictions of item nonresponse at all. Further analyses of the effects of using a traditional regression imputation method showed no real changes in reliability scores (both alpha values and factor loadings) in both surveys. Correlations between etiological variables and delinquency suffered no substantive change. Although item nonresponse on offending is significantly related to background variables such as age, gender and immigrant background, these variables cannot explain individual differences in item nonresponse on offending. Therefore, the results from regression analyses of offending are not likely to be biased because of the item nonresponse on offending. Furthermore, it seems that imputation, when the total amount of missing data for multivariate analyses is about 10%, does not alter results of multivariate regression analyses of offending. Social desirability Social desirability could be subscribed as the tendency of a respondent to be less willing to admit attitudes and behaviour with a rather threatening character, i.e. attitudes and behaviour that are less socially accepted. As a consequence, researchers tend to underestimate the individual frequency of committing delinquent acts. Social desirability is therefore a way of lying (see further).This validity problem can however be actively studied with

63

Self-reported crime.book Page 64 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

the help of social desirability scales, especially designed to measure this tendency. Most of the time this tendency is measured by asking respondents about behaviours that almost all of us have ever committed. The basic philosophy is that respondents having high scores on the social desirability items can be omitted from further analysis. Adorno et al (1950) already has used such types of scales to discern social desirability from fascist attitudes among respondents. Within the field of criminology such scales are also known (Hindelang, Hirschi, Weis, 1981). In the Netherlands, we have only seen few researchers use social desirability scales. Rovers and Wouters (1996) and Rovers (1997) are among the few examples in the Dutch context. Concerning the study of social desirability and its effects as a response set, diverse research strategies are known. Rovers (1997) dealt with the problem of social desirability by eliminating respondents with the highest scores (highest quartile) on a social desirability scale. This approach is useful, although obviously a sufficiently large sample size is necessary. Another approach is to see the social desirability scale as a control variable, and therefore to statistically control for, or model, social desirability. This method is useful when sample size is not sufficiently large. Social desirability is then taken into account within the model. One Belgian study tried to measure the effect of social desirability on explanatory models of self reported delinquent behaviour using a causal modelling approach (Pauwels, Pleysier, 2005). It was found that social desirability decreases frequencies of offending, while the effects of theoretical variables (from social bonding theory) remained unchanged. Acquiescence The concept of acquiescence is not new as well. Acquiescence is the disposition of a respondent to respond systematically to questions in another way than he would if the questions were offered in another form (Billiet, McClendon, 2000, 610). Respondents are asked to indicate their level of agreement on a set of items. Acquiescence – or agreeing bias – is the tendency to agree with the items offered, independent of the direction in which the item is formulated (positive or negative). The presence of acquiescence when using unbalanced scales (all negative or all positive) could lead to systematic overestimation or underestimation of the covariance between latent concepts (Mirowsky, Ross, 1991). In Belgium the above-mentioned study by Pauwels and Pleysier (2005) shows that acquiescence is also present in self-reported delinquency studies. This effort has never been replicated so it remains unclear to what extend results can be generalized.

64

Self-reported crime.book Page 65 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

Lying on offending Lying is a serious threat to the validity of all surveys. Within the framework of self-report studies respondents can lie about offending and lye about police contacts due to offending. Junger (1989) has examined the validity of self-reports among ethnic minorities and found that ethnic minorities significantly more often lye about official police contacts due to offending. In Belgium, no such empirical analysis has been conducted. One possible reason is that self-report studies guarantee anonymity to the respondents. The use of exact frequencies and memory’s tricks Questions on offending often are posed in several ways: first a general question is used, then follow-up questions are asked: frequencies of committing the offence within a time frame, usually within the past twelve months. It is sometimes argued that questions about exact frequencies only reveal “pseudo-exactness” because of the problem of telescoping. However, as far as we know, research on the effects of telescoping on self reported offending has never been conducted in Belgium and the Netherlands. Issues on reliability and validity In The Netherlands, a lively debate was held in the eighties and nineties about the reliability of self reported delinquency items and scales. Following Huizinga and Elliott (1986), Bruinsma (1991, 1994) tested the ’test-retest’ reliability of a classroom based self report study. He concluded that the relative stability of findings and contextual conclusions in selfreports could be questioned. Bruinsma repeated a self reported delinquency questionnaire two weeks after he conducted the original study in the same classroom and found that only 5,2% of the respondents reported exactly the same results. Bruinsma also questioned the internal consistency of delinquency scales because of low reported Cronbach's alpha values. Not everybody agreed with the harsh criterion Bruinsma used. Heijden et al (1994) argued that Bruinsma’s demands on reliability are so strict that doubt raised whether these demands fit social sciences. Nijboer (1995) on the other hand, argued to take these findings seriously, and suggested that more research was needed in this field. Another way of testing reliability is the use of Mokken scaling analysis (Rovers, 1997). Rovers used this method to create “serious offending” sub scales. Hessing and Elffers (1995) have shown very huge difference between self-reports and official known data. Their conclusion towards the validity of self-report studies was very

65

Self-reported crime.book Page 66 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

critical: self-reports inform about the willingness to reports behaviour, but do not inform about real behaviour. A statement that can be questioned, as Hessing and Elffers (1995) applied the self-report methodology on adults, while in most cases the method is used on adolescents. It is known that self-report studies perform better on a young population. Adults are on average more skeptical towards issues as confidentiality and anonymity.

V.

General conclusion on good practices in Belgium and the Netherlands

In general the method of self-report, its strong and weak points are known in Belgium and the Netherlands. More research into problems and prospects of this methodology has been conducted in the Netherlands. Issues on measurement (reliability and validity studies) have only recently gained considerable interest within social science research in Belgium and the Netherlands, compared to scholars in Anglo-Saxon countries. Within this contribution, we have highlighted several studies and contributions and have addressed what we consider to be key issues on measurement and methodology. Researchers using this methodology are aware of the methodological weaknesses in their surveys but should nevertheless invest more in the development of alternatives to get insight in the magnitude of bias arising when applying a self report instrument. Although we accept the method of measuring adolescent offending by means of self reports, it remains important to continuously improve efforts to understand possible sources of bias, and especially the magnitude of bias. When evaluating the self-report method, we suggest making a clear distinction based on the objectives of the self-report methodology. Selfreports can be used to estimate prevalence of adolescent offending (epidemiology) and to test theories on offending. With regard to the estimation of prevalence of offending, the sampling procedure is of uttermost importance. It is important to get more detailed information on the magnitude of sampling bias. Non-observational error should therefore be studied more intense then has been the case nowadays (e.g. household based surveys versus school based surveys). Future research should consider the possibilities of conducting mixed mode surveys, in order overcome the problem of non-coverage. One example was found in the WODC monitor. More knowledge is needed on the existence of differences between households that exclusively have a mobile phone, or landline phone, with respect to sampling of adolescents within households.

66

Self-reported crime.book Page 67 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

When the aim of a self report study lies in testing criminological theories, our general attitude towards the self-report methodology is relatively optimistic, but only when serious efforts are made to evaluate both reliability and validity of concepts and measurement. This means that, aside efforts made to avoid bias, one should consider also to measure and model bias and its effects on empirical outcomes. The use of advanced statistical tools allows for such research strategy. Although this is time consuming and rather expensive, survey quality and the general validity of the results will improve. When testing theories of offending, we cannot but conclude that there are very few alternatives: official data, such as police records do not provide us with background variables and social mechanisms (theoretical variables) necessary to test theories. Most selfreport studies are cross-sectional. We know very little on differences between cross-sectional data and longitudinal data and the potential effect of these differences on empirical tests of theories of delinquent behaviour. ‘Longitudinal’ analyses that are conducted are by large repeated measurement studies. On the other hand, and this is especially the case in Belgium, self-report studies that measure covariates of self-reported offending, largely pay attention to risk factors associated with offending. Although the identification of risk-factors has been of major importance, we suggest that more attention should be paid to move beyond the risk-factor approach and to develop and test theories that include the main mechanisms that are responsible for observed correlations between demographic background variables, risk-factors and offending.

References ADORNO T.W., FRENKEL-BRUNSWIK E., LEVINSON D.J., SANFORD R.N., 1950, The Authoritarian Personality, New York, Harper. ASSAEL H., KEON J., 1982, Nonsampling vs. sampling errors in survey research, Journal of Marketing 46, 2, 114-123. BARBARET R., BOWLING B., JUNGER-TAS J., RECHEA-ALBEROLA C., VAN KESTEREN J., ZURAWAN A., 2004, Self-Reported Juvenile Delinquency in England and Wales, The Netherlands and Spain, Helsinki, Finland, European Institute for Crime Prevention and Control. BIJLEVELD C., 2005, Methoden en Technieken van Onderzoek in de Criminologie, Den Haag, Boom Juridische Uitgevers. BILLIET J.B., MCCLENDON M.J., 2000, Modelling Acquiescence in Measurement Models for Two Balanced Sets of Items, Structural Equation Modelling, 7, 4, 608-628. 67

Self-reported crime.book Page 68 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

BJARNASON TH., 1995, Administration Mode Bias in a School Survey on Alcohol, Tabacco and Illicit Drug Use, Addiction, 90, 4, 555-559. BORN M., 1987, Évolution des rapports entre valeurs, milieu choisi et délinquance juvénile dans un intervalle de sept ans, in Coll. Changements de société et délinquance juvénile. Évolutions dans les problématiques, les conceptions, les politiques et les pratiques, Uitgeverij Acco, Leuven (rapport Sixièmes journées d’études internationales de Criminologie Juvénile, Leuven 25-27 mai 1987), 215-227. BORN M., GAVRAY C., 1994, Self-reported délinquency in Liège, Belgium, in JUNGER-TAS, J., TERLOUW G.J., KLEIN M.W. (Eds), Delinquency behaviour among young people in the western world. First results on the international self-reported delinquency study, Amsterdam, Kluwer Academic Publishers, 131-155. BRAUN M., 2003, Errors in comparative survey research: an overview, in HARKNESS, J.A., VAN DE VIJVER F.J. MOHLER P.PH. (Eds), Cross-cultural survey methods, Hoboken (NJ), John Wiley & Sons, 137-142. BRUCKERS L., 2003, Euroregionaal Jongerenonderzoek 2002, Een onderzoek naar vrijetijdsbesteding, gezondheidsbeleving en risicogedrag, uitgevoerd bij Limburgse jongeren in de 2e en 3e graad van het secundair onderwijs, Onuitgegeven, (onderzoeksrapport LUC, Centrum voor Statistiek en de provincie Limburg, studiecel, 2de directie Welzijn). BRUINSMA G.J.N., 1984, De methode van zelfrapportage van delicten, Tijdschrift voor Criminologie, 26, 295-308. BRUINSMA G.J.N., 1985, Criminaliteit als social leerproces. Een Toetsing van de differentile-associatietheorie in de versie van K-D Opp. Arnheim, Gouda Quint. BRUINSMA G.J.N., 1989, Scaling and reliability in self-reported property crimes, in KLEIN, M.W., Cross-National Research In Self-Reported Crime And Delinquency, Dordrecht, Kluwer academic publishers, 131-151. BRUINSMA G.J.N., 1991, De test-hertest betrouwbaarheid van de self-report methode, Tijdschrift voor Criminologie, 33, 245-255. BRUINSMA,G., 1992, Differential association theory reconsidered: An extension and its empirical test, Journal of Quantitative Criminology, 8, 29-49. BRUINSMA G.J.N., 1994, De test-hertest betrouwbaarheid van het meten van jeugdcriminaliteit, Tijdschrift voor Criminologie, 36, 3, 218-235. CERNKOVICH S., GIORDANO P., PUGH M., 1985, Chronic offenders, the missing cases in self-report delinquency research, Journal of Criminal Law and Criminology, 76, 3, 705-732.

68

Self-reported crime.book Page 69 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

CHRISTIAENS J., DE FRAENE D., DELENS-RAVIER I. (Eds), 2005, Protection de la jeunesse, Formes et Réformes/Jeugdbescherming, Vorming en hervorming, Bruxelles, Bruylant. CRONBACH L.J., 1942, Studies of acquiescence as a factor in the true-false test, Journal of Educational Psychology, 33, 401-415. CRONBACH L.J., 1950, Further evidence on response sets and test design, Educational and Psychological Measurement, 10, 3-31. DECKER S.H., WEERMAN F.W., 2005, European Street Gangs and Troublesome Youth Groups, Walnut Creek, CA, Altamira Press. DE FRAINE B., VAN LANDEGHEM G., VAN DAMME J., 2005, An analysis of well-being in secondary school with multilevel growth curve models and multilevel multivariate models, Quality & Quantity, 39, 297-316. DEGROOF S., SMITS W., 2006, Antisociaal gedrag bij jongeren onder de loep genomen, in ELIAERTS, C. (red.), Ernstige jeugddelinquentie: mythe of realiteit? Prevalentie, risicofactoren, justitiële afhandeling en interventies, Brussel, VUBPRESS, 25-52. DE RUYVER B., VERMEULEN G., FRANCK P., VAN DAELE L., 1992, Kansarmoede, druggebruik, criminaliteit, Gent, Onderzoeksgroep Drugbeleid, Strafrechtelijk Beleid en Internationale Criminaliteit. DE WITTE H., HOOGE J., WALGRAVE L., 2000, Jongeren in Vlaanderen: gemeten en geteld: 12- tot 18-jarigen over hun leefwereld en toekomst, een interdisciplinair onderzoek in opdracht van de minister-president van de toenmalige Vlaamse Regering, opgevolgd door de Administratie Wetenschap en Innovatie van de Vlaamse Gemeenschap, Leuven, Universitaire Pers Leuven, 2000. ENHUS E., DE PAUW W., 1996, Jongeren in de schijnwerper: een onderzoek naar leefstijl bij de scholieren van 12 tot 18 jaar in de Assese scholen, onderzoek VUB. ESBENSEN F.A., WEERMAN F.M., 2005, Youth Gangs and Troublesome Youth Groups in the United States and the Netherlands: A CrossNational Comparison, European Journal of Criminology, 2, 5-37. FARRINGTON D.P., LOEBER R., STOUTHAMER-LOEBER M., VAN KAMMEN W.B., SCHMIDT L., 1996, Self-reported delinquency and a combined delinquency seriousness scale based on boys, mothers and teachers: concurrent and predictive validity for African-Americans and Caucasians, Criminology, 34, 4, 493-517. GOEDSEELS E., VETTENBURG N., WALGRAVE L., 2000, Vrienden en hun vrije tijd. in DE WITTE H. HOOGE J., WALGRAVE L. (Reds.), Jongeren in Vlaanderen; gemeten en geteld. 12-tot 18-jarigen over hun leefwereld en toekomst (p. 149-183). Leuven: Universitaire Pers.

69

Self-reported crime.book Page 70 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

GIORDANO P.C., CERNKOVICH S.A., 1997, Gender and Antisocial Behavior, in STOFF D. M., BREILING J., MASER J. D. (Eds), Handbook of Antisocial Behavior, New York, Wiley, 496-510. GROVES R.M., 1989, Survey errors and survey costs, New York, John Wiley. HAALAND T., 2000, Vold – konflikt og gjengdannelse: en undersøkelse blant ungdom i fire byer. NIBR-rapport 2000, 14. Oslo, NIBR. HEIJDEN P.G.M. VAN DER, SIJTSMA K., HART H.’T, 1995, Self-report delinquentieschalen zijn nog steeds betrouwbaar, Tijdschrift voor criminologie, 37, 1, 71-77. HESSING D.J., ELFFERS H., 1995, De validiteit van de self-report methode in onderzoek naar regelovertredend gedrag, Tijdschrift voor Criminologie, 37, 1, 55-70. HINDELANG M., HIRSCHI T., WEIS J.G., 1981, Measuring Delinquency, Beverly Hill, Sage. HUISMAN J.M.E., 1999, Item nonresponse: occurrence, causes, and imputation of missing answers to test items, Leiden, DSWO Press. HUIZINGA D., ELLIOTT D.S., 1986, Reassessing the reliability and validity of self-report delinquency measures, Journal of Quantitative Criminology, 2, 4, 293-327. JUNGER M., 1989, Validiteit van zelfrapportage van delinquentie bij jongeren uit etnische minderheden, in DE JONG-GIERVELD J., VAN DER ZOUWEN J. (red.), De vragenlijst in het sociaal onderzoek, een confrontatie van onderzoekspraktijk en –methodiek, Van Loghum Slaterus, 153-174. JUNGER-TAS J., 1976, Verborgen jeugdelinkwentie en gerechlijke selectie, Brussels SCJM ( publ n°38). JUNGER-TAS J., HAEN MARSHALL I., 1999 The Self-Report Methodology in Crime Research in TONRY M. (Ed.) Crime and Justice. A Review of Research, Chicago, University of Chicago Press, 291-367. JUNGER-TAS J., HAEN MARSHALL I., RIBEAUD D. (Eds), 2003, Delinquency in an international perspective, The Hague, Kugler Publications. JUNGER-TAS J., TERLOUW G.J., KLEIN M. (Eds), 1994, Delinquent Behavior among Young People in the Western World; first results of the international self-report delinquency study, Amsterdam/New York, Kugler Publications. KLEIN M.W., WEERMAN F.M., THORNBERRY T.P., 2006, Street gang violence in Europe. European Journal of Criminology, 3, 413-437. KRUISSINK M., ESSERS A.A.M., 2001, Ontwikkeling van de jeugdcriminaliteit; periode 1980-1999, Den Haag, Ministerie van Justitie — WODC. KRUISSINK M., ESSERS A., 2004, Zelfgerapporteerde jeugdcriminaliteit in de periode 1990-2001, Den Haag, WODC-Cahier, 1. MARKER D., 2002, Improving Quality of Surveys, Paper presented at the International Conference on improving Surveys, Copenhagen 2002. 70

Self-reported crime.book Page 71 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

MELIS B., 1994, Jongeren in de stad. Onderzoek naar vrije tijd en conflict, Leuven, KUL, Onderzoeksgroep Jeugdcriminologie. MIROWSKY J., ROSS C.E., 1991, Eliminating Defense and Agreement Bias from Measures of the Sense of Control: A 2x2 Index, Social Psychology Quarterly, 54, 2, 127-145. NEWMAN G., 1999, Global report on Crime and Justice, New York, Oxford University Press. NIJBOER J., 1995, Het meten van delinquentie door middel van self-report, Tijdschrift voor Criminologie, 37, 1, 273-281. OBERWITTLER D., 2004, The Need for Computer-Aided Interview Techniques in School Surveys on Self-reported Delinquency, Paper presented at the 4th Annual Conference of the European Society of Criminology, Amsterdam, 25th to 28th August. PAUWELS L., 2007, Buurtinvloeden en jeugddelinquentie: een toets van de sociale desorganisatietheorie, Den Haag, Boom Juridische Uitgevers. PAUWELS L., 2008, Geweld in groepsverband bij Antwerpse jongeren. De bijdrage van schoolcontext en levensstijl, Tijdschrift voor Criminologie, 50, 1, 3-16. PAUWELS L., PLEYSIER S., 2005, Effecten van antwoordstijlen in etiologisch self-report-onderzoek, Een ‘causale modellenbenadering’, Tijdschrift voor Criminologie, 47, 1, 42-61. PAUWELS L., SVENSSON R., 2008, How Serious Is The Problem Of Item Nonresponse in Scale Constructs of Delinquency and Key Social Mechanisms? A Cross-National Inquiry of two Classroom PAPI-Self Report Studies in Antwerp And Halmstad, European Journal of Criminology (in press). PONSAERS P., PAUWELS L., 2003, De onderzoekbaarheid van jeugddelinquentie: over de selfreportmethode, in Vakgroep Strafrecht en criminologie (Red), Update in de criminologie, I, 45-70. ROVERS B., 1997, De buurt, een broeinest?, Een onderzoek naar de invloed van de woonomgeving op jeugdcriminaliteit, Nijmegen, Ars aequi Libri. ROVERS B., WOUTERS M., 1996, De nazaten van pietje bel, zelfrapportageonderzoek onder Rotterdamse kinderen, Tijdschrift voor Criminologie, 38, 1, 21-44. SHANNON, D. 2006. Chronic Offenders or Socially Disadvantaged Youth? Institutionalized Males as Missing Cases in School based Delinquency Research. Journal of Studies in Criminology and Crime Prevention, 7, 78100. SPAEY P., 2004, Violences urbaines et délinquance juvénile à Bruxelles. Les 1220 ans témoignent, Paris, L’Harmattan. RUBIN D.B., 1987, Multiple imputation for nonresponse in surveys, New York, John Wiley & Sons. 71

Self-reported crime.book Page 72 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

STOOP I.A.L., 2005, The Hunt for the Last Respondent. Nonresponse in sample surveys. The Hague, Social and Cultural Planning Office of the Netherlands. THORNBERRY T., KROHN M., 2000, The Self-Report Method for Measuring Delinquency and Crime, Criminal Justice Series, Vol. 4. VAN DER LAAN P.H., ESSERS A.A.M., HUIJBREGTS G.L.A.M., SPAANS E.C., 1998, Ontwikkeling jeugdcriminaliteit: periode 1980-1996; een tussentijds verslag Den Haag, Ministerie van Justitie-WODC, WODCOnderzoeksnotities, nr. 5. VAN KERCKVOORDE J., 1995, Een maat voor het kwaad, Leuven, Universitaire Pers Leuven. VERCAIGNE C., WALGRAVE L., MISTIAEN P., KESTELOOT C., 2000, Verstedelijking, sociale uitsluiting van jongeren en straatcriminaliteit, Leuven, Onderzoeksrapport DWTC. VERMEIREN R., SCHWAB-STONE M., DEBOUTTE D., LECKMAN P.E., RUSCHKIN V., 2003, Violence Exposure and Substance Use in Adolescents: Findings From Three Countries, Pediatrics, 111, 535-540. VETTENBURG N., 1998, Zelfgerapporteerde delinquentie bij jongeren. Een onderzoek bij 12-21 jarigen in Antwerpen. i.h.k.v., KUL, ISRD. VETTENBURG N., WALGRAVE L., 1988, Schoolervaringen, delinquentie en maatschappelijke kwetsbaarheid. Een theoretisch en empirisch onderzoek in het beroepsonderwijs, Leuven, Onderzoeksgroep Jeugdcriminologie. VETTENBURG N., ELCHARDUS M., WALGRAVE L., 2007a, Jongeren in cijfers en letters. Bevindingen uit de JOP-monitor 1, Tielt, Lannoo. VETTENBURG N., ELCHARDUS M., WALGRAVE L., 2007b, Jongeren van nu en straks. Overzicht en synthese van recent jeugdonderzoek in Vlaanderen, Tielt, Lannoo. WEERMAN F.M., 2000, De invloed van sociale bindingen op crimineel gedrag van jongeren. Sociale Wetenschappen, 43, 27-40. WEERMAN F.M., 1998, Het belang van bindingen. De bindingstheorie als verklaring van verschillen en veranderingen in delinquent gedrag, Groningen, Rijksuniversiteit Groningen. WEERMAN F.M., 2001, Samenplegen. Over criminele samenwerking en groepsvorming, Nijmegen, Ars Aequi Libri. WIKSTRÖM P.-O.H., 2004, Crime as Alternative: Towards a Cross-Level Situational Action Theory of Crime Causation, in MC COARD J. (Ed.), Beyond Empiricism: Institutions and Intentions in the Study of Crime, New Brunswick, Transaction, 1-37.

72

Self-reported crime.book Page 73 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

Appendix 1 Last year prevalence of self-reported delinquency in a recent Belgian urban self-report study (Antwerp, 2005) Vandalism (vehicle) Never Once or twice More than two times Graffiti Never Once or twice More than two times Buying stolen goods Never Once or twice More than two times Shoplifting Never Once or twice More than two times Serious theft Never Once or twice More than two times Threatening persons Never Once or twice More than two times Hit on purpose Never Once or twice More than two times Fighting outside school Never Once or twice More than two times Burglary Never Once or twice More than two times

Total sample

Boys

Girls

87.1 (2128) 6.3 (155) 6.5 (159)

81.0 (972) 9.2 (110) 9.8 (118)

93.1 (1155) 3.6 (45) 3.3 (41)

84.8 (2070) 8.4 (205) 6.8 (165)

84.5 (1012) 8.5 (102) 6.9 (83)

85.1 (1057) 8.3 (103) 6.6 (82)

92.2 (2248 4.3 (105) 3.5 (86)

87.5 (1048) 6.9 (83) 5.6 (67)

96.7 (1199) 1.8 (22) 1.5 (19)

83.2 (2032) 9.7 (237) 7.1 (174)

80.8 (970) 10.8 (129) 8.4 (101)

85.5 (1062) 8.7 (108) 5.8 (72)

94.4 (2299) 3.0 (73) 2.6 (64)

91.5 (1093) 4.5 (54) 4.0 (48)

97.2 (1205) 1.5 (19) 1.3 (16)

90.4 (2205) 4.8 (117) 4.8 (117)

86.8 (1040) 6.8 (81) 6.4 (77)

93.9 (1164) 2.9 (36) 3.3 (40)

69.9 (1711) 13.4 (327) 16.7 (409)

61.8 (745) 15.6 (188) 22.6 (272)

77.8 (966) 11.1 (138) 11.0 (137)

79.1 (1932) 11.0 (268) 9.9 (242)

70.1 (841) 14.7 (176) 15.3 (183)

87.8 (1090) 7.4 (92) 4.8 (59)

97.4 (2373) 1.2 (29) 1.4 (34)

95.7 (1144) 1.8 (21) 2.5 (30)

99.0 (1228) .6 (8) .4 (4)

Source: Pauwels, 2007

73

Self-reported crime.book Page 74 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Appendix 2 Bivariate correlates of self-reported delinquency (7)

(8)

General .710** 1 delinquency (6) Serious delinquency .409** .658** 1 (7) Life style risk (8) .424** .575** .416**

1

Violence (5)

Emotional attachment (9) Parental control (10) School social bond(11) Delinquency tolerance (12) Source: Pauwels, 2007

(5) 1

(6)

-.149** -.257** -.184** -.287**

(9)

(10)

(11)

(12)

1

-.302** -.422** -.318** -.416** .449**

1

-.221** -.337** -.231** -.366** .372** .357**

1

.416** .516** .339** .461** -.292** -.433** -.420**

1

** p < 0.01 level ()

Appendix 3: Initial self-reported delinquency-items used in the periodical WODC-self-report study 앫 앫 앫 앫 앫 앫 앫 앫 앫 앫 앫 앫 앫

Have you ever taken the bus, metro, tram or train without paying? Have you ever thrown out a window on purpose? Have you ever damaged a street lantern? Have you ever damaged or destroyed a car on purpose? Have you ever damaged a public phone on purpose? Have you ever damaged a bicycle on purpose? Have you ever damaged something on a bus, tram, metro or train? Have you ever damaged a bus stop? Have you ever changed prices in a shop to buy something for less money? Have you ever taken something from a shop without paying? Have you ever taken something from school from another student, from the school board, etc.? Have you ever stolen a bicycle? Have you ever bought or sold something of which you knew it was stolen?

74

Self-reported crime.book Page 75 Thursday, April 23, 2009 8:02 AM

Self-Report Studies in Belgium and the Netherlands

앫 Have you ever participated in a fight in public, e.g. at a football stadium, at a festival, manifestation or just on street? 앫 Have you ever threatened somebody to hit him or her? 앫 Have you ever hit somebody so badly that this person got injured (e.g. black eye bleeding from the nose, serious injuries). 앫 Do you ever carry a gun when you go out? 앫 Have you ever hurt someone with a knife or other weapon? 앫 Have you ever burgled a house, school, or something alike? 앫 Have you ever treated walls with pencils, paint or spray?

Follow-up questions on the self-reported delinquency items (When respondent has answered affirmatively on a previous asked delinquency item, follow-up questions are asked for each and every item) a) How old were you when you first did that? ……. (fill out age) 왌 I don’t know b) Have you also done that in the past twelve months? 왌 yes 왌 no 왌 I don’t know c) How many times have you done that in the past twelve months? ..... times (Give an estimation if you don’t remember the exact frequency) d) Were you alone most of th times or were you with others? 왌 alone 왌 together with others 왌 I don’t know e) In how many of those cases did the police find out about that? ..... times (Give an estimation if you don’t remember the exact frequency) f) Think of the last time you did this: why did you do it? 왌 I didn’t have a specific reason 왌 I liked it 왌 because other persons in my group were also doing that 왌 because of another reason, namely, ………………………….

75

Self-reported crime.book Page 76 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

g) How big do you think is the chance of getting caught for something like that? 왌 very big 왌 big 왌 not so big 왌 small 왌 very small

76

Self-reported crime.book Page 77 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland Janne Kivivuori

From the point of view of social, cultural, political and legal traditions, Finland is part of the Nordic cluster of countries with Sweden, Norway, Denmark and Iceland. The history of Finnish criminology is also intimately connected with Nordic co-operation. As an introduction to the topic, I will first briefly describe the history of self-report delinquency surveys in the Nordic area. I will then describe in more detail the Finnish self-report studies, with special emphasis on currently running standardised indicator systems which produce time series data by means of repeated survey sweeps.

I.

History: the Nordic tradition of self-report delinquency research1

The history of Nordic self-report research in crime and delinquency can be roughly divided into three phases: The first phase was one of enthusiasm and bold comparative designs, symbolised by the Nordic Draftee Research Programme. In the NDR project, the self-report method was applied in a sample of young males attending pre-military screenings in Copenhagen, Helsinki, Oslo and Stockholm, the capitals of Denmark, Finland, Norway and Sweden. This phase centred around the data collection years of the NDR (1961•1964), but if the pilot phase and the extended period of reporting are included, the first phase spanned the years 1959•1974, starting with a short enthusiastic paper by Andenaes et al. (1960)2 and ending with the rather gloomy monograph by Stangeland and Hauge (1974). Stangeland and Hauge’s monograph contained a concise comparative chapter (see also Anttila, Jaakkola, 1966), but a full-fledged international report never appeared. The second phase of Nordic self-report tradition, covering roughly the years from 1970 to 1990, was one of relative standstill in the development of self-report indicators. There were multiple reasons for this retrograde

1. 2.

This section is based on Kivivuori (2007a). This article was based on the first pilot sample collected in 1959.

77

Self-reported crime.book Page 78 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

development. First, technical problems with the NDR design made comparative work nearly impossible. Second, during the pre-computer and early computer age, quantitative analysis was relatively difficult and timeconsuming. Third, to the extent that quantitative research was to be exercised, victimisation research was preferred to self-report research. At least in Finland, the 1970s and 1980s saw important progress in the development of national victim surveys, while the development of selfreport crime surveys was brought to a virtual standstill. Fourth, it is possible that the NDR project failed to have any progeny because its policy use was exhausted during the 1960s. The original NDR researchers placed great emphasis on the moral idea that crime was normal (because selfreport findings showed that petty crime was so prevalent). Two titles of national NDR monographs, Norway’s Nyanser i grått (Shades of Grey, Stangeland, Hauge, 1974) and Denmark’s Kriminalitet som Normalitet (Criminality as Normality, Greve, 1972) testified to the great appeal of the finding that occasional delinquency was very prevalent. However, the NDR data harboured findings that could potentially contradict the “crime as normality” interpretation. For example, it was found that police detection likelihood reflected offending intensity (Christie et al., 1965). Nils Christie, one of the pioneers of Nordic self-report research, explicitly warned that the crime-as-normality rhetoric, appealing as it was, should not be pressed into absurdity (Christie, 1966 [1964], 59). However, he himself kept on using the normality argument (Christie, 1975, 73). The stubborn existence of the chronic offender thus haunted the early NDR researchers. I guess they sensed that an improvement of the method would shed more light on the “chronic” offender3. In a way, the moral promise of the method was fulfilled by the initial finding of high prevalence of occasional and petty delinquency. This interpretation is closely related to the fifth reason why Nordic self-report research faded after a promising start: the spirit of the times took an anti-positivist turn. The early NDR design had reflected the influence of empirically oriented American social science. This emphasis lost much of its appeal towards the end of the 1960s as researchers wanted to engage politically and started to criticise quantitative methods. In some countries, there were also citizen movements attacking governmental data collection.

3.

Subsequent methodological development of the self-report method has stressed the need to estimate both incidence and prevalence of offending (Thornberry, Krohn, 2000, 41-43). Once you start to measure incidence, the “chronic offender” always reappears. In the original NDR questionnaires, incidence measures were rudimentary or nonexistent.

78

Self-reported crime.book Page 79 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

The third and current phase can be described as the re-emergence of selfreport crime and delinquency studies in the Nordic area. This renaissance was presaged by Finland’s decision, as the sole Nordic country, to participate in the first round of the International Self-Report Delinquency Study (Aromaa, 1994), planting the seeds of further growth. In 1995, both Finland and Sweden established national self-report delinquency indicators based on nationally representative samples of ninth graders (Kivivuori, Salmi, 2005; Svensson, 2006). The Swedish and Finnish national indicator systems were established independently from each other, but they were both influenced by the ISRD. The local Swedish Örebro series (Olofsson, 1971; Ward, 1998) was also revived. And after decades of hibernation, the Norwegians also returned to self-report studies on a national scale during the early 1990s (Pedersen, Wichstroem, 1995; Storvoll et al., 2002; Pape, Falck, 2003). Like all historical phase divisions, the model of three phases in Nordic selfreport research simplifies the full complexity of actual events. The most important exception is the Danish case. In that country, Flemming Balvig and Britta Kyvsgaard conducted important self-report delinquency studies in a Copenhagen suburb in 1979, 1989, 1999 and 2005 (Kyvsgaard, 1992; Balvig, 2006). In the Nordic area, possibly only the Finnish Young Male Crime Survey (Salmi, 2008) reaches further to the past as a currently operating indicator system (more of this below). Very likely there were, during the interim years, important one-off studies. In Finland, for example, Sipilä’s (1982) study, which used a modular multiple-community sample, was important in keeping the tradition alive.

II.

Self-report delinquency indicators in Finland

In what follows, I will describe the major Finnish self-report delinquency studies. The studies will be presented in the order of national importance. I emphasize criminological projects which fulfil the following criteria: (a) the study is based on self-report survey method, (b) most of the questions are about the offences committed by the respondent, and (c) the project entails repeated sweeps4. Three indicator systems fulfil these criteria: the Finnish Self-Report Delinquency Study (FSRD), the International Self4.

This definition also excludes longitudinal surveys, which are practically nonexistent in Finland at least within criminology. Professor Lea Pulkkinen’s Jyväskylä Longitudinal Study of Personality and Social Development is perhaps the exception as its database contains self-report delinquency elements. For more information, consult Pulkkinen et al. (2002).

79

Self-reported crime.book Page 80 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Report Delinquency Study (ISRD) and the Young Male Crime Survey (YMCS). All these indicator systems have been operated in Finland by the National Research Institute of Legal Policy, which is the research arm of the Finnish Ministry of Justice. I will also briefly describe the School Health Survey, which has repeated sweeps, and the Mare Balticum study with a single measurement. These instruments have self-report delinquency items even though most of their content covers other topics (health behaviour and crime victimization). Only the most basic aspects of each system are described in the text below. All survey systems discussed below are based on anonymous responding. In all Finnish delinquency surveys, responding is legally voluntary for the participants. In practice, almost nobody refuses (at least overtly). The Finnish law does not require parental permission. It has been argued that children’s rights include that they can respond to research questions without parental permission (Kuula, 2006). The age at which children may be surveyed without parental consent has been debated. Some experts on research ethics advocate 12 as an important limit.

1.

Basic purpose of self-report delinquency studies

In Finland, the first and primary goal of conducting self-report research is to measure the full extent of crime conducted by adolescents. We know from many sources that the likelihood of an offender becoming known to the police (and therefore his or her entry to the statistics of recorded crime) varies over time. It would therefore be very problematic to rely on official statistics to describe crime trends and patterns. Official statistics are important especially when serious crime is studied, but they need to be supplemented with survey sources such as victim and self-report surveys. Second, self-report studies are regarded as important means of studying the efficacy of social control. By asking the respondent whether he or she became known to the police or to other control instances, we describe the likelihood of detection. Third, self-report indicators provide a basis for international comparisons. This rationale lies behind Finland’s decision to participate in ISRD sweeps. The national FSRD system is largely comparable with Sweden’s similar system (for more detailed discussion, see Kivivuori, 2007a). Fourth, the databases that are created by the self-report indicator systems can be used for basic research. The NRILP conducts such research itself, and makes its data available for university researchers as well5. The 5.

The NRILP withholds a one year embargo before the data are made public.

80

Self-reported crime.book Page 81 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

principle of open access is applied because the indicator funding is ultimately based on the taxpayer’s money, and extensive data utilisation is to be expected and hoped for. The NRILP regards the data of its self-report studies as part of the national research infrastructure.

2.

Finnish Self-Report Delinquency Study (FSRD)

FSRD is the national self-report delinquency indicator. It is targeted at ninth grade students (15-16 year olds). This grade level is the last time when a full age cohort can be reached in a same institution. The survey takes place at the end of the last compulsory grade of the comprehensive school. Responding is anonymous and takes place during school hours under teacher supervision. The FSRD system was created by the NRILP which has also been the main financer of the project. Most sweeps have received additional funding from the Finnish Ministry of Education and one sweep from the Academy of Finland. Presently the FSRD series covers the years 1995 to 2004. The next sweep has been conducted in the spring 2008. The FSRD is based on a nationally representative random sample of schools.

3.

International Self-Report Delinquency Survey (ISRD)

Finland is the only Nordic country which has participated in both sweeps of the ISRD in 1992 and 2006. On both occasions, Finland used a city sample (Helsinki). In 1992 this was a question of funding, and in 2006 Helsinki-based sample was chosen because we already possessed a national indicator system. In the 2006 ISRD measurement, Web survey methodology was used for the first time in Finnish self-report delinquency research. In the Nordic area, the ISRD-2 measurement was co-ordinated and financed by the Scandinavian Research Council for Criminology. The Nordic area report of ISRD-2 was written by a Finnish researcher (Kivivuori, 2007a). In 1992, the ISRD-1 was based on a non-random convenience sample (Aromaa, 1994). The 2006 sweep was based on a random sample of school classes. The sample was stratified by school district.

81

Self-reported crime.book Page 82 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

4.

Young Male Crime Survey (YMCS)

This project is a replication and an extension of the Finnish 1962 draftee study, which was part of the all-Nordic project (see above). We decided to replicate the 1962 study for three reasons. First, the intention was to extend the use of self-report delinquency surveys to slightly older persons (young adults) from the customary mid-adolescence range. Second, the study aimed to probe the possibility of creating a new standardised indicator system targeting the young male population in the draft context. Third, the study explores the difficulties and possibilities of long-term comparison of self-report delinquency. The original 1962 questionnaire contained mainly questions about property crime, while in the 2006 YMCS the offence repertory was much wider, including violence and drugs related behaviours. The new indicator was renamed as the Young Male Crime Survey. In Finland, all men are obligated to attend pre-military screening in the year they turn 18, making a complete cohort available for research at the draft situation. Young men who do not want to go to the army can choose the so-called civilian service, but they too have to attend the pre-military screening. Only females are exempt. The YMCS has targeted all men attending the pre-military screening in Helsinki (there is no sampling within this population). The YMCS has been created and financed by the NRILP. The project is based on the research permit of the Finnish Defence Forces and has enjoyed considerable support from the army staff in terms of practical arrangements.

5.

Other self-report indicators in Finland

Above, I have described self-report delinquency survey projects which have repeated sweeps and in which most of the questionnaire space is devoted to crime/delinquency. As noted earlier, there have been one-off studies using self-report delinquency items and surveys whose main topic has not been crime and delinquency. In 2002/2003, Finland participated in the Mare Balticum youth victimisation survey. That survey devoted most of its attention to crime victimisation, but it included a self-report delinquency scale which was used in Finnish research (Kivivuori, Savolainen, 2003). The MB survey was directed by Professor Frieder Dünkel from the Greifswald University in Germany and included many cities in the Baltic Sea region.

82

Self-reported crime.book Page 83 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

From 2000/2001, the Finnish School Health Survey has contained a short delinquency scale whose questions were adapted from the FSRD. This parallelism has been a source of methodological evaluation and crossvalidation. The SHS is unique in that its huge samples provide local (municipality level) data series. The FSRD and SHS thus complement one another. An interesting feature of the SHS is that the participating municipalities pay for being included in the study. Finland has participated in all sweeps of the European School Survey Project on Alcohol and Other Drugs (ESPAD). This survey system is based on anonymous self-reports and gathers data on alcohol and drugs use (von Hibell et al., 2004). The ESPAD results have been used in the crossvalidation of FSRD and ISRD projects (Kivivuori, 2007a).

6.

Theoretical background

The main aim of Finnish self-report delinquency indicators has been the description of crime trends and control efficacy trends. None of the Finnish indicator systems has been strongly connected to any particular theory. This may reflect the fact that the Finnish history of self-report method has been relatively strongly government-driven as opposed to university driven. Consequently the research interest has been more oriented toward statistical description of trends rather than to theoretical basic research of delinquency causation. Additionally, shortage of funding has excluded the option of longitudinal measurement. The resulting cross-sectional design is not well suited to testing causal hypotheses. The early draftee studies of the 1960s were non-theoretical, even though their general spirit was close to labelling theory. The ISRD-1 was mostly influenced by social control theory, and some of that heritage was transferred to the FSRD questionnaire. Later on, a short self-control scale was introduced to the FSRD to study the interaction of social control (sometimes named social capital) and self-control (Salmi, Kivivuori, 2006). The FSRD also contains questions tapping sources of socialstructural and economic strain. These link the questionnaire to strain theory, even though we lack questions about more social-psychological and interpersonal sources of strain. In the early part of this decade, questions pertaining to the learning of delinquency were inserted (for example, spending time with older males) for the purpose of studying proxy crime phenomena (see section IV-1 below). The FSRD questionnaire also includes questions tapping the cognitive abilities of the respondent, such as grades, and questions about special

83

Self-reported crime.book Page 84 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

education needs. These variables are robust correlates of delinquency and are therefore included in the FSRD.

7.

Methodological considerations

The Finnish self-report delinquency indicators are largely based on methodological work conducted in other countries (for an in-depth discussion of these foundations and their trans-cultural applicability, see Kivivuori, 2007a, 17-34). However, some methodological work has been conducted mainly on two aspects of self-report methodology: questionnaire design and special education inclusion. During the early 1990s, small-scale methodological evaluation of data collection method was undertaken (Honkatukia, 1995). That study mainly compared how questionnaire length influenced the results. An analogous comparison on a much larger scale was conducted using the parallelism of SHS and FSRD (Kivivuori et al., 2000). The SHS contains a short checklist of offences with no follow-ups, while the FSRD contained 18 offences with follow-ups. Judging from these studies, shorter checklists with few offences and no follow-up questions produce higher delinquency estimates. But the question remains, does “the more the better” principle apply to self-report delinquency studies? Follow-up questions are useful in assessing the relative triviality and severity of the acts (Thornberry, Krohn, 2000, 41-43). Very high prevalence levels may partially reflect the inclusion of extremely trivial offences. In the future, computer based survey technology probably offers a means of combining a rapid-sequence offence list with follow-ups emerging for respondents who admit an offence type. The decision to use teachers as data collectors in national systems was influenced by large scale experimental study on this topic conducted in Iceland (Bjarnason, 1995). That study suggested that data collection by teachers does not compromise the validity of the findings if the respondents seal their anonymous answers into an envelope. Use of teachers was also a means of cutting the expenses, because sending research assistants to all participating schools is expensive. However, at least in Finland, the problem of classroom supervision during data collection has been acknowledged and we are currently pondering its significance in the further development of self-report indicators. As a partial corrective, we have conducted all Helsinki-based surveys6 by using outside supervision. The impression is that for some offences outside 6.

ISRD-1 in 1992, MB in 2002, and ISRD-2 in 2006.

84

Self-reported crime.book Page 85 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

supervision produces higher prevalence levels. This again varies by offence types. Moreover, the trends of delinquency seem to be the same in both supervision modes. As shown in Figure 1, indicator systems based on researcher and teacher supervision, respectively, produce similar findings concerning the trend of shoplifting participation. The level/trend distinction is relevant from the point of view of validity assessment.

Figure 1 Lifetime prevalence of shoplifting, % of 15-16-year olds, by supervision mode, Finland

The FSRD system includes students placed on special needs education. Recently, we have analysed in detail how the inclusion/exclusion of SE students influences delinquency estimates. If SE students are excluded, the incidence of delinquency is underestimated. Moreover, if SE students are excluded, the strength of association between delinquency and economic deprivation is underestimated because children from economically deprived families are more likely to receive special needs education (Kivivuori, Salmi, 2006).

III. Main findings: delinquency trends in Finland In this section, the Finnish trends of delinquency are briefly described. The main focus is on the FSRD findings 1995-2004 (Kivivuori, Salmi, 2005). At the time of printing, the sixth sweep of the FSRD (2008) has been

85

Self-reported crime.book Page 86 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

administered, so that in near future the trend information will be extended to 2008. The trends shown by the Finnish ISRD measurements in 1992 and 2006 are highly consistent with the FSRD findings (Salmi, 2007).

1.

Delinquency trend 1995-2004

The most drastic change taking place in the decade between 1995 and 2004 was the radical drop of property offending (Figure 2). Participation in various thefts and property destruction decreased, the most marked drop occurring in the prevalence of shoplifting. Participation in violence-related offences has been more stable. The 2004 sweep indicated decreasing levels of participation in fighting in public places and beating up somebody, but there is no consistent trend. The use of marihuana or hashish increased, a trend that has been corroborated in other youth surveys. However, the upward trend seems to come to a halt. 50

47

46

45

40

40

35

35

33

33 30

30

27

25

22 22

20

15 16

15

14

16 13

10

7

7

9

8

4

5 0

Theft 1995

Destruction of property 1996

Violence 1998

Use of marih.or hash. 2001

2004

Figure 2 Last year prevalence of main offence types, % of 15-16-year olds, Finland (FSRD)

Largely as a result of property crime decrease, the percentage of adolescents who refrained from all crime types grew in number. It seems that the number of “completely law-abiding” adolescents have increased quite significantly. This finding is one of the most interesting (and to some extent, enigmatic) developments in the Finnish and Nordic crime scene. It

86

Self-reported crime.book Page 87 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

has to be interpreted carefully. First, the Finnish finding of increasing conformity is based on 14 offences with identical questions in all five FSRD sweeps. It cannot be ascertained that the abstaining respondents abstain from all forms of crime and deviance. The observation is dependent on the repertory of offences included in the questionnaire. Importantly, computer related offending has been absent from the FSRD. From ISRD-2, we know that it is relatively prevalent. It is therefore possible that the decreasing FSRD trend partially reflects crime displacement (Kivivuori, 2007a, 99100; Salmi, 2007, 29-40). Property crimes, which are often committed in public places, have been at least partially replaced by computer offences committed in private apartments. However, there are multiple factors whose co-occurrence may have resulted in property crime drop among adolescents. These trends are interconnected and overlapping, as summarized in table 1. 앫 Improvement of the general economic situation after the economic slump of the early 1990s 앫 Increasing social control in schools (closer contacts between the school and the police, and increasing CCTV use in school areas) 앫 Increasing control and surveillance of other public spaces such as shopping areas (CCTV, guards) 앫 Changes in opportunity structure: concentration of the retail sector in large malls instead of small shops 앫 Changes in routine activities: adolescents spend increasing time using computers, etc. 앫 General change of attitude: juveniles increasingly condemn criminal activity 앫 Changes in youth culture (eg. decline of graffiti) 앫 Local crime prevention projects, boosted by the national crime prevention project of 1998, may have some effect on adolescent property crime 앫 More efficient police strategy, such as community policing, may deter offenders Table 1 Possible factors related to the general decrease in property crime among Finnish juveniles 1995–2004

At the most general level, the decrease in property crime is likely to reflect the strengthening of both formal and informal social control. As a telling example, half of the FSRD-2004 schools had CCTV surveillance on school premises, a feature that has largely emerged during the observed period 1995–2004. At the same time, the propensity of schools to report offences committed by the students to the police has increased steadily. Such 87

Self-reported crime.book Page 88 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

developments are probably connected with the decrease in theft and destruction of school property (table 1). In the FSRD, crime-related attitudes are measured by items based on neutralization theory. In other words, the respondents are asked if they agree with various excuses and justifications of juvenile crime. The acceptance of such excuses and justifications of juvenile crime has steadily decreased from 1995 to 2004. Juveniles have become more punitive and less tolerant of juvenile crime. These changes in attitude may partially reflect increasing media coverage of rare cases of serious juvenile violence. One important rationale for Finnish self-report delinquency indicators has been the aim of studying the efficacy of social control. Over the period 1995–2004, the likelihood of adolescents becoming known to the police increased in the context of violent offences, shoplifting and destruction of school property. The most important factor explaining this trend is probably the increasing propensity of bystanders, victims and others to report juvenile misbehaviour to the police. Changing control practices of the police, for example community policing, may also have contributed to this trend. Regarding shoplifting, increased likelihood of police contact probably reflects improving technical surveillance and the rising propensity of storeowners and shopkeepers to report shoplifters to the police. School authorities are also more likely to report student offences to the police. The increase in recorded assault offences by juveniles in official statistics during the late 1990s, may, to some extent, be the result of increasing and more effective efforts to control adolescent behaviour. The delinquency trends described above have been similar in Nordic countries. Especially Sweden and Finland manifest nearly identical trends (if not levels). For example, participation in shoplifting, stealing at school and destroying property seem to have decreased in both countries, while participation in violence has been comparatively stable. The percentage of adolescents refraining from all offences has been increasing in both countries (for in-depth discussion of Nordic similarities, see Kivivuori, 2007a, 89-100).

2.

Delinquency of young males in 1962 and 20067

As noted above, Finland participated in the original Nordic Draftee Research Programme with a Helsinki-based self-report survey in 1962. In 2006, the NRILP decided to replicate this study in the pre-military screening of Helsinki, using an expanded questionnaire, and renamed as 7.

This section is based on the research of Venla Salmi (2008).

88

Self-reported crime.book Page 89 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

the Young Male Crime Survey. In what follows, I will describe some of the core comparative findings of the YMCS. All the figures and most of the interpretations are based on research by Venla Salmi (2008), whose work also contains a careful evaluation of the limits and problems in this kind of long-term survey comparison. Three offences decreased in prevalence: workplace theft, buying or selling stolen goods, and theft from a car (Figure 3). The decrease of workplace theft probably reflects changes in opportunity structure: in the 1960’s young people started to work earlier while today’s youths are often students. Similarly, the decrease of buying or selling stolen goods and theft from a care is likely to reflect changes in general economic conditions and opportunity structures. There is a saying in British English about getting something “off the back of a lorry”, meaning that someone has bought something he or she knew was stolen property. My guess is that the high prevalence of “theft from a car” in the early 1960s reflects taking stuff, in a very concrete manner, off the back of lorries with old-fashioned open platforms. Buying stole property and stealing from cars thus reflect the post-war austerity economy in which the conscripts of 1962 had lived all their lives.

Workplace theft

Buying/selling stolen goods

Theft from a car

Source : Salmi, 2008

Figure 3: Decreasing prevalence. Lifetime participation in offences, % of young males in Helsinki 1962 and 2006

89

Self-reported crime.book Page 90 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

There were also offences manifesting increasing prevalence (Figure 4). Drunken disturbance in public places clearly increased (even if there were some technical problems in creating an exact modern replication question). Likewise, drunken driving has increased. These two offences reflect the increasing availability of alcohol in Finnish society. Back in 1962, buying alcohol was heavily controlled. Starting from the late 1960s, a series of policy changes liberated the alcohol policy in Finland in a manner which had instant impact on violent crime (Kivivuori, 2003). Change in the prevalence of drunken driving also reflects the increasing availability of motor vehicles. The increasing prevalence of bicycle theft is difficult to interpret, but it may have something to do with better bikes being more enticing objects of theft.

Public disturbance when drunk

Drunken driving

Bicycle theft

Source : Salmi, 2008

Figure 4: Increasing prevalence. Lifetime participation in offences, % of young males in Helsinki 1962 and 2006

There were also offences whose prevalence remained surprisingly stable. For example, 41 per cent of the 1962 draftees had shoplifted, while the corresponding modern figure is 38 per cent. Lack of change is surprising given the massive changes in retail sector. The change from service-based small businesses to self-service-based supermarkets “should have” resulted in more theft, but possibly the intensification of technical surveillance has successfully replaced earlier types of monitoring. Moreover, the rising living standards may mean that there are fewer young males who are motivated to steal. 90

Self-reported crime.book Page 91 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

The results indicate a slight tendency towards polarization in the variety of offences (Salmi, 2008). The proportion of young men who did not commit any of the surveyed offences almost doubled during the research period. On the other hand, the proportion of males who committed many kinds of offences increased slightly. This suggests that the patterns of young male delinquency may have changed. This is supported by preliminary analyses which suggest that various types of offending are more intimately associated with one another in the present era than five decades ago. Increasing polarization may be accompanied by increasing versatility of deviance in the group of high frequency offenders. Indeed, all major Finnish self-report indicators (Finnish ISRD, FSRD and YMCS) suggest that the Finnish delinquency scene has become more polarised: there is an increasing majority of law-abiding youths and a small (but possibly slightly increasing) minority of versatile offenders. Needless to say, a 44-year time span poses problems for data interpretation. However, the basic idea was that difficulties notwithstanding, this long time range could also be potentially highly relevant for the understanding of contemporary crime patterns. The first YMCS sweep took place in a society which was still largely agricultural in terms of both economy and culture. The second YMCS took place in a post-industrial high-tech society. The results should be interpreted cautiously because of the limited spectrum of offences included in the 1962 study. Today, there are many offences that only science fiction writers could think of back in the 1960s.

IV. Use of self-report research in Finland 1.

Research use

As noted above, the Finnish SR indicators have been relatively atheoretical in their overall aims and design features. However, within the limits of the typical cross-sectional design, they have been put to research use. In what follows, a few examples of SR-indicator-based research are briefly mentioned. In a study of the transitory or episodic nature of delinquency (Kivivuori, 1998), special emphasis was placed on whether and how adolescents themselves conceptualize their delinquency as temporary "phase" in their lives. Phase rhetoric is used by adolescents in dual purpose: as a neutralization technique triggering episodic delinquency, and as a rhetoric justifiying spontaneous desistance..

91

Self-reported crime.book Page 92 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

A study of the link between psychosomatic symptoms and delinquency was inspired by the strain theoretical prediction that delinquency may be a successful means of reducing stress and tension. The fact that delinquency does not have any general power to reduce symptoms suggests that any such effect is situation-specific, as indeed hypothesized by strain theory (Kivivuori, 2000). Later on, the link between delinquency and depression was also studied using school health surveys (Ritakallio et al., 2006). More recently, self-report indicators have been used to study the extent and correlates of crime by proxy. Proxy crime refers to incidents where someone asks or forces another person to commit an offence for him or her, as a proxy offender. The proxy crime research project suggests that some part of juvenile offences (and therefore also the estimates produced by self-report indicators) reflect interpersonal coercion, not autonomous criminal motivation. In other words, some offenders are victims of coercion when committing the crime (Kivivuori, 2007b). The Finnish SR indicators have also been used to study the association between part-time work and delinquency (Kouvonen, Kivivuori, 2002), the link between social capital and delinquency (Salmi, Kivivuori, 2006) and the differences of delinquency between language groups in Finland (Obstbaum, 2006). The FSRD data is stored in the Finnish Social Science Data Archive, and they are available for university researchers who may wish to use them in research or education.

2.

Policy use

National level The most immediate administrative user of self-report delinquency indicators is the Ministry of Justice. For the purpose of this article, I asked a non-random selection of the leading officials and experts of the Ministry to comment on the utility of self-report research. The following discussion is based on their answers. The ministry of justice has used the series in its estimation of crime situation and trends in the country. As described by the Director General of the criminal policy department of the Ministry of Justice, [The FSRD data] are very useful [for administration]. Otherwise we would not have received reliable information on the decrease of adolescent criminal behaviour and on the polarization of youth crime. We have also used the data as a source when we have tried to predict future criminality. (Jarmo Littunen, Director General, Criminal Policy Department, Ministry of Justice.)

92

Self-reported crime.book Page 93 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

The person in charge of planning and co-ordinating local crime prevention efforts in Finland commented on FSRD as follows: The proper planning of crime prevention requires full and reliable information on the patterns and trends of juvenile crime. Crime prevention expertise is impossible without such knowledge of violence and property crime trends. (Hannu Takala, Head of Crime Prevention Unit and Executive Secretary of the National Council for Crime Prevention). Mr. Takala continued by noting that in a historical situation characterised by decreasing property crime, local crime prevention may falsely take the credit for the decrease which is actually a general societal pattern. The trend information produced by FSRD has also influenced the planning of penal reforms concerning young offenders. Especially the youth crime committee (2001-2003) and youth crime working group (2004) used selfreport findings in estimating the “full picture” of youth crime8. The work of these committees led to the introduction of the so-called juvenile punishment on a national basis. The FSRD was also used in the planning of the Finnish national violence reduction programme. Some experts also noted that valid knowledge is also useful because it pacifies the occasionally excessive worry about youth crime. In this respect, it is important to acknowledge the limits of self-report research: it does not tap into the most serious kind of violence. It is therefore possible that mass participation in delinquency is stable or decreasing, while the incidence of the most serious youth crime increases. This was actually the case in Finland during the late 1990s, when the delinquency scene became polarized (less overall delinquency, but increase of juvenile homicide). Especially the rising percentage of completely law-abiding youths, a fact revealed by the FSRD, probably had some influence on the policy reactions and public understanding of the delinquency trend. The results of FSRD sweeps have been also reported in the national media. However, some of the Ministry of Justice experts whose comments were solicited for this article doubted whether self-report findings can influence the general portrayal of juvenile crime in the media. The media offer lots of different and even contradicting information based on uncritical use of official statistics. Official statistics are reported with much greater frequency and they also provide the media with more “exciting news” than survey research. As a consequence, public messages about crime trends may sometimes be inconsistent.9 8. 9.

Comment by Janne Kanerva, Legislative Counsellor, Law Drafting Department, Ministry of Justice. This chapter is based on user comments by Ministerial Advisers Olavi Kaukonen and Aarne Kinnunen.

93

Self-reported crime.book Page 94 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Regional and local level The FSRD sample is not designed to produce valid estimates of delinquency at the lowest municipal level. The system has therefore few applications in the local level. However, the participating schools have used the study as a way of assessing the delinquency level of the school when compared with the national average for schools in similar communities (rural, suburban, urban). The School Health Survey has a very large sample including most municipalities in Finland. The SHS contains a short self-report delinquency scale adapted from the FSRD, and its results are available for the municipalities who participate in the study. The system is therefore unique in that it produces municipality-level time series. It is probable that many local authorities find the SHS data useful in monitoring the health and safety situation in their own area. There have been some efforts to use self-report delinquency research in the evaluation of local crime prevention projects. Currently, the Ministry of Justice is evaluating a local youth crime prevention programme by using the self-report method (Ryynänen, 2007). The project under evaluation is based on the notion that some juveniles offend because they have a mistaken belief that offending is very prevalent among youth. Consequently, educating young people about the true prevalence of offending should decrease delinquency. Interestingly, in this causal paradigm the self-report method can serve in the dual function of evaluating the intervention effect and producing the content of the intervention (true prevalence information).

V.

Some thoughts on problems and challenges

To conclude this chapter, I will raise some questions which have worried self-report researchers in Finland. Threats to external validity of school-based criminology One of the traditional limitations of the self-report method is that the most serious kind of delinquency lies outside its application domain. The serious offenders are relatively few and are not “caught” by studies based on sampling. When schools are used as research locations, the most frequent offenders tend to be excluded (Cernkovich et al., 1985). One of the reasons why we decided to use pre-military screening as the research site of the Young Male Crime Survey was the perceived need to conduct self-report studies outside the ubiquitous school context. 94

Self-reported crime.book Page 95 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

A related aspect is the increasing use of special needs education within the school systems. The net of special education (SE) classes widens: at least in Finland, more and more students receive special education. The FSRD takes care to include the SE classes to the sample, but we know that the higher the delinquent involvement, the greater the likelihood of absence from the research population (Kivivuori, Salmi 2006). The SE arrangements are different in various countries, posing problems for international comparisons. The problem of immigrant populations Immigrant minors may have language problems, or be absent from the regular school classes. Immigrants may have comparatively low trust towards government personnel such as researchers (for a more detailed discussion on the possible cultural variation in the validity of self-report method, see Kivivuori, 2007a, 27-31). Can adults be studied using self-report method? The rationale for establishing the YMCS, which targets 18 year old young males, was to extend the use of self-report indicators to slightly older age groups than the typical mid-adolescence age range. In the future, the prospect of studying adults by self-report surveys might be considered in Finland. A systematic international overview of prior and currently operating adult self-report studies would be very useful. Keeping apace with new forms of delinquency One challenge is related to the fact that forms of delinquent behaviour change. Old offences are committed using new technologies (for example, threats by e-mail or SMS message). Entirely new offences are created (copyright offences in the Internet). It may be coming increasingly difficult to keep track of the novel and versatile manifestations of juvenile offending. Need for interdisciplinary approaches When talking about the theoretical background or inspiration of an indicator, sociologists (like me) naturally think of the classical sociological canon: strain, control, learning and labelling theories. However, there is no a priori reason why the explanatory constructs used in self-report research should be limited to these approaches. Disciplinary blinders can obstruct the advancement of science (Laub, 2006). Lately, scales measuring selfcontrol have appeared in the questionnaires, which is a positive 95

Self-reported crime.book Page 96 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

development (and which initially raised some eyebrows in Finland as a breach of sociological hegemony). Additionally, the role of individual cognitive characteristics such as reading skills and general cognitive capacity are variables that might be inserted to self-report questionnaires. These individual-level factors could be added to survey instruments while keeping the traditional sociological variables, a fertile mix for probing both main and interaction effects. During the next decade, it is highly probable that the Finnish self-report instruments will shift from paper and pencil technique to computer-based data collection method. In this transition phase, the use of researcher supervision during data collection will probably also be re-examined from both scientific and economic points of view. During the transition phase, more methodological work is called for. It would also be helpful, in terms of instrument harmonization and measurement synchronization, if the future prospects and directions of European and international self-report cooperation would be clarified in the near future.

References ANDENAES J., SVERI K., HAUGE R., 1960, Kriminalitetshyp-pigheten hos ustraffede I. Norsk undersoekelse, Nordisk Tidskrift for Kriminalvidenskab, 48, 97•112. ANTTILA I., JAAKKOLA R., 1966, Unrecorded criminality in Finland. Kriminologinen tutkimuslaitos A:2, Helsinki. AROMAA K., 1994, Self-reported delinquency in Helsinki, Finland 1992 in JUNGER-TAS J., TERLOUW G.-J., KLEIN M. W. (Eds.), Delinquent Behavior among Young People in the Western World. First Results of the International Self-Report Delinquency Study, Kugler, Amsterdam / NewYork, 14-41. BALVIG F. 2006, Den ungdom! Om den stadigt mere omsiggribende lovlydighet blandt unge I Danmark, Glostrup, Det Kriminalpraeventive råd. BJARNASON T., 1995, Administration mode bias in school survey on alcohol, tobacco and illicit drug use, Addiction, 90, 555-559. CERNKOVICH S.A., Giordano P.C., Pugh M.D., 1985, Chronic Offenders: The Missing Cases in Self-Report Delinquency Research, Journal of Criminal Law & Criminology, 76,3, 705-532. CHRISTIE N., ANDENAES J., SKIRBEKK S., 1965, A Study of Self-Reported Crime, Scandinavian Studies in Criminology, 1, 86-116. CHRISTIE N., 1966 [1964], Hvorfor blif unge mennesker forbrytere?, in ANDENAES J., BRATHOLM A., CHRISTIE N., Kriminalitet og samfunn, Bergen, J. W. Eides Boktrykkeri. 96

Self-reported crime.book Page 97 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

CHRISTIE N., 1975, Hvor tett ett samfunn?, Oslo, Christian Ejler’s forlag. GREVE V., 1972, Kriminalitet som normalitet, Koebenhavn, Koebenhavn Juristforbundets Forlag. HIBELL B. VON, ANDERSSON B., SALME A., BALAKIREVA O., KOKKEVI A., MORGAN M., 2004, The Espad Report 2003. Alcohol and Other Drug Use Among Students in 35 European Countries, Stockholm, The Swedish Council for Information on Alcohol and Other Dugs (CAN). HONKATUKIA P., 1995, Itse ilmoitetun rikollisuuden mittaaminen, Helsinki Oikeuspoliittisen tutkimuslaitoksen tiedonantoja 21. KIVIVUORI J., 1998, Delinquent Phases. The Case of Temporarily Intensified Shoplifting Behaviour, British Journal of Criminology, 38, 4, 663-680. KIVIVUORI J., 2000, Delinquent Behaviour, Psychosomatic Symptoms and the Idea of “Healthy Delinquency”, Journal of Scandinavian Studies in Criminology and Crime Prevention, 1, 2, 121-139. KIVIVUORI J., 2003, Sudden Increase of Homicide in Early 1970s Finland, Journal of Scandinavian Studies in Criminology and Crime Prevention, 3, 1, 6-21. KIVIVUORI J., 2007a, Delinquent Behaviour in Nordic Capital Cities, Helsinki, National Research Institute of Legal Policy, Publication 227, Helsinki. KIVIVUORI, J., 2007b, Crime by Proxy. Coercion and Altruism in Adolescent Shoplifting, British Journal of Criminology,, 47, 5, 817-833. KIVIVUORI J., KARVONEN S., RIMPELÄ M., 2001, Nuorten rikoskäyttäytymisen yleisyys kyselyjen valossa, Helsinki, Oikeuspoliittisen tutkimuslaitoksen tutkimustiedonantoja 54. KIVIVUORI J., SALMI V., 2005, Nuorten rikoskäyttäytyminen 1995-2004, Helsinki, National Research Institute of Legal Policy, Publication no. 214. English Summary: Trends of Self-Reported Juvenile Delinquency in Finland, 1995-2004. [Online: http://www.optula.om.fi/uploads/ 9mtrckcfo.pdf]. KIVIVUORI J., SALMI V., 2006, The problem of missing students in school-based delinquency surveys: special education and absenteeism. Paper presented at the 6th annual conference of the European Society of Criminology, Tübingen, Germany, August 26-29. KIVIVUORI J., SAVOLAINEN J., 2003, Helsingin nuoret rikosten uhreina ja tekijöinä, Helsinki, Oikeuspoliittisen tutkimuslaitoksen julkaisuja 204. English summary: Violent Victimisation and Delinquent Behaviour among Helsinki Adolescents, 2002 [Online: http://www.optula.om.fi/ uploads/qv1jut75.pdf]. KOUVONEN A., KIVIVUORI J., 2002, Part-Time Jobs, Delinquency and Victimization Among Finnish Adolescents, Journal of Scandinavian Studies in Criminology and Crime Prevention, 2, 191-212. 97

Self-reported crime.book Page 98 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

KUULA A., 2006, Aineistojen hankinta, käyttö ja säilytys, Tampere, Vastapaino. KYVSGAARD, B., Ny ungdom? Om familie, skole, fritid, lovlydighet og kriminalitet, Jurist- og Oegonomsforbundets Forlag, Denmark, 1992. LAUB J.H., 2006, Edwin H. Sutherland and the Michael-Adler Report: Searching for the Soul of Criminology Seventy Years Later, Criminology, 44, 2, 235-257. OBSTBAUM Y., 2006, Brottslighet bland finskspråkiga och svenskspråkiga ungdomar, Rättspolitiska forskningsinstitutets forskningsmeddelanden 69. English summary: Juvenile Delinquency among the Swedish Speaking Minority of Finland. OLOFSSON B., 1971, Vad var det vi sa! Om kriminellt och konformt be-teende bland skolpojkar, Stockholm, Utbildiningsförlaget. PAPE H., FALCK S., 2003, Ungdomskriminalitet. Et fenomen i endring?, Tidsskrift for Ungdomsforskning, 3,2, 99-111. PEDERSEN W., WICHSTROEM L,. 1995, Patterns of Delinquency in Norwegian Adolescents, British Journal of Criminology, 35, 4, 543-562. PULKKINEN L., NYGREN H., KOKKO K., 2002, Successful Development: Childhood Antecedents of Adaptive Psychosocial Functioning, Journal of Adult Development, 9, 4, 251-264. RITAKALLIO M., KALTIALA-HEINO R., KIVIVUORI J., LUUKKAALA T., RIMPELÄ M., 2006, Delinquency and the profile of offences among depressed and non-depressed adolescents, Criminal Behaviour and Mental Health, 16, 100-110. RYYNÄNEN H., 2007, Harhojen purkaminen vähentää nuorten rikoskäyttäytymistä, Haaste, 7, 1, 16-17. SALMI V., KIVIVUORI J., 2006, The Association between Social Capital and Juvenile Crime: The Role of Individual and Structural Factors, European Journal of Criminology, 3, 2, 123-148. SALMI V., 2006, Helsingin nuoret rikosten tekijöinä 2006, Oikeuspoliittisen tutkimuslaitoksen tutkimustiedonantoja 73. English summary: Selfreported delinquency among Helsinki adolescents, 2007. SALMI V., 2008, Nuorten miesten rikoskäyttäytyminen 1962 ja 2006. Rikollisuuden kehitys itse ilmoitetun rikollisuuden kyselyjen valossa, Helsinki, Oikeuspoliittisen tutkimuslaitoksen julkaisuja 235. SALMI V., KIVIVUORI J., 2006, The Association between Social Capital and Juvenile Crime, European Journal of Criminology, 3, 2, 123-148. SIPILÄ, J., 1982, Nuorten poikkeava käyttäytyminen ja yhteisön rakenne, Jyväskylä, Jyväskylän yliopiston yhteiskuntapolitiikan laitoksen tutkimuksia A1. STANGELAND P., HAUGE R., 1974, Nyanser I grått. En undersoekelse av selfrapportert kriminalitet blant norsk ungdom, Oslo, Universitetsforlaget. 98

Self-reported crime.book Page 99 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency Surveys in Finland

STORVOLL E., WICHSTROEM L., KOLSTAD A., PAPE H., 2002, Structure or conduct problems in adolescence, Scandinavian Journal of Psychology, 43, 81-91. SVENSSON R., 2006, Ungdomar och brott åren 1995-2005, Stockholm, Brottsföre-byggande rådet, rapport 7. THORNBERRY T.P., KROHN M.D., 2000, The self-report method for measuring delinquency and crime, in DUFFEE D. (Ed.)., Criminal Justice 2000 Volume 4: Measurement and analysis of crime and justice, Washington, DC, 33-83. WARD M., 1998, Barn & Brott av vår tid? Självdeklarerad Ungdomsbrottslighet 1971 och 1996. En jämförelse utifrån Örebroprojektets data, Stockholms Universitet, Kriminologiska Institutionen.

99

100

a) b)

Seven sweeps 2000-2007

Local (Helsinki)

General self-report delinquency survey National

Local (Helsinki)

General self-report delinquency survey

Health

Geographical coverage National

Main focus of the study General self-report delinquency survey

2007: 50,450 (14-16 year olds)

Cross-sectional studies involving two sweeps in Helsinki. In 1992, group self-completion using anonymous paper & pencil questionnaires (researcher supervision). In 2006, group self-completion using anonymous Internet survey solution (researcher supervision). Response rate 84 % (1992) and 87 % (2006). Cross-sectional studies involving two sweeps in the premilitary screening of Helsinki. Group self-completion in draft situation, anonymous paper & pencil questionnaires. Researcher supervision. Response rate: 79 % of total population (18 year olds living in Helsinki). Cross-sectional annual surveys. The country is divided into two segments with alternating measurement years. Group self-completion in classrooms using anonymous paper & pencil questionnaires. Teacher supervision. Very large total samples. Response rate 81% (2003).

Cross-sectional study involving five sweeps of data collection 1995-2004 (next sweep: 2008). Group selfcompletion in classrooms using anonymous paper & pencil questionnaires. Teacher supervision. Response rate 85 % (2004).

2004: 5,142 (15-16 year olds)

1992: 691 (15-16 year olds) 2006: 1,354 (13-16 year olds) 2006: 2,190 males (18 year olds)

Method and response rate

Sample

These indicators are operated by the National Research Institute of Legal Policy. The SHS is operated by the National Research and Development Centre for Welfare and Health (Stakes), and basic results are published in its web pages.

School Health Survey (SHS)b

Date(s) of survey Finnish Self-Report Five sweeps 1995-2008 Delinquency Studya (FSRD) Kivivuori Salmi 2005 International Self- Two sweeps 1992 and 2006 Report Delinquency Studya (ISRD) Kivivuori 2007a Salmi 2007 Young Male Crime Two sweeps Surveya (YMCS) 1962 and 2006 Salmi 2008

Indicator system

Appendix 1 Self-report delinquency survey indicators with repeated sweeps, Finland

Self-reported crime.book Page 100 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Self-reported crime.book Page 101 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Surveys in France1 Cécile Carra

The purpose of this report is to take stock of self-reported delinquency surveys in France2. A bibliographical search on the basis of various French terms usually employed in such surveys – self-reported, -declared or -revealed delinquency (hyphenated and unhyphenated), on the one hand, and delinquency, crime, violence, deviancy, drugs, and at-risk behaviour on the other – shows how rare that type of surveys are in France. Regularly repeated studies are even more exceptional. These mostly deal with three broad subjects: juvenile delinquency, drug use and school violence, and are more or less directly inspired by that methodological option. In the first part of this report, surveys are presented according to their subject, situated in different conceptual frameworks and involving very different stakes, ranging from social control to public health, and including educational practices and the functioning of schools. Part two attempts to identify the uses to which they were put.

I.

Presentation of self-reported delinquency surveys

1.

Juvenile delinquency and safety

The first study claiming to be a self-reported delinquency survey in France was conducted by S. Roché, a director of research at the CNRS3, and led to the publication of a book in 2001. Conducted within the Institute of Political Studies of Grenoble, this survey was financed (in decreasing order) by the MAIF, a large mutual insurance company, the Institut des Hautes Études pour la Sécurité Intérieure, a research centre of the Ministry of the Interior, the Ministry of Justice, particularly the Protection judiciaire de la jeunesse (PJJ) its Juveniles Protection Directorate, the Centre de prospective de la Gendarmerie Nationale, a think tank for this particular

1. 2. 3.

Translation Helen Arnold, revised by Renée Zauberman. I wish to thank all of the people who helped me to collect the data for this study. Roché, 2000. CNRS stands for Centre National de la Recherche Scientifique, the largest French government-funded research organization.

101

Self-reported crime.book Page 102 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

police force, and the Semitag, the public transportation system of the Grenoble metropolitan area. The objective was twofold: to achieve a more accurate picture of delinquent activities among youths and to contribute to making local responses to crime and unsafety more appropriate. Two French metropolitan areas were investigated: the Grenoble and Saint-Étienne areas. 2,288 students, selected using a randomized sampling procedure, were questioned at their school between April and May 1999, with the number of respondents proportionate to the number of 13 to 19 year-olds registered in each of the schools studied. Whereas the purported subject of the study was juvenile delinquency, the information collected actually covered youths still attending school. The questionnaire was derived from the International Self Report Delinquency Project. It was enriched by questions about employment status (unemployed or employed), about the perception of institutions (the police, gendarmerie, mayor…), about values, and about victimisation. The section on delinquency asks questions about a series of acts ranging from simple vandalism to physical violence. For each act that the respondent reported to have committed, the survey asks further questions to determine specifics (age at first time, age at last time, location, accomplices, reaction of social circle…). Most of the questionnaire was administered with assistance, face-to-face, in a room provided by the school (usually the infirmary). In order to preserve the confidentiality of responses for behaviour involving violence or drugs, the corresponding parts were selfadministered: the respondent read and circled the responses him/herself. This design ensured that responses were anonymous, since there could be no visual recognition of the respondent by the interviewer, who was, moreover, unaware of the respondents’ identity, since the latter were referred by members of the school staff. At the end of each interview the questionnaire was placed in a sealed envelope and handed over, at the end of the day, to the person in charge of the study for that location. In 2004, another report on a survey supervised by S. Roché designed to study self-reported delinquency,, was produced for the Juveniles Protection Directorate (PJJ, Ministry of Justice). On the basis of selfreported delinquency surveys conducted on representative samples of youths attending school in the Grenoble area in 1999 (n = 1,300) and 2003 (n = 1,600) and of youths under judicial custody (n = 93) committed by the PJJ and living in a facility in the Isère département, S. Roché aims at comparing trends in delinquency between 1999 and 2003 and also at comparing the PJJ and school populations. The survey of the overall population was conducted using the same procedure in 1999 and 2003. Interviews lasted an average of 1½ hour for a complete questionnaire. 102

Self-reported crime.book Page 103 Thursday, April 23, 2009 8:02 AM

4

Main Findings of the First Self-Reported Delinquency Survey in France (2000)4 Generally speaking, most acts are committed by boys, by youths who play truancy, cheat in public transportation, whose parents do not oversee how they spend their time, and to a lesser extent, who come from a disadvantaged home. They are from larger families. The most active are also those who think that committing an offense is not very serious: visibly, beliefs are of some importance. The environment counts as well: those offending most frequently are from the suburbs (“les banlieues”), in neighbourhoods where anti-social behaviour and disorders are most marked. The acts committed change with age. At 13, vandalism is 2.7 times as frequent as thefts. At 19, vandalism is less frequent than theft, with 0.6 acts for one theft. In short, between ages 13 and 19, girls and boys cease to act unruly, and concentrate on acts that bring in objects and partake in their search for pleasure. Girls play a non-negligible role in reported acts, contrary to popular opinion, although it is true that boys are more active, and that girls definitely report fewer of the most violent acts (rackets, assault). A large proportion of offenses are committed by a small percentage of 13-19 year-olds. Irrespective of the type of behaviour (vandalism, theft or assault), the most active 5% of youths account for between 50 and 60% of all acts perpetrated. This corroborates the “theory of the five per cent”. The survey shows which offenses are committed first, in the course of a lifetime. They are the easiest to achieve (shoplifting), in places where practically no-one is in charge (vandalising public places). On the whole, delinquency possibly began at an earlier age at the time of this survey than 5 years earlier. The acts most frequently detected occur in places with some human presence (in buses and stores), those which make personal victims (assaults) or affect valuable personal objects (home or car). Between 15 and 20% of delinquents were caught in the act by the police. The least often detected acts are therefore those with no unwilling individual victim, or no victim at all (illegal consumption or dealing), or again, those involving vandalism against public or semi-collective places (building entrances, squares): 0.5 to 5% of perpetrators are detected.

4.

Excerpt from the author’s summary.

103

Self-reported crime.book Page 104 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

A strong correlation is found between truancy, cheating in buses and delinquency: a "no-rules" lifestyle is gradually building up: I notice that my environment is deteriorated, that I can leave school, that I can travel without a ticket. In fact, 4 times as many youths behaving along these lines carry a weapon. Unguarded premises facilitate the generalization of thefts and assaults.

2.

Drug abuse and prevention

Surveys on health seem to have introduced questions on illegal behaviour at an earlier date. The Rapport sur la mise en place du dispositif d’enquêtes en population générale sur l’usage de drogues (Report on the development of the general population drug abuse survey programme) (Beck, 1998) mentions surveys since the mid-1980s. Since the publication of this report, there have been three major surveys with a sound methodology, a large-sized sample and repeated administration. These are the ESCAPAD, ESPAD and Baromètre Santé (Health Barometer) surveys. However, it should be pointed out that none of these claim to be a self-reported delinquency survey. Behaviour they capture greatly exceed the domain of delinquency, and encompass all conducts involving a risk for individual health (smoking, drinking, tranquillizers, and so on). These surveys are definitely positioned in a public health perspective. ESCAPAD ESCAPAD – Enquête sur la Santé et les Consommations lors de l’A.P.D. (Survey on Health and Substance Consumption during the call to Preparation for Defence) – is a scheme for questioning all youths who go through their military preparation day (participation in which is practically compulsory), in all centres used by the National Military Service (DSN) on a given day (several weeks in the DOM and COM5) since the year 2000. This survey, repeated yearly, makes it possible to measure levels of consumption of a dozen legal and illegal substances: alcohol, tobacco, cannabis, psychoactive medications and other illegal drugs. It crosstabulates these figures for consumption with a series of indicators, and most importantly with socio-demographic, geographic, educational and behavioural indicators. The goal of this cross-sectional study is to provide accurate data for a specific age group and to point up trends in the habits

5.

Départements et Collectivités d’Outre-mer (Overseas départements and communities).

104

Self-reported crime.book Page 105 Thursday, April 23, 2009 8:02 AM

of adolescents and young adults. The population surveyed is therefore composed of French national residents aged 17 to 18. Paper and pencil questionnaires are self-administered and anonymous. The surveys are nationwide and are also representative at the regional or even the Département level in 35 cases. Scientific management and funding of these surveys are provided by the Observatoire Français des drogues et des toxicomanies (the French monitoring centre for drugs and drug addiction)6. These surveys are implemented by the OFDT in partnership with the DSN. The ESCAPAD survey has been declared appropriate by the National Council on Statistics (CNIS), has received the general interest label by the Labelling Committee, and has been approved by the CNIL (Commission nationale informatique et libertés, the French data protection commission). In 2005, ESCAPAD investigated a sample of 29,393 youths, of whom 50.9% were boys and 49.2% girls, aged 17, in metropolitan France, and close to 4,000 aged 17-18 in the DOM and COM. The report on the 2005 ESCAPAD survey (Legleye, Beck, Spilka, Le Nezet, 2007) shows that new questions were introduced so as to improve the description of the contexts for the most recent consumption of alcohol and cannabis, in an attempt to evaluate risk-taking such as occasional excessive consumption or driving following the consumption of psychoactive substances. This latest edition of ESCAPAD also yields information on the nature of the psychoactive drugs consumed by adolescents. This report represents the national analysis of the survey, the results of which are also analyzed on a regional basis, like the previous ones. Only the main findings of the chapter on cannabis of the 2007 report (see above) are reported here, since this is the main illegal drug habitually taken. 7

Findings on cannabis7 In 2005, close to half of 17 year-olds claimed to have smoked cannabis at some time and four out of ten had smoked it during the last twelve months. During the last thirty days, over one fourth had smoked it at least once, over one out of ten at least ten times and one out of twenty daily. The difference between boys and girls is considerable and increases with frequency, although less so than for drinking alcohol,

6. 7.

The OFDT also finances specific surveys such as the one conducted by Ballion (1999). Excerpt from the authors’ summary.

105

Self-reported crime.book Page 106 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

drunkenness and the other illegal drugs. On the average, experimenting took place at the very beginning of the fifteenth year, at very much the same time as they first got drunk. … In close to half of instances, overall, those who had smoked during the last thirty days claimed to have smoked one joint or less, and seven out of ten two joints or less. The most frequent consumers, regular or daily, generally smoke much more than the others: close to three regular smokers out of ten claimed five joints or more for the last time. Analysis of questions about the context of last consumption shows that it very often takes place in a group, outdoors or in a private home. Just as for alcohol and tobacco users, sociability turns out to be a fundamental factor in interpreting use, as shown by the fact that practically all consumption took place in the presence of friends. … The socio-economic profile of cannabis users seems similar to that of tobacco and alcohol users. Regular use of cannabis is more frequent among youths who had been left back at school or were presently in a vocational training program, or who had left the school system. It was also more frequent among youths who did not live with their family or whose parents did not live together. Last, it was clearly correlated with an economically privileged family environment. For the overall population, experimenting with and using cannabis have increased constantly and regularly since the early 1990s. However, a downward trend was found among 17 year-olds between 2002 and 2003, with a decline in experimenting and a degree of stability in the level of regular use, a trend which remains to be corroborated by later surveys. At the same time, experimenting seems to occur at a slightly earlier age, going from age 15.3 in 2000 to 15.1 in 2005. ESPAD ESPAD – European School Survey Project on Alcohol and Other Drugs – is a cross-sectional study in schools, conducted concomitantly in some thirty European countries on the basis of a shared methodology and questionnaire (copyrighted). The questionnaire focuses on use of, attitudes, and opinions about psychoactive substances. The questions show the prevalence of various sorts of consumption over a lifetime, the last twelve months and the last thirty days. This survey is coordinated by the CAN (the Swedish Council for Information on Alcohol and other Drugs). It is repeated every four years. Its repetition makes it possible to follow the evolution of consumption and its context from one generation to the next. Its enrichment from one version to the next provides specifics

106

Self-reported crime.book Page 107 Thursday, April 23, 2009 8:02 AM

on some points whose importance was uncovered in the previous analyses; moreover, its integration in a European framework is a guarantee of study international comparisons; the methodology used – a paper and pencil questionnaire filled out directly, and guaranteed anonymity – is internationally recognized and makes for reliable data collection. This survey articulates with a research project developed by the French Santé de l’Adolescent (Adolescent Health) team (first headed by F. Davidson, then by M. Choquet) of the INSERM (Institut National de la Santé et de la Recherche Médicale – National Institute of Health and Medical Research), which has been undertaking epidemiological surveys on the health of youth in France, including the consumption of psychoactive substances, since 19718. A first survey was conducted in 1995, with twenty-six participant countries; France did not participate, but the findings of the 1993 national survey9, obtained using the same type of methodology, were included in the ESPAD 95 report. The second edition took place in 1999, with thirty participant countries. France participated fully in the ESPAD 99 survey, and in the subsequent ones. The 2003 survey was done in 35 countries (including 23 members of the European Union). In France, the various sweeps are financed by the OFDT and the Ministry of Education and Research. Scientific responsibility is shared between the INSERM (M. Choquet and S. Ledoux) and the OFDT (S. Legleye) for the 1999 and 2003 sweeps. The OFDT has exclusive responsibility for the 2007 sweep. Like the previous sweep, the ESPAD 2003 survey sample is stratified according to the following criteria: type of school (junior high school, vocational, or general and technological high school), sector (public or private), located in a ZEP (Educational Priority Zone) or not, and located in a rural or urban district (commune). It is composed of 16,000 students aged 12 to 18 attending one of 400 secondary schools (junior or senior high schools in the public and private sector). Like in the previous sweep, the method used was the self-administered questionnaire completed anonymously by students in the classroom without the presence of the principal or of any teacher, to ensure confidentiality in responses. The questionnaire covers the student's leisure activities, health, behaviour, and use of psychoactive substances. Other themes are broached by ESPAD

8.

9.

… with the participation, in some cases, of other agencies, and sometimes at their initiative, as in the case of the survey that generated the report by M. Choquet et al. (1998), solicited by the aforementioned PJJ and funded by the MILDT, Mission Interministérielle de lutte contre les drogues et la toxicomanie (Mission for the Fight against Drugs and Drugs addiction). Choquet, Ledoux, 1994.

107

Self-reported crime.book Page 108 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

2003, including school experience, family life, and violence. The ESPAD 2007 survey has now been completed. 10

Main Findings of the ESPAD 2003 Survey10 앫 Experimenting with cannabis, the main illegal substance used, progresses rapidly between the ages of 12 and 18. Use of cannabis, very low between ages 12 and 13, rises considerably starting at age 14, with half of girls and two thirds of boys affected at age 18. 앫 Illegal substances other than cannabis are rarely consumed. Levels of consumption for entire life are very low: the percentages are always below 5%, except for inhalable substances (glues, solvents) and hallucinogenic mushrooms for 16-17 year-old boys. There is relatively little increase in these figures with age, and most of those who tried one of those substances did not repeat the experience. 앫 Regular consumption of cannabis (defined as at least 10 times over the year) seems to be exceptional before age 15, with less than 1% of students involved; it progresses considerably thereafter for boys (from 1% at age 14 to 21% at age18), and to a lesser extent for girls (from 1% to 7%). 앫 Trend in cannabis consumption over a 10-year period (1993-2003): experimenting and repeated use are on the rise since 1993, for girls as well as boys. Experimentation and repeated consumption rates have been multiplied by at least two for all age groups (with the exception of repeated consumption at ages 12-13, which remains exceptional). In ten years, repeated consumption in 14-15 year-old boys rose from 1.7% to 5.8% and for 16-17 year-olds, from 7.2% to 21.4%. Increments were particularly high between 1993 and 1999; they then continued, in boys. 앫 International comparisons: the proportion of students in France who claim to have used cannabis is almost twice that of the average for all countries studied (38% versus 21%); however, the proportion of French pupils claiming to have used other illegal drugs is similar (7% versus 6%).

10. Choquet, Beck, Hassler, Spilka, Morin, Legleye, 2004 and the ESPAD internet site [http://www.espad.org/]. Although ESPAD covers psychoactive substances in general, only the main findings on the consumption of illegal substances are presented here.

108

Self-reported crime.book Page 109 Thursday, April 23, 2009 8:02 AM

Baromètre Santé The Baromètre Santé is a national telephone survey measuring the state of knowledge, opinions, attitudes and behaviour of the French in the field of health (descriptive epidemiology) at time t. It covers people aged 12 to 75. Its repetition allows follow-up on trends for each topic studied: in all, over a dozen health-related subjects, and especially at-risk consumption (drinking, smoking, drug abuse) and acts of physical violence. The CFES (Comité Français d’Éducation pour la Santé: French Committee for Health Education) decided to create the Baromètre Santé in collaboration with the main national agencies in the field of health issues: the Caisse nationale de l’assurance maladie des travailleurs salariés (CNAMTS: national health insurance for salaried workers), the Ministry of Social Affairs, Health and Urban Affairs (General Directorate for Health), Délégation Interministérielle à la lutte contre les drogues et la toxicomanie, (Delegation for the fight against drugs and drug abuse) and the Mutualité sociale agricole (MSA: farmers’ mutual fund). Since its inception in replacement of the CFES, the INPES (Institut National de Prévention et d’Éducation pour la Santé: National Institute for Prevention and Health Education) manages the scientific and technical aspects of these surveys. Five such surveys covering the overall population have been conducted since 1992. The technique used is random sampling, two-stages sampling with unequal weights: random selection of households, followed by random selection of one eligible individual per household, except for the 2005 survey, when the person with the closest coming birthday was chosen. Samples were weighted on the basis of data from the INSEE (Institut national de la Statistique et des Études Économiques: National Institute for Statistics and Economic Studies) census to achieve representativity based on the following criteria: age, sex, size of the urban area and region. The questionnaire is administered by telephone. The interviewers and team leaders use the CATI (Computer-Assisted Telephone Interviewing) system. The latest sweep took place in 2005, in partnership with the CNAMTS, INSERM, the Ministry of Health and Solidarity and the OFDT. As in the previous ones, the sample was constructed by random selection from the sampling frame, which is to say the list of telephone numbers used for contacting households. 30,514 persons were interviewed, from a combination of households with a landline and others that were 'mobile– only'. Irrespective of the type of telephone, a single person from each household participated in the survey. Length of the interview ranged from an average of seventeen minutes on mobile phones to an average of forty minutes for calls on landlines. The construction of five regional

109

Self-reported crime.book Page 110 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

oversamples since 1999 makes regional use of the survey possible, in connection with regional health programs. The next sweep is planned for 2010. Main Findings of the 2005 Baromètre Santé Survey Illegal drugs In 2005, one third (32.4%) of people aged 15 to 64 claimed to have taken at least one illegal drug in the course of their lifetime. Three out of ten (30.6%) said they had used cannabis in the course of their lifetime and slightly over four out of ten (43.1%) had had an opportunity to take it. These findings confirm the specific place of cannabis among the illegal drugs found in France since the beginning of the 1990s: it is the most available and the most often-tried illegal drug. Experimenting is more exceptional for the other illegal drugs: less than 4% of people aged 15-64 had tried poppers, the most frequently used substance after cannabis. Consumption during the year did not exceed 1% for each drug. As a rule, the prevalence of consumption of illegal substances is higher for men than for women, and for people under 30. In terms of trends between 2000 and 2005, significant increases were recorded for experimenting and occasional use of ecstasy and cocaine, although the levels remained very low: around 0.5% of at least occasional users. Experimenting with cannabis rose considerably, from 24.9% to 30.6% between 2000 and 2005, but the rise did not apply to youths. Occasional use and recent use of cannabis were stable for the entire population; it declined in the youngest groups. Regular consumption increased considerably in both men and women. Physical Violence One question pertains to violent acts committed, in the hypothesis that delinquents may also suffer from mental health problems (behavioural problems in youths and antisocial personality disorders in adults, for instance). 앫 2.4 % of respondents claimed to have committed physical violence. 앫 33.8% of these had been the object of at least one assault during the past twelve months. 앫 Most were men and youths under age 25.

110

Self-reported crime.book Page 111 Thursday, April 23, 2009 8:02 AM

3.

Violence and school environment

Until recently, quantitative surveys on violence in schools have mostly been victimisation surveys, although they did include a few questions aimed at offenders. Surveys of self-reported delinquency/violence in France, developed by two teams, one led by É. Debarbieux and C. Blaya, the other by C. Carra, began in 2007. Both teams were inspired by the Effective School Literature, and concerned themselves with school climates, taking their inspiration from English language research, and particularly from Halpin and Croft (1963) and Finlayson (1973) for the former, and from American and Canadian research, and especially the work of Gottfredson (1985) and Janosz et al. (1998) for the latter. The object, violence, is not reduced to delinquent acts. One of the goals of these teams is to study the role of school contexts on the phenomenon of violence at school. Studies by the Observatoire européen de la violence scolaire (OEVS, European monitoring centre of school violence) The traditional focus Quantitative surveys on school violence have been in existence since the mid-1990s, with studies led by É. Debarbieux, Professor of educational science at the University of Bordeaux 2, and affiliated to the LARSEF (Laboratoire de Recherche Sociale en Éducation et Formation: Laboratory for Social Research in Education and Training). The older focus aimed at exploring ties between school climate, victimisations and breaking the rules at school. The original project was to repeat the initial survey11 at regular intervals in schools, so as to assess and describe the trend in the phenomenon of violence in the French educational system. This project could not be fully implemented for lack of funds: whereas a reference sample of over 14,316 students was interviewed in 86 schools, the survey could not be repeated in identical form. The questionnaire (copyright), designed primarily for secondary school students – and junior high school students in particular – but also covering grade 3 to 5 in elementary school, includes questions on rackets (as offender, victim or witness), fights, punishment suffered and reasons for it. It has been administered in France to 35,000 students (and 1,500 adults) since 1995. The same questionnaire has also been administered in Germany, Belgium, England, Spain and the Czech Republic. It has been 11. Debarbieux, 1996.

111

Self-reported crime.book Page 112 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

used in Burkina Faso, Djibouti, Mali, Senegal, Brazil, Chile and Quebec as well. Its administration in classrooms lasts about twenty minutes, without the presence of any adult from the school. In each school, either all classes or a sample of three classes per level – “the most difficult” and “the easiest” in the opinion of the school principal, plus a third, randomly selected class – are chosen by the team. Samples of schools include more establishments in educational priority zones and problem zones in proportion to their representation in the population, so as to test the correlation between socio-economic and victimisation or delinquency variables with subsequent weighting to grant sample representativeness. The sample mostly covers schools in the Aquitaine, Bouches du Rhône and Île-de-France regions. These studies, led by É. Debarbieux, with Y. Montoya and C. Blaya, are conducted in the framework of the OEVS since its creation in 1998. They are co financed by the European Commission, the Ministry of Education and the University Victor Segalen – Bordeaux 2. It is presently also financed by the Aquitaine Regional Council. Main Findings Violence at school is mostly composed of micro-violence. Questionnaires administered in 2003 do not show much deterioration of the school climate since 1995, on the average. However, the “social gap” is widening, with a significant increase in violence, and in fear of crime, and a considerable deterioration of relations between youths and adults, and of the general climate in underprivileged schools. Violence is less frequent, but it is increasingly perpetrated in groups and is therefore more violent, and felt more intensely by victims. There is unquestionably an ethnic component in relationships and in violence, according to É. Debarbieux, and this is contributed to by “social exclusion”. Although there is a discernible school effect, with some junior high schools succeeding in contradicting their predicted social destiny, these remain a minority. Students are therefore unequal in the face of violence, depending on whether they attend an “ordinary” or a “problem” school. The most recent focus The most recent focus has been developed in the framework of the OEVS, under the leadership of C. Blaya, who has conducted the French part of the International Self Report Delinquency Study. A first document has recently 112

Self-reported crime.book Page 113 Thursday, April 23, 2009 8:02 AM

been published, dealing with the technical aspect of the survey12. With this initiative, the OEVS extends its field of investigation to delinquency, violent and deviant behaviour that may occur outside of schools, and to the perpetrators involved. One of the main objectives of the national survey is to achieve an estimate of juvenile delinquency in France, and to assess its trends since the first French self-reported delinquency survey (Roché, 2000). Another major objective is to achieve international comparability of the prevalence and frequency of juvenile delinquency, so as to detect intercountry differences and similarities, and specific patterns, as well as to determine the role of national contexts in the forms taken by this phenomenon. The questionnaire was adjusted to the French context, and administered to a sample of 3,363 junior high school students aged 12 to 16. The French sample is composed of 23 schools. It is representative of school-attendees, in terms of age and sex; schools are representative with respect to their institutional category (priority and non-priority educational zones). 17 cities were chosen for France: 5 of over 100,000 inhabitants, 6 with less than 20,000 inhabitants and 7 middle-sized. The paper and pencil questionnaire was administered by the OEVS team in the classroom, using a protocol ensuring that responses are confidential and respondents anonymous. Findings are to be published in Junger-Tas, 2009. The CESDIP – Nord/Pas-de-Calais IUFM Survey This study, conducted in the framework of the CESDIP (Centre de recherches sociologiques sur le droit et les institutions pénales – CNRS), and directed by C. Carra, combined the victimization survey with a survey on self-reported violence, so as to study violence in elementary schools13. The objective was threefold : 앫 learning about respondents’ experience of violence; 앫 estimating the phenomenon for the Nord Département, where the survey took place; 앫 drawing comparisons between schools and evaluaing the role of the socio-educational environment, and especially of the school climate, in variations in the phenomenon. This survey was financed by the Nord/Pas-de-Calais IUFM and the CESDIP. It followed another study conducted in junior high schools, and which represented the first victimization survey conducted in schools in France14. 12. Blaya, 2007. 13. Carra, 2006. 14. Carra, Sicot, 1997.

113

Self-reported crime.book Page 114 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

This new concern with violence committed is derived from a conception viewing the experience of violence as constructed by the individual as victim as much as perpetrator. The particularity of this survey, as compared to the other surveys discussed in this report, is the absence of predefined itemization of violence. Categories are constructed afterwards on the basis of respondents’ answers to open-ended questions. This approach yields access to the situations and meanings ascribed to them by respondents. The in-depth questionnaire includes a section on victimisation and another, mirror-like one on selfreported violence. To limit possible lapses of memory in the description of violent situations suffered or inflicted, the questionnaire explores in detail the last time the respondent experienced one of those situations, be it as victim or as perpetrator, within a defined lapse of time: since the beginning of the school year. The schools selected constitute a representative sample of schools in the Nord département, based on two criteria: the institutional social category of the school (educational priority or violent zone) and size. A further criterion was the geographic location of the school. The sample was comprised of 31 schools, within which one class per level (with the exception of first grade) was investigated. Data from 2,073 student-questionnaires and about one hundred teacher-questionnaires were analyzed. The questionnaires were administered to pupils in their classroom by the investigators, whereas teachers were relieved from work so as to fill out their own questionnaire. It took about forty minutes to fill out the questionnaire, and the investigators carried all the material away with them so as to ensure that responses were confidential and respondents remained anonymous. 15

Main Findings on Perpetrators of Violence and School Contexts15 앫 27.1% of pupils reported having committed some violence within their school or in its immediate surroundings. 앫 Among those pupils who reported having committed some violence, there is a significant over-representation of boys, of single-parent or blended families, of large families (5 children or more), and of pupils aged 11 and more. The over-representations were found for pupils who claimed to have been victims. There is a resemblance, then, between victims and victimisers.

15. Only findings on students are reported, since the number of teachers turned out to be insufficient for reliable statistical analysis.

114

Self-reported crime.book Page 115 Thursday, April 23, 2009 8:02 AM

앫 69.9% of pupils who reported having committed some violence also claim to have been victims of violence. 앫 The socially segregated school phenomenon is very significantly correlated with that of school violence. 앫 However, violence is greater in school contexts where the study atmosphere is poor, and the educational climate is marked by frequent negative sanctions (punishment and bad notes) and by a strong feeling of unfairness, based on the perception that teachers do not give everyone equal treatment.

II.

What use for self-reported delinquency surveys?

1.

Findings are widely circulated and publicised

The findings from all of these surveys are issued in reports available on paper (see the bibliography for the main publications) and sometimes on the internet. The references for S. Roché’s reports and books, along with the table of contents, may be found on the CERAT website16; most articles and some reports are downloadable. The OEVS surveys have been published; a presentation may be consulted on the observatory site17. The CESDIP survey may be found on the site of the Nord/Pas-de-Calais IUFM website18, whereas the summary has been published as an issue of Penal Issues, downloadable on the CESDIP site19. All OFDT publications on the ESCAPAD and ESPAD surveys are available on the OFDT sites20. Each ESPAD survey led to the production of international reports to be found on a website21, a site which also furnishes direct access to the methodology used and the questionnaires administered. Reports are also available for each Baromètre santé survey; they are listed on the INPES site22, along with a detailed presentation of the surveys (and especially of their methodology). Some are directly downloadable. These surveys have all received publicity in the local and national media. However, while some have achieved genuine institutional recognition, 16. 17. 18. 19. 20. 21. 22.

http://www.upmf-grenoble.fr/cerat http://www.obsviolence.com http://www.lille.iufm.fr http://www.cesdip.org/IMG/pdf/pi_09_2006.pdf http://www.ofdt.fr and www.drogues.gouv.fr http://www.espad.org http://www.inpes.sante.fr

115

Self-reported crime.book Page 116 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

others encounter difficulties. This is particularly true of self-reported delinquency surveys, in spite of their focus on delinquency.

2.

Difficulty, nonetheless, in developing some of these surveys

One obstacle to the development of self-reported delinquency surveys lies in their cost. When directed at delinquency, the problem is acute: to uncover a large number of offenders having committed relatively serious acts, very large samples are required. Moreover, and contrary to AngloSaxon countries, these studies can neither lean on an academic cultural tradition nor on their use by public authorities. The question of the reliability of respondents’ answers is therefore all the more acute. The limits of sampling techniques are also questioned: when these surveys are carried out in schools, as is often the case, the probability of touching individuals seriously engaged in delinquency after 16 is slight. There is little appeal to such surveys, then. S. Roché’s reports do produce recommendations, however, and offer help in reflection and decisionmaking. There is, nonetheless, some interest for the findings they may yield: for instance, S. Roché has appeared at a hearing of the French senatorial investigation commission on juvenile delinquency23. These findings are also put to use within the scholarly community, and as mentioned above, they receive considerable publicity in the mass media. Whereas they provide data that contribute to public debate on safety, they apparently are not used to guide, or to evaluate, public policies, at least in this particular field. This type of methodology is new in the field of school violence, whereas the most frequent type of survey since the mid-1990s is the victimisation survey (see above). It seems premature, then, to evaluate the repercussion of surveys on self-reported delinquency or violence. However, those establishments which participated in the surveys and which so desire may take advantage of these findings. They are also used in the framework of initial and continuing training for teachers and supervisory staff.

3.

Public health: a field of its own

The Baromètre santé, ESCAPAD and ESPAD surveys are ongoing schemes carried out by national public agencies – the OFDT, INSERM, and INPES – working with the Ministry of Health. These agencies are all in charge of 23. … like É. Debarbieux, who reported on work based on a different methodology. See Carle, 2002.

116

Self-reported crime.book Page 117 Thursday, April 23, 2009 8:02 AM

producing knowledge for use in public debate and government action in the public health field. The goal of the OFDT, for instance, is to provide decision-makers, researchers and professionals with the most recent knowledge in the field of psychoactive substances and addictions. The INSERM aims at contributing to reflection on prevention policies through a better understanding of trends in use and changes in these trends induced by public policies24. The INPES surveys mostly serve its work, and especially the implementation of public health programs. References to these general population surveys may even be found in the texts presenting the March 2007 Act on crime prevention and the 2007 official enforcement instruction. Prevention campaigns on cannabis use were launched two years ago, when these surveys indicated a high rate of regular use among youths. Other examples may be cited, but the responses, or more often the non-response of the funding agencies as to their use of the surveys to facilitate reflection and decision-making lead us to be cautious. In other words, the question is: did these surveys guide public policies, or were they merely used to justify previous decisions? At the end of this overview, this remains a moot question, given the multiplicity of elements participating in the development and implementation of public policies and the complexity of the processes involved.

III. Conclusion There are very few self-reported delinquency surveys in France. Only two of these, one of which was under way at the time of writing, use the international questionnaire – adjusted to the French context – and therefore the underlying theoretical approach. The others make use of its methodological option, and include questions asking perpetrators about their practices, uses, behaviour, and attitudes. However, the theoretical perspective is very different, as are the practical stakes. Those researchers who apply this methodology find it difficult to raise funds for it, since neither public nor private actors are looking for this kind of survey. The surveys that manage to be continuous are those that are incorporated in national public agencies connected with ministries, and which have international, mostly European, ramifications as well. There are some occasional instances of use by local and national level actors, but these surveys – except perhaps in the public health field – do not seem to have sufficient weight to orient public policies. 24. See the INSERM site: http://www.inserm.fr

117

Self-reported crime.book Page 118 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

References ARÈNES J., JANVRIN M.-P., BAUDIER F., Dir., 1998, Baromètres santé jeunes 97/98, Vanves, CFES. BALLION R., 1999, Les conduites déviantes des lycéens, Rapport pour l'OFDT, Bordeaux, CADIS. BAUDIER F., ARÈNES J., Dir., 1997, Baromètre santé adultes 95/96, Vanves, CFES. BECK F., 1998, Rapport sur la mise en place du dispositif d'enquêtes en population générale sur l'usage de drogues, Paris, OFDT. BECK F., LEGLEYE S., PERETTI-WATTEL P., 2000, Regards sur la fin de l'adolescence: consommation de produits psychoactifs dans l'enquête ESCAPAD 2000, Paris, OFDT. BECK F., LEGLEYE S., PERETTI-WATTEL P., 2002, Alcool, tabac, cannabis et autres drogues illicites parmi les élèves de collège et de lycée: ESPAD, 1999 – France, Tome II, Paris, OFDT. BECK F., LEGLEYE S., SPILKA S., 2004, Drogues à l'adolescence. Niveau et contextes d'usage de substances psychoactives à 17-18 ans en France: ESCAPAD 2003, Paris, OFDT. BLAYA C., 2007, ISRD 2 Technical report. France, OEVS, Université VictorSegalen-Bordeaux 2, mars. CARLE J.-C., 2002, Délinquance des mineurs: la République en quête de respect (rapport de la commission d'enquête sur la délinquance des mineurs Paris, Sénat. CARRA C., Dir., 2006, Violences à l'école élémentaire, approche quantitative et comparative, le cas du département du Nord, IUFM du Nord / Pas-deCalais – CESDIP (CNRS). CARRA C., SICOT F., 1997, Une autre perspective sur les violences scolaires : l’expérience de victimation, in CHARLOT B., EMIN J.-C., (Dir.), Violences à l’école. État des savoirs, Paris, Armand Colin, 61-82. CHOQUET M., LEDOUX S., 1994, Adolescents: enquête nationale, Paris, Éd. INSERM. CHOQUET M., BECK F., HASSLER C., SPILKA S., MORIN D., LEGLEYE S., 2004, Les substances psychoactives chez les collégiens et les lycéens: consommations en 2003 et évolutions depuis 10 ans, Tendances, 35, Paris, OFDT/INSERM, mars. CHOQUET M., LEDOUX S., HASSLER C., 2002, Alcool, tabac, cannabis et autres drogues illicites parmi les élèves de collège et de lycée: ESPAD, 1999 – France, Tome I, Paris, OFDT/INSERM. CHOQUET M., LEDOUX S., HASSLER CH., PARÉ C., 1998, Adolescents de la protection judiciaire de la jeunesse et santé, rapport financé par la MILDT, à l'initiative de la PJJ, Paris, INSERM. 118

Self-reported crime.book Page 119 Thursday, April 23, 2009 8:02 AM

DEBARBIEUX É., GARNIER A., MONTOYA Y., TICHIT L., 1999, La violence en milieu scolaire, Tome II: Le désordre des choses, Paris, ESF. DEBARBIEUX É., 1996, La violence en milieu scolaire, Tome I: État des lieux, Paris, ESF. DEBARBIEUX É., 2006, Violence à l'école: un défi mondial?, Paris, Armand Colin. FINLAYSON D. S., 1973, Measuring School Climate, Trends in education, 30, 19-27. GOTTFREDSON D.C., 1985, The effective school battery: User manual, John Hopkins University, Psychological Assessment Resources, Baltimore. GOTTFREDSON G. D., GOTTFREDSON D.C., 1985, Victimization in schools, New York, Plenum Press. GUIBERT PH., BAUDIER PH., GAUTIER A., 2000, Baromètre santé 2000, 2 volumes, Saint-Denis, INPES, coll. Baromètre santé, GUILBERT PH., GAUTIER A., (dir.), 2006, Baromètre santé 2005, Premiers résultats, Saint-Denis, INPES, coll. Baromètre santé. HALPIN A. W., CROFT D. B., 1963, Organizational Climate of Schools, Chicago, Midwest Administration Center of the University of Chicago. JANOSZ M., GEORGES V., PARENT S., 1998, L'environnement socioéducatif à l'école secondaire: un modèle théorique pour guider l'évaluation du milieu, Revue Canadienne de Psycho-éducation, 27, 2, 285-306. JUNGER-TAS J., MARSHALL I., ENZMANN D., KILLIAS M., STEKETEE M., GRUSZCZYNSKA B., 2009, Juvenile Delinquency In Europe And Beyond: Results Of The Second International Self-Report Study, Berlin/ Heidelberg, Springer. LEGLEYE S., BECK F., SPILKA S., LE NEZET O., 2007, Drogue à l'adolescence en 2005. Niveaux, contextes d’usage et évolutions à 17 ans en France – Résultats de la cinquième enquête nationale ESCAPAD, Paris, OFDT. PERETTI-WATTEL P., 2001, Théories de la déviance et délinquance autoreportée en milieu scolaire, Déviance et Société, 3, 25, 235-256. EMCDDA, 2005, 2005 Annual report: The state of the drugs problem in Europe, Luxembourg, Office for Official Publications of the European Communities. ROCHÉ S., 2001, La délinquance des jeunes, les 13-19 ans racontent leurs délits, Paris, le Seuil. ROCHÉ S., ASTOR S., IVALDI G., TOURNIER P. V., 2000, Enquête sur la délinquance auto-déclarée des jeunes, Grenoble, CERAT-IEP. ROCHÉ S., BÈGUE L., ASTOR S., 2004, Délinquance de rue et risques routiers chez les adolescents, Grenoble, PACTE-CERAT-IEP. ROCHÉ S., BIANCHINI C., ASTOR S., 2004, La délinquance auto-déclarée de jeunes judiciarisés et d'un échantillon représentatif des 13-19 ans, Grenoble, PACTE-CERAT-IEP. 119

Researchers and name of survey S. Roché, S. Astor, G. Ivaldi, V. Tournier: Enquête sur la délinquance autodéclarée des jeunes, Grenoble, CERAT-IEP, 2000 This is the first selfreported delinquency in France ESCAPAD Health and Substance Consumption during the Military Defence Preparation Day

120

annual since 2000

Territorial scope Grenoble and SaintEtienne metropolitan areas

drug use nationwide public health survey with regional applications

Date(s) Scope of survey 1999 juvenile delinquency, in line with the international self-reported delinquency survey

Appendix 1 : Synoptic Table

Self-reported crime.book Page 120 Thursday, April 23, 2009 8:02 AM

Method and response rate

2,288 students aged 13 to 19, in 101 schools (public and private, junior high and high schools)

앫 random sampling (5%) from a list of students in each of the selected schools 앫 representative sample of 13-19 year-olds attending the schools surveyed 앫 anonymous, paper and pencil questionnaire Most of the questionnaire administered by researcher, faceto-face, in a room provided by the school (usually the infirmary); the part on assaults and drugs is selfadministered Students’ response rate: 86% 앫 all conscripts 앫 a cross-sectional, exhaustive survey 앫 age: usually 17 앫 paper and pencil, anonymous, self-administered 앫 sample questionnaires analyzed in 앫 weighting of the sample for the demographic weight of 2005: 29,393 the age group within each département French 17 앫 rate of participation in the Military Preparation Day: year-olds in 90% metropolitan France, 50.9% of whom were boys and 49.2% girls

Sample

Self-Reported Crime and Deviance Studies in Europe

five sweeps since 1992

Baromètre Santé

risk national consumption telephone public health survey with regional use

Date(s) Scope Territorial of survey scope every drug use national four public health survey years jointed with since the 1999 international survey

Researchers and name of survey ESPAD European School Survey Project on Alcohol and Other Drugs

Self-reported crime.book Page 121 Thursday, April 23, 2009 8:02 AM

121

Method and response rate

2003 edition: 앫 a cross-sectional survey sample composed 앫 stratification of the sample of schools based on: type of of 16,000 12 to 18 school; public or private; educational priority zone or year-old students not, located in a rural or urban district in 400 high 앫 respondents are representative of youths living in schools France for the ages at which the school attendance rate is very high (13-16), but less reliably so for older students 앫 anonymous, paper and pencil questionnaires selfadministered in classrooms 앫 students’ response rate: 78.2% 2005 sweep: 앫 two-level random sampling: random selection of ages 12 to 75 households from a list of telephone numbers, then 30,514 people random selection of one individual in each household interviewed, 앫 data weighted by probability of random selection within distributed the household and adjusted to figures for the 1999 between census households with a 앫 anonymous questionnaire, by telephone, with landline and those interviewers using CATI with a cell phone 앫 refusal rate: 27% for households in the phone directory, only 5.2%, for individuals, with another 7.5% dropping out during the interview. For those not in the directory, respectively 37.8%, 7.3% and 9.6%

Sample

122

C. Carra, Dir., Violences 2004 à l'école élémentaire, approche quantitative et comparative, le cas du département du Nord, Nord / Pas de Calais IUFM – CESDIP/CNRS, 2006.

Researchers and name Date(s) of survey of survey É. Debarbieux, La 1995 violence en milieu scolaire, Tome I: État des lieux, Paris, ESF, 1996.

violence and school contexts

violence and school contexts

Scope

Self-reported crime.book Page 122 Thursday, April 23, 2009 8:02 AM

Nord département

Territorial scope areas surveyed: mostly in Aquitaine, Bouches-duRhône and Île-deFrance regions

Method and response rate

sample including 앫 a sample of schools, with over-representation of those in 14,316 students the educational priority category or located in problem (mostly in junior zones, with subsequent weighting to ensure high school) from representativeness 86 schools 앫 in junior high schools, an exhaustive sample for schools with less than 300 students. For the others, composition of a sample based on 3 classes per level: the most difficult and least difficult according to the principal, plus one non-specific class 앫 paper and pencil questionnaire, self-administered in classrooms with responses confidential sample analyzed: 앫 a representative sample of the Nord département 2,073 pupils aged schools, based on two criteria: institutional category of 7 to 12 and 93 the school (educational priority or violence prevention teachers zone, no special category) and size. In addition, geographic location was considered 앫 31 schools surveyed, within which random selection of one class per level, excluding first grade 앫 anonymous, self-administered paper and pencil questionnaires 앫 administered in classrooms for pupils and elsewhere for teachers 앫 three schools refused to participate and were replaced by others with the same characteristics

Sample

Self-Reported Crime and Deviance Studies in Europe

Researchers and name of survey C. Blaya, ISRD 2 Technical report. France, OEVS, University Bordeaux 2, March 2007. This is the second selfreported delinquency survey done in France, currently undergoing analysis

Date(s) Scope of survey 2006French part 2007 of the International Self Report Delinquency Study (ISRD2)

Self-reported crime.book Page 123 Thursday, April 23, 2009 8:02 AM

123

Territorial scope national survey 17 cities selected: 5 with over 100,000 inhabitants, 6 with less than 20,000 and 7 mediumsized

Method and response rate

3,363 junior high 앫 23 schools surveyed school pupils aged 앫 for representativeness of students for age and gender 12 to 16 앫 representativeness of schools based on institutional category (educational priority or not) 앫 anonymous, paper and pencil questionnaires, selfadministered in classrooms

Sample

Self-reported crime.book Page 124 Thursday, April 23, 2009 8:02 AM

Self-reported crime.book Page 125 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany Thomas Görgen, Susann Rabold

In Germany, studies on self-reported delinquency date back to the late 1960ies/early 70ies (Kreuzer, 1975 ; Lösel, 1975 ; Quensel, 1971 ; Quensel, Quensel, 1969). These early studies – most of them directed at rather small samples of young people at schools and universities – tend to describe less severe forms of juvenile delinquency as ubiquitious and normal phenomena in young persons’ lives. They show that among young men delinquent behaviour is more widespread, more severe and more violent than among women of the same age. They also point at the fact that delinquency is not an isolated problem but often coincides with school deficits, family problems etc. (for an overview see Kreuzer, 1994a, b). Since those early stages of SRD research in Germany, a large number of studies have been published. Though meanwhile some local/regional surveys have been conducted repeatedly (or as longitudinal studies), Germany still lacks institutionalization of crime-related self-report studies – both on the victim and on the offender side. With regard to the multitude of German SRD studies on the one hand and the lack of any repeated national surveys on the other, the following report will focus on an overview of the current situation of research on self-reported delinquency in Germany; a complete inventory of four decades of German SRD research would be far beyond its scope.

I.

Current situation of research on self-reported delinquency in Germany

As in most countries, German SRD studies show a strong focus on juvenile delinquency. This report distinguishes between: 앫 delinquency studies in the general adult population, 앫 delinquency studies in child and adolescent populations, 앫 delinquency studies in special populations.

125

Self-reported crime.book Page 126 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

1.

Delinquency studies in the general adult population

The main German exception to the rule of equating SRD research with research on children and adolescents is the German General Social Survey (Allgemeine Bevölkerungsumfrage der Sozialwissenschaften – ALLBUS). Since 1980, ALLBUS has been conducted in two year intervals as a national survey on attitudes, behaviour and social structure. ALLBUS has a crosssectional design; representative samples of the German population are interviewed face-to-face (in recent years via CAPI). The ALLBUS surveys conducted in 1990 and 2000 included four items on self-reported delinquency (cf. Mehlkop, Becker, 2004 ; Becker, Günther, 2004 ; Becker, Mehlkop, 2006 ; Lüdemann, 2002). These questions touched four widespread offences: i) shoplifting, ii) fare dodging, iii) driving under the influence, and iv) tax evasion. In 1990, 3,051 community-dwelling German-speaking adults were interviewed; in 2000, sample size was 3,138. A 'sealed envelope technique' was used for the delinquency items in order to build trust and to emphasize anonymity and confidentiality.1 ALLBUS asks for lifetime prevalence / incidence only. In ALLBUS 2000, about 37% reported at least one incident of fare dodging; lifetime prevalence rates for drunk driving (27%), shoplifting (11%) and tax evasion (13%) were lower. The mass offences measured in the ALLBUS surveys were committed more frequently by men than by women; except for tax evasion, prevalence of offences declined with age. Using 1990 and 2000 ALLBUS data for shoplifting and tax-evasion, Mehlkop, Becker (2004) analyze relationships between social class and criminal behaviour. Across offences, there was no stable relationship between individuals’ social status and delinquency. Types of offences committed were affected by social status. Subjective expected probability of failure or success and internalized norms were important predictors of offending. Beside ALLBUS, SRD research in adult populations has mostly been restricted to special samples like heroin addicts. One exception to be mentioned here is a survey of 16- to 34-year olds (460 in East Germany in 1991 ; 650 in East and West Germany in 1993 ; 1,360 in East German cities in 1993) embedded in a study on crime and social change in the wake of German re-unification (cf. Boers, Class, Kurz, 1994). As the majority of German SRD studies are on delinquency in early stages of life, the paper will turn to this part of research now.

1.

Using data from ALLBUS 2000, Becker, Günther (2004) conclude that mail surveys could be a way to improve data quality in SRD research (e.g. because of exclusion of interviewer effects).

126

Self-reported crime.book Page 127 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

2.

Delinquency studies in adolescent populations

German SRD research’s focus on young people is due not only to the involvement of juveniles in criminal / delinquent behaviour but also to the relatively easy access researchers have to child and adolescent subjects (cf. Kreuzer, 1994a). Surveys can often be administered in the classroom, thus keeping resources required for conducting the surveys in a limit. The following sections will give an overview of delinquency studies among adolescents and – to a much lesser extent – children. Given the restrictions in space, it will not be possible to describe each and every study conducted in Germany. Cross-sectional surveys Since 1998, a series of SRD studies with adolescents (and children) has been conducted by Hanover-based Criminological Research Institute of Lower Saxony (Kriminologisches Forschungsinstitut Niedersachsen, KFN) (Baier, Pfeiffer, 2007 ; Wetzels, Enzmann, Mecklenburg, Pfeiffer, 2001 ; Wilmers, Enzmann, Schäfer, Herbers, Greve, Wetzels, 2002). There have now been repeated surveys in the cities of Hamburg (1998, 2000, 2005), Hanover (1998, 2000, 2006), Leipzig (1998, 2000, 2004), Munich (1998, 2000, 2005), Stuttgart (1998, 2005) and Schwäbisch Gmünd (1998, 2005) and a number of (yet) not repeated surveys in other cities, towns and rural areas. Most of these surveys were funded by the cities themselves. Samples of the KFN surveys are usually very large. In Munich, more than 8,000 9th graders2 and almost 500 4th graders have been interviewed. The self-administered questionnaire asks for 12 types of violent and property offences and uses a 12 months reference period. In the autumn of 2007, a very large survey of 9th and 4th graders has been conducted. This survey – funded by the German Ministry of the Interior – was originally planned as a nation-wide study and the final sample size reached 53,000. In the survey conducted in 2000, 69% of the sample reported at least one offence during the last 12 months; if fare dodging (as the most prevalent and least severe offence) was excluded, the overall prevalence rate was 46%. While thus delinquency as such appears to be normal, qualified property offences (like burglary or theft from cars) and severe forms of violence (robbery, blackmail, threat with weapon) are reported by only 1 to 4% of subjects. Violent offences, damage to property and severe forms of

2.

As German children usually enter school at age 6, respondents’ age can approximately be calculated as grade + 6.

127

Self-reported crime.book Page 128 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

theft were reported much more frequently by boys than girls whereas there was no gender difference with regard to fare dodging and shoplifting. KFN surveys repeatedly found connections between the social situation of juveniles and their delinquent behaviour. Delinquency is higher if families are economically disadvantaged and if educational perspectives/options of juveniles are very limited. Offending rates decline with rising school level3. KFN surveys focus upon links between social exclusion, ethnicity and violence. The year 2000 survey found significantly higher prevalence rates of severe forms of violence among juveniles of non-German origin. There was no difference in the field of severe property offences, and young migrants even reported lower rates of minor property offences than ethnic Germans. KFN surveys point at associations between experiences of parental violence in childhood and self-reported violence. Part of the higher rate of violence among young Turkish males can be explained by their victim experiences. The concept of "violence-legitimizing norms of masculinity" (cf. Enzmann, Wetzels, 2003 ; Enzmann, Brettfeld, Wetzels, 2003) – drawing on Nisbett, Cohen's (1996) 'culture of honor' – is used to explain differential levels of violence in ethnic groups; these norms are strongest among youths of Turkish and former Yugoslav origin. In recent KFN surveys, media use in general and violent video games / computer games in particular are important variables, supposed to influence school achievement, empathy, and aggressive behaviour (Mössle, Kleimann, Rehbein, Pfeiffer, 2006). A recent study by Brettfeld, Wetzels (2007) on young Muslims living in Germany focussed on attitudes but also included some questions on SRD. The sample of 2,683 9th and 10th graders in the cities of Hamburg, Cologne, and Augsburg consisted of three subsamples: non-Muslim youths without migration background (n=1,553), non-Muslim youths with migration background (n=630), and Muslim youths (n=500). With regard to four violent offences (robbery, assault without weapon, threat with weapon, extortion), the study used a 12 months reference period. Brettfeld, Wetzels (2007) found self-reported violence to be more prevalent among migrants than among ethnic Germans. The most interesting finding was that for non-Muslim youths, there was a negative relationship between religiosity and SRD. For non-Muslim youths, prevalence of self-reported violence declined from 43.2% (not at all religious) to 18.5% (very religious); for non-Muslim Germans, the respective values were 26.7% and 3.

The 2000 survey studied four types of schools (Hauptschule, Realschule, Integrierte Gesamtschule, Gymnasium) which offer different educational and vocational perspectives to their graduates.

128

Self-reported crime.book Page 129 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

14.3%. For young Muslims (29.2% vs 28.6%), there was no such connection between religiosity and violence. One important topic in German research is the discrepancy of trends in recorded crime and in self-report studies over time. Whereas – as figure 1 shows – officially recorded juvenile violence displays a rising tendency over more than two decades, this does not hold true for self-reported offences.

Source: German Federal Criminal Police Office (Bundeskriminalamt): Police Crime Statistics; Violent crime: Homicide, rape, robbery, and aggravated assault

Figure 1: Violent crime suspects per 100,000 of respective age group, Germany, 1984 – 2004

Pfeiffer, Wetzels (2006) and Baier, Pfeiffer (2008), in comparing 1998 and 2005 KFN survey results, show that this is not true for self-reports. The prevalence of victim experiences as well as self-reported offences has diminished (offender prevalence: 20.1% to 17.5%); the percentage of repeat serious offenders (5+ violent offences /12 months) was also smaller in 2005 than in 1998. The severity of reported victim experiences has decreased. Comparisons of 1998 and 2005 data also show that factors influencing reported crime have changed: willingness to report has risen, negative attitudes towards violence have grown stronger. This partially explains the rise in officially recorded juvenile violent crime; its visibility

129

Self-reported crime.book Page 130 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

and the readiness of victims to file charges against the offender have increased4.

Source : KFN student surveys, in % ; Baier & Pfeiffer, 2008

Figure 2: 12-months prevalence of self-reported juvenile violence in Munich, Stuttgart, Schwäbisch Gmünd, 1997 and 2004

Among other cross-sectional SRD studies to be mentioned are the surveys conducted by Dünkel and colleagues in Mecklenburg–Western Pomerania. Dünkel, Gebauer, Geng (2007) compare data from surveys of 9th graders conducted in Greifswald in 1998 and 2002.5 They show a significant decline of lifetime and 12 month prevalence of violent offences; this holds true not only for robbery (where the decline is mirrored in police data) but also for assault (where police data indicate a rise of the number of violent offences). In a survey of 4,028 8th graders in the city of Bochum (Feltes, Goldberg, 2006 ; Goldberg, 2006), 13.5% reported having assaulted somebody in a way that he /she needed medical treatment during the last year; 8.5% reported robbery. Prevalence of violence differed with ethnic 4.

5.

Recent German school surveys also point at a reduction of violence in schools. These findings are backed by data from insurance companies. This counters popular perceptions that school violence is growing in frequency and severity (cf. Oberwittler, Köllisch, 2004; Tillmann, Holler-Nowitzki, Holtappels, Meier, Popp, 2000). Studies in the 1990ies had already questioned the common perception of a constant increase in school violence (cf. Schwind, Roitsch, Ahlborn, Gielen, 1995 ; Schwind, Roitsch, Gielen, 1995). The 1992 sample (724 9th graders) was part of the multinational Mare-BalticumYouth-Survey conducted in Germany, Estonia, Finland, Lithuania, Poland and Sweden (Dünkel, Gebauer, Kestermann, 2005).

130

Self-reported crime.book Page 131 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

origin and was especially high among Turkish youths (27%, compared to 11% in German and 15% in Polish subsamples). In a survey of 9th graders in the cities Freiburg and Cologne in 1999 (Oberwittler, Blank, Köllisch, Naplava, 2001), 11.5% of all females and 28.9% of all males reported having committed at least one violent act (assault, robbery, blackmail, threat with weapon) in the last year. Of all offences, theft is most widespread among girls (28.0%) and boys (41.3%). German SRD studies are and have been integrated into the International Self-Report Delinquency Studies. An SRD study among youth in the southern German city of Mannheim found similar rates of offending among indigenous and migrant youth (Sutterer, Karger, 1994). The University of Hamburg6 is in charge of the German part of ISRD 2. The survey is conceived as a national multicenter study, a survey among 7th, 8th and 9th graders in one metropolitan area, one middle-sized town and 3 small towns in a rural area, covering juvenile delinquency as well as victimization. Longitudinal studies There are some German self-report studies taking a longitudinal approach. Funded by German Research Foundation (DFG), annually repeated surveys in schools in the cities of Münster and Duisburg (cf. Boers, Reinecke, 2007 ; Boers, Reinecke, Walburg, 2006 ; Weins, Reinecke, 2007) are part of an integrated cohort and panel design. In Münster, a survey among 1,949 7th graders (mean age 13 years) was started in 2000 and repeated annually until 2003. The Duisburg surveys started in 2002 (n=3,411) and are repeated until 2007 (i.e. until 18 years of age). Measures of delinquency include lifetime and 12-month prevalence of 16 offences (violence, vandalism, property and drug offences). Findings point at a slight decrease in self-reported delinquency over time. Processes of spontaneous desistance can already be found in early adolescence. In the Duisburg sample – being a city with a high proportion of migrants – the overall prevalence rates of male Turkish youths are very close to that of young men without migration background; delinquency among Turkish girls was found to be significantly lower than among comparable German female adolescents. A prospective longitudinal study conducted in Bremen (Schumann, 2003, 2004, 2007) measured self-reported delinquency over a time-span of 11 years. The sample consisted of 424 graduates of German Hauptschule and of special schools for handicapped children (Sonderschule). Data on 6.

The core research team consists of Dirk Enzmann, Franziska Seyboth-Teßmer, and Peter Wetzels.

131

Self-reported crime.book Page 132 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

education and work were recorded on a monthly basis; self-reported delinquency was measured annually and supplemented by official data on court-recorded crime. The study focussed on interactions between educational and work biography, delinquent behaviour and interventions by the criminal justice system. It found that work biography and delinquency are to a large extent independent spheres. However, decisions made by the criminal justice system were influenced by young people’s work and educational status. Harsh CJS interventions were found to impede education and vocational career and to increase the risk of greater persistence of delinquent behaviour. Between 1977 and 1996, the Department of Child and Adolescent Psychiatry of the Central Institute of Mental Health, Mannheim, conducted a prospective longitudinal study which included measures of SRD (cf. Lay, Ihle, Esser, Schmidt, 2005 ; Schmidt, Lay, Ihle, Esser, 2001). The original cohort consisted of a random sample (216 children) and a screening sample (183 children with behavioural problems) of children born in 1970 and resident in Mannheim in 1978. The sample of 399 eight-year-old children was followed up at ages 13, 18 and 25. 321 children participated up to t4 (179 ?, 142 ?). At all ages, data were obtained from parents. At ages 13, 18 and 25, structured interviews and questionnaires measuring selfreported delinquency were administered to the young subjects. The questionnaires consisted of 14 to 18 items describing delinquent and criminal behaviour. The researchers distinguish three life course types of delinquency: episodic juvenile delinquency, continued juvenile delinquency up to adulthood, and late-starting delinquency in early adulthood. They report that in early adulthood, the decline in recorded crime is larger than with regard to self-reported offences, implying that the prevalence of persistent offending is underestimated by police and court data. A study by Lösel, Bliesener (2003) started with a cross-sectional survey of 1.163 7th/ 8th graders from different school types (Hauptschule, Realschule, and Gymnasium) in the cities of Nuremberg and Erlangen. This survey served as a screening instrument for a longitudinal study with a sample of 102 boys divided into 4 subgroups (aggressive bullies, average students, socially competent students, and victims) that were investigated again after two years. Data for this smaller sample also include structured observations of behaviour, experiments, tests and evaluations by teachers. Lösel, Bliesener, Bender (2007) report that serious forms of self-reported aggression displayed high stability over time. Consumption of violent media content and an aggressive-impulsive response repertoire were significant predictors of delinquency.

132

Self-reported crime.book Page 133 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

Other longitudinal studies also integrate elements of self-reported delinquency, for example the project Wege aus schwerer Jugendkriminalität (Pathways out of serious juvenile delinquency) at the university of Tübingen. This study analyses the development of 56 juveniles who had received prison sentences of at least 10 months and were under probation (Stelly, Thomas, 2007). Self-report data are collected via semi-structured interviews.

3.

Delinquency studies with special populations

Beside these SRD studies in (age-selective) general populations, there are a number of German self-report studies addressing special populations. The reasons for limiting sampling to specific groups are manifold. They include easy access to the respective groups, the use of self-reports as an educational tool, research questions applying only to the respective populations, and the use of self-reported delinquency measures in research contexts where questions of victimization are central, but victim surveys appear not to be feasible. Examples will be given for these types; again, this cannot be a comprehensive presentation. Example 1: SRD studies with university students At a number of German universities, especially at law faculties (where criminological departments are usually located), SRD studies have been conducted with students, often with freshmen. Access to subjects is easy, and results and method can be made topics of lectures. At Giessen University, delinquency studies among law freshmen have been continuously conducted since 1976 (cf. Görgen, Kreuzer, Klein, 1995 ; Kreuzer, Görgen, Krüger, Münch, Schneider, 1993 ; Wittich, Görgen, Kreuzer, 1998). Studies conducted at different German law faculties between 1973 and 2005 (cf. Schwind, 2007 ; Schwind, Freier, Ballering, 2002) point at a long-time decline in self-reported delinquency (with physical assault going down from 39% to 23% and theft from 43% to 20%). Example 2: SRD studies with opioid addicts Self-report approaches have been used in the context of the recent – federally funded – German model project for heroin assisted treatment of opioid dependent patients7. Evaluation of this project included two 7.

Self-report studies with opiate addicts have been conducted much earlier in German research on drug-related crime (cf. Kreuzer, Römer-Klees, Schneider, 1991).

133

Self-reported crime.book Page 134 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

criminological studies, one using a quantitative SRD approach (Löbmann, 2003 ; 2006), the other based on in-depth interviews with addicts (cf. Köllisch, 2007 ; Köllisch, Kreuzer, 2006). Heroin treatment proved superior to methadone treatment with regard to offences committed by addicts in both groups; reductions were strongest for theft and drug trafficking. Survey findings paralleled trends in recorded offences. Example 3: SRD studies with incarcerated juvenile / young adult offenders The Hanover prison study (cf. Greve, Enzmann, 2003 ; Windzio, 2006) has been sponsored by the Volkswagen Foundation (1997-2003) and the German Research Foundation (2004 to date). This longitudinal study focuses on about 2.400 first-time incarcerated German males aged 14-24 in five correctional institutions in Germany. Self-reported delinquency is measured in face-to-face and telephone interviews during incarceration and after release. Panel attrition is high so that only for a part of the sample the full range of interviews can be realized. Serious offenders show higher rates of non-response. Example 4: SRD studies with nurses In some areas of basic victimological research, victim surveys are not feasible or severely limited with regard to data quality. In recent years, SRD studies with nurses have been conducted in research on elder abuse and neglect (cf. Görgen, 2006, and Rabold, Görgen, 2007, for SRD studies with nurses in residential and domestic care).

4.

German SRD research: Summary and further trends

Though Germany can look back at a rather long tradition of research and a multitude of studies, SRD research is still missing institutionalization. This has implications for interpretations of findings as well as for their political impact. Often, studies and findings are not comparable with regard to the samples and to the methods and instruments used. Most of German SRD studies are limited in their geographical scope. The focus of much of German SRD research is on questions associated with social change and social problems in Germany in the last decades. Recently, this includes the process of German re-unification (delinquency in East and West Germany and under conditions of rapid social transformation; cf. Posner, 1997 ; Heitmeyer et al., 1996 ; Kreuzer et al., 1993 ; Mansel, Hurrelmann, 1998 ; Langner, Sturzbecher, 1997), and the interconnections between delinquency and ethnicity / migration (Babka

134

Self-reported crime.book Page 135 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

von Gostomski, 2003a, b). Beside delinquency by ethnic foreigners, Germany has the rather unique phenomenon of migration by so-called resettlers, i.e. people of ethnic German origin, migrating to Germany mostly from the former Soviet-Union and from Eastern and Southeastern Europe, and gaining German citizenship. For both groups, SRD studies point at the significance of cultural factors (norms of masculinity) and experiences of domestic violence (Baier, Pfeiffer, Windzio, 2006). Few German delinquency studies integrate children as respondents; most of them focus upon substance abuse and school violence. Among the studies applying broader concepts of delinquency to children are surveys by Fuchs, Lamnek, Luedtke, Baur (2005) in Bavarian schools, surveys of 7th graders in Münster / Duisburg (Boers, Reinecke, 2007) and of 4th graders in KFN surveys 2005 (Baier, Windzio, 2006). The lack of German research on delinquency in childhood must be regarded as critical regarding the significance of child delinquency as a risk factor of persistent and severe offending in adolescence and early adulthood (and the lack of sound data on child delinquency in official statistics, cf. Brettfeld, 2006). Generally, German SRD research has demonstrated that minor forms of delinquency are a normal part of development in adolescence. Juvenile delinquency reaches its peak around age 16 (which is lower than with regard to police recorded crime). Historically, there was a rise of juvenile delinquency in the first half of the 1990ies, with a peak around 1996. According to several repeated studies, juvenile crime and violence have declined thereafter. German research has pointed at a number of risk factors of severe juvenile delinquency. Among them are low social status of family, restricted academic / educational opportunities, school absenteeism, excessive use of audiovisual media, especially violent video and computer games, bad familial climate, strained child-parent relationships, parental violence, and adherence to violence-legitimizing norms of masculinity. Connections between school types and delinquency have often been in the focus of research; this has to be seen with regard to Germany’s hierarchically stratified school system. Some studies point at social selectivity of the criminal justice system, claiming that formal social control is stronger for socially disadvantaged youth (cf. Brettfeld; Wetzels, 2003; Mansel, Raithel, 2003). Methodological studies on SRD research conducted during the last decade aim at improving self-reports. While some deal with general problems of self-reports like their capacity to produce valid data (Hermann, Weninger, 1999) or to make results comparable over time (Menzel, Peters, 1998), others compare different modes of administration (Naplava, Oberwittler, 2002) or suggest strategies to improve validity of SRD questions (Köllisch, 2002). 135

Self-reported crime.book Page 136 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Among recent developments in German SRD research is the application of multilevel models, analyzing the influence of individual and community factors (e.g. Oberwittler, 2003, 2004 ; Oberwittler, Köllisch, 2003) and the integration of delinquency measures in a long-term longitudinal study on early intervention that will include among a multitude of outcome variables (health, school achievement etc.) aspects of recorded and unrecorded crime and delinquency (cf. Maier-Pfeiffer, Pfeiffer, 2006). Compared to self-report studies in the UK, especially, German SRD research appears rather phenomenon-driven than aiming at progress in criminological theory. Influential concepts and theoretical approaches like self-control theory (Gottfredson, Hirschi, 1990) and the developmental approaches proposed by Moffitt (1993) or Sampson, Laub (1993) have been used in German SRD research, as well as more specific concepts like “culture of honour” (Nisbett, Cohen, 1996). With regard to funding, a rather diverse landscape of funding sources characterizes German SRD research. There have been a few studies – like the recent young Muslims survey by Brettfeld and Wetzels (2007) or the not yet finished nationwide KFN delinquency survey in schools, for which the Federal Ministry of the Interior has provided funds. For other largescale studies, German Research Foundation has been an important source. But funds also came from other foundations, other ministries, or from municipalities interested in a deeper understanding of the local crime and violence situation. Table 1 (Appendix) gives an overview of the funding sources of large German SRD studies since the early 1990ies.

II.

Impact of German self-report studies

Due to the very strong focus of German SRD research on adolescents, questions of impact primarily pose themselves with regard to juvenile crime.

1.

Impact on media

With regard to media coverage of crime and delinquency, the visible influence of SRD studies is rather low (cf. Pfeiffer, 2004 ; Reuband, 2004 ; Walter, 2000, 2003). As in other countries, German media mainly work according to the principle that "good news is no news". Consequently, media coverage of crime is strongly characterized 앫 by reporting on high profile cases; i.e. focussing on young serial offenders, instances of brutal and reckless offending etc.; 136

Self-reported crime.book Page 137 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

앫 by an underlying "things are always getting worse" message; i.e. conveying the information that crime in general and juvenile crime in particular show trends towards increasing prevalence, incidence, and severity; with regard to violence in schools, for example, the message of a qualitative step into hitherto unknown brutality has been conveyed by media reports again and again; 앫 by relying rather on official data (especially police data) than on results of SRD research. Of course, media coverage of crime is not a homogeneous phenomenon, and there are significant differences between quality papers (like weekly Die Zeit) and other print and audiovisual media. Conveying complex relationships between delinquency, reporting behaviour and official recording of crime to the public is a task that requires the will to do so but also needs space / time and an audience willing and able to delve into these complexities.

2.

Impact on crime-related policies

Estimating the impact of German SRD research on crime-related policies appears a complicated task. As crime policies typically are multidetermined phenomena, influences of delinquency studies are difficult to quantify, to distinguish from other influences and from mere congruence of developments. SRD research has the potential to serve as a "sobering agent" in German crime policies. Whether it already manages to do so, cannot be answered by a simple "yes" or "no". Critics (e.g. Feltes, 2006 ; Pfeiffer, 2004 ; Walter, 2003) argue that 앫 there is a far reaching immunity of politics and of media crime coverage against insights produced by criminological research in general and SRD studies in particular, 앫 crime policies as a whole are characterized by rising punitiveness and by giving priority to security at the expense of liberty, 앫 politics strategically use findings of criminological research to back preexisting positions and disregard them if they are not useful in that sense. Current German political discussions on 앫 lowering the minimum age for criminal sanctions (now: age 14), 앫 exclusively applying adult criminal law to young adults (currently: possibility to sanction 18-20 year olds according to juvenile law), and 앫 extending maximum prison sentences for juveniles to 15 years (now: 10 years) are based upon perceptions of juvenile crime as rising, brutalizing and as a genuine threat to public safety (or at least use such perceptions to back

137

Self-reported crime.book Page 138 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

their position) – i.e. perceptions that are much closer to media constructions of juvenile crime and delinquency than to findings of SRD research. On the other hand, the view that politicians are plainly ignorant of criminological research appears too simple. In the long run, there is a stable trend towards non-custodial sentences in juvenile justice in Germany. Though SRD studies may not have been able to change crime policies as a whole, several specific impacts can be established: 앫 Up to now, two "Periodic Security Reports" have been issued by the German government in 2001 and 2006, both written by researchers and both giving ample space to results of self-reported delinquency studies (Bundesministerium des Inneren / Bundesministerium der Justiz, 2001, 2006). 앫 The fact that the German Ministry of the Interior has recently funded an almost nation-wide SRD study among 4th and 9th graders can be seen as a strong indicator that SRD data are being regarded as an important basis for political decisions. 앫 Backed by SRD studies (e.g. Baier, Pfeiffer, 2007), school absenteeism is growingly being regarded as a serious problem and as an indicator of developmental problems. Interventions include schools, parents, police and courts and aim at breaking the mutually reinforcing effects of lack of supervision, opportunities for delinquent behaviour, and deficient educational and vocational perspectives (cf. Linssen, 2005). 앫 Current discussions about the future of the German school system partly take their arguments from findings relating to the distribution of delinquency in different types of schools (pointing at connections between early segregation of children into different school types and resultant concentration and accumulation of problems in Hauptschulen; cf. Baier, Pfeiffer, 2007) 앫 Criminological research in the context of heroin assisted treatment of opioid addicts has proven influential. Though it is still a matter of political controversy whether state administration of heroin to a small group of addicts that cannot be reached by therapeutic approaches will become a standard procedure, it has become a pragmatic alternative for many (local) political actors, and attitudes on this topic tend to cross political party lines. The strongest argument in favour of heroin administration has been its superior effect on frequency and severity of crime demonstrated by SRD research. With regard to results of SRD research, institutions of the criminal justice system may sometimes experience a kind of ambivalence: successfully fighting and preventing crime is their core job, but at the same time declining juvenile crime rates – as suggested by some self-reports – might 138

Self-reported crime.book Page 139 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

entail discussions about necessity and inevitability of the level of resources currently allocated to them. Here, open discussions about a changing "crime landscape" (international terrorism, internet-related crime – to mention just a few keywords), innovations in policing styles (especially the growing emphasis on prevention) and their implications for future crime policies are highly needed.

References BABKA VON GOSTOMSKI C., 2003a, Einflussfaktoren inter- und intraethnischen Gewalthandelns bei männlichen deutschen, türkischen und Aussiedler-Jugendlichen, Zeitschrift für Soziologie der Erziehung und Sozialisation, 23, 399-415. BABKA VON GOSTOMSKI C., 2003b, Gewalt als Reaktion auf Anerkennungsdefizite? Eine Analyse bei männlichen deutschen, türkischen und Aussiedler-Jugendlichen mit dem IKG-Jugendpanel 2001, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 55, 253277. BAIER D., PFEIFFER C., 2007, Hauptschulen und Gewalt, Aus Politik und Zeitgeschichte, 28, 17-26. BAIER D., PFEIFFER C., 2008, Entwicklung der Jugenddelinquenz und ausgewählter Bedingungsfaktoren seit 1998 in den Städten Hannover, München, Stuttgart und Schwäbisch Gmünd, KFN-Forschungsbericht, Nr. 104, Hannover, KFN. BAIER D. PFEIFFER C. WINDZIO M., 2006, Jugendliche mit Migrationshintergrund als Täter und Opfer. in HEITMEYER W., SCHRÖTTLE M., Hrsg, Gewalt: Beschreibungen, Analysen, Prävention, Berlin, Bundeszentrale für Politische Bildung, 240-268. BAIER D., PFEIFFER C, WINDZIO M. RABOLD, S., 2006, Schülerbefragung 2005: Gewalterfahrungen, Schulabsentismus und Medienkonsum von Kindern und Jugendlichen. Abschlussbericht über eine repräsentative Befragung von Schülerinnen und Schülern der 4. und 9. Jahrgangsstufe, Hannover, KFN - Forschungsbericht. BAIER D., RABOLD S., PFEIFFER C., WINDZIO M., 2006, Schülerbefragung 2005: Gewalterfahrungen, Schulabsentismus und Medienkonsum von Kindern und Jugendlichen in Thüringen. Abschlussbericht über eine repräsentative Befragung von Schülerinnen und Schülern der 4. und 9. Jahrgangsstufe, Hannover, KFN - Forschungsbericht. BAIER D., WINDZIO M., 2006, Gewalt unter Kindern im Schulkontext, Praxis der Rechtspsychologie, 16, 53-81.

139

Self-reported crime.book Page 140 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

BECKER R., GÜNTHER R., 2004, Selektives Antwortverhalten bei Fragen zum delinquenten Handeln, ZUMA-Nachrichten, 28, 54, 6-38. BECKER R., MEHLKOP G., 2006, Social class and delinquency: An empirical utilization of rational choice theory with cross-sectional data of the 1990 and 2000 German General Population Surveys (ALLBUS), Rationality and Society, 18, 193-235. BOERS K., CLASS M., KURZ P., 1994, Self-reported delinquency in Germany after the reunification, in JUNGER-TAS J., TERLOUW G.J., KLEIN M.W., (Eds.), Delinquent behavior among young people in the Western World, Amsterdam, Kugler, 345-355. BOERS K., REINECKE J. (Hrsg.), 2007, Delinquenz im Jugendalter. Erkenntnisse einer Münsteraner Längsschnittstudie. Münster: Waxmann. BOERS K., REINECKE J., WALBURG C., 2006, Jugendkriminalität – Keine Zunahme im Dunkelfeld, kaum Unterschiede zwischen Einheimischen und Migranten: Befunde aus Duisburger und Münsteraner Längsschnittstudien, Monatsschrift für Kriminologie und Strafrechtsreform, 89, 2, 63-87. BRETTFELD K., 2006, Umfang, Struktur und Entwicklung der Kinderdelinquenz: Befunde und Aussagekraft der polizeilichen Kriminalstatistik für Deutschland, Praxis der Rechtspsychologie, 16, 3052. BRETTFELD K., WETZELS P., 2003, Soziale Selektivität strafrechtlicher Sozialkontrolle bei Jugendkriminalität? Ergebnisse repräsentativer Dunkelfelderhebungen zur Wahrscheinlichkeit polizeilicher Registrierung delinquenter Jugendlicher, Praxis der Rechtspsychologie, 13, 226-257. BRETTFELD K., WETZELS P., 2007, Muslime in Deutschland. Integration, Integrationsbarrieren, Religion und Einstellungen zu Demokratie, Rechtsstaat und politisch-religiös motivierter Gewalt. Ergebnisse von Befragungen im Rahmen einer multizentrischen Studie in städtischen Lebensräumen, Berlin, Bundesministerium des Inneren. BUNDESMINISTERIUM DES INNEREN / BUNDESMINISTERIUM DER JUSTIZ, 2001, Erster Periodischer Sicherheitsbericht, Berlin, Bundesministerium des Inneren / Bundesministerium der Justiz. BUNDESMINISTERIUM DES INNEREN / BUNDESMINISTERIUM DER JUSTIZ, 2006, Zweiter Periodischer Sicherheitsbericht, Berlin, Bundesministerium des Inneren / Bundesministerium der Justiz. DÜNKEL F., GEBAUER D., GENG B., 2007, Gewalterfahrungen, gesellschaftliche Orientierungen und Risikofaktoren von Jugendlichen in der Universitäts- und Hansestadt Greifswald 1998-2002-2006, online: http://www.rsf.unigreifswald.de/fileadmin/mediapool/lehrstuehle/ duenkel/Schuelerbefragung_HGW_1998_2002_2006.pdf [29.2.2008]. 140

Self-reported crime.book Page 141 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

DÜNKEL F., GEBAUER D., KESTERMANN C., 2005, Mare-Balticum-YouthSurvey: Selbstberichtete Delinquenz und Gewalterfahrungen von Jugendlichen aus Ostseeanrainerstaaten im Überblick, Monatsschrift für Kriminologie und Strafrechtsreform, 88, 215-239. ENZMANN D., WETZELS P., 2003, Ethnic differences in juvenile delinquency: The role of violence legitimizing norms of masculinity, in DÜNKEL, F., DRENKHAHN, K. (Eds), Youth violence: New patterns and local responses – experiences in East and West, Mönchengladbach, Forum Verlag Godesberg, 316-345. ENZMANN D., BRETTFELD K., WETZELS P., 2003, Männlichkeitsnormen und die Kultur der Ehre: empirische Prüfung eines theoretischen Modells zur Erklärung erhöhter Delinquenzraten jugendlicher Migranten, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 43, 264-287. FELTES T., 2006, Kriminalpolitik, in LANGE H.J., GASCH, M. (Hrsg.), Wörterbuch zur Inneren Sicherheit, Wiesbaden, VS Verlag, 160-165 FELTES T., GOLDBERG B., 2006, Selbstberichtete Delinquenz, Viktimisierung und Verbrechensfurcht bei Schülern mit und ohne Migrationshintergrund, in OBERGFELL-FUCHS, J., BRANDENSTEIN, M. (Hrsg.), Nationale und internationale Entwicklungen in der Kriminologie. Festschrift für Helmut Kury zum 65. Geburtstag, Frankfurt, Verlag für Polizeiwissenschaft, 203-237. FUCHS M., LAMNEK S., LUEDTKE J., BAUR N., 2005, Gewalt an Schulen 1994– 1999–2004, Wiesbaden, VS Verlag. GOLDBERG B., 2006, Freizeit und Kriminalität bei Achtklässlern mit und ohne Migrationshintergrund, in FELTES T., PFEIFFER C., STEINHILPER G., (Hrsg.), Kriminalpolitik und ihre wissenschaftlichen Grundlagen. Festschrift für Professor Dr. Hans-Dieter Schwind zum 70. Geburtstag, Heidelberg, C.F. Müller, 2006, 861-892. GÖRGEN T., 2006, "As if I just didn't exist" – Elder abuse and neglect in nursing homes, in WAHIDIN, A., CAIN, M. (Eds), Ageing, crime and society, Cullompton, Willan, 71-89. GÖRGEN T., KREUZER A., KLEIN L., 1995, Neue Befunde aus Gießener Delinquenzbefragungen, Monatsschrift für Kriminologie und Strafrechtsreform, 78, 264-276. GOTTFREDSON M.R., HIRSCHI T., 1990, A general theory of crime,. Stanford, Stanford University Press. GREVE W., ENZMANN D., 2003, Self-esteem maintenance among incarcerated young males: Stabilisation through accommodative processes, International Journal of Behavioral Development, 27, 12-20.

141

Self-reported crime.book Page 142 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

HEITMEYER W., COLLMANN B., CONRADS J., MATUSCHEK I., KRAUL D., KÜHNEL W.,. MÖLLER R., ULBRICH-HERMANN M., 1996, Gewalt: Schattenseiten der Individualisierung bei Jugendlichen aus unterschiedlichen Milieus (2nd ed.), München, Juventa. HERMANN D., WENINGER W., 1999, Das Dunkelfeld in Dunkelfelduntersuchungen: über die Messung selbstberichteter Delinquenz, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 51, 759-766. KÖLLISCH T., 2002, Wie ehrlich berichten Jugendliche über ihr delinquentes Verhalten? Ergebnisse einer externen Validierung selbstberichteter Delinquenz, Freiburg, Max-Planck-Institut für ausländisches und internationales Strafrecht. KÖLLISCH T., 2007, Die Messbarkeit des Erfolgs kriminalpolitischer Interventionen am Beispiel der bundesdeutschen Heroinstudie, in LÖSEL, F., BENDER, D., JEHLE J.M., (Hrsg.), Kriminologie und wissensbasierte Kriminalpolitik: Entwicklungs- und Evaluationsforschung, Mönchengladbach, Forum Verlag Godesberg, 133-145. KÖLLISCH T., KREUZER A., 2006, Qualitative kriminologische Untersuchung des bundesdeutschen Modellprojekts zur heroingestützten Behandlung Opiatabhängiger, Gießen, Institut für Kriminologie an der JustusLiebig-Universität Gießen. KÖLLISCH T., OBERWITTLER D., 2004, Wie ehrlich berichten männliche Jugendliche über ihr delinquentes Verhalten?, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 56, 708-735. KREUZER A., 1975, Schülerbefragungen zur Delinquenz, Recht der Jugend und des Bildungswesens, 23, 229-244. KREUZER A., 1994a, Kriminologische Dunkelfeldforschung, I. Teil: Theorie und Methodik, Neue Zeitschrift für Strafrecht, 14, 10-16. KREUZER A., 1994b, Kriminologische Dunkelfeldforschung, II. Teil: Ergebnisse von Täterbefragungen nach Beispielen neuer Gießener Untersuchungen, Neue Zeitschrift für Strafrecht, 14, 164-168. KREUZER A., GÖRGEN T., KRÜGER R., MÜNCH V., SCHNEIDER H., 1993, Jugenddelinquenz in Ost und West: Vergleichende Untersuchungen bei ostund westdeutschen Studienanfängern in der Tradition Gießener Delinquenzbefragungen, Bonn, Forum. KREUZER A., RÖMER-KLEES R., SCHNEIDER H., 1991, Beschaffungskriminalität Drogenabhängiger (BKA-Forschungsreihe, Bd. 24), Wiesbaden, Bundeskriminalamt. LANGNER W., STURZBECHER D., 1997, "Aufklatschen, plattmachen, Zecken jagen!" – Jugendgewalt in Brandenburg, in STURZBECHER, D., (Hrsg.), Jugend und Gewalt in Ostdeutschland, Göttingen, Verlag für Angewandte Psychologie, 170-208. 142

Self-reported crime.book Page 143 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

LAY B., IHLE W., ESSER G., SCHMIDT M.H., 2005, Juvenile-episodic, continued or adult-onset delinquency? Risk conditions analysed in a cohort of children followed up to the age of 25 years,European Journal of Criminology, 2, 1, 39-66. LINSSEN R., 2005, "Ich mach' heut' blau!". Schulabsentismus: Verbreitung, Ursachen, Interventionen, Die Kriminalprävention, 9, 2, 63-66. LÖBMANN R., 2003, Der Einfluss heroingestützter Therapie auf die Delinquenz Drogenabhängiger, in MINTE, E. (Hrsg.), Neue Konzepte der Kriminalpolitik, Wiesbaden, Kriminologische Zentralstelle, 179-198. LÖBMANN R., 2006, Der Einfluss der Heroinverschreibung auf die Delinquenz Opiatabhängiger, Hannover, Kriminologisches Forschungsinstitut Niedersachsen. LÖSEL F., 1975, Handlungskontrolle und Jugenddelinquenz, Stuttgart, Enke. LÖSEL F., BLIESENER T., 2003, Aggression und Delinquenz unter Jugendlichen, Neuwied, Luchterhand. LÖSEL F., BLIESENER T., BENDER D., 2007, Social information processing, experiences of aggression in social contexts, and aggressive behavior in adolescents, Criminal Justice and Behavior, 34, 3, 330-347. LÜDEMANN C., Massendelikte, 2000, Moral und Sanktionswahrscheinlichkeit: eine Analyse mit den Daten des ALLBUS Soziale Probleme, 2002, 13, 128-155. MAIER-PFEIFFER A., PFEIFFER, C.,, 2006, Frühe Förderung von Kindern aus sozial benachteiligten Familien: das Projekt Pro Kind Niedersachen, in KERNER H.J., MARKS E., (Hrsg.), Internetdokumentation Deutscher Präventionstag Hannover [http://www.praeventionstag.de/content/ 11_praev/doku/maierpfeiffer/index_11_maierpfeiffer.html]. MANSEL J., HURRELMANN K., 1998, Aggressives und delinquentes Verhalten Jugendlicher im Zeitvergleich: Befunde aus 'Dunkelfeldforschungen' aus den Jahren 1988, 1990 und 1996, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 50, 1, 78-109. MANSEL J., RAITHEL J., 2003, Verzerrungsfaktoren im Hell- und Dunkelfeld und die Gewaltentwicklung, in RAITHEL, J., MANSEL, J. (Hrsg.), Kriminalität und Gewalt im Jugendalter: Hell- und Dunkelfeldbefunde im Vergleich,. Weinheim, Juventa, 7-24. MEHLKOP G., BECKER R., 2004, Soziale Schichtung und Delinquenz: eine empirische Anwendung eines Rational-Choice-Ansatzes mit Hilfe von Querschnittsdaten des ALLBUS 1990 und 2000, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 56, 95-126. MENZEL B., PETERS H., 1998, "Self-Reports" taugen wenig für objektive Vergleiche: über einen problematischen Versuch, die Häufigkeitsentwicklung von Delinquenz zu analysieren, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 50, 4, 560-564. 143

Self-reported crime.book Page 144 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

MOFFITT T.E., 1993, Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy, Psychological Review, 100, 674-701. MÖSSLE T., KLEIMANN M., REHBEIN F., PFEIFFER C., 2006, Mediennutzung, Schulerfolg, Jugendgewalt und die Krise der Jungen, Zeitschrift für Jugendkriminalrecht und Jugendhilfe, 3, 295-309. NAPLAVA T., OBERWITTLER D., 2002, Methodeneffekte bei der Messung selbstberichteter Delinquenz von männlichen Jugendlichen, Monatsschrift für Kriminologie und Strafrechtsreform, 85, 6, 401-423. NISBETT R.E., COHEN D.,1996, Culture of honor: the psychology of violence in the south, Oxford, Westview. OBERWITTLER D., 2003, Geschlecht, Ethnizität und sozialräumliche Benachteiligung – überraschende Interaktionen bei sozialen Bedingungsfaktoren von Gewalt und schwerer Eigentumsdelinquenz von Jugendlichen, in LAMNEK S., BOATCA M. (Hrsg.), Geschlecht – Gewalt – Gesellschaft, Opladen, Leske + Budrich, 269-294. OBERWITTLER D., 2004, A multilevel analysis of neighbourhood contextual effects on serious juvenile offending: The role of subcultural values and social disorganization, European Journal of Criminology, 1, 2, 201-235. OBERWITTLER D., BLANK T., KÖLLISCH T., NAPLAVA T. 2001, Soziale Lebenslagen und Delinquenz von Jugendlichen. Arbeitsberichte 1/2001. Freiburg: Max-Planck-Institut für Strafrecht. OBERWITTLER D., KÖLLISCH T., 2003, Jugendkriminalität in Stadt und Land. Sozialräumliche Unterschiede im Delinquenzverhalten und Registrierungsrisiko, in RAITHEL J., MANSEL J., (Hrsg.), Kriminalität und Gewalt im Jugendalter: Hell- und Dunkelfeldbefunde im Vergleich,), Weinheim, Juventa, 135-160. OBERWITTLER D., KÖLLISCH T., 2004, Nicht die Jugendgewalt sondern deren polizeiliche Registrierung hat zugenommen, Neue Kriminalpolitik, 16, 4, 144-147. PFEIFFER C., 2004, Die Dämonisierung des Bösen, Frankfurter Allgemeine Zeitung, 05.03, 9. PFEIFFER C., WETZELS P., 2006, Empirisch-kriminologische Forschung und Kriminalpolitik: das Beispiel Jugendgewalt, in FELTES T., PFEIFFER C., STEINHILPER G., (Hrsg.), Kriminalpolitik und ihre wissenschaftlichen Grundlagen: Festschrift für Hans-Dieter Schwind, Heidelberg, C.F. Müller, 1095-1127. POSNER C., 1997, Die Täterbefragung im Kontext „Sozialer Umbruch und Kriminalität“, in BOERS, K., GUTSCHE, G., SESSAR, K., (Hrsg.), Sozialer Umbruch und Kriminalität in Deutschland, Opladen, Westdeutscher Verlag, 157-185.

144

Self-reported crime.book Page 145 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

QUENSEL S., 1971, Delinquenzbelastung und soziale Schicht bei nicht bestraften männlichen Jugendlichen, Monatsschrift für Kriminologie und Strafrechtsreform, 54, 236-261. QUENSEL S., QUENSEL E., 1969, Lässt sich die Delinquenzbelastung messen?, Kriminologisches Journal, 1, 4-25. RABOLD S., GÖRGEN T., 2007, Misshandlung und Vernachlässigung älterer Menschen durch ambulante Pflegekräfte: Ergebnisse einer Befragung von Mitarbeiterinnen und Mitarbeitern ambulanter Dienste, Zeitschrift für Gerontologie und Geriatrie, 40, 366-374. REUBAND K.H., 2004, Kriminalitätsentwicklung und Medienwirklichkeit, in WALTER M., KANIA H., ALBRECHT H.-J., (Hrsg.), Alltagsvorstellungen von Kriminalität, Münster, Lit, 235-248. SAMPSON R.J., LAUB H.H., 1993, Crime in the making: Pathways and turning points through life, Cambridge, M. A., Harvard University Press. SCHMIDT M.H., LAY, B., IHLE W., ESSER, G., 2001, Bedeutung von Dunkelfelddelikten für episodische und fortgesetzte Delinquenz, Monatsschrift für Kriminologie und Strafrechtsreform, 84, 1, 25-36. SCHUMANN K.F., (Hrsg.), 2003, Delinquenz im Lebensverlauf. Bremer Längsschnittstudie zum Übergang von der Schule in den Beruf bei ehemaligen Hauptschülern, Bd 2, Weinheim, Juventa. SCHUMANN K.F., 2004, Sind Arbeitsbiographie und Straffälligkeit miteinander verknüpft? Aufklärungen durch die Lebenslaufforschung, Monatsschrift für Kriminologie und Strafrechtsreform 87, 222-243. SCHUMANN, K.F., 2007, Berufsbildung, Arbeit und Delinquenz: empirische Erkenntnisse und praktische Folgerungen aus einer Bremer Längsschnittstudie, in DESSECKER, A., (Hrsg.), Jugendarbeitslosigkeit und Kriminalität, 2. Aufl., Wiesbaden, Eigenverlag Kriminologische Zentralstelle, 43-68. SCHWIND H.D., 2007, Kriminologie: eine praxisorientierte Einführung mit Beispielen (17. Aufl.), Heidelberg, Kriminalistik Verlag. SCHWIND H.D., FREIER D., BALLERING P., 2002, Studentenbefragungen zum Dunkelfeld im Langzeitvergleich. Göttingen – Osnabrück – Bochum 1973 – 1995 – 2001, Die Kriminalprävention, 3, 99-104. SCHWIND H.D., ROITSCH K., AHLBORN W., GIELEN, B., 1995, Gewalt in der Schule: am Beispiel von Bochum, Mainz, Weisser Ring. SCHWIND H.D., ROITSCH K., GIELEN B., 1995, Gewalt in Schulen: die Bochumer Studie zur Gewalt in Schulen, Kriminalistik, 49, 10, 618-625. STELLY W., THOMAS J., 2007, Das Ende der kriminellen Karrieren bei jugendlichen Mehrfachtätern, in LÖSEL F., BENDER D., JEHLE J.M., (Hrsg.), Kriminologie und wissensbasierte Kriminalpolitik. Entwicklungs- und Evaluationsforschung,. Mönchengladbach, Forum Verlag Godesberg, 433-446. 145

Self-reported crime.book Page 146 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

SUTTERER P., KARGER T., 1994, Self-reported juvenile delinquency in Mannheim, Germany, in JUNGER-TAS J., TERLOUW G.J., KLEIN M.W. (Eds), Delinquent behavior among young people in the Western world, Amsterdam, Kluger, 156-185. TILLMANN, K.J., HOLLER-NOWITZKI, B., HOLTAPPELS, H.G., MEIER, U., POPP, U., Schülergewalt als Schulproblem: verursachende Bedingungen, Erscheinungsformen und pädagogische Handlungsperspektiven, Weinheim, Juventa, 2000. WALTER M., 2000, Von der Kriminalität in den Medien zu einer Bedrohung durch Medienkriminalität und Medienkriminologie?, in BUNDESMINISTERIUM DER JUSTIZ, (Hrsg.), Kriminalität in den Medien, Mönchengladbach, Forum Verlag Godesberg, 10-22. WALTER M., 2003, Mehrfach- und Intensivtäter: kriminologische Tatsache oder Erfindung. der Medien?, Zeitschrift für Jugendkriminalrecht und Jugendhilfe, 13, 159-163. WEINS C., REINECKE J., 2007, Delinquenzverläufe im Jugendalter: eine methodologische Analyse zur Auswirkung von fehlenden Werten im Längsschnitt, Monatsschrift für Kriminologie und Strafrechtsreform, 90, 5, 418-437. WETZELS P., ENZMANN D., MECKLENBURG E., PFEIFFER C., 2001, Jugend und Gewalt: eine repräsentative Dunkelfeldanalyse in München und acht anderen deutschen Städten, Baden-Baden, Nomos. WILMERS N., ENZMANN D., SCHÄFER D., HERBERS K., GREVE W., WETZELS P., 2002, Jugendliche in Deutschland zur Jahrtausendwende: Gefährlich oder gefährdet? Ergebnisse wiederholter, repräsentativer Dunkelfelduntersuchungen zu Gewalt und Kriminalität im Leben junger Menschen 1998–2000, Baden-Baden, Nomos. WINDZIO M., 2006, Is there a deterrent effect of pains of imprisonment? The impact of "social costs" of first incarceration on the hazard rate of recidivism, Punishment & Society, 8, 341-364. WITTICH U., GÖRGEN T., KREUZER A., 1998, Wenn zwei das gleiche berichten.... Beitrag zur kriminologischen Dunkelfeldforschung durch vergleichende Delinquenzbefragungen bei Studenten und Strafgefangenen, Mönchengladbach, Forum-Verlag Godesberg.

146

앫 앫 앫 앫 앫

앫 fear of crime 앫 victimisation (in general, school and family) 앫 delinquency 앫 drug use 앫 school absenteeism

Criminological mainly 2000 Research Institute funded by of Lower Saxony municipalities Wilmers, Enzmann, Schäfer, Herbers, Greve, Wetzels (2002) “Schülerbefragung“ (Student survey) München Hamburg Hannover Leipzig Friesland

München Hamburg Kiel Hannover Leipzig Stuttgart Schwäbisch Gmünd 앫 Wunstdorf 앫 Lilienthal

Geographical coverage 앫 앫 앫 앫 앫 앫 앫

Date Main focus of of the study survey 앫 fear of crime 앫 victimisation (in general, school and family) 앫 delinquency 앫 drug use

Funding organisation

Criminological mainly 1998 Research Institute funded by of Lower Saxony municipalities Wetzels, Enzmann, Mecklenburg, Pfeiffer (2001) Schülerbefragung (Student survey)

Institution/Principal investigators/ Study Title

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting, burglary, harassment, vandalism, threat with weapon, car theft 앫 based on Delinquency Scale (Lösel 1975)

앫 N=10,460 앫 cross-sectional 앫 50.7 males, 앫 survey of 9th graders 앫 self-administered ques49.3 % females tionnaire 앫 Mean age=15.2 앫 response rate: 87.7 % (of (range 13-19) pupils in participating schools)

Measures of delinquency

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting, burglary, harassment, vandalism, threat with weapon, car theft 앫 based on Delinquency Scale (Lösel 1975)

Sample

앫 N=16,190 앫 51.7 % males, 48.3 % females 앫 Mean age=15.2

앫 cross-sectional 앫 survey of 9th graders 앫 self-administered questionnaire 앫 response rate: 81.2 %

Method and response rate

Appendix: Table 1: Overview of main German SRD studies since 1990

Self-reported crime.book Page 147 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

147

148 2005

ministry of culture (Thuringia)

Criminological Research Institute of Lower Saxony Baier, Pfeiffer, Windzio, Rabold (2005) Schülerbefragung (Student survey)

앫 victimisation (in general, school and family) 앫 delinquency 앫 drug use 앫 school absenteeism

앫 victimisation (in general, school and family) 앫 delinquency 앫 drug use 앫 truancy

Date Main focus of of the study survey

mainly 2005 funded by municipalities

Funding organisation

Criminological Research Institute of Lower Saxony Baier, Pfeiffer, Windzio, Rabold (2005) Schülerbefragung (Student survey)

Institution/Principal investigators/ Study Title

Self-reported crime.book Page 148 Thursday, April 23, 2009 8:02 AM

München Stuttgart Dortmund Kassel Oldenburg Schwäbisch Gemünd Peine (district) Soltau-Fallingbostel (district) Wallenhorst/Belm Lehrte (district)

앫 Thuringia (federal state)

앫

앫

앫

앫

앫 앫 앫 앫 앫 앫

Geographical coverage

앫 cross-sectional 앫 survey of 4th and 9th graders 앫 self-administered questionnaire 앫 response rate 9th grade: 85.1 % 앫 response rate 4th grade: 74.8 %

앫 cross-sectional 앫 survey of 4th and 9th graders 앫 self-administered questionnaire 앫 response rate 9th grade: 73.8 % 앫 response rate 4th grade: 71.4 %

Method and response rate

Survey of 9th grade pupils 앫 N=2,720 앫 47.1 % males, 52.9 % females 앫 Mean age=14.9 Survey of 4th grade pupils 앫 N=613 앫 49.5 % males, 50.5 % females 앫 Mean age=10.4

Survey of 9th grade pupils 앫 N=14,301 앫 49.8 % males, 50.2 % females 앫 Mean age=15.1 Survey of 4th grade pupils 앫 N=5,529 앫 49.9 males, 50.1 % females 앫 Mean age=10.3

Sample

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting,, break and enter (car or building), vandalism, threat with weapon, car theft, fare dodging, driving without licence, graffiti

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting, break and enter (car or building), vandalism, threat with weapon, car theft, fare dodging, driving without licence, graffiti 앫 4th grade only school violence (12 months prevalence and incidence): hit or beat sb., tease sb., destroy sb’s property., blackmail sb. to give money or other things

Measures of delinquency

Self-Reported Crime and Deviance Studies in Europe

Crime 1998 Prevention 2002 Council of 2006 MecklenburgWestern Pomerania

2003 2004

Greifswald University Dünkel, Gebauer, Geng (2007) Mare Balticum Youth Survey

Bochum University Feltes, Goldberg (2006)

victimisation delinquency drug use xenophobia

앫 victimisation 앫 delinquency

앫 앫 앫 앫

Date Main focus of of the study survey

Funding organisation

Institution/Principal investigators/ Study Title

Self-reported crime.book Page 149 Thursday, April 23, 2009 8:02 AM

Bochum and Herne

Greifswald

Geographical coverage

앫 12 months prevalence 앫 robbery, coercion, assault, use of weapons, fare dodging, graffiti-spraying, sale of pirate copies, theft, insult, fun fighting / nonhostile fighting

앫 N=4,028 앫 cross-sectional 앫 49.3 % males, 앫 survey of 8th graders 앫 self-administered ques50.7 % females tionnaire 앫 Mean age=13.9 앫 complete surveys 앫 response rate 8th grade: 98.,5 % (of pupils in participating schools)

Measures of delinquency

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting, break and enter (car or building), vandalism, threat with weapon, car theft, fare dodging, driving without licence, graffiti, sale of pirate copies, theft, selling drugs, severe harassment, arson

Sample

1998 앫 N=1,529 앫 49.2 % males, 50.8 % females 앫 Mean age=15.0 (age range 1317) 2002 앫 N=724 앫 49.4 % males, 50.6 % females 앫 Mean age=15.3 (age range 1418) 2006 앫 N=832 앫 48.2 % males, 51.8 % females 앫 Mean age=15.2 (age range 1318)

앫 cross-sectional 앫 survey of 9th graders 앫 self-administered questionnaire 앫 complete surveys 앫 response rate: 86.4 % (1998), 87.9 % (2002), 81.7 % (2006)

Method and response rate

Self-Reported Delinquency Studies in Germany

149

1999

Max-Planck-Insti- German tute for Foreign and Research International Penal Foundation Law Oberwittler, Naplava, Blank, Köllisch Soziale Probleme und Jugenddelinquenz im sozialökologischen Kontext (Social problems and juvenile delinquency in ecological perspective)

Freiburg, Köln 앫 cross-sectional 앫 survey of 8th, 9th and 10th graders 앫 self-administered questionnaire 앫 response rate (Freiburg): 86.,1 % (of pupils in participating schools) 앫 response rate (Köln): 84.9 % (of pupils in participating schools)

앫 delinquency 앫 impact of social context (neighborhoods) on delinquency

앫 cross-sectional 앫 face-to-face interviews 앫 response rate: 45.2 %

Method and response rate

Mannheim

Geographical coverage

앫 delinquency

Date Main focus of of the study survey 1991

Funding organisation

Max-Planck-Institute for Foreign and International Penal Law Sutterer, Karger (1994)

Institution/Principal investigators/ Study Title

Self-reported crime.book Page 150 Thursday, April 23, 2009 8:02 AM

150 Freiburg 앫 N=1,886 앫 46.8 % males, 52.2 % females Köln 앫 N=3,445 앫 46.5 % males, 53.0 % females

앫 N=300 앫 60.7 % males, 39.3 % females 앫 Age range: 14 to 20 years

Sample

앫 life time prevalence 앫 prevalence/incidence last 12 months 앫 sale of pirate copies, graffiti-spraying, vandalism, break and enter (car or building),shoplifting, theft, drug dealing, assault, severe assault, blackmail, robbery, car theft

앫 based on ISRD-Instrument

Measures of delinquency

Self-Reported Crime and Deviance Studies in Europe

20002003

Münster University German / Bielefeld Univer- Research sity Foundation Boers, Reinecke Jugendkriminalität in der modernen Stadt (Juvenile crime in the modern city)

Measures of delinquency

Münster

앫 longitudinal (7th grade) 앫 7th grade (sample 앫 life time prevalence 앫 self-administered quescharacteristics at 앫 12 months prevalence/incitionnaire t1 for sample dence 앫 additional official police retained at t4, 앫 robbery, blackmail, assault, data 2003) shoplifting, break and 앫 response rate in 2000: 앫 N=813 enter (car or building), 69.0 % 앫 43 % males, vandalism, car theft, theft, 앫 56.1 % drop outs in com57 % females handbag robbery, selling parison to original drugs, drug offences cohort

Sample

앫 delinquency

Method and response rate

앫 N=2,683 앫 cross-sectional 앫 53.1 % males, 앫 survey of 9th and 10th graders 49.9 % females 앫 self-administered ques앫 Mean age=15.7 tionnaire 앫 additional computerassisted telephone interviews of the muslim population, self-administered questionnaires for students and qualitative interviews of Islamic organisations 앫 response rate: 79.9 % (of pupils in participating schools)

Geographical coverage

앫 integration of Hamburg, muslims Köln, 앫 attitudes of Augsburg Muslims towards their religion, democracy and politically motivated violence 앫 SRD: robbery, assault without weapon, threat with weapon, extortion

Date Main focus of of the study survey 2005

Funding organisation

Hamburg UniverFederal sity Ministry of Brettfeld, Wetzels the Interior (2007) Muslime in Deutschland (Muslims in Germany)

Institution/Principal investigators/ Study Title

Self-reported crime.book Page 151 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency Studies in Germany

151

Funding organisation

Münster University German / Bielefeld Univer- Research sity Foundation Boers, Reinecke Jugendkriminalität in der modernen Stadt (Juvenile crime in the modern city)

Institution/Principal investigators/ Study Title 20022006

delinquency

Date Main focus of of the study survey

Self-reported crime.book Page 152 Thursday, April 23, 2009 8:02 AM

Duisburg

Geographical coverage 앫 longitudinal 앫 survey of 7th and 9th graders (only 2 measurements 9th grade) 앫 self-administered questionnaire 앫 response rate 7th grade in 2002: 60. % 앫 48.1 % drop outs (status: 2005) in comparison to original cohort 앫 response rate 9th grade in 2002: 50.4% 앫 34.1 % drop outs (2003) in comparison to original cohort

Method and response rate

7th grade (sample characteristics at t1 for sample retained at t4, 2005) 앫 N=1,715 앫 43.8 % males, 56.2 % females 앫 Mean age=12.9 9th grade (sample characteristics at t1 for sample retained at t2, 2003): 앫 N=1,732 앫 46.1 % males, 53.9 % females 앫 Mean age=15.0

Sample

앫 life time prevalence 앫 12 months prevalence/incidence 앫 robbery, blackmail, assault, shoplifting, break and enter (car or building), vandalism, car theft, theft, handbag robbery, selling drugs, drug offences

Measures of delinquency

Self-Reported Crime and Deviance Studies in Europe

152 .

Funding organisation

German Research Foundation,

Institution/Principal investigators/ Study Title

Schumann et al. (2003) Berufsbildung, Arbeit und Devianz (Vocational education, work, and deviance)

19881989 19921993 1995 1997 2000

interactions between educational and work biography, delinquency and interventions by the criminal justice system

Date Main focus of of the study survey

Self-reported crime.book Page 153 Thursday, April 23, 2009 8:02 AM

Bremen

Geographical coverage

Sample

앫 N=424 (1992/ 1993; sample characteristics at t1) 앫 56.6 % males, 43.4 % females 앫 Mean age=16.8 앫 N=333 (2000)

Method and response rate

앫 longitudinal 앫 face-to face interviews, self-administered questionnaire for delinquency questions 앫 official data on courtrecorded crime 앫 response rate first wave (survey of potential graduates): 58.2 % 앫 21.5 % drop outs in comparison to original cohort (N=424)

앫 12 months prevalence/incidence 앫 robbery, theft at workplace, handling stolen goods, assault, severe assault, shoplifting, fraud, theft from vehicle, vehicle theft, sale of pirate copies, theft, selling drugs, using goods / devices without owner's permission, driving without licence, illegal possession of firearms, vandalism, forgery of documents, unlawful entry, hitand-run driving, driving under the influence, illegal arms trade

Measures of delinquency

Self-Reported Delinquency Studies in Germany

153

Federal Ministry for Education, Science and Technology, German Research Foundation,

Federal Criminal Police Office

Erlangen-Nuremberg-University Lösel, Bliesener (2003)

Funding organisation

Lay, Ihle, Esser, Schmidt (2001) Risk conditions and developmental patterns of mental disorders from childhood to early adulthood

Institution/Principal investigators/ Study Title

154 2003 2005

19771996

Mannheim

Erlangen, Nuremberg

앫 victimisation 앫 delinquency 앫 social information processing

Geographical coverage

delinquency

Date Main focus of of the study survey

Self-reported crime.book Page 154 Thursday, April 23, 2009 8:02 AM

앫 12 months prevalence/incidence 앫 delinquency Scale (Lösel 1975), property and violence offences 앫 teacher reports about physical and verbal aggression and general aggressiveness

앫 cross-sectional (N=1163) combined with longitudinal (N=102), 7th and 8th graders 앫 self-administered questionnaire 앫 response rate: approx. 91.5 % (of pupils in participating schools)

앫 N=1,163 앫 52.8 % males, 47.2 % females, 앫 Mean age=14.0 앫 Follow up (20 months later) with 102 males

Measures of delinquency

앫 prevalence/incidence since last survey 앫 prevalence/incidence since last survey 앫 fare dodging, driving without a licence, evasion of entrance fees (cinema, concert), retaining found/borrowed goods, carrying weapons, forgery, selling stolen goods, burglary (adding and abetting), blackmail, arson, slitting car tyres, damaging fuel tanks, reporting innocent persons to police

Sample

앫 longitudinal 앫 N=321 (sample 앫 self-administered quescharacteristics at tionnaire and structured t1) interviews 앫 55.8 % males, 앫 survey of parents at age 44.2 % females 8, 13, 18 and 25 of their children 앫 survey of 13, 18 and 25 year-old persons 앫 additional official police data 앫 19,5 % drop outs in comparison to original cohort

Method and response rate

Self-Reported Crime and Deviance Studies in Europe

Self-reported crime.book Page 155 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland Susan McVie

This report presents evidence on the existing research from Great Britain (England, Scotland and Wales) and Ireland on self-report delinquency (SRD) surveys. Key aims of the report are to summarise the main features of the research from these four jurisdictions and to highlight issues of good practice and methodological excellence. The information contained in this report has been compiled through a mixture of literature reviews, on-line searches and personal contributions by many experienced researchers and distinguished academics. The report represents an extensive (but not exhaustive) review of around 30 major self-report delinquency studies carried out in Great Britain (GB) and Ireland since the 1960s (see Appendix 1). The report focuses on largescale studies (including at least several hundred participants), conducted at either national or local level, which have a primary focus on offending behaviour. The report does not contain details of the many self-report drug-use surveys carried out in this jurisdiction, or the large number of smaller studies that have also contributed to this field. In addition, it was not possible within this scope of this report to provide a review of the research findings from these studies, which are various and wide-ranging. This report is structured in five parts. The first section provides a brief introductory overview of the historical development of SRD surveys carried out in Great Britain and Ireland. Section two gives a more detailed synopsis of these studies, including their geographical coverage, age focus, funding sources, aims, theoretical developments and policy focus. The third section discusses some of the methodological issues raised by SRD studies, such as survey content, methodological integrity, issues of validity and reliability, and other examples of good practice. This is followed by a section on the usage of survey data by Government (local and national) and other organisations in informing decision making or development of public policy, academic and non-academic dissemination and media coverage. The final section ends with some general conclusions and recommendations about effective methodology and good practice for these types of study.

155

Self-reported crime.book Page 156 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

I.

Historical overview

Measurement of crime in Great Britain and Ireland has largely relied upon official statistics (particularly police recorded crime rates and criminal convictions data) and national victimisation surveys (in particular, the British, Scottish and Northern Ireland Crime Surveys). Nevertheless, these countries have an established record in conducting SRD surveys and can boast some of the most seminal studies in this field. Since the early 1960s, there have been in the region of 30 major offending studies carried out in the jurisdictions of England, Scotland, Wales and Ireland, many of which have adopted a longitudinal approach and therefore, in themselves, represent several separate surveys. Taken together, the surveys reported here have collected data from more than 140,000 individuals and contributed a vast body of evidence on the extent, nature, correlates and causes of criminal offending in these four countries. Appendix 1 provides a series of tables presenting general information about each of these studies, in chronological order, including the main researcher(s) involved, when the surveys were carried out, the main focus of the study, geographical coverage, sample size, methods employed and (where known) the response rates. The first major SRD surveys carried out in Great Britain in the 1960s were, to a large extent, inspired by the proliferation of American studies which started in the 1940s but which gained popularity in the 1950s with the publication of research by Short and Nye (1957). The first major study to be carried out in England, the Cambridge Study in Delinquent Development, is perhaps the most enduring longitudinal study of offending world-wide and has contributed significantly to the field of developmental and life-course criminology (Farrington, 1995). Although it might be argued that it has some limitations compared to the standards of contemporary survey methodology (it has a small sample size, focuses exclusively on males and has considerable sample bias), the Cambridge Study has provided a model of criminal careers research that few other studies have managed to achieve. The breadth of data collected from a wide range of sources, the quality of the information collected and the longevity of the study, which spans a 40 year period, has earned it the distinction of being one of the most pioneering studies in criminology. Although the earliest study dates back to 1961, Figure 1 shows that the bulk of SRD surveys in Great Britain and Ireland have taken place far more recently, towards the late 1990s and into the 2000s (Appendix 1 provides specific details). The number of studies emerging during the period of the New Labour government (since 1997) is particularly noteworthy, and a point that shall be returned to later in this report. Over time, there have 156

Self-reported crime.book Page 157 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

been some fairly significant changes in survey methodology and design. The very earliest studies tended to focus only on males, although from 1975 onwards most of the SRD research in Great Britain and Ireland has included both male and female participants. Sample sizes in the earliest studies were generally restricted to a few hundred (the survey by Belson being an exception); whereas, the more recent surveys, even those restricted to sub-national level, have tended to aim for much larger sample sizes. In addition, there has been a definite trend towards longitudinal and repeated cross-sectional, rather than snapshot, surveys during the last 10 years. The potential for understanding the problem of criminal offending has never been greater, particularly with the significant developments in analytical techniques which now permit complex statistical analyses and data modelling.

Figure 1: Number of SRD surveys commencing in GB and Ireland since 1960

1.

Geographical coverage

With one or two notable exceptions, the early SRD studies in Great Britain and Ireland were most commonly limited to local geographies i.e. based in cities or sub-parts of cities. The trend since the early 1990s has been towards far more national surveys, however. Figure 2, below, illustrates the territorial scope of these surveys across the four jurisdictions of Great Britain and Ireland. It is evident that the majority of research has involved either national or local surveys within England. Wales is well represented amongst the national surveys, although this is largely due to Wales being ‘added-on’ to English surveys, which are designed, administered and analysed by within the Home Office. Wales has had few independent SRD surveys. Ireland has been particularly poorly represented in SRD terms, although this has improved in very recent years with the administration of two independently organised national surveys. Scotland has had eight SRD

157

Self-reported crime.book Page 158 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

studies in total, of which four have been national studies. However, only one of the national surveys was an independent Scottish survey, and then it was principally a study of victimisation rather than offending. Overall, England has the biggest share of the funding, and therefore the greatest input into how research is designed and what theoretical and/or policy orientation it has taken.

Figure 2: Geographical scope of the SRD surveys in Britain and Ireland

2.

Age focus

One characteristic of these surveys that has changed little over time is the predominant focus on young people in their early to mid teenage years. Figure 3 presents a summary of the number of studies carried out in Britain and Ireland that have included respondents between ages 5 and 30+. The emerging ‘age-crime-survey curve’ indicates that most of the SRD surveys have focused their attention on early adolescents. This is reflected in the sampling techniques that have largely involved either a census approach to sampling school years or random samples of schools within local authority areas. Exceptions to this include the early Irish studies by McQuoid and the Home Office funded studies by Graham and Bowling (1995), Flood-Page et al. (2000) and Budd et al. (2005), all of which included a broader age range in their samples. The Sheffield Pathways out of Crime Study (Bottoms et al. 2004) and the study of ‘crimes of everyday life’ by Karstedt and Farrall (2006) are unusual in that they have focused specifically on offending in the adult years.

158

Self-reported crime.book Page 159 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

Figure 3: Age focus of SRD surveys in Britain and Ireland

The focus on the early to mid teenage years reflects an increasing preoccupation since the early 1990s within both developmental criminology and youth justice policy in England and Wales, as well as elsewhere, with the risk factor prevention paradigm. The notion of ‘early intervention’ for offenders has led to a predominant focus on collecting data which may identify risk and protection factors at or prior to the peak years of the age-crime curve to aid our construction of predictive models of offending at this limited time point. It is arguable that this focus has rather obscured the development of longer term approaches to understanding the causal processes and developmental mechanisms associated with change and stability in offending over the life course such as that advocated by Sampson and Laub (1993). The recently developed longitudinal studies are in their infancy, however, and it remains to be seen whether they will receive the necessary funding to facilitate such a lifecourse approach. The Cambridge Study has been instrumental in advancing theories about criminal careers, and the bulk of its contribution to the literature has focused on the predictive role of the early risk factors (for a full review of publications from the Study see Farrington et al. 2006). While there are arguments for and against this paradigm (see Case 2007) it is clear that most of the new breed of longitudinal SRD surveys in Great Britain and Ireland have also focused in broadly this direction. Nevertheless, figure 3 indicates that relatively few studies have included children under the age of 10 in their samples, so it is difficult to see how far such theories can develop.

159

Self-reported crime.book Page 160 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

3.

Sources of funding

The surveys in Great Britain and Ireland have been funded by a diverse range of organisations; however, central government has proved to be the biggest sponsors of SRD surveys, particularly in England and Wales, ensuring their development and continuation. The Home Office has funded, either in whole or part, many studies aimed at assisting in the development and/or evaluation of criminal justice policy, such as the early youth lifestyles surveys (Graham, Bowling, 1995 ; Flood-Page et al. 2000), the Offending Crime and Justice Survey (Budd et al., 2005b), the Prisoner and Community Sentence Criminality Surveys (Budd et al., 2005a), the On Track youth lifestyles surveys (Armstrong et al. 2005) and, of course, the Cambridge Study (Farrington et al., 2006). The cross-disciplinary importance of some of the SRD surveys is also evident, with several receiving funding from the Department of Health, such as the Belfast Youth Development Survey (McCrystal et al., 2006), the Mental Health study of children and young people (Green et al., 2005) and, again, the Cambridge Study. Government departments in both Scotland and Ireland have also been major sponsors of SRD studies, although Scotland is the only country of the four to have failed to receive any government funding for a national SRD study. Research councils, including the Economic and Social Research Council, and charitable organisations, such as the Nuffield Foundation and the Joseph Rowntree Foundation, have contributed to the funding of several major SRD studies, including the Edinburgh Study of Youth Transitions and Crime (Smith et al., 2001), the Peterborough Adolescent and Young Adult Development Study (Wikstrom, Sampson, 2006), SPOOCS (Bottoms et al., 2004) and the Communities that Care Youth Survey (Anderson et al., 2002). On the whole, however, the major sponsors of SRD studies in the United Kingdom have been local or national government, which clearly has implications for the design and focus of these surveys (see below). The funding of longitudinal studies is something of a contentious issue in Great Britain and Ireland. Although there is strong support from research funders to sponsor longitudinal studies, almost all of the current longitudinal studies in this jurisdiction have had to rely on sponsorship from more than one source. At present, there is no mechanism for continued programmatic funding for longitudinal SRD studies and, as a result, it is unclear whether all of the current research studies will continue much beyond the teenage years of the individual cohorts involved.

160

Self-reported crime.book Page 161 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

4.

Survey aims

The specific aims of the SRD studies carried out in Great Britain and Ireland vary widely, and are too lengthy to describe in detail in the context of this report. However, a number of broad themes emerge from the more recent studies that are worthy of note here. Around a third of the surveys aim to ‘measure the prevalence and frequency of offending’ amongst their sampled populations, particularly those surveys that are funded directly by Government. However, a larger proportion of the SRD surveys listed in Appendix 1 are concerned with ‘exploring the causal factors or mechanisms associated with offending’ and of contributing to a deeper criminological understanding in this field. Most of the longitudinal studies aim to contribute to theory development, although this is a less common objective for the cross-sectional surveys. A significant number of studies aim to identify early risk and protective factors that might be valuable in the development of effective youth crime policy and the design of targeted prevention initiatives; however, few give the aim of pursuing a life-course investigation. Although those studies that are funded by Government mention broad aims about contributing to policy, these rarely provide a specific indication as to what this will involve. An exception is the On Track youth survey, which aimed to specifically evaluate the implementation of this major government initiative (Armstrong et al., 2005).

5.

Theoretical development

Several of the SRD studies carried out in Great Britain and Ireland claim to be working towards alternative theoretical explanations for offending and/ or antisocial behaviour. Four of the studies mentioned in this report have published details of new or emerging theories based on their research. These four perspectives are summarised below : 1. Although it was not originally designed to test any one theory of offending, the Cambridge Study in Delinquent Development has facilitated the development of the integrated cognitive antisocial potential (ICAP) theory to explain the importance of different risk and protective factors (Farrington, 2005). The ICAP theory is a synthesis of other theoretical perspectives, including strain, labelling, control, learning and rational choice theory. It relates to the potential, rather than the propensity, of an individual to commit an antisocial act and assumes that the transition from potential to actual offending relies on individual cognitive processes.

161

Self-reported crime.book Page 162 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

2. Wikstrom is using the PADS+ study (part of the SCoPiC research network) to further develop his Situational Action Theory (Wikstrom, 2005). This theory combines individual and environmental explanations of crime. It attempts to explain offending by addressing the complex interactions between individual characteristics (such as moral values and executive capabilities which influence choice) and environmental characteristics (including external monitoring, human friction and available temptation), in the context of different situational mechanisms (e.g. the individual’s perception of alternatives). The crime prevention implications of this theory focus largely on reducing individual propensity to offend by influencing moral values and enhancing individual capacity to exercise self-control. 3. Like the Cambridge Study, the Edinburgh Study of Youth Transitions and Crime was not designed to test any one theory of offending, but rather develop a synthesis of existing theories. The Edinburgh Study uses Hirschi’s social control theory as a foundation, but supplements this with aspects taken from a range of other theoretical traditions including Lemert’s labelling theory, Bandura’s social learning theory and Sutherland’s differential association theory (Smith, McVie, 2003). The emerging theoretical model is one of Negotiated Order, where emphasis is placed on engagement and interactions with the broader social environment as the key basis for understanding individual turning points, particularly in the context of formal and informal social controls (McAra, 2006). 4. The Belfast Youth Development Study is mainly influenced by a combination of primary socialisation theory, social network theory, and stage sequential models of drug use development. The research team are currently in the process of developing a new theoretical perspective which focuses on the role of self-regulation in developing moderate (controlled) alcohol and drug use patterns amongst teenagers (Percy, in press).

6.

Policy focus

Given the extent of Government funding for SRD surveys in Great Britain and Ireland, it is not surprising that these studies have declared a strong policy focus. Even those that have been funded by Research Councils and other funders, however, have tended to intimate their value in policy terms. Almost universally, these studies have proposed to inform policy by furthering our understanding of the causes of a range of offending and delinquent behaviours and to make recommendations for effective crime prevention strategies and interventions. The nature of these policy recommendations is not always very specific, although a good number

162

Self-reported crime.book Page 163 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

have highlighted the identification of early risk factors as a primary focus. Several studies have identified particular contexts which are of primary interest; for example, the social environment (PADS+), the family and neighbourhood (E-Risk) and the agencies of formal social control (the Edinburgh Study). While some studies have clearly been aimed at informing policy at the national or universal level (e.g. the OCJS, the Northern Ireland surveys, the Welsh Youth Survey, the national CTC survey and the MORI youth surveys), other SRD studies in Great Britain and Ireland have also made efforts to develop effective policy recommendations at the local level. For example, the Edinburgh Study, PADS+, BYDS and SPOOCS have all secured support (and in some cases financial backing) from local authorities and criminal justice organisations on quid pro quo that they feed back into the policy processes of these local agencies. In addition to its national survey, the Communities that Care (CTC) organisation, through the support of the Joseph Rowntree Foundation, has carried out a large number of local SRD surveys among young people in an attempt to develop targeted programmes of prevention specific to the needs to local residents. In the United Kingdom, the national election of the New Labour government in 1997 was closely followed by the establishment of devolved parliaments in Wales and Scotland in 1999, and later in Northern Ireland in 2007. The recent national SRD surveys carried out in Northern Ireland and Wales are, therefore, very strategically important to understanding the patterns and profiles of crime and delinquency in these United Kingdom regions compared to England. Following devolution in Scotland and Wales, youth crime policies have followed very different tracks. Since devolution, policies in Wales have become much more liberal and childcentred; whereas, in Scotland, policies have tended towards a more punitive turn (although there are indications that this may change following the election of a new Nationalist government in 2007). It is notable that the Scottish policies of the New Labour government were developed in the absence of any government sponsored national SRD surveys. Despite their apparent policy focus, few of the SRD studies in Great Britain and Ireland have been used to specifically evaluate the effectiveness of policies in relation to young people and offending. Exceptions to this are the MORI surveys, which were commissioned by the Youth Justice Board and used to monitor the performance and effectiveness of the youth justice policy; the On Track survey, which was used to evaluate the Government’s multiple intervention initiative targeted at children and families in high deprivation/high crime areas in England and Wales (Armstrong et al., 2005); and the Welsh survey which was used to evaluate the Promoting 163

Self-reported crime.book Page 164 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Prevention initiative in Swansea (Haines, Case 2004). In addition, the BYDS is used by the United Kingdom Focal Point on Drugs (which coordinates the United Kingdom submission to ECMDDA) as a key national indicator for the prevalence and patterns of drug use among the general population in Northern Ireland.

II.

Methodological issues

1.

Survey content

SRD surveys carried out in Great Britain and Ireland have consistently covered a wide range of behaviours, although few have included the same list of crimes or offences, and many use different wording or questionnaire design. The types of offences included in surveys tend to range from minor acts of delinquency, more commonly thought of as forms of anti-social behaviour, through to more serious forms of criminal offending. The forms of delinquency covered fall into six broad categories: property crime (including graffiti, vandalism, fire setting, vehicle theft and breaking into buildings); violence (including bullying, assault, robbery and weapon use); substance use (including smoking and drinking and illegal drug use); fraud and dishonesty (such as not paying the correct fare on public transport, handling stolen goods and illegal use of credit cards/cheque books); other forms of theft (such as theft from shops, school, work or home) and more general anti-social behaviour (which includes difficult behaviour at school, posing a public nuisance, etc). A few surveys have included questions on less ‘traditional’ criminological topics, such as animal abuse, gang fighting, pick-pocketing and benefit fraud. The study by Karstedt and Farrall (2006) was particularly unusual in focusing solely on ‘moral’ crimes such as making false insurance claims and jumping red lights. Nevertheless, the majority of questioning tends to be at the more trivial end of the spectrum in relation to offending behaviour. A general criticism of SRD studies is their failure to include very serious crimes, such as rape or murder, and the Great Britain and Ireland studies are no exception. Farrington (2001) has argued that this limits the value of studying criminal careers, and that a standard ‘index’ of serious offences should be developed for inclusion in all SRD surveys. The majority of SRD studies have been restricted to asking a general prevalence question (i.e. have you done…?), during the period ‘ever’ and/or ‘in the last year, followed by a frequency question (i.e. how many times have you …?). However, some of the longitudinal studies (such as the OCJS, 164

Self-reported crime.book Page 165 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

PADS+ and the Edinburgh Study) have included a range of additional follow-up questions. These questions include age of onset and desistance, contact with the police or other criminal justice organisations, more specific details of the offences committed (e.g. nature of items stolen or damaged, injuries inflicted, weapon used etc), involvement of cooffenders, crime location and relationship to victim. Studies focusing on health behaviours have also included follow up questions on attempts to quit and dependency (e.g. the study of Mental Health in Children and Young People). In addition to a range of standard individual socio-demographic questions (which tend to be similar across the SRD studies), the surveys carried out in Great Britain and Ireland have included many additional questions aimed at developing explanatory models and testing causal hypotheses. The prevailing risk factor paradigm has ensured that many of these questions are of a similar nature, for example questions on early upbringing, family structure and dynamics, parental supervision, school-related factors, individual personality measures, conventional and unconventional leisure activities, peer associations and behaviours, and so on. These sorts of generalised questions are similar to those found in surveys across a range of countries, including the instruments used in the International Study of Self-Reported Offending. The Communities that Care organisation has been particularly instrumental in developing a range of standard questions aimed at measuring the most common risk and protective factors, although these are marketed rather than distributed freely. Several of the recent SRD studies in GB and Ireland have set themselves apart by developing more innovative methodologies, including the collection of data from a range of different objective sources. The PADS+ study has collected a unique set of data about individuals’ exposure to different environments and the social settings within those environments in which they take part, in order to further develop Wikstrom’s Situational Action Theory. The E-Risk study focuses on the complex nature versus nurture debate, by surveying twins and collecting DNA material as well as a wide range of social survey data. SPOOCS is collecting detailed evidence on the factors underlying the process of desistance, which are often neglected by other SRD studies. While the OCJS and the Edinburgh Study have collected more detailed than usual information about individual contacts with the police and other criminal justice organisations.

165

Self-reported crime.book Page 166 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

2.

Methodological integrity

The SRD studies listed in Appendix 1 vary in terms of their methodological integrity and data quality. However, the proliferation of private sector social survey organisations during the 1990s and early 2000s has helped to improve the approach of researchers to large-scale survey design. Most Government funded social surveys are carried out by private sector social research organisations, which are governed by industry-led standards of conduct and, therefore, apply high degrees of rigour to their methodologies. Examples of good practice in the social research industry include: the development of robust sampling methods; the use of computerised data collection methods (such as Computer Assisted Self Interviewing or CASI), reducing problems of item non-response and improving the reliability of responses to offending questions; weighting techniques to deal with issues of response bias and selection effects; and technical reports published as standard, detailing information on study design and methodology. It cannot be claimed that SRD surveys carried out by academic researchers are less methodologically robust than those carried out by private sector organisations, but it is the case that they have tended to be less transparent about design and administration. Academic research is not governed by any standard set of guidelines, other than institutional ethical protocols. The peer review process is considered rigorous enough to ensure that only the highest quality research is funded, although peer reviews are never made public and academic studies rarely publish the detailed methodological details of their research. The release of precious data (providing the opportunity for validation of results) is also somewhat contentious, with only two of the academic studies included in this review (the Cambridge and Edinburgh studies) having lodged data with the United Kingdom Data Archive for general use. One problematic aspect of some SRD surveys in Great Britain and Ireland (especially those that are not Government tools) is that they still ask respondents to complete paper and pencil questionnaires, which are uneconomical in terms of researcher time and effort, are known to increase respondent error and contribute to missing data. Researchers involved in developing the Welsh youth survey, however, used an innovative method known as Interactive Computer-Based Self-Completion interviewing (ICSI) (Haines, Case, 2004). Consultation with young people in the United Kingdom has shown that positive engagement with the research process is enhanced if young person-centred methods are employed rather than more ‘formal’ types of adult enquiry such as traditional interviews (Kirby, 2002). Following extensive piloting, Haines and Case (2004) found that the 166

Self-reported crime.book Page 167 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

computer interactive questionnaire not only increased response and completion rates but produced highly positive feedback. The authors argue that the application of computer-based multi-media technologies to questionnaire design and completion represents the next stage of development in social research. Despite common problems of missing data, participation bias and sample attrition in longitudinal studies, weighting strategies tend not to be applied to academic datasets (exceptions include the Edinburgh Study and the ERisk study which both use weights, and the BYDS which uses multiple imputation to deal with missing data). Few academic studies have published detailed methodological papers or released detailed technical reports (the Cambridge and Edinburgh studies are, again, exceptions).

3.

Validity and reliability

Questions about the validity and reliability of self-report studies are often raised due to concerns that respondents will under or over-report their offending, either deliberately (through concealment or exaggeration) or inadvertently (through forgetfulness or inaccurate recall). Farrington (2001) provides a review of the issues of validity in terms of self-reports of offending. Briefly, the key issues are: do self-reports accurately estimate the number of crimes committed or the prevalence or frequency of offenders ? And what are the best external criteria against which to compare selfreported offending ? The most common method of validation for self-reports is comparison with other sources of data, usually official records of arrest or conviction, but comparison with peer, teacher or parental reports is also valuable. Datalinkage is more often carried out for longitudinal rather than crosssectional surveys although, in Great Britain and Ireland, few Governmentsponsored SRD surveys have incorporated this method and, as a result, validation is often not possible. The Cambridge study has provided some of the best evidence of the value of linking datasets to validate self-report measures of offending. The study included additional data from parents and teachers up to around age 14 or 15, peer ratings and partner interviews, as well as comparisons with official data on convictions. Farrington (2001) reports that concurrent validity was good amongst Cambridge Study respondents (that is, the majority of those convicted had reported their offences); however, predictive validity was even more impressive as a high proportion of those who admitted offending went on to be convicted.

167

Self-reported crime.book Page 168 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Most of the contemporary longitudinal studies include elements of data linkage in their design, which indicates the potential for these studies to contribute further to the evidence about concurrent and predictive validity. The E-Risk study includes links to school records, hospital birth records, ACORN (a geo-demographic indicator), census data and a survey of neighbourhoods. The BYDS has incorporated a survey of family members (parents and older siblings) for some sample members. PADS+ includes school records at the individual level and a range of data at the neighbourhood level, including a community survey, census and land use data and police recorded crime. Moreover the Edinburgh Study includes record data from schools, social work, children’s hearings, police and criminal convictions for individuals; a survey of parents; teacher’s assessments; and a GIS with neighbourhood level data from census, police recorded crime and a community survey. Most of these studies are at an early stage, however, and none has yet published findings on validity through the use of multiple sources of data. A further issue that relates to validity is the attrition of respondents from longitudinal studies. The use of weighting to adjust for survey nonresponse is discussed above; however, such techniques cannot always diminish the impact of high levels of attrition (McVie, Norris, Raab, submitted). High response rates are essential and it is particularly important to retain the involvement of ‘high risk’ participants, who are generally the most difficult to locate and secure cooperation from. Amongst the longitudinal studies carried out in Great Britain and Ireland response rates vary widely, although this appears to bear some relationship to the size and method of the study. After 40 years of study, the Cambridge Study continued to secure the participation of 93% of the original 411 respondents, which compares to a response rate of 81% after only six years amongst the 4,300 respondents of the Edinburgh Study. Response rates are generally higher amongst school-based populations, but often fall dramatically when attempts are made to follow participants after leaving school (as was the case with the Edinburgh Study). In relation to ‘good practice’, it is not always transparent from study websites or publications exactly how response rates have been calculated, such as whether they relate to population response or sample retention. Internal reliability of self-report instruments is harder to determine, although there is evidence that the current studies in Great Britain and Ireland have taken steps to minimise survey error through poor questionnaire design. Most of the survey instruments are stated to be based on tried-and-tested measures, and comparison of data both between and within interviews are reported to indicate a high degree of internal validity. One particularly important model of questionnaire development is that of 168

Self-reported crime.book Page 169 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

the Communities that Care organisation, which has spent many years developing their questionnaire (Anderson et al., 2002). The survey is designed to maximise validity by ensuring complete anonymity; however, respondents are removed if they breach any of a series of conditions that brings the validity or honesty of their responses into question. Nevertheless, there has been little systematic testing of reliability, precision and content validity of instruments to the level recommended by Huizinga and Elliott (1986), and this is an aspect of methodology that could be greatly enhanced in Great Britain and Ireland. Finally, techniques such as the inclusion of a ‘bogus drug’ and comparison of individual data between sweeps are common ways of testing response reliability. The BYDS research team used ‘recanting’ analysis to assess the reliability of teenage self-reports of drug use behaviour. In this method, temporal inconsistencies in reports of lifetime drug use were used to identify and classify reporting errors and to model the characteristics of young people prone to over-report and under-report licit and illicit drug use (Percy et al., 2005). Further work is ongoing to identify methods for correcting identified reporting errors within the data. Exploratory and confirmatory factor analyses have been conducted to examine the structural validity of key measures within the data set (Percy et al., in press).

4.

Other examples of good practice

As well as the examples of good practice illustrated above, there are some further elements of research design utilised by researchers in Great Britain and Ireland that deserve mention in this report. For the most part, SRD studies have been carried out under strictly controlled conditions and administered in person by trained interviewers (rather than by teachers or through postal surveys). The application of rigorously consistent interview techniques and a low turnover of highly trained and experienced research staff are particular examples of good practice (PADS+). Some studies also provide targeted help and support to children with reading difficulties to ensure successful completion of the questionnaire (Edinburgh Study, BYDS). Lengthy questionnaire development and pilot testing appears to be a common feature of research preparations and most questionnaires tend to include a range of tested, validated and theory-based measures (in particular, CtC). One particular example of good practice is the establishment of a ‘panel’ of children who provide critical appraisal and review of the questionnaire each year (BYDS). Large-scale Government

169

Self-reported crime.book Page 170 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

surveys adopt techniques such as CASI and Audio-CASI as standard, although this is rare amongst academic studies (the use of ICSI in the Welsh youth survey being an important exception). Ensuring quality at the data entry is also important, and several research teams reported carrying out blind double-entry and coding checks, in addition to standard data validation checks, to minimise keying errors and ensure quality in the dataset. Funded research in Great Britain and Ireland must adhere to ethical guidelines, and this is evident in the strict consent procedures adopted by those researching populations aged below 16. In addition, most studies place heavy emphasis on confidentiality and data security. There are particular examples of good practice in monitoring and tracking young people, which includes developing methods for anticipating attrition and using extensive efforts to trace and secure compliance from participants, thereby maximising sample retention (PADS+, Cambridge). There is also added value in collecting as detailed information as possible at point of contact, and using additional data from official records and peers to track cohort members (Edinburgh Study). The benefits of a long period of follow-up are demonstrated most particularly by the Cambridge Study which, despite its limitations, continues to make a powerful contribution to both theory and policy. Users of the Cambridge Study have also provided guidance to other criminologists through the publication of a large number of methodological papers and articles, which is highly valuable. Another particularly useful survey in this respect is the report of the On Track evaluation, which makes a number of very clear and practical recommendations for good practice in schools-based SRD survey research.

III. Usage and policy impact of SRD surveys It is more difficult to summarise the usage and impact of SRD survey data in Great Britain and Ireland than to describe the surveys and assess their methodological strengths. In a wide-ranging review of the surveys carried out in these four jurisdictions, it was not clear – even to the research teams themselves – to what extent their research had had a tangible impact on policy and, if so, what that impact actually was. Nevertheless, a review of the major policy transformations in the jurisdictions concerned provides some clues as to the value of SRD studies in shaping youth justice policy, in particular. In the early 1990s, youth justice policy in England and Wales began a period of marked transition. In 1993, as shadow home secretary, Tony Blair 170

Self-reported crime.book Page 171 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

was impressed by Bill Clinton's successful attempts to repoliticise crime to positive electoral effect in the US, and came back to Britain promising to be tough on crime, tough on the causes of crime. During the early to mid-1990s, senior New Labour politicians increasingly used punitive rhetoric and symbolism in their criminal justice policies and, in particular, their youth justice policies which were characterised by increasing intolerance. In May 1997, following a highly successful election campaign, New Labour published a White Paper entitled No More Excuses: A New Approach to Tackling Youth Crime in England and Wales. So began a 10 year period of major policy reform that would apparently be driven by an ‘evidence-based’ approach. The proliferation of SRD studies in the late 1990s and early 2000s has been largely facilitated by support from central Government and, therefore, there can be little doubt that such studies have played a role in the policy making process over that time. At the heart of government policy on criminal justice is the identification of ‘risk factors’, facilitated largely by studies (such as the Cambridge Study) which have focused on the predictors, causes and correlates of childhood offending and antisocial behaviour (see Farrington, 1997, 2006). New Labour's transformation of the youth justice apparatus (in particular, the introduction of the multiagency Youth Offending Teams) was almost entirely underpinned by constructions of children and families at most risk. The notion that offenders could be predicted from early in life resulted in a series of ‘preemptive initiatives’, whereby young people were targeted on the basis not of an offence they had committed but on the basis of a ‘condition’, a ‘character’ or a ‘mode of life’ that was judged to pose a risk. Proposals included targeting the under 5s with so-called ‘baby ASBOs’. In Scotland, following the onset of devolution in 1999, youth justice policies which had previously been welfarist in orientation very much took the same turn as those driven by New Labour in England. The Edinburgh Study has strong connections with both local and national policy makers, and the research team have published a series of policy-oriented digest reports and made numerous conference, seminar and private presentations to Ministers, senior civil servants and local authority representatives. The findings advocate a return to welfare principles and a model of universal social and economic regeneration, rather than punitive disposals targeted at a specific young people (McAra, McVie, 2005, 2007); however, it is only in recent months with the inception of a new Nationalist government that such policies have started to turn. The direct impact of the Edinburgh Study findings is, however, unmeasurable. The PADS+ research team has direct communication with officials at the Home Office, the Youth Justice Board and local government officials, as 171

Self-reported crime.book Page 172 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

well as some direct contact with individual Members of Parliament. The project is still in the early stages of dissemination, however, and no tangible impact is yet evident. The CtC initiative has published a series of reports that have been widely read, and have directly prompted meetings between CtC representatives and the Home Office. Nevertheless, CtC representatives claim that the effect of such contact is “hard to say”. In Ireland, the delay of devolution has meant that Irish politicians have not been so influenced by the punitive turn of the United Kingdom Government, and they are only now starting to look to other jurisdictions to evaluate youth justice policy. Perhaps for this reason, the BYDS team have had somewhat greater success in influencing policy, playing an active role in the development of the New Strategic Direction for Alcohol and Drugs (2006-2011) in Northern Ireland. The findings from the study have also been used by the Advisory Council on the Misuse of Drugs. In addition, the findings of the Irish International Self-Report Delinquency survey have been taken on board by the EU who funded the research, although the actual policy impact is again uncertain (O'Mahony, personal communication). Interestingly, Welsh devolution resulted in a distinctly different framework for youth justice services in Wales to that in England. Although crime and security were not devolved, education and community safety were and youth justice policy has been developed by the Welsh Assembly Government through these channels. As in England there are local, multidisciplinary Youth Offending Teams, but the funding arrangements and the configuration of adjacent services vital to the prevention of youth offending are very different. The focus has very much been on prevention and effective community-based disposals. The youth surveys in Wales appear to have had some direct impact on local policy development through a positive evaluation of the Promoting Prevention initiative, which aimed to prevent youth offending in Swansea (Case, Haines, 2004). Overall, however, the direct impact of SRD surveys on policy across the four jurisdictions is not clear, in part due to the lack of transparency in the political decision making process. While the United Kingdom Governments markets themselves as ‘intelligent customers’ and proclaim to build policy on nothing but the very highest evidence base, there continues to be a veil of secrecy around which research findings translate into policy and exactly how this is achieved.

172

Self-reported crime.book Page 173 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

IV. Conclusion This report has provided a broad review of around 30 major self-report delinquency surveys carried out in the four countries that make up Great Britain and Ireland. From these four jurisdictions have emerged a large number of influential SRD studies, contributing both to theoretical knowledge and to policy development, particularly in recent years in relation to youth justice. Compared to other European countries, these countries are in an enviable position in terms of the wealth of scientific evidence available on the prevalence, frequency, nature and causes of crime; however, there are important lessons to be learned from this review of methodology and good practice, and critical questions to be asked about the direction and usage of this SRD research. From a methodological point of view, the studies reviewed here demonstrate many strengths and highlight the benefits of improving standards of research practice. Many lessons have been learned from the involvement of private sector research organisation in terms of complex sampling methods, use of computerised data collection instruments, strategies for dealing with missing data and the publication of detailed technical reports. Although it is not claimed here that academic research is less methodologically robust, there is scope for improvement in terms of survey administration, greater transparency in design and administration, and more ready release of data for secondary analysis or validation of results. Nevertheless, there are some particular examples of good practice that deserve notice here, including: the retention of highly trained, experienced interviewers ; provision of targeted help and support to less able respondents; testing research instruments on established ‘panels’ of similarly aged respondents; the use of user-friendly computerised questionnaires; adherence to strict ethical guidelines and data protection requirements; and the development of methods for reducing sample attrition and improving individual tracing in longitudinal studies. With the exception of the Cambridge Study, few of the studies mentioned in this report have published extensively on issues of validity or reliability. However, most of the contemporary longitudinal surveys are still in their infancy and the potential for contributing to this field of literature is considerable, particularly given the high degree of linkage to data on convictions, geographical context and school records, for example. Unfortunately, this is rarely done in the case of large-scale government sponsored research studies, which form the bulk of the SRD research in Great Britain and Ireland. A related problem is the lack of harmonisation across surveys in terms of the questions asked on self-report offending. It is not proposed here that SRD studies should adopt a standard or formulaic 173

Self-reported crime.book Page 174 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

design; however, it would be in the interests of developmental criminology to have a well-tested and widely agreed upon set of questions that might be lodged in an easily accessible (and free-of-charge) ‘Question Bank’, for example. This might include a standard index of serious offences, as advocated by Farrington (2001). Such measures would facilitate the systematic testing of reliability, precision and content validity of survey instruments used in Great Britain and Ireland, as recommended by Huizinga and Elliott (1986). One particularly important issue that is raised by this report relates to the policy usage and impact of SRD surveys. The predominant focus of the existing SRD studies reflects a preoccupation with early intervention and risk factor identification, which dovetails neatly with current Government criminal justice policy under New Labour. Few of the studies conducted in Great Britain and Ireland proclaim themselves to be following a more general life-course developmental approach. It is difficult to dissociate the primary focus of such studies from the broader political context, since the majority of this research has been and continues to be largely Government funded. Policy interest in SRD research is undoubtedly strong, but whether such studies have influenced the direction of government policy during the New Labour era, or merely contributed in a passive way to the evidence base, is unclear as there is a general lack of transparency about how such research findings are used in the policy making process. It remains to be seen whether the new generation of longitudinal studies will attempt to address a broader set of questions about crime across the life-course and, indeed, whether they will receive the funding to do so. Finally, although this report has largely discussed the SRD studies from England, Ireland, Scotland and Wales as one body of research, there are distinct and important differences between these four jurisdictions. The bulk of the research has been conducted in England; and much of that conducted in the other three countries has been designed, administered and analysed by teams of researchers within England. Yet, the devolved governments have an increasing desire to commission independent scientific evidence on which to base their policies. Crime remains at the heart of political debate in all four jurisdictions, but so far the provision with regards to SRD research has been patchy. Scotland, for example, has had no Government sponsored national survey of offending. The time seems ripe for a broad review of such surveys and the development of a United Kingdom strategy which clarifies the purpose of such research and produces a harmonised baseline of measures, against which the more theoretically focused academic studies can compare themselves. Importantly, funding mechanisms must be found to ensure that SRD studies which do not directly contribute to current Government policy can 174

Self-reported crime.book Page 175 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

be facilitated and that longitudinal research can continue beyond the teenage years to ensure longer-term study of offending behaviour.

References ANDERSON B., BEINART, S., LEE S., UTTING D., 2002, Youth at Risk?, London, Communities that Care. ANDERSON S., KINSEY R., LOADER I. SMITH, C., 1994, Cautionary Tales, Aldershot, Avebury. ARMSTRONG D., HINE J., HACKING S., ARMAOS R., JONES R., KLESSINGER N., FRANCE A., 2005, Children, risk and crime: the On Track Youth Lifestyles Surveys, London, Home Office Research Study 278. BOTTOMS A., SHAPLAND J., COSTELLO A., HOLMES D., MUIR G., 2004, Towards Desistance: Theoretical Underpinnings for an Empirical Study, The Howard Journal of Criminal Justice, 43, 4, 368-389. BREEN J., MANNING U., O’DONNELL I., O’MAHONY D., SEYMOUR M., 2006, The International Self-Reported Delinquency Study: Results from the Survey in the Republic of Ireland, Unpublished report to the Department of Justice, Ireland. BUDD T., COLLIER P., MHLANGA B., SHARP C., WEIR G., 2005a, Levels of selfreport offending and drug use among offenders: findings from the criminality surveys, London, Home Office Online Report 18/05. BUDD T., SHARP C., MAYHEW P., 2005b, Offending in England and Wales:First results from the 2003 Crime and Justice Survey, London, Home Office Research Study 275. CASE S., 2007, Questioning the evidence of risk that underpins evidenceled youth justice interventions, Youth Justice, 7, 2, 91-105. CASE S., HAINES K., 2004, Promoting prevention: evaluating a multi-agency initiative of youth consultation and crime prevention in Swansea, Children and Society, 18, 355-370. FARRINGTON D.P., 1995, The development of offending and anti-social behaviour from childhood: Key findings from the Cambridge Study in Delinquent Development, Journal of Child Psychology and Psychiatry, 36, 929-964. FARRINGTON D.P., 1997, Human development and criminal careers, in MAGUIRE M., MORGAN R., REINER R. (Eds), The Oxford Handbook of Criminology (2nd edition), Oxford, Clarendon Press, 361-408. FARRINGTON D.P., 2001, What has been learned from self-reports about criminal careers and the causes of offending?, London, Home Office Report.

175

Self-reported crime.book Page 176 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

FARRINGTON D.P., 2005, The integrated Cognitive Antisocial Potential (ICAP) theory, in FARRINGTON D.P. (Ed.), Integrated Developmental and Life-course theories of offending Advances in Criminological Theory, Volume 14, New Brunswick, NJ, Transaction, 73-92. FARRINGTON D.P., 2006, Understanding and PreventingYouth Crime, in Muncie, J., Hughes, G., MCLAUGHLIN, E. (Eds), Youth Justice: Critical Readings, London, Sage, 425-430. FARRINGTON D.P., COID J.W., HARNETT L.M., JOLLIFFE D., SOTERIOU N., TURNER R.E., WEST D.J., 2006, Criminal careers up to age 50 and life success up to age 48: new findings from the Cambridge Study in Delinquent Development, 2nd Edition, London, Home Office Research Study 299. FLOOD-PAGE C., CAMPBELL S., HARRINGTON V., MILLER J., 2000, Youth Crime, Findings from the 1998/99 Youth Lifestyles Survey, London, Home Office Research Study 209. GRAHAM J., BOWLING B., 1995, Young People and Crime, London, Home Office Research Study 145. GREEN H., MCGINNITY A., MELTZER H., FORD T., GOODMAN R., 2005, Mental Health of Children and Young People in Great Britain 2004: Summary Report, London, Office for National Statistics. HAINES K., CASE S., 2004, Promoting prevention: a multi-agency initiative to prevent youth offending through consultation in Swansea Schools, Youth Justice, 4, 2, 117-132. HUIZINGA D., ELLIOTT D.S., 1986, Reassessing the reliability and validity of self-report delinquency measures, Journal of Quantitative Criminology, 2, 4, 293-327. JAMIESON J., MCIVOR G., MURRAY C., 1999, Understanding Offending Among Young People, Edinburgh, The Stationery Office. KARSTEDT S., FARRALL S., 2006, The moral economy of everyday life: markets, consumers and citizens, British Journal of Criminology, 46, 1011-1036. KIRBY P., 2002 Participatory research with children and young people, in FRANKLIN B. (Ed) The Handbook of Children’s Rights: Comparative policy and practice, 2nd Edition, London, Routledge. MCARA L., 2006, Negotiated order: gender, youth transitions and crime, British Society of Criminology E-Journal, Volume 6 http:// www.britsoccrim.org/volume6/005.pdf. MCARA L., MCVIE S., 2005, The Usual Suspects? Young People, Street-Life and the Police, Criminal Justice, 5, 1, 5-36. MCARA L., MCVIE S., 2007, Youth Justice? The Impact of System Contact on Patterns of Desistance from Offending, European Journal of Criminology, 4, 3, 315-345.

176

Self-reported crime.book Page 177 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

MCCRYSTAL P., HIGGINS K., PERCY A., 2006, School exclusion, drug use and delinquency in adolescence, Journal of Adolescence, 29, 829-836. MCQUOID J., 1994, The Self-Reported Delinquency Study in Belfast, Northern Ireland in JUNGER-TAS J., TERLOUW G.-J., KLEIN M. W. (Eds.) Delinquent Behavior among Young People in the Western World. First Results of the International Self-Report Delinquency Study, Kugler, Amsterdam/New-York, 65-101. MOFFITT T., E-RISK TEAM, 2002, Contemporary teen aged mothers in Britain, Journal of Child Psychology and Psychiatry, 43, 6, 727-742. MVA, 2002, Young people and crime in Scotland: Findings from the 2000 Scottish crime survey, Edinburgh, Scottish Executive. PERCY A., 2008, Moderate adolescent drug use and the development of substance use self-regulation, International Journal of Behavioural Development, 32, 5, 451-458. PERCY A., MCALISTER S., HIGGINS K., MCCRYSTAL P., THORNTON M., 2005, Response consistency in adolescent drug use self-reports, Addiction, 100, 189-196. PERCY A., MCCRYSTAL P., HIGGINS K., 2008, Confirmatory Factor analysis of the Adolescent Self-Report Strengths and Difficulties Questionnaire, European Journal of Psychological Assessment, 24, 1, 43-48. SAMPSON R.J., LAUB J.H., 1993, Crime in the making: Pathways and turning points through life, Cambridge, MA, Harvard University Press. SHORT J.F., NYE F.I., 1957, Reported behaviour as a criterion of deviant behaviour, Social Problems, 5, 207-213. SMITH D.J., MCVIE S., 2003, Theory and Method in the Edinburgh Study of Youth Transitions and Crime, British Journal of Criminology, 43, 1, 169195. SMITH D.J., MCVIE S., WOODWARD R., SHUTE J., MCARA L., 2001, The Edinburgh Study of Youth Transitions and Crime: Key Findings at Ages 12 and 13, Edinburgh, Study of Youth Transitions and Crime Research Digest No. 1. WIKSTRÖM P.O., 2005, The Social Origins of Pathways in Crime: Towards a Developmental Ecological Action Theory of Crime Involvement and Its Changes, in FARRINGTON, D.P (Ed.), Integrated Developmental and Life-Course Theories of Offending. Advances in Criminological Theory, Volume 14, New Brunswick, NJ, Transaction, 211-246. WIKSTRÖM P.O., SAMPSON R.J. (Eds), 2006, The explanation of crime: context, mechanisms and development, Cambridge, Cambridge University Press. Youth Justice Board, 2003, Youth Survey 2003. [http://www.yjb.gov.United Kingdom/Publications/Resources/Downloads/ YouthSurvey2003.pdf]. 177

178

1967

1975

W.A. Belson

R. Mawby

Sub-national: Sheffield (England)

Sub-national: London (England)

National: England, Wales and Scotland

Geographical coverage Sub-national: London (England)

Method and response rate

Longitudinal survey involving nine sweeps of data collection (at ages 8, 10, 12, 14, 18, 21, 25, 32 and 48) from boys living in one inner-city area of South London. Individual interviews were supplemented by parental interviews (ages 8-14), teacher questionnaires (ages 8-14) and peer ratings (ages 8 and 10). Female partner interviews at age 48. Consistently high response rates (95% at age 18; 94% at age 32 and 93% at age 48). Additional data collection from criminal records. 808 males, One off survey of a representative sample of boys aged 15 – 21 from Great Britain, using individual interviews. 71% response rate. 1,425 males, One off survey involving a representative sample aged 13-16 of boys from London, using individual interviews. 86% response rate. 591 males and One off survey involving group-based self females, aged completion in one school, paper and pencil 13-15 questionnaires. 80% response rate.

411 males aged 8-9, followed up to age 48

Sample

Details of some of the earlier surveys are taken from David P. Farrington’s (2001) report for the Home Office ‘What has been learned from self-reports about criminal careers and the causes of offending?’

General selfreported delinquency General selfreported delinquency General selfreported delinquency

1963

H.D. Willock

1.

Main focus of study General selfreported delinquency

Researcher(s) & study Date(s) of title survey 1961 – 2004 D.J. West and D.P. Farrington: Cambridge Study in Delinquent Development

Appendix 1: Summary of major self-report delinquency surveys in Great Britain and Ireland1

Self-reported crime.book Page 178 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

General selfreported delinquency

General selfreported delinquency

1992-1993

J. McQuoid

J. Graham, B. Bowling : 1992-1993 Young People and Crime

General selfreported delinquency

1990

1989

Main focus of study General selfreported delinquency General selfreported delinquency

S. Anderson, R. Kinsey, I. Loader, C. Smith : The Edinburgh Crime Survey J. McQuoid

Researcher(s) & study Date(s) of title survey D. Riley, M. Shaw 1983

Self-reported crime.book Page 179 Thursday, April 23, 2009 8:02 AM

One off survey involving individual interviews. 71% response rate.

751 males and females aged 14-15 892 males and females aged 11-15 One off survey involving group-based self completion in four local schools. Response rate not reported.

Method and response rate

Sample

Sub-national: Belfast 310 males and One off survey involving individual interviews, conducted as a pilot survey for the first (Northern Ireland) females aged 14-21 International Self-Report Delinquency Study. 95% response rate. Sub-national: Belfast 883 males and One off survey involving individual interviews, (Northern Ireland) females aged representing the Northern Ireland component of 14-21 the first International Self-Report Delinquency Study. Face to face interviews using paper and pencil questionnaires. Adopted the ‘modified random walk’ method to select the sample. 92% response rate. National: England 2,529 males One off survey involving individual interviews and Wales and females, with a national random sample of 1721 aged 14-25 individuals living in private households, plus a booster of 808 young people from ethnic minorities. Face to face interviews using paper and pencil questionnaires, based on the ISRD-1 survey instrument. Supplementary life-history interviews carried out with 21 ‘desisters’. 64% response rate.

Sub-national: Edinburgh (Scotland)

Geographical coverage National: England and Wales

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

179

Main focus of study General selfreported delinquency

General selfreported delinquency

Researcher(s) & study Date(s) of title survey C. Flood-Page, S. 1998-1999 Campbell, V. Harrington, J. Miller : The Youth Lifestyles Survey

J. Jamieson, G. McIvor, 1996-1999 C. Murray : Understanding Offending amongst Young People

Self-reported crime.book Page 180 Thursday, April 23, 2009 8:02 AM

180 Sub-national: two small towns in Scotland

Geographical coverage National: England and Wales Follow up to the survey carried out in 1992-93, reported by Graham and Bowling. One off survey involving individual interviews, using computer assisted self-completion interviewing (CASI). Household survey design. Based largely on instrument used by Graham and Bowling. 69% response rate. One-off survey involving group self-completion of paper and pencil questionnaires in six secondary schools by 1,274 pupils. The sample was drawn from two Scottish towns with crime rates close to the national average, one with higher than average unemployment rates and the other with lower than average unemployment rates. This was followed up by 276 in-depth interviews with young people aged 14-15, 18-19 and 22-25. Interviews were also conducted with seven police officers, seven social workers and five teachers.

4,848 males and females aged 12-30

1,274 males and females, aged 14-16

Method and response rate

Sample

Self-Reported Crime and Deviance Studies in Europe

Mental health and National: England, behavioural Wales and Scotland disorder

1999, 2004 H. Green, A. McGinnity, H. Meltzer, T. Ford, R. Goodman : Mental Health of Children & Young People in Great Britain

Geographical coverage Sub-national: Edinburgh (Scotland)

Main focus of study General selfreported delinquency

Researcher(s) & study Date(s) of title survey D.J. Smith, L. McAra, 1998-2004 S. McVie : The Edinburgh Study of Youth Transitions and Crime

Self-reported crime.book Page 181 Thursday, April 23, 2009 8:02 AM

Longitudinal study involving six annual sweeps of data collection. Census of all school pupils aged 11/12 in 1998. Group self-completion mainly in classrooms (40 schools) using paper and pencil questionnaires, with absentees followed up at home. Some individual interviews at ages 13 and 17. Survey of parents at age 15 and teacher’s questionnaires at age 13. Additional individual data collected from official records of children’s hearing system, social work, schools, police and criminal convictions. Supported by a GIS system (using census and police recorded crime data) and Community Survey. 96% response rate at age 12,81% at age 17. Repeated cross-sectional survey. Multi-stage stratified random sample drawn from child benefit records held by DWP Child Benefit Centre. Survey of children living in private households. Those aged 11-16 filled in a selfcompletion questionnaire on a laptop. Additional information collected from interviews with parents/carers and a postal survey of teachers. 65% response rate overall, 27% of which were 11-16 year olds.

4,300 males and females aged 12, followed up to age 17

3,344 males and females, aged 5-16, surveyed in 2004

Method and response rate

Sample

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

181

182 General selfreported delinquency

Crime victimisation

T. Moffitt, A. Caspi, L. 1999-2008 Arsenault : E-Risk Study

MVA : The Scottish Crime Survey Youth Component

1993, 1996, 2000

Main focus of study General selfreported delinquency

Researcher(s) & study Date(s) of title survey Youth Justice Board : 1999-2005 The MORI Youth Surveys

Self-reported crime.book Page 182 Thursday, April 23, 2009 8:02 AM

National: Scotland

National: England and Wales

Geographical coverage National: England and Wales

Method and response rate

Seven annual repeated cross-sectional surveys using the MORI schools omnibus questionnaire. Group-based self-completion in school classrooms. Random selection of schools within regions, selected proportional to size. Random selection of classes within schools. Paper and pencil questionnaires. Response rates not reported. 2,232 males Longitudinal study of twins involving four sweeps (at age 5, 7, 9 and 12), forming part of and females, age 5, followed the SCoPiC research network. National up to age 12 probability sample using the twins register as the sampling frame. Children at high risk for disruptive behaviour disorders oversampled. Home visits with face to face interviews, tests, and observational data collection by research workers. SRD questions included for first time at age 12. Supplementary data sets include OFSTED school records, hospital birth records, ACORN Census data and SCoPIC survey of neighbourhoods. 96% response rate at age 10. 400-500 males Repeated cross-sectional victimisation survey of and females, private households, where any children in the aged 12-15, at household aged 12-15 were identified and asked each sweep to complete a paper questionnaire. No data linkage to other sources. 70% response rate at each sweep.

Approx 5,000 males and females, aged 11-16, at each sweep

Sample

Self-Reported Crime and Deviance Studies in Europe

2002

Date(s) of survey 2000

B. Anderson, S. 2000-2001 Beinart, S. Lee, D. Utting : Communities that Care Youth Survey

Researcher(s) & study title T. Budd, P. Collier, B. Mhlanga, C. Sharp G. Weir : The Prisoner Criminality Survey T. Budd, P. Collier, B. Mhlanga, C. Sharp G. Weir : The Community Penalties Criminality Survey

General selfreported delinquency

Main focus of study General selfreported delinquency amongst prisoners General selfreported delinquency amongst offenders with a community sentence

Self-reported crime.book Page 183 Thursday, April 23, 2009 8:02 AM

National: England, Wales and Scotland

National: England and Wales

Geographical coverage National: England and Wales Cross-sectional survey of prisoners in England and Wales. Random probability sample of all those sentenced to prison or young offender institutions during a two month period in 2000. 90% response rate. Cross-sectional survey of offenders given a community sentence in England and Wales. Random probability sample of those sentenced to a Community Rehabilitation Order, a Community Punishment Order, a Community Punishment and Rehabilitation Order or a Drug Treatment and Testing Order. Over-sample of women, black and minority ethnic offenders, DTTO sentencees and persistent offenders. 53% response rate. One-off national survey, but linking directly to many local studies carried out in local ‘Communities that Care’ designated areas. Based on a sample of schools drawn from the National Foundation for Educational Research school register. Two stage sample design: 41 LEAs selected with probability proportional to size (ages 11-16 approx) stratified by region; then within each LEA, five schools selected with probability proportional to size. Group selfcompletion in school classrooms using paper and pencil questionnaires. 41% of sampled schools participated. Individual response rate not reported. No additional data collected.

1,884 males, aged 15 or over

15,000 males and females, aged 11-16

1,578 males and females, aged 16 or over

Method and response rate

Sample

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

183

184

A. Percy, I. McCrystal : Belfast Youth Development Study

Main focus of study General selfreported delinquency

2001-2007 Substance use and (funding up general selfto 2009) reported delinquency

Researcher(s) & study Date(s) of title survey D. Armstrong, J. Hine, 2001 A. France : The On Track Youth Lifestyles Survey

Self-reported crime.book Page 184 Thursday, April 23, 2009 8:02 AM

Sub-national: Belfast, Downpatrick and Ballymena (Northern Ireland)

Geographical coverage Sub-national: 24 On Track areas of high deprivation and high crime across England and Wales

Method and response rate

Cross-sectional survey forming the basis for evaluation of On Track, part of the Home Office Crime Reduction Programme. One survey so far, but plans for repeat surveys. Primary, middle and secondary schools in neighbourhoods participating in On Track were sampled, so results are not nationally representative. Groupbased self-completion of questionnaires in school classrooms, mainly administered by teachers. No linkage to other data sources. School response rates were 74% for secondary schools, 79% of primary school and 100% of middle schools. Individual participation rates averaged 83%. Longitudinal study involving seven annual 4,620 males and females, sweeps (with one further sweep planned). Census of all school pupils aged 11/12 in 2000 in aged 11-12, followed up to the 3 sites. Includes 4500 in mainstream school age 18-19 + booster sample of 120 outside mainstream education. Group self-completion using paper and pencil questionnaire in school classrooms (43 schools). Separate family survey (parents and older siblings) for a sample of cohort members. 84% response rate overall.

Approx 30,000 males and females, aged 7-16

Sample

Self-Reported Crime and Deviance Studies in Europe

P.O. Wikström : Peterborough Adolescent and Young Adult Development Study (PADS+)

Main focus of study ‘Everyday’ moral crimes

2002-2006 General self(funding up reported to 2011) delinquency

Researcher(s) & study Date(s) of title survey S. Farrall, 2002 S. Karstedt

Self-reported crime.book Page 185 Thursday, April 23, 2009 8:02 AM

1,807 males and females, aged 25-65

Sample

Method and response rate

One-off cross-sectional study involving comparative elements with West and East Germany. Based on a national random sample of households, using CAPI questionnaires administered by interviewers. Response rate not published. Longitudinal study forming part of the SCoPiC Sub-national: 716 males research network, involving four annual sweeps and females, Peterborough and of data collection to date (with three further surrounding villages aged 12, (England) followed up to sweeps planned). Census of all school pupils age 16 (plans aged 11/12 in 2002. Group self-completion with to follow to 4 to 1 interviewer led questionnaire, 1 to 1 21) interview, 1 to 1 Space-Time Budget interview (spatial time diary) and 1 to 1 interviewer guided executive/emotive control laptop exercise. Separate survey of parents at age 12. Additional individual data collected from school records. Supported by Community Survey, census variables, land use data and police crime data. 72% response rate, with 98% retention rate amongst responders.

Geographical coverage National: England and Wales

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

185

186 General selfreported delinquency

S. Case, K. Haines : The Welsh Youth Survey

2003

Main focus of study General selfreported delinquency

Researcher(s) & study Date(s) of title survey D. Armstrong, J. Hine, 2001 A. France : The On Track Youth Lifestyles Survey

Self-reported crime.book Page 186 Thursday, April 23, 2009 8:02 AM

National: Wales

Geographical coverage Sub-national: 24 On Track areas of high deprivation and high crime across England and Wales Cross-sectional survey forming the basis for evaluation of On Track, part of the Home Office Crime Reduction Programme. One survey so far, but plans for repeat surveys. Primary, middle and secondary schools in neighbourhoods participating in On Track were sampled, so results are not nationally representative. Groupbased self-completion of questionnaires in school classrooms, mainly administered by teachers. No linkage to other data sources. School response rates were 74% for secondary schools, 79% of primary school and 100% of middle schools. Individual participation rates averaged 83%. One-off national survey based on an opportunity sample drawn from secondary schools across Wales, representing 14 of the 22 Local Authority areas. Group self-completion in school classrooms using a modified version of CASI, interactive computer-based self-interviewing (ICSI). No linkage to other sources of data. Response rates not reported.

Approx 30,000 males and females, aged 7-16

3,088 males and females, aged 11-16

Method and response rate

Sample

Self-Reported Crime and Deviance Studies in Europe

2003-2008

2004 Communities that Care : The Safer London Youth Survey

A.E. Bottoms, J.M. Shapland : Sheffield Pathways out of Crime Study

Researcher(s) & study Date(s) of title survey D. Wilson, 2003-2006 C. Sharp, A. Patterson : Offending Crime and Justice Survey

General selfreported delinquency

General selfreported delinquency

Main focus of study General selfreported delinquency

Self-reported crime.book Page 187 Thursday, April 23, 2009 8:02 AM

Approx 12,000 males and females, aged 10-65 in 2003; 5,000 aged 10-25 at 2004 onwards

Sample

Method and response rate

Repeated cross-sectional survey, with partial longitudinal panel design. Survey of private households, with boosted samples of 10-25 years olds (in 2003, for follow up in later sweeps) and ethnic minorities. Face to face interviews involving computer assisted personal interviewing (CAPI), computer assisted selfinterviewing (CASI) and computer assisted audio self-interviewing (A-CASI). Administered by interviewers. Response rate at 2005 (most recent) was 83% for panel sample and 70% for fresh sample. No collection of additional data. Sub-national: South 300 males and Longitudinal study forming part of the SCoPiC Yorkshire (England) females, aged research network. Following a sample of 20, followed offenders born in 1983, with at least two up to age 24 convictions, living in South Yorkshire. Individual 1-1 interviews with offenders carried out by academic researchers at nine-month intervals. Individual response rate not known. Sub-national: 11,458 males One-off cross-sectional survey of young people London (England) and females, living in six areas of London, forming part of a aged 11-15 crime prevention programme funded by the Government Office for London. Group-based self completion using paper questionnaires. Sample drawn from secondary schools and pupil referral units. Response rate not reported. No additional data collected.

Geographical coverage National: England and Wales

Self-Reported Crime and Delinquency Surveys in Great Britain and Ireland

187

Main focus of study General selfreported delinquency

General selfreported delinquency

Date(s) of survey 2003-2004, 2005, 20062007

2006

Researcher(s) & study title Northern Ireland Office : Northern Ireland Crime and Justice Survey

D. O'Mahony, M. Seymour, I. O'Donnell : Ireland ISRD-2

Self-reported crime.book Page 188 Thursday, April 23, 2009 8:02 AM

188 National: Northern Ireland and the Republic of Ireland

Geographical coverage National: Northern Ireland Repeated cross-sectional national survey, forming the NI equivalent of the Offending Crime and Justice Survey in England and Wales. Sample in 2005 included interviews in private households; interviews with offenders on probation and interviews with offenders in custody. Random selection of participants across the 3 populations. Individual 1-1 interviews, using CASI and Audio-CASI. 66% response rate. One-off national survey forming part of the ISRD-2. A multi-stage clustered design using stratification and random probability sampling. Five geographic areas chosen according to size and degree of urbanization, consisting of one large city (population size 495,781), one medium-sized city (population 123,062) and three small towns (population sizes 32, 500, 31,577 and 17,000). In each area, schools were randomly selected from a list of schools available from the Department of Education and Science. Group self-completion in school classrooms. 60% of sampled schools participated. Individual response rate not reported. Limited additional information collected from teachers.

Approx 3,000 to 3,500 males and females each year, aged 10 and over

1,570 males and females, aged 13-15

Method and response rate

Sample

Self-Reported Crime and Deviance Studies in Europe

Self-reported crime.book Page 189 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy Simona Traverso, Giada Cartocci, Giovanni Battista Traverso

Over the last twenty years the use of self-reports as tools of research in the field of criminology has increased at both international level, where they had already been in use for some time1, and nationally. Until the mid 1990s the criminological research scene in Italy was distinctly barren, with the exception of the studies carried out by Olivieri (1982) and Mariani (1987), both of which were restricted to the use of drugs, and the work of Ambroset and Pisapia (1980) on the social distribution of delinquency among Milanese youths.

I.

Overview of the survey conducted

Nowadays, however, self-report research focusing on the analysis of deviant behaviour is in continuous increase, to the extent that in this paper we report 21 studies (carried out between 1980 and 2007) that have made a valid contribution to the national and international body of literature on this subject (Appendix).

1.

Ambroset, Pisapia, 1980

Aim : the research aimed to achieve a quantitative evaluation of the dark number phenomenon in the context of deviant behaviour analysis. Sample : 198 subjects of between 15 and 19 years of age, with an equal number of males and females, most of whom were students (87.8%). Method : self-report questionnaire analyzing the most common deviant behaviours such as, for example, acts of vandalism, running away from home, sexual abuse, driving without a license, use of drugs, pickpocketing, shoplifting, snatching, stealing, possession of firearms.

1.

See for example: Porterfield (1946); Short, Nye (1957); Elmshorn (1965); Le Blanc, Coté (1974); Junger-Tas (1977); Zimmermann, Broder (1980); Canter (1982); Elliot, Huizinga (1983); Dunford, Elliot (1984); Gomme et al. (1984).

189

Self-reported crime.book Page 190 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Results : the data highlighted that Milanese youths do not commit socially significant deviant acts, even if they appear to make a substantial contribution to the dark number in relation to milder offences. Moreover, their behaviour was relatively stable in time, with the exception of aiding and abetting, pickpocketing and running away from home, which occurred sporadically. Concerning the relationship between deviant behaviour and gender, the results highlighted that there was an equal distribution between males and females for most crimes, while females were more often involved in pickpocketing and illegal possession of firearms (probably connected to the fact that they are more frequently the subjects of violence), the use of drugs and running away from home. The data highlighted: 앫 an increase of crimes committed by youths not responsible (< 14 years), 앫 an increase of crimes committed by girls, 앫 an increase of recidivism. Funding agencies : Institute of Criminology, Milan University.

2.

Olivieri, 1982

Aim : to quantify the spread of drugs in secondary schools in Verona. Sample : 16,844 secondary school students (8,136 boys and 8,708 girls) aged from 14 to 21. Method : self-report questionnaire containing three questions about the use of drugs. Results : the results showed that 10.7% of the students interviewed had used drugs at least once in the previous year and that males were much more addicted to substance abuse (13.3% vs 8.3%) than females. The study also highlighted that the use of drugs increased with age. Regarding the specific use of hard drugs, the survey revealed that 2.2% of the students had used them in the previous year, with a clear prevalence of males (3.1% vs 1.2%), especially in the last years of school. Funding agencies : Comune di Verona and Cassa di Risparmio di Verona, Vicenza e Belluno.

3.

Mariani, Protti, 1987

Aim : analysis of students’ behaviour in relation to the consumption of alcohol and other drugs. Sample : 1,718 secondary school students of both sexes, aged from 14 to 20 years, living in the Valle d’Aosta. 190

Self-reported crime.book Page 191 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

Method : self-report questionnaire containing 152 items. This study is similar to that conducted in Michigan during the project called “Monitoring the Future”, co-financed by National Institute on Drugs Abuse, and the Institute for Social Research of Michigan University. Results : this study revealed that 9% of the students interviewed admitted having used illegal substances at least once in their lives. Concerning alcohol consumption, 79% said they had consumed alcohol in the last 12 months, while in the same period 6%, 1.2% and 1.5% of the students (mainly males) had used respectively marijuana, cocaine or heroin. The key age was between 13 and 14, when 80% of students started to drink alcohol and smoke; between 14 and 17 there was an increase in the consumption of marijuana and hashish and from 18 students began to use cocaine and heroin. Funding agencies : Assessorato della Sanità e Assistenza Sociale.

4.

Gatti et al., 1991

Aim : the authors carried out a pilot self-report delinquency study in Genoa as part an international comparative research project (ISRD-1). Sample : 137 student and non-student youths, including 74 boys (54%) and 63 girls (46%) of 14-21 years of age (103 secondary school students attending scientific high school (liceo scientifico), or technical or vocational institutes and 34 “non-students” entrusted to social services) were interviewed between April and July 1990. Method : face-to-face interviews carried out by distributing a self-report questionnaire. The questionnaire is divided into three parts: the first is based on socio-demographic characteristics of students (age, sex…); the second ones is based on the deviant behaviours categorised in five types (property offences, violence against objects, violence against persons, drug offences, other youth-related offences); the third parts investigates the age onset of the commission of crime and the frequency of deviant behaviour during the last year. Results : firstly, the research highlighted that illegal behaviour is quite common among young people, although the mostly frequently committed crimes are not serious ones. The most common deviant behaviours are: travelling without a ticket (81.7%); illegal duplication of recording cassettes (80.3%); stealing (53%). Violence against persons is quite rare except engagement in riots (29%). Secondly, the difference between male and female deviance was confirmed, especially in relation to the type of crime committed, female deviance being

191

Self-reported crime.book Page 192 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

less serious (running away from home, obscene callings, graffiti, act of vandalism, stealing at home). Regarding the relationship between deviance and social class, although not univocal, the results showed that while on the one hand the “nonstudents”, belonging to the more deprived social classes, committed a greater number of deviant acts, on the other hand young people attending the vocational institute seemed to commit more crimes than those attending high school. Funding agencies : the research was supported by the Italian National Research Council (CNR).

5.

Russo, 1992

Aim : to test the applicability of the self-report technique in young people from Messina. Sample : 109 students and non-student youths: 90 students attending scientific high school, or technical or vocational institutes, of both sexes, aged from 14 to 21 years, and 19 “non-students” (18 boys and 1 girl) of whom 9 attended a social services facility, 9 were employed and 1 was attending a vocational course. Method : face-to-face interviews carried out by distributing a self-report questionnaire used by Gatti et al. (1991). Results : the youths interviewed had frequently been involved in deviant behaviour and some forms of deviance (such as travelling by bus without a ticket) were so common that they almost represented the rule. Males and non-students admitted a higher level of deviance (such as violence against objects, stealing, possession of firearms). Among the students deviant behaviour was reported more frequently by vocational school students. Finally, the author points out that drug consumption is less common in Messina than in Genoa. Funding agencies : the research was supported by the Ministero dell’Istruzione, dell’Università e della Ricerca.

6.

Traverso et al., 1994

Aim : the aim of the research was to test the applicability of the self-report technique in Italy. Sample : 94 secondary school students of both sexes (50 boys and 43 girls), aged from 14 to 21 years, living in Siena.

192

Self-reported crime.book Page 193 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

Method : face-to-face interviews carried out by distributing a self-report questionnaire used by Gatti et al. (1991). Results : the results obtained demonstrated that young people frequently commit acts of petty crime. Acts of vandalism and overtly violent acts appear to be rare, except during the period of the Palio horse race. Regarding substance abuse, only a few cases of use of soft drugs came to light, while alcohol abuse was found to be quite common in both sexes and raised some concern. With reference to gender, this study highlighted that female participation in petty crime is much higher in reality than suggested by official statistics. In relation to age, the data showed that older youths admitted to committing more serious illegal behaviour in groups. Funding agencies : the research was supported by the Italian National Research Council (CNR).

7.

Gatti et al., 1994

Aim : analysis of the data obtained from research in three Italian cities (Genoa, Siena, Messina) as part of the project ISRD-1. Sample : 1,009 secondary school students of both sexes (536 boys and 473 girls), aged from 14 to 19 years, living in the abovementioned cities. Method : the data were collected by means of “face-to-face” interviews. The questionnaire used was the same that it has been used in Genoa, Messina and Siena studies (Gatti et al., 1991). Results: the comparison of the research results in three cities (Siena, Genoa and Messina) showed that considering the total number of delinquent behaviours, the highest percentage was registered by Messina (72.2%), next came Genoa (64.5%) followed by Siena (56.6%). Drug abuse and property offences reached their highest percentage in Genoa; instead, in Siena, there was a low frequency of cases of graffiti, vandalism and driving without a license. The research showed that, while there was a difference between male and female pattern of deviance, there was not as great as official statistics suggested (graffiti, for example, showed a substantially similar frequency in both sexes). Finally, the data highlighted the fact that maladjustment, failure at school and especially family break-ups were among the most influential risk factors involved in the onset of deviant behaviour. Funding agencies : the research was supported by the Italian National Research Council (CNR).

193

Self-reported crime.book Page 194 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

8.

Genta et al., 1996

Aim : this study aimed to document the frequency and characteristics of bully/victim behaviours in primary and middle schools of two areas in Italy: Florence and Cosenza, representative of Central and Southern Italy respectively. Sample : 1,379 pupils took part in the study. In Florence (Central Italy) there were 246 pupils (126 boys, 120 girls) from 5 primary schools, and 538 pupils (299 boys, 239 girls) from 4 middle schools. In Cosenza (Southern Italy) there were 298 pupils (160 boys, 138 girls) from 4 primary schools, and 297 pupils (146 boys, 151 girls) from 4 middle schools. The age range was 8 to 11 years in primary school and 11-14 years in middle school. Method : the questionnaire, distributed at each school in May 1993, contained 28 single or multiple choice questions, and was filled in by pupils anonymously. It closely followed the design of those used by Olweus (1991) and Whitney and Smith (1993), with minor adaptations for the Italian version. Results : the bully/victim problem was appreciable in every school in which the questionnaire was administered. A large part of bullying appears to be dyadic (one child bullying another) and a large part involves several children (boys and girls): boys tend to bully both boys and girls, often with physical direct aggression but also with direct verbal aggression; girls tend to bully other girls, often with direct verbal aggression and spreading rumours. Whereas physical bullying tends to decline with age, verbal aggression remains high at older ages and indirect types of aggression, especially spreading rumours, actually increases in frequency. Levels of bullying were high in our two survey areas of Florence and Cosenza: in primary schools rates were particularly high in Florence; in middle schools the rates of bullying were noticeably higher among girls in Cosenza. If we compare these data on the amount of bullying with that from England (Sheffield), the total percentages in the Italian sample are about twice as high. Could methodological or procedural differences explain these findings ?, ask themselves the authors. Their answer is that this seems unlikely. The methodology used corresponds closely to that of Whitney and Smith (1993) in their English survey, with an identical definition of bullying and virtually identical questions asked. Nevertheless there could be some relevant cultural differences between the two countries. The semantic value of the term “bullying” may not correspond completely to the Italian translation of prepotenze. The latter may reflect a wider range of phenomena than the former. In order to verify this hypotesis in a small 194

Self-reported crime.book Page 195 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

sample of 45 primary school children the definition of bullying was read aloud by teacher and children were asked to name this type of behaviour. The Italian term most used by children was prepotenze, followed by violenze. In a second sample of 40 children the effect of the semantic value of this term has been tested administering the same questionnaire with the term violenze (prepotenze vs violenze). The data indicate that the use of term violenze does not make any significance differences in results. It is our opinion that it might be useful to replicate the survey in the near future to test the reliability of the results. Funding agencies : financial support of Prof. A. Fonzi and Dr. M.L. Genta (Department of Psychology, Florence University, Italy).

9.

Baldry, 1998

Aim : the aim of the study was twofold: to measure the prevalence of bullying in a middle school in Rome by administering a self-report questionnaire and to understand and interpret the results of the questionnaire by adopting a qualitative method. Sample : 183 students of both sexes (87 females and 96 males), aged from 11 to 14 years. Method : in the first part of the study, the author used the Italian version of the self-report questionnaire on bullying originally developed by Olweus. Results : pupils reported being bullied both “sometimes” or more often and “once a week” or more: three times more often than peers from other countries. This research reported an even more disturbing prevalence of bullying others at least “sometimes”: five times more frequently than international samples and almost double that of previous Italian investigations. This proportion increased further when referring to frequent episodes of bullying others. The most frequent types of bullying behaviours which girls experienced were: name calling, teasing, having malicious rumors spread. Boys were more often threatened, physically attacked and called nasty names. This difference from the findings of international research could be related to the semantic meaning of the word used in Italian to refer to “bullying” (prepotenze) which has broader meaning in the Italian language, mainly referring to any behaviour that expresses the exercise of power over others. Another possible explanation is that the school from which the students were recruited was located in a problematic area of Rome. Funding agencies : Department of Social and Developmental Psychology, Faculty of Psychology, Rome University.

195

Self-reported crime.book Page 196 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

10. Baldry, Farrington, 1999 Aim : the aim of this research was to investigate the prevalence of bullies and victims and the types and places of bullying. Sample : 113 girls and 125 boys aged between 11 and 14 years. Students came from a representative middle school in Rome. Method : the questionnaire was based on the Italian version (Genta et al., 1996) of the original questionnaire developed by Olweus (1993). Results : bullying was found to be quite prevalent among Italian pupils. Over half of all students had bullied others, and nearly half had been bullied. Boys bullied more than girls, and girls were somewhat more likely than boys to be bullied sometimes or more often. Boys were more likely to suffer direct bullying such as being threatened or physically hurt. Surprisingly, boys were almost as likely as girls to suffer indirect bullying such as being rejected and having rumours spread about them. In this sample both boys and girls were bullied by one or several boys. Most of the bullying took place in the classroom, with lesser amounts in the corridors or playground. Boys were more likely than girls to be bullied in the toilets. Funding agencies : University of Cambridge, Institute of Criminology.

11. Baldry, Farrington, 2000 Aim : to investigate the personal characteristics and parental styles of bullies/delinquents, only bullies, only delinquents and controls (nondelinquent/non-bullies). Sample : 238 middle schools students of both sexes (125 boys and 113 girls), aged from 11 to 14 years, living in Rome. Method : this research used the bullying questionnaire of Genta et al. (1996) and the self-report delinquency questionnaire of Gatti et al. (1994). Results : bullying and delinquency were more common among boys than among girls. However, while bullying did not vary significantly with age, delinquency increased with age. Bullying and delinquency were especially common in boys and older students. Only bullies were younger, while only delinquents were older, suggesting that bullying might be an early stage of a developmental sequence leading to delinquency. Only bullies and only delinquents had different parenting correlates; only bullies had authoritarian parents and disagreed with their parents, whereas only delinquents had conflictual and unsupportive parents. This suggested that bullying and delinquency are not simply different behavioural manifestations of the same underlying construct. Funding agencies : University of Cambridge, Institute of Criminology. 196

Self-reported crime.book Page 197 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

12. Baldry, 2003 Aim : the study aimed to investigate the relationship between bullying and victimisation in schools and exposure to interparental violence in a nonclinical sample of Italian youngsters. Sample : 1,059 Italian children (48.5% girls and 51.5% boys), aged from 8 to 15 years, recruited in the city and province of Rome. Method : students were asked to sit separately to prevent conferring, talking or helping when filling in the original bullying questionnaire developed by Olweus (1993). Results : results indicated that bullying and victimisation are associated with domestic violence, though exposure to domestic violence per se does not fully predict behaviour at school, especially in the case of boys. A father harming the child was found to be a risk factor for both bullying and victimisation, as well as being a boy rather than a girl. Funding agencies : the research was supported by the Italian National Research Council (CNR).

13. Eurobarometer, 2003 Aim : to investigate through self-report questionnaire areas such as health, food and alcohol and safety. Sample : the number of participants, aged 15 years and over, was 1,006. Method : the standard Eurobarometer covers the population of the respective nationalities of the European Union Member States, resident in each of the Member State. The basic sample design applied in all Member States is a multi-stage, random (probability) one. In each EU country, a number of sampling points was drawn with probability proportional to population size (for a total coverage of the country) and to population density. For doing so, they were drawn systematically from each of the “administrative regional units”, after stratification by individual unit and type of area. They thus represent the whole territory of the Member States. In each of the selected sampling points, a starting address was drawn, at random. Further addresses were selected as every Nth address by standard random route procedures, from the initial address. In each household, the respondent was drawn, at random. All interviews were face-to-face in people’s home and in the appropriate national language. For each country a comparison between the sample and the universe was carried out. Results : The results underline that Italy with Portugal and Spain had the lowest percentage of person who had drunk alcohol (wine, beer, spirits,

197

Self-reported crime.book Page 198 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

other alcoholic drinks) in the previous four weeks with a rate of 55.4%, 50.9% and 49.9% respectively (Figure 1).

Source : Eurobarometer 186 "Health, Food and Alcohol and Safety"

Fig. 1: Drinking Patterns of Europeans, 2002

Of those who drank, Italians with Portuguese and Spaniards did so the most, measured by days on which alcohol was drunk. The intensity of drinking, that was to say, how many drinks were consumed at one sitting, also varied from country to country; in particular, Italy, ranked last at 1.49 drinks per day (Figure 2).

Source : Eurobarometer 186 "Health, Food and Alcohol and Safety"

Fig. 2 : Drinking Patterns of Europeans, 2002

198

Self-reported crime.book Page 199 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

The consumption of different types of alcohol – beer, wine or spirits – also differed from country to country. Italy, which was one of the countries that had the highest number of days upon which alcohol was drunk, also had the highest number of days upon which wine was drunk (12.35), suggesting that drinking wine was a social event belonging to everyday life. Italy was also one of the top countries for drinking “mainly/only when eating”. About the average age at which Europeans began drinking is 14.57 years with a range of 12.18 for Italy (Figure 3). There is a relationship between age of first drink and what percent of a country’s national drank in the previous four week. In general, looking at those countries in which a higher percentage have drunk alcohol in the previous four week, the age at which people began to drink is slightly lower and in general the research underlines that the age of the first drink has been continuously decreasing.

Source : Eurobarometer 186 "Health, Food and Alcohol and Safety"

Fig. 3: Average age of first drink, 2002

14. Baldry, 2004 Aim : this study aimed to empirically investigate Italian middle school students’ attitudes towards bullies and victims according to the way in which bullying took place (alone or in a group) and whether the gender of bullies and victims in question was the same as or different from that of respondents.

199

Self-reported crime.book Page 200 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Sample : 117 students (49 boys and 68 girls), aged 11-12, recruited from a middle school in Italy, were randomly allocated to one of four independent groups according to experimental conditions: individual bullying among girls, individual bullying among boys, group bullying among girls, group bullying among boys. Method : participants watched one of four versions of a video according to experimental conditions, showing a brief standardised bullying episode taking place at school; they then had to fill in a self-report questionnaire measuring the dependent variables: respondents’ positive or negative judgments towards the bully and the victim shown in the video and to what extent the victim was blamed for what had happened. The bullying episode, chosen as a baseline stimulus, was played in four different ways according to the experimental conditions: a) a girl bullied by a group of five girls; b) a girl bullied by one girl alone; c) a boy bullied by a group of five boys; d) a boys bullied by a boy alone. Results : results indicated that students have positive attitudes towards the victims of bullying and tend not to blame them for what has happened; moreover, male respondents blamed the victim less when the bullying took place among boys rather than among girls; female respondents, on the other hand, blamed the victim less when the bullying took place among girls rather than among boys. Gender differences also indicated that when a girl is bullied by a group of students rather than by a single student, she is blamed more; the reverse applied in the case of male victims. Funding agencies : Department of Clinical Psychology, Free University of Amsterdam, The Netherlands.

15. Baldry, 2004 Aim : the study aimed to investigate, through a cross-sectional study, the separate effects of direct and indirect bullying and victimisation on internalising symptoms, moderated by having a positive relationship with the father and/or with the mother. Sample : 661 adolescents (54.2% boys and 45.8% girls) recruited from ten different middle schools in Rome and suburbs, aged from 11 to 15 years. Method : bullying and victimisation were measured using the Italian version of the original bullying questionnaire developed by Olweus (1993). Three classes from each of the ten schools involved were then randomly selected, for a total of 30 classes. Results : findings from this study showed that the mental and physical health of Italian middle school students is affected by involvement in bullying and victimisation at school.

200

Self-reported crime.book Page 201 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

Bullying was reported by a high proportion of students, especially boys. Indirect bullying was reported evenly by boys and girls. With regard to victimisation, gender differences occurred only for direct victimisation (boys reporting higher prevalence rates). Concerning the internalisation of problems, girls tend to react internally, and therefore feel more depressed and sad, whereas boys tend to react overtly to problems encountered at school. Finally, none of the interactions with the mother or father were significant, and there is evidence that bullying and victimisation affect the mental and physical health of children independently of gender. Being a girl is per se correlated with worse health conditions, but does not vary according the various bullying and victimation instances experienced. Funding agencies : the research was partially supported by the National Research Council (CNR).

16. Italian population survey on alcohol and other drugs (IPSADItalia), 2006 Aim : this research was aimed at documenting the frequency and characteristics of use/abuse of legal and illegal substances in order to create new strategies of prevention. Sample : was represented by the collection of the different samples of surveys led in 2001, 2003 and 2005. Sample age ranges in between 15 to 64 years. The sample was built by a stratification of 103 Italian provinces divided by factors concerning their geographical location, population density and the prevalence of drug abuse measured through a multivariate index (SMAD Index) (Di Fiandra, Mariani, 1984). Method : the study used a self-report questionnaire mailed to a sample of responders and it investigated their socio-demographic characteristics and their use/abuse of legal and illegal substances. Afterwards, a simple random sample of the provinces was built from which the final list of interviewees was extracted. In accordance with other studies such as De Leeuw’s (1990) research and more recently the EMCDDA project2, the authors considered this methodology as a valid and reliable instrument to investigate some private issues. Furthermore, the self-report questionnaire mailed was characterised, in their opinion, by a minor error of measurement.

2.

De Winter et al., 2000.

201

Self-reported crime.book Page 202 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Results : About the use/abuse of drugs, during the period between 20012005, IPSADItalia data showed that there was an increase in the general population of consumption of cannabis (the interviewees that used cannabis during 2001 at least once in a lifetime were 22% while during 2005 the rate of consumers was 32%). 앫 Heroin – The abuse of heroin during the last 30 days remained stable in times while abuse of cocaine, in the same period, showed a slight increase. The regions that showed a significant increase of phenomena once or more times during the last 12 months are Lazio in the abuse of cannabis (10,6%), Lombardia in the abuse of cocaine (4,7%) and Liguria in the abuse of heroin (0,7%). 앫 Alcohol – The research underlined a decrease of consumption in general population (overall males) during the period between 20032005. However, it was important to highlight the presence of new consumption models among youths characterised by a younger average age for first time consumption. 앫 Tobacco – The study showed a general diminution during the period 2001-2005 (at least once during the last 12 months: 36% vs 32%). 앫 Smoke – About relationship between smoke and gender, the research underlined an increase of female smokers (+6,4%) compared to a diminution of males smokers (-6,6%).

17. European school survey project on alcohol and other drugs (ESPADItalia), 2006 Aim : this research aimed to document the frequency and characteristics of use/abuse of legal and illegal substances in order to create new strategies of prevention. Sample : was constituted by the collection of the different samples of surveys led during the period between 2001- 2006. The sample was representative of secondary school students. Method : The ESPAD survey was conducted with a standardised methodology including an identical questionnaire to provide as comparable data as possible. Data were collected by group-administered questionnaire in schools on nationally representative samples of classes. Teachers or research assistants collected the data. The students answered the questionnaires anonymously in the classroom. The self-report questionnaire investigated some socio-demographic characteristics of secondary school students and the use/abuse of legal and

202

Self-reported crime.book Page 203 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

illegal substances. The study concerned the consumption of drugs during their lives, in the last 12 months and in the last month. Results : The ESPADItalia research underlined that the consumption of drugs among students increased with the age and this result was present during the whole period considered (2001-2005). The use of cannabis (once or more times during the last 12 months) was referred by 24.5% of the students while the rate of the abuse of cocaine and heroin, at the same time, was respectively 4% and 1.6%. Within these overall prevalence rates, the prevalence of users with more problematic substance abuse habits (more than twenty times in one year) is lower for cocaine (12%) and heroin (16%) than for cannabis (26%). In contrast to the results of IPSAD Italia, that analyzed the general population, ESPADItalia underlined an increase of alcohol abuse among students: during 2000 the prevalence was 64,7% while during 2006 was 69,6%. Another important result was the increase of cigarette use above-all by females. About poliabuse, this phenomena appeared quite prevalent and the rate was 22% during the period between 2005-2006. Finally, it is important to underline that in both studies, IPSAD and ESPADItalia, the most prevalent abuse was represented by the consumption of cannabis without other associated drugs.

18. Coluccia, 2006 Aim : the research project aimed to collect data on young people in Siena and on deviant behaviour among them. Sample : 1,215 secondary school students of both sexes (645 boys and 564 girls), aged from 14 to 21 years, resident in the province of Siena. Method : the study was carried out using the self-report technique. Two questionnaires were used; one aimed at students and the other at their families. The two different types of questionnaire investigated four different areas: 앫 lifestyle 앫 the relationship between students and school and among students (analyzing phenomena such as bullying) 앫 perception of the family about deviant behaviour 앫 analysis of deviant behaviour Results : the research revealed that the most common deviant behaviours were: the consumption of soft drugs and spirits; regular involvement in

203

Self-reported crime.book Page 204 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

fights; dangerous behaviour and the habitual expression of physical and verbal violence towards people close to them (peers and teachers). This study also revealed a significant coincidence between the behaviour self-reported by the students and their behaviour as reported by their families, suggesting that the families had a high level of knowledge and awareness of their children. This research was based on the Segre’s Theory (1996) that aimed to unify the different explanations of deviant behaviour (Strain Theory; Theory of Social Control; Matza Theory; Theory of Social Disorganisation; Theory of Labelling) in a single complex theory. Funding agencies : Province of Siena.

19. Aspetti della vita quotidiana, ISTAT, 2006 Aim : the multi-objective investigation, conducted during 2006, was based on the administration of self-report questionnaire analyzing different areas of life (school, work etc.) including the use/abuse of alcohol. Sample: the sample, aged over 11 years of age, was represented by 49,000 individuals belonged to 835 Italian towns. Method : administration of self-report questionnaire to the family members. The sample design started from a stratification of all towns out of which clusters of families were drawn up systematically from sociodemographic lists. Information was collected through PAPI technique (Paper and Pencil Interview) and through interviews with all family members. Results : the results highlighted that the model of alcohol consumption is “moderate” in Italy because, in most cases, people drank during lunch or dinner. In general, the phenomena of alcohol consumption showed a stable trend in the last nine years with a slight increase during 2001 but the most important results were that there was a significance increase among youths and above-all among women (Figure 4). During the period between 19982006, the consumption of alcohol in the general population was stable (around 70%), while the research underlined an increase (57.6% to 59%) among women aged 20 to 24 years. However, Italy, in contrast with other countries, was at the bottom for alcohol consumption during the last 12 months (Figure 5). The most important result was the consumption of alcohol among youths at a younger age, with a range between 11-15 years: 18.6% of youths claimed to have drunk during the last year. The study highlighted that in youths between 14-17 years of age, there was an important increase of drinking between meals: 20.5% in 2006, compared to 12.6% in 1998. This result showed a pattern of consumption of drugs

204

Self-reported crime.book Page 205 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

similar to that ofthe Northern European countries: a greater use of alcohol between meals and an increase of binge drinking.

Source : ISTAT

Figure 4: Youths aged 14-17 drinking between meals according to gender

Source : ISTAT 2007

Figure 5: Population percentage, aged 15 years and over, drinking during the last 12 months per country

20. Gatti et al., 2007 Aim : this study attempted to assess the frequency and features of deviant behaviour among youths in Italy by analyzing a large, representative sample of students in 15 cities/towns of different sizes located in different geographical areas. Sample : 7,278 students of both sexes (3532 boys and 3746 girls) attending two different types of school: primary schools (age group 12-14) and 205

Self-reported crime.book Page 206 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

secondary schools (high school, technical institutes and vocational institutes – age group > 14). Method : the Italian survey used a city-based sample comprising 15 cities/ towns. Normally, three primary schools and three secondary schools (one school for each type of study) were sampled for each city. A self-report questionnaire compiled between 16.02.2006 and 03.06.2006 was used. A pilot study was conducted in the city of Brescia in order to assess the reactions of students, questionnaire administration time and any difficulties that might arise. Results : analysis of the results revealed that 45.7% of the youths interviewed admitted having broken the law at least once in their lives and 31.3% admitted having done so in the previous year. The offences most frequently admitted were: participation in group fights (27.8% of the total), shoplifting (19.5%), act of vandalism (16.3%), computer hacking (10.8%) and carrying a weapon (10.5%). With regard to drugs-related crimes, the survey recorded a frequency of 4.2% of drugs dealing; the use of LSD/ heroin/cocaine was admitted by 2.3% of respondents, while that of XTC/ speed was admitted by 2.0%. The results reported were descriptive; it was not claimed that they provided any observations of causal nature, which would require the application of multivariate analyses. What emerged from these results was that deviant behaviours were more widespread than the official statistics indicate; that they were more frequent among males than females; that they increased with age, and that they were more common among youths attending vocational institutions and less common among high school students. The prevalence of unlawful behaviours did not appear to be correlated in a homogeneous, systematic and significant manner with the size of the city/ town of residence or with its geographical location (north, south or center), as some deviant behaviors were prevalent in some cities and others in others. Being a member of a youth gang or deviant group implied a far greater likelihood of committing crimes and of using alcohol and drugs; it also seemed to constitute a significant risk factor for victimisation (with the exception of bullying). Funding agencies : the research was partially supported by the Ministero dell’Istruzione, dell’Università e della Ricerca.

206

Self-reported crime.book Page 207 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

21. Melossi et al. 2008 Aim : this research aimed to examine possible links between modes of socialisation – especially their problematic and conflictual aspects – and the incidence of deviance and criminal behaviour. Sample : 335 students (177 males and 158 females) of a mean age of 13-14 years attending 19 third year classes of three middle schools in Bologna. Method : self-report questionnaire containing 108 items that investigated three different variables of students: a) socio-demographic characteristics; b)socio-economical factors; c) family background. Results : this research did not identify a greater inclination to commit crimes among foreigners and there did not seem to be a consistent positive relationship between being foreign and an increased rate of self-reported deviance. The risk factors that seemed most likely to lead to the onset of deviant behaviour were male sex, a diminished sense of wellbeing, and weak family ties: factors that can relate to both Italians and foreigners in Italy. In Melossi’s study, the theoretical frameworks considered most useful to explain his results were the Theory of Cultural Conflict (Sellin, 1938), the Social Control Theory (Hirschi, 1969) and the Theory of Labelling (Becker, 1987; Matza, 1976). Funding agencies : project co-financed by the Ministero dell’Istruzione, dell’Università e della Ricerca.

II.

Policy implications

It is important to underline that in the case of bullying, the Government established a toll free telephone number (numero verde antibullismo) that helps people to tackle this important problem that has emerged only in recent years. The data underline that the current number of calls is 12,874 with an average of 70 per day. Calls are above-all made by parents (36.1%) and teachers (21.2%). The rate of students is only 13.4%. 53.1% of calls ask for advice on correct behaviour while 8.9% of the callers only want to be heard. The Ministry of Education conducts a research “Listening Project” (Progetto ascolto) in primary school (sample is represented by 11,000 students) to verify the diffusion of this phenomena. The results state that there is a good relationship among students (95%), but at the same time 74% of pupils have assisted at least once in the act of bullying among students.

207

Self-reported crime.book Page 208 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

In this way, important positive initiatives about bullying prevention (promoted by Guiseppe Fioroni, Minister of Education) are been created at national level. Government organises campaigns of communication and dissemination of information targeting not only teachers but also families and students in order to gain a better understanding of the characteristics of this phenomena. In particular, discussions in primary schools are generally conducted among teachers and students while the secondary schools organise campaigns of information. In addition, Permanent Regional Observatories have been established to quantify the prevalence of bullying. Finally, the operative strategies are of two types: promotion of strategy of prevention and action against bullying with a constant monitoring of the phenomena. About use/abuse of alcohol and drugs among the students, the ESPADItalia data, contained in the Annual Report of Parliament (2006), shows that the prevention activities are different in relation to the part of country considered: in 21.2% of southern schools and islands, there is no prevention project; in the north and in the centre of the country some prevention activities are present with a rate of 9.6% and 10.8% respectively. It is possible to underline a different typologies of prevention program: in vocational institutes there are larger initiatives with adequate teacher training and the presence of counselling for students in case of difficulty. Moreover, the prevention activities are addressing the family with the aim to improve family communication and to determine the main risk factors of drug abuse. Another objective is to pay attention to risk groups such as immigrants, students with school problems and maladjustment, to give valid support.

III. Good practice About good practice, it is necessary to say that in our scientific literature there are no specific studies talking about the methodological aspects of self-report studies. What is possible to underline is that the main part of researches mentioned above have been carried out under strictly controlled conditions and administered in person by trained interviewers. Some studies also provide targeted help and support to children with reading difficulties to ensure successful completion of the questionnaire.

208

Self-reported crime.book Page 209 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

IV. Conclusion Findings from this overview show that the use of self-report delinquency studies in Italy is on the increase and provide more in-depth information about the extent of some deviant phenomena. Specifically, the results highlight that 앫 although deviant behaviours are rather common, most of the crimes committed are minor; the most important risk factors involved in the onset of deviant behaviours are maladjustment, scholastic failure and especially family break-up; 앫 female deviance was found to be more frequent than reported by the official judicial statistics; 앫 among the types of deviant behaviour analysed, it came to light that: a) bullying is increasing and involves both males and females, albeit in different ways; b) the use of alcohol and spirits begins at an early age; c) the use/abuse of soft drugs is concentrated in the 13-14 years age group; d) the use of hard drugs is prevalent from eighteen year onwards. Concerning prevention, during the last few years, Government has tried to improve the diffusion of much more detailed information concerning the main social problems, in particular of youths, such as violence in school (bullying) and use/abuse of legal and illegal substances with many programs and a toll free help line.

References AMBROSET S., PISAPIA G., 1980, Numero oscuro della devianza e questione criminale, Verona, Bertani Editore. BALDRY A.C., 1998, Bullying among Italian middle school students, School Psychology International, 19, 4, 361-374. BALDRY A.C., 2003, Bullying in schools and exposure to domestic violence, Child Abuse and Neglect, 27, 7, 713-732. BALDRY A.C., 2004a, ’What about bullying?’. An experimental field study to understand students’ attitudes towards bullying and victimisation in Italian middle schools, British Journal of Educational Psychology, 74, 116. BALDRY A.C., 2004b, The impact of direct and indirect bullying on the mental and physical health of Italian youngsters, Aggressive Behavior, 30, 5, 343-355.

209

Self-reported crime.book Page 210 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

BALDRY A.C., FARRINGTON D.P., 1999, Types of bullying among Italian school children, Journal of Adolescence, 22, 3, 423-426. BALDRY A. C., FARRINGTON D. P., 2000, Bullies and delinquents: Personal Characteristics and parental styles, Journal of Community & Applied Social Psychology, 10, 1, 17-31. BECKER H.S., 1987, Outsiders. Saggi di sociologia della devianza, Torino, Edizioni Gruppo Abele. CANTER R.J., 1982, Sex differences in self-report delinquency, Criminology, 20, 373-93. COLUCCIA A., 2006, I comportamenti giovanili nelle relazioni e nel disagio, Milano, FrancoAngeli. DE LEEUW E.D., 1990, Data quality in mail, telephon and face to face surveys: a quantitative review, Technical Report no 1°, Response effects in surveys, Vrije Universiteit, Amsterdam. DE WINTER M., COHEN P., LANGEMIJER M., VICENTE J., HARTNOLL R. 2000, Methodological study to compare the effect of different methods of data collection on the prevalence of self-report drugs use in General Population Surveys, EMCDDA Scientific Report, Project CT. 97.EP.02, http:// www.emcdda.europa.eu/attachements.cfm/att_1388_EN_selfreported_surveys_report.pdf. DI FIANDRA T., MARIANI F., 1984, Sistema di monitoraggio dell’abuso di droghe, Progetto finalizzato tossicodipendenze, Consiglio Nazionale delle Ricerche, Roma. DUNFORD F.W., ELLIOT D.S., 1984, Identifying career offenders with Selfreport data, Journal of Research in Crime and Delinquency, 21, 1, 57-86. ELLIOT D.S., HUIZINGA D., 1983, Social class and delinquent behaviour in a National youth panel 1976-1980, Criminology, 21, 149-177. ELMSHORN K., 1965, Study in Self-Reported Delinquency among School Children in Stockholm, in CHRISTIANSEN K.O., Ed., Scandinavian Studies in Criminology, vol. I, Tavistock, London, 117-146. ESPADITALIA (European school survey project on alcohol and other drugs), 2006, in Relazione annuale al Parlamento sullo stato delle tossicodipendenze in Italia, Ministero della Solidarieta Sociale, Roma, http://www.solidarietasociale.gov.it/nr/rdonlyres/3e62241f-4791-4e55b17a-800f48c4faf2/0/relazione_parlamento_2006.pdf. GATTI U., MARUGO M.I., MALFATTI D., TARTARINI, E.., 1991, La diffusione dei comportamenti devianti tra i giovani: una ricerca sulla popolazione genovese mediante la tecnica dell’autoconfessione (self-report), Rassegna Italiana di Criminologia, 2, 387-412. GATTI U., FOSSA G., LUSETTI E., MARUGO M.I., RUSSO G., TRAVERSO G.B., 1994, La devianza 'nascosta' dei giovani. Una ricerca sugli studenti di tre città italiane, Rassegna Italiana di Criminologia, 2, 247-267. 210

Self-reported crime.book Page 211 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

GATTI U. et al., 2007, Self-Reported Juvenile Delinquency in Italy, paper presented to the conference of the Italian Society of Criminology, Gargnano, Ottobre 2007. GENTA M. L. et al., 1996, Bullies and victims in schools in central and southern Italy, European Journal of Psychology of Education, 11, 97-110. GOMME I. M., MORTON M. E., WEST W.G., Rates, 1984, Types and Patterns of Male and Female Delinquency an Ontario County, Canadian Journal of Criminology, 26, 3, 213-224. HIRSCHI T., 1969, Causes of delinquency, University of California Press, Berkeley. IPSADITALIA (Italian population survey on alcohol and other drugs), 2006, Relazione annuale al Parlamento sullo stato delle tossicodipendenze in Italia, Ministero della Solidarieta Sociale, Roma, http:// www.solidarietasociale.gov.it/nr/rdonlyres/3e62241f-4791-4e55-b17a800f48c4faf2/0/relazione_parlamento_2006.pdf. ISTAT, 2006, The Multi-objective Investigation “Aspetti della vita quotidiana” (Aspects of everyday life), ISTAT, Roma. JUNGER-TAS J., 1977, Delinquency and Judicial selection in Belgium, in FRIDAY P., STEWART W. L., Eds., Youth Crime and Juvenile Justice, New York/London, Praeger 70-95. LE BLANC, COTÉ G., 1986, Milieu et style de vie, régulation sociale, fonctionnement psychologique et conduite délinquante: une comparaison de l’adolescence des jeunes de 14-15 ans en 1974 et en 1985, in TREMBLAY R., LE BLANC M., SCHWARTZMANN A. E., (Dir.), La conduite délinquante des adolescents à Montréal (1974-1985): étude descriptive et prédictive, École de Psycho-éducation, Montréal, École de Criminologie. Linee di indirizzo per la prevenzione e la lotta al bullismo, Roma, Ministero di Grazia e Giustizia, Presidenza del Governo Italiano. MARIANI F., PROTTI M. A., 1987, Atteggiamenti e comportamenti degli studenti delle scuole secondarie superiori della Valle d’Aosta nei confronti del consumo di tabacco, alcool ed altre droghe, Regione Autonoma Valle d’Aosta, 1987. MATZA D., 1976, Come si diventa devianti, Bologna, Il Mulino. MELOSSI D., DE GIORGI A., MASSA E., 2008, Minori stranieri tra conflitto normativo e devianza: la seconda generazione si confessa?, Sociologia del Diritto, 2, 99-130 OLIVIERI D., 1982, La diffusione della droga nelle scuole secondarie superiori di Verona: indagine statistica e risposta casualizzata, Vicenza, La grafica & Stampa s.r.l.,. OLWEUS D., 1991, Bull/victim problems among schoolchildren: basic facts and effects of a school-based intervention programme, in Pepler D., 211

Self-reported crime.book Page 212 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Hillsdale R. K., Eds., The development and Treatment of childhood aggression, NJ, Erlbaum, 411-448. OLWEUS D., 1993, Bullying at school: what we do and what we can do, Blackwell, Oxford. PORTERFIELD A. L., 1946, Youth in trouble. Studies in delinquency and despair with plans for prevention, Fort Worth, Leo Portishman Foundation. RUSSO G., 1992, La devianza autorilevata. Indagine condotta tra i giovani messinesi, Rassegna Italiana di Criminologia, 3, 515-535. SEGRE S., 1996, Una teoria integrata della delinquenza giovanile alla luce della recente letteratura nordamericana, in Segre S., Ed., La devianza giovanile. Cause sociali e politiche di prevenzione, Milano, FrancoAngeli, 23-39. SELLIN T., 1938, Culture, conflict ant crime, New York, Social Science Research Council. SHORT J.F., NYE F.I., 1957, Reported behaviour as a criterion of deviant behaviour, Social Problem, 5, 207-213. TRAVERSO G.B. et al., 1994, I risultati di uno studio pilota sulla delinquenza giovanile condotto a Siena con la tecnica dell’autorilevazione, Rassegna Italiana di Criminologia, 3, 397-410. WHITNEY I., SMITH P.K., 1993, A survey of the nature and the extent of bullying in junior, middle and secondary schools, Educational Research, 35, 1, 3-25. ZIMMERMAN J., BRODER P. K., 1980, A comparison of different delinquency measures derived from self-report data, Journal of Criminal Justice 8, 3,147-162.

212

Date(s) of survey 1980

Olivieri: La diffusione 1982 della droga nelle suole secondarie superiori di Verona: indagine statistica e risposta casualizzata Mariani, Protti: 1987 Atteggiamenti e comportamenti degli studenti delle scuole secondarie superiori della Valle d’Aosta nei confronti del consumo di tabacco, alcool ed altre droghe

Researchers(s) & study title Ambroset, Pisapia: Numero oscuro della devianza e questione criminale Verona

Aosta

Deviant behaviour

Substance abuse

Main focus of the Geographical study coverage Milan Deviant behaviour

1,718 (14-20 years)

Self-report questionnaire. The key age was between 13 and 14, when 80% of students started to drink alcohol and smoke.

Self-report questionnaire. The data highlighted: a) an increase of crimes committed by youths not responsible (< 14 years); b) an increase of crimes committed by girls; c) an increase of recidivism. Self-report questionnaire. The study highlighted that the use of drugs increased with age and that there was a clear prevalence of males.

198 (15-19 years)

16,844 (14-21 years)

Method and response rate

Sample

Appendix. List of Self reported Delinquency Surveys in Italy

Self-reported crime.book Page 213 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

213

Researchers(s) & study title Gatti et al.: La diffusione dei comportamenti devianti tra i giovani: una ricerca sulla popolazione genovese mediante la tecnica dell’autoconfessione (self-report) Russo: La devianza autorilevata. Indagine condotta tra i giovani messinesi Traverso et al.: I risultati di uno studio pilota sulla delinquenza giovanile condotto a Siena con la tecnica dell’autorilevazione Gatti et al.: La devianza nascosta dei giovani. Una ricerca sugli studenti di tre città italiane

214 Deviant behaviour

Deviant behaviour

Deviant behaviour

1992

1994

1994

Genoa, Siena, Messina

Siena

Messina

Main focus of the Geographical study coverage Deviant Genoa behaviour

Date(s) of survey 1991

Self-reported crime.book Page 214 Thursday, April 23, 2009 8:02 AM

1,009 (14-21 years)

94 (14-21 years)

109 (14-21 years)

137 (14-21 years)

Sample

Self-report questionnaire. The comparison of the research results in three cities (Siena, Genoa and Messina) shows that considering the total number of delinquent behaviours, the highest percentage is registered by Messina (72.2%), next comes Genoa (64.5%) followed by Siena (56.6%).

Self-report questionnaire. Males and nonstudents admitted a higher level of deviance (such as violence against objects, stealing, possession of firearms). Self-report questionnaire. The results obtained demonstrated that young people frequently commit acts of petty crime.

Self-report questionnaire. The research highlighted that illegal behaviour is quite common among young people, although the mostly frequently committed crimes are not serious ones.

Method and response rate

Self-Reported Crime and Deviance Studies in Europe

Bullying

Bullying

Bullying

Bullying

Baldry: Bullying among 1998 Italian middle school students

Baldry, Farrington: 1999 Types of bullying among Italian school

Baldry, Farrington: 2000 Bullies and delinquents: personal characteristics and parental styles

Baldry: Bullying in 2003 schools and exposure to domestic violence

Rome

Rome

Rome

Rome

Main focus of the Geographical study coverage Bullying Florence, Cosenza

Researchers(s) & Date(s) of study title survey Genta et al.: Bullies and 1996 victims in schools in central and southern Italy

Self-reported crime.book Page 215 Thursday, April 23, 2009 8:02 AM

1,059 (8-15 years)

238 (11-14 years)

238 (11-14 years)

Self-report questionnaire. Bullying was found to be quite prevalent among Italian pupils. Over half of all students had bullied others, and nearly half had been bullied. Self-report questionnaire. Bullying and delinquency were especially common in boys and older students. Only bullies were younger, while only delinquents were older, suggesting that bullying might be an early stage of a developmental sequence leading to delinquency. Self-report questionnaire. Results indicated that bullying and victimisation are associated with domestic violence, though exposure to domestic violence per se does not fully predict behaviour at school.

Self-report questionnaire. Levels of bullying were high in our two survey areas of Florence and Cosenza: in primary schools rates were particularly high in Florence, but in middle schools the rates of bullying were noticeably higher among girls in Cosenza. Self-report questionnaire. Pupils reported being bullied both “sometimes” or more often and “once a week” or more.

1,379 (8-14 years)

183 (11-14 years)

Method and response rate

Sample

Self-Report Delinquency in Italy

215

216

2003

Bullying

Alcohol abuse

Deviant behaviour

1994

Rome

Italy

Genoa, Siena, Messina

Main focus of the Geographical study coverage Deviant Siena behaviour

Date(s) of survey 1994

Baldry: ‘What about 2004 bullying?’ An experimental field study to understand students’ attitudes towards bullying and victimisation in Italian middle schools

Eurobarometer: Health, food, and alcohol and safety

Researchers(s) & study title Traverso et al.: I risultati di uno studio pilota sulla delinquenza giovanile condotto a Siena con la tecnica dell’autorilevazione Gatti et al.: La devianza nascosta dei giovani. Una ricerca sugli studenti di tre città italiane

Self-reported crime.book Page 216 Thursday, April 23, 2009 8:02 AM

Self-report questionnaire. The results obtained demonstrated that young people frequently commit acts of petty crime.

Method and response rate

Self-report questionnaire. The comparison of the research results in three cities (Siena, Genoa and Messina) shows that considering the total number of delinquent behaviours, the highest percentage is registered by Messina (72.2%), next comes Genoa (64.5%) followed by Siena (56.6%). 1,006 Self-report questionnaire. The results highlights (15 years and over) that Italy has the lowest rate of person who have drunk alcohol in the previous four week and it’s also one of the top countries for drinking “mainly /only when eating” 117 Self-report questionnaire. The results indicated (mean 11/12) that students have positive attitudes towards the victims of bullying and tend not to blame them for what has happened; moreover, male respondents blamed the victim less when the bullying took place among boys rather than among girls.

1,009 (14-21 years)

94 (14-21 years)

Sample

Self-Reported Crime and Deviance Studies in Europe

Multi-objective 2006 Investigation, ISTAT: Aspetti della vita quotidiana Coluccia: I 2006 comportamenti giovanili nelle relazioni e nel disagio

2006

IPSADItalia

Italy

Italy

Deviant behaviour

Italy

Lifestyle

Legal and illegal substance abuse

Legal and illegal substance abuse

2006

Italy

Main focus of the Geographical study coverage Bullying Rome

Date(s) of survey 2004

Researchers(s) & study title Baldry: The impact of direct and indirect bullying on the mental and physical health of Italian youngsters ESPADItalia

Self-reported crime.book Page 217 Thursday, April 23, 2009 8:02 AM

1,215 (13-19 years)

49,000 (11 year and over)

Self-report questionnaire. The research revealed that the most common deviant behaviours were: the consumption of soft drugs and spirits; regular involvement in fights; dangerous behaviour and the habitual expression of physical and verbal violence towards people close to them (peers and teachers).

Self report questionnaire. The consumption of drugs and alcohol among students increases with the age and this result is present during the whole period considered (2001-2005). Self-report questionnaire mailed. There is an increase of consumption of cannabis; about abuse of alcohol, the research underlines a diminution of consumption in general population (overall males) during the period between 2003-2005 Self-report questionnaire. The results highlights that the model of alcohol consumption in Italy is “moderate”.

Self-report questionnaire. The mental and physical health of Italian middle school students is affected by involvement in bullying and victimisation at school.

661 (11-15 years)

National secondary school students sample (2001-2006) National sample aged between 15-64 years (2001, 2003, 2005)

Method and response rate

Sample

Self-Report Delinquency in Italy

217

Date(s) of survey 2007

In press

Researchers(s) & study title Gatti et al.: Self-report juvenile delinquency in Italy

Melossi et al.: Minori stranieri tra conflitto normativo e devianza: la seconda generazione si confessa?

218 Deviance and juvenile migrants

Bologna

Main focus of the Geographical study coverage Deviant Italy behaviour

Self-reported crime.book Page 218 Thursday, April 23, 2009 8:02 AM

Method and response rate

Self report questionnaire. The results reported are descriptive. Deviant behaviours are more widespread than the official statistics indicate; that they are more frequent among males than females; that they increase with age, and that they are more common among youths attending vocational institutions and less common among high school students. 335 Self report questionnaire. This research did not (mean 13/14 years) identify a greater inclination to commit crimes among foreigners.

7,278 (12-21 years)

Sample

Self-Reported Crime and Deviance Studies in Europe

Self-reported crime.book Page 219 Thursday, April 23, 2009 8:02 AM

Self-Report Delinquency in Italy

Summary of Self-Report Delinquency Surveys conducted in Italy

General SRD Alcohol and drugs Alcohol and drugs

Level International European

Sample National

ISRD

ISRD

Bullying Timeline 1955-59 1960-64 1965-69 1970-1974 1975-1979 1980-1985 1985-1989 1990-1994 ISRD1 1995-1999 2000-2003 2004-2007 ISRD2

ESPAD

Regional/ Local ISRD

ESPADITALIA IPSADITALIA ISTAT (ITA) ITA EUROBAROME TER (ITA) ITA

EUROBAROMETER ESPADITALIA IPSADITALIA ISRD 2 (ITA)

ITA ITA ISRD 1 (ITA) ITA ITA ISRD2

219

Self-reported crime.book Page 220 Thursday, April 23, 2009 8:02 AM

Self-reported crime.book Page 221 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden Lina Andersson

I.

Overview of surveys of self-reported delinquency in Sweden

Self-reported delinquency surveys signify that people are asked about their participation in criminal acts. This is usually done through selfadministered questionnaires or structured interviews (telephone or face to face) (Coleman, Moynihan, 1996). The principle is the same as in all surveys; if you want to know something you ask (Hindelang et al., 1981). But there is an important difference between surveys in general and selfreported delinquency surveys in specific. In self-reported delinquency surveys people are asked to answer questions about acts that are punishable offences, this might increase the risk for bias. The difference between surveys in general and self-reported delinquency surveys denotes that methodological question should be treated extra carefully regarding self-reported delinquency (Junger-Tas, Marshall, 1999). This text gives a description of the situation concerning self-reported delinquency surveys in Sweden. Methodological issues will be handled, and also closely related fields such as financial funding and practical as well as scientific use of the surveys. The method of self-reported delinquency as a way to measure criminality started as a complement to the data accessible from police and court records, because of all the problems associated with the official data. An inventory of surveys of self-reported delinquency carried out in Sweden shows that the self-report method in many ways still is just a complement to the official data. A close search has been done using databases, references from relevant literature and information from experienced researchers in the field and so on, but only around 20 studies with a satisfactory degree of scientific quality has been found. Compared with research based on police and court records is this a sparse result. All identified surveys with delinquency index have been included in the overview. In general surveys with questions only about drugs and without other offenses are not in the overview. The national drug survey (The Swedish Council for Information on Alcohol and Other Drugs, CAN, 2005) is an exception and is included even though it doesn´t contain any

221

Self-reported crime.book Page 222 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

questions about criminality, because of the important role it´s been playing in the history of Swedish self-reported delinquency surveys. It has for example been a source of experience and knowledge concerning methodology and research techniques.

Figure 1. Timeline for Swedish self-reported delinquency surveys

The first Swedish self-reported delinquency survey was conducted in the end of the 1950-ies and published in 1960 (Nyqvist, Strahl, 1960) and the last one conducted in 2006 (ISRD-2) and not yet published. The overview also shows that there has been two waves of Swedish self-report studies, the first one in the 60-ies and the beginning of the 70-ies. In the 1990-ies there was a second wave that has continued since then. Between 1974 and 1990 there was a gap of 17 years without any known self-report studies (except the national drug survey, the CAN study). One explanation to the limited extent of Swedish self-report studies is the generally high trust in the official police and court statistics and a belief that the statistic is of high quality in Sweden and sufficiently valid as a measure of the levels and structure of criminality. This would mean that the need for self-reported delinquency as an alternative measure has not been regarded as high (Lindström, 1990). Further possible explanations given by Kivivuori (2007) are among others technical difficulties with quantitative research in general and self-reported delinquency surveys

222

Self-reported crime.book Page 223 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

specifically, and an anti-positivist time spirit which didn´t work together with the empirically orientated self-report surveys.

1.

Three themes

A simplified analysis of the identified studies shows three different themes or types of purposes of the studies. Investigating the extent and structure of juvenile delinquency seems to be the head purpose in the biggest part of the studies. This is the case in both national, regional and also local surveys. Another purpose seems to be the question of finding possible explanations of delinquency, or at least factors that correlate with criminal and deviant behaviors. This purpose seems to be more common in the national studies or at least in large research projects than in local and regional studies. The third type of research question found in the Swedish self-report studies is focusing on special forms of deviant behavior such as violence or drugs, but these examples are just a few. Many of the Swedish studies have used the same questionnaire for their data collection, also the sampling procedures and other research design aspects are similar. Therefore the studies will mostly not be presented individually; instead the overview will focus on some issues that will give a general picture of the method of self-reported delinquency surveys in Sweden.

2.

Territorial scope

There are two national self-report surveys with national representative samples in Sweden. Both of these partially have the intention of measuring the development of criminality or other deviant behavior over time. So they are both done repeatedly with the same sampling technique and questionnaires. One is the CAN school survey, investigating juvenile’s consumption of alcohol, narcotics and tobacco and has been conducted every year since 1971 (CAN 2005). The CAN school survey does not include any questions about criminality except for use of narcotics, which is illegal in Sweden. So the other national representative self-report survey is the only one measuring general delinquency valid in a national perspective. This study started in 1995, the results was first published as a doctoral thesis in 1999 (Ring, 1999), and has been conducted every second year since then (National Council for Crime Prevention, BRÅ, 2006). More common is the local or regional self-report surveys. These surveys are sometimes conducted by local or regional authorities and sometimes by 223

Self-reported crime.book Page 224 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

outside researchers who might have a special interest in a city or an area or just using the region to enclose their population. Some of the surveys made by the regional official authorities include only a small number of questions about delinquency in surveys where the general intention is to measure health, living conditions and ways of living by the youths in the region. The region in these cases consists of the biggest administrative unit after the national level (21 in the whole of Sweden). Regions also seem to be able to co-operate in these matters. In the search for self-report surveys four regions were found where the same questionnaire is used and comparisons between the regions are possible (The county of Örebro, the county of Uppsala, the county of Södermanland and the county of Västmanland). At this moment there are not any comparative studies published but it will probably happen in the future. In the local studies the research populations consists of youths in the municipalities, which is the smallest administrative unit in Sweden. Contact was taken with all the municipalities of Sweden, but local selfreport studies have only been found in the three biggest cities, (Stockholm, Göteborg and Malmö), and one other large city (Umeå). Within the frame of the local survey in the capital of Sweden, Stockholm, a few suburb municipalities, not belonging to the municipality of Stockholm, have been included in the research. In all these four different local studies the surveys are repeated regularly, and partly aim to measure the development over time. But there are differences between them when it comes to the time when the study started. In Stockholm the first sampling was drawn in 1972, but by then the questionnaire only contained questions about drugs and it was not until year 2000 that questions about other criminal behaviour were included. In the other three cities the self-report surveys started much later. When it comes to local studies done by outside researchers there are principally four studies that should be mentioned. The first one is an extensive research project named The Örebro project, where the selfreport survey is only one part of the study together with register studies and qualitative research. A first pilot was done in 1965 and published in a doctoral thesis 1971 (Olofsson, 1971). In 1974 there was a second selfreport survey published within the same project with the same purpose, but with a slightly different questionnaire (Dunér, Haglund, 1974). A replication of the study from 1974 was published in 1998 ; the replication was not a part of the extensive research project, though the aim was to use the earlier study to find a measure of the development of crime levels over time (Ward, 1998). All these three studies took place in the middle sized town Örebro, but the researchers were connected to Stockholm University. The second local self-reported delinquency survey conducted 224

Self-reported crime.book Page 225 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

by outside researchers is also connected with a large research project including more parts then the self-reported delinquency survey, The Stockholm Project. The purpose of the project is to investigate crime in an urban environment, with a special interest in Stockholm. The researchers are not mainly connected to the local authorities in the municipality of Stockholm; instead they are working at the National Council for Crime Prevention and the Department of Criminology at Stockholm University. Results from the survey are presented in different publications (Wikström, 1990 ; Dolmén, Lindström, 1991 ; Wikström, 1991 ; Martens, 1992 ; Lindström, 1993), but in this text only the report where the selfreported delinquency survey is the main focus is being presented (Dolmén, Lindström, 1991). A self-report study about youth´s participation in violence is a third example of a local study by outside researchers. Here the choice of city for the investigation is motivated by current problems with youth violence in this special area (Stattin et al., 1995). The Swedish participation in The Second International Self-report Delinquency Study (ISRD-2) can also be labeled as a local survey done by outside researchers. The main part of the sample consists of youths from schools in Stockholm, but there are also two smaller samples from one middle sized town and one rural area. Stockholm was selected because one of the purposes was to compare the results between the Scandinavian capitals. The areas for the two smaller samples were randomly selected and should be seen as a way to narrow the study. It is hard to define if the division of local surveys into the two categories, local or outside researcher, has a substantial meaning. A reasonable assumption might be that the objects of the research are different, where surveys conducted by local researchers aim at develop local practical preventive strategies, while surveys conducted by outside researchers have more theoretical and scientific purposes. This pattern is seen in the studied Swedish self-report survey, but a closer analysis would be desired.

3.

Contents

As might have been noticed above some of the Swedish self-report studies have special focuses. These focuses are concerning drugs, violence or general health issues. Drugs are the most common direction. Among others is the national drug school survey (CAN 2005), but also most of the local surveys done by the local authorities had drugs as their main interest before they added questions about other deviant acts such as criminality (the Stockholm, Malmö, Göteborg surveys). These surveys include questions

225

Self-reported crime.book Page 226 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

about alcohol, narcotics, tobacco, sniffing and sometimes gambling. Violence is the focus of two self-report surveys, one of them is mentioned above and investigates general violence (Stattin, 1995) and the other one concentrates on the issue of youths carrying illegal weapons to school or outside home during their free time (Elofsson, 1995). Health issues are the special focus in some of the regional surveys. So in these surveys delinquency is regarded as a health issue. The rest of the Swedish self-report studies have their interests in delinquency in general. Which kinds of questions are included in these studies does tend to differ though. Questions about trivial acts, such as stealing fruits from gardens, going with a train or bus without paying, running out from a café without paying and so on, are more common in the earlier studies during the 60-ies and 70-ies. At least in the later surveys these questions are not included in the general delinquency index. When it comes to violence the difference between the early and the later studies is the other way around. Questions about youth´s possible participation in robbery and assault are not included before the survey in 1974 (Dunér, Haglund). Later studies also include questions on violent crimes such as group fight and carrying weapon. Except for trivial crimes and different kinds of violence other common fields of questions are theft, vandalism and fraud. Sexual crimes and computer related crimes are rarely included in the questionnaires.

226

227 X

X X X X X

X X X X X X

Violence X X X X X X X X X

Theft X X X X X X X X X X X X X X

X

X X X X X

Vandalism X X X X X X X X X X X X X

-

X X X X

Fraud X X X X X X X X X X

-

-

Sexual X X X X

Table 1: Contents of the Swedish Self-reported delinquency surveys

Nyquist, Strahl (1960) Forsman, Gentz (1962) Elmhorn (1969) Olofsson (1971) CAN (1971-...) Werner (1972) Dunér, Haglund (1974) Dolmén, Lindström (1991) Rolfson (1994) Stattin (1995) Elofsson (1995) ADAD Ward (1998) Ring (1999) The Stockholm Survey (2000) Sarnecki (2001) The Umeå Survey (2005) The Malmö Survey (2006) Tuvblad (2006) Life and health – youth (2007) ISRD-2

Self-reported crime.book Page 227 Thursday, April 23, 2009 8:02 AM

-

X X X

Traffic X X X X X X X X X -

X

X X X X X

X

X

Narcotics Computer X X X X X X X X X X -

X

X X X X

Bullying X X X X

X

X X X X X

Victimisation X X X X X X

Self-Reported Delinquency in Sweden

Self-reported crime.book Page 228 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Even though the Swedish self-reported delinquency surveys differ in some parts concerning which fields they include in their questionnaires, they are similar in many ways as well. It seems like they have been inspired by each other, learned from each other’s mistakes and they also refer to each other's studies. When describing the selection process of constructing the questionnaires, the researchers are both referring to other earlier Swedish (and Scandinavian) surveys and typically youth crimes according to the police data. In the only national representative survey of general self-reported delinquency, The Ninth Grade Survey, the construction of the questionnaire was very much inspired by the questionnaire used in the first ISRD survey (Ring, 1999). The accordance is clear both regarding the delinquency index and the background questions. In two other surveys the exact same questionnaire was used, so the ISRD questionnaire played an impact even there (Sarnecki, 2001 ; Tuveblad, 2006) The Ninth Grade Survey has further more been an important source of inspiration for other delinquency surveys. The three big local surveys (the Stockholm, Göteborg and Malmö surveys) have for example used the delinquency index from the Ninth Grade Survey as one of the starting points in the construction of their own delinquency index. But the main focus in these surveys is drug habits, so they are in general more similar to the national drug survey (CAN). The principal picture is that the similarity in questionnaires between the different surveys is a good thing. There are many reasons why already constructed questionnaires should be utilized again: They are already tried and validated so the re-use saves time and resources. Similarities also enhance the possibilities of comparisons between different surveys. But the re-use also means a risk. When there are no new instruments constructed there are no development. New types of delinquent behavior, as for example computer related crimes, might not be included in the surveys to the same extent as would be desired. The similarities between the different Swedish self-reported delinquency surveys, with both the advantages and disadvantages followed, are also shown in the section below.

4.

Research techniques

The research populations in the found self-report studies of delinquency are young people living in Sweden. In the two national surveys the population is all the youths in the defined year group and in the regional and local studies the populations are the youths in these special areas. The youths investigated in these surveys are almost exclusively teenagers. The

228

Self-reported crime.book Page 229 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

two first Swedish self-report surveys do differ from the others though, since their samples consist of university students. The suggested reason for this is that these studies are more or less pilots or attempts to test the self-report method as a way of investigating criminality or try it as a data collection method. They are small studies and the samples (98 respectively 166 individuals) were constructed from practical and efficiency motives (Nyquist, Strahl, 1960 ; Forsman, Gentz). Nyquist and Strahl (1960) was also a part of a Scandinavian project were the aim was to get comparable results, which required comparable samples (Kivivuori, 2007). For the rest of the studies (there is one more exception: ADAD, which will be described more later on) the samples consist of school children from year 6 in compulsory school to year 2 in senior high school (Swedish gymnasium). Some of the studies have separate samples for different year groups, for example one sample from year 8 in compulsory school and one from year 2 in senior high school, and some do concentrate on one particular year group. For example, Ring (1999) and the following replications (BRÅ, 2006) which should be considered as one of the most important studies of self-reported delinquency in Sweden, restricted their population to ninth grade in compulsory school. This study is also usually referred to as the Ninth Grade Survey. The sizes of the studies also differ, and varies between 542 (Dunér, Haglund, 1974) and 19,665 (CAN, 2005) sampled individuals. The national representative repeated Ninth Grade study (Ring, 1999 ; BRÅ, 2006) has a sample size between 5,000 and 10,000 per year. Some of the local studies are not based on samples, because the population sizes are limited, they are instead total investigations. When it comes to response rate the overall picture looks reasonable, but the variations between different studies are rather big. The early studies have response rates at almost 100%, while the later ones have much lower frequencies. For example the Swedish part of The Second International Self-report Delinquency Study (ISRD-2), where the data was collected in 2006, has a response rate at 80%1. The same tendency of decreasing response frequencies over time is also shown when the different years of the national representative Ninth grade surveys are compared. The first time the data was collected in 1995 the response rate was 95%, but this number has decreased every time the data collection have been conducted and was in year 2005 86%. Whole school classes refusals to participate in the studies, or individual youngsters who refused participation, or were absent from school because of sickness or truancy are common explanation 1.

Comparing the number of pupils in the sampling frame with the number of usable questionnaires. When only looking at the participating classes the missing cases are less.

229

Self-reported crime.book Page 230 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

of the missing cases. The general decreasing trend is seen as mainly a sign of survey tiredness. The usual way of administering the studies of self-reported delinquency in Sweden is to use schools or school classes as sampling units and let the pupils answer the questionnaires during school time. All the studies found for this overview (again except the ADAD questions) have used self administered questionnaires to get the answers from the youths. The questionnaires have been answered at almost the same time for all the participating classes in the same school. Most of the times in classrooms with all the youths in one class sitting together or in an assembly-hall or something similar connected to the school. The perception here is that many of the studies conducted after the first data-collection of the Ninth grade Survey in 1995 (Ring, 1999) have learned from this study and the study's carefully preparations, and have because of that used the same methods for data collections. There has for example not been any personnel from the research groups present in the classrooms, instead teachers or other personnel from the school has handed out the questionnaires to the pupils and are the ones supervising the lessons. The personnel have been instructed in what kind of information they should give to the pupils and that they are not allowed to walk around in the classroom looking at the pupils questionnaires, this to keep the pupil's anonymity. Ring (1999) tested possible effects of teacher and researcher supervision by comparing results from a sample of 500 pupils, and he did not find any significant differences between the two groups. A common procedure is also that the youths puts their filled questionnaire in an individual envelope and seal it by themselves, so that there will be no risk for the teacher to read the pupils answers. Olofssons (1971) and Dunér, Haglund (1974) are the only studies of self-reported delinquency (again except the ADAD interview) were the pupils weren’t kept anonym for the researchers either. As mentioned before the self-report studies done by Olofsson and Dunér, Haglund were parts of a bigger research project where police data and other information about the participating youth were necessary, so a totally anonymous survey was not possible. The exception from the general patterns that have been mentioned several times above: The ADAD interview is not a single study of self-reported delinquency. Instead it is a collection of data from institutionalized youths, with the major aim to plan the care at the institution. But most of the youths also agree for their information to be kept in a database which can be used by researchers with permission2. This data collection is done by 2.

In 2006 709 youths of 740 interviewed agreed for their interview to be kept in the register.

230

Self-reported crime.book Page 231 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

face to face interviews and is for obvious reasons not anonymous. Sometimes the youths in the ADAD material are a bit older compared with the main part of the rest of the self-report studies; 19 to 20 years old.

5.

Validity and reliability controls

There have been some validity and reliability tests linked with the Swedish self-report surveys. Most of them are small and have the form of pilots for the main survey, but still there are some interesting results worth mentioning. In two of the earlier surveys, tests were done that probably had some impact on the method in later surveys and are relevant for the discussion of validity of the self-report methodology. Elmhorn (1969) did among other things a control of the effect of anonymity, where she found that the pupils who were not anonymous reported lower levels of criminality. These results were especially explicit concerning more serious crimes. To what extent the not anonymous pupils were informed that no one outside the research group would read their answers is not clear. Olofsson (1971) conducted a survey where the pupils weren't anonymous for the researchers, but where they were carefully informed that the identification was confidential for everybody outside the research group. Olofsson tested the validity of her results by comparing the data with police records. The test shows some problems; some of the boys that write in the questionnaire that they have been reported to the police are not in the police register and there are also boys in the register who don't report this in the questionnaire. Olofsson (1971) still draws the conclusion that the results have high criteria validity though the survey and the police register show a satisfactory degree of accordance. A newer validation study done using police records is a criteria validation control of the ADAD instrument in 2007 (Shannon, Innala, 2007). The control shows good agreement between the two data sources. Another early validation by Elmhorn (1969) is that she conducts a small study with follow up interviews of 40 boys. The boys filled in the questionnaires in the same way as was done in the main survey and were afterwards interviewed about their interpretations of the questions. Elmhorn found that most of the questions had been understood the way they were supposed to, but there were some questions which had been misunderstood by a few boys. Many of the surveys conducted after Elmhorn (1969) are referring to these follow up interviews, so hopefully the mistakes that were done in the survey from 1969 are not repeated at the same extent in the modern surveys.

231

Self-reported crime.book Page 232 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Before and during the first sweep of the national representative selfreported delinquency survey, the Ninth Grade Survey, the researcher did a few tests of reliability and validity (Ring 1999). Among other things he did observations in the classrooms during a pilot study and noted that the pupils took the task very serious. He also encouraged the pupils to ask questions if there was something in the questionnaire they didn´t understand or that they regarded as ambiguous, but the pupils only asked a few questions. This pilot study also tested four different types of questionnaires before the final questionnaire was constructed. As mentioned above a further test was the comparison between teacher and researcher supervised groups where no significant differences were found. Several studies have done modified test-retest controls, where they included two batteries of questions supposed to measure the same thing, one battery in the beginning of the questionnaire and one at the end. Comparisons show reasonable accordance (Ring, 1999; Ward, 1998 ; Lindström, Dolmén, 1991). There are also examples where the validity studies are not only pilots connected to the main surveys. In two cases data from the ADAD instrument is being used, of which one is mentioned above (Shannon, Innala, 2007). The other one compares the ADAD data from institutionalized youths with results from the National representative Ninth grade survey. The purpose with the study is to meet and go further with one of the most common critique against self-report surveys: that surveys based on school or community samples only measure limited delinquency by relatively conformist youths, as they exclude institutionalized youths. One of the conclusions in the study is, that even though the two samples differs in some areas, the difference is not big enough and the number of institutionalized youths is so restricted that their inclusion or exclusion from surveys based on school or community samples doesn’t have a substantial impact on the results (Shannon, 2006).

6.

Funding agencies

The general situation with funding is that the institute or organization conducting the study also is the funder. The local studies performed by researchers connected to the local authorities are also funded by the local authorities. Sometimes special research institutions are the one performing the data collection and/or the analyses for the authorities, but the local authorities are in these situations still the funders. For the regional selfreport studies, the situation is the same. The regional authorities are most of the times both doing the research and paying for it. For these cases the

232

Self-reported crime.book Page 233 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

surveys are connected to the health and medical services, which also are the divisions of the regions that are funding the research. The two national representative studies of self-reported delinquency in Sweden are, in the same way as the local and regional studies are linked to the local and regional authorities, linked to the state authorities. The connection is more indirect in these cases though. The CAN study is performed and funded by the CAN. CAN is an independent organization but receives most of their financial support from the Swedish government. The national representative Ninth grade survey (BRÅ, 2006) is today carried out by the Swedish National Council for Crime Prevention (BRÅ). BRÅ´s function is as the government’s body of expertise within the judicial system. So the Ninth Grade Survey of self-reported delinquency is also indirectly funded by the government. The two first data-collections for this study were however done at the Department of Criminology, Stockholm University, and these were financed by two different research funds, more or less independent from the government (Ring, 1999). When it comes to the studies performed by outside researchers on a local sample/population the funding is the following: The Örebro project (Olofsson, 1971; Dunér, Haglund, 1974) was financed by The Swedish National Agency for Education and carried out by researchers from the department of Psychology, Stockholm University. The later replication (Ward, 1998) was done with economic support from The Scandinavian Research Council for Criminology. Also the data-collection for the Swedish part of the Second International Self-report Delinquency Study (ISRD-2) was funded by The Scandinavian Research Council for Criminology, but the rest of the work with the data was with support from the Department of Criminology, Stockholm University. The survey within the Stockholm project (Dolmén, Lindström 1991) was conducted at the National Council for Crime Prevention (BRÅ) and in co-operation with the Department of Criminology, Stockholm University, and was also financed by these institutes.

II.

The use of the surveys

This section is mainly based on short telephone interviews with some key persons involved in the different studies of self-reported delinquency and political official servants. So the information presented is in most of the cases not available in any written documents and it cannot be assured that the key persons are representative for all the possible perspectives on these issues.

233

Self-reported crime.book Page 234 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

1.

Practical use

Many of the examined surveys have as one of the purposes for the research that the results, and the knowledge the results gives, will work as a foundation for progress of the work and activities in their field and to develop crime preventive strategies. But it is hard to define in what way and in what extent this ambition has been fulfilled. From interviews with researchers and political official servants on local, regional and national level, it seems like the implementation of research results from studies on self-reported delinquency is most easily done at the local level. Many of the surveys have questions that measure possible risk and protection factors for delinquent behavior, and have also found results that might be used in the progress of preventive strategies. Most of these factors belong to fields that are connected with work at the local level and activities that are positioned under the responsibility of local authorities. School environment, parental support and drug problems are examples of this. Interviews with local researchers show good examples of situations where the research results have been used in the local policy making and also in the concrete preventive work. When published reports include propositions for preventive strategies it seems like the policymakers and local institutions can use this to improve their activities. How large the possibility is for this implementation depends in a great way on relationships and opportunity for co-operation between different parts of the community. In one of the municipalities where a survey was performed, and the researchers came up with suggestions of how the knowledge could be used, schools wrote new action plans on how they should prevent delinquency among their pupils. Parental guidance programs based in schools were started and the co-operation and communication between the social services and the schools was improved. Also the pedagogic activities around narcotics, alcohol and tobacco were enhanced. In another municipality where the researchers suggestions were more situational preventive orientated, one of the school buildings are now under reconstruction. This will get a new physical school environment, it will be less anonymous and opportunities for adult control over the youths will be greater. The researchers meant in their report, that these changes might decrease violence and bullying in the school. Of course also the national surveys could influence preventive activities. Even though these activities are mainly linked to the local authorities, the national surveys can work as sources of knowledge and inspiration for the local policy makers. To which extent this is done is hard to find out, the researchers at the national level cannot possibly know which local policymakers read their report and were inspired by it. Instead the practical 234

Self-reported crime.book Page 235 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

use of national studies of self-reported delinquency discussed here is how they are used at the national level of policy making. The general picture is that the governmental official servants find the surveys important and of high quality. Self-report surveys are a necessary alternative to the official statistics and a good measure of size and structure of delinquency and also a good opportunity to find out about risk and protection factors. The government can use the results, and knowledge got from the research, at a structural level. Especially the fact that both the national self-report studies are repeated surveys that show development over time seems to be appreciated. The time series are used to evaluate national investments and to understand structural trends. Results from the studies are also used as arguments in the national political debate. A frequently mentioned problem in discussions with people working at public authorities in Sweden is the division of responsibilities between different levels of government and the difficulties with co-operation between different levels. For example health issues falls under the regional level, while social problems such as addiction problems, poverty and school problems are issues concerning the municipality. The division of responsibilities seems to be relevant when it comes to the use of self-report surveys as well. Many of the results found in the regionally conducted surveys concerns issues that belong to the local level and can therefore not be practically handled by regional authorities. The ambition is to cooperate between the levels of authority, but most of the times this only works in policy making and not for the concrete activities. Of course there are also examples where these impediments have been mastered.

2.

Scientific use

A clear picture of the scientific development appears when looking at a great part of the performed Swedish surveys of self-reported delinquency. In early studies it is easy to observe a lot of methodological problems, which are more or less disappearing when the Swedish researchers are getting more experience of the self-report method. In that perspective most of the self-reported delinquency surveys have had scientific importance; they have been a part of an important development of an essential method in the field of Criminology. Some of the studies have also been of scientific use when it comes to theory testing, finding possible explanations to juvenile delinquency and broaden the picture of criminal behavior among youths. Worth mentioning are among others the Örebro project (Olofsson, 1071 ; Duner, Haglund, 1974), Dolmén, Lindström (1991) and Ring (1999). The two national representative surveys have also been important

235

Self-reported crime.book Page 236 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

because they have offered a high quality picture of criminal prevalence and drug use over time (BRÅ, 2006; CAN, 2005).

3.

Media's reception

The relationship between criminality and media, and between criminological research and media is a complex issue, and an own field of research. There are for example research that shows a negative correlation between trends of space given to criminality in the press and crime trends shown from research (Estrada, 1999). The discussion below is just a short summary of the picture of media's reception given by the interviewed key persons. Interviews with the researchers show an ambivalent relationship to the media. The local studies seem to get a lot of attention from local media, newspapers, radio and TV, which the researchers in a general matter appreciate. But at the same time there are examples when media focus on parts of the studies that the researchers wouldn't choose, and the picture of delinquency that is being shown to the public is not the same as the picture the researchers focus. An example might be that media are writing about the great degree of drug use prevalence found in a study, while the researchers rather would spread results, shown in the same study, concerning low levels of school violence or suggestions for preventive strategies.

III. Main findings In this section a small selection of findings from the Swedish self-reported delinquency surveys will be presented. Results from different studies are not comparable in a proper way because they differ concerning samples, questionnaires and techniques (see appendix). Figure 2 shows prevalence from the first wave of Swedish self-report surveys. The contents of the questionnaires from the first wave are more alike, so these surveys are more comparable with each other than to the surveys from the second wave starting in the 1990's. The data presented in Figure 2 concern all the delinquency questions included in the questionnaires. The prevalence levels vary between 81% (Werner, 1972) and 97 % (Nyquist, Strahl, 1960). Even if the delinquency indexes in the first wave are more similar to each other than to the surveys in the second wave, a significant part of the diverse prevalence levels could probably be explained by what kind of items are included in the questionnaire. For example the questionnaire of 236

Self-reported crime.book Page 237 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

Nyquist and Strahl (1960) include items such as stealing fruit from gardens and stealing money from home, which is not the case in Werner (1972).

Figure 2: Lifetime prevalence’s, all the delinquency items

Findings from the national representative Ninth grade survey show decreasing crime trends from 1995 to 2005. The decreasing trends are most distinctive regarding the prevalence of theft, from 66% in 1995 to 52% in 2005, and vandalism, from 46% in 1995 to 31% in 2005. For violence and narcotic related behavior the prevalence are more stable over time and the levels are also lower then for theft and vandalism (BRÅ 2006).

Source: BRÅ (2006)

Figure 3: Prevalence last 12 months, ninth graders in Sweden (%)

237

Self-reported crime.book Page 238 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Source: BRÅ (2006)

Figure 4: Crime trends, prevalence, last 12 months, ninth graders in Sweden (%)

IV. Summary and conclusions In his text different issues related to the methodology of self-reported delinquency surveys in Sweden have been described; general research questions as well as more detailed research techniques. An overview of conducted self-report delinquency surveys in Sweden show a long history, with the first survey carried out already during the end of 1950´s. But the Swedish history of self-report is irregular whereas there is a gap of 17 years between 1974 and 1992. Three types of main purposes are shown in the studies; to investigate the general prevalence, structure and trends of criminal behavior; to test criminological theories; and to investigate special fields of delinquent behavior such as drugs or violence. Comparisons between the different surveys show some differences and a development over time, both concerning research techniques and contents in the questionnaires. The early self-reported delinquency surveys include more questions about trivial acts, such as stealing fruits from gardens, going with a train or bus without paying, running out from a café without paying etc., than the later surveys do. The later surveys include questions about violence, which is not done in the early studies. Although the surveys are very similar in some parts, it is noticeable that the researchers have been inspired from each other and have learned from each other’s 238

Self-reported crime.book Page 239 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

experiences. Some of the questionnaires of the later surveys are the same, and when the questionnaires are not identical the delinquency indexes are very much alike. The advantages of this are significant because of the saved resources and possibilities for comparison. At the same time, changes in the questionnaire are necessary, new behaviour need to be included, a development that the re-use of already constructed questionnaires in new surveys might restrict. Criminal acts related to the computer and internet techniques are examples of behaviour rarely included in the questionnaires used in the surveys, but often discussed as common among youths. In the beginning of this text the self-report method in Sweden was classified as a complement to police and court data. If the ambition with the method is not just to be a complement, but an important source of information used in policy making and the development of preventive strategies, it is probably necessary to meet these challenges and improve the questionnaires with new behaviour. Steps towards these improvements are also made. In preparation for the next wave of data collection of the national representative Ninth Grade Survey in the autumn of 2008 the questionnaire has been reviewed and updated.

References BRÅ, 2006, Ungdomar och brott åren 1995-2005. Brå-rapport 2006, 7, Stockholm, Brottsförebyggande rådet. CAN, 2005, Skolelevers drogvanor 2005. CAN rapport #90, Stockholm, CAN. COLEMAN C., MOYNIHAN J., 1996, Understanding crime data. Haunted by the dark figure, Maidenhead, Open University Press. DOLMÉN L., LINDSTRÖM P., 1991, Skola, livsstil och brott: ungdomar i åtta Stockholmsskolor, Stockholm, Brottsförebyggande rådet. DUNÉR A., HAGLUND B., 1974, Tonårspojkar och brott: kunskaper, attityder och faktiskt beteende, Stockholm, Utbildningsförl för Skolöverstyrelsen. ELMHORN K., 1969, Enkätundersökning av faktisk brottslighet bland skolbarn, SOU 1969, 1, Stockholm. ELOFSSON S., 1995, Beväpnad?: en studie av högstadie- och gymnasieelever, Stockholm, Fritid Stockholm. ESTRADA F., 1999, Ungdomsbrottslighet som samhällsproblem. Utveckling, uppmärksamhet och reaktion, Avhandlingsserie nr. 3, Kriminologiska institutionen, Stockholms universitet. FORSMAN H., GENTZ C.-F., 1962, Kriminalitetsförekomsten hos presumtivt ostraffade. En enkätundersökning, Nordisk Tidsskrift for Kriminalvidenskab, 50, 318-324.

239

Self-reported crime.book Page 240 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

HINDELANG M., HIRSCHI T., WEIS J., 1981, Measuring Delinquency, Beverly Hills, CA, Sage. JUNGER-TAS J., MARSHALL I., 1999, The Self-Report Methodology in Crime Research, in TONRY M., Ed., Crime and Justice. A Review of Research, Chicago, University of Chicago Press, 291-367. KIVIVUORI J., 2007, Delinquent Behaviour in Nordic Capital Cities, Scandinavian Research Council for Criminology & National Research Institute of Legal Policy, Finland. LINDSTRÖM P., 1993, School and delinquency in a contextual perspective, Stockholm, Brottsförebyggande rådet. MARTENS P., 1992, Familj, uppväxt och brott, Stockholm, Brottsförebyggande rådet. NYQUIST O., Strahl I., 1960, II. Svensk undersökning, Nordisk tidsskrift for Kriminalvidenskab, 48, 113-117. OLOFSSON B., 1971, Vad var det vi sa! Om kriminellt och konformt beteende bland skolpojkar, Stockholm, Utbildningsförlaget. RING J., 1999, Hem och skola, kamrater och brott. Akademisk avhandling, Stockholm, Kriminologiska institutionen Stockholms universitet. ROLFSON M., 1994, Unga på drift. Om sociala normer och kontroll i Rosengård, Akademisk avhandling, Lund, Lunds universitet. SARNECKI J., 2001, Undersökning av självdeklarerad brottslighet och annat problembeteende. Tyresö våren 1998, 1999 och 2000, Stockholm, Kriminologiska institutionen, Stockholms Universitet. SHANNON D., 2006, Chronic Offenders or Socially Disadvantaged Youth? Institutionalized Males as Missing Cases in School-based Delinquency Research, Journal of Scandinavian studies in Criminology and Crime Prevention, 7, 1, 78-100. SHANNON D., INNALA C., 2007, En kriterivalidering av brottsfrågorna i svenska ADAD-instrumentet. Allmän SiS-rapport 2007:2, Stockholm, Statens institutionestyrelse. STATTIN H., 1995, Våld bland 15-åringar I en svensk stad. En enkätunderökning i maj 1995. In Forskningsrådnämnden, Det obegripliga våldet. Forskare om ungdomar och våld, Stockholm, Forskningsrådsnämnden. TUVEBLAD C., 2006, Genetic and enviormental influences on antisocial behavior from childhood to emerging adulthood. Akademisk avhandling, Stockholm, Karolinska intitutet. WARD M., 1998, Barn och Brott av vår tid? Självdeklarerad Ungdomsbrottslighet 1971 och 1996. En jämförelse utifrån Örebroprojektets data, Stockholm, Kriminologiska institutionen Stockholms universitet.

240

Self-reported crime.book Page 241 Thursday, April 23, 2009 8:02 AM

Self-Reported Delinquency in Sweden

WERNER B., 1972, Den faktiska brottsligheten, Nordisk tidsskrift for Kriminalvitenskab, 59, 105-141. WIKSTRÖM P.-O., 1990, Crime and Measures against Crime in the City, Stockholm, Brottsförebyggande rådet. WIKSTRÖM P.-O., 1991, Sociala problem, brott och trygghet, Stockholm, Brottsförebyggande rådet.

241

242

1959 Elmhorn (1969) Survey of actual criminality among school children (SOU 1969:1) Olofsson (1971) 1968 On delinquency and conformity among school boys

1962 Forsman, Gentz (1962) Criminality among unpunished

Researchers and Dates of survey study title (Author’s translation of title) Nyquist, Strahl 1959 (1960) Criminality among unpunished, II, Swedish survey 98 university students (males and females)

Sample

Method and response rate

Cross-sectional survey, one sweep. Anonymous self administered paper and pencil questionnaire during a lecture at the university. 100% response rate. 164 university Cross-sectional survey, one sweep. General self-reported Local: students (males Anonymous self administered paper and delinquency survey. University of Validate the results Gothenburg and females) pencil questionnaire during a lecture at the university. from Nyquist & 98.8% response rate. Strahl (1960) General self-reported Local: Stockholm 1,800 school Cross-sectional survey, one sweep. delinquency survey. children (males Anonymous self administered paper and pencil questionnaire in class-rooms during and females, aged 9 up to 14) school time. Researcher supervised, self-sealed individual envelope. 95.5% response rate. General self-reported Local: Middle size 549 school The survey was a small part of a large delinquency survey town Örebro children (only longitudinal study with different kinds of data. Identify risk factors. males, age 15) The Örebro project. Not anonymous for the researchers (code system), self administered paper and pencil questionnaire, in class rooms during school time, supervised by outside researchers. Self sealed individual envelope. Absentees were followed up at an extra occasion.

Geographical coverage

General self-reported Local: delinquency survey. University of Uppsala

Main focus of the study

Appendix 1: Summary of self-reported delinquency surveys in Sweden

Self-reported crime.book Page 242 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Dunér, Haglund (1974) Teenage boys and delinquency. Knowledge, attitudes and behavior

1971

Werner (1972) The 1963 / 1964 actual criminality

Researchers and Dates of survey study title (Author’s translation of title) The CAN School Every year survey since 1971 National

Trends in drug use among youths.

Sample

General self-reported Local: Middle size 542 school boys delinquency survey. town Örebro Identify risk factors.

Varies between (1971-2005) 2,937 and 19,667 (males and females, age 12, 15 and 17) General self-reported Local: Stockholm 12,300 draftees delinquency survey. and Malmö. (only males, age 18-26)

Geographical coverage

Main focus of the study

Self-reported crime.book Page 243 Thursday, April 23, 2009 8:02 AM

An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. Response rates vary between 84 and 90%. Anonymous survey in connection with the enlistment for compulsory military service. Self administered questionnaires supervised by outside researchers. 95,7% and 98.4% response rates in the two different regions. The survey was a small part of a large longitudinal study with different kinds of data: The Örebro project. Not anonymous for the researchers (code system), self administered paper and pencil questionnaire, in class rooms during school time, supervised by outside researchers. Dropped out pupils were looked up and asked to participate in the survey. Self sealed individual envelope. Absentees were followed up at one extra occasion.

Method and response rate

Self-Reported Delinquency in Sweden

243

244

Rolfson (1994) 1990 Youth on drift: social norms and social control in Rosengård 1995 Stattin (1995) Violence among 15-year olds in a Swedish town Elofsson (1995) 1995 Armed? A survey of school youths

Researchers and Dates of survey study title (Author’s translation of title) Dolmén, 1990 Lindström (1991) School, life-style and delinquency

Geographical coverage

Local: suburb in the municipality of Malmö

Method and response rate

Cross-sectional design, one sweep. Anonymous self administered paper and pencil questionnaire in class-rooms during school time. Cross-sectional design, one sweep. Anonymous self administered paper and pencil questionnaire in class-rooms during school time. Response rates vary around 90%.

The survey was one part of a large research project: The Stockholm project. It was a crosssectional design, one sweep. Anonymous self administered paper and pencil questionnaire in class-rooms during school time. Researcher supervised, self-sealed individual envelope. 95% response rate. 456 school Cross-sectional design, one sweep. children (males Anonymous self administered paper and and females, age pencil questionnaire in class-rooms during 13 up to 16) school time. Teacher supervised.

792 school children (males and females, age 15)

Sample

Local: Middle size 876 school town Sundsvall children (males and females, age 15) 3,325 school Local: Youth carrying children (males weapon to school Municipality of and females, age and in their free time Eskilstuna and 13 up to 19) municipality of (special part of survey concerning Vallentuna life habits among youths)

Degree and circumstances of youth violence

Social norms violations among youths

General self-reported Local: Stockholm delinquency survey. Compare different parts of Stockholm.

Main focus of the study

Self-reported crime.book Page 244 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Geographical coverage

Sample

711-848 School children (males and females 1416)

5,600-6,000 School children (males and females, age 1516)

General self-reported Local: the middle 1,650 school delinquency survey. size town Örebro children (males Replicate Dunér & age 15) Haglund (1974).

Main focus of the study

General National delinquency. Linkage between delinquency and youth’s social situation. Sarnecki (2001) 1998, 1999 and General delinquency Local: The 2000 municipality of Study of Tyresö self-reported delinquency in (Stockholm suburb) Tyresö spring 1998, 1999 och 2000

Researchers and Dates of survey study title (Author’s translation of title) Ward (1998) 1996 Children and crime of our time? Self-reported delinquency 1971 and 1996 Ring (1999) Home 1995 and 1997 and school, peers and crime

Self-reported crime.book Page 245 Thursday, April 23, 2009 8:02 AM

Replication of Dunér & Haglund (1994), but not the same project. An anonymous crosssectional survey, with self administered questionnaires in class-rooms during school time. Supervised by outside researchers. No follow up for absentees. 98% response rate. An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. Response rates vary between 95% and 97%. An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Response rates vary between 86% and 89%.

Method and response rate

Self-Reported Delinquency in Sweden

245

246

Life habits survey, 2004 (The first Umeå municipality sweep of the survey was in (2005) 1996, but 2004 was the first year with questions about criminality in the questionnaire) The Malmö survey 2006 (2006)

Researchers and Dates of survey study title (Author’s translation of title) The Stockholm Every year survey between 1972 and 2006. Questions about criminality are included from year 2000.

7,378 school children (males and females, age 12, 15 and 17)

Local: The municipality of Umeå

Local: The municipality of Malmö

Life habits among youths

Life and drug habits among youths

Sample

Varies between 8,400 and 13,000 (the years when questions about criminality have been included) (males and females, age 1516 and 17-18) 7,249 school children (males and females, age 13 up to 19)

Geographical coverage

General self-reported Local: Stockholm delinquency survey with special focus on drugs

Main focus of the study

Self-reported crime.book Page 246 Thursday, April 23, 2009 8:02 AM

An anonymous repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. Response rates vary around 80%.

Anonymous repeated survey cross-sectional survey. Self administered computer based questionnaires in school during school time. Teacher supervised. Response rates vary between 78% and 84%

An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Response rates vary between 72% and 93%.

Method and response rate

Self-Reported Crime and Deviance Studies in Europe

1999 and 2002 Tuvblad (2006) Genetic and environmental influences on antisocial from childhood to emerging adulthood Life and health – 2005 and 2007 young

Researchers and Dates of survey study title (Author’s translation of title) The Gothenburg 2004 survey (2004) Local: The municipality of Gothenburg

General self-report delinquency survey

Life situation and life Regional: Four habits among different regions youths. (The county of Örebro, the county of Uppsala, the county of Södemanland and the county of Västmanland).

General delinquency National (not representative) (only applies the small SRD part of a large project)

Geographical coverage

Main focus of the study

Self-reported crime.book Page 247 Thursday, April 23, 2009 8:02 AM

Method and response rate

Around 10,000 school children in every region (males and females, age 13, 15 and 17)

An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. The response rates vary between 78% and 84%.

An anonymous repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. 81% response rate. 1,480 twin pairs The survey was a small part of a large (males and longitudinal study with different kinds of data: females, aged 13- The twin study. Not anonymous self 14 and 16-17) administered questionnaires.

10,000 school children (males and females, age 15 and 17)

Sample

Self-Reported Delinquency in Sweden

247

248

ADAD

ISRD-2 in Sweden

Researchers and study title (Author’s translation of title) BRÅ (National Council for Crime Prevention) The ninth grade survey National

Geographical coverage

2006

To get international comparable measures of delinquency

Local: One big city municipality (main sample), one middle sized town and one rural area No crossThe main purpose is National (not to gather representative) sectional data gatherings, data information for exist from year planning the care given to 1997 institutionalized youths, but is also used for different kinds of research

Every second General self-report year since 1999 delinquency survey (continuing Ring 1999) Not 2007

Dates of survey Main focus of the study

Self-reported crime.book Page 248 Thursday, April 23, 2009 8:02 AM

A majority of all institutionalised youths. In year 2006 709 out of 740 (58%) (males and females aged 15 up to 21)

2,911 school children (males and females from age 13 up to 16)

Varies between 5,000 and 10,000 (males and females, aged 15-16)

Sample

An anonymous, repeated survey with a crosssectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. Response rates vary around 85%. An anonymous survey with a cross-sectional design. Self administered questionnaires in class-rooms during school time. Supervised by teachers. Individually sealed envelopes. Opportunity for follow ups. 78% response rate. Non anonymous face to face interviews by staff at the institutions.

Method and response rate

Self-Reported Crime and Deviance Studies in Europe

Self-reported crime.book Page 249 Thursday, April 23, 2009 8:02 AM

On the Authors Marcelo F. Aebi is Professor of Criminology and Vice-director of the School of Criminal Sciences at the University of Lausanne, Switzerland. Currently, Marcelo F. Aebi is also a consultant expert of the Council of Europe and the European Commission, as well as the Executive Secretary of the European Society of Criminology. His main research topics include juvenile delinquency, comparative criminology, corrections, methodology, drugs and crime, and victimization and self-reported delinquency studies. His most recent books are: 앫 Temas de criminología, Madrid, Dykinson, 2008. 앫 Comment mesurer la délinquance, Paris, Armand Colin, 2006. He is also one of the co-authors of the European Sourcebook of Crime and Criminal Justice Statistics, The Hague, Boom Juridische Uitgevers, 2006, 3rd edition. Lina Andersson is a doctoral student at the Department of Criminology, Stockholm University. Her main field of research is Self-report methodology in Delinquency research. Her doctoral thesis will be an evaluation and validation of Self-report methodology in Sweden. Among other things, Lina participated in the Swedish fraction of the ISRD2, where she collected the data and is a co-author of both a national report with the results from the Swedish survey and the Swedish national chapter in the international publication Juvenile Delinquency in Europe and Beyond: An International Perspective on Key Issues and Causes (Springer, in print). Cécile Carra is a Lecturer in Sociology at the Institut universitaire de formation des maîtres (Teachers Training School) of the Université d'Artois and a researcher at the Centre de recherches sociologiques sur le droit et les institutions pénales. She is the head of the research team RECIFES, which specialises in the study of norms and deviance in the field of education. After her studies on juvenile delinquency in working-class neighbouhoods, her current interest is on school violence that she mainly explores with self-reported delinquency and victimisation surveys. Her recent publications are: 앫 Violences à l'école élémentaire. Ce que révèlent les déclarations des enseignants et de leurs élèves, Paris, PUF, coll. Education et Société, 2009. 앫 Ecole et violences, Paris, La documentation française, coll. Problèmes politiques et sociaux, 2002 et 2006, (avec D. Faggianelli). 앫 Délinquance juvénile et quartiers « sensibles ». Histoires de vie, Paris, l’Harmattan, Coll. Déviance et Société, 2001. 249

Self-reported crime.book Page 250 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Giada Anna Maria Cartocchi has received her Degree in Psychology, with major in Child Psychology, at the University of Florence. She postgraduated in Systemic Psychotherapy and is involved in teaching and research within the chair of Forensic Psychiatry of the University of Siena. She authored some articles in the field of Criminology and Forensic Psychiatry. Thomas Görgen is Professor at German Police University in Muenster and head of the Department of Criminology and Crime Prevention. Among his main research interests are topics related to victimization, old age and crime, domestic violence, crime prevention, juvenile crime, homicide, sexual offences, attitudes related to crime and crime control. Some recent publications: 앫 Interdisziplinäre Kriminologie: Festschrift für Arthur Kreuzer zum 70. Geburtstag, Frankfurt a.M., Verlag für Polizeiwissenschaft, 2008 (Editor with K. Hoffmann-Holland, H. Schneider, J. Stock). 앫 Misshandlung und Vernachlässigung älterer Menschen in privaten Pflegebeziehungen, in A. Dessecker A., Egg R., (Hrsg.), Gewalt im privaten Raum: aktuelle Formen und Handlungsmöglichkeiten, Wiesbaden: Kriminologische Zentralstelle, 2008, 125-152. 앫 Organisierte Kriminalität und Terrorismus – unvereinbare Phänomene oder gefährliche Allianzen ?, in Informationszentrum Sozialwissenschaften, (Hrsg.), Kriminalsoziologie + Rechtssoziologie, 2008, 1, 9-25 (with D. Schröder). Janne Kivivuori is the research director of the Criminological Unit at the National Research Institute of Legal Policy, Helsinki. His research has centred on two major topics: juvenile delinquency and homicide. His recent work has appeared in the British Journal of Criminology, European Journal of Criminology, and Journal of Scandinavian Studies in Criminology and Crime Prevention. Commissioned by the Scandinavian Research Council of Criminology, he recently authored the Nordic area report of the second sweep of the International Self-Report Delinquency Study ("Delinquent Behaviour in Nordic Capital Cities"). He has been centrally involved in the creation of the major Finnish crime monitoring systems such as the Finnish Self-Report Delinquency Study, Young Male Crime Survey, and the Finnish Homicide Monitoring System. He teaches criminology courses at the University of Helsinki as an adjunct professor of sociology. Susan McVie is a Senior Research Fellow in the School of Law at the University of Edinburgh in Scotland. She is Co-Director of the Edinburgh Study of Youth Transitions and Crime; Network Leader for CJ-QUEST,

250

Self-reported crime.book Page 251 Thursday, April 23, 2009 8:02 AM

On the Authors

which forms part of the Scottish Centre for Crime and Justice Research; and Director of the Applied Quantitative Methods Network. Susan’s work focuses on the application, implementation and improvement of quantitative methodologies within social research, particularly within the field of criminology. She has been involved in developing and managing several large-scale surveys, both cross-sectional and longitudinal, and has acted as a specialist advisor to many Government departments, private organisations and voluntary bodies. Susan’s interests include youth crime, deviance and substance use; comparative crime trends and patterns; criminal careers and life-course criminology; systems of justice, including transitions from juvenile to adult criminal justice systems; and neighbourhood effects on offending. She has published extensively in these fields and is currently in the process of co-writing a book entitled Delivering Youth Justice. Some recent publications: 앫 Criminal careers and young people in Barry M., McNeill F., Eds., Youth Offending and Youth Justice, London, Jessica Kingsley Research Highlights in Social Work Series, in press. 앫 Youth justice? The impact of system contact on patterns of desistance from offending, European Journal of Criminology, 2007, 4, 3, 315-345 (with L. McAra). 앫 Criminal Justice Transitions, Edinburgh Study of Youth Transitions and Crime, Research Digest No. 14, 2007, (with L. Mc Ara). Lieven Pauwels has a PhD in Criminology from Ghent University. He is Professor of Criminology at the Department of Criminal Law & Criminology, co-director of the Research Group Social Analysis of Security. His interests are situated in the domains of the social ecology of crime, adolescent offending, the study of measurement error in surveys, crossnational comparative research. Some of his articles have appeared in International Review of Victimology, European Journal of Criminology, Crime & Delinquency. Recent Publications: 앫 Individual Differences In Adolescent Life Style Risk By Gender And Ethnic Background: A Test in Two Urban Samples, European Journal of Criminology. 2009, 6, 1, 5-23 (with R. Svensson). 앫 Disentangling neighbourhood- and school contextual variation in serious offending: assessing the effect of ecological disadvantage, in Cools, M, De Kimpe, S., De Ruyver, B., Easton, M., Pauwels, L., Ponsaers, P., Vande Walle, G., Vander Beken, T., Vander Laenen, F., Vermeulen, G., Eds., Governance of Security Research Papers Series I,

251

Self-reported crime.book Page 252 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

Contemporary Issues in the Empirical Study of Crime, Antwerpen, Maklu, 2009. 앫 Is a risky lifestyle always “risky”? The interaction between individual propensity and lifestyle risk in adolescent offending: A test in two urban samples. Crime & Delinquency. OnlineFirst, published on September 24, 2008, DOI:10.1177/0011128708324290 (with R. Svensson). 앫 Geweld in groepsverband onder Antwerpse Jongeren: de rol van schoolcontext en leefstijl, Tijdschrift voor Criminologie, 2008, 50, 1, 316. Stefaan Pleysier is a lecturer in Research Methodology and Statistics and senior researcher at the ‘Research Centre in Security, Safety and Society’ (KATHO University College) and the ‘Leuven Institute of Criminology’ (Katholieke Universiteit Leuven). He has recently obtained his PhD degree in Criminology involving a study on the conceptualisation and measurement of ‘fear of crime’. Furthermore, he is a member of the Advisory Board and the Technical Working Group of the Belgian Safety Monitor, and has published on measurement issues and quality assessment in large surveys as the Safety Monitor. Recent publications: 앫 Crime victim and Insecurity Surveys in Belgium and the Netherlands, in Zauberman R. (Ed.). Victimisation and insecurity in Europe. A Review of Surveys and their use, Brussels, VUBPress, 2008 (with Lieven Pauwels), 39-64. 앫 Victimes de délinquance et insécurité: Les enquêtes en Belgique et aux Pays-Bas, in Zauberman R. (Ed.). Victimation et Insécurité en Europe. Un bilan des enquêtes et de leurs usages, Paris, L'Harmattan, 41-65. 앫 Temporal invariance in repeated cross-sectional ‘fear of crime’ research, International Review of Victimology, 2005, 12, 3, 273-292, (with L. Pauwels, G. Vervaeke, J. Goethals). 앫 Assessing cross-cultural validity of fear of crime measures through comparisons between linguistic communities in Belgium. European Journal of Criminology, 2005 2, 2, 139-159 (with L. Pauwels). Susann Rabold is a Research Associate at the Kriminologisches Forschungsinstitut Nierdersachsen in Hannover. She holds an MAequivalent degree in sociology from Dresden University of Technology. Her research interests are juvenile delinquency, quantitative methods and elder abuse and neglect. Some recent publications: 앫 Ethnische Unterschiede im Gewaltverhalten von Jugendlichen – Die Struktur von Freundschaftsnetzwerken als Erklärungsfaktor, in Hillmann F., Windzio M., Hrsg., Migration, residentielle Mobilität und 252

Self-reported crime.book Page 253 Thursday, April 23, 2009 8:02 AM

On the Authors

Stadtentwicklung, Opladen, Budrich UniPress Ltd., 2008, 313-331, (with C. Baier). 앫 Gewalt und andere Formen abweichenden Verhaltens in Förderschulen für Lernbehinderte Zeitschrift für Pädagogik, 2008, 1 118-1141 (with D. Baier). 앫 Sind Freizeitzentren eigenständige Verstärkungsfaktoren von Gewalt? Zeitschrift für Jugendkriminalrecht und Jugendhilfe, 2008, 19, 258-268 (with C. Pfeiffer, D. Baier). Philippe Robert is a sociologist, directeur de recherches émérite at the Centre national de la recherche scientifique (CNRS). He is the founder and former head of the Centre de recherches sociologiques sur le droit et les institutions pénales (CESDIP), the main French research centre on crime. He has also founded the Groupe européen de recherches sur les normativités (GERN), an European scientific network. Within the Coordination Action CRIMPREV, he coordinated the workshop of which this volume is one of the outcomes. His current fields of research are criminal sociology, measuring crime, victimisation and insecurity surveys. Among his recent publications: 앫 Sociologia do crime, Petropolis, Vozes (www.editoravozes.com.br), 2007. 앫 Sociologie du crime, Paris, La Découverte (www.editionsladecouverte.fr), Coll. Repères, 2005. 앫 Bürger, Kriminalität und Staat, Wiesbaden, VS Verlag für Sozialwissenschaften/GmbH (www.vs-verlag.de), 2005. Giovanni Battista Traverso received his Degree in Medicine and Surgery at the University of Genoa. A specialist in Clinical Criminology and Forensic Medicine at the University of Genoa, he is a Professor of Forensic Psychiatry and Legal Medicine in various schools of specialization of the Siena Medical School (Neurophysiopathology, Psychiatry, Clinical Psychology, Legal Medicine) and a Visiting Professor at the Clarke Institute of Psychiatry at the University of Toronto. He is an active member of the Italian Society of Forensic Medicine of the American Academy of Psychiatry and Law, of the European Association of Psychology and Law. Simona Traverso received her Degree in Medicine and Surgery and her specialization in Psychiatry at the University of Siena. As Assistant Professor, she teaches forensic psychiatry in Siena Medical School, and is a Professor of Social Medicine at the Political Sciences Faculty (Degree in Sciences of the Social Service) of the University of Siena. She participates in some research projects with others Italian universities analysing, for

253

Self-reported crime.book Page 254 Thursday, April 23, 2009 8:02 AM

Self-Reported Crime and Deviance Studies in Europe

example, intimates' homicide (PAR 2004), juvenile delinquency through the use of self report questionnaire (MIUR 2006).. She authored many national and international papers about mental illness and violent behaviour, homicide-suicide in Italy, intimates' homicide, juvenile delinquency. Renée Zauberman is a senior researcher at the Centre de recherches sociologiques sur le droit et les institutions pénales, a research centre affiliated to the French Centre National de la Recherche scientifique. Her interest lies is norms and deviance, especially criminal justice and crime. While she has worked on the French Gendarmerie, her current focus is on victimisation surveys. She has published : 앫 Zauberman R. Ed., Victimisation and Insecurity in Europe ; A review of Surveys and their Use, Brussels, VUBPress, 2008. 앫 L’évolution de la délinquance d’après les enquêtes de victimation. France, 1984-2005, Déviance & Société, 2008, 32, 4, 435-472 (avec Ph. Robert, S. Névanen, E. Didier). 앫 Police, Minorities and the French Republican Ideal, Criminology, 2003, 41, 4, 1065-1100 (with R. Lévy).

254