AIC_2007 Provided by University Library of Tübingen Transcribed with Tesseract

This work is protected by copyright or related property rights but available in Open Access. The right of use and especially the right to reproduction is only granted within the legal limits of copyright law or due to the consent of the copyright holder.

 Annales internationales de criminologie, 45 ANNALES INTERNATIONALES DE CRIMINOLOGIE INTERNATIONAL ANNALS OF CRIMINOLOGY ANALES INTERNACIONALES DE CRIMINOLOGIA ANNEE 2007 (volume 45-1/2) SOCIÉTÉ INTERNATIONALE DE CRIMINOLOGIE Ministère de la Justice 12, rue Charles Fourier 75013 — Paris phone 33.(0)1.45.88.00.23 fax : 33.(0)1.45.88.96.40 e-mail crim.sic@wanadoo.fr web site http://perso.wanadoo.fr/societe.internationale.de.criminologie ANNALES INTERNATIONALES DE CRIMINOLOGIE Organe officiel de la Société Internationale de Criminologie Président de la Société : Tony PETERS (Belgique) Secrétaire général de la Société : George PICCA (France) COMITÉ DE RÉDACTION DES ANNALES Rédacteur en chef Prof. Georges KELLENS (Belgique) adj. Prof. Michaël DANTINNE (Belgique) Rédacteurs Associés Prof. Edmundo Oliveira (Brésil) Prof. Louise Shelley (USA) Prof. Setsuo Miyazawa (Japon) La redacciôn agradece a la Dra. Maria Jesüs Guardiola Lago su colaboracion en la traduccion de los textos en castellano. The editors want to thank Dr. Maria Jesûüs Gardiola Lago for her assistance in Castilian. Merci à Maria Jesûs Gardiola Lago, post-doctorante, pour son aide en castillan. Table de matières Table of contents Indice XV* Congrès mondial de criminologie : entre passé et avenir XV Congreso Mundial de Criminologia : entre el pasado y el futuro XVth World Criminology Congress : between the past and the future . . e »s + e « » e » e e s + » » » e . e « » e * * »s » * 6 « « + « « « — — « — e e Georges Picca The criminological profession : the development of a job mar- ket for criminologists in (Flanders) Belgium Johan Goethals . . e + e » e + + + « < — « » = + « < © « s. s. + e e . — e « « e » » e e « « © e 19 Early predictors of late onset offenders Georgia Zara, Ph.D., David P. Farrington, PhD. .......….. 37 Long-term effects of investment in infant health care on homi- cide in Italy 57 Uberto Gatti, Hans M.AÀ. Schadee, Richard E. Tremblay.. Foreign juvenile offenders in Germany, Italy, France and Spain : Theories, profiles and notes for a social policy César San Juan, Fernando Bermejo, Estefania Ocâriz . . . « e e T1 — 5— XVe CONGRES MONDIAL DE CRIMINOLOGIE : ENTRE PASSÉ ET AVENIR George Picca ‘ Le prochain Congrès mondial de criminologie (Barcelone 20-25 juillet 2008) est l’occasion de souligner l’origine et l’actualité des Congrès pré- cédents, organisés depuis plusieurs décennies par notre Société, avec le concours actif de responsables nationaux, gouvernements et universités. Décembre 1934 : au cours d’une réunion tenue à Paris, au Ministère de la justice, présidée par le Ministre et, à laquelle participaient des délé- guêés de nombreuses Sociétés et, Instituts spécialisés de divers pays, est décidée la réalisation d’un premier Congrès mondial, afin de fédérer ces différentes initiatives. Il est également proposé de tenir un prochain Congrès mondial à Rome, en 1938. Présidé par le Premier Président de la Cour de Cassation d’Italie, ce premier Congrès de criminologie rassemble une fédération de vingt neuf Sociétés et Instituts nationaux. Dans cette perspective encourageante, un nouveau Congrès est prévu au Brésil, à Rio de Janeiro et San Paulo. Malheureusement la guerre mondiale interdit, alors, toute communi- cation internationale et, par suite, dans les domaines concernés. Il fau- dra, dès lors, attendre 1949 et 1950 pour que la démarche initiale des fondateurs puisse se développer. C’est alors que la Société Internationale de Criminologie également, s’organise et étend son influence. Ses statuts lui permettent, désormais, d’être honorée d’un Statut consultatif par les Nations Unies (Conseil Économique et Social) ainsi que par l'UNESCO. La Société Internationale de Criminologie, Organisation non gouver- nementale (ONG), a ainsi pour vocation, depuis reconnue, d’assurer le développement des sciences, dans leur application au phénomène crimi- nel, dans les différentes sociétés humaines. Elle contribue également à faciliter la coordination des Associations et Instituts nationaux dans le domaine considéré. Quatorze Congrès mondiaux se sont, depuis déroulés, tant en Europe que, dans les Amériques et en Asie. Ils ont, chacun réunis au moins un millier de participants de diverses disciplines ou praticiens de la justice pénale, venus des différentes régions du monde, favorisant, ainsi, les échanges multiculturels et professionnels entre juristes, sociologues, psy- ( Secrétaire général de la Société Internationale de Criminologie _7— chologues et, responsables de police et de la justice. Quels enseignements peut-on aujourd'hui en tirer’ Outre leur rayonnement géographique, ces Congrès ont permis des échanges multidisciplinaires entre, d’une part, responsables et, praticiens de la police, de la justice pénale et, d’autre part, des chercheurs en scien- ces sociales. La recherche scientifique est ainsi désormais confrontée aux exigences des politiques anticriminelles et, aux réalités dans les diverses évolutions des délinquances. Longtemps trop éloignés dans leur périodicité, ces Congrès sont aujourd’hui prévus tous les trois ans, prenant, ainsi en compte, de manière plus efficace l’accélération de la connaissance dans les sciences sociales mais, également les variations qui affectent les différentes ten- dances de la criminalité. À cet égard, les thèmes traités au Congrès de Budapest (1993) «Socio-political change and crime :a challenge of the 21st century», Séoul (1998) « Crime and justice in a changing world : Asian and global perspectives» et Rio de Janeiro (2003) « Reducing crime and promoting justice : Challenges to Science, Policy and Practice», apparaissent signi- ficatifs. Ils témoignent, en effet, de la prise en compte des transforma- tions des diverses formes de la délinquance (criminalité urbaine, crimi- nalité des affaires, trafic de stupéfiants), en même temps que de l’évolution des connaissances dans le domaine des sciences sociales et de l’adaptation des politiques anticriminelles et de la prévention sociale, dans un contexte de globalisation. Quel va être demain, l’apport du prochain Congrès dans le contexte de ces diverses évolutions? Le Congrès de Barcelone a retenu pour thème central «Criminalité et criminologie : recherche et action». Il est destiné à s'inscrire dans l’évolution des précédents Congrès en mettant l’accent à la fois, sur les données principales de l’évolution de la crimi- nalité dans le monde d'une part, et les domaines d’action de la crimino- logie. La problématique qui fonde ce thème général et, les principales pers- pectives qu’il offre à la réflexion du Congrès sont destinés ainsi à se développer suivant trois grandes directions principales : — Quel nouvel avenir proposer pour la criminologie, en l’état des mena- ces actuelles (et à prévoir) que révèle la criminalité dans le contexte actuel de la globalisation? — Comment la criminologie doit-elle se développer utilement, dans les domaines respectifs de la recherche et de l’action? — Les réponses, jusqu’ici apportées par la criminologie et la justice pénale à ces interrogations sont elles adaptées (ou insuffisantes), _ 8- compte tenu des tendances actuelles (et à prévoir) de la criminalité et des problèmes que posent la prévention et le contrôle social? S'agissant de la criminalité transnationale, en plein développement, 1l importe de prendre en compte la mesure des tendances dominantes actuelles de la criminalité, dans un contexte de globalisation économi- que, sociale et culturelle. Le développement du crime organisé, (crimi- nalité économique, cybercriminalité), ainsi que du terrorisme, du trafic d’êtres humains et de stupéfiants, exigent une coopération interétatique, et transcontinentale et le développement de politiques supranationales en vue de l’adaptation des législations et des politiques criminelles natio- nales Il importe de souligner, en second lieu, le développement de la crimi- nalité urbaine qui est une conséquence du développement croissant de l'urbanisation, dans différentes régions du monde. Ce développement appelle une réflexion renouvelée sur la relation entre ce phénomène social et la criminalité, notamment dans le cadre de la mondialisation. Cela suppose en particulier, une réflexion sur les problèmes spécifiques posés par le développement de grandes mégapoles, sur l’insécurité des populations mais également de l’influence de la diversité des cultures et des phénomènes migratoires. Cela implique, notamment, la mise en oeuvre de nouvelles stratégies (police, prévention) et une rénovation des politiques urbaines (city planning) Il importe, enfin de souligner que, non seulement, La criminalité a augmenté le nombre de victimes, mais également eu pour conséquence de susciter un intérêt nouveau pour les victimes et, la place qui doit leur être réservée dans le procès pénal, ce qui plaide en faveur d’une moder- nisation des procédures. Victimisation, victimologie et droits des victi- mes ouvrent de nouveaux champs à l’évolution de la justice pénale. C’est dans ce sens que se développent diverses procédures de réparation, com- pensation ou médiation. Par suite, 1l importe de prendre en compte l'existence et le développement d’'associations de victimes qui sont autant de partenaires nouveaux dans le déroulement du procès pénal et parfois constituent des groupes de pression en faveur d’une justice, désormais qualifiée de «restaurative». Telles sont les grandes lignes du thème général du XV* Congrès mon- dial de criminologie. C’est, par suite, un champ particulièrement vaste qui s’offre à la réflexion, à la confrontation des politiques pénales ainsi qu’aux divers domaines de la recherche scientifique. Nul doute que ce Congrès devrait, par voie de conséquence, offrir des perspectives nouvelles aux responsa- bles des stratégies sociales, des actions de prévention et des politiques anticriminelles. — 9 C’est dans cet espoir que nous formons le souhait que ce XV* Con- grès mondial suscite, à la fois l’intérêt par le thème choisi et, rassemble de nombreux participants. Dans cette attente nous remercions particu- lièrement, l’Université de Barcelone et l’Université Ramon Llull de leur contribution active à l’organisation du Congrès. Nos remerciements vont également aux diverses autorités qui contribuent à cette organisa- tion. Le choix de la ville de Barcelone, à la fois lieu privilégié de culture et de modernité, est un gage supplémentaire de succès. — 10 — XV CONGRESO MUNDIAL DE CRIMINOLOGIA : ENTRE EL PASADO Y EL FUTURO George Picca ‘ El prôximo Congreso Mundial de Criminologia (Barcelona, 20-25 de julio de 2008) brindarä la oportunidad de resaltar el origen y la actua- lidad de los Congresos anteriores, organizados desde hace varias déca- das por nuestra Entidad, con la participaciôon activa de responsables nacionales, gobiernos y universidades. Diciembre 1934: durante una reuniôon efectuada en Paris, en el Ministerio de Justicia, presidida por el Ministro y en la que participaban delegados de diferentes Entidades e Institutos Especializados de varios paises, se tomô la decisiôn de llevar a cabo un primer Congreso Mundial a fin de federar las diferentes iniciativas. Asimismo, se propuso efectuar el Congreso Mundial en Roma, en 1938. Presidido por el Primer Presidente del Tribunal Supremo de Italia, este primer Congreso de Criminologia reûne una federaciôn de veinti- nueve Entidades e Institutos Nacionales. En esta perspectiva alentadora, afecta a un nuevo Congreso en Brasil, en Rio de Janeiro y San Paulo. Desafortunadamente, la guerra mundial impide toda comunicaciôn internacional y, por consiguiente, en los sectores atañidos. Habrâ que esperar hasta 1949 y 1950 para poder impulsar la acciôn inicial de los fundadores. Es en ese periodo que la Sociedad Internacional de Crimi- nologia se organiza y extiende su influencia. Sus estatutos le permiten, a partir de entonces, ser privilegiada con un Estatuto Consultivo por la Naciones Unidas (Consejo Econômico y Social), asi como por la UNESCO. La Sociedad Internacional de Criminologia, un Organismo No Gubernamental (ONG), tiene por vocaciôn, desde su fundaciôn, garan- tizar el desarrollo de las ciencias, aplicaändolas en el fenômeno criminal de las diferentes sociedades humanas. Asimismo, contribuye a facilitar la coordinaciôn de las Asociaciones e Institutos Nacionales en el sector atañido. ( Secretario general de la Sociedad Internacional de Criminologia — 11 — Ya suman catorce los Congresos Mundiales efectuados, tanto en Europa como en América y Asia. Cada uno de ellos reunidô, por lo menos, mil participantes de diversas disciplinas o procuradores de la jus- ticia penal, provenientes de diferentes regiones del mundo, favoreciendo asi el intercambio multicultural y profesional entre juristas, socidlogos, psicôlogos y responsables de la policia y la justicia. ;Qué enseñanza nos han dejado? Ademâs de su expansiôn geogräfica, dichos Congresos han permitido intercambios multidisciplinarios entre responsables, agentes policiales y procuradores de la justicia penal por un lado y, por otro, investigadores en ciencias sociales. De esta manera, hoy en dia, la investigacion cienti- fica se ve confrontada a los imperativos de las politicas anticriminales y a las realidades de las diferentes evoluciones de la delincuencia. Tras periodos de realizaciôn bastante distantes durante mucho tiempo, hoy dia, estos Congresos se efectüan cada tres años, lo que per- mite considerar de una manera mâs eficiente la aceleracion del conoci- miento en las ciencias sociales, pero también las variaciones que afectan las diferentes tendencias de la criminalidad. ÀA este respecto, los temas tratados en el Congreso de Budapest (1993) “Socio-political change and crime : a challenge of the 21* cen- tury”, Seul (1998) “Crime and justice in a changing world: Asian and global perspectives” y Rio de Janeiro (2003) “Reducing crime and pro- moting justice : Challenges to Science, Policy and Practice”, son signif1- cativos. Revelan, en efecto, la consideraciôn de las transformaciones de las diferentes formas de delincuencia (criminalidad urbana, criminalidad financiera, trâäfico de estupefacientes), a la par que la evoluciôn de los conocimientos en el ambito de las ciencias sociales y de la adaptaciôn de las politicas anticriminales y de la prevenciôn social, en un contexto de globalizaciôn. Y el dia de mañana, (qué beneficios traerâ el prôximo Congreso en el contexto de estas diferentes evoluciones? El Congreso de Barcelona tiene por tema central “Criminalidad y criminologia : investigaciôon y acciôn”. Estä destinado a entrar en la evoluciôn de los precedentes Con- gresos, enfatizando al mismo tiempo sobre los datos principales de la evoluciôn de la criminalidad en el mundo y los ambitos de acciôn de la criminologia. La problemätica que plantea este tema general, asi como las princi- pales perspectivas que presenta dicho tema como àambito de reflexion durante el Congreso estän destinadas a desarrollarse siguiendo tres direcciones principales : — 11n— — iQué nuevo futuro se debe proponer a la criminologia, en las condi- ciones de amenaza actuales (y a prever) que revela la criminalidad en el contexto actual de la globalizaciôn? — (Cômo desarrollar la criminologia de una manera ûtil, en los respec- tivos ambitos de la investigaciôn y la acciôn? — (Las respuestas aportadas hasta ahora por la criminologia y la justicia penal a estas cuestiones se adaptan o son insuficientes, teniendo en cuenta las tendencias actuales (y a prever) de la criminalidad y los problemas que plantean la prevenciôn y el control social* En lo que respecta la criminalidad transnacional, en pleno desarrollo, es importante tener en cuenta la medida de las tendencias dominantes actuales de la criminalidad, en un contexto de globalizaciôn econômica, social y cultural. El desarrollo del crimen organizado, (criminalidad eco- nômica y cibercriminalidad), asi como del terrorismo, el träfico de seres humanos y estupefacientes, exigen una cooperaciôn interestatal y trans- continental, y el desarrollo de politicas supranacionales con miras a la adaptaciôn de las legislaciones y las politicas criminales nacionales. En segundo lugar, es importante señalar el desarrollo de la crimina- lidad urbana como una consecuencia del desarrollo creciente de la urba- nizaciôn en las diferentes regiones del mundo. Dicho desarrollo conlleva una nueva reflexiôn sobre la relaciôn entre este fenômeno social y la cri- minalidad, en particular, en el marco de la mundializaciôn. Esto supone, en particular, una reflexiôon sobre los problemas especificos planteados por el desarrollo de grandes megâäpolis, la inseguridad de las poblacio- nes, pero también la influencia de las diferentes culturas y los fenômenos migratorios. Esto implica, sobre todo, la implementaciôn de nuevas estrategias (policia, intervenciôn) y una renovaciôn de las politicas urba- nas (city planning). Por ültimo, es importante señalar que no sôlo la criminalidad ha aumentado el nûmero de victimas, sino también ha debido suscitar un interés nuevo para las victimas y el lugar que se les debe reservar en el proceso penal, lo que exige una modernizaciôn de los procedimientos. Victimizaciôn, victimologia y derechos de las victimas abren nuevas perspectivas para la evoluciôn de la justicia penal. En este sentido se desarrollan diferentes procedimientos de reparaciôn, compensaciôn o mediaciôn. Por consiguiente, es importante tener en cuenta la existencia y el desarrollo de asociaciones de victimas que toman parte en el desa- rrollo del proceso penal y, en ocasiones, constituyen grupos de presiôn en favor de una justicia calificada, actualmente, como “restauradora”’. Tales son las grandes lineas del tema general del XV Congreso Mun- dial de Criminologia. — 13 — Por consiguiente, es un ambito particularmente amplio que se presta a la reflexiôn, la confrontaciôn de las politicas penales, asi como a los diversos Ambitos de la investigaciôn cientifica. Nadie duda que este Congreso deberia brindar, por consiguiente, nuevas perspectivas a los responsables de las estrategias sociales, las actuaciones de prevenciôn y las politicas anticriminales. Con esta esperanza, deseamos que este XV Congreso Mundial suscite el interés por el tema elegido y convoque a numerosos participantes. En espera de ello, agradecemos muy particularmente a la Universidad de Barcelona y la Universidad Ramon Llull su participaciôn activa en la organizaciôn del Congreso. Asimismo, agradecemos a las diversas auto- ridades su participaciôn a dicha organizaciôn. La elecciôn de la ciudad de Barcelona, lugar privilegiado de cultura y modernidad, es una prueba de éxito adicional. — 14 — XVth WORLD CRIMINOLOGY CONGRESS : BETWEEN THE PAST AND THE FUTURE Georges Picca ‘ The next World Criminology Congress (Barcelona, 20th-2Sth July 2008) is an opportunity to underline the origin and topicality of the con- gresses that our society has organised for several decades with the active cooperation of national officials, governments and academic institu- tions. December 1934: During a meeting in Paris at the French Justice Ministry, chaired by the Minister and attended by delegates from a large number of societies and specialist institutes from several countries, a decision was made to hold a first world congress to pool the initiatives already taken. À proposal was also made to hold another world con- gress in 1938. Chaired by the first presiding judge of the Italian Supreme Court, this first criminology congress brought together twenty-nine soci- eties and national institutes. Following this encouraging start, another congress was planned for Brazil, in Rio de Janeiro and Säo Paulo, unfortunately the international dialogue broke down with the onset of war, even in academic circles, so it was not until 1949 and 1950 that the founders’ original idea was able to bear fruit. This was also when the International Society for Criminology was organised and extended 1ts influence. Having drawn up its articles of association, it was honoured with consultative status by the United Nations (Economic and Social Council) and Unesco. Thus, the role of the International Society for Criminology, an internationally recognised NGO, was defined as developing the sciences in their application to the phenomenon of crime in different human societies and facilitating coor- dination between national associations and institutes in the field of crim- inology. Fourteen world congresses have been held since those days, in Europe, the Americas and Asia, each attracting at least a thousand par- ticipants from various disciplines, including criminal justice practition- ers, from different parts of the world, thereby fostering multicultural and professional dialogue between jurists, sociologists, psychologists, ( General Secretary, International Society for Criminology. _— 15 — police officers and representatives of justice systems. What lessons can we draw from them today? Beyond their geographical influence, the ISC congresses have pro- vided a forum for police officers and criminal justice officials, on one side, to discuss issues with social science researchers, on the other. Sci- entific research now has to deal with the demands of anti-crime policies and new developments in crime. For many years our congresses were spread too far apart but now that they are held every three years they can effectively reflect the acceleration of knowledge in the social sciences and also the changes affecting criminal trends. The themes of the con- gresses held in Budapest (1993), “Socio-political change and crime: a challenge of the 21st century”, Seoul (1998), “Crime and justice in a changing world: Asian and global perspectives”, and Rio de Janeiro (2003), “Reducing crime and promoting justice : challenges to science, policy and practice”, are significant because they reflect the debate over the changes in different forms of crime (urban crime, business crime and drugs trafficking) along with the development of knowledge in the social sciences and the adaptation of anti-crime and social welfare policies in a context of globalisation. What will our next congress contribute in the context of all the changes that we are facing? The central theme chosen for Barcelona 1s “Crime and criminology : research and action’”, which fits in with the evolving themes of previous congresses by stressing both the main fea- tures of the development of crime around the world and the spheres of criminology. Three main issues underlie this general theme, offering plenty of scope for discussion : — Where should criminology be heading given the current (and forecast) threats of crime in today's context of globalisation? — How should criminology develop in the fields of research and action? — Have criminology and the criminal justice system responded appropri- ately to these questions given the current (and forecast) trends in crime and the problems raised by social welfare and control? In terms of transnational crime, which is developing fast, it is impor- tant to assess the current dominant trends in crime in a context of eco- nomic, social and cultural globalisation. The development of organised crime (economic crime and cyber-crime), terrorism, human trafficking and drugs trafficking require inter-state and transcontinental coopera- tion and the development of supranational policies to adapt national legislation and policies on crime. In the second place, it is important to stress the development of urban crime as a result of increasing urbanisation around the world, which requires a rethink of the relationship between the concentration — 16 — of populations and crime, especially in the context of globalisation, and particularly the specific problems raised by the spread of large cities, the fear of crime and also the influence of cultural diversity and population migration. Among other things, this implies the implementation of new strategies (policing and crime prevention) and urban policies (city plan- ning). Finally, we must bear in mind that crime has both increased the number of victims and created a new interest in them and the role that they should be given in criminal trials. This is an argument in favour of modernising judicial procedures. Victimisation, victimology and victims” rights are opening up new areas for the development of criminal justice and various reparation, compensation and mediation procedures are being introduced. It is therefore important to take account of the exist- ence and spread of victims’ groups, which are new partners in the crim- inal trial process and often lobby in favour of so-called “restorative” justice. Such will be the main strands of the overall theme of the XVth World Criminology Congress. So there will be plenty of room for discussing and comparing crim- inal policies and the fields of scientific research and no doubt this Con- gress will offer new ideas to those in charge of social policy and strate- gies for preventing and fighting crime. We therefore hope that the theme will generate lots of interest and that this XVth World Congress will attract a large number of participants. We would particularly like to express our gratitude to Barcelona and Ramon Llull universities for their active contribution to the holding of the event. Our thanks also go to the various authorities that are helping to organise it. The choice of Barcelona, a city symbolising both culture and modernity, is a further guarantee of success. _— 17 — THE CRIMINOLOGICAL PROFESSION : THE DEVELOPMENT OF A JOB MARKET FOR CRIMINOLOGISTS IN (FLANDERS) BELGIUM Johan Goethals ‘* THE HISTORY OF THE ACADEMIC CRIMINOLOGY IN BELGIUM The birth of criminology as a social and normative science goes back to the end of the 19'” century, when religious and meta-religious definitions and explanations of crime, and later on, the classic para- digm of the individual as a rational being, were challenged and sub- stituted by explanations from developing social and medical sciences like sociology, psychiatry and psychology. The legal dogmatic (classi- cal) position that human acts are to be judged as results of free will was given up, and the new developing sciences connected crime with its natural causes (the aetiological tradition). In Belgium and abroad, this scientific revolution was strengthened by political and societal concerns about the social turmoil at the end of the 19th and the begin- ning of the 20th century, a social turmoil that challenged the govern- ment to strengthen its control on society and to protect itself and soci- ety against dangerous persons and events (the social defence movement). This social and criminal policy included a need for scien- tific explanations of criminal behaviour and for active interventions in the field of social problems (doctrine of social defence), and it explains the creation of criminology as a scientific university programme in the first half of the 20th century. In Belgium, the Catholic University of Leuven created its criminology programme in 1929, followed by the Free University of Brussels (Université Libre de Bruxelles) in 1936, and the State Universities of Liège and Ghent in 1938. Their academic criminological programmes were focussing on ‘the criminal and his behaviour’ from a multidisciplinary perspective : not only penal law and penal procedure were taught, but also forensic psychology and psychiatry. Moreover, as the criminological training was aimed at pre- paring future practitioners in the field (lawyers, prison staff…), it also included topics like protection of youth, forensic medicine, forensics, (*) Professor of Criminology, LINC (Leuven Institute of Criminology), Catholic Univer- sity of Leuven — 19 — and prophylaxis of delinquent behaviour. The emphasis on the social defence ideology resulted in strongly policy oriented academic pro- grammes, with more attention directed to penal practice than to fun- damental and theoretical research. These new criminological institutes were attached to the law faculties, their students being civil servants already working in the criminal justice system, and the content of the training overwhelmingly legal and clinical. Other European countries as France, Italy, the Scandinavian countries, the Netherlands and Ger- many have witnessed similar developments, although in different degrees. For example, some universities in the Netherlands (Utrecht, U. Amsterdam) created chairs and/or criminological institutions that were more research-oriented and offered only a limited amount of optional courses in the law faculties, with no specific aims to instruct magistrates, court officials or other civil servants working in the crim- inal justice system. Things changed fundamentally in the sixtiess when (European) criminology was strongly influenced by the developments in the Amer- ican criminology : in the second half of the century criminology turned from a rather legal and clinical subdiscipline into a subdiscipline of sociology. The findings of the dark number approach and the influ- ence of the phenomenological tradition added a new paradigm to the classic aetiological and social defence based paradigms in Europe : the symbolic interaction and labelling perspectives that strongly threatened the basics of the aetiological tradition. Moreover, the development of the phenomenological tradition was successful in weakening the nar- row deterministic approach of the aetiological tradition. According to this approach, human behaviour is not only determined by natural causes, but it is also shaped by meanings that humans attach to the situation. This tradition invaded European criminological thinking and resulted in a much more sociologically oriented criminology in the six- ties and seventies. It not only included the labelling tradition, but later on also developed variants of radical and critical criminology, and more recently new paradigms of peacekeeping and restorative crimi- nology (e.g. Burke, 2005). It was during this period that Belgian uni- versities modified their programmes into more ‘critical’ programmes, focusing on social and justice institutions rather than on the individual criminal. Moreover, these new criminology programmes attracted a new kind of students. Where in the first period academic criminolog- ical education was designed to train criminal justice workers, since the sixties gradually other interested students, not working in the criminal justice system, showed interest in criminological matters. Until the end of the sixties, universities offered criminology courses as an additional master programme to interested masters in other disciplines, such as — 20 — lawyers, psychologists, sociologists and medical doctors. At the end of the sixties (1968), the Leuven University, followed by the other Dutch speaking universities (Free University of Brussels and the University of Ghent) decided to add a bachelor (candidature programme) to the master. The French speaking universities did not. With this, Belgium imitated what already existed in Canada (Ecole de Criminologie of the Université de Montréal) and the U.S. (School of Criminology, Univer- sity of Berkeley). During that period, it was less clear what kind of jobs criminologists could perform on the job market, as the majority of them were not professionals involved in the criminal justice system, but outsiders to the criminal justice system, frequently taking a critical stance to the system. À new period started recently. Our society faced important challenges : increasing crime rates that are considered in many countries by a large part of the population as one of the main social problems. An important consequence of this is a growing need for advanced knowledge about new forms of crime (fraud, corporate crime, organised crime, cybercrime ..) and for a ‘more appropriate approach’ to crime. Due to their research and publications about crime and their activities in the criminal justice system, criminologists brought the victim back in the criminal justice system, and they instigated new forms of dealing with crime (victim care, mediation, restorative justice, administrative jus- tice, privatisation of punishment…) and new types of sanctions and pre- vention. All these developments necessitated basic scientific research and created new employment opportunities. THE PARTICULAR SITUATION OF BELGIUM In addition to these general developments, the scientific criminol- ogy in Belgium and the position of the professional criminologist are strongly influenced by some particularities of the Belgian state. Since 1970 Belgium is a federal state, composed of three language commu- nities (Dutch, French and German), and three regions (the Flemish region, the Walloon region and the region Brussels capital). Its popu- lation is about 10.000.000 inhabitants, 60% of them living in Flanders, 30% in the Walloon region and 10% in the region Brussels-capital. The “communities” were originally mainly established to protect the spe- cific cultural characteristics of the Dutch-speaking, French-speaking and German-speaking inhabitants in the country. Later on their com- petences were expanded to include personal matters (health system, family policy, youth protection, etc.), education and the use of lan- guage. Because one of the competences of the communities 1s educa- _— 21 — tion, the university training is organised by the Flemish and the French Communities respectively. In each part of the country five uni- versities organize full university programmes, three of them including criminology programmes : three Flemish universities (Ghent, Leuven and Brussels) offer full bachelor and master programmes in criminol- ogy, three French speaking universities (Louvain-la-Neuve, Brussels and Liège) only a master degree. The other universities do not offer criminology degrees. Not only the global structure of university crim- inology training differs between the communities, also the content and the entry to the criminology programmes. At the Walloon universities students with a bachelor in law, psychology and other university degrees can enrol in a criminology master programme; bachelor pro- grammes in criminology are not organised. At the three Flemish uni- versities, three-year bachelor programmes are organised along with a one year master programme. While bachelors in criminology have a direct access to the master, other professional bachelors (social work, societal security, nursing, etc.) and academic bachelors and masters (psychology, sociology, law, etc.) can enrol after a transition pro- gramme. Recently the structure of academic training changed from a two years candidature and two years master (licentiate) programme to a three years bachelor and one year master programme. However, the following description of developments in the job market is based on the former training structure, with a two years candidature and a two years master programme. What the effect of the launching of academic bachelors on the job market will be remains to be seen. Another particularity of the Belgian situation 1s the division of pow- ers between the federal and the community/regional levels. Criminal jus- tice essentially remains a federal competence, with the federal level being in charge of the judiciary, the police, the prison system, alternative sanc- tions, after-care and mentally ill offenders, at least where the control aspects are concerned. Treatment, social work, social aid, and health matters are competencies of the communities. This creates difficulties and tensions between both levels of power as far as working with offenders and victims is concerned, but also job opportunities at the ‘bridge’ between both political levels. THE CRIMINOLOGICAL JOB MARKET The remainder of this paper describes the job market for criminol- ogists graduated at the Katholieke Universiteit Leuven (K.U.Leuven). The Leuven Institute of Criminology (LINC) has been monitoring the professional position of criminologist-alumni of the K.U.Leuven since _— 22 — 1980. It is not clear whether the professional situation of the alumni of the other Flemish and Belgian universities 1s similar to this of the Leuven alumni, due to the lack of monitoring studies carried out by the other universities. The university of Ghent published a study cov- ering the alumni 1993-94 to 1997-98 (n=202) (Decorte, 2000) and the Free University of Brussels about their alumni 1969-1998 (Eliaerts, 2000). Due to differences in reference period and the set-up of the questionnaires, the results of these sweeps cannot be compared with the results of the Leuven questionnaires. Neither 1s information avail- able of the graduates in the Walloon region. It is unsure whether the information collected on the Leuven graduates can be generalised to the whole Flemish region, and to the Walloon region. The K.U.Leuven administered a first mail questionnaire in 1978 (Van Kerckvoorde & Peters, 1980) describing the situation of the alumni 1967-1977 (n=262 respondents), followed by six more sweeps covering the alumni 1978-1983 (n=192) (Peters & Van Kerckvoorde, 1984), 1982-1985 (n=224) (Van Kerckvoorde & Puttaert, 1988), 1986- 1989 (n=174) (Puttaert & Van Kerckvoorde, 1992), 1990-1993 (n=113) (Goethals et al., 1997), 1994-1998 (n=145) (Goethals & Pleysier, 2000) and 2001-2004 (n=288) (Neys, 2007). Questions included the entry in the job market, the actual job situation (type of function, salaries, sat- isfaction with the job), experienced problems and an evaluation of the training programme at the university. The surveys included a cover let- ter that emphasised the basic justification for the study, the survey itself, and a return envelope. Minimal changes were added to the ques- tionnaire during the several sweeps. As non response rates in mail sur- veys can be very high (60-70%), several recommendations of the Total Dillman method (1978) were applied : format; sending a new question- naire to respondents who did not reply — twice during the last sweep —, phoning a sample of non respondents to check the reasons of non responding, and collecting some basic information regarding their job situation, in order to compare the groups of respondents and non- respondents and to rule out that both groups are different. These data revealed no differences between respondents and non-respondents, suggesting that the interrogated sample covered the whole population very accurately. Alumni were approached one to three years after graduation. This means that collected information could cover a ref- erence period of one to three years after graduation. Only in the last sweep more detailed information was collected for each year sepa- rately. In the last sweep alumni were approached by an e-mail questionnaire; only those who could not be reached by e-mail were sent a mail questionnaire. The mean response rate was 80% (77% for the last sweep) of the known respondents (respondents whose address __ 23 — was unknown or who lived abroad not included). Only the response rate of the survey 1994-1998 was low : 54%. CHARACTERISTICS OF THE SAMPLES Today, about 70% of the criminology alumni are female (from 51% females in the Leuven cohort 1982-85, 55% 1n the cohort 1986-89, to 71% in the cohort 2001-2004). The mean age of the Leuven samples is 27-28 year, with a minimum age of 23 and a maximum age of 71, and 90% of the respondents under age 30. As the bachelor/master structure has been organised only from 2004 onwards, the respondents performed a study course of (one or two years of) candidature and two years of licentiate. Students could start their criminology studies via a first candidature after having finished secondary school (at age 18), via a one year candidature after having obtained a non-university degree, mainly social work, orthopedagogics and nursing, or after another master degree in Law, Psychology, Education, Social Sciences or Medicine. The results of the last sweep show that a large group of criminologists actually obtain(ed) a second degree (not necessarily a university degree) : of the last sweep 40% graduated in criminology after having obtained another degree, frequently social work, nursing, orthopedagogics and law, and 46% obtained an additional degree dur- ing or after their graduation in criminology. Additional degrees include a licence to teach in secondary school, police training, and other university degrees in public management, sexology, social and cultural anthropology, and business and economics. Only 23% obtained only their criminology degree. RESULTS Type of job Are alumni of the Leuven criminology training programmes doing really ‘criminological jobs'? Respondents were asked in what area they were working, and to indicate whether this was a clinical, policy oriented of administrative job. From table 1 the following conclusions can be drawn. First, at the moment of the administration of the sur- vey (one to three years after graduation), about half of the criminol- ogy alumni are employed in core criminal justice jobs. Table l dis- plays the findings from six Leuven sweeps. It can be concluded from the most recent cohort that more than half of the respondents (57%) are employed in the criminological core-sector, and another 14% in __ 24 _ areas narrowly related to the criminological area. One out of three (30%) is employed outside the criminological and related professions. Together, about three out of four alumni are employed within the criminal justice system. Similar conclusions can be drawn from the Ghent sweep (Decorte, 2000). Further, table 1 shows that since 1990 the percentage of alumni employed in the core areas has strongly increased, although a decrease was registered at the end of the nine- ties (from 65% in 1990-93 to 52% in the 1994-98 cohort). Concur- rently, the rate of criminologists employed in the related fields decreased, and those working in other (non criminology related) fields followed an inverse U-shaped curve. (Non-significant) differences were found between criminologists with a criminology degree and others with criminology and other degrees, more criminologists of the former group being employed in core criminological areas (62%, 40% and 58% 1n the three areas; X?=3.25, df=2, p<.20). That only 40% of the criminologists employed in the related areas are one degree crim- inologists 1s due to the fact that in this group a large group of (female) social workers (having also a criminology degree) are employed. Their social work or other second degree allow them to enter the job market through criminology related jobs. Further, sig- nificant differences are found according to sex : female criminologists are overrepresented in the related areas (90% are women versus 62% en 75% in de other areas : X?=7.00, df=2, p=.030). TABLE 1 : Area of employment (in percentages) area 1978-81 1982-85 1986-89 1990-93 1994-98 | 2001-04 criminological field 29.4 28.1 23.5 64.6 51.7 57.0 criminology 41.2 34.3 27.0 13.3 19.3 14.0 related Not crimi- 29.4 37.5 49.4 22.1 28.9 30.0 nological Table 2 displays the job situation at year 1, 2 and 3 after gradu- ation for the cohort 2001-2004. The data show that half of the alumni immediately did find a criminological job, and that only a small group moved from another job to a core criminological job : from the group that was employed in the criminolog:cal field, 92% remained employed in that area between year 1 and year 2. The stability per- centage for the year 1-2 for the other groups was 84% and 80%. Between year 2 and year 3 the stability percentages were 97%, 88% — 25 — and 85%. These figures also suggest that a large part of the alumni (30%) never enter the criminological job market, and that a large per- centage of the criminologists employed in the related fields remain employed in that field, at least during the first three years of their professional career. The meaning of this stability 1s not clear yet : either the employment situation is very motivating during the first years, or job opportunities are scarce and moving from one job to another risky. TABLE 2 : The stability of the employment during the first three years Year ! year 2 year 3 Criminological field 52 54 60 16 15 12 Criminology related No criminological job 32 31 28 Police, youth protection and corrections, including community sanctions, absorb the largest groups of alumni. Table 3 displays what must be understood by core, related and non related areas. Employ- ment in the core criminological fields includes jobs in the institutions of the criminal justice system such as courts“) (victim assistance ; meditation); police (police intervention and crime investigation ; staff functions; victim assistance ; managing positions:; policy profiling; crime analysis, and behavioural analysis at the federal police level); private security; corrections (managing positions; pol- icy staff functions at the justice department, after-care); the moni- toring of alternative sanctions; youth protection and dealing with problem youngsters (staff functions, policy staff functions); prevention ; mediation ; restorative justice functions in prisons and in the community; policy staff functions at policy departments; and criminological research and training. Related functions are jobs working with marginal groups (homeless; street corner work.….), social work, social aid, education and teaching. Non related func- tions include press, non criminology related civil servants, public health, commerce, finance and assurance. () The function of magistrates in Belgium is restricted to those holding a law degree. — 26 — TABLE 3 : Area of employment (in percentages) 1994-98 2001-2004 Core criminological sector Police 22 22 Courts 7 Youth protection and problems ]1 Corrections 16 Criminological research Prevention/private security Subtotal 32 37 Related Care for marginal groups; social work 16 11 School and education Subtotal 19 14 Not related Press Health care 16 16 Civil servant not criminologist Commerce and finance Subtotal 29 30 Based on the number of employed alumni (94%) We further measured the content of the job, whether the job was rather clinical, policy oriented, administrative or research/training ori- ented. In the last cohort, about one out of four respondents (26%) report doing a clinical job, although frequently in combination with policy, administration, education and training. “Clinical” here refers to working with individual offenders, victims and family (after-care; treatment ; individual counselling). Clinical functions incorporate the rather classic supervision of offenders (‘parole’ officer functions, but stressing the ‘social’ aspect of the reintegration of offenders into soci- ety rather than the ‘control’ aspect), and new functions like supervis- ing offenders with community sanctions, promoting restorative justice in prisons, mediation between offenders and victims, working with __ 27 — juvenile delinquents in ambulatory and residential (drug) pro- grammes, mental health programmes, etc… . Policy oriented func- tions include managing positions with police, prevention, youth pro- tection, and corrections, and the ‘criminologist as bridge between the department of Justice and the departments of Welfare and Public Health’. One out of three describes his job as policy or staff function, another 23% as administrative duties, and 14% as training and research. It can be concluded that the percentage of alumni working in policy functions 1s increasing, in contrast with the percentage of alumni working in clinical and administrative and implementation jobs. TABLE 4 : Type of job (in percentages) Sweep Type of job 1978-81 1982-85 | 1986-89 | 1990-93 | 1994-98 | 2001-04 Clinical 38.2 30.6 23.2 39.2 30.8 26.0 Policy/staff/ 18.4 12.7 19.4 19.6 25.3 38.0 direction Administration/ executive 23.5 35.8 55.1 23.7 23.4 23.0 Research, 19.9 20.7 12.3 17.5 20.4 14.0 training, teaching À less positive finding is that, although a respectable number of alumni are hired on a solid contractual basis, a large and growing per- centage has only temporary contracts. This phenomenon is mainly due to the professional market of security contracts concluded between the government and cities, the dependence of many criminological pro- grammes and functions on state funding on an annual basis, and the carrying out of research functions, which are concluded on a yearly basis. (@) See Jean Proulx’ description of the role of the ‘Montréal’ criminologist _— 28 — TABLE 5 : Type of contract (in percentages) Sweep* 1978-81 1982-85 1986-89 1990-93 1994-98 Independent 8.0 6.3 4.6 0.9 4.4 Indeterminate 57.3 57.1 68.1 54 64.1 Determinate 8.8 13.0 17.3 38.9 26.4 Weak contractual basis 25.7 23.4 9.8 6.2 5.0 * not examined in the last sweep Entry in the job market How much time does it take to find employment in the criminolog- ical sector? How much time are alumni unemployed? Respondents were asked whether they experienced periods of unemployment by indicating on a monthly time frame their job situation. Respondents 2001-04 with only a criminology degree do not differ from graduates with more than a criminology degree (no significant t-test differences), so both groups are collapsed. The mean number of months of full employment, partial employment, and not voluntary unemployment are not different. Às can be expected the number of months of voluntary unemployment during the first year differs (t = 2.000, df= 175, p<.03). This difference reflects the time invested in additional courses and degrees after having gradu- ated in criminology. The first year after graduation is composed of 8 months of full-time employment, 0.6 months voluntary partial unem- ployment, 0.4 months of forced partial unemployment, 0.9 months of voluntary unemployment and 2.1 months of forced unemployment. It can be concluded that our graduates face a (short period) of forced unemployment during 18% of the first year period after graduation. However, not all respondents faced forced unemployment : only 24% was unemployed during a short period of time (max. three months), 10% during four to six months, and 5% ten to twelve months. For 13% of them the unemployment can be seen as problematic (more than six months of forced unemployment). After one year, 85% of the alumni are full time employed, 5% voluntary partially unemployment; 3% forced partially unemployed, 2% voluntary unemployed and 4% involuntary unemployed. The second year after graduation 1s composed of 10.8 months of full employment, 0.4 months of voluntary partial employment; 0.3 months of forced partial unemployment, 0.2 months of voluntary unemploy- ment and 0.3 months of forced unemployment. These graduates are — 29 _— unemployed during 3% of their second year. Two years after gradua- tion, 92% of the alumni are full-time employed, and 8% mentioned forced unemployment. The third year after graduation 1s composed of 11.2 months of full employment, 0.3 months of voluntary partial unemployment; 0.1 month of forced partial unemployment, 0.2 months of forced unemployment and 0.2 months of voluntary unemployment. These graduates are unemployed during 2% of their third year after graduation. Only 3% mentioned (partial) unemployment. TABLE 6 : Employment at the time of the administration of the questionnaire (l to 3 years after graduation) (in percentages) Sweep 1978-81 1982-85 | 1986-89 | 1990-93 | 1994-98 | 2001-04 Full time 73 85 87 94 94 92 employment Voluntary partial Forced partial Voluntary unemployed forced unemployed 17 ]1 studying Satisfaction with the job Respondents were asked whether they were satisfied with their job, whether they estimated their job as criminologically relevant, how they were paid, and whether they wished to move to another job. Table 7 dis- plays the degree of satisfaction of the alumni with the quality of the job (the categories satisfied and moderately satisfied were collapsed). In the cohort 2001-04, 95% of the respondents showed themselves satisfied with their actual job, and the rate of satisfaction has increased from 45% in the first sweep (1978-81) to 95% in the last one. However, one out of three of the respondents (31%) is actively looking for another job. This percentage is lower than in the preceding sweeps (37% and 52%). The reasons that our respondents are looking for another job are partly job related : the function 1s not relevant from a criminological point of view (53%), not (enough) policy oriented, inadequately salaried, too far from home, or not really at a university level. However, the level of satisfac- — 30 — — — tion is not related to the job area. (X” 2.977, df = 2, ns), sex (X? = 0.00, df=1, ns) nor type of degree (X? = .24, df=1, ns). TABLE 7 : Satisfaction with the job situation Sweep 1978-81 1982-85 1986-89 1990-93 1994-98 | 2001-04 (very) satisfied 45 48 83 90 89 95 Not satisfied 55 52 17 10 ]1 % looking for * 52 37 37 37 24 another job * not measured Table 2 displayed the professional situation of the alumni based on our (external) evaluation of the employment situation. In addition we asked the respondents to give their subjective appreciation of their job as criminology related. One out of three respondents (38%) describe their job as strongly criminology related, 33% as fairly criminology related, and one out of three (29%) as not criminology related. Not sur- prisingly, this appreciation 1S strongly related to the area of employment : the more the employment 1s situated in the core area, the more the employment is estimated relevant (8%, 30% and 63% not relevant : X’?=57.41, df=4, p=.00). The same differences are found between the type of job (criminal research and training are estimated less relevant (51%) than the other types of jobs (71%, 78% and 87%)(X?=18.52, df=6, p=.004), and between male en female respondents : 55% of the males and 32% of the females describe their job as strongly criminology related (31% versus 24% as not criminology related (X?=6.73, df=2, p<.04). We further asked the respondents whether their job required a uni- versity degree. This was the case for about half of the groups. Moreover, comparing the different sweeps shows that the number of criminology functions requiring a university degree 1s increasing (32% in the sweep 1986-89, 46% in the sweep 1990-93) (not asked in the last sweep). In the Belgian job market this increase can be explained by the creation of criminological functions by the government (staff functions in preven- tion- and security programmes on city and state level; in alternative sanctions; mediation; etc.…). Criminologists with only one degree are more frequently working on university degree level (63% versus 51%.), the latter group having more opportunities to start a professional career _— 31 — in the area of study they graduated in, often social work or another non-university degree. The earlier sweeps showed the same tendencies. Finally, questions were asked about the salaries of the criminologists. Seven out of ten (68%) of the respondents estimate their salaries as the normal salaries for university graduates; one of three is underpaid. À correlation is found between salary and area of profession : relatively more people working in clinical functions (frequently salaried as social workers, not on a university level) (74%) are paid under the university salary level than criminologists working in administrative functions — — (43%), research and training (33%) and policy functions (10%) (X? 55.62, dfg=4, p=.000). Further differences are found between males and females (male criminologists are better paid : X°=5.78, df=3, p<.05). No differences were found between areas of profession (core areas, related areas and not related areas) (X?= 3.62, df=4, p<.50). SOME CONCLUSIONS Although all this information about the employment market of Flemish criminologists seems positive, some more negative comments should be made. First, although we made efforts to contact as many respondents as possible (several attempts to contact respondents and checks on the non response categories), a large part of the respondents could not be reached or did not respond to the surveys. It can be assumed that exactly this part of the population has experienced the greatest difficulties to find a job in the criminological sector, and that it is exactly because of these experiences that they refrained from responding to the questionnaire. Our picture of the criminological job market might therefore be too positively skewed. Other observations suggest that finding a criminological job is not straightforward : the amount of jobs explicitly defined as jobs for criminologists is scarce; employers are not familiar with the criminology degree; the competition with other types of social scientists for the jobs in the criminological job market is strong; many employment contracts are concluded on an annual basis, and frequently the job is described as rather bureaucratic. Second, the criminological job market in Belgium has profited from the growing crime rate and the creation of new modalities of interven- tion. The employment situation during the last decennium was the result of a cyclical prosperous interplay of pull and push factors on the crim- inological job market. The creation of a series of new functions and services (prevention, mediation, victim care and assistance, and other functions in the local and federal administrations) went hand in hand with a decrease in the number of alumni, with as a result a strong posi- __ 32 _ tion of criminologists on the employment market. However, the actual strong increase in numbers of interested students and the still growing number of alumni criminologists threatens this positive image that has been scheduled here. It is to be feared that once all new created func- tions are filled, new alumni will face more problems to find appropriate employment. The creation of new job opportunities for criminologists is a function of the close collaboration between academics and policy makers through consulting and research, and vice versa through the hiring of policy makers for academic teaching. This close relationship is of utmost importance to narrow the gap between academia and the job market, and is one of the main conditions to create criminological functions. In Belgium, academics have influenced criminal justice policy for youths and adults; they had a very strong influence on the youth protection law and its modifications; they stimulated the creation of a forensic social work field, and sustained the creation of a new prison legislation. Recently, Belgian universities imported the restorative justice philosophy in the Belgian criminal justice system, in a first phase through a series of field experiments to examine the introduction of mediation in the criminal justice system, in a later phase through the creation of ‘restor- ative justice jobs’ in and outside the Belgian prisons (counsellors, medi- ators). All these, and many others are examples of the strong interrela- tionships between academia and policy. The surveys have taught a lot about the content of the criminolog- ical job market. They show what professional roles of criminologists can be. Criminologists are researchers, able to study social problems like (different forms of) crime, objective and subjective insecurity, and the effects of different types of interventions. As such, criminologists should be well trained in research methods, both quantitative and qualitative. They can be hired as researchers at universities, state departments, the criminal justice system, and private corporations. Second, criminologists are analysts of penal, procedural and correc- tional policy, able not only to analyse and evaluate policy in these areas, but also to design and develop it. These criminologists can work for political parties, parliament, local and federal governments, the police, corrections, the press, pressure groups and social organisations. Third, criminologists are well trained for functions where law and behavioural approaches intersect. Criminologists, being trained in law and social science, can function as bridges between magistrates and courts — not acquainted with behavioural approaches —, and the health system — not acquainted with the legal system —. As mentioned above, this 1s a growing need in the Belgian political context. Such functions are case managers for inmates in prisons, case managers for alternative — 33 — sanctions, coordinators in teams treating mentally disordered offenders in psychiatric institutions, prevention workers in municipal prevention teams, etc… À fourth type of criminologist 1s the classic clinical crim- inologist trained to perform person oriented criminological interven- tions (Vervaeke, 2001). Criminological interventions can be distin- guished from non criminological oriented interventions by the goals of the intervention, the underlying theories, the methods to reach these goals, the setting in which the intervention takes place, and the rela- tionship between the involved parties. According to these criteria crim- inological interventions focus on the conflicts caused by rule trans- gressing or behaviour provoking damage to persons and goods, private or public welfare, and the prevention of these conflicts, taking into account that the definitions and experiences of conflicts can be largely different between groups. Criminological interventions are based on diagnostic assessment (crime analysis, analysis of criminal careers, assessment of effects of victimisation, risk assessment), and behav- ioural interventions in the different steps of the prevention pyramid : counselling and advice as primary prevention, all types of early inter- vention programmes as secondary prevention, and many types of intervention on the tertiary prevention level : e.g. victim care, interven- tions in case of family violence, crisis intervention with police officers, detention planning and intervention programmes for inmates, parole sexual supervision, offender-victim mediation, pre-therapy with offenders, and supervision of alternative sanctions. We believe a criminologist should master all the necessary scientific knowledge enabling him to act as a specialist in one of these fields. Criminologists should, from a scientific point of view, be able to define, to analyse and to report on criminological problems and pol- icy, and to critically evaluate policy. They must be able to perform individual and organisation (policy) oriented criminological interven- tions. This necessitates an academic programme that offers an out- standing intellectual endeavour for students, a broad multidisciplinary and contextualising approach of crime, deviance and social control; a qualitative transfer of knowledge based on empirical research on crime and crime control; a broad training in methods of scientific research, problem solving, and jurisprudence; a thorough knowledge of policy and policymaking, and the possibility to acquire a series of vocational skills to work with offenders, victims and other groups involved 1n criminal justice. __ 34 _ BIBLIOGRAPHY BURKE, R.H. (2005). An introduction to Criminological Theory. Cullompton : Wil- lan. DECORTE, T. (2000). “De intrede in de arbeidsmarkt van criminologen afgestu- deerd aan de Universiteit Gent”. In VANDERBORGHT, J., VANACKER, J. & Maes E. (eds), Criminologie. De wetenschap, de mens, pp. 101-126. Brussels : Politeia. DILLMAN, D.A. (1977). Mail and telephone surveys. The total Design Method. New York; Wiley. ELIAERTS, C. (2000). “De intrede in de arbeidsmarkt van criminologen afgestu- deerd aan de Vrije Universiteit Brussel”. In VANDERBORGHT, J., VANACKER, J. & MaEs E. (eds), Criminologie. De wetenschap, de mens. pp. 127-132. Brussels : Politeia. GOETHALS, J., Goris, P. & Maes, E. (1997). “De intrede in de arbeidsmarkt van criminologen afgestudeerd aan de K.U.Leuven in de jaren 1990 tot 1993”, Panopticon, 19, 268-292. GOETHALS, J. & PLEYSIER, S. (2000). “De intrede in de arbeidsmarkt van crimi- nologen afgestudeerd aan de K.U.Leuven”. In VANDERBORGHT, J., VANACKER, J. & Maes E. (eds), Criminologie. De wetenschap, de mens. pp. 81-1000. Brussel : Politeia. Neys, S. (2007). De ‘nieuwe’ mensen van het recht : criminologen. Loopbaanonder- roek bij afgestudeerde criminologen tijdens 2000-2004. Master thesis University of Leuven. PETERS, T. & VAN KERCKVOORDE, J. (1984). “De intrede in de arbeidsmarkt van recent aan de K.U.Leuven afgestudeerde criminologen”. Panopticon, 4, 93- 110. PROULX, J. (2006). Profession criminologue. Montréal : Les Presses de l’Université de Montréal. PUTTAERT, M. & VAN KERCKVOORDE, J. (1992). “De intrede in de arbeidsmarkt van criminologen afgestudeerd aan de K.U.Leuven in de jaren 1986 tot en met 1989”. Panopticon, 12, 268-292. VANDERBORGHT, J., VANACKER, J. & Maes E. (2000). Criminologie. De weten- schap, de mens. Brussel : Politeia. VAN KERCKVOORDE, J. & PETERs, T. (1980). “Tien jaar criminologie aan de K.U.Leuven. Studieloopbaan en beroepssituatie van de afgestudeerden”. Pan- opticon, 1, 470-480. VAN KERCKVOORDE, J. en PUTTAERT, M. (1988). “De intrede in de arbeidsmarkt van criminologen, afgestudeerd aan de K.U.Leuven in de jaren 1982 tot en met 1985”. Panopticon, 8, 472-496. VERVAEKE, G. (2001). “Persoonsgerichte criminologische interventie : beschouwin- gen over de specificiteit”. In DUPONT, L. & HUTSEBAUT, F. (eds). Herstelrecht tussen toekomst en verleden, pp. 559-576. Leuven : Universitaire Pers. __ 35 — ABSTRACT Starting from the history of academic Criminology in Belgium, the author presents the job market for criminologists as it appears from monitoring the professional position of criminologist-alumni of the K.U. Leuven since 1980. Type of job, area of employment, stability of the employment during the first years, time of unemployment, satisfac- tion of the job, are surveyed and discussed, and conclusions are pro- posed for the future market of criminological jobs. RÉSUMÉ Après un historique de la criminologie académique en Belgique, l’auteur présente le marché de l’emploi des criminologues tel qu'’il appa- raît du suivi des diplômés réalisé à l’Université de Louvain (K.U.L.) depuis 1980. Les enquêtes portent sur le type d’emploi, le secteur d’acti- vité, la stabilité d’emploi dans les premières années, le temps d’attente, la satisfaction du travail. Les chiffres sont suivis d’une discussion et de conclusions sur les conditions futures de la demande et de l’offre d’emploi de criminologues. RESUMEN Después de una exposiciôn de la historia de la criminologia acadé- mica en Bélgica, el autor presenta el mercado laboral de los criminôlo- gos, tal y como aparece del seguimiento de los diplomas realizado en la Universidad de Lovaina (K.U.L.) desde 1980. La investigacion hace referencia al tipo de empleo, al sector de actividad, a la estabilidad labo- ral en los primeros años, al tiempo de desempleo y a la satisfacciôon en el trabajo. Estas cifras se completan con un debate y conclusiones sobre las futuras condiciones de oferta y demanda laboral de los criminôlogos. —- 36 — FEARLY PREDICTORS OF LATE ONSET OFFENDERS Georgia Zara, Ph.D. ‘ David P. Farrington, Ph.D. ” Criminal career research is concerned with the onset, persistence and desistance of antisocial and offending behavior. Traditionally most stud- ies (!) have focussed their attention on childhood antisocial conduct and adolescent delinquency, regarding adult criminal onset as superfluous and rare, due to the established thought that virtually all adult offenders were antisocial children and delinquent adolescents. Longitudinal findings ‘ show, however, that offending is not pre- dominantly an adolescent phenomenon. Blumstein and colleagues ‘ noticed in their longitudinal data that there were a substantial number of adult offenders (40% to 50%) who had not been delinquent as juve- niles. In a Swedish follow-up study, 25% of men with a criminal record experienced their first official conviction at age 21 or older‘*), and they represented 9% of the total male sample . In another study‘® carried out by Wolfgang and his colleagues, the proportion of adult offenders, without juvenile criminal records, and responsible for a significant amount of crimes, was 24%. What emerges from their follow-up‘” is that, despite adult-only offenders committing the same number of crimes as juvenile-only offenders, and fewer than the persistent offend- (*) Associate Protessor ot Criminal Psychology, Department of Psychology, Turin Uni- versity. (**) Professor of Psychological Criminology, Institute of Criminology, Cambridge Uni- versity. () RosINs, L.N. & RUTTER, M. (1990) (Eds.). Straight and devious pathways from child- hood to adulthood. New York : Cambridge University Press. (?) FARRINGTON, D.P. (1986). Age and crime. In M. Tonry & N. MORRIS (Eds.). Crime and Justice (Vol. 7) (pp. 189-250). Chicago : University of Chicago Press, p. 235. (3) BLUMSTEIN, A., COHEN, J., ROTH, J.A., & VISHER, C.A. (1986) (Eds.). Criminal careers and “Career criminals” (vol. 2). Panel on Research on Career Criminals, Committee on Research on Law Enforcement and the Administration of Justice, Commission on Behavioral and Social Sci- ences and Education, National Research Council. Washington, DC : National Academy Press. (9) STATTIN, H., MAGNUSSON, D., & REicHEL, H. (1989). Criminal activity at different ages : À study based on a Swedish Longitudinal Research population. British Journal of Criminology, 29, 368-385. G) STATTIN, H. & MAGNUSSON, D. (1996). Antisocial development : A holistic approach. Development and Psychopathology, 8, 617-645. (6) WOLFGANG, M.E., FiGL10, R.M., & SELLIN, T. (1972). Delinquency in a birth cohort. Chicago : University of Chicago Press. () WoLFGANG, M.E., THORNBERRY, T.P., & FiGL10, R.M. (1987). From boy to man, from delinquency to crime. Chicago : University of Chicago Press. __ 37 — ers, adult-only offenders were more likely to be involved in serious crimes. Also McCord‘), in the Cambridge-Somerville Youth Study, identified 41 boys without any juvenile criminal record who were con- victed for serious crimes as adults. In addition, Eggleston and Laub ©), examining data from the Federal Bureau of Investigation “, found that 72% of those individuals arrested for Crime Index offences were adults (aged 18 or older), and that 56% of the violent crime arrestees were aged over 25. In light of these findings, and looking presently at the contemporary criminal situation in Western society, it is striking that the proportion of adults who step over the criminal threshold at a later age is signifi- cantly increasing, to the extent that “adult onset offenders comprised an average of 50 percent of the adult offender population”“!. This may, partly, explain why the mean age of the prison population is rising. For instance, in EFuropean prisons in September 2004, the mean age of inmates was 34, with an age range of 28 to 40?), International criminal statistics and research studies report, not only that the greatest proportion of crime is committed by adult individuals aged 21 or over°), but that a steep rise in the recidivism trend is attrib- utable to adult offenders #. Shepherd and Whiting (2006), using data from a 2003 cohort, found that differences in re-offending depended on age, and while between 2000-2003 the re-offending rates for 18-20 and 21-24 year olds slightly decreased (69% re-offended in 2000 and 67% did so in 2003), at the same time there was a significant increase in re- offending by individuals aged 35 or over (39% re-offended in 2000 com- pared to 43% in 2003)“°). Analysing self-reported life events and the offending history of a sample of 1,011 Chinese male prison inmates, who 8 McCorp, J. (1978). A thirty-year follow-up of treatment effects. American Psycholo- gist, 33, 284-289. () EGGLESTON, E.P. & LauB, J.H. (2002). The onset of adult offending : A neglected dimension of the criminal career. Journal of Criminal Justice, 30, 603-622. (0) Federal Bureau of Investigation (2000). Crime in the United States, 1999. Washing- ton, DC: U.S. Government Printing Office. (D EGGLESTON, E.P. & Laus, J.H. (2002). op. cit., p. 614. (2) Penological Information Bulletin (2006). TANEVA, I. (Ed.). Council of Europe, Stras- bourg, Nos. 25-26, p. 69. (13) Apay, R.H. (2003). Aging prisoners : Crisis in American corrections. Westport, CT : Praeger. MARVELL, T.B., & Moopy, C.E. (1997). Age-structure trends and prison populations. Journal of Criminal Justice, 25(2), 115-124. (4 Bupp, T., SHArP, C. & MAYHeEw, P. (2005). Offending in England and Wales : First results from the 2003 Crime and Justice Survey. (Home Office Research Study 275). London : Home Office Research, Development and Statistics Directorate. (15) SHEPHERD, A., & WHITING, E. (2006). Re-offending of adults : results from the 2003 cohort. London : Home Office Statistical Bulletin, 20/06, p. 6. __ 38 — were between 30 and 39 years of age, Chu“!°) demonstrated that these incarcerated men were likely to have late onset (the mean onset age was 22) and late ages of decreasing criminal involvement (after the age of 30). The involvement of adult individuals in offending behavior is a seri- ous problem, and it should be studied more explicitly. Addressing why adults start criminal careers, and why they persist, may help to under- stand how to prevent adult criminal onset. CRIMINAL CAREER RESEARCH Much adult antisocial and criminal behavior is likely to have some of its roots in childhood difficulties ”. In the last decade there has been a considerable increase in understanding of the links between childhood antecedents of adult behavior and of problematic adjustment later in life. Looking backwards, the majority of adult persisting offenders man- ifested conduct disorders as children. Looking forwards from childhood, however, only a small percentage of children with conduct disorders and antisocial behavior persist in a maladjustment pattern during their adulthood “8), À well-known finding of longitudinal studies 1s that early antisocial onset is likely to exhibit robust links with late and continuing offending. Early onset is related to : a high risk of being involved in a lengthy crim- inal career; a large number of convictions and more heterogeneous types of offences; a higher probability of violence; and a higher frequency of offending‘°’. Such empirical information provides evidence of the importance of sound early interventions to tackle those early risk factors related to late criminal behavior. Despite the extensive research on early criminal careers, very little 1s known about the development of antisocial behavior late in life. West and Farrington “ stated that “[pJerhaps rather surprisingly, it appears that no other investigators have tried to compare the characteristics of (16) CHu, M.M. (2002). Incarcerated Chinese men, their life experiences and criminal onset. British Journal of Criminology, 42, 250-266. ( FARRINGTON, D.P. (1995). The Twelfth Jack Tizard Memorial Lecture : The develop- ment of offending and antisocial behavior from childhood — Key findings from the Cam- bridge Study in Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929-964. (8) RoBINs, L.N. (1966). Deviant children grown up : À sociological and psychiatric study of sociopathic personality. Baltimore : Willlams and Wilkins. (19) PIQUERO, A.R., FARRINGTON, D.P., & BLUMSTEIN, À. (2007), Key issues in criminal career research. Cambridge : Cambridge University Press. (20) WesT, D.J. & FARRINGTON, D.P. (1977). The delinquent way of life. London : Heine- mann, p. 121. — 39 __ juvenile delinquents with those of persons convicted for the first time as adults”. Some studies have dismissed the possibility of an adult onset ‘?), defining it as an artefact of official measurement “?. Some- times, when the issue‘?) was addressed, the definition of adult onset coincided with late adolescence “*), thus missing that proportion of indi- viduals who start offending later in their life. Only recently have a few researchers ?* devoted their attention to adult criminal careers, specifically analysing those proximal correlates that are likely to have a negative influence on an adult individual’s life, such as the loosening of social bonds or informal controls, by family disruption or unemployment *. The present study 1s the first attempt to explore whether adult crim- inal behavior could be predicted and potentially nipped in the bud ‘“”), and to identify whether there are early correlates that, in childhood or adolescence, act as a buffer against negative or antisocial influences, so as to delay the beginning of a criminal career and defer it into adult- hood. Why does an individual start a criminal career in adulthood? How early can future criminal behavior be predicted? Which psychological features affect late antisocial offending? These are three of the main questions that will be addressed in this article in order to explain the patterning over time of an adult criminal CarceTr. (21) MOFFITT, T.F. (2003). Life-course-persistent and adolescent-limited antisocial behavior : À 10-year research review and a research agenda. In B.B. LaHry, T.E. MOFFITT, & A. CasPI (Eds.). Causes of conduct disorder and juvenile delinquency (pp. 49-75). New York : Guilford Press. (22) MoFFITT, T.E. (2006b). Life-course persistent and adolescent-limited antisocial behav- ior. In D. CICCHETTI & D. COHEN (Eds.). Developmental psychopathology (2"% ed., pp. 570- 598). New York : Wiley. (23) KLEVENS, J., RESTREPO, O., RoCA, J., & MARTINEZ, À. (2000). Comparisons of offend- ers with early- and late-starting antisocial behavior in Colombia. International Journal of Offender Therapy and Comparative Criminology, 44(2), 193-203. (24) KROHN M.D., THORNBERRY, T.P., RIVERA, C., & LE BLANC, M. (2001). Later delin- quency careers. In R. LoEBER & D.P. FARRINGTON (Eds.). Child delinquents (pp. 67-93). Thousand Oaks CA : Sage Publications. (25) GOMEZ-SMITH, Z., & PIQUERO, A.R. (2005). An examination of adult onset offending. Journal of Criminal Justice, 33, 515-525. @6) LauB, J.H. & SAMPson, R.J. (2003). Shared beginnings, divergent lives : Delinquent boys to age 70. Cambridge, MA : Harvard University Press. @N SUTTON, C., UTTING, D, & FARRINGTON, D.P. (2006). Nipping criminality in the bud. The Psychologist, 19(8), 470-475. — 40 — ADULT OFFENDING Adult criminality can be predicted with a certain degree of success. In 1985, using a variety of predictive techniques, Farrington “ illus- trated that it was possible to identify at age 10 those boys who subse- quently became convicted as juveniles and then as adults. Adult criminality certainly does not emerge suddenly. It is character- ised by a delayed criminal onset, which 1s likely to be influenced by fac- tors that, in certain individuals, delay or supress early antisocial mani- festations that instead surface later in life. Early criminogenic factors may not always produce their full effects until early adulthood. In some cases there exists a sort of latent process (‘“sleeper effect”) whose out- comes are only seen later in life, which could be a quite daunting dis- covery ?). Elander and colleagues ° conducted a study on a high-risk sample of individuals, and their findings suggested that what was defined as apparently late onset criminality (being convicted only after their 22"“ birthday) was mediated by factors such as childhood disturbances, ado- lescent conduct disorders, and mental problems that placed those indi- viduals at an enduring risk of criminality that was fully expressed only late in life (3 There is an extensive body of research on how psychological factors affect human behavior ” especially among juveniles ®. Nervousness and anxiety (such as being worried for no reason, being shy, weary, withdrawn and socially isolated, avoiding situations that cause nervous- ness, tenseness or a sense of inadequacy) seem to have, especially in (28) FARRINGTON, D.P. (1985). Predicting self-reported and official delinquency. In FAR- RINGTON, D.P. & TARLING, R. (Eds.). Prediction in Criminology (pp. 150-173). Albany, N.Y. : State University of New York Press. (29) WEST & FARRINGTON, (1977), op. cit. (G0) ELANDER, J., RUTTER, M., SIMoNOFF, E., & PickLES, À. (2000). Explanations for apparent late onset criminality in a high-risk sample of children followed up in aduit life. British Journal of Criminology, 40, 497-509. (3 ELANDER, et al., 2000, op. cit., p. 506. G2) DAMASIO, A.R. (1994). Descartes’ Error : Emotion, Reason, and the Human Brain. USA : Putnam Publishing. B3 KEAN, J. (1999). Children’s temperaments and their relation to behaviour and emo- tion across different contexts. Paper presented at the Karly Childhood Council Annual Con- ference, Rotorua, New Zealand, April 29-May 2. McCBURNETT, K. LaHry, B.B., Frick, P.J. RiscH, C., LOEBER, R., HART, E.L. CHRIST, A.G., & HANSON, K.S. (1991). Anxiety, inhibition, and conduct disorder in children : II. Relation to salivary cortisol. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 192-196. OLLENDICK, T.H., SELIGMAN, L.D., & BUTCHER, À.T. (1999). Does anxiety mitigate behavioral expression of severe conduct disorder in delinquent youth? Journal of Anxiety Disorders, 12, 565-574. — 41 — childhood, a proximal protective impact against riskxy and dangerous stimuli. In the Cambridge Study, other things being equal, a shy and timid temperament tended to prevent a boy from becoming delinquent, and this was one of the very few instances in which an apparent adverse fea- ture was associated with an encouraging outcome 6. However, this shielding effect seems to be a negative protective influence °°) because in the long run it may leave the individual vulnerable and unprepared to deal with external, stressful and antisocial influences C6). Resistance to hazards may derive from controlled exposure to risk, rather than its avoidance 3”). Recently researchers 8) have shown considerable interest in the role of anxiety and nervousness in the expression of disruptive behavior, also because of its neurobiological foundation ° Gray’s (1973)(%) conceptualization of behavioral activation and inhibition systems has been applied to understanding various forms of psychopathology , including internalizing‘? and externalizing problems *. His theory suggests that two distinct neurobiological systems operate in opposi- tion to one another. One seems to be involved in the manifestation of aggressive behavior (the Behavioral Activation System or BAS), while G4) WEST and FARRINGTON, (1977), op. cit. p. 115. G5) KERR, M., TREMBLAY, R.E., PAGANI, L., & VITARO, F. (1997). Boys’ behavioral inhi- bition and the risk of later delinquency. Archives of General Psychiatry, 54, 809-816. G6) RUTTER, M. (2003). Crucial paths from risk indicators. In B.B. LAHEYy, T.E. MOFFITT, & À. CAsPI (Eds.). Causes of conduct disorder and juvenile delinquency (pp. 3-24). New York : The Guilford Press. G7) RUTTER, M. (2006). Implications of resilience concepts for scientific understanding. Annals of the New York Academy of Sciences, 1094(1), 1-12, p. 1. 68) OLLENDICK, et al. (1999), op. cit., p. 566. G9) Gray, J.A. (1976). The behavioural inhibition system : À possible substrate for anx- iety. In M.P. FELDMAN & À. BROADHURST (Eds.), Theoretical and experimental bases of the behaviour therapies (pp. 3-41). London : Wiley. Gray, J.A. (1987). The psychology of fear and stress (2nd ed.). Cambridge : Cambridge University Press. (40) Gray, J.A. (1973). Causal theories of personality and how to test them. In J.R. ROyCE (Ed.). Multivariate analysis and psychological theory (pp. 409-463). New York : Academic Press. (4 MEyEr, B., JOHNSON, S.L., & CARVER, C.S. (1999). Exploring behavioral activation and inhibition sensitivities among college students at risk for bipolar spectrum symptoma- tology. Journal of Psychopathology and Behavioral Assessment, 21, 275-292. MINEKA, S., WATSON, D., & CLARK, L.A. (1998). Comorbidity of anxiety and unipolar mood disorders. Annual Review of Psychology, 49, 377-412. (42) KAGAN, J., REZNICK, J.S., & SNIDMAN, N. (1987). The physiology and psychology of behavioral inhibition in children. Child Development, 58, 1459-1473. (43) ARNETT, P.A., SMITH, S.S., & NEWMAN, J.P. (1997). Approach and avoidance moti- vation in psychopathic criminal offenders during passive avoidance. Journal of Personality and Social Psychology, 72, 1413-1428. CARVER, C.S., & WuniTe, T.L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment : The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319-333. __ 42 — the other is implicated in the production of anxious behavior (the Behavioral Inhibition System or BIS). It is likely that higher levels of activity in the BÂS, accompanied by a parallel increase in the activity in the BIS, should result in less severe behavioral manifestations of conduct disturbance. Despite various studies providing contrasting findings ** this theory seems to address the role of mitigating psycho- logical factors upon externalizing behavior. Walker and colleagues ° found that anxiety buffered or counterbalanced the effects of conduct disturbances, in the sense that children with comorbid conduct disor- ders, who were also anxious, later reported fewer police contacts and were rated by peers as less aggressive than children with only conduct disorders. Ollendick and colleagues‘* evaluated whether comorbid anxiety would exercise a protective influence in severely conduct-disordered incarcerated male and female adolescents, in regard to the age of the first offence, number of offences, and severity of initial and most recent offences. Contrary to expectation, their findings diverged from the ones reported in the literature‘“”, in the sense that comorbid anxiety disor- ders in their incarcerated boys and girls did not significantly delay the onset of the first known delinquent act, did not lessen the overall fre- quency of offences, and did not moderate the severity of delinquent acts. One possible explanation of why the mitigating effect of anxiety on their measures of disturbance was not found may be because their investiga- tion focused on particularly strong and very virulent forms of conduct disturbance, which were probably less amenable to the modulating effects of anxiety and nervousness. It is of vital interest then to continue exploring which early psycho- logical and background factors would prove to be specific attributes of those who were first convicted as adults, and which similarities or dis- crepancies there are in comparison with factors that particularly char- acterize either non-offenders or early antisocial onsetters. This fits in with the claim that early individual and psychological features have (44) ANDERSON, J.C., WILLIAMS, S., McGEE, R, & SiLva, P.A. (1987). DSM-III disorders in preadolescent children. Archives of General Psychiatry, 44, 69-76. VERHULST, F.C., & VAN DER ENDE, J. (1993). “Comorbidity” in an epidemiological sample : À longitudinal perspective. Journal of Child Psychology and Psychiatry, 34, 767-783. (45) WALKER, J.L., LAHEY, B.B., RUsso, M.F., FRICK, P.J., CHRIST, M.A., MCBURNETT, K. LOEBER, R., STOUTHAMER-LOEBER, M., & GREEN, S. (1991). Anxiety, inhibition, and conduct disorder in children : I. Relations to social impairment. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 187-191. (46) OLLENDICK, et al. (1999), op. cit. (4 WALKER et al. (1991), op. cit. WOOLSTON, J.L., ROSENTHAL, S.L., RIDDLE, M.À., SPARROW, S.S., CICCHETTI, D., & ZIM- MERMAN, L.D. (1989). Childhood comorbidity of anxiety/affective disorders and behavior dis- orders. Journal of the American Academy of Child and Adolescent Psychiatry, 25, 707-713. _— 43 — enduring effects on people’s lives‘*9), and that some factors are likely to reflect an antisocial potential ‘°) while others may be caused by circum- stances outside the person own control (s0) This study uses the prospective longitudinal data collected in the Cambridge Study in Delinquent Development (CSDD), and attempts to concentrate specifically on early predictors of adult offending onset. THE CAMBRIDGE STUDY IN DELINQUENT DEVELOPMENT (CSDD) The CSDD is a prospective longitudinal survey of the development of antisocial and offending behavior of 411 South London boys, mostly born in 1953. It aims to describe the development of criminal behavior of inner-city males, to investigate how 1it could be predicted in advance, to explain why juveniles begin to commit delinquent acts, why they con- tinue or discontinue offending into adulthood, and why offenders tend to desist from crime once they reached their mid-twenties ©". The Study began in 1961. The males in the sample have been followed up from when they were 8 to age 48. The majority of the boys (n = 399) were chosen by taking all the boys who were attending 6 state primary schools within a one-mile radius of the research office established for the purpose of this study. Twelve other boys were drawn from a local school for the educationally subnormal with the aim of attempting to make the sample more representative of the male population of the area. Most of — the boys were of British origin and White in appearance (n — 357, or 87%), and were brought up by parents who themselves grew up in Eng- land, Wales or Scotland. Of the remaining 54 boys, 12 were Black, hav- ing a parent of West Indian or African origin; the other 42 boys were White and of non-British origin. “On the basis of their fathers’ occupa- (48) HUESMANN, R.L., EroN, L.D. & Dusow, E.F. (2002). Childhood predictors of adult criminality : are all risk factors reflected in childhood aggressiveness? Criminal Behaviour and Mental Health, 12, 185-208. OUIMET, M. & LE BLANC, M. (1996). The role of life experiences in the continuation of the adult criminal career. Criminal Behaviour and Mental Health, 6, 73-97. (49) FARRINGTON, D.P. (2005). The Integrated Cognitive Antisocial Potential (ICAP) the- ory. In D.P. FARRINGTON (Ed.). Integrated developmental and life-course theories of offending. (Advances in criminological theory, Vol. 14, pp. 73-92). New Brunswick, NJ : Transaction, p. 76. (60) FARRINGTON, D.P. (2003a). Advancing knowledge about the early prevention of adult antisocial behaviour. In D.P. FARRINGTON & J.W. Cormp (Eds.). Farly prevention of adult anti- social behaviour (pp. 1-31). Cambridge, UK : Cambridge University Press. ($1) FARRINGTON, D.P., Coip, J.W., HARNETT, L., JOLLIFFE, D., SOTERIOU, N., TURNER, R. and 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. London : Home Office (Home Office Research Study No. 299). __ 44 — tion when they were aged 8, 94% of the boys could be described as work- ing class (categories III, IV or V on the Register General’s scale, describ- ing skilled, semiskilled, or unskilled manual workers), in comparison with » (52) the national figure of 78% at that time . Almost all boys were living in traditional two-parent families with both a father and a mother figure. THE PRESENT STUDY The present study focusses its attention on late starters, individuals who begin their official criminal career (i.e. the first conviction) at age 21 or later. The primary purpose is to explore the early precursors of late onset offending. Firstly, late starters (LS) (n = 38), early starters (ES) (n = 54), medium starters (MS) (n = 75), and non-offenders (NO) (n = 236) were contrasted to analyse whether adult offenders were exposed in childhood and adoles- cence to factors that may have had a buffering effect against early delin- quent onset. According to official data, early starters were defined as those individuals with a criminal onset in childhood (between ages 10 and 14); medium starters were those individuals who began their criminal career dur- ing the adolescent years (between ages 15 and 20); and non-offenders were those who had no criminal record. The average age of first conviction for an early criminal career was 13.23, while for a late career it was 30.51. Secondly, the interest was to explore why adult offenders were con- victed, while non-offenders continued to remain unconvicted, and EMS offenders (ES and MS) were convicted both earlier and later in life. This 1s the first time that a prospective longitudinal study has been used to investigate early psychological and background correlates of late criminal onset. It 1s the first time that appropriate control groups of unconvicted individuals and offenders (ES and MS) are employed. Ît is the first time that childhood, adolescent and teenage risk factors are analysed, including a wide range of standardised measures derived from participants themselves, their parents, teachers, peers, psychologists, social workers, and official criminal records. HYPOTHESES 1. Adult criminal behavior can be predicted in advance. (32) FARRINGTON, D.P. (2003b). Key results fdrom the first forty years of the Cambridge Study in delinquent development. In T.P. THORNBERRY & M.D. KROHN (Eds.), Taking stock of delinquency : An overview of findings from contemporary longitudinal studies (pp. 137-183). New York : Kluwer Academic/Plenum Publishers, pp. 138-139. — 45 — 2. There are childhood risk factors that play a significant role in delay- ing offending until adulthood. 3. Adult onset and early onset criminal behavior differ in the constella- tion of risk factors that are present in childhood and adolescence, making adult onsetters more similar to non-offenders earlier in their life. MEASUREMENT For the present research, each variable was dichotomised into the worst quarter of boys (e.g. the quarter with the poorest housing) versus the remainder. The rationale behind this procedure was to compare the importance of different variables, and also to encourage a risk factor approach, which helps in targeting intervention efforts (53). Dichotomous variables do not necessarily lessen the measured strength of association, and do not contain inherently less information than other types of measurement. They actually equate the sensitivity of measurement of all variables, and make it possible to compare the pre- dictive strengths of explanatory variables “*. PROCEDURE The most important risk factors, assessed in childhood (ages 8-10), adolescence (ages 12-14), and the teenage years (ages 16-18), were first examined separately for each age range, and for the comparisons between groups. Their significance as predictors has been supported across numerous studies °°’. It was then possible to explore the extent 63) FARRINGTON, D.P. (2006a). Comparing football hooligans and violent offenders : Childhood, adolescent, teenage and adult features. Monatsschrift für Kriminologie und Straf- rechtsreform, 3, 193-205. (4) FARRINGTON, D.P., & LOEBER, R. (2000). Some benefits of dichotomization in psychi- atric and criminological research. Criminal Behaviour and Mental Health, 10, 100-122. (°5) KEENAN, K., & SHAw, D.S. (2003). Starting at the beginning. Exploring the etiology of antisocial behavior in the first years of life. In B.B. Laxey, T.E. MOFFITT, & A. CASPI. (Eds.). Causes of conduct disorder and juvenile delinquency (pp. 153-181). New York : The Guilford Press. LOEBER, R. (1990). Development and risk factors of juvenile antisocial behavior and delinquency. Clinical Psychology Review, 10, 1-41. LOEBER, R., SLOT, N.W., & STOUTHAMER-LOEBER, M. (2006). À three-dimensional, cumu- lative developmental model of serious delinquency. In P.-O. WIKSTRÔM & R.J. SAMPSON (Eds.). The explanation of crime: Context, mechanism and development (pp. 153-194). Cambridge : Cambridge University Press. MOFFITT, T.E., & CasPI, À. (2003). Childhood predictors differentiate life-course persist- ent and adolescent-limited pathways, among males and females. Development and Psychopa- thology, 13, 355-375. __ 46 — to which risk factors were associated with criminal onset at different stages of individual development. The analysis proceeded in two phases. First, odds ratio (OR) analysis was employed to identify which predictors had significant influences on participants’ criminal careers. The OR provides information about the existence, direction, and strength of an association between two (target and comparison) groups regarding the likelihood of an event occurring6°. It is not influenced by the prevalence of risk factors or offending. An OR of 2.0 or greater, roughly indicating a doubling of the risk, shows a strong relationship. Second, the variables, which in the previous analysis were identi- filed as potential explanatory factors in the aetiology of late adult offending, were subjected to a series of logistic regression tests. The aim of logistic regression 1s to build multi-variable explanatory mod- els that best account for observed variation in the outcome variable. To identify which of the risk factors was a significant independent predictor of criminal onset, the variables were entered into a forward stepwise logistic regression model, and only those predictors with a significant score were retained (the cut-off point for significance being p = 0.05). RESULTS Childhood features In line with previous scientific literature 67 individual and family factors mainly predicted differences among criminal onsets in childhood. Table 1 shows childhood features of criminal onset for LS, ES, and MS. It was found that late starters were significantly different from the other groups in some aspects of their childhood lives. For example, 52.6% of late starters, 48% of medium starters, 53.7% of early starters, and 27.5% of non-offenders came from poor housing (e.g. neglected accommodation, poor interior, etc.). Compared with unconvicted males, late starters were significantly more likely to live in these deprived home conditions (OR = 2.9; CI = 1.5-5.9). Unsurpris- ingly, large family size (i.e. boys with five or more siblings) and coming from a disrupted family (i.e. both permanent or temporary separation from a parent up to age 10) were significant predictors of future crimi- (36) FARRINGTON & LOEBER, (2000), op. cit. (°7) LOEBER, R.. & FARRINGTON, D.P. (Eds.) (2001). Child delinquents. Thousand Oaks CA : Sage Publications. _— 47 — nality °8), Respectively 31.6% of late starters and 16.1% of non-offenders came from a large family (OR = 2.4, CI = 1.1 - 5.2), and 28.9% of late starters and 15.3% of non-offenders lived in a disrupted family (OR = 2.3, CI = 1.03 - 5.0). 48.5% of late starters and 30.4% of non-offenders had siblings with behavioural problems (OR = 2.2, CI = 1.03 - 4.5). Late starters were not significantly different from EMS offenders on these factors. Verbal and non-verbal IQ (intelligence), low junior attainment, lack of concentration, and nervousness had the most significant impact on future offending. Late starters had a lower verbal IQ (OR = 2.1, CI = 1.01 - 4.47) and non-verbal IQ (OR = 2.6, CI = 1.3 - 5.5) in comparison with non-offenders. They also were more likely to have low junior attainment (OR = 2.7, CI = 1.2 - 6.0) and lacked concentration (OR = 2.3, CI = 1.01 - 5.1). LS offenders were also more likely than non- offenders to be physically neglected (OR = 3.1, CI = 1.1 - 8.6), impulsive (OR = 2.1, CI = 1.0 - 4.4), and to be rated as dishonest by peers (OR = 2.2, CI = 1.0 - 5.1) by age 10. Interestingly, the risk factor of being nervous or withdrawn was quite exceptional in being associated with a decreased likelihood of early delinquency; those individuals who were rated as particularly nervous were more likely to become late starters. 38.9% of late starters and 25.2% of non-offenders were rated as nervous (OR = 1.9, CI = .91 - 4.0). Despite the finding not being significant (on a two-tailed test), this predictor deserves further investigation. When late starters were com- pared with EMS offenders, being nervous was significantly associated with adult onset (OR = .34, CI = .15 - .76); hence, a delay in the onset of a criminal career was more likely to occur when the individual was nervous, withdrawn, or inhibited. Adolescent features Table 2 shows adolescent (age 12-14) features of the groups. For example, when late starters were compared with non-offenders, the sig- nificant predictors were high anxiety, low verbal IQ, and early school leaving. Late starters were significantly more likely to be highly anxious ado- lescents compared with non-offenders (OR = 3.1, CI = 1.2 - 8.1). It was more likely that late starters had a lower verbal IQ (OR = 2.1, CI = .94 - 4.5) and left school earlier (OR = 2.3, CI = 1.1 - 4.9) than non-offend- 68) MURRAY, J., & FARRINGTON, D.P. (2005) Parental imprisonment : Effects on boys' antisocial behavior and delinquency through the life-course. Journal of Child Psychology and Psychiatry, 46, 1269-1278. __ 48 — ers. No significant differences relating to these risk predictors were found between late starters and EMS offenders. Teenage features In the teenage years (age 16-18), the most important individual dif- ference between late starters and non-offenders was that the late starters had a high level of neuroticism. 40% of late starters were highly neurotic in comparison with 22% of non-offenders (OR = 2.4, CI = 1.2 - 4.9). Late starters were also more likely than non-offenders to express anti- establishment attitudes (OR = 2.3, CI = 1.1 - 4.9) and to commit motor- ing offences (OR = 2.7, CI = 1.1 - 6.6). Late starters and EMS offenders were not significantly different in these predictors. REGRESSION ANALYSES The forward stepwise logistic regression employed in this study makes it possible to examine whether an underlying factor (e.g. poor housing), which predicts the outcome of interest (e.g. criminal onset), is statistically significant once other underlying factors are controlled. Table 4 shows the variables that were selected in the model in order of their strength of prediction (i.e. their contribution to the predictive power of the model), the change in the likelihood ratio chi-squared (LRCS), and the partial odds ratio (OR). Predictors of late starters versus non-offenders The most important age 8-10 predictors of late starters were poor housing and low non-verbal IQ. At age 12-14 high anxiety (rated by teachers) and early school leaving were independent predictors of late offending. The most important age 16-18 predictors were high neuroti- cism (measured by the Eysenck Personality Inventory), an anti-establish- ment attitude, and motoring offences. Of all these predictors, poor hous- ing, high anxiety, an anti-establishment attitude and high neuroticism were the most important. DISCUSSION This study used prospective data from the Cambridge Study in Delinquent Development to investigate adult criminal onset, and the extent to which early adult criminality could be predicted. Differently — 49 _ from other adult onset studies ©°), this is the first attempt to identify early predictors of late onset. Before anyone was convicted, the future delinquents differed from the non-delinquents in several respects. Ât age 8-10, late onset offenders were more likely than unconvicted males to be nervous or withdrawn, impulsive, lacking in concentration and restless. They also had lower verbal and non-verbal intelligence. They came from disrupted families and had siblings with behavioral problems. At age 12-14, late starters were different from unconvicted males being more anxious, with a lower IQ and an earlier age of leaving school. At age 16-18 the main differ- ences between the two groups related to the late starters’ high neuroti- cism, anti-establishment attitude, and motoring offences. Late onset offenders were also markedly different from EMS offend- ers from childhood, in their level of nervousness, neuroticism and anx- iety. Those individuals who were most nervous and inhibited were, early in their life, less likely to get involved with others in daring and risk tak- ing behavior. These may be important reasons why their antisocial onset was delayed until age 21 or later, when the early protective influence of nervousness Was overturned by more reactive and externalising responses to life circumstances. In line with other studies‘6®, these findings show that psychological variables such as nervousness, anxiety, and neuroticism have a mitigat- ing or a negative protective impact upon early antisocial or deviant influences. More research 1s needed on why the protective effects of anx- iety/nervousness wear off in adulthood. Chung and colleagues ‘ (2002) found that those individuals who at ages 10 to 12, manifested symptoms of anxiety and depression, and aggression, were more likely to be late starters. Family, peer and school experiences during late childhood did not discriminate between late starters and non-offenders, probably because living in a neighbourhood at high risk for drugs, where these late onsetters resided, may have contributed in some degree to this later divergence in offense trajectories ‘6°). (59) EGGLESTON & LAUB, (2002), op. cit.; GOMEZ-SMITH & PIQUERO, (2005), op. cit. (60) KERR, M., TREMBLAY, R.E., PAGANI, L., & VITARO, F. (1997). Boys' behavioral inhi- bition and the risk of later delinquency. Archives of General Psychiatry, 54, 809-816. MoOFFITT, T., CASPI, À., HARRINGTON, H., & MILNE, B.J. (2002). Males on the life-course- persistent and adolescence-limited antisocial pathways : Follow-up at age 26 years. Develop- mental and Psychopathology, 14, 179-207. OLLENDICK, et al. (1999). op. cit. (61) CHuno, 1.-J., HirL, K.G., HAWKINs, J.D., GILCHRIST, L.D., & NAGIN, D.S. (2002). Childhood predictors of offense trajectories. Journal of Research in Crime and Delinquency, 39, 60-90. (62) CHUNG, et al. (2002), op. cit., p. 82. Kerr and colleagues‘*) showed that behavioural inhibition protected both disruptive and non-disruptive boys against involvement in delin- quency. Their account of inhibited children, in contrast to the uninhib- ited, resembles closely the description of anxious and nervous boys in our data. Behavioral inhibition is the tendency to display signs of wea- riness and react fearfully to strange people, novel social stimuli and sit- uations‘*), or to the threat of punishment or danger . Inhibited or nervous children tend to be more prone to avoid those conditions per- ceived as risky, hazardous, confrontational, and dangerous. This shield, in the short term, may defend them from delinquent peers and attrac- tions, but in the long run, when faced with more life pressures and demands that come along with the experience of growing up, it may lose its protective power. In contrast, Eggleston and Laub“° in their endeavour to explore adult criminal onset, reached the conclusion that “late onset offenders may not need to be analysed separately from persistent offenders. [.….] and that informal social bonds in adulthood explain variation in crime above and beyond criminal propensity”. In other words, they suggested that the predictors of adult offending were similar for all adult offend- ers, independently of past delinquency (persisters and adult offenders), which implies that the correlates of adult onset were the same for those who continue from juvenile to adult offending. Their statistically signif- icant variables were adult ones, such as employment stability, adult offending peers and seriousness of peer offending, drugs use, and so forth, which “seem logical since proximal factors should most immedi- ately influence a person’s life and decisions”. Tackling crime successfully will depend on how early it is addressed, and given the human and financial costs of adult criminality, policy makers cannot certainly just wait until individuals become adult offend- ers before intervening‘6”). Looking backwards from adulthood, attention should be directed towards the early roots of adult vulnerability ‘ Many adult onsetters in this study had already been subject to early stress, anxiety, and difficulties that seem to have constituted some lia- (63) KERR, M., et al. (1997). op. cit. (64) KAGAN, J., REZNICK J.S., & SNIDMAN, N. (1988). Biological basis of childhood shy- ness. Science, 240, 167-171. (65) Gray, J.A. (1990). Brain systems that mediate both emotion and cognition. Cognitive Emotion, 4, 269-288. (66) EGGLESTON & LAUB, 2002, op. cit. pp. 611-614. (67) SHERMAN, L.W., FARRINGTON, D.P., WeLsH, B.C., & MACKENZIE, D.L. (Eds.) (2002). Evidence-based crime prevention. London, UK : Routledge. (68) MAUGHAN, B. & KIM-COHEN, J. (2005). Continuities between childhood and adult life. British Journal of Psychiatry, 187, 301-303. FARRINGTON, D.P. (2000). Psychosocial predictors of adult antisocial personality and adult convictions. Behavioral Sciences and the Law, 18, 602-622. __ 51 — bility to, however late, criminal onset. Obviously, as Rothbart and Bates ‘°°) suggested, subsequent developmental experiences were required to transform these features into risk factors. It would follow that, by being aware of significant early predictors and risk factors for future offending, it may be possible to work successfully to divert people from a criminal career. Âs is often the case, this present study is not without limitations. It focussed attention on a sample of working-class men from South Lon- don. It could be interesting to examine whether and to what extent a late onset is likely to occur in middle or upper class environments; in different geographical contexts; in different types of criminal offending. It could also be particularly challenging for future empirical work to investigate the influence of gender on different criminal onsets (70) Future research should also control for self-reported delinquency (SRD) involvement. In this study SRD data measured at ages 12-14 and 16-18 suggested that late starters were not active offenders. Only 12.8% of non-offenders and 15.8% of late starters self-reported antisocial acts at age 12-14. At age 16-18, self-reported delinquency was admitted by 11.6% of non-offenders and 21.6% of late starters. To the extent that juvenile involvement in antisocial minor acts seems quite normative dur- ing the adolescent years‘"), we decided not to remove from the analysis those few cases of late-offenders and of non-offenders who self-reported (72) adolescent delinquency, so as to have a more representative sample of the offending and non-offending population. In accordance with the scope of prospective research, it may be use- ful to delve into the neuropsychology of antisocial behavior, and to explore the neural basis of social behavior at different points of devel- opment, so as to provide some insights into the influence of human brain structures and functions on late onset offending. While much work still remains to be done, and much more about human behavior needs to be learned, nevertheless the present research shows the extent to which adult offending can be predicted. We hope (69) ROTHBART, M.K., & BaTes, J.E. (1998). Temperament. In W. DAMON & N. EISENBERG (Eds.), Handbook of Child Psychology. Social, Emotional and Personality Development (vol. 3, pp.105-176). NewYork : JohnWiley. (0) KRATZER, L., & HopGINs, S. (1999). A typology of offenders : A test of Moffitt’s the- ory among males and females from childhood to age 30. Criminal Behaviour and Mental Health, 9, 57-73. (D MOFFITT, T.E. (1993). Adolescence-limited and life-course-persistent antisocial behaviour : À development taxonomy. Psychological Review, 100, 674-701. (2) MoFFITT, T.E. (2006a). À review of research on the taxonomy of life-course persistent versus adolescent-limited antisocial behavior. In F.T. CULLEN, J.P. WRIGHT, & K.R. BLEVINS (Eds.). Taking stock : The status of criminological theory (Advances in criminological theory, Vol. 15, pp. 277-311). New Brunswick, NJ : Transaction. __ 52 — that it represents the beginning of a multidisciplinary field of investiga- tion that may provide scientists, governments, and practitioners with a better understanding of the causal mechanisms linking psychological risk factors and delayed criminal behavior. TABLE 1 : Childhood predictors Odds Age Groups 8-10 % of: Ratio Groups NO LS MS ES LS/INO (236) (38) (75) (54) Predictors Nervousness 25.2 38.9 16.9 18.8 1.9 behaviour 30.4 48.5 49.3 51.2 2.2* Sibling problems Physical neglect 6.1 16.7 18.9 27.5 3.1* Poor housing 27.5 52.6 48.0 53.7 2.9* Large family size 16.1 31.6 34.7 40.7 2.4* Disrupted family 15.3 28.9 30.7 37.0 2.3* Low non-verbal IQ 18.2 36.8 26.7 46.3 2.6* 19.7 34.2 32.4 33.3 2.1* Low verbal IQ Low attainment 15.2 32.4 32.4 44.0 2.7* Lacks concentration 14.0 27.0 29.3 29.6 2.3* Impulsive 19.9 34.2 24.0 46.3 2.1* Dishonest 17.8 32.3 35.3 40.5 2.2* * 95% confidence interval (CI) does not include 1. Note : Some numbers on individual outcomes might be lower than the total n because of missing cases. NO = non-offenders; LS = late starters; MS = medium starters; ES = early starters. __ 53 — TABLE 2 : Adolescent predictors Age Groups Odds 12-14 % of: Ratio NO LS MS ES LS/NO Groups (236) (38) (75) (54) Predictors High anxiety 6.8 18.4 14.7 7.4 3.1* 16.5 28.9 29.7 37.7 2.1* Low verbal IQ Early school leaving 51.3 711 69.3 87.0 2.3* * 95% confidence interval (CI) does not include 1. Note : Some numbers on individual outcomes might be lower than the total n because of missing cases. NO = non-offenders; LS = late starters, MS = medium starters; ES = early starters. TABLE 3 : Teenage predictors Age Groups Odds 16-18 % of: Ratio Groups NO LS MS ES LS/NO (236) (38) (75) (54) Predictors High neuroticism 21.5 39.5 35.1 26.4 2.4* Anti-establishment 17.3 32.4 40.3 32.7 2.3* 9.3 21.6 26.4 28.8 2.7* Motoring offences * 95% confidence interval (CI) does not include 1 Note : Some numbers on individual outcomes might be lower than the total n because of missing cases. NO = non-offenders; LS = late starters; MS = medium starters; ES = early starters. __ 54 — TABLE 4 : Logistic Regression Analyses : Late Starters (1) vs. Non-offenders (0) LRCS Partial Age Predictors OR Change* 8-10 Poor housing 8.98 003 2.73 005 Low non-verbal IQ 4.81 028 2.38 024 12-14 4.67 031 2.48 072 High anxiety Early school leaving 3.63 057 2.06 063 16-18 High neuroticism 4.27 039 2.10 052 Anti-establishment attitude 3.39 066 2.13 058 8-18 Poor housing 8-10 9.83 002 3.22 002 High anxiety 12-14 4.97 026 2.90 047 Anti-establishment attitude 18 3.70 055 2.08 080 High neuroticism 16 2.91 088 2.01 082 Note : Forward stepwise analyses using likelihood ratio method. LRCS = Likelihood Ratio Chi-Squared. OR = Odds Ratio. * When predictor added to equation. ABSTRACT In the Cambridge Study in Delinquent Development, 411 South Lon- don males have been followed up from age 8 to age 48. The main aim of this article is to investigate the predictors of late onset offenders : those first convicted after the 215 birthday. The most important predic- tors were poor housing and low non-verbal intelligence at age 8-10; high anxiety and early school leaving at age 12-14; and high neuroticism and an anti-establishment attitude at age 16-18. It is concluded that high anxiety and nervousness may have protected children from offending at an early age, because they did not get involved with other children in daring or risk-taking activities. However, the protective effects wore off in adulthood. Low intelligence and attainment were also important pre- dictors of late onset offenders. __ 55 — RÉSUMÉ Dans l’étude du développement délinquant de Cambridge, 411 hom- mes du sud de Londres ont été suivis de l’âge de 8 ans à celui de 48. L’objet essentiel du présent article est d’examiner les prédicteurs de délinquance tardive, menant à une première condamnation après leur 21e anniversaire. Les principaux prédicteurs ont été un habitat pauvre et une faible intelligence non verbale à l’âge de 8 à 10 ans; une anxiété élevée et un abandon précoce de la scolarité entre 12 et 14 ans; et une forte tendance névrotique et une attitude d’hostilité à l’ordre établi à 16 à 18 ans. La conclusion est qu’une anxiété et une nervosité élevées ont pu pro- téger les enfants de la délinquance dans un premier temps, parce qu'’ils n’étaient pas impliqués avec d'autres enfants dans des activités osées et risquées, mais qu’à l’âge adulte ces effets protecteurs ont disparu. Une faible intelligence et de faibles accomplissements représentent aussi des prédicteurs importants de délinquance d’apparition tardive. RESUMEN En el estudio del desarrollo del delincuente de Cambridge, se han seguido a 411 hombres del sur de Londres, de edades comprendidas entre los 8 y los 48 años. El objeto principal del presente articulo es exa- minar los predictores de delincuencia tardia, con una primera condena después de su 21° cumpleaños. Los principales predictores han sido un hâbitat pobre y una inteli- gencia no verbal débil a la edad de 8 a 10 años; una elevada ansiedad y un abandono precoz de la escolaridad entre los 12 y los 14 años; y una fuerte tendencia neurôtica y una actitud de hostilidad al orden esta- blecido entre los 16 y los 18 años. La conclusiôn es que una ansiedad y un nerviosismo elevados han podido proteger a los niños de la delincuencia en un primer momento, porque no se implicaban con otros niños en actividades osadas y arries- gadas, pero a la edad adulta estos efectos protectores han desaparecido. Una débil inteligencia y un escaso cumplimiento de objetivos son tam- bién predictores importantes de la delincuencia de apariciôn tardia. __ 56 — LONG-TERM EFFECTS OF INVESTMENT IN INFANT HEALTH CARE ON HOMICIDE IN ITALY Uberto Gatti ( Hans M.A. Schadee ‘”) Richard E. Tremblay ”” INTRODUCTION Criminal phenomena often arouse immediate reactions. Public opin- ion may be alarmed by serious criminal acts, which are often sensation- alised by the mass media and sometimes exploited for political ends. In reality, however, measures and policies adopted in the wake of emo- tional reactions risk being irrational and are unlikely to succeed in reducing crime or protecting citizens against the consequences of crimi- nal behaviour. In assessing the efficacy of political and social strategies aimed at reducing crime, it is frequently overlooked that the effects of such strat- egies are often felt only years later, on a time-scale that goes beyond that of political consensus-building or the electoral advantages hoped for by the policy-makers. The present study aims to evaluate the long- term effects of investments in health and social services for children, and in particular the possible preventive effects of such investments on vio- lent crime. « Investments in welfare” are those which loosen the bonds between the citizen and the labour market, and which compensate for financial disadvantages, thereby making health and educational services available to those people who would otherwise be unable to afford them. The relationships between such investments and crime rates have generally been examined with synchronous data. Messner and Rosenfeld (1997), for example, showed that investiments in welfare were able to reduce the anomic pressure and, consequently, homicide rates in the various nations considered. In an investigation involving several countries, Pampel and Gartner (1995) examined the effect of demographic changes — particularly (*) University ot Genoa, Italy. (**) University of Milano-Bicocca, Italy. (***) GRIP-University of Montreal, Canada. __ 57 — increases in the number of young people — on homicide rates. These authors found that, in countries where institutions for collective social protection were weak, an increase in the number of young people led to a rise in the homicide rate, while no such effect was recorded in coun- tries where these institutions were strong. On comparing homicide rates in various countries, Savolaien (2000) found that the effect of economic inequality on this type of crime was non-existent in countries with high welfare investments. In this regard, Fiala and LaFree (1988) conducted an interesting study relating rates of child abuse and child murder in several countries to a range of social and economic conditions, in order to ascertain which social characteris- tics are linked to higher rates of such crimes. These authors found that one of the most significant predictors of child murder was the propor- tion of mothers who worked because of economic necessity, while child- murder rates were lower in countries where more women were involved in higher education and professional work. However, the large propor- tion of working women did not exert a negative effect in countries with a highly developed system of social support for mothers and families: indeed, such support reduced the number of child murders. For instance, in Sweden and Denmark — countries which invest heavily in assistance for mothers in difficulty, poor families and children — child murder rates were very low, even though the percentage of working women was high. In an analysis of imprisonment rates in the various states of the USA, Beckett and Western (2001) discovered that incarceration was more frequent in states where the welfare system was weaker. These authors conducted a multiple regression analysis to ascertain whether investments in welfare (social assistance, unemployment benefits, etc) diminished the detention rate in the state, while controlling for the main socio-demographic variables. Their results demonstrated that the nega- tive relation between welfare and imprisonment had increased over the years, becoming significant in the 1990s. De Fronzo (1983,1996, 1997) conducted a series of studies on the relationship between the financial level of welfare aid and crime rates. In an early (1983) study of 39 US metropolitan areas, this author found that, after controlling for other social and economic factors, the level of public assistance to poor families correlated negatively with the rates of various crimes, including homicide, sexual assault and burglary. The author therefore claimed that reducing public assistance would lead to a rise in the rates of these crimes. In a subsequent study, De Fronzo (1996) examined the effect of the mean level of social assistance, per per- son, in 141 US cities, on burglary rates, while controlling for a whole range of socio-economic variables. Only three variables proved to be sig- __ 58 — nificant predictors of burglary rates : the size of the population, the per- centage of families with solitary mothers, and the mean financial level of social aid. While the first two variables correlated positively with bur- glary rates, welfare expenditure correlated negatively : the higher the level of AFDC (Aid to Families with Dependent Children), the fewer the burglaries in the city. Using the same sample of cities, De Fronzo (1997) also found that the level of social assistance (AFDC) had a direct negative impact on homicide rates; moreover, it exerted an indirect neg- ative effect on these rates through its association with family status (by reducing the number of families headed by a single mother). The high level of economic support therefore reduced homicide rates through two different causal mechanisms. By means of a time-series study over the period 1948-1985 in the whole of the USA, Devine et al. (1988) sought to discover whether cer- tain socio-economic factors (unemployment, inflation), some forms of social control (number of prison inmates) or the amount of aid to the poor influenced the rates of homicide, robbery and burglary. Their results showed that, when controlling for changes in socio-economic fac- tors, social control and criminal opportunity, variations in the amounts allocated to social assistance were negatively associated with changes in burglary rates and marginally negatively associated with robbery rates, but did not correlate significantly with homicide rates. AIl of the above results are enlightening and fall within the frame- work of the broader question of the effects of welfare intervention that has been debated in recent decades (Patterson, 1994). Indeed, such intervention is regarded by some as being counter-productive, foster- ing dependence and passivity, stifling effort and promoting marginal- isation (Murray, 1984; Mead, 1992); others, however, regard it as an indispensable tool for reducing the deprivation of those who belong to the marginal sectors of society, and, in the final analysis, for improv- ing their social adaptation (Collins and Haber, 1997; Kenworthy, 1999). Awareness of the effects of welfare intervention acquired particular importance in the USA, when in 1996, under the Clinton administra- tion, a radical change in welfare policy took place. In that year, the Aid to Families with Dependent Children programme was abolished, and was replaced by the more restrictive Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), which limited the period of eligibility for aid and imposed prescriptions in the field of work. Às a result, the number of families on assistance, which had peaked at 5 mil- lion in 1994, had been halved by June 1999 (Ku, 2001) and further declined to about 2 million in 2001. According to The Economist (2001), the success of this reform went beyond the most optimistic forecasts, __ 59 — while other commentators judged it to be a substantial failure (Jenks and Swingle, 2000; Burke, 2001; Cancian, 2001). Other experts, how- ever, viewed the results of the reform as partly positive and partly neg- ative. Zaslow et al. (1998), for instance, see the results as positive in those cases in which mothers actually succeed in finding a job and increasing family income, but as potentially negative when mothers are unable to find or keep a job once the maximum permitted time on assistance has been reached. Fuller et al. (2002) monitored a sample of 948 mothers with small children following their entry into the new wel- fare programmes, and concluded that many mothers actually did find employment, though their earnings were rather limited and their jobs somewhat irregular. In spite of working, many mothers remained in dif- ficult financial circumstances, and one mother out of eight claimed that the family sometimes had to skip meals or eat less for lack of money. Moreover, the mothers spent less time with their children and took less care of them, and a rather high percentage of mothers suffered from depression. The results of this study, which was conducted by research- ers from the Universities of Berkeley, Stanford, Columbia and Yale, indicate that the 1996 reform did not achieve the objectives set, and that inducing mothers to work does not, in itself, have a positive effect on the development of their children. Having analysed the principal evaluation studies, Morris (2002) claims that the new welfare structure may be advantageous for children but potentially harmful to adolescents; since the reform, smoking, alco- hol use, small-time delinquency and maladjustment at school appear to have increased among teenagers, who receive less supervision from their families. À particular negative effect of the reform has been highlighted by Paxon and Waldfogel (1999), who claim that, as a result of reduced aid to families in difficulty, cases of child abuse have increased (Geen et al. 2000). The heated debate and the research efforts aimed at assessing the results of aid programmes emphasize the importance of welfare systems in the politics of a nation, especially with regard to their repercussions on families and child development. It is important to understand which system is best able to provide health and educational support for child development, since poverty, deprivation and abandonment may have extremely harmful effects both on the children themselves and on society in general. From a criminological point of view, it has been hypothesised (Vila, 1994) that all of the factors which enhance child development, such as healthcare, education and parenting, have a long-term effect, and are able to reduce crime a generation later (“nurturance” hypothesis). At the same time, all else being equal, factors that hinder the development — 60 — of the individual in early life will exacerbate crime in the long term at the level of the general population (Vila, 1994). To test this hypothesis, Savage and Vila (1997a) carried out analyses, with countries as units, aimed at ascertaining whether indicators of child welfare were able to predict crime levels 20 years later. In their first analysis, which involved a limited number of countries (N varied from a minimum of 13 to a maximum of 55, depending on the analysis) the authors found that — after controlling for a whole range of variables such as economic growth, criminal opportunity and the percentage of unsolved crimes — the index of child welfare, as measured in 1960, cor- related negatively with crime rates measured in 1980. These findings therefore confirmed the “nurturance” hypothesis. In a subsequent anal- ysis conducted on a larger sample of countries (N between 31 and 86, depending on the analysis), this hypothesis was not confirmed for crime in general, but only for crimes of violence. In a further study, again car- ried out with a sample of countries, Savage and Vila (1997b) examined the relationships between changes in childcare services and changes in crime rates recorded in a later period. The dependent variables in the various regression analyses consisted of changes in the rates of homi- cide, violent crime, crime in general and theft over a 10-year period (from 1980 to 1990). The independent variables consisted of indicators reflecting variations in children’s health conditions, such as the infant mortality rate and mean life expectancy, over a period of 10 years (from 1960 to 1970). In addition, changes in the educational sphere, such as primary school enrolments (from 1965 to 1975) and teacher-pupil ratios (from 1970 to 1980), were taken into account. The choice of the 10-year period within which changes were observed in the health, social and scholastic spheres was determined by the time required for such changes to have an effect on children who might later commit crimes as adults. Thus, variations in infant mortality rate, which regard early childhood, were recorded 20 years before the period in which changes in crime rates were observed, while changes in educational patterns were observed 10 years before. Âs covariates, the authors considered measures of economic growth and inequality, factors that reflected the opportunities to commit crimes (changes in the number of television sets and automobiles per capita, etc) and demographic factors. Initially, a series of simple regression analyses was carried out, relating each dependent variable to each index of change in childcare. Obviously, there was a risk that any statistical associations found might be spurious, and that changes in crime rates might be due to variations in economic growth, opportunities for crime, demographic factors, etc. To obviate this possible specification error (omission of variables that should have been included), the authors con- _— 61 — ducted a further series of analyses; this time, they added the 11 co-var- lates indicating change one by one and checked that the estimates of the child welfare parameters remained stable. The results of this second analysis confirmed the “nurturance” hypothesis and indicated 1) that improvements in primary school enrolments and in the number of hos- pital beds available per capita were negatively associated with changes in homicide rates; 2) that improvements in infant mortality rates, life expectancy at birth and the number of hospital beds were negatively associated with changes in the rates of violent crime, and 3) that improvements in life expectancy at birth were negatively associated with changes in the rates of theft. Thus these studies relating changes in time to one another confirmed that improvements in health and social care for children were able to reduce crime rates in the long term, all other conditions being equal. RESEARCH METHOD Hypothesis The present research aims to ascertain whether differences in the quality of health and social services for children, as recorded in the 1970s in the Italian provinces, influenced the rates of homicide 10 and 20 years later, while controlling for economic factors. This means inves- tigating whether a long-term preventive effect on homicide can be dis- cerned in those provinces where investment in childcare services was greatest. The underlying rationale 1s that better childcare may produce adolescents and young adults who are socially better adapted, and there- fore less likely to commit crimes. The study examines the relationships between infant mortality (in the first year of life) and per capita Gross Domestic Product on the one hand, and the rates of homicide over a period of two decades (from 1974 to 1997) on the other. The variables were measured at three time- points : in the middle of the 1970s (1974-1977), 1980s (1984-1987) and 1990s (1994-1997)‘), and are composed of aggregate data for the 95 Italian provinces. In this analysis infant mortality is interpreted in two ways. On the one hand, it is seen as a general (negative) indicator of the state of health and well-being of the child population; in this way, it enables us indirectly to assess the quality of assistance during pregnancy, childbirth (WIt should be noted that, given the relative stability of the phenomena observed, the data regarding the four years in each decade can be regarded as reflecting the situation exist- ing over a longer time-span, including a few years before and after the period considered. __ 62 — and during the first year of life in a given province. On the other hand, infant mortality is interpreted as an indicator of investment in services (health, social, etc) for children. Such services, however, may impinge not only upon early infancy, but also upon childhood in general; indeed, we may suppose that a reduction in infant mortality reflects a policy that is more oriented to children’s needs, not only in the field of health- care, but also in the social and educational spheres. That infant mortal- ity can be regarded as an indicator of all welfare investments is corrob- orated by the (negative) correlation seen between infant mortality and per capita welfare expenditure in the various municipalities. As no fig- ures for this expenditure are available at the provincial level, we consid- ered data collected at the regional level. These were taken from research by ISTAT (National Institute of Statistics), which carried out a detailed analysis of welfare expenditure, as calculated from the budgets of a large sample of Italian municipalities in 1995, 1997 and 1999 (Censis, 2000). We were therefore able to ascertain that infant mortality recorded in the various regions in 1995 was linked to per capita welfare expenditure. This (negative) correlation was highly significant (-.75; p < .001 for 1995). Infant mortality also had a negative correlation with the percent- age of the municipalities’ total expenditure that was allocated to welfare (-.67; p < .01 for 1995) . On the basis of the above observations, we formulated the following hypotheses : Hypothesis I. — The value of a given variable 1s explained by the value of the same variable regarding the previous period. This connec- tion does not appear to be particularly interesting, although it is impor- tant to assess how far a variable 1s explained by the same variable in the previous period. It may be assumed that the variance explained by the values recorded in the previous periods will be high for what concerns the Gross Domestic Product, since this variable partly reflects the phys- ical structure of the economy, while it will not be so high with regard to infant mortality, since this variable 1s more susceptible to change. As far as homicide is concerned, the explained variance should be fairly high, given the relative stability of the socio-cultural factors underlying this type of crime. Hypothesis 2. — In each of the three periods, there should be a neg- ative association between infant mortality and Gross Domestic Product, C The correlation between infant mortality and municipality spending on welfare is partly spurious, since both are related to the economic prosperity of the province. However, controlling for per capita GDP (per region, in 1995) there remains a correlation of -.24 between infant mortality and welfare expenditure 1f per capita expenditure is considered, and a correlation of -.38 if the percentage of the municipality's budget allocated to welfare 1s con- sidered. — 63 — in that the prosperity of the region should imply better healthcare. The correlation within a period means that the good economic conditions can bring about an improvement in healthcare for children in a rela- tively short time, thereby reducing infant mortality. Hypothesis 3. - On controlling for economic conditions, a higher infant mortality rate will lead to a higher homicide rate 20 years later. This involves testing the “nurturance” hypothesis, according to which an improvement in childcare will lead to a decrease in criminal behaviour among those young people who have been better looked after as chil- dren. The “nurturance”’ hypothesis suggests that investment in childcare at time tl will have an effect on crime rates at time t3, when the children become young adults, bearing in mind that this is the age-group most frequently involved in homicide. Thus, the higher the level of well-being of children is, the lower the level of homicide crime will be 20 years later. The effect of infant mortality (which is an expression of the quality of the services provided during maternity and early infancy) is therefore assessed 20 years later by analysing the data for four years in each dec- ade. It should be pointed out, however, that this effect involves a longer time-span than the four years considered, in that infant mortality in the period 1974-1977, for example, reflects health and social services for children in the first few years of life, and not only in the neonatal period. Clearly, the relationship between infant mortality and crime rates recorded 20 years later has to be controlled for the economic factor. It is well known that infant mortality has a negative correlation with the level of economic prosperity (Hales et al., 1999) of a given geographical area. The association between infant mortality and crime could there- fore be spurious, reflecting the well-established association between pov- erty and homicide at the territorial level. Evaluating the 10-year effect serves to check that any effect exerted by the independent variables on the dependent variables 20 years later is not the result of effects that take place within a shorter time-span, and which might not come to light if the relationships were only evaluated over a 20-year period. By introducing an intermediate period into the analysis (in this case referring to the 1980s), we can ascertain whether the 20-year relationship remains, in spite of the two measurements car- ried out 10 years apart, or whether it disappears as a result of the intro- duction of the two shorter periods. __ 64 — Variables. For each of the 95 Italian provinces, we recorded the following data regarding the three periods 1974-1977°), 1984-1987 and 1994-1997) : 1 The number of homicides reported, including murder, manslaughter and infanticide, per 100,000 of the population (Homi1); Per capita Gross Domestic Product, in millions of Italian lire, as an indicator of the economic well-being of the population (GDP). The data for the periods 1974-1977 and 1984-1987 were multiplied by the coefficients taken from the consumer price indexes for the families of manual and office workers in order to convert them into values corresponding to purchasing power in the period 1994- 1997. Infant mortality : deaths during the first year of life, per 1000 live births (Mort). AIl data were gathered from the appropriate ISTAT (National Insti- tute of Statistics) publications. For all three variables, we used the means of the data recorded in four years for each of the periods considered : 1970s (1974-1975-1976- 1977), 1980s (1984-1985-1986-1987) and 1990s (1994-1995-1996-1997). The numbers 7, 8, and 9 in the tables refer to the 1970s, 1980s and 1990s, respectively. STATISTICAL ANALYSIS AND RESULTS Table 1 reports the means of the four central years in each of the three decades for the different variables. As will be seen, the mean infant mortality rate falls steadily over a 20-year time-span, from 19.3 to 5.9 deaths in the first year of life, per 1000 live births, while the per capita Gross Domestic Product increases from 13.9 to 28.0 million lire (in terms of 1994-1997 purchasing power). The homicide rate remains stable from the 1970s to the 1980s, but increases between the 1980s and the 1990s. C) With regard to the provincial rates of homicide, we used the data for 1978 instead of 1977: for the 1977 data were not available in the ISTAT publication of giudicial statistics for that year. __ 65 — TABLEAU 1 : Mean and standard deviation (N=95) 1974-1977 1984-1987 1994-1997 Mean D.S. Mean D.S. Mean D.S. Infant mortality (0-1 y. 19.518 4.058 9.576 2.727 5.858 1.574 per 1000 live births) Gross Domestic Prod- uct per inhab. (millions 13.865 3.432 23.009 5.221 28.006 7.510 of lire) Reported homicides 1.676 1.361 1.675 1.754 2.402 2.320 per 100.000 inhab. (*) (*) With regard to the provincial rates of homicide, we used the data for 1978 instead of 1977; for the 1977 data were not available in the ISTAT publication of judicial statistics for that year. With regard to the relationships among the variables (Table 2), it will be seen that the correlation between Gross Domestic Product and infant mortality 1s always negative and significant. Moreover, the correlation between Gross Domestic Product and homicide is also negative. The correlations between infant mortality and homicide are always positive and significant. The correlations among the values of the same variable in different periods show that the strongest autocorrelations are for Gross Domestic Product, followed by those concerning homicide, and finally by the correlations regarding infant mortality. Tableau 2 : Correlations among the variables (N=95) Homi.74- Homi Homi. I. Mor. J Mor. l Mor. G.D.P. G.D.P G.D.P. 78 84-87 94-97 74-77 84-87 94-97 74-77 84-87 94-97 Homicide 1.000 74-78 (+) Homicide 783+#* 1.000 84-87 Homicide 664*** 793*** 1.000 94-977 Inf. Mort 572+#+ 541++* 651*** 1.000 74-77 Inf. Mort 412+++ 472*** 520*** 609*** 1.000 84-87 Inf. Mort 94-97 385*** 418*** 461*** 570*** 713*** 1.000 Gr.Do.Pro 74-77 -.492*** -.483*** 416*** -.622*** -.547++* -.621*** 11.000 _— 66 — Tableau 2 : Correlations among the variables (N=95) Homi.74 Homi Homi. Il Mor. L Mor. I Mor. G.D.P G.D.P. G.D.P 78 84-87 94-97 74-77 84-87 94-97 74-77 84-87 94-97 Gr.Do.Pro. 84-87 -.565*** 531*** -.491 *** -677*** -.592#** -.656*** 937*** 1.000 Gr.Do.Pro. 94-97 -492+++ -.523*** -.488*** -.678*** -.486*+* __656*** 872%** 895*** 1.000 *** p < 0.001 (two-tailed). ** p < 0.01 (two-tailed). * p < 0.06 (two-tailed). (+) With regard to the provincial rates of homicide, we used the data for 1978 instead of 1977; for the 1977 data were not available in the ISTAT publication of judicial statistics for that year. MODEL In order to evaluate the models suggested by the hypotheses, we used two programmes : MIM (Edwards, 2000) and LISREL (Jôreskog and Sorbom, 1988). As the hypotheses listed above do not specify all the connections among the variables, MIM was used to select the significant relation- ships and to exclude those that were either inexistent or had low signif- icance, through a process of systematic research (Edwards and Havränek, 1987). In this model, the exclusion threshold excluded all coefficients with a value of p>.005. The result was a model with a chi* value of 36.3 with 22 degrees of freedom (p = .028). The model thus found was subsequently examined in detail by means of the LISREL programme. Moreover, unlike in the model implicitly used by MIM , we gave a causal direction to the relations from Gross Domestic Product to infant mortality within the 1980s and 1990s, in conformity with our basic hypotheses. Examination of the relationships in this model suggested that the mort8 Æ GDP9 relationship could be eliminated; this resulted in the final model (Fig 1 and related estimates in Tab. 3), which has a chi” value of 40.5, with 23 degrees of freedom (p = .014). The difference between the two models in terms of chi* and degrees of freedom (chi” of 4.5 with 1 degree of freedom, p = .041) rep- resents the significance of the path coefficient that has been removed. In (9 Within each decade MIM uses a covariance selection model. also known as a Gaus- 2000. sian graphical model, in which the links between the variables are undirected. Cf. Edwards. — 67 — the model presented here, all the remaining coefficients are significant at a level of p < .01, apart from the relation between infant mortality in the 1970s and GDP in the 1990s, which 1s significant at a level of p < .05. However, before removing the relation mort& Æ GDP9, this rela- tion (mort7 Æ GDP9) was significant (p = .0044); for this reason, it has been kept in the model. If this were also to be removed, we would have a LISREL model with a chi” value of 45.1 with 24 degrees of freedom (p = .006), which would be barely acceptable. The estimates of the other parameters in that model change only marginally in comparison with the estimates reported here, and do not change at all in the part of the model that is of most interest here, that is to say, the part concerning the relations between infant mortality and homicide rates. The model reported here (Graph..1 and Tab. 3) concerns the rela- tionships among infant mortality, homicide and wealth in the Italian provinces in the three periods examined, and treats the data as the result of a complex process that develops over time. The regression parameters that link the various phenomena (e.g. the fact that a lower infant mor- tality rate in the 1970s leads to a lower rate of homicide in the 1990s) are interpreted as support for the hypotheses formulated earlier. In Fig- ure 1, we can observe both the relationships among the variables within the same period and the relationships among the variables among dif- ferent periods separated by 10 and 20 years. Graph. 1 Causal model for homicide (N=95) SEVENTIES EIGHTIES NINETIES 78 2 HOMICIDE 7 22 HOMICIDE 8 2> HOMICIDE 9 57 /31 INF.MORT.7 | -— 38 INF.MORT. 8 52 INF.MORT. 9 62 13 41 15\33 G.D‘.l".…7 | 84 80 Chi” = 40.45 D.F. = 23 P =.014 __ 68 — TABLEAU 3 : Parameter estimates of de causal model of homicide REGRESSION EQUATIONS anni 74-78 anni 84-87 anni Dependent 94-97 Variables GDP9 Homi7 L Mort7 , GDP7 Homi& AMn8Ï GDP8 Within seventies Homi7 = 57*m7 33 (.08) Mort7 = -.62*g7 39 (.08) Within eighties and between seventies and eighties Homi8 = .78*h7 61 (.06) Mort8 = .38*m7 43 -.33*28 (.11) (.11) GDP8 = - 15*m7 84*g7 89 (.04) (.04) within nineties and between seventies, eighties and nineties Homi19 = -31*m7 .62*h8 69 (.06) (.06) Mort9 = .52*m8 - 41*g9 64 (.08) (.08) GDP9 = - 13*m7 81 80*g8 (.06) (.06) Legend : Mort(m)= infant mortality; GDP(g)= Gross Domestic Product; Homi(h)=homicide Standard errors of parameters (above) in brackets. In the set of data referring to the 1970s, Hom17 and GDP7 are not directly connected; the link passes through Mort7. This means that the (negative) relationship between Gross Domestic Product, which repre- sents the wealth of the province, and homicide 1s indirect, in that the impact of Gross Domestic Product on crime is mediated by the quality of childcare. Thus, when welfare investments (of which infant mortality here is an indicator) are taken into account, the (negative) correlation between wealth and homicide disappears. When the regressions on the variables in the preceding periods are taken into account, the negative relationships between wealth and infant mortality seen in the 1980s and 1990s may be interpreted either as causal — 69 — — greater wealth leading to lower infant mortality — or, more neutrally, as the result of a partial negative correlation among the residuals of the regressions for Mort8 and GDPB8, and Mort9 and GDP9, due to causal mechanisms involving variables that are not included in the model. The “nurturance” hypothesis 1s confirmed by the results obtained, for a higher infant mortality rate in the 1970s 1s followed by a higher rate of homicide in the 1990s. Thus, greater investments and higher quality childcare are followed by a reduction in homicide crime 20 years later, when the generation targeted by these services reaches adulthood. This effect is fairly strong (.31) and highly significant (p < .0001). It is note- worthy that our results confirm the ‘“nurturance” hypothesis, in spite of the fact that our data analytic approach differred from that of Vila et.al. (1997a, 1997b). Our results also revealed a negative correlation between infant mor- tality and Gross Domestic Product at a distance of 20 years. This rela- tionship indicates that investment in childcare services increases the wealth of the province 20 years later, when the children targeted by such investment become adults. This path coefficient is, however, weaker (-.13), the level of significance being much lower (p = .030) than that of the relation between infant mortality and homicide; further con- firmation is therefore needed. If these relationships were to be con- firmed, it could be maintained that a lower level of investment in child- care services involves a ‘“hidden cost”, a phenomenon discussed by Keating (1999), who pointed out that a lack of investment in services to support development in early infancy leads to a whole range of undesirable consequences, including a shortage of the “human capital” necessary for economic growth and innovation. The relation between infant mortality in the 1970s and Gross Domes- tic Product in the 1980s displays a fairly similar value (-.15, p=.0004). This relation however, is not confirmed 10 years later, in the following period. If confirmed, the results showing a beneficial effect of childcare investment on the economy would make a new and interesting contri- bution to the debate on welfare. This aspect, however, falls outside the scope of the present research; in our study, Gross Domestic Product constitutes a control variable that is essential in ascertaining the true role of childcare investment, in that it eliminates spurious relationships between infant mortality and homicide, due to the dependence of both on poverty. Âs has already been observed discussing the correlations, each of the three variables is partly explained by the values of the same variables in the preceding period. The regression parameters also reveal a stronger — 70 — association among the three observations of Gross Domestic Product and among the three observations regarding crime, while they reveal a relatively weaker association among the three observations of infant mortality. Infant mortality therefore seems to be the factor that is most amenable to rapid change, a consideration that may be interesting from the political and administrative point of view. The influence of the eco- nomic factor on infant mortality, within the same period, also explains the weaker link among the values of infant mortality over time. CONCLUSIONS The results obtained are in agreement with the hypothesis that the rate of infant mortality influences the rate of homicide 20 years later. In this respect, they lend support to the “nurturance” hypothesis, according to which better childcare yields benefits in the long term, when the chil- dren targeted for intervention become adults. In spite of the fact that the relationships within each 10-year period have been taken into account, the 20-year link persists. This seems to indicate that its effect 1s specific, and that it concerns the positive consequences of an improvement in development during the first few years of life. The data that we have examined do not reveal a direct link between homicide and the wealth of the territory. The indirect relationship between Homi7 and GDP7 seen in the model 1s probably spurious. Indeed, this relationship does not reappear within the 1980s and 1990s, in which the relationships are controlled for the values recorded in the previous decade. It is possible that, if the same control were applied using 1960s values, the relationship between Homi7 and Mort7 would be revealed to be spurious. and that the indirect relationship between Homi7 and GDP7 would disappear. We feel that our findings cast some doubt on the commonly accepted association between poverty and vio- lent crime at the territorial level, and that this association should rather be interpreted as a complex relationship. The results obtained show that the consequences of investment in health and social services for children do not necessarily become appar- ent in the short term, but rather are manifested over the long period. While ecological studies frequently analyse the relationships between social factors and crime by measuring the respective levels of these phe- nomena at the same point in time, our study underscores the importance of assessing the deferred effects of such factors. As Sampson (2000) points out, criminological research has adhered to an excessively static logic in studying crime rates from a spatial perspective, and has repeat- edly compared structural factors (poverty, divorce, etc) with crime rates, — 71 — without identifying the mechanisms at work beneath the associations found. It is rather surprising that criminological studies at the individual level often take the time factor into account, while studies at the com- munity level frequently neglect this dimension. This is all the more sur- prising in the light of the fact that analysis over time can help us to establish the direction of the relationships, and makesit possible to exclude retroactive effects of crime on the other variables. The results obtained refer to three periods of four years each, and require further confirmation. Although provisional, the conclusions to which this study points are interesting both from the point of view of the analysis of the factors that underlie crime and from the point of view of the evaluation of health and social intervention. From the standpoint of criminological theory, our results on homi- cide are in agreement with Vila’s general paradigm of criminal behav- iour (1994). According to this view, in addition to other environmental factors (criminal opportunity, deterrence, social control, etc), it is impor- tant to take into account everything that affects the well-being of chil- dren, in that deficient care during childhood (in terms of health services, social and educational intervention, skills and attitudes of parents, etc) will impact negatively on the socialisation process and, in the long run, favour crime-oriented behaviour. This paradigm therefore links crime prevention to issues of public health, education and childcare, which are often overshadowed by the importance attached to phenomena such as drug abuse, youth gangs, organised crime, etc (Vila, 1994). From the point of view of supportive intervention, our data reveal the possible preventive effects that can be exerted by health and social programmes for children. What is particularly relevant, however, 1s that the effects of such intervention cannot be observed within a short space of time, but will only emerge over a longer period. The practical impli- cations of this observation are considerable, in that social policy deci- sions (if indeed the results of different policies are actually evaluated) are based on current data that do not take into account long-term effects. If our data are confirmed, we will have to conclude that reduc- ing investments in services for children implies an increase in homicide years later, while increased expenditure in this sector will lead to reduced rates of homicide in the long term. While our results appear interesting, the conclusions should be looked at with caution owing to the limits intrinsic in the type of data and analyses used. Since aggregate data are used the results cannot automatically be transferred to the individual level, as this would engen- der the risk of committing an “ecological fallacy”. This risk urges cau- tion in forming interpretations. It should not, however, put a stop to — 72 — research of an ecological nature, in that, as demonstrated by Sampson (1992), similar problems also exist with regard to studies carried out at the individual level’. Moreover, developments in the structural theory of causality that have taken place over the last 15-20 years (Pearl 2000) and the debate on causality in macro-economic analyses, independently of the individual microfoundations of economics (Hoover 2001)‘°), also suggest caution in disregarding causal analyses or interpretations on the grounds that they use aggregated data. Finally, the meaning of the independent variable, infant mortality, raises some problems. Although controlling for the economic variable (per capita gross domestic product) eliminates the possibility of a whole range of spurious relationships, we are unable to state unequivocally that this variable 1s really an indicator of the quality of health and social services for young children. Nevertheless, the finding of a statistically significant link between infant mortality and homicide 20 years later, controlling for economic factors is an interesting one. While needing more specific val- idation by means of further data and other methods of analysis, this finding fits in with the hypothesis initially formulated. REFERENCES BECKETT, Katherine, and WESTERN — Bruce, (2001), “Governing Social Marginality : Welfare, Incarceration, and the Transformation of State Pol- icy”. Punishment and Society, 3, 1, pp. 43-59. BURKE, Amy D., (2001), “Ending Welfare, Continuing Poverty”. Dissent, 48, 4, pp. 90-95. CANCIAN, Maria, (2001), “Rhetoric and Reality of Work-based Welfare Reform”. Social Work, 46, 4, pp. 309-314. COLLINS, Ann, and ABER J. Lawrence, (1997), Issue Brief 1: How Welfare Reform Can Help or Hurt Children. NewYork: Columbia University, National Center for Children in Poverty. GIn order to illustrate what he defines as the “individualistic fallacy”, Sampson cites the correlation between race and violence that is often encountered at the individual level. This correlation might be spurious and linked to contextual factors : indeed, it may be observed that, while only 7% of poor whites live in areas of extreme poverty, some 40 % of poor blacks live in areas of this kind, and that in addition, poor blacks live in areas with high rates of broken families, while poor whites (even those from broken families) live in areas where family stability is relatively high (Sullivan, 1989). The basic explanation might there- fore be community-related and linked to the different social capital present in the various arecas. (6) Per capita gross domestic product, which is also used here, is a paradigmatic example of a macro-economic variable that plays an important role in macro-economic theories; how- ever, its significance at the individual level is unclear. — 73 — CENSIS, (2001), Rapporto annuale sulla situazione sociale del Paese. Roma : CEN- SIS (WWw.censis.it). DEFRONZO, James, (1983), “Economic Assistance to Impoverished Americans. Relationship to Incidence of Crime”. Criminology, 21, 1, pp. 119-136. DEFRONZO, James, (1996), “Welfare and Burglary”, Crime and Delinquency, 42, pp. 223-230. DEFRONZO, James, (1997), “Welfare and Homicide”, Journal of Research on Crime and Delinquency, 34, 3, pp. 395-406. DEVINE Joel A., SHELEY Josef F., and SMITH M. Dwayne, (1988), “Macroeco- nomic and Social-Control Policy influences on Crime Rate Changes, 1948- 1985”. American Sociological Review, 53, pp. 407-420. EDWARDS, David, (2000 2d ed.), Introduction to graphical modelling (Springer, New-York - Heidelberg) EDWARDS David, Havränek, T., (1987), “A fast model selection procedure for large families of models”, Journal of the American Statistical Association, 82, pp. 205-213. FIALA Robert and LAFREE Gary, (1988), “Cross-National Determinants of Child Homicide”. American Sociological Review, 53, pp. 432-445. FULLER Bruce, KAGAN Sharon Lynn, LorB Susanna, et al., (2002), New Lives for Poor Families? Mothers and Young Children Move through Welfare Reform. The Growing Up in Poverty Project - Wave 2 Findings, California, Connecti- cut, and Florida. Technical Report. http ://pace.berkeley.edu/ pace_new_release.html. GEEN Rob, FENDER Lynn, MARKOWITZ Teresa, and LEOS-URBEL Jacob, (2000), Assessing the Effects of Welfare Reform on Child Welfare Caseloads. Wash- ington, DC : Urban Institute. HALES Simon, HOWDEN-CHAPMAN Philippa, SALMOND Clare, WoOoDWARD Alistair MACHENBACH Johan, (1999), “National Infant Mortality Rates in Realation to Gross National Product and Distribution of Income” The Lancet, 354, 2047. JENKS Christopher, and SWINGLE Joseph (2000) “Without a Net”. The American Prospect, vol. 11, n. 4, January, 3 JôRESKOG, Karl G. e SÔRBOM, Dag, (1988), LISREL7 À guide to the program and applications, SPSS Inc. Chicago, Ill. KEATING, Daniel P., (1999), “Developmental Health as the Wealth of Nations” In KEATING Daniel P., HERTZMAN Clyde (eds.) Developmental Health and the Wealth of Nations. New York : The Guildford Press, pp. 337-347. KENWORTHY, Lane, (1999), “Do Social-Welfare Policies Reduce Poverty? À Cross-National Assessment”. Social Forces, T7, 3, pp. 1119-1139. Ku, Inhoe, (2001), “The Effect of Welfare on Children’s Education”. The Social Service Review, 75, 2, pp. 245-270. MESSNER, Stefen F. and Rosenfeld Richard, (1997), “Political Restraint of Market and Levels of Criminal Homicide : À Cross-National Application of Istitu- tional-Anomie Theory”. Social Forces, 75 (4), 1393-1416. MEap, Lawrence, (1992), The New Politics of Poverty: The Non-working Poor in America. New York : Basic Book. MORRIS, Pamela, (2002), “The Effect of Welfare Reform Policies on Children” Social Policy Report, XVI, 1, pp.4-19. www.sred.org/spr.html. __ 74 — MurRaY, Charles, (1984), Loosing Ground : American Social Policy, 1950-1980. New York : Basic Book. PAMPEL, Fred and Gartner Rosemary, (1995), “Age-structure, socio-political institutions, and national homicide rates”. Furopean Sociological Review, 11, pp. 243-260. PATTERSON, James T., (1994), America's Struggle against Poverty, 1900-1994. Cambridge, Mass. : Harvard University Press. PAXON, Christina and WALDFOGEL Jane, (1999), Work, Welfare, and Child Mal- treatment. National Bureau of Economic Research working paper no. 7473, Cambridge, MA : NBER. PAXON, Christina and Waldfogel Jane, (2001), Welfare Reforms, Family Resources, and Child Maltreatment (Www.jepr.org/book/pdf/IncentivesWaldChap2.pdf) SAMPSON Robert J. (2000), “Whither the Sociological Study of Crime”. Annual Review of Sociology, 26, 711-714. SAVAGE Joanne and VILA Bryan, (1997a), “Lagged Effects of Nurturance on Crime : À Cross-National Comparison”. Studies on Crime and Crime Preven- tion, 6, 1, pp. 101-120. SAVAGE Joanne and ViLA Bryan, (1997b), Changes in Child Welfare and Subse- quentcrime rate trends: a Cross-national Test of the Lagged Nurturance Hypothesis. SAVOLAIEN Jukka, (2000), “Inequality, welfare state, and homicide : further sup- port for the institutional anomie theory”. Criminology, 38, 4, 1021-1039. The Economist, (2001), United States. America’s Great Achievement : Welfare Reform. London, Aug. 25, pp. 25-27. VILA Bryan, (1994), “A General Paradigm for Understanding Criminal Behavior : Extending Evolutionary Ecological Theory”. Criminology, 32, 501-549. ZASLOW Martha. TouT Kathryn, SMITH Sheila, MOORE Kristin, (1998), Implica- tions of the 1996 Welfare Legislation for Children : À Research Perspective. Social Policy Report. Society for Research in Child Development, vol. XII, n. 3, pp. 1-35. ABSTRACT According to Vila’s (1994) perspective, better maternal and infant health care will be reflected in reduced crime rates a generation later. We tested this hypothesis by exploring the relationship between the rates of infant mortality and the lagged rates of homicide (20 years later), controlling for gross domestic product in the 95 Italian provinces. The variables are measured at three points in time : in the middle of the sev- enties, in the middle of the eighties and in the middle of the nineties. We used LISREL to evaluate the model suggested by the hypotheses; the “nurturance”’ hypothesis was corroborated by the results on homicide. __ 75 — RÉSUMÉ Dans la perspective de Bryan Vila (1994), une meilleure santé de la mère et de l’enfant se reflète dans une réduction des taux de crimes à la génération suivante. Les auteurs ont testé cette hypothèse en explorant la relation entre les taux de mortalité infantile et les taux de condamna- tions pour homicide (20 ans plus tard), en tenant compte de la variable de contrôle du produit domestique brut per capita, qui pourrait inter- férer avec la première variable. Les variables ont été mesurées à trois moments : au milieu des années soixante-dix, quatre-vingt et quatre- vingt-dix, dans les 95 provinces d’Italie. Le programme LISREL a été utilisé et a permis de corroborer l’hypothèse de l’approvisionnement (Cnurturance») de la délinquance, à long terme, par le manque de soins au premier âge. RESUMEN En la perspectiva de Bryan Vila (1994), una mejor salud de la madre y del niño se refleja en una reducciôn de la tasa de crimenes en la gen- eraciôn siguiente. Los autores han sometido esta hipôtesis a prueba explorando la relaciôn entre las tasas de mortalidad infantil y las tasas de condenas por homicidio (20 años mâs tarde), teniendo en cuenta la variable de control del Producto Interior Bruto por câpita que podria interferir con la primera variable. Las variables han sido medidas en tres momentos : a mediados de los años setenta, ochenta y noventa, en las 95 provincias de Italia. El programa LISREL se ha utilizado y ha per- mitido corroborar la hipôtesis del aprovisionamiento («nurturance») de la delincuencia, a largo plazo, por la falta de cuidados en la primera infancia. —— 76 — FOREIGN JUVENILE OFFENDERS IN GERMANY, ITALY, FRANCE AND SPAIN e ° THEÉORIES, PROFILES AND NOTES FOR A SOCIAL POLICY César San Juan ” Fernando Bermejo Estefania Ocâriz Keywords : juvenile delinquency immigration legal multicultural. PHASE 1: CRIMINOLOGICAL ANALYSIS OF THE IMMIGRANT JUVENILE OFFENDER WITHIN THE EUROPEAN CONTEXT 1. — Introduction The European Union 1s gradually becoming a multicultural society. It already is. Within its borders there have been some major advance- ments in favour of this supranational reality, such as the free movement of EU citizens or the European monetary unit. Other factors of change originate from outside Europe, one of the most notable being the influx of immigrants from other cultures. In the middle of the 20th Century, immigrants were seen as tempo- rary residents whose aim in life was to return to their country of origin once their financial situation had improved. In more recent times, immi- grants have preferred to settle in the host society together with their families. Generations of immigrant descendants are therefore establish- ing themselves in multicultural societies where, sadly, exclusion and dis- crimination seem to predominate over intermixing and integration. Immigrants, and their descendants too, live worryingly often in a sit- uation of social disadvantage with respect to their hosts, and this 1s reflected, among other aspects of life, in crime. In the case we are con- cerned with, the media distribute news and “official figures” which instil within public opinion the idea that minors of foreign origin are danger- () This research has been financed by European Commission ( AGIS program). (*) César San Juan is Assistant Director of the Basque Institute of Criminology and Director of the DMS International Research Centre. Basque Country University. cesar.san- juan@ehu.es. __ 77 — ous. From the Criminological perspective, it is necessary to undertake studies to establish offender profiles and to objectively reflect the nature of criminality among foreign minors in relation to juvenile delinquency as a whole. 2. — An approach to foreign juvenile offenders in Germany, Spain, France and Italy 2.1. — Figures Within juvenile crime as a whole, in Spain the rate of foreigners increased by 13 points between 1998 and 2002 (going from 8 to 21%), the most significant increases occurring during the later years of the period, coinciding with the increase in immigration. This leads us to pre- dict that this proportion of crime will continue to rise over the next few years if the issue of immigrant integration is not adequately addressed. In France, during the same period, the rate remained around 15% for criminal offences; in Italy, between 1995 and 2001 the rate fell by 6 points, from 28 to 22%. On comparing the situation in the 3 countries we note a drop in Italy and a rise in Spain, both converging towards a rate of 21-22%, whilst France remains at a lower level. In terms of the countries studied, it is worth noting that the crime index (or CI : the number of crimes per 10,000 inhabitants) for minors of foreign origin is significantly higher than that for native inhabitants. So, the CI for Italian minors (14-17 years old) is 11/10,000 whilst for foreigners it 1s 22.5 (including 1llegal immigrants), in all, over double the index for Italian minors. Considering the total number of crimes, offences and class 5 misde- meanours, in the case of France the CI for foreign minors is more than three times that for French minors (ratio 3.5 to 1). This CI continues to be higher in any of the three types, but 1t 1s almost 4 to 1 for committing an offence. With regard to the data obtained from Spain, between 1998 and 2002, foreign minors committed offences proportionally in a ratio of 9 to | with respect to Spanish nationals, although these figures should be treated with great caution as they require a series of demographic cor- rections. 2.2. — Socio-demographic profile With regard to the foreign minors arrested in Spain, one basic factor is their geographical proximity, as almost half of them came from __ 78 — Morocco (49%) and in a much smaller proportion from Algeria (11%) and Romania (9%). In Italy, of the total number of foreign minors held in Juvenile Reception Centres during 2003, 34% came from Romania, 16% from Morocco and 13% from Serbia-Montenegro. In the case of France, the majority of foreign minors in prison come from Eastern Europe and the Maghreb and whilst the figure for Magh- rebis dropped by half between 1993 and 2002, the minors from Eastern Europe multiplied by 10 during that period. There is a notable presence of unaccompanied Romanian minors. In short, there 1s evidence of a significant presence of juvenile offend- ers originating from the Maghreb and from Eastern Europe, mainly from Morocco, Algeria, Romania and the former Yugoslavia, many of them unaccompanied. It is a fact corroborated by the vast majority of criminological stud- ies on juvenile delinquency that males represent around 90%. However, in Italy, of the foreign minors in custody, there is an unusual proportion of girls, 36% in 1999, a rate which dropped to 27% in 2001. 2.3. — Analysis of the criminality of foreign juvenile offenders : Criminal typologies On analysing offences grouped into types of crime, in the case of Italy, we find significant differences between Italian and foreign juvenile offenders, as with the latter there 1s greater criminal activity with regard to crimes against property, whilst their rate with regard to violent crimes remains at around 8%. In Germany there is a certain over-representation of juvenile offend- ers of foreign origin with regard to property crimes, but their involve- ment in violent crimes 1s also significant, and is particularly related to ex-Yugoslavs and Turks, linked into a culture of honour which legiti- mises and even extols the use of violence among males, considering it a sign of manliness. Paradoxically, this “culture of violence” 1s shared with young people of the extreme right, neo-Nazis, who are radically opposed to immigration. There 1s also a clear tendency towards violence in a specific type of German people, although of foreign origin : these are naturalised ex-Soviets who have serious problems with integration. The concept of “foreign” 1s too broad a term which can refer to very diverse situations but which, in the absence of more specific data such as nationality, serves as a research variable. The Italian and Spanish reports make references to local studies which do in fact show evidence of correlations between nationality and higher rates in certain types of crime. So, by way of example, a study included in the Italian report showed that Algerian and Moroccan minors were over-represented in relation to those from other origins, in crimes such as the theft of wal- lets and cars. 2.4. — Explanatory theories of juvenile delinquency 2.4.1. — Culture conflict theory This theory focuses on the specific tensions experienced by many children of immigrants, who find themselves at the crossroads of adopt- ing the cultural values of their senior family members or integrating into the receiving country, with different values. Pressurised by their parents and by the society in which they live, they must face the dilemma of choosing between family or social confrontation, amounting to pressures which can lead them to crime. In the case of France, problems of integration increase in proportion with cultural differences and the negative experience, on the part of the parents, of their status in the receiving country. In this respect, there 1s the specific situation of the Maghrebi minors, whose ancestors came from the colonies. According to Esterle-Hedibel (1999), the delinquency of these young people is not so much the result of cultural differences between their families and the receiving society, but of the experience of their ancestral culture as being undervalued. These minors, therefore, experience double socialisation, from their parents and from their environment, and this can be contradictory and lead to axiological dilemmas and loyalty issues, or to identity or inter- generational conflicts. The culture conflict theory, endorsed by a great amount of empirical research, provides an explanation for delinquency specifically in immi- grant minors, but not on its own; it must be combined with other the- oretical approaches : 2.4.2. — Social structure theories These highlight the fact that ethnic minorities, in general, suffer material insecurity and situations of social disadvantage with respect to other groups, which would lead them to a greater extent into property crime. This situation of social disadvantage in immigrants 1s evident in the countries dealt with in this study. In Germany, various notable studies have been undertaken by the Criminological Research Institute of Lower Saxony, which indicate that the over-representation of foreigners and ethnic minorities within crime — 80 — as a whole is in many cases linked to socio-economic variables (level of education, unemployment, etc.). In France, the unemployment rate is very high among immigrants. The work carried out by the parents is not very skilled and is very poorly paid. Unemployment also has a considerable effect on the first- born of the Ist generation, including those who are qualified. The per- ception the children have of the family’s living conditions and the zero expectations of social mobility, tend to reinforce in them the idea that idleness and criminal behaviour are also productive, especially drug traf- ficking. À quantitative study of these socio-economic factors undertaken by Roche (2001 and 2004) on self-disclosed crime, showed that a correla- tion could be established between socio-economic factors (which partic- ularly affect young people of foreign origin) and the increased tendency towards crime. This theory may provide an adequate explanation mainly for crimes against property, but not for violent crimes, such as those legitimised by the “culture of honour” particular to Turkish and Balkan young offend- ers 2.4.3. — Ecological theories These theories all claim that the environment can have a decisive influence on human behaviour, also on criminality. On analysing the backgrounds of juvenile offenders it 1s clear that many live in deprived areas, in conditions of overcrowding…. Whilst strictly linked to the the- ories of social structure, the ecological theories focus on psychosocial aspects such as the effect of the surroundings or anomie. Specifically, this approach seems to make sense in France. We find differences between the French and the foreigners residing in these con- flictive areas. On comparing both groups of minors, the French seem more prone to delinquency as, in accordance with anomie theory, they have higher quality-of-life expectations than the immigrants, and there- fore experience greater frustration when they move to these problematic districts, reinforcing their antisocial behaviour. In contrast to the French minors, the crime rate among young Magh- rebi offenders remains the same regardless of their place of residence, and therefore the correlation between the suburban habitat and crime has not been established. The environment in itself does not seem to have a particular influence on the criminal behaviour of the young Maghrebis. The subjective experience of the environment, however, does. In this respect Sicot (2000) distinguishes between two contrasting categories of offenders of foreign origin in France : the “rebellious”, — 81 — habitual offenders whose families are resentful of France and see their environment as a hostile means of relegation, and the “conformists”, one-off offenders whose family circle accepts the situation to which its members have adapted. 2.4.4. — Criminal personality theory Several authors, such as Cabello and Fernändez (s.d.) in Spain, or Cario (1999) and Frechette and Leblanc (1987) in France, support the idea that the offender 1s not born with a criminal or antisocial person- ality, genetically inherited, but that their personality is developed in an environment which modulates its traits according to their personal assimilation of experiences… in the course of a complex maturing proc- Ess It is worth mentioning that criminal personality theories are particu- larly important for explaining - if only in part - the incipient criminality of minors, as it 1s during infancy and adolescence that their main traits are formed. We would therefore like to highlight that during this critical period of development, both precociousness and the failure to apply prompt treatment, lead to a high risk of developing an antisocial per- sonality, and thus an increase in criminal behaviour. 2.4.5. — Labelling theory Some studies (although other research, such as the study carried out by Dittmann and Wernitznig (2003) in Bavaria, found no negative dis- crimination towards foreigners) such as the works of Roché (2001 and 2004), in France, or Avilés (2002a and 2002b), in Spain on the police, indicate the fact that in police, judicial and prison institutions in the countries studied there 1s evidence of discrimination and hostile practices towards ethnic minorities. Let us, for example, refer to Levy and Zau- berman (1998), in France, who indicate the existence in the police force of hostile practices towards foreigners, in the form of greater control and heightened suspicion due to their greater visibility, and a generalised racist rhetoric, which generates in minors of immigrant origin recipro- cated hostility towards the police authorities, which are seen not as a protective body, but as coercive in particular, and towards the system in general. This hostility can lead them to commit “expressive” crimes in response to their discontent. This iter criminis would constitute a clear example of “self-fulfilled prophecy” and would confirm, at least in part, the arguments of the Labelling Approach, according to which label- ling an individual as a delinquent can initiate his/her criminal behaviour. __ 82 — On a judicial level, the greatest discriminatory situation for minorities 1s to be found in preventive detention, a measure which 1s experienced much more frequently by foreigners than by nationals. In France, there is double the amount of foreign minors in preventive detention with respect to French minors. The response of the ministers concerned 1s that in order to grant provisional freedom, an important criterion 1s the availability of a fixed address, which 1s much less common among for- eigners, and this would explain to some extent their over-representation in internment centres in Germany, Spain, France and Italy. 3. — The second generation of immigrants 3.1. — Conceptual approach Broadly speaking, several authors, such as Avilés (2002a and 2002b), Capdevila (2000), or Moreno, (2002), consider this second generation as being made up of all the descendants of immigrants, both those born in the country of origin and those born in the destination country. Strictly speaking, the second generation consists of individuals who were born in a particular country but some of whose older family mem- bers are foreign, and it would therefore be incorrect to class them as immigrants. In this report we have chosen to adopt the latter approach, keeping this term as it 1s the most widely used. 3.2. — Incidence of crime among second generation minors In the case of Germany, whilst in 1st generation immigrants (who arrived during the 50s and 60s) there is no sign of an increase in their crime rates in comparison with Germans, several studies relating to the 70s and 80s showed that young people from the 2nd and 3rd generations of immigrants had a crime incidence which was double that of the Ger- mans. In France, a study by M. Giraud (1993), carried out in Lyons, emphasised the high amount of crime committed by minors of foreign origin, particularly from the Maghreb. 3.3. — Aetiology This second generation has been born and has developed and been brought up in the host society, becoming familiar with many of its val- ues, but in return they must face a series of discriminatory barriers which they did not expect to encounter (and which are more easily __ 83 _ accepted by foreigners), and this can lead to various psychosocial problems : Cultural conflict : The second generation can become victims of cul- tural conflict when minors have experienced double socialisation (from their parents and from the society in which they live) which can provoke the dilemma of having to choose between a family conflict (if they adopt the values of their society rather than those of their older family mem- bers) or social exclusion (if they adopt the values of their parents and reaffirm themselves as members of a minority group). This dilemma can create psychological tensions and identity conflicts in the minor, which can in turn lead to the emergence of antisocial behaviour. Differences in the social structure : The situations of social disadvan- tage in which many immigrants find themselves usually continue into the second generation, with the difference that immigrants tend to accept it as a “toll”, whilst the second generation tend to find the con- tinuation of their social exclusion intolerable, laying the foundations which legitimise subsequent antisocial and criminal behaviour. Social discrimination : Although this second generation was born in the host society and has adopted 1ts values, many of them are discrim- inated against in many different areas and treated as if they were for- eigners, restricting their rights, limiting their employment expectations, their access to living accommodation, etc., logically and rightfully increasing their frustration against an excluding society. Experiencing this situation can create and develop problems of identity and anomie, laying the foundations for future antisocial behaviour. So, in France 1t 1s clear that, with the same amount of qualifications, descendants of immigrants are seen to be more affected by unemploy- ment than the ‘“native” French, mainly due to ethnic discrimination. Cognitive factors : Situations of social disadvantage or discrimination can be experienced in very different ways. Certain conformist or hostile attitudes towards their situation will therefore emerge (these are being seen more often among second generation minors). These attitudes can emerge and grow from within the very bosom of the family or, subse- quently, within the peer group (in accordance with subcultural theories). Consequently, comparing the aforementioned study by Sicot (2000), it is believed that for minors of the second generation, the subjective attitude of the family prevails over the objective conditions : the con- formists and the rebellious/resentful (see section 1.4.3) suffer similar sit- uations of social disadvantage but show a different incidence of crimi- nality. __ 84 — PHASE Il : CRIMINOLOGICAL PROFILE OF THE IMMIGRANT JUVENILE OFFENDER : A RETROSPECTIVE STUDY. 4. — Procedure The analysis developed below is the result of a study of the files of sentencing recorded in the Juvenile Courts of the Autonomous Commu- nity of the Basque Country (San Sebastiän, Vitoria and Bilbao) from January 2001 to December 2003, specifically of the psychological reports produced by psychosocial teams and the actual content of the judgment. We were able to access all files of convicted immigrant minors (N =168), of which we have been able to analyse 122 psychological reports. Consequently, going beyond a representative sample, practically the whole world of immigrant juvenile offenders corresponding to the period indicated is described. In parallel to this, a comparison group was formed consisting of non-immigrant juvenile offenders, with an equal number of files and for the same period of time, selected using a sys- tematic random procedure. Once the files had been selected, the information contained within them was then extracted, identifying a series of key variables associated with the criminal and psychosocial profile of the minors, the analysis of which 1s shown below. 5. — Criminal profile and record — — The majority of files analysed (N 168) of immigrant juvenile offenders corresponded to males (92.3%). The average age at which the first crime was committed was 16. 66% of the minors did not reoffend, 15% went on to commit a second crime, whilst the remaining 20% of minors reoffended between 3 and 10 times. In terms of the nature of the crimes, it is worth mentioning that in 47.6% of the cases, the minor confronted their victim in some way when they committed their first crime (by threatening, insulting or assaulting the victim). With regard to the measures adopted, in 53.6% of the cases these were implemented in a semi-open prison, 23.8% in a closed prison, whilst in 21.4% of cases the mediation process involved no penal meas- ures at all. __ 85 — 6. — Delay in response and situation of the minor One of the guiding principles of the models for delivering measures in Juvenile Justice is the immediacy and reciprocity of the response. As established in the Justice Plan for the Basque Country, the time period between the acts and their consequences must be very narrow so that the greater part of the educational element of the response 1s not lost. For this reason, we have created a delay in response indicator, cal- culating the time elapsed between when the crime was committed and the judgment reached, obtaining an average delay of 328 days, although with a high typical deviation (226.95). Furthermore, a quartile analysis showed that 25% suffered a delay equal to or less than 175 days, 50% equal to or less than 255 days and finally, 75% equal to or less than 423 days. But these clarifications do not soften the fact that there is an excessive delay in response. Of the population of foreign minors analysed, practically half (49.2%) live with family members. Whilst of the unaccompanied minors, practically all of them are Moroccan (83.9%) and Algerian (11.3%), among the accompanied minors there is a great diversity of origin (Colombia, Ecuador, Cape Verde, Romania, USA, France, etc.), with a notable presence of Portuguese, representing almost 32% of the accom- panied minors. With regard to the place of residence, almost half live in a juvenile centre in accordance with judicial measures, specifically 49.2%, and 44.3% live in private homes. Just 8 minors have no fixed address. In terms of adapting to the place of residence, over half of the minors, almost 60%, have adapted. To make this variable operational, indicators such as good behaviour, adhering to rules, showing respect for others you live with, treating property with care, etc., have been taken into consideration. 7. — Family relationships In almost all cases, both parents of the minors are alive, 12% have at least one of them and only 3 minors are orphans. Over two-thirds of this population maintain a relationship with their parents, 41.5% with both, compared to 30.3% who have no kind of relationship with their parents. Of the minors who have a relationship with their parents, for the majority (64.2%) this is habitual and not sporadic. With regard to the residence of the parents of immigrant juvenile offenders, we can see that almost half of the minors’ parents reside in their country of origin (49.2%), a figure which corresponds to the one we have obtained for unaccompanied minors. Of the remainder, 42.4% of the parents reside at the same address as the minor and only in 10 cases do the parents reside at a different address. In terms of parental control, the data shows that in the majority of cases, in practically 78%, the parents exercise no form of control over the minors. The remainder either exercise inadequate control (17.2%) or only in 4.9% of cases exercise good control. Another aspect to consider is the relationship with their siblings. Over half of those who answered this question, specifically 56.4%, have a relationship with all their siblings, whilst in 36.4% of cases, they have no relationship with any of them. In terms of the frequency of the rela- tionship the minors have with their siblings, we can see that of those who do have a relationship, it 1s habitual in 70% of cases. The siblings of almost half of the immigrant minors studied (48.6%) reside in the country of origin and 42.2 % of the immigrant juvenile offenders live with their siblings. 8. — Vulnerability factors With regard to the economic situation, a large majority (77.4% of the families of the minors) experience a high level of economic vulnerability. In order to assess the economic situation of the families of the minors, we have taken into account whether the parents work, receive social benefits, have their own home, etc. With regard to the regularisation of their residence, over half do not have the relevant permits, compared with 47.5% who are in the fortu- nate position of being regularised. In terms of groups of friends, two-thirds of the population analysed keep the company of risk groups, 1.e. they receive no schooling, con- sume some kind of toxic substance, display antisocial behaviour (rob- bery, theft, fighting), live in an unstable environment, etc. With regard to their schooling, it 1s worth noting that 58.7% of those interviewed do not go to school. Of the minors who attend school, almost 70% show a good level of adapting to school or workshops. This level has been based on the evaluation of behaviour in terms of time- keeping, regular attendance, etc. Therefore, although only a few of the minors study, most of these have adapted well to the school environ- ment. In addition to the level of schooling, the study has also looked at the working activity of these minors. Over 91% are not working, although, as was the case with the students, of the few who do work it can be seen that most of them have adapted positively to the workplace. __ 87 — Considering the above data as a whole, 50% of them do not study or work, thus creating a special risk situation. We shall conclude this section by stating that almost two-thirds of these minors, 63%, consume some kind of toxic substance. 9. — Profile of the immigrant juvenile offender The aim of this descriptive analysis was to outline a profile of the immigrant juvenile offender based on the prevalence of the most signif- icant psychosocial traits. Within a general profile of social exclusion, the most significant risk factors concern the lack of parental control, within a framework of a broken family unit, and alternative socialisation based on the integration of the minor into a peer group prone to transgressing the law : — Male aged 16 at the time of committing his first crime. — 50% are unaccompanied minors and have no residence or work per- mit — Almost half live in juvenile centres, reception houses, internment cen- tres, etc., and 6% have no fixed address. — 41% have not adapted to their place of residence. — 30% of the immigrant juvenile offenders have no relationship with their parents and for those who do maintain a relationship, it is spo- radic in 36% of cases. — 36% have no relationship either with their siblings. — The majority have no family control whatsoever (almost 78%). — In 77% of cases, situations of serious family insecurity are reported. — 66% have friends in a social risk situation. — Over half receive no schooling and, of those who do, 32% have not adapted to this schooling. — Only 8% work, but almost all of these (89%) have adapted to the workplace. — 62% consume some kind of toxic substance. 10. — Differences between immigrant and non — immigrant juvenile offenders To obtain a better understanding of the idiosyncratic characteristics of the immigrant juvenile offender, this group must be compared with another group of non-immigrant juvenile offenders. The whole sample 1s therefore made up of 239 minors from both groups. __ 88 — In order to analyse whether significant differences existed between the two groups, an ANOVA was carried out on a factor indicating the main differences (significant if p>0.005) in the family and school environments : — Immigrant minors have less of a relationship with their parents (F = 11.158, p < 0.001) and siblings (F = 10.967, p < 0.001). — Parental control is less in the group of immigrants (F = 17.290, p < 0.000). — Family insecurity is less in the group of immigrants (F = 36,497, p < 0,000). — Non-immigrant minors receive more schooling (F = 7.588, p < 0.006). On the other hand, the group of immigrants has adapted more to schooling (F = 17.105, p < 0.000). — No differences have been found in terms of the delay in response from the Courts or in the age at which the first crime is committed — Non-immigrants commit a significantly higher number of crimes than immigrants (F = 9.775; p<0.002), they reoffend more. In terms of the typology of crime, we highlight in figure 1 those crimes which have been committed more frequently and can cause greater public alarm or police involvement (where A is burglary; B is robbery with violence or intimidation; C is resistance to authority; D is injuries and E is drug trafficking). The differences between both groups are more evident in violent robbery, resisting the authorities and drug trafficking. Figure 1: Type of crime and group 20 151 104 O non-immigrants 5f E immigrants f The hypothesis suggested to us by these exploratory analyses is that vulnerability factors are strongly related to the stability of family ties. Specifically, the group of unaccompanied foreign minors (UFMs) has been identified by all Juvenile Justice Administrations as a group which —0 must receive special attention through the implementation of an inten- sive set of socio-educational resources : a very complex challenge pre- cisely due to their lack of association with a positively influential social network. In order to analyse, specifically, the differential aspects of the UFMSs with respect to the other immigrant minors and to the group of juvenile offenders born in Spain, we have identified, through an ANOVA, three distinct groups : I. Accompanied immigrant minors. IM. Unaccompanied immigrant minors and I. Non-immigrant minors. On comparing averages with Scheffe's post-hoc multiple comparison test we have been able to verify that in the variables relationship with parents (F = 22.742; p < 0.000), parental control (F = 14.620; p < 0.000), contact with high-risk friends (F = 8.818; p < 0.000), and con- sumption of toxic substances (F = 9.121; p < 0.000), which can be con- sidered as risk factors especially conducive to crime, there are no differ- ences between accompanied immigrants and non-immigrants, but they do show significantly different averages from the group of unaccompanied immigrants. This group shows a combination of a low level of parental control, more relationships with high-risk friends and a greater consump- tion of toxic substances. It is surprising that, with this profile, they are not the group of minors who reoffend most (F = 4.856), which are the non-immigrants (x = 2.80). The problem lies in the fact that the immi- grant unaccompanied minors perpetrate a greater number of crimes involving direct confrontation with the victim ( x = 1.35) than the group of immigrant accompanied minors ( x = 0.52; p< .003) and the group of non-immigrant minors (x =1.19; p < .005), which can lead to the crea- tion of a greater feeling of insecurity in the community, due to their vis- ibility. 11. — Conclusions and some notes for intervention with foreign juvenile offenders In the middle of the 90s a new protagonist appeared on the scene of migration flows : the minor. This new characteristic of migration is forc- ing various European countries to devise a new legal, penal and social system to manage, penalise/educate and integrate minors who begin a career of crime on European soil and, moreover, to protect their rights. One of the first conclusions which must be presented as a result of this research 1s that the correlation between crime and immigration is in no way supported by intrinsic factors relating to the fact that they are —— 90 — foreign. But rather by a series of extrinsic factors such as the socio-eco- nomic conditions or the treatment they receive from society and recep- tion institutions, which they would share with other disadvantaged native groups. This being clear, we must remember that the only impor- tant differences which could distinguish foreign juvenile offenders from national juvenile offenders, are above all related to the accumulation of psychosocial stresses and to limited means of social participation, wors- ened by the fact that they possess, from the outset, two counterproduc- tive characteristics : being a minor and being an immigrant (from Cuesta & San Juan, 2005). On a global scale, there seems to be a development of what we could call the infantilisation of exclusion (in addition to its feminisation) in a context within which there is also a great visibility of the immigrant minor marked by public alarm. These considerations aside, the reality 1s that with an initial review of European documentation, we can identify a growing set of crimino- logical problems associated with the immigrant minor among which we are going to mention, by way of example, the growing seriousness of the cases of crime in Italy, the control of the Ceuta and Melilla borders in Spain, inter-ethnic confrontations in the UK, the decline of peripheral areas with over-representation of immigrant populations and children of immigrants in France, Belgium and Germany, etc. In fact, an element of each of these problems is occurring to a greater or lesser extent in the whole of Europe. We must point out that, whilst earlier in this study we have put for- ward a number of theories to explain the over-representation of foreign- ers in juvenile crime, none of these can be used in isolation to explain this complex phenomenon. À broad, ecosystemic, multilevel theoretical framework which takes into account cultural differences and conflicts, the disadvantages of the social structure and the influences of a hostile environment (both physical -deprived areas-, and human -discrimina- tion-), will enable us to obtain a better understanding of this prevalence gnd, more importantly, to implement the necessary measures to reduce it Of the problems described, we believe that the case of Unaccompa- nied Foreign Minors (UFMSs), a group which is becoming increasingly large in Europe, merits particular attention, due to their situation of heightened vulnerability. In Spain, France and Italy a large number of these minors come from the Maghreb and the Balkans, mainly from Morocco, Albania and Romania. From an operational point of view, we believe it appropriate to dis- tinguish between two large groups of UFMs: on the one hand those _— 91 — who come in search of work with a plan and show a positive attitude towards integrating into society, and on the other hand those who have grown up on the street or who have no migration plan, a minority group but whose criminological profile corresponds to a large extent to that of habitual juvenile offenders, and who therefore require psychoso- cial measures in order to integrate. In the European context there is no overall policy for dealing effec- tively with receiving and protecting UFMSs in a position of social risk. The social, legal and administrative treatment of these minors 1s heter- ogeneous and action initiatives are often limited to various local or international organisations involved in the resocialisation of these minors, and often clash with the action principles of Governments. These reports all underline the lack of a reception and support sys- tem to enable them to integrate into society and to provide them with an adequate education. On the contrary, during their process of adap- tation, their situation of insecurity and their link to criminal and mar- ginal environments are reinforced, making it possible for them to turn into habitual offenders or become victims of sexual exploitation. If we are to put the interests of the minors first, in view of the insecure con- ditions of their society of origin and the absence of any known or struc- tured family guidance, repatriation should be limited to very specific cases. However, the mass influx of these minors, combined with the lim- ited human and material resources available, make it impossible for all UFMs arriving in the EU to be received adequately. The current situa- tion, therefore, points towards an uncertain future. Another phenomenon which we have also wanted to focus upon is that relating to minors known as “second generation immigrants”. This is a group in a position of risk which in Belgium, Germany and, above all, in France has exploded as a result of a history of tension between their status as Furopean citizens and being deprived of “EFuropean expectations” (Cano, 2006). In the rest of Europe, it is only a question of time before conflicts break out if the appropriate measure are not taken. The position of social disadvantage seems to prevail among a group in which establishing their social identity is also, in itself, a poten- tially stressful and complex process. Taking into account that, in line with current migration trends, the majority of immigrants coming into European Union countries intend to settle here, the number of individuals belonging to these new gener- ations is going to increase considerably in the near future. In Germany, for example, they are already talking about “third generation immigrants” : Multiculturalism can be a fascinating source of cultural __ 99 _ wealth, but the ghettoisation of society 1s also a possible outcome. It is today. In the case of Spain and Italy, it will be important to study the devel- opment of the level of social and labour discrimination affecting these second generation immigrants in countries with a greater tradition of receiving immigrants, such as France or Germany, and the measures implemented to mitigate this discrimination if there is the political will to avoid making the same mistakes. As Cano (2006) points out, if sat- isfactory medium or long-term integration of these descendants of immi- grants is not carried out, this could result in the establishment of groups of adolescents affected by their lack of roots and frustration, as a con- sequence of the continuous social exclusion to which they can be sub- jected. One of the first conclusions that we can form from the study data is that the immigrant juvenile offenders share a series of risk factors as no significant differences have been found, in terms of the criminal profile, with respect to minors born in Spain. What does stand out 1s that it is unaccompanied minors who have perpetrated crimes which can cause greater public alarm. However, we have been able to verify that in the indicators which can be the most informative, such as reoffending (understood as the number of offences), there are no differences between immigrant unaccompanied minors and minors born in Spain who, in fact, commit more crimes than immigrant accompanied minors. The data gathered in this research suggests to us the importance of paying special attention to the context of departure, to the underlying structural causes for the need to migrate. An analysis of the personal migration plans of minors, their expectations, motivations, etc., must be undertaken. It must be emphasised that the context of arrival is charac- terised as being a scenario with a culture which is alien to these minors, which can create an identity crisis, which will in turn condition their process of integration and will determine whether their future will focus on the myth of return or on settling into the new society. Depending on their life history, it is possible that the idea of the future will have lost all meaning when they find themselves under pressure to address their day-to-day needs. However, that said, this type of analysis is necessary if there is to be an attempt to formulate overall policies of a preventive nature. It 1s also worth pointing out, in addition to those mentioned in rela- tion to the unaccompanied foreign minors and the second generation, other emerging problems such as those minors, whether immigrants or not, who come of age whilst under judicial measures or those minors, __ 93 _ again whether immigrants or not, who from an environment of protec- tion enter the system of reform. In all cases, it seems to be a priority for actions to be based around three fundamental aspects : — Schooling, in Europe, from a preventive perspective, must provide an education system that provides descendants of immigrants with the same opportunities for integration as the rest of the school popula- tion. The political agenda, in this respect, has certain outstanding ISSuEs. Furthermore, with regard to employment and vocational training, in order to prevent situations of economic insecurity, giving rise to crime, we propose measures which would enable immigrant minors of working age to work and train, with the opportunity to undergo work placements, have contracts with reduced hours to allow them to con- tinue their training or so that the companies employing them can ben- efit from tax allowances… Finally, the family, the framework for the basic socialisation of the minor. The loss of family influences should be avoided as far as pos- sible, and where this is not possible, alternatives should be sought, such as guardians in native families, a system of mentors, etc… In short, to identify stable figures of influence who can provide them with the parental control we talked about earlier as a factor which can protect them from situations of marginalisation and exclusion. Nevertheless, it 1s essential to explore somewhat bolder strategies and to propose original approaches which can be flexible according to the resources and, of course, can adapt to the circumstances of the minor, which must guide penal measures, almost always before defining the punishment for the offence. So, it 1s clear that this 1s a complex problem which requires compre- hensive, cross-border strategies and policies, as compartmentalised solu- tions are, with the utmost certainty, doomed to failure. We believe it 1s essential to adopt a systematic approach to analysis and intervention where we must identify a series of levels from the individual to the mac- rosystemic, spanning interpersonal, group, school, family, work and lei- sure relationships, also addressing religious aspects, as well as everything relating to the parameters of the current structural conditions from a politicalleconomic point of view on the one hand, and from the /egal/ social point of view on the other. Clearly, this working hypothesis will depend on the will of both gov- ernments and citizens, as otherwise it would become evident that having no plan for the future is not just a shortcoming of the young offender. __ 94 — REFERENCES AviLÉs Farré, J. (2002a) “La cuestiôn de la inmigracion en España” [online]. Available from : http ://www.gees.org/articulo/292/ [Accessed 25 October 2006] AVILÉS Farré, J. (2002b) ‘“Tendencias del delito en España” [online]. Available from : http ://portal.uned.es/pls/portal/url/ITEM/ 0B1818CODB06EA84E040660A32697765 [Accessed 14 February 2007] CABELLO Vâäzquez, I., y FERNANDEZ Garcia, M* J. (s/d) “Menores en conflicto social”. In Save the Children, Menores y conflicto social, 21- 22 April, Cerezo de Abajo- Segovia. CAPDEVILA ! CAPDEVILA, M. (2000) “Los menores extranjeros indoc- umentados no acompañados (M.E.I.N.A.) exigencia de nuevas repues- tas” [online]. ZN Universidad Pontificia, ZZ Congreso sobre la Inmigracion en España España y las migraciones internacionales en el cambio de siglo, 5-7 October 2000 Madrid. Available from : http ://www.oijJ.org/ doc/documental_806_es.pdf [Accessed 30 April 2007] CANO, M. A. (2006) “Algunas reflexiones criminolôgicas sobre el fenômeno de la violencia juvenil en Francia”. Revista Electrônica de Ciencia Penal y Criminologia [online], 8 (4): 1-31. Available from : http ://criminet.ugr.es/recpc/08/recpc08-04.pdf [Accessed 30 April 2007] CARIO, R. (1999) “Jeunes délinquants. À la recherche de la socialisa- tion perdue”. Paris : L’Harmattan, coll. Transdisciplines. CUESTA, J.L. de la & San Juan, C. (2005) “Juvenile delinquency and immigration in Western Europe : Criminal Responsability versus Social Responsability”, International Annals of Criminology (43, 'm.) : 35-50. DITTMANN J., WERNITZNIG B. (2003) “Strafverfolgung und Sanktion- ierung bei deutschen und ausländischen Jugendlichen und Heranwach- senden. Eine Untersuchung am Beispiel des Einbruchsdiebstahls”, Monatsschrift für Kriminologie und Strafrechtsreform (86) : 195-205. ESTERLE-HEDiBEL, M. (1999) “La marque de l’étranger. La construc- tion d’identités délinquantes chez les jeunes d’origine maghrébine”, in J- M Bessette (ed), Crimes et cultures. Paris : L’Harmattan. FRECHETTE, M. and LE BLANC, M (1987) “Le syndrome de la person- nalité délinquante”, Revue Internationale de Criminologie et de Police Technique, 87 (2) : 133-146. GiraUD, M. (1993) “Les mineurs délinquants sur l’agglomération lyonnaise en 1990”. Research for the Groupe de Recherche sur Sociali- sation (GRS). Bron : Université Lumière Lyon2 __ 95 — LEvy, R. y ZAUBERMANN R. (1998) “La police et les minorités visibles : les contradictions de l’idéal républicain” in Y. Cartuyvels, Poli- tique, police et justice au bord du futur. Paris : L’Harmattan. MORENO RODENAS, P. (2002) “Reflexiones en torno a la segunda gen- eraciôn de inmigrantes y la construcciôn de la identidad”, Ofrim Suple- mentos [online] 10: 10-95. Available from http ://sauce.pntic.mec.es/ -smarti4/docs/n10_cap1_reflexiones.pdf [Accessed 5 June 2004] RocH£, S. (2001) “La délinquance des jeunes - Les 13-19 ans racon- tent leurs délits”. Paris : Seuil. RocHÉ, S. (2004) “Ethnicité et délinquance des jeunes en France : une question politique à la lumière des résultats d’une enquête auto- déclarée” En Revue Internationale de Criminologie et de Police Tech- nique et Scientifique (1l): 3-28. SicoT, F. (2000a) “Délinquance juvénile et immigration. Recherche exploratoire selon une méthode ethnographique”. Toulouse : Université Toulouse-Le Mirail. SicoT, F. (2000b) “Enfants d’immigrés maghrébins : rapport au quartier et engagement dans la délinquance”, in Les cahiers de la sécu- rité intérieure (42) : 87-108. ABSTRACT This research aims to study in depth the phenomena of immigrant juvenile offenders within the context of Europe. It consists of two very distinct phases : PHASE I: We show the most important social and legal aspects related with the foreign juvenile offenders in Spain, Germany, France and Italy. Furthermore, special attention will be paid to the analysis of minors who are the children of immigrants, who were born and have been brought up within the host society, their society, but who must deal with situations of discrimination and social disadvantage. PHASE II: This phase undertakes a retrospective study of crime recorded in the Juvenile Courts of the Autonomous Community of the Basque Country from January 2001 to December 2005. The analysis of the sentencing in those cases will enable us to form a comprehensive view of the crime profile of the immigrant juvenile offender in this area of Spain, which may be comparable with other European regions. — 96 — RESUMÉ Soucieux d'étudier en profondeur le phénomène des jeunes délin- quants immigrés dans le contexte européen, les auteurs ont, dans une première phase, pris en compte les principaux aspects sociaux et juridi- ques concernant les jeunes délinquants en Espagne, en Allemagne, en France et en Italie. Ils ont porté une attention spéciale à l’analyse des mineurs enfants d’immigrants, nés et élevés dans la société d’accueil, leur société, mais dans laquelle ils ont dû faire face à des situations de dis- crimination et de désavantage social. Dans une deuxième phase, une étude rétrospective des dossiers de condamnation de jeunes immigrants et de jeunes non immigrants des tri- bunaux de la jeunesse de la Communauté autonome du Pays basque de janvier 2001 à décembre 2005 permet de dresser un profil criminologique du jeune délinquant immigrant en Espagne, susceptible de comparaison avec les autres régions de l’'Union européenne. RESUMEN Preocupados por estudiar en profundidad el fenômeno de los jôvenes delincuentes inmigrantes en el contexto europeo, los autores han tenido en cuenta, en una primera fase, los principales aspectos sociales y juri- dicos en relaciôn con los jôvenes delincuentes en España, Alemania, Francia e ÎItalia. Han prestado una atenciôn especial al anâlisis de los niños de inmigrantes que han nacido y se han criado en la sociedad de acogida, su sociedad, pero en la cual han debido hacer frente a situa- ciones de discriminaciôn y de desventaja social. En una segunda fase, un estudio retrospectivo de los expedientes con- denatorios de jôvenes inmigrantes y de jôvenes no inmigrantes de los Juzgados de Menores de la Comunidad Autônoma del Pais Vasco, de enero de 2001 a diciembre de 2005, permite elaborar un perfil crimino- lôgico del joven delincuente inmigrante en España, susceptible de com- paraciôn con las otras regiones de la Uniôn Europea. __ 97 _