Victimisation and Insecurity in Europe : A Review of Surveys and Their Use [1 ed.] 9789054874959

Citation preview

Victimisation and Insecurity.book Page 1 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe A Review of Surveys and their Use

Victimisation and Insecurity.book Page 2 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity.book Page 3 Thursday, December 4, 2008 12:33 PM

Renée Zauberman (ed.)

Victimisation and Insecurity in Europe A Review of Surveys and their Use Contributions from Mike Hough, Lieven Pauwels, Stefaan Pleysier, Joachim Obergfell-Fuchs, Amadeu Recasens i Brunet, Philippe Robert, Giovanni Sacchini, Rossella Selmini, Renée Zauberman

This edited book is the outcome of a seminar held within a Coordination Action Assessing Deviance, Crime and Prevention in Europe, (CRIMPREV, www.crimprev.eu) funded by the European Commission (contract 028300 under FP6) and coordinated by the Groupe européen de recherches sur les normativités (GERN, Centre national de la recherche scientifique, CNRS, France).

Victimisation and Insecurity.book Page 4 Thursday, December 4, 2008 12:33 PM

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

Cover design: Koloriet, Leefdaal Book design: Style, Hulshout Print: DCL Print & Sign, Zelzate © 2008 VUBPRESS Brussels University Press VUBPRESS is an imprint of ASP nv (Academic and Scientific Publishers nv) Ravensteingalerij 28 B-1000 Brussels Tel. ++ 32 2 289 26 50 Fax ++ 32 2 289 26 59 E-mail [email protected] www.vubpress.be ISBN 978 90 5487 495 9 NUR 824 Legal Deposit D/2008/11.161/039 All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.

Victimisation and Insecurity.book Page 5 Thursday, December 4, 2008 12:33 PM

Contents

1. Surveys on victimisation and insecurity in Europe Renée Zauberman

7

2. Crime victims and insecurity surveys in Belgium and the Netherlands Lieven Pauwels, Stefaan Pleysier

39

3. Research on victimisation and insecurity in Britain Mike Hough

65

4. Research on victimisation and insecurity in France Philippe Robert

87

5. Crime victims and insecurity surveys in Germany Joachim Obergfell-Fuchs

105

6. Research on victimisation and insecurity in Italy Giovanni Sacchini, Rossella Selmini

127

7. Victimisation surveys in Spain and Portugal Amadeu Recasens i Brunet

149

8. On the authors

167

5

Victimisation and Insecurity.book Page 6 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity.book Page 7 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe Renée Zauberman

In the framework of the 6th FPRTD, the European Commission financed a coordination action, Assessing Deviance, Crime and Prevention in Europe (CRIMPREV). This coordination action is made up of six thematic workpackages, one of which is devoted to problems of Methodology and Good practices.

I.

Introduction

1.

The Methodology and Good Practices workpackage

The objective is to come up with a listing of the significant implementations of instruments of knowledge about crime and of their uses. We shall therefore • map the situation in Europe, • identify the good – as well as the bad – practices, and • specify elements of comparison within the European zone. The last half-century witnessed the appearance of powerful new instruments for knowledge about crime. Their particularity consists of emancipating research from institutional data in which the study of crime has traditionally been shut away. Not only have these instruments renewed – at least partially – the scientific knowledge we had of crime but, even more, they can provide important help in decision-making. For all that, their introduction has taken place quite variably in the various countries of Europe. In addition, the mastery of these instruments is rather unequally divided up insofar as the number of confirmed specialists is limited; consequently, the uses made of them are more or less relevant. Finally, non-scientific users often have only limited knowledge of these tools’ potential. Consequently, there is room in CRIMPREV for a workpackage devoted to listing the most significant implementations of these methods in the principal European countries and the use made of them… in the hope of

7

Victimisation and Insecurity.book Page 8 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

perfecting and distributing both states of knowledge and catalogues of good practices. The responsibility for the workpackage is shared by two scientific centres, the French Centre de recherches sociologiques sur le droit et les institutions pénales (CESDIP) and a Barcelona group, a European network of local municipalities concerning security, the European Forum for Urban Security (FESU) and a regional coordination organism of local security programmes, Città sicure, in order to properly place it in the interface of the academic world and that of potential users. Four methods have been selected: • surveys in the overall population on victimisation and insecurity; • surveys in the overall population on self-reported crime; • the comparison of survey data and that coming from institutional sources such as police statistics; • evaluative research of safety policies. For each of these methods, we proceed according to the same protocol in five phases: 쑱 Phase 1: elaboration of a report grid; choice of a general rapporteur in charge of synthesising the information gathered; and rapporteurs – a half-dozen per topic – in charge of drawing up the state of knowledge and uses in different countries where the method is sufficiently developed. Of course, given problems of availability of the prospective experts, we cannot cover them all. We are not aiming at exhaustiveness but rather at presenting a reasonably representative selection of what is being done in the European zone, especially in the main countries. 쑱 Phase 2: each rapporteur writes up a review concerning the country or zone of which he/she is in charge and these documents are then circulated among all rapporteurs. 쑱 Phase 3: meeting of all rapporteurs, national and general, with the promoters of the workshop in a seminar where the reports are presented and discussed. 쑱 Phase 4: the general rapporteur writes a synthesis of the reports and discussions. 쑱 Phase 5: validation of this document by the workpackage promoters and dissemination in the form of a 50-page booklet in English and French. 쑱 Phase 6 : editing all contributions for publication in two volumes, one in English – presentend here – and the other in French.

8

Victimisation and Insecurity.book Page 9 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

2.

The workshop devoted to surveys on victimisation and insecurity

We present here the synthesis of reviews devoted to the first workshop, pertaining to surveys on victimisation and insecurity. The team for this first workshop was comprised of: – Mike Hough (King’s College, London) for England and Wales, – Lieven Pauwels (RijksUniversiteit Gent) and Stefaan Pleysier (Expertisecentrum Maatchappelijke Veiligheid KATHO University College associated with the Katholieke Universiteit Leuven) for the Netherlands and Belgium, – Joachim Obergfell-Fuchs (Kriminologischer Dienst Baden-Württemberg) for the Federal Republic of Germany, – Amadeu Recasens i Brunet (Universitat Barcelona) for Spain and Portugal, – Giovanni Sacchini (Città Sicure) and Rossella Selmini (Città Sicure and Università degli studi di Macerata) for Italy, – Philippe Robert (CNRS/CESDIP) for France, – Renée Zauberman (CNRS/CESDIP), general rapporteur. The national rapporteurs were commissioned by the workpackage’s steering committee in July 2006 and handed in their reports at the beginning of 2007. Then a three-day seminar brought together the promoters, general rapporteur and national rapporteurs for a presentation and discussion of the reports, so as to draw upon them for feeding transverse lessons into the general report. This final synthesis thus depends on the national reports as much as on the seminar discussions.

3.

Origins and development of surveys on victimisation and insecurity

When crime begins to constitute a social problem – in Europe of the second half of the 18th century – the problem of its measurement emerged immediately and it was spontaneously expressed in terms of counting courtroom activity. From the absolutist period of Enlightenment on – Archduke Peter-Leopold commissioned a survey on the activity in Tuscan courts before the adoption of the Leopoldina1; Councillor Montyon carried out another survey of convictions in the jurisdiction of the Paris 1.

Da Passano (1995).

9

Victimisation and Insecurity.book Page 10 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Parliament2 – but it was at the very beginning of the 19th century that counting became systematic. At a varying pace depending on the country, the most serious sentences at first, then all sentences, then all prosecutions and finally commitments to prison were subjected to reckoning. The French Compte général de l’administration de la justice3, published regularly as of 1827, would become the reference for European debates on measuring crime throughout the 19th century. This statistical production served both to count the activity of the courts and to measure crime4. From the beginning, however, doubts were raised about the relevance of the latter use: did all situations open to being qualified as criminal come to be known to the judge in charge of determining which concrete cases corresponded to the abstract incriminations of the Law? Clearly not – for a number of reasons. How then to be content with data that were limited to listing the cases known to the judge or his auxiliaries, the public prosecutor or police, in charge of passing them on to him? It was for a long time assumed that the available figures ‘represented’5 reasonably well all the cases that, for one reason or another, had eluded the judge. At very most, the choice was gradually made to use the data produced as upstream as possible in the institutional process on the grounds that those were closer to the commission of the crime and therefore, their informational loss was lesser. Following the advice of Thorsten Sellin (1931), police records rather than judicial reckonings became the source of crime knowledge par excellence. However, persistent doubts as to the sufficiency of those institutional reckonings for knowing crime led to looking for terms of comparison in the non-criminal records of the same events. Thus the police counts of homicides were compared to the statistic of causes of death held by the health agencies of all developed countries; or the statistics on shoplifting with inventory shortage, i.e., the difference between a physical inventory and an accounting inventory… This effort to find terms of comparison to crime statistics in other sectors reached its peak in the studies on the costs of crime, already encouraged by the National Commission on Law Observance and Enforcement (1931), called the Wickersham Commission. However, in spite of all the ingenuity thus expended, the remedy could

2. 3. 4. 5.

Lecuir (1974). Perrot, Robert (1989); Robert (1993). Coll. (1998). In the sense that a limited population ‘represents’ a parent population when it reproduces its most specific characteristics in a sufficiently accurate way. In the 1840s, Quetelet (1848, reprinted 1984) asserted establishing, in certain conditions, this representativeness of crime statistics in relation to the crime committed.

10

Victimisation and Insecurity.book Page 11 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

apply only in the limited number of cases where it turned out to be possible to find terms of comparison. In addition, it often allowed for only highly approximate comparisons between overall orders of magnitude. Meanwhile, crime statistics remained the major tool of knowledge about crime; but doubts as to their relevance swelled at a moment when public safety policies were loosing ground in face of an unprecedented rise in petty property crime, concomitant with the entry into consumer society6. Schematically, it was realised that the chances of recording an event depended on the victims’ propensity to inform the official services (reporting), the priority that the latter attached to its detection and the ease or difficulty of this detection (visibility of the author and/or his act), and finally of the willingness and ability of these institutions to handle this case effectively7... so that the ability of the official figures to ‘represent’ behaviours committed is an empirical and, moreover, variable8 matter. It appeared less and less satisfying to construct knowledge about crime on such uncertain bases. So that producing ad hoc data instead of limiting themselves to borrowing that of the administrators was the alternative researchers finally came up with. How to carry out large-scale surveys on crime? Schematically, two approaches were tried successively: first, questioning population samples to know who had committed such and such an offence; later, inversely, asking who had been the victim. The Victimisation Surveys methodology, was inaugurated in the 1960s by experiments carried out by Al Biderman, Philip Ennis and Al Reiss Jr for a presidential commission on crime, called the Katzenbach Commission (The President’s Commission…1967)9. Its fortune was subsequently extraordinaire. No other kind of crime research had mobilised, and in a lasting way, so many specialists, funds and methodological ingenuity. For the first time, criminal science built autonomous data bases on a large scale. After becoming routine in the United States since the 1970s, and in England and Wales a decade later, the production of victimisation surveys also intensified – at a less systematic pace, nonetheless, and to varying degrees – in different Western European countries... Since the end of the Eighties, there has also been an ambitious programme of international victimisation surveys (ICVS) launched by Jan Van Dijk, Patricia Mayhew

6. 7. 8. 9.

On this crisis in security policies, cf. Robert (1999). Cf. for example Bottomley, Coleman (1976); Robert (1977); Bottomley, Coleman (1981). Cf. for example Robert (1991). Cantor, Lynch (2000).

11

Victimisation and Insecurity.book Page 12 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

and Martin Killias10. This kind of survey is used not only on the national or supranational levels, but the localisation of public safety policies has also led to implementing them at more local stages. New survey techniques were made use of to nurture the various local, national, and even international ‘observatories’. Nevertherless, the practice of surveys on victimisation and insecurity is not at all homogeneous in Western Europe and, on the contrary, varies considerably from one country to the next, in quantity as well as in quality. Even though such surveys exist, the use that can be made of them also reveals considerable differences from one instance to another. Therefore, it is useful to compare, in these respects, the situation in a certain number of European countries. We could not select them all, and not all of them present the same interest for the subject. Moreover, we sought contrasted situations with respect to the recourse to those surveys. We also wanted to select a range of countries corresponding to different legal and institutional traditions. We finally decided on the bloc of founding members of the EEC to which were added England and Wales – where recourse to these surveys is long-standing and quite developed – and the Iberian countries. The eight countries reviewed make up an uninterrupted territorial entity representing 62% of the population of the Union in its current format. They also represent four fifths of the (ten) countries whose research centres are participating in the coordinated action. We regret only having not succeeded in introducing into this range at least one Scandinavian nation, owing to the early development of victimisation surveys in some of them like Sweden. On the basis of the analysis grid proposed to the national rapporteurs, we shall first attempt to describe the carrying out of victimisation and insecurity surveys and its diversity. In a second part, we shall discuss the uses that are made of them by the different categories of users.

10. Van Dijk, Mayhew, Killias (1990); Frate, Zvecic, Van Dijk (1993); Van Dijk, Mayhew (1992); Mayhew, Van Dijk (1997); Van Kesteren, Mayhew, Nieuwberta (2000); a specifically European version was developed in 2005 by the Gallup Institute with financing from the European Commission (Cf. Van Dijk, Manchin, Van Kesteren, Nevala, Hideg [2007]).

12

Victimisation and Insecurity.book Page 13 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

II.

Conducting surveys on victimisation and insecurity

The diversity of practices and the lessons that must be drawn from them can be illustrated with three characteristics of the surveys: – their scope, from victimisations to the feeling of insecurity and fear of crime; – their concentration or dispersal – their degree of soundness.

1.

The scope

The situation is quite contrasted depending on whether one considers victimisation or the sense of insecurity. Victimisation The range of events covered is more or less the same everywhere: vehicle crime, from thefts to vandalism, burglaries, thefts, violence, damage to property. Most of the surveys agree on distinguishing between those that affect all the members of a household or living unit (vehicle crime, burglaries) and those that are more personal (thefts, violence); nonetheless, this distinction has practical importance mainly for surveys designed to interview more than one individual per household. Of course, these kinds of events are looked into with more or less details: • While most surveys distinguish between the theft of and theft from vehicles11, not all, on the other hand, dwell on car vandalism; some make explicit the theft of two-wheeled vehicles, in other cases, we do not know clearly whether these are taken into account. • Some surveys are limited to residential burglaries; others include those in other places (such as holiday homes or business premises). • Some make a distinction between a wide variety of thefts, generally according to their location, whereas others are more succinct. • Finally, damage to other property than cars is not always taken into account, or only partially. • Violence constitute the category that poses the most problems owing to the large diversity of situations that the term can cover; the basic

11. This was not, however, the case of French national surveys on household living conditions (Enquêtes permanentes sur les conditions de vie, EPCV).

13

Victimisation and Insecurity.book Page 14 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

problem is that of sexual violence12 or else violence by intimates that can either be passed over, or included explicitly in the questioning on assaults or else, be the subject of specific questions. With a general question on violent victimisation, one never knows whether the respondents will mention these particular kinds of violence; with specific questioning, on the other hand, there is more chance of seeing them emerge. The reserve that often accompanies their acknowledgment sometimes has entailed the adoption of particular interviewing arrangements: the telephone survey is often presented as being more discreet than a face-to-face interview except by using a CASI procedure for the latter (the interviewer gives his or her laptop to the person being interviewed, asking him/her to type the answers). These variations have important consequences: the events that the survey tries to grasp are often of minor importance and have happened in a more or less distant past; experience has shown that their recollection is stimulated more by detailed interrogations than by overly-general questions. However, an excess of factual details can weary the respondent and reduce the reliability of the responses. Moreover, concern about the optimal size of the questionnaire often obliges restricting the interrogations on the subjective repercussions of the victimisations although these provide valuable information, or those on the context of daily life which are often cited as key variables for understanding the level of victimisation and its impact. Nonetheless – besides observing a trend towards standardisation of victimisation questions13 –, the range of events likely to enter into this type of survey cannot be expanded much further. Certainly, some reports mention attempts at studying shoplifting in small shops, some frauds, notably consumer fraud or business crime, without overlooking the multiple incidents of school life that specific surveys generously detail… But these attempted enlargements – which sometimes encounter serious wording obstacles – do not go very far. They quickly run up against a double barrier that is difficult to surmount: – questions on victimisations should be phrased in a language that is likely to be understood, in approximately the same way, by all respondents; – above all, someone has to be likely to consider him- or herself the direct victim of the offence, which is not the case with victimless crime or indirect victimisation14. 12. Of course, surveys focussing on violence against women generously detail the variety of sexual victimisations. 13. Cf. for example the Obergfell-Fuchs report for Germany (this volume). 14. Nor, moreover, when victimisation makes the victim disappear as in the case of successful homicide.

14

Victimisation and Insecurity.book Page 15 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

Although the range of victimisation cases likely to enter into surveys is somewhat clear, and although the standardisation of these interrogations is not out of reach15, conversely, the picture is much more confused when it comes to insecurity. Insecurity All the victimisation surveys also include questions on insecurity but in most of the reviewed countries16, there are many other surveys devoted specifically to this latter topic. They are less complicated to put together than victimisation surveys and also less expensive; finally, they do not compete with the traditional forms of crime measurement like public crime statistics and therefore do not elicit the same reticence from their administrators. This abundance is not, for all that, synonymous of wealth of information. Most of the rapporteurs have voice highly critical assessments of these investigations. On the one hand, the questions used are rarely standardised, and one often has trouble weighing their significance. On the other hand, even for the standardised questions, thinking about insecurity has not progressed sufficiently for dispelling all ambiguities in responses. In the United Kingdom, sizeable studies17 have recently re-examined this whole field from every angle and important research can also be noted in the Netherlands and Belgium18 and a heated controversy in Germany19. A tremendous effort has to be accomplished if we want the multitude of polls on insecurity to yield ubstantive additional knowledge.

2.

Concentrating or dispersing surveys

The countries reviewed in the CRIMPREV seminar display a good picture of the variety of European situations: sometimes one single survey dominates the whole landscape; elsewhere, this landscape is much more fragmented, and finally, a series of intermediary instances can be observed. 15. Nonetheless, the balance between factual questions and questions of opinion can vary considerably within the same victimisation module depending on whether the survey’s focus is more on ‘measuring’ crime or on the analysis of the respondents’ actual victimisation experience. 16. Cf. for example the Sacchini, Selmini report for Italy, Robert for France, as well as Recasens I Brunet for the two Iberian countries (this volume). 17. Cf. for example, in a very long list, Ditton, Farrall (2000) or Jackson (2005). 18. Pauwels, Pleysier (2005); Pleysier, Pauwels, Vervaeke, Goethals (2005); Venderveen (2006). 19. Obergfell-Fuchs report for Germany (this volume).

15

Victimisation and Insecurity.book Page 16 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

The Netherlands, Belgium and England and Wales well represent one extreme of this palette. Although the English rapporteur does mention a few local experiences20 – whose impact have often been notable as much in the public debate as in the advancement of scientific knowledge – they are hardly numerous, often old, and the British Crime Survey (BCS) of the Home Office Research, Development, and Statistics Directorate occupies an increasingly hegemonic position. In the Netherlands, two national systems – that of the Central Statistics Bureau, which, since 1980, has followed the experiments of the Justice Ministry’s research department, and, above all, the massive Politiemonitor Bevolking (PMB) of the Ministry of the Interior, created in 1993 – dominate the field at all levels, from national to local. Their planned merger is likely to further reinforce their hegemony. On a lesser scale, the situation is quite comparable in Belgium: local experiments run by the universities of Gent and Liege have been followed by the Ministry of the Interior’s (large) Security Monitor, copied after its Dutch homologue and present at the federal and local levels21. To this model should be attached the Italian case: most victimisation surveys are carried out by the Istituto nazionale di statistica (ISTAT) but, in fact, the mechanism is more recent and much less solidly established; it does not correspond to the investment of a central actor that would be in a hegemonic position since only a few regions – primarily Emilia-Romagna through Città sicure – and a few cities have really entered the field and they are the only real users of the surveys. Inversely, federal surveys are rare and sporadic in Germany22 and, for the most part, found at the urban level at the initiative either of research institutes or, more recently, local governments in the framework of their crime prevention and safety programmes. At the level of the Spanish Estado central, only a few old attempts of the Centro de investigaciones sociologicas (CIS) are mentioned meant to carry out victimisation surveys whereas, since 1983, Barcelona has developed a system of annual surveys that was enlarged in 1999 to include all of Catalonia. In Portugal, the experiments undertaken by the Ministry of Justice in the early 1990s do not seem to have prospered; henceforth, it is the participations in the ICVS that play the role of nationwide surveys. A surveys mechanism has been developed by an Observatorio permanente de Segurança run by the University for the Municipality of Porto23. 20. Sparks, Genn, Dodd (1977); Smith, Gray (1985); Kinsey (1984); Jones, MacLean, Young (1986). 21. At least to that of the towns that signed a security contract with the federal Ministry of the Interior. 22. Often linked to reunification and a concern for comparison between old and new Länder or else to specific populations like women. 23. Agra, Quintas, Fonseca (2001); Agra, Queiros (2007).

16

Victimisation and Insecurity.book Page 17 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

The French situation is intermediate: after the isolated experiment of a national survey of academic initiative in the mid-1980s, the National Institute for Statistics and Economic Studies (INSEE) has been developing, since the mid-1990s, an annual survey on the living conditions of households including a module on victimisation; it has been recently considerably overturned and remodelled at the initiative of the Ministry of the Interior. But this survey is hardly in a hegemonic situation: there are thematic national surveys, often more solid (albeit less regular); there is still a large regional survey with regular sweeps in Île-de-France and a fair number of local surveys (of which the development prospects are not yet clear). When all is said and done, the range of victimisation surveys that appears in the countries under study can be summed up as follows: • general surveys o ➠ national, o ➠ regional, o ➠ local; • thematic surveys concerning particular populations (young people, women…). Obviously, the entire range is not present in all the countries under review. Moreover, certain national surveys are constructed so as to ensure a regional or local representativeness, either directly – as is henceforth the case for the BCS, whose results must be representative down to the 43 police forces level, or for the Dutch and Belgian security monitors – or by optional enlargement of the sample (in favour of certain regions or certain cities) as with ISTAT surveys. The EU ICS has provided for an automatic enlargement for the capitals of the countries included in the campaign. To this must be added the International Crime Victim Survey (ICVS) which covered, depending on the sweep, a variable number of countries24. Reading national reports reveals very low use of these international surveys, essentially owing to the small size of samples that cannot compete with those of national surveys. The Netherlands are an exception: the interest shown in international surveys is probably due to the central role played by this country and by Jan Van Dijk in the creation of the ICVS. A specifically European version (EU ICS) was developed in 2005 by the Gallup Institute with financing from the European Commission. Even before, elements on victimisation and insecurity were to be found in the 1996 Eurobarometer (44.3) and in the European Social Surveys of 2002, 2004 and 2006. Behind this variety of arrangements, stand variable sources of funding. 24. Certain times, only a country’s cities are the object of the survey.

17

Victimisation and Insecurity.book Page 18 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Generally, for the most part, they are at the national level: most often, the Ministry in charge with internal affairs (in England, the Netherlands, Belgium, in France recently), sometimes the Ministry of Justice (in the Netherlands and in France initially, Portugal), specific ministries for thematic surveys (in Germany, France, Spain, Italy for women), in certain cases, national statistic institutes (France, Italy). By no means insignificant contributions from regional or local governments (France, Catalonia, Belgium, England initially; Germany, Portugal) and finally by research funding agencies (in Germany, England and France initially) should also be mentioned. The high cost of these surveys is a major issue: they require large samples to avoid overly open confidence intervals, especially for infrequent types of victimisation. While national and even regional governments can easily assume such costs, the charge often turns out to be much heavier for municipalities or even large urban areas – with the exception of Barcelona, however, which has financed an annual survey for more than twenty years. The extent of costs has two kinds of consequences: on the one hand, it can stand as an obstacle to resorting to victimisation surveys by local governments which, however, often would need them; on the other hand, the hegemonic position of state or regional administrations can incline them, in certain countries, to limit the dissemination of the data to the scientific community or to filter beneficiaries. By contrast, surveys devoted specifically to insecurity, as they are less costly, have a much more diversified range of sponsors, even including the media or polling institutes.

3.

Varying soundness

Various criteria can be used here: the sample, the instrument, the administration of the questionnaire, the recall period. The sample The larger it is, the less the confidence intervals will be open and the more reliable the results will be. From this point of view, the Politiemonitor Bevolking (PMB) probably constitutes a peak with its 88,000 respondents (out of a population of 16 million25), but the Anglo-Welsh BCS attains a high level especially since it 25. In the purity of principles, one should disregard age classes that are not submitted to the inquiry. As these are always more or less the same (the under 14 or 15) and their share not differing much from one country to another, one can settle for overall populations as an proxy indicator.

18

Victimisation and Insecurity.book Page 19 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

increased to 47,000 respondents (for a population of 53 million); the same goes for the Belgian security monitor (40,000 respondents for a population of 10 million) or for the Catalan survey (approx. 14,000 for a population of 6 million) and even the surveys of the Italian ISTAT (20,000 households resulting in 60,000 respondents26 for a population of 58 million). In France, on the contrary, the samples of INSEE surveys are fairly low (6,000 households giving some 11,000 respondents for a population of 59 million in continental France). The recent Ministry of the Interior involvement granted only limited priority to enlarging the sample (rising to some 13,000 households and 26,000 respondents) preferring to concentrate its efforts on the reformatting of the questionnaire. It is on the side of regional surveys in Île-de-France (11,000 households and individuals for a population of 11 million) and Baromètres Santé (30,000 for the 2005 edition) that one finds samples of sufficient size. The instrument As has already been pointed out, overall, the questions about insecurity are not considered terribly satisfying. For victimisation, everything depends on the scope of the interrogations: in general, the surveys ask questions about the circumstances of the event and whether or not victims have reported the incident to the police27. The impact – especially subjective – of victimisation is included less systematically and more or less skilfully. Finally, the quality of the instrument also rests on its ability to steer between two contradictory requisites: the interview must not last too long for fear of wearying or exasperating the respondent; and yet the questionnaire must not only delve into the factual details of the victimisation incident but still leave enough room for more subjective information on the way the incident was actually experienced and on its context as well (for example, on neighbourhood characteristics and problems), failing which the exploiting of the survey will be reduced to dry counting, ultimately not particularly enlightening. Administering the questionnaire Sending a mail questionnaire – the least expensive technique – seems to be widespread only in Germany where many surveys have been carried out by 26. Besides the enlargements commissioned by some regions (11,000 for Emilia Romagna). 27. Although most surveys take a more or less detailed interest in reporting to the police, reporting towards other participants (insurance, health services, security equipment suppliers…) is less systematic.

19

Victimisation and Insecurity.book Page 20 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

university institutes with limited means. The principal disadvantage is the response rate: not only is it low, but it is also quite variable in the different population segments. Face-to-face interviewing – increasingly often with the assistance of a computer (CAPI) – constitutes the most expensive technique. The Home Office resorts to it, as do INSEE, one of the Dutch national surveys and certain German surveys (those having sufficient funding). Their quality depends on that of the available network of interviewers. The multiplication of anti-intrusion systems in homes is pointed out by the English report as an increasingly significant problem. Computer-assisted telephone interview (CATI) is currently the most widely spread: in Scotland and France for the regional and local surveys and for most of the national thematic surveys in Catalonia, Italy, Belgium, for the Dutch Politiemonitor Bevolking, and Portugal). Its success is basically due to a consistent savings compared to face-to-face. On the other hand, it is jeopardized – if not by the development of subscriptions in unlisted numbers that are easily got round by random digit dialing – by the multiplication of households equipped only with mobile phones (so-called ‘mobile-only’28). Respondents are in fact drawn from lists of landlines subscribers, and there are no similar lists for mobiles. Drawing up complementary random samples of ‘mobile-only persons’ is not absolutely impossible29, but it runs up against a great many limitations. In addition, the mobile phone allows for much shorter interviews than those conducted over fixed telephones. We must still mention that surveys on the victimisation of students or schoolchildren are generally administered in classrooms in the form of a printed questionnaire that is filled out under the supervision researchers and/or teachers. Here, the main problem is parental authorisation, which is often considered a necessary precondition, especially for the youngest schoolchildren. A similar method – without the same authorisation problems naturally – is used for the surveys linked to military service (in Switzerland) or the defence call-up preparation day (ESCAPAD surveys in France), but the progressive abandoning of armies of conscripts in favour of professionals does not allow for envisaging much development of this case. 28. The Belgian report estimates their proportion at 15% of the households (similar to what is put forward for France; see also, with the EUICS, Hideg, Manchin, 2005). Several specific studies (Pleysier, Vervaeke, Goethals [2006]; Beck, Arwidson, Firdion, Jaspard, Grémy, Warszawski [2001]; Beck, Legleye, Peretti-Watel [2005]) placed the accent above all on the particularities of these ‘mobile-only population’ (younger, better educated…). 29. Beck, Guilbert, Gautier (2007, 29-31, 34-37).

20

Victimisation and Insecurity.book Page 21 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

English and German reports draw attention to the disturbing decline in response rates: the BCS has managed to maintain it at 75% in the last decade, but the increase in polls of all kinds ends up wearying the solicited population. Moreover, adding to this difficulty are further obstacles to specific administration methods such as the growing number of households with no fixed landlines or the proliferation of devices preventing entrance to residential buildings. In this respect, the German report remarks that the few rare telephone surveys conducted in the country did not give higher response rates than the average rates of its postal surveys. As a matter of fact, the crucial problem is not so much the response rate as the bias aroused by unevenly distributed non-responses30: the seeming preservation of a constant response rate masks, for example, in England and Wales, a localised drop precisely in the areas with a high crime rate31. The recall period The shorter the recall period, the fewer problems of memory loss or of telescoped time frames, thus making findings all the more reliable. The standard period is a year (England and Wales, Catalonia, the Netherlands, Belgium, Italy, Germany), with some alternate, longer periods such as two or three years, five years or entire lifetime (Germany, ICVS, Italy), particularly in surveys on specific populations. In France, however, INSEE surveys question over two years, and the regional and local surveys generally over three years. Finally, the quality of the surveys depends considerably on the ability of the commissioning bodies to secure the backing of proficient scholars. This point will be considered in the following section where we will deal with different uses of victimisation surveys.

III. The uses of surveys on victimisation and insecurity As for what uses are made of these surveys, one may first consider those in the vein of public policies, then those concerning the media, and finally those taking place in the academic world. Beforehand, a word must be said about the dissemination of results and their publication.

30. Pauwels, Pleysier (this volume). 31. Hough (this volume).

21

Victimisation and Insecurity.book Page 22 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

1.

Surveys and public policies

Mobilisation of surveys results in the process of decision-making and evaluation of public policies ranges from intensive to low. It is systematic in England and Wales: not only does the BCS seem to have become the ordinary reference for measuring crime but, moreover, its results are used systematically in support of the evaluation of policies developed by the Labour government following, for example, the Crime and Disorder Act of 1998. From them are extracted indicators of the crime level, confidence in the police and justice, and victim and witness satisfaction. Table 1, in Mike Hough’s report for England and Wales (this volume), shows the importance of indicators borrowed from the BCS (italicised) in evaluating the compliance with objectives of public security policies. This is even truer starting from the moment that the sample was increased so as to ensure the representativeness of the survey at the level of each regional police department (following the Police Reform Act of 200232). Some analysts33 were even inclined to detect a certain tension between this managerial culture that imposes a highly centralised grid of indicators and a ‘hands off’ management style which allows a good deal of discretion for individual decisions to local agencies. The English report also mentions several cases of direct influence of survey results on discussion and public policies. The two Merseyside and Islington surveys fostered the resistance of the Labour municipalities to the Conservatives’ safety policy and, over the longer term, contributed to forging the Labour criminal policy. The Policy Studies Institute survey fuelled the debate on the policing strategies of the London police. The discovery of the emblematic role of certain ‘signal crimes’ in producing fear of crime seems to have retained the attention of police chiefs and, through them, of political leaders, and contributed to the definition of specific programmes. In the Benelux, the situation is somewhat comparable to that of the BCS but less systematic: in Belgium, the Security Monitor is linked explicitly to the local security contracts passed between the federal state and towns; as for the Politiemonitor Bevolking, it constitutes an integral part of the police organisation. In Catalonia, the survey is part of the Generalitat’s statistics plan but its concrete use does not seem clear. The Ajuntament de Barcelona clearly used it in an early phase to assert itself as the sole actor possessing precise information on safety in the agglomeration and to determine safety policy 32. Hope (2005). 33. Cf. for example Crawford (2001).

22

Victimisation and Insecurity.book Page 23 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

priorities. In a more recent phase, its concrete use seems less obvious. As for the Estado central, it did not seem to invest the production of this kind of data, settling for insecurity surveys that tended to be, moreover, unsystematic. In France, it was just recently that the results of national surveys were integrated into national ‘observatories’ or research institutes such as those on destitute urban areas (ONZUS, Observatoire national des zones urbaines sensibles) and, even more, on crime (OND, Observatoire national de la délinquance). Nonetheless, it does seem that they occupy in fact the rather secondary position of a junior partner beside police statistics. The Île-deFrance regional surveys appear in an observatory at the disposal of the Regional Council whose use is not well known. As for the local surveys, they have often been commissioned to integrate nascent local observatories; but their heads – especially accustomed to handling administrative figures coming from the police, school system, housing projects, fire department, etc. – do not easily manage to make use of data as distinctive as general population survey results. We find no notable impact of the surveys on public policies in Germany – except sometimes, notes the rapporteur sarcastically – to justify savings measures. However, it must be pointed out that most of the recent local surveys have been commissioned by municipal authorities in support of prevention and security programmes, without fully knowing their concrete use. In Italy, finally, no use was found at the state level, but certain regions (primarily Emilia Romagna, which has provided itself with a specific study tool in Città sicure) and a few municipalities (particularly Modena and Bologna) use fairly regularly regional or local oversampling that they request from ISTAT. To sum up, the national states and certain regions (such as Catalonia, Emilia Romagna, Île-de-France) integrate – more or less – surveys for the piloting of crime prevention and safety policies. This is also the case with a certain number of cities and agglomerations, but this time with much greater uncertainties, both about funding – which seems to have been easily endured by Barcelona for the past quartercentury, but which weighs heavy on other budgets – and about the abilities for integrating such indicators. Certain rapporteurs (Germany) seem sceptical as to the development of surveys at this level of government; others, on the contrary, deem it assured. We can cite the case of the Belgian security monitor in the cities under security contracts, but this time with steering and funding from the federal state. It does seem that, in the long term, the fortune of such local contracts presupposes participation of regional or national governments in the financing and implementation 23

Victimisation and Insecurity.book Page 24 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

techniques of these surveys. Nonetheless, such participation presupposes choices: determining which towns will benefit from them and according to what criteria is crucial, an issue all the thornier since the towns most affected by crime are often most lacking in resources for funding surveys. On the other hand, thematic surveys on specific populations – violence against women, young people, especially schoolchildren – seem to get coverage everywhere and often have notable impact, as has been seen with the three Spanish surveys on violence against women. Similarly, the German report emphasises the large impact of surveys on school violence in pedagogic milieus and educational managers.

2.

Surveys and media

It is fairly difficult to get an overall idea of the media coverage of surveys across the various countries under review. It seems to be closely tied to the news: if a survey is conducted when insecurity is publicly debated – e.g. during an election campaign, or when a spectacular incident occurs (such as serious violence in schools) – or if it deals with a sensitive topic such as gender violence – it may receive extensive coverage. But such attention seems to be fleeting and sporadic rather than systematic. The situation also differs depending on whether the victimisation surveys have become the ordinary source of information on crime or are still secondary to police statistics. It is much more frequently publicised in the former case than in the latter.

3.

Surveys and the academic world

In this respect again, the situation is quite variable depending on the country. In Germany, not only are a large number of surveys of academic origin, but also a system without strong university involvement seems unimaginable to the point that one of the reasons that allegedly slowed down the setting up of a regular national survey was the difficulty in determining which academic institution would run it. In Belgium and the Netherlands, academic involvement also seems strong: not only did university centres initiate the first surveys, but a large number of these still carry out secondary analyses of survey data. It is noteworthy that the Belgian Federal Police, which commissions the surveys, puts aside special funds to encourage academics to work on this kind of data34. 34. Of course, these also seek funds for their work on surveys from agencies specifically devoted to financing university studies such as the federal Belgian Science Policy.

24

Victimisation and Insecurity.book Page 25 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

In England and Wales, not only have local surveys been of academic origin, but very important secondary analyses have been conducted by scientific teams receiving data and financing from the Home Office. In Italy, on the contrary, university involvement seems slight35 – even though the ISTAT national survey was designed by an academic – and it is study centres dependent upon regional governments (Città sicure or CISIS) that ensure the main scientific involvement. It appears even slighter both for the Spanish surveys and for the experiments of the Portuguese Ministry of Justice. In France, while certain surveys are run by scientific bodies and consequential secondary analyses are conducted, on the other hand, access to data is in some cases difficult for researchers – from the Île-de-France Regional Council, for instance – and, in general, there is a certain distance between the academic world and the bodies commissioning surveys. However, good cooperation between academics and the public statisticians from the INSEE who are running the national survey should be mentioned. In all cases – even in the countries that seem fairly well off, such as the United Kingdom, the Netherlands, France and Belgium –, there are unfortunately insufficient numbers of scholars fluent in quantitative methods capable of developing secondary analyses on these data and combining their outcomes with the state of knowledge in criminal sociology. As an illustration, the box below gives a few examples of secondary analyses of victimisation and insecurity surveys data.36 Examples of research based on the secondary analysis of survey data We can first note methodological studies such as Gabry Vanderveen’s analysis of insecurity indicators and their significance (2006) as well as that of Jonathan Jackson on the validation of new insecurity measurements (2005). Lieven Pauwels and Stefaan Pleysier (2005) studied the cross-cultural validity of insecurity measures. Stefaan Pleysier, Lieven Pauwels, G. Vervaeke, J. Goethals (2005) analysed the invariance of estimates in surveys on insecurity that use complex instruments such as factorial analyses or scales. Helmut Kury (1994)36 studied the impact of the wording of questions in surveys. R. Schnell and F. Kreuter (2000) analysed cases in which very similar surveys produce different results.

35. However, one mentions Transcrime of the universities of Trento and Sacro Cuore in Milan. 36. Cf. also Kury et al. (2000).

25

Victimisation and Insecurity.book Page 26 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Other research has concentrated on victimisation and the victims. Robert J. Sampson and Byron W. Groves (1989), like Nicolas Herpin and Hugues Lagrange (2005), as well as Tim Hope (2006), analysed the social and territorial distribution of victimisation and insecurity. E. Stephan (1976), like H.D. Schwind et al. (1975, 2001) has used German victimisation surveys to test the theory of social disorganisation. Johan Van Wilsem (2003) carried out a multilevel (individual, neighbourhood, city, country) analysis of the effects of context on victimisation. Ken Pease’s research (1993, 1998) led to really ‘discovering’ the depth of repeated victimisation. Renée Zauberman, Philippe Robert and Marie-Lys Pottier (2004) constructed profiles of victims and victimisation by combining circumstances of the incident, ways of living through it and of reacting to it. Some are focused more on insecurity. Karin Wittebrood (2001) carried out a multilevel analysis of the feeling of insecurity in industrialised countries; Philippe Robert and Marie-Lys Pottier (2004) analysed the change in security concerns of the French middle classes in the late 1990s. We can also mention research concerning the police. Wesley Skogan (1994, 2007) studied police-public contacts and the evaluation of police performances, based on results of the BCS; similarly, Philippe Robert, Renée Zauberman and Marie-Lys Pottier (2003) analysed, starting from surveys in Île-de-France, the reporting by victims to the police from the angle of a face to face encounter between laymen and professionals. Some research has focussed on attitudes towards criminal justice. Mike Hough and Julian Roberts (2007) studied, citizens’ confidence in it and attitudes towards sentencing, based on the BCS. Cooperation with the academic world crucially keeps bodies commissioning and using surveys to misevaluate their contributions and limits. It is also important for the designing of surveys. Without in-depth experience of secondary analyses and sustained acquaintance with sociological literature on crime, the elaboration of protocols can run up against serious difficulties. Pauwels and Pleysier (this volume), for example, stress that the importance of the instrument’s stability is often neglected: a varying instrument makes the analyst unable to tell whether a variation in the results is to be attributed to its instability or to a real change in phenomena that it pretends to be measuring. From this they deduce that an alteration – however minimal – should always be preceded by research allowing for balancing out the gain procured by the change and the loss resulting from the break in the continuity of observations, also allowing for determining how to maintain, despite all, the continuity of the series. 26

Victimisation and Insecurity.book Page 27 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

Finally, this cooperation is critical for the contribution of surveys to the understanding of crime: without in-depth secondary analyses, without a cross-interpretation between survey results and time-honoured criminal sociology, settling for pseudo-results – as is often the case regarding insecurity – or drawing only a meagre share of a survey’s potential contributions are serious risks, which would be particularly regrettable for such costly investigations. Hope (2006) has thus put forward that, for years, the drop in burglaries was attributed to programmes fighting repeat victimisation, whereas – on the basis of knowledge in urban sociology on social and spatial segregation and of the criminal opportunities theories – it could rather be accounted for by the protective strategies developed by potential victims. By mobilising the sociology of professions and of the police, Robert, Zauberman and Pottier (2003) showed that one could often explain the difference between survey and police statistics by the differing perspectives between a professional (the police officer) and a layman (the victim).

4.

Dissemination and publication

In most countries (England, the Netherlands, Belgium, Italy), results are regularly accessible, at least on Internet sites. The situation is more complicated in Germany owing to the variety of surveys: the most important ones are published, but a certain number of them carried out in a purely academic framework or the smallest local surveys are often barely distributed. In the Iberian peninsula, dissemination does not seem very systematic, and publication even less so. Finally, in France, the situation is variable: although the results of surveys conducted in academic environment are systematically published and posted on the Web, it is much more succinct when it comes to national surveys (except through academic secondary analyses), and certain regional or local campaigns are not published or hardly distributed. The reader is referred to the appendix for a list of publications mentioned in the national reports and of websites containing survey results.

IV. Conclusion 1. In a quarter-century, surveys on victimisation and insecurity have developed considerably but quite irregularly, depending on the countries. In addition, the range between general surveys and those 27

Victimisation and Insecurity.book Page 28 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

2.

3.

4.

5.

delving into a specific topic, between international, national surveys and regional or local arrangements, between surveys covering both victimisations and insecurity and polls concerning only the latter is, in final analysis, quite diversified. We recommend widespread dissemination amongst users of the palette of existing studies. Although we are quite close to having an approximately standard list of victimisations studied, the same cannot be said for insecurity: in this latter case, the protocols used are far from standardised and quite often run up against serious criticisms. We recommend a serious effort in the standardisation of protocols, especially concerning insecurity. The solidity of surveys depends in large part both on the size of the sample – a good number are insufficient for providing tight enough confidence intervals – and the stability of the instrument. Without the latter, it is difficult to know whether a change in results describes a modification in the real situation or whether it is only the artefact of an uncontrolled modification of the instrument. Even if these problems are settled, serious threats remain as to the future of these surveys with the rising rate of non-responses, as those can be contained only at additional cost. We recommend that serious attention be paid to the problems of the sample size, questionnaire administration and, above all, stability of the protocols used, as well as a systematic examination of new problems such as the rate of non-responses or the increase in the number of households without a fixed landline. In the end, rare are the countries where these surveys are integrated into a decision-making and evaluation process. Several of them only manage to consider them as the junior partner of traditional police statistics. Finally, a good number of national, regional and local governments are having trouble in really taking into account the investigations that they nonetheless commissioned. We recommend that transverse discussion structures allow policymakers at different levels of government to maximise the use of these surveys. Still, this material – when of sufficient quality – can provide for a large number of scientific studies likely to give profound new insight into the knowledge of crime. And the development of such research – beyond what is currently carried out even in well-equipped countries – is necessary for avoiding misuse of these surveys and even for improving their quality. If this scientific production generally remains too scarce, it is initially owing to the overly small number of researchers with a 28

Victimisation and Insecurity.book Page 29 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

quantitative methodology orientation capable of both mastering the relevant scientific literature and working expertly on this type of data. We recommend giving priority to the development of high-level research on victimisation and insecurity surveys.

References AGRA C. DA, QUINTAS, J., FONSECA, E., 2001, De la sécurité démocratique à la démocratie de sécurité: le cas portugais, Déviance et Société, Les politiques de sécurité et de prévention en Europe, 25, 4, 499-513. AGRA C. DA, QUEIROS C., 2007, Research on fear of Crime in Portugal, Contribution to FP6 Coordinated Action CRIMPREV, Workpackage 4 Perceptions of crime and insecurity. Workshop on The Role of Social and Cultural Transformation in Constructing Contemporary Insecurities, Esslingen. BECK F., ARWIDSON P., FIRDION J.-M., JASPARD M., GRÉMY I., WARSZAWSKI J., 2001, L'avenir des enquêtes téléphoniques face à l'évolution des télécommunications, in Droesbeke J.J., Lebart L. (dir.), Enquêtes, modèles et applications, Paris, Dunod, 285-293. BECK F., GUILBERT PH., GAUTIER A. (dir.), 2007, Baromètre santé 2005, Saint-Denis, INPES. BECK F., LEGLEYE S., PERETTI-WATEL P., 2005, Aux abonnés absents: liste rouge et téléphone portable dans les enquêtes en population générale sur les drogues, Bulletin de méthodologie sociologique, 86, 5-29. BOTTOMLEY K., COLEMAN C., 1976, Criminal Statistics: The Police Role in the Discovery and Detection of Crime, International Journal of Criminology and Penology, 4, 33-58. BOTTOMLEY K., COLEMAN C., 1981, Understanding crime rates. Police and public roles in the production of official statistics, Farnborough, Gower. CANTOR D., LYNCH J.P., 2000, Self-Report Surveys as Measures of Crime and Criminal Victimisation, in DUFFE D., MCDOWALL D., GREENMAZEROLLE L., MASTROFSKI S. (eds.), Measurement and Analysis of Crime and Justice, Washington DC, National Institute of Justice, 85-138. COLL., 1998, La statistique judiciaire; son histoire et ses usages scientifiques, Déviance et Société, XXII, 2, numéro spécial donnant suite à un colloque du GERN organisé en 1995. CRAWFORD A., 2001, Les politiques locales de prévention de la délinquance en Angleterre et au Pays de Galles: nouvelles stratégies et nouveaux développements, Déviance et Société, 25, 4, 427-458.

29

Victimisation and Insecurity.book Page 30 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

DA PASSANO M., 1995, Le statistiche giudiziare come strumento di politica penale nel Granducato di Toscana, Materiali per una Storia delle Cultura Giuridica, 25, 1, 43-58. DEL FRATE A.A., ZVEKIC U., VAN DIJK J.J.M.(eds.), 1993, Understanding Crime. Experiences of Crime and Crime Control, Rome, UNICRI. HERPIN N., LAGRANGE H., 2005, La victimation de proximité, les précautions et la peur. Étude sur la cohésion sociale de voisinage, Revue économique, 56, 2, 285-312. HOPE T., 2006, Mass Consumption, mass predation – private versus public action? The case of domestic burglary in England and Wales, in Lévy R., Mucchielli L., Zauberman R. (dir.), Crime et insécurité: un demi-siècle de bouleversements. Mélanges pour et avec Philippe Robert, Paris, L'Harmattan, 45-61. HOUGH M., 2008, Research on victimisation and insecurity in Britain, this volume. HOUGH M., ROBERTS J.V., 2007, Public Opinion and Criminal Justice: the British Crime Survey and Beyond, in Hough M., Maxfield M.G. (eds.), Surveying Crime in the 21st Century, Cullompton, Willan Publishing, 199-221. JACKSON J., 2005, Validating New Measures of the Fear of crime, International Journal of Social Research Methodology, 8, 4, 297-315. JONES T., MCLEAN B., YOUNG J., 1986, The Islington Crime Survey. Crime, victimisation and policing in inner-city London, Aldershot, Gower. KINSEY R., 1984, Merseyside Crime Survey: First Report, Liverpool, Merseyside Metropolitan Council. KURY H., 1994, The influence of the specific formulation of questions on the results of victim studies, European Journal on Criminal Policy and Research, 2, 4, 48-68. KURY H., OBERGFELL-FUCHS J., WÜRGER M., 2000, Gemeinde und Kriminalität. Eine Untersuchung in Ost- und Westdeutschland, Freiburg, iuscrim. LECUIR J., 1974, Criminalité et moralité: Montyon, statisticien du Parlement de Paris, Revue d'histoire moderne et contemporaine, juilletseptembre, 445-493. MAYHEW P., VAN DIJK J.J.M., 1997, Criminal Victimisation in Eleven Industrialised Countries : Key Findings from the 1996 International Crime Victims Survey, Den Haag, WODC. OBERGFELL-FUCHS J., 2008, Crime victims and insecurity surveys in Germany, this volume. PAUWELS L., PLEYSIER S., 2005, Assessing Cross-Cultural Validity of Fear of Crime Measures through Comparisons between Linguistic Communities in Belgium, European Journal of Criminology, 2, 2, 139-159. 30

Victimisation and Insecurity.book Page 31 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

PAUWELS L., PLEYSIER S., 2008, Crime victims and insecurity surveys in Belgium and the Netherland, this volume. PEASE K., 1993, Individual and community influences on victimisation and their implications for crime prevention, in FARRINGTON D.P., SAMPSON R.J., WIKSTROM P.-O. (eds.), Integrating Individual and Ecological Aspects of Crime, Stockholm, National Council for Crime Prevention, 323-338. PEASE K., 1998, Repeat Victimisation: Taking Stock, Crime Detection and Prevention Series Paper 90, London, Home Office. PERROT M., ROBERT Ph., 1989, Compte Général de l'Administration de la Justice criminelle en France pendant l'année 1880 et rapport relatif aux années 1826 à 1880, Genève, Paris, Slatkine. PLEYSIER S., PAUWELS L., VERVAEKE G., GOETHALS J., 2005, Temporal invariance in repeated cross-sectional ‘fear of crime’ research, International Review of Victimology, 12, 3, 273-292. PLEYSIER S., VERVAEKE G., GOETHALS J., 2006, Hoe ‘representatief’ is deVeiligheidsmonitor? Over de accuraatheid van telefonische surveys in Vlaanderen, Panopticon, 27, 4, 107-114. POLICE STANDARDS UNIT, 2006, Guidance on Statutory Performance Indicators for Policing 2006/07, London, Home Office. PRESIDENT’S COMMISSION ON LAW ENFORCEMENT AND ADMINISTRATION OF JUSTICE, 1967, The Challenge of Crime in a Free Society, Washington DC., US Gov. Printing Office. QUETELET A., 1984, Sur la statistique morale et les principes qui doivent en former la base, Déviance & Société, 8,1, 13-41 (1° édition : 1848). RECASENS I BRUNET A., 2008, Victimisation surveys in Spain and Portugal, this volume. ROBERT Ph., 1977, Les statistiques criminelles et la recherche, Déviance et Société, 1, 1, 3-27. ROBERT Ph., 1991, "Quand le sociologue utilise les statistiques criminelles"... ou comment concevoir le crime?, in ROBERT PH., EMSLEY C. (eds.), Geschichte und Soziologie des Verbrechens, Hamburg, Centaurus-Verlagsgesellschaft Pfaffenweiler, 29-34. ROBERT Ph., 1993 V° statistiques judiciaires, in ARNAUD A.-J. (dir.), Dictionnaire encyclopédique de théorie et de sociologie du droit, 2ème édition corrigée et argumentée, Paris, LGDJ, 582-583. ROBERT Ph., 1999, Le citoyen, le crime et l'État, Paris-Genève, Droz. ROBERT Ph., 2008, Research on Victimisation and insecurity in France, this volume. ROBERT Ph., ZAUBERMAN R., POTTIER M.-L., 2003, La victime et le policier: point de vue profane et point de vue professionnel sur la délinquance, Sociologie du travail, 45, 3, 343-359.

31

Victimisation and Insecurity.book Page 32 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

SACCHINI G., SELMINI R., 2008, Research on victimisation and insecurity in Italy, this volume. SAMPSON R.J., GROVES B.W., 1989, Community Structure and Crime: Testing Social-Disorganization Theory, American Journal of Sociology, 94, 4, 774-802. SCHNELL R., KREUTER F., 2000, Untersuchungen zur Ursache unterschiedlicher Er-gebnisse sehr ähnlicher Viktimisierungssurveys, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 52, 1, 96-117. SCHWIND H.-D., AHLBORN W., EGER H.J., JANY U., PUDEL V., WEIß R., 1975, Dunkelforschung in Göttingen 1973/1974, Wiesbaden, Bundeskriminalamt. SELLIN T., 1931, The basis of the crime index, The Journal of Criminal Law and Criminology, 22, 335-356. SMITH D.J., GRAY J., 1985, Police and People in London, London, Gower. SPARKS R.F., GENN H.G., DODD D.G., 1977, Surveying Victims. A Study of the Measurement of Criminal Victimisation, Chichester-New York- BrisbaneToronto, John Wiley & Sons. STEPHAN E., 1976, Die Stuttgarter Opferbefragung. Eine kriminologischviktimologische Analyse zur Erforschung des Dunkelfeldes unter besonderer Berücksichtigung der Einstellung der Bevölkerung zur Kriminalität, Wiesbaden, BKA. VAN DIJK J.J.M., MANCHIN R., VAN KESTEREN J., NEVALA S., HIDEG G., 2007, The Burden of Crime in the EU. Research Report: A Comparative Analysis of the European Crime and Safety Survey (EU ICS) 2005, Brussels. VAN DIJK J.J.M., MAYHEW P., 1992, Criminal Victimisation in the Industrialised World: Key Findings of the 1989 and 1992 International Crime Surveys, The Hague, Ministry of Justice. VAN DIJK J.J.M., MAYHEW P., KILLIAS M., 1990, Experiences of Crime across the World. Key findings from the 1989 International Crime Survey, Deventer – Boston, Kluwer. VAN KESTEREN J., MAYHEW P., NIEUWBEERTA P., 2000, Criminal Victimisation in Seventeen Industrialised Countries. Key Findings from the 2000 International Crime Victims Survey, The Hague, WODC. VANDERVEEN G., 2006, Interpreting Fear, Crime, Risk and Unsafety, Boom Legal Publishers. WITTEBROOD K., 2001, Onveiligheidsbeleving en slachtofferschap in westerse geïndustrialiseerde landen. Een multilevel-analyse, Tijdschrift voor criminologie, 43, 2, 144-157. VAN WILSEM J., 2003, Crime and Context: The Impact of Individual, Neighborhood, City and Country Characteristics on Victimisation, ICS Dissertation Series. ZAUBERMAN R., ROBERT Ph., POTTIER M.-L., 2004, Profils de victimes, profils de victimations, Déviance et Société, 28, 3, 369-384. 32

Victimisation and Insecurity.book Page 33 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

Publications or sites giving access to surveys 쑱 On the international surveys,

• for the ICVS, cf. the UNICRI site www.unicri.it • for the Gallup Europea survey, see. the EU ICS site www.europeansafetyobservatory.eu. 쑱 For Belgium,

• concerning the security monitor, see. the Federal Police site www.polfed-fedpol.be/pub/veiligheidsMonitor/2006/ monitor2006_nl.php ou www.polfed-fedpol.be/pub/veiligheidsMonitor/2006/monitor2006_fr.php • for the university-run local surveys: o Hebberecht P., Hofman H., Philippeth K., Colle P., 1992, Buurt en Criminaliteit. Vanden Broele/v.z.w, Politeia. o Hebberecht P., Pelfrene E., Lippens R., 1986, Onveiligheid – Criminaliteit – Burgerwacht, Eindrapport, Gent, Rijksuniversiteit Gent, School voor Criminologie, Onderzoek uitgevoerd in opdracht van de Minister van Binnenlandse Zaken. 쑱 For England and Wales,

• The various editions of the British Crime Survey are on the Home Office site http://www.homeoffice.gov.uk/rds/bcs1.html • For regional or local surveys: o Jones T., Maclean B., Young J., 1986, The Islington Crime Survey: Crime, Victimisation and Policing in Inner City London, Aldershot, Gower. o Kinsey R., 1984, Merseyside Crime Survey: First Report, Liverpool, Merseyside Metropolitan Council. o Smith D.J., 1983, Police and People in London, A survey of Londoners, London, Policy Studies Institute. o Sparks R., Genn H., Dodd D., 1977, Surveying Victims, London, Wiley. 쑱 For France,

• for the first national survey and local surveys in Toulouse and Épinay: o Zauberman R., Robert Ph., 1995, Du côté des victimes; un autre regard sur la criminalité, Paris, L’Harmattan. • for the Amiens survey: o Zauberman R., Robert Ph., Pottier M.-L., 2000, Risque de proximité ou risque lié au style de vie; enquêtes et évaluation de la sécurité urbaine, Les Cahiers de la sécurité intérieure, 42, 4, 193-220. 33

Victimisation and Insecurity.book Page 34 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

• for the EPCVMs: o the site of the National Institute for Statistics and Economic Surveys (INSEE) http://www.insee.fr/fr/ffc/docs_ffc/irsoc045.pdf o certain elements – especially for recent years – are also found in the reports of the National Crime Observatory (OND) www.inhes.interieur.gouv.fr and those of the National Observatory of Urban Problem Areas (ONZUS) http://www.ville.gouv.fr/ infos/dossiers/observatoire-des-zus.html • for the local surveys in Lyons, Saint-Denis, Gonesse and Aubervilliers and for the first two (2001, 2003) regional surveys in Île-deFrance, cf. the CESDIP site www.cesdip.org/-Etudes-et-Donnees-Penales.html • for succinct information on the other Île-de-France regional surveys (2005, 2007): www.iaurif.org/fr/ressources_doc/publications/publicationsrecentes/notesrapides/securite.htm • for the Health Barometer (containing certain elements on the victimisation of physical violence), cf. the site of the National Institute for Prevention and Health Education (INPES) www.inpes.sante.fr • for the ESPAD and ESCAPAD surveys of young people (containing certain elements on the victimisation of physical violence), cf. the sites of the Interdepartmental Mission for the Fight Against Drugs and Addiction (MILDT) www.drogue.gouv.fr and the National Drugs and Addictions Observatory (OFDT) www.ofdt.fr • for the surveys on violence against women: o Jaspard M., Brown E., Condon S., Firdion J.-M., FougeyrollasSchwebel D., Houel A., Lhomond B., Maillochon F., Saurel-Cubizolles M.J., Schiltz MA., 2003, Les violences envers les femmes en France; une enquête nationale, Paris, La Documentation française. • there is no publication of all the surveys on school violence, but one can refer to a summary article and references quoted in o Debarbieux É., 2004, Les enquêtes de victimation en milieu scolaire. Leçons critiques et innovations méthodologiques, Déviance & Société, 28, 3, 317-333. • for national surveys (Agoramétrie et al.) on the security preoccupation: o Robert Ph., Pottier M.-L., 1997, ‘On ne se sent plus en sécurité’. Délinquance et insécurité. Une enquête sur deux décennies, Revue française de science politique, 47, 6, 707-740. o Robert Ph., Pottier M.-L., 2004, Les préoccupations sécuritaires: une mutation?, Revue française de sociologie, 45, 2, 211-242.

34

Victimisation and Insecurity.book Page 35 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

o Robert Ph., Pottier M.-L., 2006, Security, Law and Order: a Profound Transformation?, Revue française de sociologie, 47, annual English selection, 35-64. 쑱 For Germany,

• Aben R., 1992, Kriminologische Regionalanalyse Lübeck in Koch K.F., Ed., Kriminalitätslagebilder, Wiesbaden, Bundeskriminalamt, 305-341. • Arnold H., Korinek, L., 1991, Victimisation, attitudes towards crime and related issues: Comparative research results from Hungary in Kaiser G., Kury H., Albrecht H.-J., Eds., Victims and criminal justice, Freiburg, Max-Planck-Institut für Strafrecht, 51, 99-121. • Baier D., Pfeiffer C., Windzio M., Rabold S., 2006, Schülerbefragung 2005: Gewalterfahrungen, Schulabsentismus und Medienkonsum von Kindern und Judgendlichen, Hannover: Kriminologisches Forschungsinstitut Niedersachsen www.kfn.de/schuelerbefragung.shtml • Boers K., 1991, Kriminalitätsfurcht, Pfaffenweiler, Centaurus. • Boers K., Gutsche G., Sessar K., Eds., 1997, Sozialer Umbruch und Kriminalität in Deutschland, Opladen, Westdeutscher Verlag. • Bundesministerium für Familie, Senioren, Frauen und Judgend, 2004, Lebenssituation, Sicherheit und Gesundheit von Frauen in Deutschland, Berlin, BMFSFJ http://www.bmfsfj.de/Kategorien/publikationen.html • Dölling D., Feltes T., Heinz W., Kury, H., Eds., 2003, Kommunale Kriminalprävention – Analysen und Perspektiven, Holzkirchen, Felix. • Dreher G., Kury H., Obergfell-Fuchs J., 2005, Bevölkerungsumfragen in ländlichen und städtischen Regionen – Kriminalitätsanalysen und gezielte proaktive Maßnahmen am Beispiel der Stadt Rottweil, in Bannenberg B., Coester M., Marks E., Eds., Kommunale Kriminalprävention, Mönchengladbach, Forum Verlag Godesberg, 169-188. • Feltes T., Goldberg B., 2006, Selbstberichtete Delinquenz, Viktimisierung und Verbrechensfurcht bei Schülern mit und ohne Migrationshintergrund, in Obergfell-Fuchs J., Brandenstein M., Eds., National und international Entwicklungen in der Kriminologie. Festschrift für Helmut Kury zum 65. Geburtstag, Frankfurt/Main, Verlag für Polizeiwissenschaft, 203-237. • Forschungsgruppe Kommunale Kriminalprävention Baden-Württemberg, 1998, Viktimisierungen, Kriminalitätsfurcht und Bewertung der Polizei in Deutschland Monatsschrift für Kriminologie und Strafrechtsreform, 81, 2, 67-82. • Funk W., Ed., 1995, Nürnberger Schüler-Studie 1994, Regensburg, S. Roderer.

35

Victimisation and Insecurity.book Page 36 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

• Hermann D., Dölling, D., 2001, Kriminalprävention und Wertorientierungen in komplexen Gesellschaften, Mainz, Weisser Ring. • Kräupl G., Ludwig H., 1993, Wandel kommunaler Lebenslagen, Kriminalität und Sanktionserwartungen. Bevölkerungsbefragung in einer städtischen Region Thüringens1991/92 (Jenaer Kriminalitätsbefragung), Freiburg, Max-Planck-Institut für Strafrecht. • Kreuzer A., Görgen T., Krüger R., Münch V., Schneider, H., 1993, Judgenddelinquenz in Ost und West, Bonn, Forum Verlag Godesberg. • Kury H., 1991, Victims of crime – Results of a representative telephone survey of 5.000 citizens of the former Federal Republic of Germany, in Kaiser G., Kury H., Albrecht H.-J., Eds., Victims and criminal justice, Freiburg, Max-Planck-Institut für Strafrecht, 50, 265-304. • Kury H., Dörmann U., Richter H., Würger M., 1992, Opfererfahrungen und Meinungen zur Inneren Sicherheit in Deutschland, Wiesbaden, Bundeskriminalamt. • Ludwig H., Kräupl G., 2005, Viktimisierung, Sanktionen und Strafverfolgung. Jenaer Kriminalitätsbefragung über ein Jahrzehnt gesellschaftlicher Transformation Mönchengladbach, Forum Verlag Godesberg. • Oberwittler D., Blank T., Köllisch T., Naplava, T., 2001, Soziale Lebenslagen und Delinquenz von Judgendlichen. Arbeitsberichte 1/2001, Freiburg, Max-Planck-Institut für Strafrecht. • Oberwittler D., Köllisch T., Naplava T., Blank T., 2002, MPI-Schulbefragung Breisgau/Markgräfler Land 2000, Freiburg, Max-PlanckInstitut für Strafrecht. • Plate M., Schwinges U., Weiß, R., 1985, Strukturen der Kriminalität in Solingen, Wiesbaden, Bundeskriminalamt. • Schwind H. -D., Ahlborn W., Eger H.J., Jany U., Pudel V., Weiß, R., 1975, Dunkelfeldforschung in Göttingen 1973/74, Wiesbaden, Bundeskriminalamt. • Schwind H.-D., Fetchenhauer D., Ahlborn W., Weiß, R., 2001, Kriminalitätsphänomene im Langzeitvergleich am Beispiel einer deutschen Großstadt. Neuwied, Kriftel, Luchterhand. • Schwind H.-D., Roitsch K., Ahlborn W., Gielen B., Hrsg., 1995, Gewalt in der Schule, Mainz, Weisser Ring. • Stephan E., 1976, Die Stuttgarter Opferbefragung, Wiesbaden, Bundeskriminalamt. • Teske R. H. C., Arnold, H., 1991, A comparative victimisation study in the United States and the Federal Republic of Germany: A description of the principal results, in Kaiser G., Kury H., Albrecht H.-J., Eds., Victims and criminal justice, Freiburg, Max Planck Institut für Strafrecht, 51, 3-44. 36

Victimisation and Insecurity.book Page 37 Thursday, December 4, 2008 12:33 PM

Surveys on victimisation and insecurity in Europe

• Wetzels P., Enzmann D., Mecklenburg E., Pfeiffer C., 2001, Judgend und Gewalt, Baden-Baden, Nomos. • Wetzels P., Greve W., Mecklenburg E., Bilsky W., Pfeiffer C., 1995, Kriminalität im Leben alter Menschen, Stuttgart, W. Kohlhammer. • Wilmers N., Enzmann D., Schaefer D., Herbers K., Greve W., Wetzels P., 2002, Judgendliche in Deutschland zur Jahrtausendwende: Gefährlich oder gefährdet, Baden-Baden, Nomos. 쑱 For Italy, • for the national victimisation surveys carried out by the Istituto nazionale di statistica, see. the ISTAT site www.istat.it/. For those concerning violence against women, cf. in particular www.istat.it/ giustizia/. • see. also the sites of the CISIS http://www.cisis.it/attivita/pubblicazioni/ vol004.html, Città sicure http://www.regione.emilia-romagna.it/ sicurezza/, and Transcrime http://www.transcrime.unitn.it/tc/1.php 쑱 For the Netherlands

• for the Central Bureau of Statistics surveys: http://statline.cbs.nl/statweb/start.asp?DM=SLen&LA=en&lp=applet&THEME=5470 • for the Politiemonitor Bevolking: http://www.politiemonitor.nl/tabellenreport.pdf 쑱 For Portugal, the Ministry of Justice surveys are found in:

• Crucho de Almeida M.R., 1992-1994, Inquérito de vitimaçao, Lisboa, Ministerio da Justicia, Gabinete de estudos e planeamento. • for the surveys carried out in Porto, cf. e.g.: o Negreiros, J. 1999 Vitimaçâo criminal na cidade do Porto, Observatorio Permanente de Segurança, Vol.III, duplicated. 쑱 For Spain,

• the surveys of the Centro d’investigaciones sociologicas are found on its site www.cis.es • for the Enquesta de seguretat publica de Catalunya, cf. that of the Department of the Interior of the Generalitat www.gencat.net/interior • a summary of earlier surveys by the Ajuntament de Barcelona in Sabaté Delgado J., 2005, L’Enquesta de victimitzacio de barcelona i de l’Area metropolitana, Barcelona, IERMB. • for the surveys on violence done to women, cf. the site of the Instituto de la mujer www.mtas.es/mujer • for surveys on school violence: www.defensordelpueblo.es/ informes2.asp www.gencat.net/interior/docs/text_integre.pdf www.gencat.net/interior/departament/publicacions/estudis/ resum_informe_ECESC_2005-2006_cat.pdf 37

Victimisation and Insecurity.book Page 38 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity.book Page 39 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands Lieven Pauwels, Stefaan Pleysier

In this contribution a synthetic overview of the crime victimisation and insecurity surveys in Belgium and the Netherlands is offered. We will focus on the background, scope, and scientific and technical organisation of the most important crime and victim surveys in both countries. From the observation that these kinds of surveys are sometimes lacking in attention towards methodological issues and measurement problems, we will stress research that invested in quality assessment and methodological improvement of the measurement instruments and methodologies used. Also, attention will be devoted to the use and implementation of these surveys in policy, at various governmental levels. We will discuss research and scientific assessments that concentrated on delineating good and bad practices, in the past, and for future research to come.

I.

Overview of crime victim and insecurity surveys in Belgium

Although crime and insecurity have already some tradition as the subject of criminological and sociological research, in recent years they have reached a prominent place in both the political discourse and scientific research. In Belgium, as in other European countries, there is undeniably a growing interest in this topic within the social and political sphere, and a heightened sensitivity in the media (Crawford, 2002). It is fair to say that, in Belgium, insecurity has become a hot public opinion issue from the 1991 general elections onwards; ever since, the issue appears more prominent in electoral campaigns and debates, on the political agenda and among policy makers in general (Hebberecht, Duprez, 2002). Although we would not go as far as arguing that ‘feelings of insecurity’ or even ‘fear of crime’, are imposed on the public by Vlaams Blok especially in the Flemish Community, it is hard to disagree with the influence this party had on

39

Victimisation and Insecurity.book Page 40 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

setting, and prioritizing, the ‘crime and insecurity’ issue on the political agenda1. On a quasi-regular base, this evolution has gained momentum with each new election, with the growing significance of Vlaams Belang, with the Dutroux case and its aftermath, and with a series of ‘senseless violence’ incidents within the last years. The growing importance of crime and insecurity in politics and policy is paralleled by an important external catalyst; media attention. Crime, insecurity and fear of crime are ‘mediagenic’ and literally hot triggers in newspapers and television programs. The amount of editorials and letters on this topic in Belgian newspapers, the number of reality shows related to crime or deviance, and ‘crime watch’ programs on TV channels are just some of the indicators of this. The attention to public safety issues, broadly from the 1990’s onward, in Belgium and other countries, hides a longer tradition of scientific and government-funded policy research. The genealogy of the national crime and victim survey is situated within the Commission of Law Enforcement and Administration of Justice, established in 1965 during the administration of U.S. president Johnson (Ditton, Farrall, 2000). After some pilot studies, the first National Crime Victimisation Survey (NCVS) was administered in 1973. Other national crime and victim surveys, such as the British Crime Survey (BCS), were soon to follow. Although a large body of research concentrates on crime, victimisation and ‘fear of crime’ internationally, this is hardly the case in Belgium (Devroe, Eliaerts, 2006; Goethals et al, 2002; Pleysier et al, 2002). Apart from local studies in the inner city of Ghent (Hebberecht et al, 1986; 1992; 1996), Liege (Janssen et al, 1981; Lemaître, 1982), and a small Dimarso-Gallup opinion poll in 1984 (Uniop, 1985), the first general population victim survey at the federal level was not conducted prior to 1997. In 1997 the Belgian Security Monitor (Veiligheidsmonitor) originated; the Security Monitor is the official national crime and victim survey, conducted by the Federal Police, under the authority of the Ministry of the Interior. 1.

Vlaams Blok (since 2004 Vlaams Belang) is an extreme right-wing party in Flanders; since the general elections of November 1991, where they reached as much as 10.6 percent for the Chamber of Representatives, their electorate rose to a little over 12% in 1995, and even 15.3% in the general elections of June 1999. ‘Improving people’s security’ and ‘being tough on crime’ has been a political priority of Vlaams Belang from the 1991 general elections onwards, closely related to what is perhaps their most lucrative theme, the ‘impossibility of a multicultural society’. Maddens describes this as a textbook example of ‘symbolic politics’: suggestive and emotional slogans are applied to unleash and stir up underlying and more existential feelings of anxiety (Maddens, 1998; Billiet, De Witte, 2001).

40

Victimisation and Insecurity.book Page 41 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

Since a first sweep in 1997, the Security Monitor was held in 1998, 2000, 2002, 2004 and 2006. In what follows, the Security Monitor will be discussed at length. Furthermore, Belgium also participated in the 1989, 1992 and 2000 sweeps of the International Crime Victim Survey (ICVS), the 1996, 2000 and 2002 sweeps of the Eurobarometer ‘Public Safety’, and the 2005 European Crime and Safety Survey (EUICS). Within this contribution, marginal attention will also be paid to these surveys. And finally, we will also mention some other, national and supranational, surveys, where the main interest is not crime victimisation or insecurity, but have nevertheless topical questions on crime or ‘fear of crime’.

1.

The Security Monitor

Background The Belgian Security Monitor, inspired by the Dutch Police Monitor (cf. infra), is in essence a federal, repeated cross-sectional, victim survey, using Computer Assisted Telephone Interviewing (CATI). It is designed to inform the development of new policies in the response to crime and victimisation and to evaluate past policy strategies, and is therefore a crucial counterbalance next to the official crime statistics (Ponsaers et al, 2001; Van Den Bogaerde, Van Den Steen, 2006). In 1997 the Minister of the Interior ordered a first sweep of the Security Monitor at the federal level, and at the local level, in those communities that had a safetycontract (veiligheids- en samenlevingscontract) with the federal government or were ‘pilot police zones’. The second sweep, conducted in 1998, introduced a postal survey as a possible, and cheaper, alternative to the telephone mode, in +/- 70 communities. After a thorough evaluation (cf. infra), it was decided not to offer the postal survey alternative in future sweeps. In order to create more time and budgetary space for further analyses and research, and the implementation of the Security Monitor in local communities, data collection is spread every two years since 1998. Scope The federal Security Monitor is based on a national, stratified random sample that allows for comparison in space, between regions, provinces and municipality types. The local Monitors are executed in all 73 communities with a safety-contract and in the police zones those communities belong to. Although it is a cross-sectional survey design and

41

Victimisation and Insecurity.book Page 42 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

not panel research, comparisons over time are possible as well2; we wish to stress, however, that the methodological conditions for valid trend analyses are, both with panel and cross-sectional designs, very stringent (Taris, 2000; Ditton et al, 2005; Pleysier et al, 2005). In order not to jeopardize comparability in time and space, the questionnaire of the Security Monitor remained largely unchanged since 2000. Comparability and continuity of the questionnaire measurement instruments, the question order, and the methodology used, are conditions sine qua non for valid comparisons over time. Limited changes to the questionnaire were allowed, but only after serious consideration and supported by scientific research (e.g. Billiet et al, 1998; Vanderhallen et al, 2000; Ponsaers et al, 2001; Molenberghs et al, 2005). Obviously, as it is a federal survey, questionnaires are in Dutch, for the Dutch-speaking respondents (living mostly in Flanders and to a small extent in Brussels), and in French for the French-speaking (living mostly in Wallonia and in Brussels). The questionnaire contains five different question modules; (1) neighbourhood problems, (2) fear of crime, (3) victimisation and reporting behaviour, (4) contact and police functioning, and (5) socio-demographic and background variables. Within the confined limits of this contribution, the construction of the questionnaire will not be discussed in detail; we will however give a short overview of the question blocks concerning victimisation and ‘fear of crime’. The questionnaire can be downloaded integral on the web pages of the federal police3. Also, descriptive results of each round of the Security Monitor (1997-2006), both as a federal report and as separate reports for each of the participating communities or police zones, together with a highlight summary, and a manual and reading guide can be consulted online on the web pages of the federal police4. Victimisation is questioned using a filter question concerning victimisation during the last twelve months; the victimisation of several possible 2.

3.

4.

Traditionally, within longitudinal research one can discern panel studies from repeated measurement surveys. In a panel design, a group of persons is followed over time and therefore the same people are questioned at each measurement point. In repeated measurement designs, different samples are drawn in an identical way, repeatedly within a given interval. The Dutch version can be consulted on: http://www.polfed-fedpol.be/pub/ veiligheidsMonitor/2006/notas/vragenlijst_2006.pdf; the French version on: http:// www.polfed-fedpol.be/pub/veiligheidsMonitor/2006/notes/questionnaire_2006.pdf. The results of the latest 2006 Safety Monitor can be found on: http://www.polfedfedpol.be/pub/veiligheidsMonitor/2006/monitor2006_nl.php (Dutch web pages) or on http://www.polfed-fedpol.be/pub/veiligheidsMonitor/2006/monitor2006_fr.php (French web pages).

42

Victimisation and Insecurity.book Page 43 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

personal and household crimes is asked (see Table 1). Also, additional questions, if relevant, concerning the frequency, place and time of victimisation, and whether or not, and where, this crime victimisation was reported to the police authorities, are asked. Within the Security Monitor reports, the results are presented in general and, if significant, by sociodemographic and other background variables (sex, age groups, occupational activity, education, region and year). Sexual crimes are questioned more in detail, with a description of attempted rape, rape, verbal sexual threat, unwanted sexual behaviour, and exhibitionism. Also, questions on the offender and the consequences are asked. Household crimes Vandalism car Vandalism (not on car, burglary excluded) Theft of car Theft from car Attempted burglary (without theft) Theft of bicycle Domestic burglary (with theft) Theft of motor cycle Vandalism motor cycle Personal crimes Threat with physical violence Theft from person without violence Traffic related crime Physical violence Crimes against the sexual morals Theft out of the home, victim absent Theft from person with threat/violence Table 1. Crime victimisation within the last 12 months, as measured in the Belgian Security Monitor

The measurement of ‘fear of crime’ in the Monitor is in many ways indicative for the general malaise in ‘fear of crime’ literature and research (Pauwels, Pleysier, 2005; Pleysier et al, 2004). Despite the weight of the publications that have appeared since the empirical ‘discovery’ of ‘fear of crime’ in the first U.S. crime surveys, it is striking that the content and theoretical ‘body’ are less imposing. Ferraro & LaGrange argue that ‘conceptual cloudiness and inappropriate operationalization taints the majority of this literature, thereby distorting the meaning and utility of the fear of crime concept’ (1987, 76); although perhaps somewhat pessimistic, the bottom line of this critique is not outdated, and has continued to be 43

Victimisation and Insecurity.book Page 44 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

repeated. In fact, Ditton & Farrall (2000) state in their overview of the literature that, despite more than 30 years of research on the subject ‘fear of crime’, surprisingly little can be said about the phenomenon. An important factor underlying and explaining at least partially the above conclusion is the conservatism inherent to the era –‘the heyday of survey research’- when ‘fear of crime’ research originated. Large-scale crime and victim surveys were developed and administered, mainly by telephone, in order to have a more accurate estimate of the victimisation rate and the ‘fear of crime’ (Block, 1993)5. As these large scale surveys have a longitudinal design and, therefore, are administered on a recurrent base, conservatism not only points to the dominant methodology –telephone surveys- but within that methodology to the item level as well. The concept ‘fear of crime’ in these surveys is predominantly measured by a single, ‘standard’ item: ‘How safe do or would you feel being out alone in your neighbourhood at night?’ (Ditton, Farrall, 2000 ; Hale, 1996). Ever since, numerous authors have questioned this conservative tenacity with profound persistence, resulting in recent years in a growing tendency towards using scaling techniques as a far better way of measuring a complex and multidimensional concept as ‘fear of crime’. As far as the conceptualisation and measurement of the ‘fear of crime’ concept in the Belgian Security Monitor is concerned, the classic criticism in the international literature is at least partly applicable here as well. We refer to Table 2 for a description of the ‘fear of crime’ measurement in the Safety Monitor. The ‘fear of crime’ item set used in the Monitor contains a single, general or ‘formless’ item on ‘how safe the respondent feels’ – without reference to crime, a specific offence or a situation-, tapping more into a general feeling of unease and the emotional component of a broad ‘fear of crime’ conception. This general single-item question precedes four questions tapping the behavioural component of ‘fear of crime’; the respondent is asked whether he or she avoids certain areas in the neighbourhood, does not open the door for strangers after dark, hides valuable things at home, or avoid leaving home after dark, because he or she does not consider it safe otherwise. Since 2000, these questions are completed with four items concerning the individual’s risk perception of becoming -within the next 12 months- a victim of a burglary, of physical violence, of a theft without violence, and of a traffic related crime. One 5.

Also, since most national crime and victim surveys are by origin ‘governmental instruments’, questions rose whether “the empirical ‘discovery’ of fear of crime was invoked to support ‘tougher’ neo-conservative political agendas and policies” (Jackson, 2005, 6; Lee, 1999). Lee (1999) has labelled this as the governance-throughfear strategy of neo-liberal governments; or “in other words, ‘fear’ is the legitimation of surveillance, punishment and punitive laws” (Vanderveen, 2006, 213).

44

Victimisation and Insecurity.book Page 45 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

could say this item set represent a more cognitive component of ‘fear of crime’. And finally, one question concerning the so called ‘altruistic fear’ is added: if respondents have children, they are asked whether they prevent their children from going places because they fear for their safety. Research on fear of victimisation has concentrated exclusively on personal fear (fear for one’s own safety) while overlooking the fear that individuals have for the safety of others (e.g., children, spouses, friends) (Warr, 1992, 724). Variable

FOC

V63 V57

Does it sometimes happen that you feel unsafe? Is this…? Does it sometimes happen that you avoid certain areas in your neighbourhood because you do not consider them safe? Is this…? Does it sometimes happen that you do not open the door for strangers in the evening or at night because you do not consider it safe? Is this…? Does it sometimes happen that you hide valuable things at home? Is this…? Does it sometimes happen that you avoid leaving home if it is dark? Is this…? (always/often/sometimes/seldom/never)

V58

v59 V60

item

Table 2. ‘Fear of crime’ item set in the Belgian Security Monitor

As a final note on the measurement of ‘fear of crime’, and the operationalisation in the Security Monitor is no exception to that, we could not agree more with Ditton et al (2000, 154-255), in their contention that ‘the so-called fear of crime is –to an unknown degree- a function of the types of questions that are asked, and the way that they are posed’. Organisation The Security Monitor is a project runned by the Department of the National Database (Directie van de Nationale Gegevensbank), as part of the General Department of Operational Support (Directie van de Operationele Ondersteuning), of the Federal Police. The project resorts under the responsibility of the Minister of the Interior. The fieldwork of the Monitor is outsourced, by means of a call for proposals every round, to a research bureau. The project is monitored and guided by an Advisory Board (composed of researchers, representatives of all policy levels, federal and local police services,…) and occasionally, a Working Group in order to deal with more technical issues of large scale survey research. Due to the bi-annual, and 45

Victimisation and Insecurity.book Page 46 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

not annual, fieldwork, budgetary space is created to propose and attract scientific research on a variety of relevant aspects, with the aim to upgrade the scientific value and overall quality of the survey. We will discuss some of this research below, under ‘Methodology’ and ‘Use and Implementation of the Security Monitor’. On demand, raw data and codebooks are available to interested researchers and research groups in order to perform their own analyses. Although this possibility has not been used frequently up until now, it is our firm believe that the availability of the data –and therefore the transparency of the instrument- is a major step forward in the proclamation of the Security Monitor as an important and indispensable data source on crime and insecurity. Methodology The growing importance of large-scale general surveys, not in the least within politics and the media, has undoubtedly increased the awareness and concerns about the quality, validity and reliability of survey data. Although traditionally survey quality and measurement issues are not a number one priority in crime and victim surveys, it is of the utmost importance to pay attention to the potential threat these forms of bias can produce (Dillman et al, 2002 ; Couper, de Leeuw, 2003 ; Stoop, 2005). In this respect, Groves (1989) distinguishes four types of potential survey error, i.e. sampling error, coverage error, nonresponse error and measurement error. Sampling error results from using a sample instead of surveying the entire population. A coverage error results from the failure to give all the units of the population on which information should be gathered a nonzero chance to be included in the sampling frame. Nonresponse error results from the fact that not all the units of the gross sample actually participate in the survey. Finally, different processes related to the instrument, the interviewer, the respondent, and the data collection mode contribute to measurement error (Braun, 2003, 137-138). In this section of our contribution, we wish to pay attention to these potential threats to the survey quality of the Security Monitor. Sampling and coverage problems The sampling design of the Security Monitor is complex and multi-staged: in 2004, the total number of succeeded interviews was 41,017 (both federal and local). The sample for a local Monitor is a simple random sample from the Belgacom register of telephone landlines at the level of the community or police zone. Fieldwork for the local Monitors continues until 350 respondents are questioned. The federal Monitor contained 12,000 46

Victimisation and Insecurity.book Page 47 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

succeeded interviews, and was partly composed with respondents out of the local Monitors, and added with a separate, federal sample. In order to obtain an overall federal Monitor, representative for the total population and all relevant sub samples (regions, provinces and community types), this separate random sample, stratified proportional to the regions, provinces and community types, of telephone numbers is drawn and added. Next, individuals within the selected households (telephone numbers) are chosen using the ‘birthday method’: interviewers will interview the person (15+ and speaking either French or Dutch in a sufficient way) in the household whose next birthday is closest to the date of the interview. This ensures a random selection of eligible respondents within the households (telephone numbers) selected. And finally, post-stratification is used to weigh subgroups based on their known age and gender distribution, in order to correct for any bias owing to the sampling and administering of the survey. Despite these efforts, it is hard to say whether the Security Monitor can be described as a truly representative sample. Perhaps an even bigger challenge as far as sampling and coverage problems are concerned, is the mode of interviewing in the Security Monitor (Pleysier et al, 2006). The target population in the Security Monitor are individuals living in Belgium, 15 years and older with a reasonable knowledge of Dutch or French. Using telephone interviewing however, confronts the survey design with serious, and expanding non-coverage problems; non-coverage is a consequence of the difference between the target population and the population one is effectively able to reach, the frame population (Stoop, 2005). Non-coverage can be seen as a problem in the Security Monitor, since the register used to sample is confined to landline telephone numbers only. In recent years however, the number of respondents that are no longer accessible by a landline, but only by mobile phone has increased exponentially. Overall, the number of individuals in Belgium that can not be contacted by a landline phone number –and therefore have a zero chance to be included in the sampling frame- is estimated around 15% of the total population (Pleysier et al, 2006). This percentage is however vastly different depending on subgroup characteristics: within subgroups like young couples without children, or single parents, or young singles, the percentage of potential respondents no longer reachable by landline can easily amount up to 50% and more. According to many scholars, this kind of error is generally seen as the largest problem of error in survey research (Assael, Keon, 1982 ; Marker, 2002).

47

Victimisation and Insecurity.book Page 48 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Nonresponse error Non-response may be due to the inability to make contact with the selected sample unit, or because the sample unit is not able, not capable or not willing to participate in the survey (Couper, de Leeuw, 2003 ; Pleysier, Billiet, 2005). Based on this information, and additionally, on what one considers being eligible and ineligible respondents, response, non-contact and refusal rates can be calculated in myriad ways (Stoop, 2005). The (non)response, non-contact and refusal rates of a survey are generally considered to be a major – or the major – quality criterion. However, there is no agreed-upon benchmark figure or threshold value or standard to evaluate the response rate obtained in a specific survey. A higher response rate is usually preferable to a lower one, but this is insufficient guidance for evaluating the response rate of a specific survey (Loosveldt e.a., 2004, 69). Also, a single-minded focus on response rates is not a fruitful approach for a more fundamental reason (Stoop, 2005). What is important is not so much the (non-)response rate an sich, but the potential bias non-response can cause. Bias due to non-response is a function of the amount of nonresponse and the differences between respondents and non-respondents. This logic implies that efforts to maximize a study’s response rate during data collection should not be extended at any cost; Simply increasing the response rate for a study does not necessarily reduce the possibility of nonresponse error (Couper, de Leeuw, 2003, 159). A central question for the evaluation of the response quality is therefore whether those who cooperated with the interview differ from those who did not take part, in characteristics relevant to the survey (Loosveldt e.a., 2004, 70). The calculation of the response rate in the Security Monitor could be done in a more profound way, using stronger criteria (Pleysier et al, 2006). Moreover, the response rates in the Security Monitor have shown to be rather instable over time – 72% in 1998, 58.6 in 2000, 74.2% in 2002 and 66.4% in 2004 – and serious geographical differences can be found – 63% in the province of West-Flanders compared to 78% in the province Luxembourg (2004 Security Monitor). These observations, combined with the topics and type of questions asked in crime and victim surveys, stress the crucial importance of a continued investment in the phenomenon of non-coverage and non-response. At this moment, the Federal Police has issued a federal Agora-programme research to the Catholic University of Leuven and the University of Hasselt, sponsored by the Belgian Science Policy, on the question how to deal with and overcome non-coverage and non-response problems in a telephone survey such as the Security Monitor.

48

Victimisation and Insecurity.book Page 49 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

Measurement error Measurement error is caused by a difference between the true score of a respondent and the score that in reality is assigned to the respondent. Some important causes of measurement issues are discussed below with regard to Belgian victim studies. This can be due to a variety of reasons concerning the measurement instrument or mode used, the interviewer, the respondent, etc. Interviewer bias The Security Monitor is a large-scale survey in which, in the 2004 sweep, 237 interviewers are conducting more than 40,000 interviews during five months of fieldwork. Although the technical report of the Security Monitor signals no significant interviewer differences, closer examination reveals that significant differences between the number of ‘don’t know’ and ‘no answer’ exist, which is indicative for the existence of interviewer bias (Pleysier et al, 2006). Also, about 50% of the 21,454 Dutch respondents were interviewed by as little as 22 interviewers. These interviewers were able to deliver more than 300 succeeded interviews each; eight of them had over 500 interviews, with a maximum of no less than 823 succeeded interviews for one interviewer. Also, previous research had already shown that interviewers followed up interview instructions in a rather varying way (Billiet et al, 1998). Therefore, it was suggested to distribute the number of interviews more evenly over to total amount of interviewers active in the fieldwork, in order to avoid interviewer bias (Pleysier et al, 2006). Measurement invariance Above, we have signalled a growing tendency in recent years towards using scaling techniques for measuring complex concepts or attitudes. The Security Monitor picked up on this evolution and offers the possibility to measure several concepts, such as ‘fear of crime’, using an item set instead of a single item. Choosing more complex measures, however, gives rise to a number of ‘new’ issues related to measurement error, which remain until today fairly underreported in criminological literature. Measurement invariance is such an issue (Pleysier et al, 2003; 2004; 2005 ; Pauwels, Pleysier, 2005). Central to scaling techniques and factor analysis is the assumption that a number of observed or ‘manifest’ items represent an underlying, ‘latent’ concept. It is then assumed that this latent concept is responsible for the variance between the observed items. However, classic factor analysis does not take into account possible within-sample

49

Victimisation and Insecurity.book Page 50 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

differences between subgroups of the sample; different scores between groups on, for example, ‘fear of crime’, might therefore be due to ‘true’ differences on ‘fear of crime’, or due to systematic bias in the way different groups respond to certain items. Elsewhere, we stated that if cross-cultural comparisons are to be valid, the assumption of measurement invariance across cultural groups has to be fulfilled. The federal Security Monitor is based on a national sample comprising roughly equal numbers of Flemishspeaking and French-speaking respondents. For this reason, bias can occur, on the one hand because of poor translations, or on the other hand because of ‘true’ cross-cultural differences and divergent sensitivity to the measurement instrument at hand. Secondary analyses using structural equation modelling have shown that the ‘fear of crime’ item set in the 1998 Security Monitor is sufficiently invariant (Pauwels, Pleysier, 2005). However, ignoring potential cross-cultural bias would invalidate all comparisons between different cultural or linguistic groups, and therefore causes contextual interpretation to be unfounded. As large scale longitudinal and cross-national or cross-cultural surveys gain attention within criminological research, it is our firm belief that the issue of measurement invariance will become more prominent in the future6. Indeed, Welkenhuysen-Gybels argues that perhaps one of the greatest challenges in conducting an international survey is the invariance of the measurement instruments (i.e. the survey questionnaire) across all subpopulations (e.g. countries, languages or cultural groups) under study (Welkenhuysen-Gybels, 2003,1; Billiet, 2002). Item parameter drift Within longitudinal (panel or repeated cross-sectional) research, comparing scores, e.g. ‘fear of crime’ levels, across time, is bound to this assumption of measurement invariance as well (Pleysier et al, 2005). According to this very logic, different scores on, for example, ‘fear of crime’ items across time are again not necessarily the reflection of true differences or a tendency, but possibly the outcome of a different or changed 6.

Obviously, cross-cultural invariance is an essential precondition, but it is one of many; there is a general need for measurement invariance between any groups that are to be compared within the survey sample. Within ‘fear of crime’ research, for example, gender bias is potentially a serious issue: a substantial part of the ‘fear of crime’ literature has empirically found and discussed different ‘fear of crime’ levels between men and women. To our knowledge, the assumption of measurement invariance is hardly ever tested (Pleysier et al, 2004). Therefore, gender differences in ‘fear of crime’ are not necessarily a reflection of true differences, but possibly a reflection of a different understanding of the central concept (and its measurement) between men and women.

50

Victimisation and Insecurity.book Page 51 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

understanding of the concept or one or more of its constitutive indicators. The latter is referred to as ‘item parameter drift’, as “one of the forms of differential item functioning (DIF) that can affect objective tests” (Bock et al, 1988, 275). We suggested to refer to the assumption of ‘temporal invariance’, as a specific form of measurement invariance, where items are tested on their potential parameter drift, or their differential functioning across time. Item parameter drift has been studied in ‘fear of crime’ item sets, based on the 1998, 2000 and 2002 sweeps of the Security Monitor, and the 1999, 2000 and 2002 sweep of the APS-SCV survey (cf. infra) ; analyses showed that latent ‘fear of crime’ constructs in both surveys are not completely free from item parameter drift and therefore are not temporally invariant (Pleysier et al, 2005). Although, in that contribution, substantive outcome was subordinate to an illustration of the importance of temporal invariance, the main concern should be clear: not checking temporal invariance or the reliability of survey instruments seriously undermines all substantive and contextual interpretation and therefore threatens the validity of the study at stake. Considering again the growing importance of large-scale longitudinal, international surveys within criminology, such as the ICVS or EUICS, it is our firm belief that the application of these techniques not only surpasses the ‘fear of crime’ domain, but also will undeniably gain importance within criminological research and crime and victim surveys in general. Acquiescence The concept of acquiescence is not new as well. Cronbach (1942, 1950) called the tendency of survey respondents to agree with items independent of the content of the item a response set or style factor. It is “the disposition of a respondent to respond systematically to questions in another way than he would if the questions were offered in another form” (Billiet, McClendon, 2000, 610). Respondents are asked to indicate their level of agreement on a set of items. Acquiescence – or agreeing bias – is the tendency to agree with the items offered, independent of the direction in which the item is formulated (positive or negative). The presence of acquiescence when using unbalanced scales (all negative or all positive) could lead to systematic overestimation or underestimation of the covariance between latent concepts (Mirowsky, Ross, 1991; Pleysier et al, 2003). Researchers are advised to use balanced scales whenever possible, so they can take the effect of acquiescence into account and check on the possible interference of style factors with background variables (such as SES, age, gender, etc.). So far, such research has not been conducted on Security Monitor data. However, a small-scale test in a questionnaire on fear of crime at the

51

Victimisation and Insecurity.book Page 52 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

University Campus of Heverlee (Pleysier et al, 2003) has illustrated the possibility of this kind of measurement error. It is not clear to what degree this could potentially effect substantial results in the Security Monitor. Use and implementation of the Security Monitor Stimulation of the use of the Security Monitor in Belgium The government had as an explicit goal to use the Security Monitor on the evaluation of policy. Therefore however, correct use of the data is crucial. To improve correct use of the measurement instrument, several research projects have been conducted. In this case it is worth mentioning the ‘Action-research’ (Ponsaers et al, 2001) that was conducted in order to improve understanding of how the Security Monitor is used and could be used, more specifically at the local level. Recently, the research project ’Mini-Monitor‘ on demand of the Federal Science Policy, is finalized. In the Flemish and Walloon part of the country, researchers have conducted research and advised on the further development of the Satefy Monitor. Ponsaers (University of Ghent), Lemaître and Born (Liege University) were the promoters of this recently finished project7. Key objectives were to organise focus groups in which the instrument was carefully discussed. In this research project the exclusive use of landline phones has once again been identified as a major problem. A suggestion was made to conduct face-to-face interviews to discover possible bias (mixed mode survey). Divergence between police records and results from the Security Monitor It is also worth mentioning that a small scale research project has been conducted by a research team of the University of Hasselt, where they were trying to find out how official crime data stemming from police registrations and results from the Security Monitor can be used and simultaneously (Vandersmissen et al, 2004). This is especially relevant for local policy makers, who sometimes see themselves confronted with different trends based on analyses of both data. Valid relations between phenomena stemming from the Security Monitor and the official police statistics can only be studied when the data from the Security Monitor and the police records could be matched at the individual level. At the aggregate level, there is a potential danger to commit the so called ecological fallacy (Robinson, 1950). To analyse the relation between ‘fear

7.

A summary can be found on: http://193.191.208.76/belspo/home/publ/pub_ostc/ agora/raggg104_nl.pdf

52

Victimisation and Insecurity.book Page 53 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

of crime’ and officially recorded crime, multilevel analysis is sufficient (see below). One important source of differences between both data is a consequence of the way both data sets are created. To illustrate this problem we focus on the way both data sets are composed:

Source: Vandersmissen et al, 2004

Figure 1. Divergence between police records and the Security Monitor

Vandersmissen et al (2004) suggest further analyses to validate the data of the Security Monitor. From a scientific point of view Vandersmissen et al (2004) suggest to further examine where crimes are being reported; questioning the exact place of crime commitment in the Security Monitor is crucial for a comparison between survey data and official records. It became clear that future research should take this discrepancy serious, especially when comparisons are being made at the local level. Besides that, differences between penal definitions and questions used in surveys impede correct comparison. Multilevel analysis on the Belgian Security Monitor Researchers are increasingly realising the fact that the analysis of largescale surveys has to gain with the use of more complex methodologies and strategies of analysis. One increasingly popular technique is the use of multilevel modelling, as a simultaneously and statistically correct way to analyse contextual (area) effects, effects of individual level variables on

53

Victimisation and Insecurity.book Page 54 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

crime victimisation and ‘fear of crime’. To test the possibilities and to evaluate the potential gain of using this methodology, a project of the Federal Science Policy ordered a research that was conducted by Hajnal et al (2000). Data were drawn from the 1998 Security Monitor. The main reason for using multilevel modelling is the fact that the Belgian Security Monitor is a multi-stage sample, which implies a certain degree of clustering. It is known that (geographical) clustering can affect the substantial results from bivariate and multivariate analyses on the data. Therefore ‘fear of crime’ and victimisation were explained by using variables measured at the individual and municipal level. Results of the individual analyses of ‘fear of crime’ (avoiding behaviour) and of victimisation did not deviate seriously from the results of the combined multilevel analysis. The results show that if we try to explain feelings of lack of safety and victimisation among individuals, aggregate police statistics (in this case four types of facts recorded by the police) play no role whatsoever. Another striking result was that an increase in the percentage of foreigners in a community increases ‘fear of crime’, without having an impact on victimisation. Multilevel analyses are however very scarce in Belgium, in contrary to the situation in the Netherlands (cf.infra).

2.

APS-SCV Survey

Another important repeated cross-sectional survey is the APS-SCV survey (Administratie Planning en Statistiek – Sociaal Culturele Veranderingen) of the Flanders Authority. In contrast to the Security Monitor, the APS-SCV survey’s main interest is not crime or victimisation; since 1996, it is an annual ‘barometer’ of socio-cultural changes among Dutch speaking inhabitants of the Flemish Community or the Brussels Capital Region. The main interest, therefore, is a broad one, dealing with values, attitudes and opinions of the Flemish on a number of relevant topics. There are some other important differences compared to the Security Monitor. The APSSCV survey is not a federal survey, but concentrates on only one part of the federal state, i.e. Flanders. Furthermore, the data are gathered in face-toface surveys (representative sample of 1,500 respondents) with a questionnaire in Dutch only. As the APS-SCV is an annual survey, some of the question blocks rotate; this is the case for the ‘fear of crime’ item set. This set of eight Likert type items (see Table 3) was used in the questionnaire of the 1999, 2000, 2002 and 2004 sweep of the survey. Previously, we reported on secondary analyses investigating the temporal invariance assumption of the ‘fear of crime’ items of the 1999, 2000 and 2002 round (Pleysier et al, 2005).

54

Victimisation and Insecurity.book Page 55 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

item

Variable

FOC

V90_1

Nowadays it is not safe to send children out alone in the streets. Out of fear for robbery, as I get in my car, I always immediately lock it. In the evening you ought to be extra careful on the street. In the past ten years, the streets have become more unsafe. The police is no longer able to protect us from criminals. In the evening and at night, I no longer open the door if the doorbell rings. These days, a burglary alarm is no excessive luxury. When on holiday, I do not dare to leave my home unprotected. (totally agree/agree/agree nor disagree/disagree/totally disagree)

V90_2 V90_3 V90_4

v90_5 V90_6 V90_7 V90_8

Table 3. ‘Fear of crime’ item set in the APS-SCV survey

3.

Supranational surveys

The International Crime Victims Survey (ICVS) began development in 1987, in order to improve international comparative knowledge and research, and is by now the most far-reaching programme of fully standardised surveys looking at householders’ experience of crime in different countries. Belgium participated in the 1989, 1992 and 2000 sweeps of the International Crime Victim Survey (ICVS). Data of the ICVS are available online8. At this very moment the ICVS has been conducted over a period of 12 years in 25 industrialised countries, with more than one sweep in most of the participating countries. Over a slightly shorter time span, there have also been surveys in 46 cities elsewhere around the world. All told, there have been about 140 sweeps of the ICVS around the world. These have involved interviews with over 200,000 respondents, of which 110,000 were interviewed in industrialised countries. In Belgium, contrary to the situation in the Netherlands (see below), the data of the ICVS have never been used for more detailed analysis. Belgium also participated in the 1996, 2000 and 2002 sweeps of the Eurobarometer ‘Public Safety’9. Also worthwhile mentioning is the fact that Belgium participated in 2005 in the European Crime and Safety Survey 8. 9.

http://ruljis.leidenuniv.nl/group/jfcr/www/icvs/Index.htm Key findings can be consulted online at http://ec.europa.eu/public_opinion/archives/ ebs/ebs_100_en.pdf

55

Victimisation and Insecurity.book Page 56 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

(EUICS), conducted by Gallup-Europe. The EUICS is a tool for measuring the volume and nature of crime in Europe. Detailed information is also on line available10. To achieve an accurate picture of crime in the EU, surveybased comparative measurement of a large sample of the European public is required, in order to build a sound knowledge-base of European crime trends, and to provide tools for evidence-based policy research related to the basic security of European citizens. The EUICS consortium combines leading European research centres with a proven track of data collection and analysis. Each consortium member has been working on previous projects collecting, analysing and evaluating policy options based on comparative surveys of crime. The project follows on from the International Crime Victims Surveys (ICVSs) which all members of the consortium have previously participated in, and developed a new instrument, the European Crime and Safety Survey (EUICS). The latter addresses new needs for European comparative data for policy-makers.

II.

Overview of crime victim and insecurity surveys in the Netherlands

1.

The climate of ‘management of fear’ in The Netherlands

To introduce the current context in the Netherlands, a critical analysis conducted by Rene van Swaaningen (2005) on the safety discourse in big cities in the Netherlands, is used as a guideline. In the Netherlands, a safety discourse has been developing since the nineties, but became very explicit around the turn of the millennium. Van Swaaningen (2005) states how politicians began to describe safety in a negative sense. Political parties, as happened in Belgium, started to emphasize that safety is about getting rid of those groups (living in the periphery of society), which form a threat to the general public feeling of safety. Results from crime victim surveys revealed that the perception of urban disorder strongly affects feelings of unsafety. The notions of safety and security are interpreted through ethnic lines. Van Swaaningen stresses the importance of three events, namely September 11th 2001, the day of the attack on the World Trade Centre in the USA, May 6th 2002, the day when Pim Fortuyn (a Dutch right wing politician, the leader of LPF), was murdered, and November 2nd 2004, the day when film director Theo Van Gogh was murdered by a young DutchMoroccan radical Islamist. Fortuyn introduced the term ’liveability’ 10. see: http://www.gallup-europe.be/euics/Xz38/default.htm

56

Victimisation and Insecurity.book Page 57 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

(leefbaarheid in Dutch), suggesting that liveability had declined seriously in The Netherlands. The political agenda of all governing parties was seriously stirred up and liveability became a major point in the political agenda. This shift in attention from ‘crime prevention’ to ‘managing fear’ was explicitly visible in the local fight against crime in large cities. Van Swaaningen (2005) describes the example of the Rotterdam Safety Plan, where key terms such as fighting incivilities and street crime are the heart of the plan, which consists of very extreme and sometimes unrealistic goals. Striking is the fact that van Swaaningen (2005) reports that the Rotterdam safety index has been showing a decrease in crime and incivilities since the start and implementation of the plan. It is within this context that crime and ‘fear of crime’ are being monitored as never before.

2.

The CBS victim survey

The CBS victim survey (Enquête Rechtsbescherming en Veiligheid, Central Bureau of Statistics) is nowadays an integral part of the POLS (Permanent Onderzoek Leefsituatie, Permanent Research Life Situation) is the longest running victim survey in the Netherlands (Bijleveld, 2005). In 1980 it was decided to change the existing WODC victim survey into the CBS survey. Since that time, some changes have been conducted. The sample frame is the Municipal Basic Administration and the questionnaires are handled through face-to-face interviews, contrary to the situation in Belgium. Units of analysis are households and within every household an individual is selected randomly for interviewing. The questionnaire consists of following issues: ‘fear of crime’, victimisation, reporting behaviour, crime prevention measures, and also social assistance.

3.

The Police Monitor

The Police Monitor (Politiemonitor Bevolking, PMB) had a serious impact on the development of the Belgian Security Monitor. It can be said that the Belgian national survey was largely inspired by this Dutch example. The Police Monitor uses a very large sample; the latest sweep consisted of ca. 88,000 respondents. The Police forces use the results of this large-scale survey to measure appreciation of their functioning. Traditionally ‘fear of crime’ and crime victimisation are measured. In every police region the survey is conducted. Since 1993 the survey took place every second year, but since 2001, the survey is conducted every year. The Police Monitor uses telephone interviewing. 57

Victimisation and Insecurity.book Page 58 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

4.

The Dutch Safety Monitor

The secretaries-general of the Departments of the Interior, Royal Relations and Justice and the director-general of the Central Bureau of Statistics (CBS) have recently (early March 2007) signed an agreement which is called Stroomlijning Veiligheidsmonitors. In this agreement rules and appointments have been written down concerning the development of one joint Safety Monitor. This project aims at reducing the existing and aforementioned large scale surveys to one general survey, in which crime victimisation, ‘fear of crime’, neighbourhood problems and opinions on police functioning will be questioned. This common survey will be the first Dutch general survey on the aforementioned issues. The survey will be held at the national, regional and local level and will consist of some facultative and obligatory modules. The project aims at a participation of at least 50% of the 31 largest municipalities, 50 middle range municipalities and small municipalities, during 2007. Later, this percentage will be augmented.

5.

International survey participation of the Netherlands

Similarly to the Belgium case, the Netherlands have been participating in several sweeps of the ICVS. To be correct: the Netherlands joined ICVS in the 1989, 1992, 1996 and 2000 sweeps. Furthermore, the Netherlands have played a crucial role in the development and reporting of the key findings on the ICVS, with support of the Dutch WODC (Scientific Research and Documentation Centre) and the NSCR (Netherlands Institute for the Study of Crime and Law Enforcement) (Van Kesteren et al, 2000; Nieuwbeerta, 2002).

6.

The use of crime victim survey data in the Netherlands

Crime victim surveys have been used frequently in recent years for both multivariate theory based analyses and analyses concerning the improvement of measurement issues. Especially the number of multilevel studies to measure the so-called contextual effects of area characteristics exceeds by and large the use of multilevel modelling in Belgium. In the Netherlands, the focus has been lying on testing the relationship between indicators of social capital and victimisation. These studies seem to suggest the existence of small but substantial contextual effects of area characteristics on crime victimisation. For a good overview of studies we refer to the doctoral dissertation of Johan Van Wilsem (2005) and to the 58

Victimisation and Insecurity.book Page 59 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

work of Karin Wittebrood (2000, 2001). Nieuwbeerta (2002) edited a highly recommendable book, which presents results from the ICVS in a comparative perspective.

7.

Measurement issues in the Netherlands

Researchers in the Netherlands have by and large good experience with method issues, due to the detailed cooperation and analysis of data stemming from the ICVS, Police Monitor and CBS survey. In what follows, we bring some recent research under attention that deals with method issues. Goudriaan (2006) has recently examined in detail reporting behaviour of crime victims. In her doctoral dissertation she has developed and tested a socio-ecological model of reporting behaviour, in which aspects of the social and ecological context are taken into account. Also, the work of Vanderveen (2006) is of specific interest to researchers interested in methodological issues in surveying crime and insecurity in general. Vanderveen (2006) has devoted her doctoral thesis to the measurement of ‘fear of crime’. She carefully distinguishes attempts to improve traditional survey scales, measuring several dimensions of the complex construct of ‘fear of crime’, but proposes an original alternative for the improvement of conceptualising and measuring different dimensions of fear and insecurity. Vanderveen (2006) explicitly criticises the traditional conceptualisation of the complex concept of ‘fear of crime’. Further, she pays a lot of attention to the evaluation of measurement instruments in function of their reliability and validity. Vanderveen distinguishes two strategies concerning the measurement of the multidimensional concept of fear of crime. The first strategy departs from existing measurement instruments and stresses continuity and improvement of traditional indicators. She argues that traditional instruments do not allow for expressing ‘fear of crime’ in a single index, as the concept is multidimensional by definition. What is needed when following the traditional measurement is the detailed analysis of reliability and validity of indicators and constructs. The second strategy, and this is a big challenge, is to start all over and create a new measurement instrument as a part of a conceptual analysis of how people perceive unsafety. One advantage of the traditional strategy is the fact that one can compare results over time. Which strategy to use, depends on the need of the researcher. Vanderveen’s new instruments (2006) are useful when the researcher’s interest is lying in the explanation of individually observed differences, rather then the mere observation of percentages of respondents that are afraid of situations. The new measurement instrument is based on the measurement of different attitudes and personal judgement of situations. 59

Victimisation and Insecurity.book Page 60 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Vanderveen (2006) tested and evaluated her alternative strategy and reported in detail about this evaluation in her dissertation.

III. General conclusion on good practises in Belgium and the Netherlands It is fair to say that the development of a federal instrument measuring crime and insecurity emerged rather late in Belgium, compared to other European countries and the situation in the USA. Monitoring safety is done at the national level since 1997 in Belgium, inspired by the Dutch Police Monitor. In the Netherlands several instruments already existed and serious efforts to repeatedly monitor ‘fear of crime’ go back to the early eighties. Large scale instruments exist, of which the Dutch Police Monitor has about 80,000 respondents. Nowadays, an attempt is made to integrate these existing large-scale datasets into one joint survey, which is also called the Safety Monitor. The use of results by (local) policy makers, both in Belgium and the Netherlands, is not unproblematic. One of the reasons is the observed difference between survey results and results of official data. Policy makers need to be thoroughly aware of the possibilities and restrictions of the use of survey instruments when evaluating local policies. Action research can be seen as a successful way to stimulate correct use of data stemming from victim surveys. Issues of measurement (reliability and validity studies) have only rather recently been of considerable interest to social scientists in Belgium and the Netherlands. Within this contribution, we have highlighted several studies and contributions concerning key issues of measurement, and using stateof-the-art methodologies, which can be inspiring to develop good practises. Future research should pay attention to evaluate the possibilities of conducting mixed mode surveys, in order overcome the problem of noncoverage. More knowledge is needed on the existence of differences between households that exclusively have a mobile phone, or landline phone, with respect to victimisation and ‘fear of crime’. Researchers in Belgium and The Netherlands are fully aware of the conceptual complexity of measuring fear and crime; they invest in testing existing instruments and creating new instruments. The issue of comparability of results is one major point that will influence decisions to conduct major changes in questionnaire designs and questions asked. We have to continue to realise that without thorough investment in conceptual and measurement issues, contextual and policy-relevant conclusions remain somehow problematic. Although this implies serious efforts and also financial costs we can not but conclude that such investments are in favour of a correct use of survey results. 60

Victimisation and Insecurity.book Page 61 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

References ASSAEL H., KEON J., 1982, Nonsampling vs Sampling errors in Survey Research, Journal of Marketing, 45, 114-123. BIJLEVELD C. 2005, Methoden en Technieken van Onderzoek in de Criminologie. Den Haag, Boom Juridische Uitgevers. BILLIET J.B. 2002, Cross-cultural equivalence with structural equation modeling. In HARKNESS J.A., VAN DE VIJVER F.J.R., MOHLER P. PH. (Eds.).Cross-Cultural Survey Methods. New Jersey, Wiley-Interscience. BILLIET J.B., DE WITTE H., 2001, Wie stemde in juni 1999 voor het Vlaams Blok en waarom, Tijdschrift voor Sociologie, 22, 1, 24, 5-35. BILLIET J.B., MCCLENDON M.J. 2000, Modelling Acquiescence in Measurement Models for Two Balanced Sets of Items. Structural Equation Modelling. 7, 4,608-628. BILLIET J.B., PLEYSIER S., PICKERY J. HAJNAL I. 1998, Methodologische ondersteuning van de veiligheidsmonitor. Bulletin nr. 1998/30 van het ISPO, 69 pp. + bijlagen. BLOCK R. 1993, A Cross-National Comparison of Victims of Crime, Victim Surveys of Twelve Countries. International Review of Victimology 2, 183207. BOCK R.D., MURAKI E., PFEIFFENBERGER W. 1988, Item Pool Maintenance in the Presence of Item Parameter Drift. Journal of Educational Measurement, 25, 275-285. BRAUN M. 2003, Errors in comparative survey research: an overview In HARKNESS J.A., VAN DE VIJVER F.J.R., MOHLER P. PH. (Eds.). Crosscultural survey methods, Hoboken (NJ), John Wiley & Sons, 137-142. COUPER M., DE LEEUW E. 2003, Non-response in cross-cultural and crossnational surveys In HARKNESS J.A., VAN DE VIJVER F.J.R., MOHLER P. PH. (Eds.) Cross-cultural survey methods, Hoboken (NJ):John Wiley & Sons157-178. CRAWFORD A. (ed.) 2002, Crime and Insecurity. The governance of safety in Europe, Cullompton, Devon, Willan Publishing. CRONBACH L.J., 1942, Studies of acquiescence as a factor in the true-false test, Journal of Educational Psychology. 33, 401-415. CRONBACH L.J. 1950, Further evidence on response sets and test design, Educational and Psychological Measurement. 10, 3-31. DEVROE E., ELIAERTS CHR., 2006, Die criminografie-schaarste in België, in DEVROE E., BEYENS E., ENHUS E., (Eds.), Zwart op wit? Duiding van cijfers over onveiligheid en strafrechtsbedeling in België, Handbook criminografie, VUB-Press Politeia, 13-28. DILLMAN D.A., ELTINGE J.L., GROVES R.M., LITTLE R.J.A. 2002, Survey Nonresponse in Design, Data Collection and Analysis. In GROVES R.M., 61

Victimisation and Insecurity.book Page 62 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

DILLMAN D.A., ELTINGE J.L., LITTLE R.J.A. (Eds.) Survey Nonresponse, Hoboken (NJ), John Wiley & Sons, 3-26. DITTON J., FARRALL S. (Eds.) 2000, The Fear of Crime, Aldershot, Ashgate, The International Library of Criminology, Criminal Justice & Penology. DITTON J., FARRALL S., BANNISTER J. GILCHRIST E. 2000, Crime surveys and the measurement problem: Fear of Crime In JUPP, V., DAVIES P., FRANCIS P. (Eds.) Doing Criminological Research, London, Sage, 399-413. FERRARO K., LAGRANGE F. R. 1987, The Measurement of Fear of Crime Sociological Inquiry 57, 70-101. GOETHALS J., DE BOYE A., FIJNAUT C. 1999, Criminaliteit en onveiligheidsgevoelens in de Limburgse ex-mijngemeenten, Leuven, Universitaire Pers Leuven. GOETHALS J., PONSAERS P., BEYENS K., PAUWELS L., DEVROE E., 2002, Criminografisch onderzoek in België, in BEYENS K., GOETHALS J., PONSAERS P., VERVAEKE G., (Eds.), Criminologie in actie, Brussels, Politeia, 189-206. GOUDRIAAN H. 2006, Reporting Crime. Effects of social context on the decisions of victims to notify the police, Leiden, NSCR. GROVES R.M. 1989, Survey errors and survey costs, New York, John Wiley. HAJNAL I., PICKERY J., BILLIET J. 2000, Multilevelanalyse op de gegevens van de Veiligheidsmonitor, Onderzoeksverslag van de Afdeling voor Dataverzameling en Analyse, KULeuven, Faculteit Sociale Wetenschappen. HALE, C. 1996, Fear of Crime: A review of the literature, International Review of Victimology 4, 79-150. HEBBERECHT P., HOFMAN H., PHILIPPETH K., COLLE P. 1992, Buurt en Criminaliteit, Vanden Broele/v.z.w. Politeia. HEBBERECHT P., PELFRENE E., LIPPENS R. 1986, Onveiligheid – Criminaliteit – Burgerwacht, Eindrapport, Gent, Rijksuniversiteit Gent, School voor Criminologie, Onderzoek uitgevoerd in opdracht van de Minister van Binnenlandse Zaken. JANSSEN C., KEGELS M.L., KELLEN M., SOUGNE L. 1981, La peur et la ville. Rapport d'une enquête sur le sentiment de sécurité des Liégeois. Liège: Université de Liège, Faculté de Droit, d'Economie et de Sciences Sociales. KESTEREN J. VAN, MAYHEW P., NIEUWBEERTA P. 2000, Criminal Victimisation in Seventeen Industrialised Countries. Key findings from the 2000 International Crime Victims Survey, The Hague, WODC. LEE M., 1999, The fear of crime and self-governance : towards a genealogy, Australian and New Zealand Journal of Criminology, 32, 3, 227-246. LEMAÎTRE A., 1982, Le sentiment de sécurité des Liégeois, Liège, U. Liège, mémoire inédit de licence de criminologie. 62

Victimisation and Insecurity.book Page 63 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Belgium and the Netherlands

LOOSVELDT G., CARTON A., BILLIET J. 2004, Assessment of survey data quality: a pragmatic approach focused on interviewer tasks, International Journal of Market Research, 46, 1, 65-82. MADDENS B. 1998, De partijkeuze en de beleidsthema’s. In SWYNGEDOUW, M., J. BILLIET, A. CARTON & R. BEERTEN (Red.), De (on)redelijke kiezer. Onderzoek naar de politieke opvattingen van Vlamingen, Verkiezingen van 21 mei 1995, Leuven, Acco, 139-158. MARKER D., 2002, Improving quality of surveys, Copenhagen, Paper presented at the International Conference on Improving Surveys. MIROWSKY J., ROSS C.E. 1991, Eliminating Defense and Agreement Bias from Measures of the Sense of Control: A 2x2 Index, Social Psychology Quarterly, 54, 2, 127-145. MOLENBERGS G., VANDERMISSEN V., THIJS H., 2005, De Veiligheidsmonitor : Validatie, Analyse en Adviezen, Hasselt, Limburgs Universitair Centrum, Centrum voor Statistiek. NIEUWBEERTA P. (Ed.) 2002, Crime Victimisation in comparative perspective. Results from the International Crime Victims Survey, 1989-2000, Den Haag, Boom Juridische uitgevers. PAUWELS L., PLEYSIER S. 2005, Assessing cross-cultural validity of fear of crime measures through comparisons between linguistic communities in Belgium, European Journal of Criminology, 2, 2, 139-159. PLEYSIER S., BILLIET J.P., 2005, Data quality assessment in ESS Round 2. Between whishes and reality : Closing the gap?, Report of data quality assessment of ESS Round 2 PLEYSIER S., PAUWELS L., VERVAEKE G., GOETHALS J. 2005, Temporal invariance in repeated cross-sectional ‘fear of crime’ research, International Review of Victimology, 12, 3, 273-292. PLEYSIER S., VERVAEKE G., GOETHALS J. 2002, Het ‘onveiligheidsgevoel’ onderzocht. Groeipijnen van een onderzoekstraditie in wording In BEYENS K., GOETHALS J., PONSAERS P., VERVAEKE G. (Eds.) Criminologie in actie, Brussel, Politeia, 189-206. PLEYSIER S., VERVAEKE G., GOETHALS J. 2004, Cross-cultural Invariance and Gender Bias when Measuring ‘Fear of Crime’, International Review of Victimology, 10, 3, 245-260. PLEYSIER S., VERVAEKE G., GOETHALS J. 2006, Hoe ‘representatief’ is deVeiligheidsmonitor? Over de accuraatheid van telefonische surveys in Vlaanderen, Panopticon, 27, 4, 107-114. PLEYSIER S., VERVAEKE G., GOETHALS J., BILLIET J. 2003, Assessing measurement invariance in ‘fear of crime’ instruments In VANDERHALLEN M., VERVAEKE G., VAN KOPPEN P.J., GOETHALS J. (Eds.) Much ado about crime. Chapters on Psychology and Law, 75-90. Brussel, Politeia. 63

Victimisation and Insecurity.book Page 64 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

PLEYSIER S., VERVAEKE G., GOETHALS J., PAUWELS L. 2003, ‘Acquiescence’. Het meten en modelleren van volgzaamheid in twee gebalanceerde ‘angst voor criminaliteit’-schalen Panopticon, 24, 6, 547-562. PONSAERS P., VERVAEKE G., GOETHALS J. (Eds.) 2001, DeVeiligheidsmonitor. Behoeftendetectie bij de bevolking, Brussel, Politeia. ROBINSON, W.S. 1950, Ecological Correlates and The Behavior Of Individuals, American Sociological Review, 15 (3), 351-357. STOOP I.A.L. 2005, The Hunt for the Last Respondent. Nonresponse in sample surveys, The Hague, Social and Cultural Planning Office of the Netherlands. TARIS, T.W. 2000, A primer in longitudinal data analysis, London, Sage. UNIOP, 1985, Onveiligheid, een idee of werkelijkheid, Brussels, UNIOP. VAN DEN BOGAERDE E., VAN DEN STEEN I. 2006, Veiligheidsmonitor 2006. Handboek Veiligheidsmonitor, Brussel, Federale Politie, Algemene Directie Operationele Ondersteuning, Directie van de nationale gegevensbank, Dienst Beleidsgegevens. VAN SWAANINGEN, R. 2005, Public safety and the management of fear, Theoretical Criminology 9, 289-305. VAN WILSEM, J. 2003, Crime and Context, The Impact of Individual, Neighborhood, City and Country Characteristics on Victimisation ICS Dissertation Series. VANDERHALLEN M., PLEYSIER, S., VERVAEKE, G., GOETHALS, J. 2000, Verbetering van de vragenlijst van de Veiligheidsmonitor en meer in het bijzonder de module politiefunctioneren. Panopticon, 21, 3, 276-281. VANDERSMISSEN V., THIJS H., BRUCKERS, L., MOLENBERGHS G. 2004, Wetenschappelijk advies over de vergelijkbaarheid van de veiligheidsmonitor en de criminaliteitsstatistieken, Onderzoeksrapport Leuvens Universitair Centrum. VANDERVEEN G. 2006, Interpreting Fear, Crime, Risk and Unsafety, Boom Legal Publishers. WARR M., 1992, Altruistic fear of victimisation, Social Science Quaterly, 73, 723-736. WELKENHUYSEN-GYBELS J. 2003, The detection of differential item functioning in Likert score items, Doctoraal proefschrift, Katholieke Universiteit Leuven, Faculteit Sociale Wetenschappen, Departement Sociologie. WITTEBROOD K. 2000, Buurten en geweldscriminaliteit. Een multilevelanalyse, Mens en maatschappij 75, 92-107. WITTEBROOD, K. 2001, Onveiligheidsbeleving en slachtofferschap in westerse geïndustrialiseerde landen Een multilevel-analyse. Tijdschrift voor criminologie 43, 144-157.

64

Victimisation and Insecurity.book Page 65 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain Mike Hough

This paper summarises survey research on victimisation and on insecurity in Britain over the last thirty years. It focuses on the British Crime Survey (BCS), for the simple reason that this survey has, since its first report (Hough and Mayhew, 1983), been the main source of information on these topics in Britain. It should be said at the outset that the British Crime Survey is a misnomer, because it has covered only England and Wales, with the exception of the first sweep. I make some passing reference to work in Scotland and Northern Ireland, both of which have different administrative and criminal justice systems, but of the most part, I shall be talking about experience in England and Wales. The paper falls into two unequal parts. In the first two thirds of the paper I shall take stock of what surveys have been mounted on victimisation and ‘fear’ both nationally and locally. Secondly I shall discuss the political and academic uses found for this body of research, and see what promise the future holds for crime surveys. I have written this paper as someone who was closely involved with the origination and design of the survey, and in the analysis of the first two sweeps. I contributed a module to the third sweep, and had overall management responsibility for the 1992 survey. Since then I have been less closely involved, designing and analysing modules as an academic contractor, and conducting limited secondary analysis. The paper does not claim in any way to be an ‘official’ account of the BCS.

I.

Taking stock of surveys of victimisation and ‘fear’

1.

National victimisation surveys

The first national survey to carry questions about victimisation was the General Household Survey – a governmental omnibus survey – which included a few questions about experience of burglary throughout the

65

Victimisation and Insecurity.book Page 66 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

1970s. This provided some – albeit limited – insight into reporting and recording rates for one type of crime. However, it was not until the early 1980s that the BCS began to provide a more comprehensive account of crimes against individuals and their property. Since then the BCS has been the main national survey on crime issues, although there have been surveys in both Scotland and Northern Ireland. The impetus for the BCS was undoubtedly the US National Crime Victimisation Survey1. The Home Office intermittently considered the possibility of a nationwide survey in Britain throughout the 1970s. This was ruled out at the time, however, mainly on cost grounds, but also because of anxieties about the political snares that survey estimates of unrecorded crime might create. But in 1980 the Research and Planning Unit of the Home Office and its Crime Policy Planning Unit mounted a strong case for a survey. The case rested on three arguments: • police records were an unreliable index of crime • a survey might help correct public misperceptions about crime levels, trends and risks and provide a more informed picture of crime • a national survey would give a boost for criminological research and theory. It was agreed in July 1981 that the British Crime Survey (BCS) should proceed. A budget of £230,000 was allocated – roughly equivalent to £700,000 (or ?1m) at present prices. The survey methodology was developed at some speed – borrowing on the US National Crime Victimisation Survey and the expertise of American technical consultants, including Wes Skogan, Al Biderman and Al Reiss. The first sweep of the survey was conducted in early 1982, with a nationally representative sample of 10,905 respondents in England and Wales (Hough, Mayhew, 1983). At the time, it appeared to the BCS designers that there were two models that the BCS could adopt. The ‘social indicators’ model had been pursued by the NCVS. The aim here was to create a new index of crime that served simply as a social indicator. The alternative model was to design the survey as a criminological research instrument, with enough flexibility to address a range of policy-relevant research questions. We decided to adopt this model, the defining feature of which was a modular structure. The questionnaire comprised: • A ‘main questionnaire’ asking a fixed series of screening questions about victimisation • A number of ‘victim forms’ collecting details about victims’ experience

1.

See Hough et al., (2007) for a fuller account of the origins of the survey.

66

Victimisation and Insecurity.book Page 67 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

• Various ‘follow up’ questionnaires asked questions about crime-related topics, which could be changed from sweep to sweep. The follow-up questionnaires are asked of various sub-samples, whose size depends on the degree of precision required for the topic covered. Information collected in follow-up questionnaires has included: people’s sense of insecurity; their experience of the police; attitudes to punishment; crime preventive behaviour; and self-reported drug use. Scotland carried out a parallel survey, using the same contractors and an almost identical questionnaire. Following the first BCS in 1982, the survey was repeated two years later in 1984 (Hough, Mayhew, 1985); a third sweep followed in 1988 (Mayhew et al., 1989), and a fourth in 1992 (Mayhew et al., 1993). It was then conducted in alternate years until 2000, when the sample size doubled – to about 20,000 (Kershaw et al., 2000). In 2001 it moved to being a continuous ‘rolling’ annual survey with 40,000 interviews conducted throughout the year (Kershaw et al, 2001; Simmons et al., 2002; Simmons, Dodd, 2003).2 One reason for the increased sample size was to provide more reliable measures of different forms of violent crime. Another was the perceived need for numerical ‘performance indicators’ relating to levels of public confidence in the police in each of the 43 police force areas in England and Wales. The much larger sample size allows tolerably precise survey estimates for overall household and personal crime at individual police force level. The possibility of developments of this sort had never occurred to us in the early days of the BCS, when we promoted the survey largely as a research tool, and in no sense as a means of performance management. We shall return to this issue in Part 2. After the first sweep in 1982, the Scotland Crime Survey has tended to diverge increasingly from the BCS as carried out in England and Wales. Surveys have been carried out in 1988, 1993, 2000, 2003 and 2005. The most recent one, in 2005, moved to telephone interviewing. It encountered problems in assembling an adequate sample, missing both those people who listed themselves as ex-directory and those households who only used mobile phones (Hope, 2005). Figure 1 summarises BCS findings on crime trends in England and Wales. The BCS shows a steady rise until 1995, followed by an equal steady fall.

2.

Originally the survey was mounted in the first three months of the year, respondents being asked to report on their experiences of crime in the previous calendar year. They are now asked about their experiences in the past twelve months. The decision to move to a continuous rolling survey was a function of the scale of the enterprise: no survey company could undertake to conduct 40,000 interviews in three months at reasonable cost.

67

Victimisation and Insecurity.book Page 68 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Most types of crime have fallen since the mid-1990s, and falls in burglary and vehicle crime have been marked.

Source: Walker et al. (2006) Note: the BCS count of ‘All BCS crime’ is numerically larger, for any given year, than the number of crimes recorded by the police in the equivalent crime categories. The graph shows indexed trends, showing percentage changes from levels in 1981. For this reason the ‘All BCS crime’ trend line is lower than the others – because the percentage rise and fall over time has been shallower.

Figure 1: Crime trends in England and Wales from 1981

Victims became increasing more likely to report crimes to the police in the 1980s and early 1990s, with the spread of household insurance and the growth in phone use. Thereafter, the reporting rate has been broadly static, with the exception of violence, for which reporting rates have continued to rise. This is marked in domestic violence – suggesting a cultural shift towards intolerance of this offence. The police started to record proportionately fewer crimes in the early 1990s – probably owing to the prevailing performance management regime, with a heavy emphasis on meeting quantitative targets. This trend was reversed in the late 1990s, when a greater emphasis was placed on fuller recording; and in 2002 the National Crime Recording Standard introduced an explicit policy of full recording, in particular taking victims’ reports at face value. The upward turn in trends in recorded crime thereafter is illusory – with some exceptions. 68

Victimisation and Insecurity.book Page 69 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

2.

Local victimisation surveys

The earliest local crime survey was the London survey conducted by the Cambridge Institute of Criminology in the early 1970s and funded by the Home Office (Sparks, Genn, Dodd, 1977). This was an impressive piece of work, which addressed almost all the key issues relating to surveys of victimisation. It was conceived of as a piece of exploratory academic research, however, and its results did not achieve any significant penetration into policy thinking. The 1980s saw various local surveys. David Smith and colleagues mounted a large-scale study of crime and policing in London (Smith, 1983). Whilst this focussed on policing, it covered both experience of crime and fear of crime. It achieved a significant political impact, though largely as a result of its highly critical assessment of aspects of policing in London. The mid 1980s saw three local crime surveys, all modelled quite closely on the BCS. Farrington and Dowds (1985) published the results of a survey in Nottinghamshire and surrounding areas; they set out to explain why this area consistently returned very high recorded crime rates. They convincingly demonstrated that the area had a slightly higher crime rate than its neighbours, but a much higher rate of recording of reported crimes. Kinsey (1984) reported on the Merseyside Crime Survey, covering the large Northern conurbation with the city of Liverpool at its centre; and Jones et al. (1986) reported on a survey of the London Borough of Islington3. These two surveys achieved a considerable political impact. Although their methods were largely derivative from the BCS, they were skilfully marketed as distinctively different, and were used to demonstrate that crime risks were typically greatest for those with least resources to respond to crime. They made a significant contribution to ‘Left Realist’ criminology, which in turn had a big impact of New Labour’s emerging thinking on criminal policy. The political message from the surveys for the Labour Party was that crime was a significant problems for their traditional electoral supporters, and if Labour was to take power from the Conservatives, it needed to have an electorally convincing position on crime. They added weight to the political position that is well encapsulated by Tony Blair’s (1992) promise to be ‘tough on crime, tough on the causes of crime’. Although further attempts were made to mount further local crime surveys, or to boost local samples of the BCS, no significant ones actually got off the ground in the 1990s – excluding a series of surveys

3.

Islington’s population was around 175,000 at the time.

69

Victimisation and Insecurity.book Page 70 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

conducted for the Metropolitan Police in London, concerned largely with public ratings of the police. To my knowledge only one further large-scale local crime survey has been carried out since the turn of the century, by myself and colleagues (Fitzgerald et al., 2002). This ‘Policing for London Study’ partly replicated David Smith’s earlier London Survey, whilst also being designed to dovetail with the London sub-sample of the BCS. The overall message from this work was that the overall decline in crime in London since the mid-1990s had not been accompanied by improvements in public satisfaction with the police. We attributed this in part to the perverse impact of performance targets, which had had the effect of focussing police attention on ‘volume crime’ at the expense of less serious forms of crime and disorder which were nevertheless of significance to the public.

3.

Surveys on ‘fear of crime’

Although the London survey by Sparks and colleagues included questions relating to insecurity, it was the BCS that put ‘fear of crime’ on the criminological and political map in Britain. I have placed ‘fear’ in quotes, as it a term of questionable value. It is used in Britain as a shorthand label for a wide range of rather different emotions and reactions to crime, ranging from genuine fearfulness, on the one hand, to more diffuse anxiety about crime risks, on the other, and even to concern about crime as a social problem. Early sweeps of the BCS argued that levels of worry about crime were sufficiently high, at least for some groups, to constitute a problem in its own right (see eg Hough and Mayhew, 1983, 1985). The early BCS reports offered what in hindsight seem rather naïve analyses of fear of crime, with an implicit suggestion that people’s worries about crime were irrational and exaggerated. We never stated this explicitly in the early BCS reports, though the subtext was perfectly clear. It is reasonable to think of this perspective on ‘fear’ as being a fairly direct import from the US. This theme was picked up initially with enthusiasm by national and local politicians and by the police. However, detailed analysis of later sweeps has substantially dented empirical support for the ‘irrational fears’ viewpoint (e.g. Hough, 1995). In particular, the notion that the elderly (generally at low risk) are most fearful is underpinned only by trends in the measure of ‘feeling unsafe on the streets’ – which may reflect their more generalised anxieties about negotiating public space. On other measures, the BCS consistently shows that people with the highest levels of concern about crime are those who have most to worry about. For example, young men

70

Victimisation and Insecurity.book Page 71 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

worry most about their cars being attacked – reflecting both the way in which youth lifestyle and housing access put them at risk, and particular emotional and financial investment that young men make in their cars. My (1995) conclusions were in broad terms, that “the right people worried about crime” relative to risk, and that it was hardly an empirical matter to decide whether they were worrying the right amount4. Despite this, the doctrine that ‘fear of crime is a problem in its own right’ has established itself firmly within policy thinking both within the police and amongst local Crime and Disorder Reduction Partnerships. By contrast, it has been largely dropped from national political debate since 1992. The reasons are not hard to find: when the two main political parties are vying for poll position as tough on crime, neither are likely to suggest that public concerns are exaggerated. And, of course, some politicians may genuinely believe, as I do, that the evidence is flimsy that fear of crime is disproportionate to the real scale of the problems. The BCS has shown beyond doubt that levels of worry associated with particular forms of crime have fallen in line with falling crime since the mid-1990s. For example, the proportion who are “very worried” about burglary has halved from 1994 (see Figure 2).

Source: British Crime Survey Question: How worried are you about....having your home broken into and something stolen?

Figure 2 : Worry about burglary : % very worried 4.

This was ingeniously mis-reported in the centre-right press as “people were right to worry about crime” – an important difference!

71

Victimisation and Insecurity.book Page 72 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Local or smaller scale surveys have also focussed on fear of crime. The local crime surveys associated with Left Realist criminology in the 1980s unsurprisingly, given the analysis I have offered above, provided an analysis which turned its back on the thesis that fear was a problem in its own right. Feminist criminology equalled tended towards scepticism that anxieties about crime were exaggerated. Two further strands of work on ‘fear’ and insecurity need discussion. Jason Ditton, Stephen Farrall, Jon Jackson and colleagues have mounted an extensive programme of work that exposes conventional concepts of ‘fear’ to critical scrutiny (eg Ditton et al., 1988, 1999, 1999b; Ditton, Farrall, 2000a, 2000b; Ditton et al., 2003, 2005; Farrall et al., 1997; Farrall and Ditton, 1999, Farrall et al., 2004, Jackson, 2004, 2005, 2006; Jackson et al., 2006). Their work, funded by several Economic and Social Research Council grants, has relied on a mixture of survey work and qualitative interviewing, based largely but not exclusively in Scotland. (They have also mounted a three-wave crime survey in Trinidad.) Key themes to emerge from this work are: • ‘fear’ remains poorly conceptualised • Fear, worry, anxiety and even concern are treated as synonyms • If fear is an emotion – or a mental event – surveys should measure its frequency, rather than its intensity5 • Surveys have very largely failed to contemplate other reactions to crime • Anger in particular has been overlooked, and depending on how you measure it, it is more prevalent than fear6. • Expressed levels of fear are unstable, and it is unclear to what extent this reflects measurement error or genuine variability over time. Secondly, work by Martin Innes and colleagues (Ditton, Innes, 2005; Innes, 2004; Innes et al., 2005) on the ‘signal crimes perspective’ (SCP) has proved of considerable significance, in policy as well as academic terms. People’s sense of security (or insecurity) is the anchoring concept for this work. The key idea is that certain disorder and crime incidents have an especially profound impact upon collective insecurity in an area because they function as signals to people, acting as interpretative indicators of the presence of other risks and threats (Innes, 2004). The perspective has important implications for policing in that if one of the central police objectives is to establish safer communities that also feel safer, then policing had to engage with such matters, and tackle ‘signal crimes’ – however apparently trivial these might appear, relative to forms of objectively more serious crime. 5. 6.

See Hough (2004) for a brief commentary on this. The BCS now routinely reports results on anger about crime.

72

Victimisation and Insecurity.book Page 73 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

SCP found particular resonance at the start of the century with senior police officers, who were becoming increasingly concerned that public confidence in the police was showing no signs of rising in response to falling crime. The reverse was the case, and despite a decade of downward crime trends a majority of the British public remained convinced that crime was rising (see Figure 3). Senior police officers initially, followed thereafter by central government polticians, argued that this so-called ‘reassurance gap’ needed to be closed as quickly as possible, and the Signal Crimes Perspective was incorporated into an ambitious demonstration project entitled the National Reassurance Policing Programme. The aim was to establish a form of local policing that systematically identified signal crimes and disorders that were undermining people’s sense of security, and then took remedial action. A Home Office evaluation judged the programme a success (Tuffin et al., 2006) and a modified version of the programme has been rolled out nationally, as Neighbourhood Policing.

Little more

Lot more

Source: British Crime Survey Question: I would like to ask whether you think that the level of crime in the country as a whole has changed over the past two years. Would you say there is more crime, less crime or about the same amount (since two years ago)?

Figure 3 : Public perceptions of crime trends

The work of Ditton, Farrall, Innes and Jackson has brought increasing sophistication to research on ‘fear’, and a greater recognition of the full complexity of the range of factors that go to construct people’s sense of security – or lack of it – in their homes and neighbourhoods. 73

Victimisation and Insecurity.book Page 74 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

4.

Thematic surveys

Household surveys of victimisation necessarily provide an incomplete picture of crime. In England and Wales, the Home Office has increasingly pursued a strategy of assembling a patchwork of thematic surveys to cover the gaps in coverage of the BCS. Unlike the BCS, most of these are episodic rather than continuous surveys. One of these is the Youth Lifestyles Survey, which provides information on both victimisation and offending for a sample of teenagers and young adults. Two sweeps have been conducted, and a third is planned. A second is the Commercial Victimisation Survey. Surveys of retail and manufacturing premises were conducted in England and Wales in 1994 and 2002, and a further one is planned (Mirrlees-Black, Ross, 1995; Shury, Speed, Vivian, Kuechel, Nicholas, 2005). There is also the Offending and Criminal Justice Survey, a survey of young people’s experience as victims and offenders (Budd et al., 2005a; Budd et al., 2005b; Hayward, Sharp, 2004; Sharp et al., 2006; Wood, 2005). There has also been a – methodologically more problematic – Arrestees Survey, which includes drug testing and self-report interviewing of offenders in police stations. There is also the Citizenship Survey, which has some relevance to the criminal justice system. Further ways are being examined for covering the crime experience of teenagers aged under 16 and that of institutional populations such as students.

5.

Methodological issues

Whether national sample surveys have a long-term future is debatable (see Cantor, Lynch, 2007, for a good discussion). On the one hand, one can envisage a continuing demand for survey-based information. On the other, the technical challenges are likely to get harder. The main problem is one of falling response rates. In the 1980s it was reasonable to expect at least an 80% response rate for a national survey. Now it can be a struggle to achieve much over 70%. The British Crime Survey has actually managed to retain relatively high response rates – stable at around 75% for the last decade. However, an important change in sample design – the move to a lower proportion of respondents from inner cities in 20007 – actually masks a fall in response rates. Looking ahead, it is likely to become increasingly hard 7.

When the BCS extended its sample to 40,000 in order to provide estimates at police force level, the strategy of oversampling inner cities was abandoned. Obviously a new weighting system was used to restore representativeness.

74

Victimisation and Insecurity.book Page 75 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

to secure public co-operation for a long, complex interview process. Low response rates are likely to be highest in those high-crime areas where accuracy is most needed. The main factor depressing response rates must be the increasing competition from both commercial and state bodies for people’s cooperation in surveys and other forms of consultation. Securing respondent compliance is also particularly hard in the increasing number of gated communities and blocks of flats which are protected by screening systems (human or technological concierges). There are solutions to falling response rates. One is to invest more heavily in call-backs – but that carries a price tag. Equally, paying incentives to respondents is an option – but a costly one, when samples are as large as those of the BCS8. Another vulnerability of the BCS is the fact that it excludes people under 16. It has become increasingly clear that teenagers are a very heavily victimised group, especially for crimes such as street robbery. In the early 1990s a booster sample of young teenagers was introduced into the BCS (Aye Maung, 1995), but this was not particularly successful within the confines of a survey designed to be about adult experience. However, the importance of including estimates for teenagers under 16 means that plans are being made to revisit this. Crime surveys clearly need to adapt themselves to capture new crimes. The BCS traditionally has avoided questions about fraud – largely because of definitional difficulties associated with the threshold between incompetence and dishonesty9. Whether it can continue to do so is questionable. Certainly the growth of e-crimes and identity theft means that there has to be coverage of these crimes, and plans for this are in hand. Questionnaires were originally administered through pen-and-paper personal interviews (PAPI); since the 1994 sweep, the survey has involved computer assisted personal interviewing (CAPI).10 This has undoubtedly improved the survey’s technical quality, reducing interviewer error in administering a highly complex questionnaire. A by-product of CAPI was the potential to allow respondents to use the computer themselves to answer questions of a particular sensitive nature – a technique is known as Computer Assisted Self Interviewing (CASI.) CASI has been used in all 8.

An experiment in the use of low-value incentives – a book of postage stamps – found that there was an initial positive effect which wore off. This suggests a ‘survey company’ effect of some sort. A higher incentive would probably have more durable effects. 9. However, recently it has developed questions on identity fraud. 10. In CAPI, interviewers enter responses into a laptop. The questionnaire is a computer program which specifies the questions, the range and structure of permissible answers. Routing instructions determine which questions are asked, and in what order.

75

Victimisation and Insecurity.book Page 76 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

sweeps since 1994 for respondents aged 16-59, covering self-reported drug-taking, sexual victimisation, domestic violence, and being offered and buying goods known to be stolen.

II.

The use of victimisation and insecurity surveys

Looking back on twenty five years of BCS reports, and those from other crime surveys, it is hard now to see how the police, policy and research communities managed without crime surveys for such a long time. Criminologists now take for granted that for a wide range of crimes we have a pretty accurate grasp of the number of offences recorded by the police, the number that go unreported, and those that are reported but unrecorded. We have a good understanding of the factors that affect reporting and recording rates, and a reasonable knowledge of how these have changed over time. This knowledge, which is now taken for granted as part of our understanding of crime, was simply a matter of speculation before the BCS.

1.

Policy uses

The BCS – and indeed local crime surveys – have first are foremost been tools designed for political and administrative use. Despite all the political risks in charting the ‘dark figure’ the BCS rapidly established itself as a valuable source of information for the Home Office and the Home Secretary. Its value has always lain partly in the fact that the survey was controlled by the Home Office, and thus the Home Office had exclusive use of an authoritative source of information about a wide range of topics related to criminal justice. Its central value, however, lay – and continues to lie – in its use as a social indicator about crime trends. The survey has managed to provide reliable information about crime trends, especially over the last ten years, when successive changes have been made to the ways in which the police record crime statistics. These changes have had the effect of making the police statistics more inclusive and more complete, and the BCS has offered a corrective index of crime. Unfortunately, and ironically, the way in which the media and politicians have exploited the divergent trends have probably created a deep public mistrust in crime statistics of any sort. An important development since 1997 has been the growing use of the BCS in the Government’s system of performance management targets, associated with New Labour’s enthusiasm for New Public Management 76

Victimisation and Insecurity.book Page 77 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

methods. The Home Office is subject to seven Public Service Agreement targets, with a range of PSA measures and performance indicators against which their performance is measured. Table 1 summarises these. The italicised measures in the table are drawn from the BCS. It can be seen the survey has become of central importance in performance management, in relation both to crime targets and to a range of attitudinal measures relating to ‘fear’ and confidence in justice. This has had a significant impact on the way that the BCS has developed. First, the fact that a wide range of variables are used to measure performance over time means that the flexibility of the survey has been considerably limited, and policy customers have become increasingly involved in the analysis of ‘their’ questions. Equally important, the capacity of the BCS team to carry out innovative exploratory analysis – which was a key strength of the early sweeps – has been sharply curtailed. A survey which was once a nimble and flexible research machine has now become a performance management juggernaut. The current priority for the Home Office researchers working on the BCS is to analyse the data to yield headline estimates of crime trends and of other measures used in performance management. The time available for more thoughtful, research-driven analysis is getting increasingly limited. Moreover, demands on the future development of the survey now centre on how it might deliver finer-grained local performance management data. There is less talk of expanding the BCS along more theoretical lines, for instance to fill ‘knowledge gaps’ about ordinary victimisation, or help better explain changing patterns of risks. It is easy to be overly-pessimistic about the approach to the management of public services in England and Wales. What Table 1 illustrates, amongst other things, is that there is now an astonishing range of information available to politicians and criminal justice managers about the performance of the criminal justice system. Whilst there are – inevitably, some would say – perverse effects associated with quantitative target setting, bemoaning these effects is a luxury not available to most member states – because comparable information is simply not available for those jurisdictions.

77

Victimisation and Insecurity.book Page 78 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

11

PSAs PSA MEASURES 2005/06–2007/08 2005/06– 2007/ 08 a

BCS: overall crime

STATUTORY PERFORMANCE INDICATORS 2006/07

KEY PRIORITIES 2006 –2009

4a

Reduce overall crime by 15% by 2007-08 and more in high crime areas

4b

1. Reduce crime by 15% and further in high crime areas

b

Recorded crime BCS comparator1

5b 5f

violent crime acquisitive crime

5e

life-threatening / gun crime value of cash and confiscation orders

8c

a 2. Reassure the public, reduce the fear of crime and b anti-social behaviour, and build confidence c in the criminal justice system without compromising d fairness

e

BCS risk of personal crime (proxy) BCS risk of household crime (proxy)

BCS: worry about becoming a victim

10a BCS fear of crime

BCS: feeling that ASB is a big problem

10b

BCS: thinking local police do a good job

2a

BCS perceptions of antisocial behaviour BCS: thinking local police do a good job

BCS: confidence in CJS effectiveness HOCS: CJS 3b agencies treat people 3c equally including 3d disproportion ality measures

Tackle serious / organised crime through improved intelligence and information sharing between partners

victim satisfaction by ethnicity searches leading to arrest by ethnicity detection for violent crime by ethnicity

Provide every area in England and Wales with dedicated, visible, accessible and responsive neighbourhood policing teams; and reduce public perception of antisocial behaviour

11. The BCS comparator is the count of recorded crimes that can be compared directly with a sub-set of BCS crimes.

78

Victimisation and Insecurity.book Page 79 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

PSAs PSA MEASURES 2005/06–2007/08 2005/06– 2007/ 08 f

BCS: victim and witness satisfaction

STATUTORY PERFORMANCE INDICATORS 2006/07

KEY PRIORITIES 2006 –2009

1a-1e satisfaction of victims (for burglary, violent crime, vehicle crime and RTCs) 3a satisfaction of victims (for racist incidents) 11a police officer time spent on frontline duties 8a domestic violence offences leading to arrest

3. Improve the A Number of delivery of justice offences brought by increasing to justice number of crimes where offender brought to justice to 1.25 million

6b

4. Reduce the harm caused by illegal drugs…

10c Perceptions of local drug use/drug dealing (BCS)

a

Drug Harm Index (DHI)

…including b increasing the number of drug misusing offenders entering treatment through the CJS

Drug misusing offenders entering treatment

5. Reduce unfounded asylum claims as part of a wider strategy to tackle abuse of the immigration laws and promote controlled legal migration

Unfounded asylum claims

a

7a

percentage of offences brought to justice percentage of sanction detections

Bring more offences to justice in line with the Government’s PSA

79

Victimisation and Insecurity.book Page 80 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

PSAs PSA MEASURES 2005/06–2007/08 2005/06– 2007/ 08

STATUTORY PERFORMANCE INDICATORS 2006/07

6. Increase volun- a tary and community engagement, especially amongst those at risk of social exclusion

Voluntary activity by those at risk of exclusion

b

Voluntary and community sector contribution to delivering public services

a

Perceptions of 3e racial discriminatio n

b

Perceptions of component of 2a (enhanced community BCS measure of confidence) cohesion

7. Reduce race inequalities and build community cohesion

KEY PRIORITIES 2006 –2009

police recruits from minority ethnic groups

component of 2a (enhanced BCS measure of confidence) SPIs not linked directly to SR04 PSA measures

1. 2.

3g

percentage of female officers 9a number of people killed/seriously injured in RTCs 12a delivery of efficiency targets 13a police officer sickness 13b police staff sickness

Reproduced from ‘Guidance on Statutory Performance Indicators for Policing 2006/07 ANNEXES’ BCS-related measures are italicised.

Table 1 : PSAs, SPIs and NPP priorties

What contribution has the BCS made to the development of crime policy? There are some examples of approaches to crime prevention that are very closely associated with the BCS, notably the targeting of repeat victims. More generally, its contribution has probably been greatest in the field of situational crime prevention, one of the prerequisites of which is detailed, 80

Victimisation and Insecurity.book Page 81 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

crime specific analysis of the circumstances and characteristics of offending. However, the last decade has seen growing use of the attitudinal data collected by the survey, and I personally welcome the focus now given to measures of confidence in justice. To date, policy has drawn less on a different strand of BCS analysis that I associate with Rob Sampson initially and more recently with Tim Hope and colleagues on the role of communities in crime control. Whatever the reasons for the steep falls in crime shown by the BCS, one consequence has been the progressive concentration of crime problems on those most disadvantaged communities with fewest resources to take action.

2.

Academic uses

The BCS has supported some highly creative academic analysis, especially in the 1980s and 1990s, the most significant of which have been: The ‘discovery’ of repeat victimisation (Pease, 1993, 1998) ‘Fear’ (Maxfield, 1984; Hough, 1995, Ditton, Farrall, 2007) Attitudes to sentencing and confidence in justice (Hough, Roberts, 2007) Communities and crime/ the social distribution of crime (Sampson, Groves, 1989, Hope, 2007) Policing (Skogan, 1994, 2007) It must be said, however, that academic exploitation of the survey has been small, relative the opportunities for secondary analysis. There is some promising work in progressing, exploiting the analytic possibilities afforded by fourteen years’ of data, covering a twenty-five year period (eg Jackson, National Centre for Social Research), but the case is overwhelming, in my view, for more pump-priming investment to draw in academics to secondary analysis.

3.

The Future

There are several challenges facing the BCS. The development that has enabled it to really ‘earn its keep’ has been the adoption of various BCS measures as national and local performance indicators. This represents both a strength and a risk. The more that the survey becomes part of central government’s apparatus for performance management, the less flexibility there will be for other work, both in terms of space in the questionnaire, and analytic capacity. A further challenge is to be found in the survey’s public credibility. Early BCS reports could be read as critical commentary on police crime statistics.

81

Victimisation and Insecurity.book Page 82 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Amongst informed commentators and the academic community, the reports were generally very well-regarded, and there was no challenge to their technical excellence.12 However, since 1995, when the BCS began to show ‘good news’ for the Government in terms of falling crime, the BCS has increasingly been seen as yet another example of Government ‘spin’. The (groundless) impression that results have been massaged was compounded by the rise in police figures, due first to ‘counting rules’ changes in 1998/ 99, and then the introduction of the National Crime Recording Standard. Perhaps the biggest threat to national crime surveys is the emergence of the ‘localism’ agenda. The government and both opposition parties are flirting with forms of localism in which, to varying degrees, power will be delegated, or returned, to local politicians. This is an attractive prospect, given that the last two decade of increasing centralisation have clearly demonstrated that central government cannot run public services through strategies of centralised ‘command and control’. The BCS, of course, is increasing becoming an emblem of centralised control, as it becomes embedded in the government’s performance management system. If central government lays claim to crime control, then – in the 21st century – it probably needs a large national survey to support crime management. If responsibility for crime control were ever delegated effectively to county or municipality level, the case for a national survey would clearly be weakened. It is of some significance that both of the two recent reviews of crime statistics have stressed the importance of improving the quality and accessibility of local crime statistics recorded by the police.

References AYE MAUNG N. 1995, Young people, victimisation and the police: British Crime Survey findings on experiences and attitudes of 12 to 15 year olds, Home Office Research Study No. 140, London, HMSO. BUDD T., SHARP C., MAYHEW P. 2005, Offending in England and Wales: First results from the 2003 Crime and Justice Survey. Home Office Research Study No. 275. BUDD T., SHARP C., WEIR G., WILSON D., OWEN N. 2005, Young people and crime: findings from the 2004 Offending, Crime and Justice Survey, Home Office Statistical Bulletin, 20/05 12. This is not to suggest that early publications were without their critics. They were criticised by some academics for their treatment of fear of crime, for example, and their risks analyses, which were seen to disregard the disproportionate burden of crime on inner city communities.

82

Victimisation and Insecurity.book Page 83 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

CANTOR D., LYNCH J. 2007, Addressing the challenge of cost and error in victimisation surveys, in HOUGH, M. AND MAXFIELD, M. (Eds.) Surveying Crime in the 21st Century. Cullompton, Willan Publishing, 281-302. DITTON J., BANNISTER J., GILCHRIST E., FARRALL S. 1999, Afraid or Angry? Recalibrating the ‘fear’ of crime. International Review of Victimology, 6, 83-99. DITTON J., CHADEE D., KHAN F. 2003, The stability of global and specific measures of the fear of crime: Results from a two wave Trinidadian longitudinal study, International Review of Victimology, 9, 49-70. DITTON J., FARRALL S. 2000a, The Fear of Crime, Aldershot, Ashgate. DITTON J., FARRALL S. 2000b, Refining National Survey Questions for Use at the Local Level, International Journal of Police Science and Management, 3, 9-19. DITTON J., FARRALL S., 2007, The British Crime Survey and the fear of crime in HOUGH M., MAXFIELD M., (Eds.), Surveying crime in the 21st century, Collompton, Willan, 223-241. DITTON J., FARRALL S., BANNISTER J., GILCHRIST E. 1998, Measuring Fear of Crime, Criminal Justice Matters, 31, 10-12. DITTON J., FARRALL S. BANNISTER J., GILCHRIST E., PEASE K. 1999, Reactions to Victimisation: why has anger been ignored? Crime Prevention and Community Safety. 11, 37-54. DITTON J., INNES M. 2005, The role of perceptual intervention in the management of crime fear, in TILLEY, N. (Ed.) The Handbook of Community Safety and Crime Prevention, Cullompton, Willan. DITTON J., KHAN F., CHADEE D. 2005, Fear of crime quantitative measurement instability revisited and qualitative consistency added: Results from a three wave Trinidadian longitudinal study International Review of Victimology, 12, 3, 247-271. FARRALL S., BANNISTER J., DITTON J., GILCHRIST E. 1997, Questioning the Measurement of the Fear of Crime, British Journal of Criminology, 37, 657-678. FARRALL S., DITTON J. 1999, Improving the Measurement of Attitudinal Responses: an example from a crime survey, International Journal of Social Science Methodology, 2, 55-68. FARRALL S., GADD D., 2004, The frequency of the fear of crime, British Journal of Criminology, 44, 1, 127-132. FARRINGTON D., DOWDS E.A. 1985, Disentangling Criminal Behaviour and Police Reaction, in FARRINGTON D.P., GUNN J. (Eds), Reaction to Crime: The Public, The Police, Courts and Prisons, Chichester, John Wiley, 4172. FITZGERALD M, HOUGH M., JOSEPH I., QURESHI T. 2002, Policing for London, Cullompton, Willan Publishing. 83

Victimisation and Insecurity.book Page 84 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

HAYWARD R., SHARP C. 2004, Young people and anti-social behaviour: Findings from the 2003 Crime and Justice Survey Home Office Findings 245. HOPE S. 2005, Scottish Crime and Victimisation Survey Survey: calibration exercise – a comparison of survey methodologies. Research report for the Scottish Executive. Edinburgh, Scottish Executive http:// www.scotland.gov.uk/Resource/Doc/47121/0020932.pdf HOPE T. 2007, The Distribution of Household Property Crime Victimisation, in HOUGH M., MAXFIELD, M. (Eds.) Surveying Crime in the 21st Century Cullompton, Willan Publishing, 99-124. HOPE T., TRICKETT A., 2004, La distribution de la victimation dans la population, Déviance & Société, 28, 3, 385-404. HOUGH M., 1995, Anxiety about Crime: Findings from the 1994 British Crime Survey. Home Office Research Study No. 147. London, Home Office HOUGH M., 2004, Worry about crime: mental events or mental states?, International Journal of Social Research Methodology, 7/2, 173-76. HOUGH M, MAXFIELD M., MORRIS M., SIMMONS J. 2007, The British Crime Survey after 25 years, in HOUGH, M., MAXFIELD, M. (Eds.) Surveying Crime in the 21st Century. Cullompton, Willan Publishing, 7-31. HOUGH M., MAYHEW P. 1983, The British Crime Survey: first report, Home Office Research Study No. 76 London, HMSO HOUGH M., MAYHEW P. 1985, Taking Account of Crime: findings from the second British Crime Survey, Home Office Research Study No. 85, London, HMSO. HOUGH M, MAYHEW P., 2004, L’évolution de la criminalité à travers deux décennies du British Crime Survey, Déviance & Société, 28, 3, 267-284. HOUGH M., ROBERTS J. V. 2007, Public Opinion and Criminal Justice: the British Crime Survey and Beyond, in HOUGH M., MAXFIELD M. (Eds.) Surveying Crime in the 21st Century. Cullompton, Willan Publishing, 199-221. INNES M. 2004, Signal crimes and signal disorders: notes on deviance as communicative action, British Journal of Sociology, 55/3, 335-56. INNES M. et al. 2005, Signal Crimes and Reassurance Policing, Volumes 1 & 2. Guildford, Surrey Police. JACKSON J. 2004, Experience and expression: Social and cultural significance in the fear of crime British Journal of Criminology, 44, 6, 946-966. JACKSON J. 2005. Validating new measures of the fear of crime, International Journal of Social Research Methodology, 8, 4, 297-315. JACKSON J. 2006, Introducing fear of crime to risk research, Risk Analysis, 26, 1, 253-264.

84

Victimisation and Insecurity.book Page 85 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Britain

JACKSON J., ALLUM N., GASKELL G. 2006, Bridging levels of analysis in risk perception research: The case of the fear of crime Forum, Qualitative Social Research, 7, 1, Art 20. JONES T., MACLEAN B., YOUNG J. 1986, The Islington Crime Survey : Crime, Victimisation and Policing in Inner City London, Aldershot, Gower. KERSHAW C., BUDD T., KINSHOTT G., MATTINSON J., MAYHEW P., MYHILL A., 2000, The 2000 British Crime Survey. England and Wales, London, Home Office, HO Statistical Bulletin. KERSHAW C., CHIVITE-MATTHEWS N., THOMAS C., AUST R., The 2001 British Crime Survey. First Results. England and Wales, London, Home Office, HO Statistical Bulletin. KINSEY R. 1984, Merseyside Crime Survey: First Report, Liverpool, Merseyside Metropolitan Council. MAXFIELD M. 1984, Fear of Crime in England & Wales. London, HMSO, Home Office Research Study 78. MAYHEW P., ELLIOTT D., DOWDS L., 1989, 1988 British Crime Survey, London, HMSO, Home Office Research Study 111. MAYHEW P., AYE MAUNG N., MIRRLEES-BLACK C., 1993, 1992 British Crime Survey, London, HMSO, Home Office Research Study 132. MIRRLEES-BLACK C., ROSS A. 1995, Crime against retail and manufacturing premises: Findings from the 1994 commercial victimisation survey Home Office Research Study, 146. London, Her Majesty’s Stationery Office. PEASE K. 1993, Individual and community influences on victimisation and their implications for crime prevention In FARRINGTON D. P., SAMPSON R. J., WIKSTROM P-O. (Eds.), Integrating Individual and Ecological Aspects of Crime Stockholm, National Council for Crime Prevention, 323-338. PEASE, K. 1998, Repeat Victimisation: Taking Stock Crime Detection and Prevention Series Paper 90, London, Home Office. SAMPSON R.J., GROVES B. W. 1989, Community structure and crime: testing social disorganisation theory, American Journal of Sociology, 94, 774802. SHARP C, ALDRIDGE J., MEDINA J., May 2006, Delinquent youth groups and offending behaviour: findings from the 2004 Offending, Crime and Justice Survey. Home Office Online Report 14/06 SHURY J., SPEED M., VIVIAN D., KUECHEL A., NICHOLAS S. 2005, Crime against retail and manufacturing premises: Findings from the 2002 commercial victimisation survey. Online Report 37/05. London, Research, Development and Statistics Directorate, Home Office. (Http:/ /www.homeoffice.gov.uk/rds/onlinepubs1.html). SIMMONS J. AND COLLEAGUES, 2002, Crime in England and Wales, 2001/ 2002, London, Home Office, HO Statistical Bulletin. 85

Victimisation and Insecurity.book Page 86 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

SIMMONS J., DODD T., 2003, Crime in England and Wales, 2002/2003, London, Home Office, HO Statistical Bulletin. SKOGAN W. G. 1994, Contacts Between Police and the Public: A British Crime Survey Report Home Office Research Series, London, Her Majesty’s Stationary Office. SKOGAN W. G. 2007, Survey Assessments of Police Performance, in HOUGH, M., MAXFIELD, M. (Eds.) Surveying Crime in the 21st Century, Cullompton, Willan Publishing, 165-181. SMITH D. J. 1983, Police and People in London I. A Survey of Londoners, London, Policy Studies Institute. SPARKS R, GENN H, DODD D. 1977, Surveying Victims,London, Wiley. TUFFIN R., MORRIS J., POOLE A. 2006, An evaluation of the impact of the National Reassurance Policing Programme, Home Office Research Study 296. London, Home Office. WALKER A., KERSHAW C., NICHOLAS S. 2006, Crime in England and Wales 2005/06, Home Office Statistical Bulletin 12/06. London, Home Office http://www.homeoffice.gov.uk/rds/pdfs06/hosb1206.pdf WOOD M. 2005, The victimisation of young people: findings from the Crime and Justice Survey, 2003, London, Home Office Findings 246.

86

Victimisation and Insecurity.book Page 87 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France Philippe Robert

In France, no great effort was made in the early stages to improve upon or complete the official data on crime. Until very late in the sixties, the state of research on crime remained quite anemic and the rare attempts that were made focused mainly on juvenile delinquency. Moreover, the monopoly of the official data enjoyed the legitimacy of a long and glorious statistical tradition – which began with the creation of the Compte général de l’administration de la justice criminelle in 1827 and with the entire current of moral statistics initiated by Quetelet and Guerry…. – to the extent that the resurgence of an empirical sociology of crime was set in motion with Davidovitch’s1 work, by applying to court data the techniques of contemporary statistical analysis. The situation however was to change. At the end of the 70s, the Ministry of Justice renounced the Compte général that it had been producing for a century and a half and withdrew from the area of criminal statistics, leaving the field clear for the development of the more recent police statistics2. However, from the early sixties, unbridled consumerism led to an escalation in crime involving mainly property crime. It is only in the mid70s that the concern with safety3 began to emerge, bringing in its wake a revival of thinking on the sources of information on crime. To limit oneself to the figures yielded by police institutions seemed increasingly less satisfactory in the sense that it was the activity and the politics of policing which were being called into question: were they still capable of ensuring the safety of people and their property? Nevertheless, the initial surveys were tentative and isolated. Although there were many studies on the cost of crime from the early 70s4 and their findings even met with a good response, they had more to say on tax fraud and on homicide than on crime involving thefts, burglaries or assaults, which are at the very heart of insecurity. The self-reported delinquency surveys were more or less unknown. A few attempts to survey victimisation 1. 2. 3. 4.

See in particular Davidovitch, Boudon (1964). Although started in 1950, they were only published from 1972 onwards. For the reasons of this time-lag see Robert (1999), passim and in particular 59-111. See for example Robert, Godefroy (1978).

87

Victimisation and Insecurity.book Page 88 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

were on the contrary made in the latter half of the ’70s: local monographs5 and studies on specific target populations of victims, such as women and migrants6. In particular, the Comité d’étude sur la violence, la criminalité et la délinquance had commissioned a polling institute to conduct a survey that contained some questions – very perfunctory – on victimisation, but focused more – and with better technical quality – on the feeling of insecurity and on the neighbourhood issues7. Manifestly, the standard practices were at that time more or less unknown in France. Paradoxically however, it happens that the research on victimisation and insecurity is now quite well-stocked, especially if we take into account its rather tardy start. We will first take stock of what has been accomplished, before examining the use made of these surveys.

I.

Taking stock of victimisation and insecurity surveys available in France

1.

National victimisation surveys

The situation effectively changed in the ’80s. A team from the Centre de recherches sociologiques sur le droit et les institutions pénales (CESDIP) systematically besieged the field of victimisation surveys. Three features distinguished the enterprise: an initial desire to fully master the research done in other countries8; the drawing up of a systematic programme starting with a qualitative research and leading to a national survey followed by local investigations; and lastly a shift in focus: the victimisation survey had been conceived as a supplement to or substitute for criminal statistics – thus as an instrument of measurement. This approach was conferred lower priority by the CESDIP and it was the understanding of victims9 as social actors through the construction and analysis of typologies that was given pride of place. 5. 6. 7.

8. 9.

For example, Morange (1979). For instance, Dootjes-Dussuyer et al. (1980); Soubiran (1981). Peyrefitte (1977), III, 390. The survey was conducted on quota samples of 1500 individuals (in reality, only 1292 interviews could ultimately be used). The Committee had not done an in-depth analysis of the survey it had ordered. The data was later used for secondary analysis (Lagrange [1985]), which revealed that there were very few questions on victimisation as compared with the questions on the feeling of insecurity and on the neighbourhood. In particular Zauberman (1982, 1985b). The emergence of this trend can be seen for example in Zauberman (1985b).

88

Victimisation and Insecurity.book Page 89 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

The qualitative study10 was not simply a preliminary to the drawing up of a questionnaire for a national survey. Composed of about forty retrospective interviews with a diversified population of victims, it enabled the devising of hypothetical typologies, which were to provide guidelines for later analyses. The 1986 CESDIP survey constituted the first nationwide victimisation survey in France11. A series of screening questions were initially incorporated into an omnibus survey12. Among the victims thus identified, sub-populations were formed – with specific sampling for each type of victimisation, based on frequency – and 1,138 interviews were conducted. The questionnaire borrowed from similar surveys done in other countries, but also made full use of the results of the qualitative research; it also drew greatly upon a series of surveys undertaken by the centre during the previous decade on social attitudes towards Justice, crime and the criminal13 – the questionnaire was formulated so as to analyse the various ways people experienced victimisation according to their attitudes and social positions. Despite this kick off, a decade went by before fresh, systematic national surveys were conducted14: expertise and knowledge did not make up for the lack of funds, and hence regular surveys could not be carried out. The doors were finally unlocked when the National Institute of Statistics and Economic Surveys (INSEE) joined a concerted effort of the major European statistical institutes. In 1996, it launched annual surveys on the living conditions of households that contained a module on victimisation. With the INSEE, interviews are always conducted face-to-face, this time on representative samples of households15, but the thrust reversed to the classical approach: the main focus was now on counting victims – and secondarily victimisation incidents. Questions on opinions and attitudes 10. Robert, Zauberman (1985). 11. Zauberman et al. (1990) ; Zauberman, Robert (1990) ; Zauberman (1991) ; but especially Zauberman, Robert (1995). It is the Ministry of Justice that financed the first national survey with funds from the national research budget. 12. 11,156 interviewees were chosen from quota samples of the target population. 13. Robert, Faugeron (1978); very close to the French tradition of the social psychology of attitudes (Jodelet [1989]). 14. Polling institutes have generously dotted their surveys with questions on victimisation, but they are not very useable as they are dispersed and neglectful of the state of the art thinking about insecurity and victimisation. 15. Approximately 6,000 with, in the first year, one interview per household, then as many as three. The 2004 and 2005 surveys interviewed 6,351 and 13,872 households respectively, but the 2006 survey included only 6,000. Semi-rotating samples were used (half the samples of one survey had already been rotated during the previous sweep).

89

Victimisation and Insecurity.book Page 90 Thursday, December 4, 2008 12:33 PM

.

Victimisation and Insecurity in Europe

were reduced to the feeling of insecurity and to opinions on the local environment, housing, and transportation. Despite this modified angle, it was nevertheless possible to compare the findings to those of the CESDIP national survey after a ten-year gap16, and moreover to contrast them with police data. On the other hand, extracting typologies of victims, although possible, gave less significant and less stable results, as items on opinions and attitudes were scant17. Afterwards, the serialisation of national surveys and their comparison with police data was systematically taken up as and when the INSEE surveys were given access to18. Why was it that ten years had to be waited after the first survey before regular national studies became available? Such research is expensive – because of the size of the samples required in order to grasp the rarest types of victimisation – and the natural sponsors – the police and justice departments, the former in particular – have shown little enthusiasm for the development of this new source of data. Information on crime until then referred to official statistics alone. The idea of losing this monopoly aroused an unspoken apprehension. In 1999, the Ministry of the Interior, generator of police statistics, pretended an interest in victimisation research by asking INSEE to carry out a more detailed investigation in order to make a more in-depth comparison with these statistics. However, the dissemination of the initial results revealed an enormous quantitative discrepancy between the survey findings and the official data, and aroused strong opposition within the police force. The Ministry then decided to stop analysing the survey data. In the last few years however, its efforts have been more sustained, but its initiative really served to disrupt the continuity of the only nationwide survey programme in existence for the past ten years. A broader protocol was substituted for the small modules on victimisation included in the periodic surveys on the living conditions of households (EPCV) by INSEE, but which ostensibly does not take into account earlier French work and does not seek to maintain the continuity of the data series. We can however still consider as national surveys, French participations in the international surveys initiated by Jan Van Dijk, Pat Mayhew and Martin 16. The rapid growth of violent victimisation, in the span of a decade, was highlighted, but also its relatively low levels in absolute figures, especially in comparison with the impressive, but relatively stagnant figures of property victimisation. 17. These comparisons, which have been published, also deal with typological analysis (Robert Zauberman, Pottier, Lagrange, [1999, 2001]). 18. See for the period 1985-2001, Lagrange, Pottier, Zauberman, Robert (2004); a fresh serialisation is being carried out currently, which integrates the surveys conducted from 2002 to 2006.

90

Victimisation and Insecurity.book Page 91 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

Killias: Ministry of Justice funded the French involvement in the first sweep19. Later, the Ministry of the Interior financed further French participations, but the small size of the samples renders their use shaky, so that they have been only occasionally analysed in the French context20. In 2005, a sweep conducted by Gallup was financed by the European Commission (Van Dijk et al., 2007). We can also find questions on victimisation in the 1996 Eurobarometer (44.3) and in the European Social Surveys of 2002, 2003, 2004 and 2006, with funding for the French participation by the Ministry of Research and, scientific supervision by the Centre d’études de la vie politique française (CEVIPOF).

2.

Local victimisation surveys

Let us now move from the national to the local level. CESDIP’s programme in the ’80s included a third stage, a local one. It took the form of two telephone surveys realised in 1989, one in Épinay, an outlying district of Paris, the other in the Toulouse urban community21. These two urban surveys used a simplified version of the questionnaire constructed during the 1986 national survey. Financed by the Délégation interministérielle à la ville22, their aim was to help in developing urban safety monitoring agencies. The indecisiveness shown by the DIV when it came to funding stopped progress from being made in the first half of the ’90s. Here too the deadlock was broken from the outside. In 1999, an urban research organ devised by the CNRS enabled the CESDIP team to undertake a pilot survey23 in a small regional metropolis, Amiens. The usual results – prevalence, incidence, and victim typologies – were combined with observations on the characteristics of the different urban zones. For the residents of residential neighbourhoods, victimisation was at worst the correlate of some form of lifestyle-associated careless behaviour, and easily avoidable. For those living in isolated, poor suburbs, on the contrary, it was an aspect of local life from which they could not escape because they did not have the means to move to a more peaceful but more expensive neighbourhood.

19. 1,500 individuals were interviewed on the telephone; see Zauberman (1993). 20. Approximately 1,000 individuals interviewed on the telephone; see Zauberman (1993); Grémy (1997). 21. 1,780 and 1,576 interviews respectively; Robert Zauberman (1991). 22. An governmental agency devoted to urban policies 23. Telephone survey (Lagrange, Peretti, Pottier, Robert Zauberman [2000]; Zauberman, Robert, Pottier [2000]) of a sample of 1,156 interviewees conducted on the telephone.

91

Victimisation and Insecurity.book Page 92 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

In 2005, this study served as the basis for CESDIP surveys for five members of the French Forum for Urban Security (the Lyons urban community, and the towns of Saint-Denis, Gonesse, Aubervilliers and Aulnay-sous-Bois in the Paris region)24. However, these local efforts never resulted in regular survey programmes as in Barcelona. The increase in the number of these undertakings seems, moreover, to come up against one difficulty, namely to resolve the funding problems endemic to such surveys, especially as local governments, who primarily face the most acute safety problems – and whose potential need for such surveys is greatest – are also frequently the poorest. The Amiens survey also served as the basis for a survey of the Île-de-France region organized by CESDIP for the Institut d’aménagement et d’urbanisme.de la région Île-de-France (IAURIF) in 200125, and which the latter replicated without modification in 200326, 200527 and 2007. Thus it is at the regional level that surveys are conducted with most regularity with, moreover, a well-structured questionnaire and a goodsized sample. However, this was just one region – although the most important one – and only for-sweeps. The continuity of this programme remains uncertain, firstly because the financing of each new sweep depends on a political deliberation of the Regional Assembly, and secondly because of the risk of the protocol rapidly becoming outdated, insofar as there no longer exists any organised collaboration between the managers of the procedure and the state of the art sociology of crime. Modules on victimisation, more or less developed, also figure, and have done so for a long time, in a whole series of local monographs although devoted mainly to some other subject. This was the case with the feeling of insecurity: Lagrange, later joined by Roché, operated in the Rhône-Alpes region28; Duprez29 concentrated on the Northern region and in collaboration with INSEE, worked on a sampling of urban and suburban problem areas. More recently Roché (2000) has included modules on victimisation in a survey on incivilities in two urban sites. 24. Depending on the area, telephone survey of 1,000 to 5,000 (Lyon agglomeration) individuals aged 15 years and above (Névanen et al., [2006]; Didier et al., [2006a & b]; Robert et al., [2006], Zauberman et al, [2006].). 25. Institute for urban planning and development of the Paris Île-de-France region telephone survey in January 2001 based on a sample of 10,504 interviewees; Pottier et al., (2002). 26. Heurtel (2004); for a more detailed analysis, see Fouquet et al. (2006). 27. IAURIF (2006). 28. 1292 individuals of 17 years and above interviewed one-to-one; Lagrange, Roche, (1987-89). 29. Duprez (1996); also see Duprez, Leclerc-Olive, Pinet (1996).

92

Victimisation and Insecurity.book Page 93 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

3.

Thematic surveys

In the end, the least unstable programmes are to be found with the thematic surveys. Large scale nationwide surveys – but on specific population samples, like youth and women – include more or less developed sections either on violence30 or on forced sexual intercourse among the youth31 or on violence towards women32. No doubt, some of them are episodic (violence towards women) or occur in irregular sweeps (school violence), others have managed to incorporate a module on victimisation (more or less comprehensive, generally quite limited) in the larger monitoring schemes carried out in areas where the need for information is very great (health, youth) and where no properly organised and powerful enough institutional statistics can successfully resist the development of extra-institutional data sources. On the whole – but thematic surveys often evade this situation – the development of good victimisation surveys is hampered by the tendency of sponsors and polling institutes to keep their distance from the scientific community and take little or no notice of its work in order to improve their own research approach. We find ourselves in the position where those who decide on the surveys and often those who implement them have themselves no experience of fieldwork and do not participate in the secondary analyses, which are developed from the data collected by the victimisation surveys. As for the research centres that undertake the serialisation and comparisons with other data on crime or the secondary analyses of victimisation data, they sometimes find it difficult to access the 30. Regularly publishing a Baromètre santé adulte [adult health barometer] (18-75 years) since the early ’90s, the Comité français d’éducation pour la santé, in 1997-98 conducted a telephone survey of adolescents between 12 and 19 years (random national sample of 4,115 interviews), which includes some questions on violent victimisation (Arènes, Janvrin, Baudier [1998]). In 1999-2000, the two surveys were combined into a single barometer (CFES [2000]) of individuals between 12 to 75 years (telephone survey on a random national sample of 13,685 interviews) and comprising modules on violent victimisation 31. Lagrange, Lhomond, (1997, 145-151). A questionnaire on their sexual behaviour was administered in institutions, generally schools, to 6,182 adolescents aged 15-18. It included questions on forced sexual behaviour. The survey was financed by the Agence national de recherches sur le SIDA (National agency of research on AIDS). 32. Jaspard (2003). Carried out on the telephone on a national sample of 6,970 women between 29 to 59 years of age, the study is on violence experienced after 18 years, sexual violence throughout life and violence in the last 12 months. The Service for women’s rights attached to the Employment and Solidarity Ministry provided the funds.

93

Victimisation and Insecurity.book Page 94 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

survey data, or else manage to do so only after excessive delays. There exists a sort of curious division of labour that tends to keep the world of scholars off limits from participating in public surveys. It is difficult to have a strict separation between the victimisation surveys and those on the feeling of insecurity: the former normally have questions on the concern with and fear of crime and the latter often include questions on victimisation. Analytically however, it is better to address them separately.

4.

Surveys on the feeling of insecurity

When we started to discuss the subject in the late 1970s, insecurity was a controversial topic; no surveys had been conducted until then. Everyone however, spontaneously agreed that the concern with safety was excessive in relation to the risk, even if they were deeply divided as to the conclusions to be drawn. As a result, the initial empirical work undertaken was more inclined to ascertain whether the feeling of insecurity was not the result of an alarmist discourse on the part of the authorities and the media in order to divert attention from more serious problems, such the growing unemployment. But from this perspective, the results were not very conclusive33. So experts on community life took up the task again for another long decade. The increase in the number of monographs eventually threw some light on the role played by the quality of micro-social relationships and networks, and also the role of integration in community life34. During the 90s, the question of its relationship with crime, until then neglected, again surfaced; but the data available, fragmentary and not particularly homogenous, was not able to shed light on it. Once again more theories were advanced than data produced to enable testing of these theories35. The problem was somewhat resolved when a collection of national surveys on the social concerns of the French spanning a quarter of a century, were unearthed. They were not especially designed for the study of insecurity, but by taking some precautions, we were able to use them for this purpose36. It was then decided to undertake a re-evaluation of the merits of a distinction between fear and concern, introduced in the United States in 33. Coing, Meunier (1980). 34. The most important monographs were successfully completed by Lagrange and Roché (1987-89; Roché [1993]; Lagrange, [1985]) and by Duprez and Hedli (1992); Duprez (1996); Duprez, Leclerc-Olive, Pinet (1996). 35. Robert, Pottier (1997a). 36. Robert, Pottier (1997b).

94

Victimisation and Insecurity.book Page 95 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

1971, but largely under-utilised in France until then. Fear – an anxiety of being victimised, oneself or one's intimates – was analysed as a consequence of the perceived risk – hence of the surrounding crime, especially violence – filtered through the prism of self-attributed vulnerability. Concern is a more abstract apprehension of crime as a social problem and responds to a more complex configuration. It is dominated by ambient socio-economic conditions: it is striking that it appeared at a moment when full-employment was fast losing ground, and that it affected the more regions that were the most vulnerable in this respect. Although individuals concerned with safety are not necessarily more directly or more immediately at risk of becoming victims of crime, they feel however more vulnerable to a possible jeopardizing of their social status. This distinction is used in the most recent studies to explain why the feeling of insecurity appears, depending on the perspective from which it is observed, as both correlated to and yet so unrelated to the risk of victimisation. The more recent research37 has used the same database to show how the middleclass, which had earlier resisted it, was drawn into this concern with safety from the mid-90s. In the end, the sub-field of surveys on the feeling of insecurity did not suffer from the same constraints as the victimisation surveys. It merged more easily with ‘the democracy of polls, which characterizes the contemporary French political regime. It did not threaten the monopoly of institutional statistics as it measured a ‘feeling’ and not crime. Here too, however, we note the constant difficulty experienced by those who commission or conduct surveys to take into account the state of knowledge produced by academics and its evolution. Lastly, these surveys – much less costly than victimisation ones – are frequently undertaken by organizations – the press, polling institutes, independent administrative commissions or research centres – independently of official departments that produce criminal statistics. It is moreover a programme financed by the major industries in the energy sector, which has constituted the most coherent and the most durable38 investigation on the feeling of insecurity39.

37. Robert, Pottier (2004, 2006). 38. Agoramétrie (formerly AESOP), is an annual survey conducted from 1977 to 2004; a national stratified quota sample of some 1,000 persons inteviewed on the telephone. However, this programme seems to have been interrupted after nearly three decades. 39. There exists moreover a survey (monthly survey with a national stratified quota sample of some 1,000 individuals aged 18 years and above) – carried out by SOFRES for the newspaper, Le Figaro – conducted regularly over nearly three decades (1974-1987, 1988…..), and very effective in measuring the variations in the concern about safety.

95

Victimisation and Insecurity.book Page 96 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

5.

Technical characteristics of the surveys

International, regional or local surveys, the Health Barometer as well as surveys on the feelings about crime are conducted on the telephone. All the others, and which are also the oldest, are direct face-to-face interviews, including those of INSEE, which is able to mobilise its large network of interviewers. The use of the telephone has increased in recent years, especially because it has reduced the cost of surveys, but it has encountered an increasing number of obstacles as more and more households no longer have a landline connection and only have cell phones40. In both cases, the use of the computer (CAPI or CATI) is becoming more pervasive and makes repeated surveys more reliable, quicker and more manageable. The postal questionnaire method – less costly but also less reliable – is on the other hand not used in French surveys41. As for surveys on school students, they are generally administered in schools and are monitored by the researchers themselves. Though some regional surveys – for example those conducted in Île-deFrance (the Greater Paris region) – are based on large samples, most local and national surveys comprise smaller samples, which results in more widely open confidence intervals, at least in the case of the rarest types of victimisation. This is the chief limitation of victimisation surveys by the INSEE and it is surprising that the Ministry of the Interior, in its recent involvement in this programme, has not given long-term priority to a significant expansion of the samples.. If thematic surveys naturally cover a specific range of victimisations, surveys covering a broader-spectrum are concerned, in more or less greater detail, with property crime (namely thefts, burglaries, vandalism) and violent crime. Earlier attempts to broaden their scope to include consumer or white-collar crime have not prospered. With regard to the feeling of insecurity, the survey questionnaires do not always use the various standard tried-and-tested formulations: hence the analysis of results and especially their comparison is not facilitated. This again appears as the frequently encountered negative effects of the gap between the commissioning agencies and pollsters on the one hand, and the academic world on the other.

40. The latest Health barometer contains a sub-sample of these ‘exclusive cell phones’ with a simplified questionnaire (Beck, Guilbert, Gautier, [2007]). This solution is only possible at the national level as at present it is not possible to localize the residence of the cell phone holders. 41. Obergfell-Fuchs, Kury, Robert, Zauberman, Pottier (2003).

96

Victimisation and Insecurity.book Page 97 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

If the surveys on the feeling of insecurity have a variety of sponsors, including the press, and even research institutes, on the other hand victimisation surveys – whose cost is much more onerous – are always commissioned by the central government ministries or by regional as well as local governments, but the latter seem to have great difficulty in ensuring the funding of the surveys that they desire to take up. Until the recent take over by the Interior Ministry, the EPCV surveys were commissioned by INSEE and the thematic surveys often by the research funding state agencies. The scientific or technical piloting of the surveys corresponds to very split situations. In certain cases – most of the thematic surveys, and also some of the general surveys – it is under direct responsibility of the research institutes, in others it is ensured by the big quantitative survey agencies such as INSEE. Lastly, in many cases it is monitored directly by the commissioning agency’s own subdivisions.

II.

The use of victimisation and insecurity surveys

If the scientific uses are relatively easy to describe, the others correspond to a less clear situation. A distinction should therefore be made between them.

1.

Scientific uses

A very small number of research centres work on this type of data. Some of them undertake surveys in their field of expertise – generally oneoff surveys, although they are sometimes replicated – and analyse them directly, and occasionally even use their data later for secondary analyses42. Others use the data from surveys, in which they have not participated, to undertake only secondary analysis43.

42. For example, Lagrange, Roché (1987-89) (Centre d’études et de recherche sur l’administration et le territoire, CERAT) ; Duprez et al. (1996) (Laboratoire de sociologie du travail, de l’éducation et de l’emploi, LASTREE) ; Jaspard (2003) (Institut de démographie de l’Université de Paris I, IDUP) ; Lagrange et al. (1997) ; Debarbieux (2004) ; Carra, Sicot (1997). 43. Thus for the Observatoire sociologique du changement (OSC), Herpin, Lagrange (2005) ; for the Centre de recherche en économie et statistique (CREST), Fougère, Kramarz, Pouget (2005a, b) ; also see Peretti-Watel (2000).

97

Victimisation and Insecurity.book Page 98 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Lastly, yet others conduct the surveys themselves, and then analyse the results; but they can also use survey data for secondary analysis, whether or not they have carried out the surveys themselves44. This scientific production has developed: standard tools for analysing data in the form of rates (prevalence, incidence, concern, fears, etc.) that are used in all research work; multivariate analysis – for example to generate typologies of victims or of victimisations – that are hardly used except by research centres; secondary analyses of which the most developed example currently is the hypothesis – although not tested fully as yet – of a tripartite division between urban centres not so affected by the feeling of insecurity, although greatly affected by victimisation, residential zones that are facing only limited crime problems and whose inhabitants are not very concerned on this score, and underprivileged zones that are under great pressure of delinquency and also beset by a strong feeling of insecurity.

2.

Non-scientific uses

As far as the central government is concerned, the use of specific surveys by concerned government departments – Health for example, or Education – does not seem to pose a problem at least in terms of communication, as proved by the widely circulated results of the survey on violence against women. We have less information on their operational uses, for example at the level of the monitoring agencies; thus it seems that the Ministry of Education is finding it difficult to give equal weight to survey data on violence in schools alongside data furnished by the school principals. The results of polls and surveys on the feeling of insecurity sometimes figure in the abundant media coverage of insecurity, especially when they are the outcome of polls rather than surveys. But reference is made primarily to findings on concern, which are easy to represent through global data, less to fear that appears fragmented, depending on the risk run and the vulnerability felt by the different groups. As for the results of the general victimisation surveys, especially nationwide ones, for a long time they were not used at this level of government, which was not really in the market for it. They have now begun to appear in the toolbox of the national monitoring agencies 44. Thus for the OSC and CESDIP, Robert et al. (1999, 2001) ; Lagrange et al. (2004) ; for CESDIP, Robert (2006) ; Robert Pottier (1997, 2004, 2006) ; Robert Zauberman (1995, 1991, 2006) ; Robert et al. (2006) ; Névanen et al. (2006) ; Didier et al. (2006a, b) ; Pottier et al . (2002) ; Zauberman, Robert, (1985) ; Zauberman et al. (1990, 2000, 2004, 2006).

98

Victimisation and Insecurity.book Page 99 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

(Observatoire national de la délinquance on crime or Observatoire national des Zones urbaines sensibles on destitute urban areas). However, this use is not very appropriate: after having ignored the existence of such methods of investigation, government departments claim to have invented them ex nihilo by sometimes referring to overseas examples, practically never to national achievements. In other words, they relegate them to the rank of junior partner, placed alongside the institutional data, as if in keeping with a trend45. This difficulty probably arises from a two-fold weakness of these institutions: firstly, these monitoring agencies do not mobilise scientific capabilities, or do so minimally – rather they shy away from them. Secondly, they cannot shake off – in fact if not by law – close administrative supervision, which uses them as foil for official discourse on crime and safety. If we now come to the regional government level, the only region that regularly uses such data – Île-de-France – incorporates them in an ‘observatory’ run by its research institute, IAURIF. One difficulty arises perhaps from the fact that the regional government level does not have direct authority in the matter of security, except for school equipment and public transportation. Lastly, at the local level, a few urban communities and cities, have started, after innumerable difficulties, to commission victimisation and insecurity surveys. We do not have enough hindsight at the present, to know if these local governments will manage without too much trouble to combine survey results with institutional statistics, which they have been so far supplying to the nascent ‘observatories’. On the other hand, for the moment we find very few attested uses of victimisation and insecurity surveys for the evaluation of public safety policies or more specifically of the various programmes that compose them. It is true that that if the term ‘evaluation’ has been very much in vogue since the late 80s, it generally applies to the simple internal audits, realized by the inspectorates of the different government departments, on the basis of institutional data alone46. Here the overriding trait that characterizes the entire issue is the extreme difficulty for a strong administrative culture to really take into consideration extra-institutional data whose production and usage it is not able to properly master.

45. For instance, Observatoire national de la délinquance (2005, 2006, 2007). 46. See for example, Robert (2003).

99

Victimisation and Insecurity.book Page 100 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

References ARÈNES PH., JANVRIN M.-P., BAUDIER F. (Dir.) 1998 Baromètre Santé Jeunes 97-98, Paris, Centre français d’éducation pour la santé (CFES). BECK, F., GUILBERT PH., GAUTIER A. 2007, Baromètre santé 2005, SaintDenis, INPES. CARRA C., SICOT F. 1997, Une autre perspectives sur les violences scolaires : l’expérience de victimation, in CHARLOT B., EMIN JC, (Dir.), La Violence à l’école ; état des savoirs, Paris, Armand Colin, 61-82. CENTRE FRANÇAIS D’ÉDUCATION POUR LA SANTÉ (CFES) 2000, Baromètre Santé 2000, Paris, Centre français d’éducation pour la Santé (CFES). COING H., MEUNIER C. 1980, Insécurité urbaine, une arme pour le pouvoir, Paris, Anthropos. DAVIDOVITCH A., BOUDON R. 1964, Les mécanismes sociaux des abandons de poursuite, L’Année sociologique, 111-244. DEBARBIEUX É. 2004, Les enquêtes de victimation en milieu scolaire ; leçons critiques et innovations méthodologiques, Déviance et Société, 28, 3, 317-333. DIDIER E., NÉVANEN S., ROBERT PH., ZAUBERMAN R. 2006a, Enquêtes locales 2005 sur la victimation et l’insécurité (Gonesse), Guyancourt, CESDIP (www.cesdip.com). DIDIER E., NÉVANEN S., ROBERT PH., ZAUBERMAN R., 2006b Enquêtes locales 2005 sur la victimation et l’insécurité (Aulnay-sous-Bois), Guyancourt, CESDIP, (www.cesdip.com). DOOTJES-DUSSUYER I., AUTHIER D., BERGER P., LHOMOND B. 1980, Les femmes victimes de sévices au cours de leur vie maritale, Lyon, Université Claude-Bernard, Institut Alexandre Lacassagne. DUPREZ D. 1996 La dramaturgie de la relégation ; vie quotidienne et sociabilités dans les cités, Profils, 41, 55-73. DUPREZ D., HEDLI M. 1992, Le mal des banlieues ? Sentiment d’insécurité et crise identitaire, paris, L’Harmattan. DUPREZ D., LECLERC-OLIVE M., PINET M. 1996, Vivre ensemble ; la diversité des quartiers sensibles à l’épreuve de la vie quotidienne, Lille, LASTREE. FOUGÈRE D., KRAMARZ F., POUGET J. 2005a, Délinquance et mobilité résidentielle, Revue économique, 56, 2, 313-336. FOUGÈRE D., KRAMARZ F., POUGET J. 2005b, L’analyse économétrique de la délinquance : une synthèse de résultats récents, Revue française d’économie, 19, 3, 3-55. FOUQUET A., LOTODÉ H., NÉVANEN S., ROBERT PH., ZAUBERMAN R. 2006, Victimation et insécurité en Île-de-France ; deuxième enquête de l’IAURIF, Guyancourt, CESDIP (www.cesdip.com).

100

Victimisation and Insecurity.book Page 101 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

GRÉMY J.P. 1997, Les Français et l’insécurité ; trois sondages réalisés en 1996 sur l’insécurité et ses remèdes, Paris, Institut des hautes études de la sécurité intérieure (IHESI). HERPIN N., LAGRANGE H. 2005, La victimation de proximité, les précautions et la peur. Étude sur la cohésion sociale de voisinage, Revue économique 56, 2, 285-312. HEURTEL H. 2004, Victimation et sentiment d'insécurité en Île-de-France. Résultats de la 2ème enquête – 2003, Paris, IAURIF. IAURIF 2006, Troisième enquête de victimation en Île-de-France : les résultats Note rapide 411, Paris. JASPARD M., BROWN E., CONDON S., FIRDION J.-M., FOUGEYROLLASSCHWEBEL D., HOUEL A., LHOMOND B., MAILLOCHON F., SAURELCUBIZOLLES M.-J., SCHILTZ M.-A. 2003, Les violences envers les femmes en France. Une enquête nationale, Paris, La Documentation française. JODELET D., (Dir.) 1989, Les représentations sociales, Paris, PUF. LAGRANGE H. 1985, Réponses à l’insécurité, analyse secondaire d’une enquête réalisée en 1976 à la demande du Comité d’étude sur la violence présidé par A. Peyrefitte, Paris, Ministère de l’Urbanisme et du logement (MUL). LAGRANGE H. 1995, la civilité à l’épreuve. Crime et sentiment d’insécurité, Paris, PUF. LAGRANGE H., LHOMOND B., COLL. CALVEZ M., LEVINSON S., MAILLOCHON F., MOGOUTOV A., WARSZAWSKI J. 1997, L’entrée dans la sexualité ; le comportement des jeunes dans le contexte du sida, paris, La Découverte. LAGRANGE H., PERETTI P., POTTIER M.-L., ROBERT PH., ZAUBERMAN R. 2000, Les enquêtes sur les risques urbains ; étude de préfiguration, Guyancourt, CESDIP. LAGRANGE H., POTTIER M.-L., ROBERT PH., ZAUBERMAN R. 2004, Enquêtes de victimations et statistiques de police : les difficultés d’une comparaison, Déviance et Société, 28, 3, 285-316. LAGRANGE H., ROCHÉ S. 1987-89, Baby alone in Babylone ; deux perspectives d’analyse du sentiment d’insécurité : systèmes d’attitudes et formes de sociabilité en milieu urbain, Saint Martin d’Hères, CERAT. MORANGE E.R. 1979, La criminalité réelle à Aix-en-Provence, Aix, Université de droit, d’économie et de sciences sociales d’Aix-Marseille. NÉVANEN S., DIDIER E., ROBERT PH., ZAUBERMAN R., 2006, Enquêtes locales 2005 sur la victimation et l’insécurité (Aubervilliers), Guyancourt, CESDIP (www.cesdip.com). OBERGFELL-FUCHS J., KURY H., ROBERT PH., ZAUBERMAN R., POTTIER M.-L., 2003, Opferbefragungen in Deutschland und Frankreich. Unterschiedliche Konzeptionen und Vorgehensweisen, Monatsschrift für Kriminologie und Strafrechtsreform, 86, 1, 59-73.

101

Victimisation and Insecurity.book Page 102 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

OBSERVATOIRE NATIONAL DE LA DÉLINQUANCE, 2005, Premier Rapport annuel, Saint-Denis, Institut national des hautes études de sécurité (ministère de l’Intérieur). OBSERVATOIRE NATIONAL DE LA DÉLINQUANCE 2006, Rapport 2006, SaintDenis, Institut national des hautes études de sécurité (ministère de l’Intérieur),. OBSERVATOIRE NATIONAL DE LA DÉLINQUANCE, 2007, Rapport 2007, SaintDenis, Institut national des hautes études de sécurité (ministère de l’Intérieur) (www.inhes.intérieur.gouv.fr) PERETTI-WATEL P. 2000, L’inscription du sentiment d’insécurité dans le tissu urbain, Les Cahiers de la sécurité intérieure, 39, 201-221. PEYREFITTE A. (Dir.) 1977, Réponses à la violence, annexes au Rapport du Comité d’étude présidé par Alain Peyrefitte, Paris, Documentation française,. POTTIER M.-L., ROBERT P., ZAUBERMAN R. 2002, Victimation et insécurité en Île-de-France. Les résultats de la première enquête (2001) Rapport final, Paris/Guyancourt, IAURIF/CESDIP (www.cesdip.com). ROBERT PH. 1999, Le citoyen, le crime et l’Etat, Genève-Paris, Droz. ROBERT PH. 2003, The evaluation of prevention policies, European Journal of Crime, Criminal Law and Criminal Justice 1, 114-130. ROBERT PH., 2006, Die Paradoxen Kombinationen von Viktimisierung und Unsicherheit in der Region Île-de-France, in OBERGFELL-FUCHS J., BRANDENSTEIN M., (Hrsg), Nationale und Internationale Entwicklungen in der Kriminologie, Frankfurt-am-Main, Verlag für Polizei Wissenschaft Dr Clemens Lorei, 315-331. ROBERT PH., FAUGERON C. 1978, La justice et son public ; les représentations sociales du système pénal, Paris-Genève, Masson-Médecine & Hygiène. ROBERT PH., GODEFROY T. 1978, Le coût du crime ou l’économie poursuivant le crime, Paris-Genève, Masson-Médecine & Hygiène. ROBERT PH., POTTIER M.-L. 1997a, Sur l’insécurité et la délinquance, Revue française de science politique, 47, 5, 630-644. ROBERT PH., POTTIER M.-L. 1997b, "On ne se sent plus en sécurité". Délinquance et insécurité. Une enquête sur deux décennies, Revue française de science politique, 47, 6, 707-740. ROBERT PH., POTTIER M.-L. 2004, Les préoccupations sécuritaires : une mutation ?, Revue française de sociologie, 45, 2, 211-242. ROBERT PH., POTTIER M.-L. 2006, Security, Law and Order : a Profound Transformation ?, Revue française de sociologie, 47, annual english selection, 35-64. ROBERT PH., ZAUBERMAN R. 1985, Les victimes entre la délinquance et l'Etat, Revue de l'Institut de Sociologie, 1-2, 9-45.

102

Victimisation and Insecurity.book Page 103 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in France

ROBERT PH., ZAUBERMAN, COLL. LEW FAI P. 1991, Enquêtes locales de victimation ; deux tests en milieu urbain, Paris, CESDIP. ROBERT PH., ZAUBERMAN R. 2006, Insécurité et traitement policier des victimations, in COLL., Une criminologie de la tradition à l’innovation. En hommage à Georges Kellens, Bruxelles, De Boeck, Larcier, 147-169. ROBERT PH., ZAUBERMAN R., NÉVANEN S., DIDIER E. 2006, Enquêtes locales 2005 sur la victimation et l’insécurité (Saint-Denis), Guyancourt, CESDIP, (www.cesdip.com). ROBERT PH., ZAUBERMAN R., POTTIER M.-L., LAGRANGE H. 1999, Mesurer le crime ; entre statistiques de police et enquêtes de victimation (19851995), Revue française de sociologie, XL, 2, 255-294. ROBERT PH., ZAUBERMAN R., POTTIER M.-L., LAGRANGE H. 2001, Measuring Crime: Police Statistics and Victimisation Surveys (1985-1995), Revue française de sociologie, 42, supplement (Annual English Selection), 133174. ROCHÉ S. 1993, Le sentiment d’insécurité, Paris, PUF. ROCHÉ S. 2000, La théorie de la ‘vitre cassée’ en France ; incivilités et désordres en public, Revue française de science politique, 50, 3, 387-412. SOUBIRAN F. slnd, Les migrants en tant que victimes d’infraction,. SOUBIRAN F. 1981, Travailleurs immigrés victimes d’infractions au système pénal, sl. VAN DIJK J.J.M., MANCHIN R., VAN KESTEREN J., NEVALA S., HIDEG G., 2007, The Burden of crime in the EU. Research Report : A comparative analysis of the European crime and safety survey (EU-ICS), Bruxelles, Gallup Europe. ZAUBERMAN R. 1982, Grandes enquêtes en recherche pénale et difficultés de réalisation : réflexions complémentaires à propos des enquêtes de victimisation, Déviance & Société, 6, 3, 281-309. ZAUBERMAN R. 1985a, Les victimes : étude du crime ou sociologie du pénal ?, L’Année sociologique, 35, 31-60. ZAUBERMAN R. 1985b, Sources d’information sur les victimes et problèmes méthodologiques dans le domaine, in COLL., Recherches sur la victimisation, Strasbourg, Colloque de l’Europe, 19-66. ZAUBERMAN R. 1991, Various Faces of the Victims, in KAISER G.,KURY H., ALBRECHT H.J., Victims and Criminal Justice, Freiburg i B., Max-PlanckInstitut für ausländisches und internationales Strafrecht, 567-594. ZAUBERMAN R. 1993, The International Crime Survey in France : Gaining Perspective, in ALVAZZI DEL FRATE A., ZVECIC U., VAN DIJK JJM., (Eds)., Understanding Crime. Experiences of Crime and Crime Control, Roma, UNICRI, 307-319.

103

Victimisation and Insecurity.book Page 104 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

ZAUBERMAN R., NÉVANEN S., ROBERT PH. 2006, Enquêtes locales 2005 sur la victimation et l’insécurité (Communauté urbaine du Grand Lyon), Guyancourt, CESDIP, (www.cesdip.com). ZAUBERMAN R., ROBERT PH. 1990, Victims as Actors of Social Control: an Empirical Inquiry in France and Some Implications, International Review of Victimology, 1, 2, 133-152. ZAUBERMAN R., ROBERT PH. 1995, Du côté des victimes, un autre regard sur la délinquance, Paris, L’Harmattan. ZAUBERMAN R., ROBERT PH., PEREZ-DIAZ C., LÉVY R. COLL. MACIOSCEK J. 1990, Les victimes, comportements et attitudes, enquête nationale de victimation, Paris, CESDIP. ZAUBERMAN R., ROBERT PH., POTTIER M.-L. 2000, Risque de proximité ou risque lié au style de vie ; enquêtes et évaluation de la sécurité urbaine, Les Cahiers de la sécurité intérieure, 42, 4, 193-220.

104

Victimisation and Insecurity.book Page 105 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany Joachim Obergfell-Fuchs

Surveys on crime and insecurity have been carried out in Germany since about the early 1970s and have reached a numerical peek in the 1990s. One can roughly differentiate three phases: – until the late 1980s mainly consisting of victim surveys with a local scope, – from the late 1980s until the early 1990s, focussing on the German reunification with national and regional surveys, often comparing East and West Germany – since the mid 1990s, local victim surveys serve mainly as a tool for the creation and evaluation of community crime prevention programs. Considering all victim and fear of crime surveys in Germany is almost impossible, many of them are not disseminated in any way. In particular this is the case for studies in the third period or for special subgroups like juveniles, women, students, etc. Often such surveys are carried out as for example master theses and they might be behold only by chance. Because of increasing crime rates measuring victimisation and fear of crime among juveniles has become an important topic in Germany. Such thematic surveys are frequently designed as student surveys because schools represent the easiest access to juveniles of all social strata. But often such thematic surveys do not measure victimisation or feelings of insecurity, their aim is getting information on self-reported delinquency, therefore they are not included here. Table 1 (see appendix) gives a chronological overview of the surveys included into the following analyses, this sample can be regarded as a quite representative cross-section of all German victimisation surveys.

I.

Synthetic stocktaking of the surveys

The first (published) victim survey in Germany dates 1973, from then until 1990 the frequency of such surveys was quite low with a mean of about 2 surveys within 5 years. But in the early 1990s a steep increase occurred, up to about 5 surveys per year were conducted and since then, the quantity 105

Victimisation and Insecurity.book Page 106 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

remained on this higher plateau, which is about the quadruple of the numbers of the 1980s. Most of the 34 surveys on victimisation and insecurity shown in table 1 (see appendix), were designed by research institutes to answer special scientific questions, some others, especially those in later years, have been developed in cooperation between local governments and research institutes, their major goal was local policy planning. Until now periodic national or statewide victimisation surveys are still lacking in Germany. On a local level, some efforts were made to develop such periodic surveys (e.g. studies 3, 4, 12, 13, 20; the numbers refer to the studies in table 1), but either they cover only a more or less narrow period of time (12, 13) or the intervals between the particular inquiries is quite long (3, 21), hence, longitudinal interpretations might be rather flawed. A similar approach can be seen with respect to some thematic surveys like the student surveys by the Criminological Research Institute of Lower Saxony. Here, the studies in 1998 (23), 2000 (28), and 2005 (34) follow the same methodological pattern in order to allow longitudinal interpretations. Most studies do not only include questions on victimisation and fear of crime, frequently, they also comprise topics like attitudes to punishment, assessment of the police and legal authorities, or the perception of incivilities. Some polling institutes temporarily or even continuously ask for some criminological aspects in their questionnaires (e.g. Wohlfahrtssurvey, GESIS; Sozioökonomisches Panel, Deutsches Institut für Wirtschaftsforschung DIW; ALLBUS, GESIS; or Die Ängste der Deutschen, Infocenter R+V Insurances). Here, particular questions on fear of crime are embedded in other research topics like the social or economical situation or general aspects of wellbeing. Hence such studies cannot be regarded as insecurity surveys in the narrow sense of the term therefore they are not included in table 1. Some aspects of these surveys will be addressed below.

1

Territorial scope of the surveys

Among the 34 surveys taken into the analysis, 8 are national surveys (8, 10, 11, 14, 18, 19, 31, 33), only one survey has a strictly regional scope (27) while the half of all surveys are strictly local. Seven surveys combine different aspects, 5 of them have a combination of local and regional scope (12, 13, 17, 28, 34) which generally means they cover more than one city in larger region, e.g. of different community sizes, in order to compare urban and rural areas. In such surveys, regional aspects themselves mostly play an only subsidiary role. Although the study by Kreuzer et al. (4) has regional aspects (Universities of Giessen, Jena, and Potsdam), a description 106

Victimisation and Insecurity.book Page 107 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

of the territorial scope is impossible due to a diversity in students’ home addresses. Two other surveys were international cooperation projects, one of them comparing cities (29), respectively city areas in Europe, another one drew an international comparison of different regions (5). A more in-depth analysis shows that the majority of the surveys (24 out of 34) refers to general victimisation surveys, this means a more or less comprehensive and representative sample of the general population has been surveyed. Another 8 surveys are designed as thematic student surveys, most of them refer to samples of youngsters in public schools. Another single survey is designated only to women and represents the first and sole large scale survey on women’s victimisations with more than 10,000 interviews (31). Finally one survey on victimisation and fear of crime among university students was added to this analysis (4). When looking at the territorial scope of each of these four subgroups it is remarkable that about the half of the general surveys on victimisation and fear of crime are local studies, most of them are designed as single site surveys, this means they comprise one single community. Some other local surveys are designed as multi site studies (e.g. 23), investigating at least two communities, normally in order to compare the results. Except the survey on women (31) all national studies are general surveys, e.g. national studies on victimisation of students are up to now not existing. While local single site studies are more or less dominating the category of the general surveys (n = 11 of 24), the picture with student surveys is somewhat more heterogeneous. Local aspects are dominating too, but the portions of single site (n = 3), multi sites (n = 2) or regional surveys (n = 3) are more comparable. The reasons for the dominance of spatially limited surveys are manifold: sampling is much easier than with national surveys, results are easier interpretable, and, above all, the surveys are cheaper. Not the least, many of the local surveys are commissioned by local authorities. Characteristics of national surveys In Germany, there were only a few national surveys during the last decades. In 1989 the first national German victimisation survey has been carried out as part of the first wave of the International Crime and Victimisation Surveys ICVS (8). During the time of this survey, the fall of the Berlin Wall and the opening of the borders to the GDR occurred, which opened the floor for subsequent comparisons of victimisation and feelings of insecurity between East and West. This led to three further comparative victim surveys (10, 11, 14). In 2005 Germany participated for the second time in the meanwhile 5th wave of the ICVS (33), this has been the sixth

107

Victimisation and Insecurity.book Page 108 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

national survey on general victimisation and fear of crime within since 1989. Some of the national surveys were not only used for theory testing or explorative analyses, they often served as a basis for comparison for other, e.g. local, studies (e.g. 18). The main idea hereby was to gather representative national data in order to compare results. An example is the commission of a national survey by the Research Group Community Crime Prevention in Baden-Wuerttemberg in 1995. Sometimes such surveys also served as methodological tools, e.g. for the testing of the reliability and validity of parts of the questionnaire. For example the study by Heinz et al (1997, 19) became such a methodological survey, although primarily not intended, because of significant variances in victimisation rates between two parts of the survey. In 2003 the first national survey on victimisation and fear of crime among women commissioned by the Federal Ministry for Family, Seniors, Women, and Youth has been realized (31). Besides a vast representative sample (n = 10,000), special subgroups like migrants, prostitutes, or prison inmates have been included. This survey allowed a comprehensive view on the situation of women in partnerships and under special living conditions and their specific risk of becoming a victim of violent acts. For long times German criminologists accentuated the need for continuous periodical national victimisation surveys modeled on the National Crime Survey in the USA or the British Crime Survey. Many attempts came to grief, mainly because of a lack of financial means. But also the controversy between the criminological research institutes who should be responsible for the scientific leadership of such surveys did not promote the endeavor. Characteristics of local surveys With local crime and fear of crime surveys one can clearly distinguish two phases: A first one, until the late 1990s was designated to theory testing by means of local surveys. Theories like urban ecology or defensible space took center stage of the research designs. This has changed at the beginning of the 1990s when the major aim of the surveys became delivering data for community policy planning and community crime prevention. Most of the local surveys cover mid-size communities between 100,000 und 600,000 inhabitants, such community sizes have the advantage that the cities are big enough for having a measurable crime problem and hence are sufficiently interesting for scientific purposes. Furthermore, administrative districts are mostly large enough to allow sound analyses, but they are small enough to have their particular appearance. In larger cities this is not always guaranteed because some administrative districts

108

Victimisation and Insecurity.book Page 109 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

themselves have dimensions of mid-size communities, whereas smaller cities often do not have seizable crime problems. However, in the last 15 years even smaller cities (much less than 100,000 inhabitants) became an object of scientific research on victimisation and fear of crime, mostly due to crime prevention schemes. Almost all of the local surveys cover spatial analyses and spatial comparisons of different areas within the community in order to assess peculiarities in crime load and fear of crime. The vast majority of the thematic surveys measuring fear of crime and victimisation among juveniles (student surveys) are locally bound too. This is mostly due to sampling procedures, normally, some schools and within these schools some classes are selected and surveyed, a procedure which leads more or less automatically to a local approach. Because strictly regional surveys are quite seldom, it is not useful to designate an own paragraph to this subgroup. As mentioned above, most of them focus on different communities within a circumscribed region and hence the surveys are rather self-contained regional but a couple of local surveys with the same questionnaire and the same methodology in nearby communities. Only two typical regional studies are depicted in table 1: The survey by Arnold et al. (5) compares victimisation and fear of crime in the state of Baden-Wuerttemberg (Germany) with the region of Baranya in Hungary and the state of Texas in the USA. Therefore, it is not only one of the first German large-scale victim surveys, but also one the few studies with supranational comparison (see study 29 as well). The other regional survey by Oberwittler et al. (27) is a completion of their student survey in Freiburg and Cologne in order to compare urban and rural aspects of victimisation and crime among students. Summing up, most of the German victim surveys during the last 30 years were locally or regionally limited, rather expensive national surveys were rare. A bundle of surveys during this time period was dedicated for special purposes like victimisation of juveniles, realized as student surveys, often combined with self-reported delinquency. Most times the main idea of these studies was to give information on the crime development and crime load of juveniles under particular social situations.

2

Funding

Funding of the German surveys on victimisation and fear of crime is extremely heterogeneous. Among the main funding agencies are different kinds of foundations, above all the German Research Foundation (DFG) but also other smaller foundations. Further important funding agencies for such criminological research are the European Union (e.g. surveys 29, 33),

109

Victimisation and Insecurity.book Page 110 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

national ministries as well as state ministries and national departments like the German Federal Criminal Police Office (BKA). While these institutions often fund national surveys or at least studies with a more comprehensive scientific background, local authorities became important financial backers for surveys in the area of community crime prevention. Together with local associations or local private institutions, e.g. financial institutes, many of these surveys are funded. Up to now, public authorities and independent foundations are the most common financial backers for victimisation and fear of crime surveys, the influence of private institutions is still very small in this kind of German research. There are huge differences in funding between surveys which are part of a contract research project and those being a project by a research institute itself. While the former are normally fully financed by the customer, the latter are often at best partly financed by means of fundraising. Examples of contract research projects are the national survey on victimisation of women (31) and the national surveys by Wetzels et al. (14) or Heinz et al. (19), but also many of the local studies with respect to community crime prevention projects (20, 21, 22, 30). In total, only a minority of the studies is based on contract research, some are developed in cooperation between authorities and research institutes, e.g. the first German-German victim survey (10). This means that most surveys are at least partly if not fully financed by the research institute itself, a fact which might explain the frequently limited extension of the surveys and the frequent use of mail survey methodology.

3

Scientific and technical leadership

In most of the described studies, 21 out of 34, scientific and technical leadership was the duty of the organizing research institute. This means, the research institute designs the survey, it conducts it and it looks for possible funding. Because, as mentioned above, funding is quite limited, an outsourcing of sampling or data analysis is, due to its considerable costs, quite rare. In only 8 of 34 studies, almost all of them national surveys, technical leadership was given to commercial polling institutes. The advantages for such an outsourcing are the use of a standardized methodology (e.g. ICVS, 8, 33), the possibility of participating at surveys with extremely large sample sizes planned by the polling institute anyway (e.g. 10, 18, 19) and the use of comprehensive methodological abilities, e.g. special sampling techniques like CATI or random walk. Such commercial polling institutes often dispose of a large experienced interviewer staff and of extensive regional logistics.

110

Victimisation and Insecurity.book Page 111 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

In four studies (21, 22, 26, 30), technical leadership was in the hand of the commissioning community or police authority while the scientific leadership rested with the participating research institute. All of them were victimisation surveys in the field of community crime prevention. The reasons thereof were a sound knowledge of the particular local situation and an easier feasibility on the side of the local authorities. In some surveys scientific and/or technical leadership has been shared by different research institutes, often it was about large-scale studies, for example a couple of German-German comparative victimisation surveys. Besides leadership, many surveys used methodological counselling of central research institutes, especially the Center for Survey Research and Methodology (ZUMA) at Mannheim is often contacted for such purposes.

4

Technical characteristics

The following section deals with technical characteristics (sampling procedures and administration) but also the measurement of victimisation and fear of crime. Sampling Half of the surveys depicted in table 1 are based on random samples from official registration files. This is a rather cheap and common way of sampling, mostly used in surveys with smaller budget, especially in the field of community crime prevention. Less frequent and only used in general victimisation surveys are random walk designs, this is particularly the case with national surveys under the technical leadership of polling institutes (10, 11, 14, 18, 19). Here, the so-called ADM-design, developed by the representation of private market and private institute conducting sociological and economical research, is used. Almost all thematic student surveys (15, 16, 23, 25, 27, 28, 34) used a stratified sampling procedures, where schools and classes are randomly selected and then all students within a class are interviewed. This approach ensures an approximated representativeness. Exceptions are the comprehensive surveys by Kreuzer et al (4), they used a total census of university freshmen, and by Feltes and Goldberg (32), who adopted a total census of all students participating at a project day. Until now, telephone surveys are quite rare in German research on victimisation and fear of crime, in former years this was due to a low private telephone density in East Germany, but nowadays other reasons like a lack of particular expert-knowledge in the research institutes or a lack of technical equipment for CATI might play an important role. 111

Victimisation and Insecurity.book Page 112 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Generally, almost all sampling procedures are based on random selection and aim at representativeness, only one study partly consisted of selected samples (31), here female migrants, prostitutes, and female prison inmates were surveyed. So the data reflect a quite high methodological standard of victim surveys in Germany. Administration Besides sampling, the used method is another important technical aspect of the surveys. In German victimisation surveys face-to-face interviewing and mailed questionnaires are the most frequent methods. While 10 studies in table 1 are based on face-to-face interviews (1, 2, 6, 10, 11, 14, 18, 19, 29, 31), 12 others used mailed questionnaires (4, 5, 7, 9, 12, 17, 20, 21, 22, 24, 26, 30). Discussing the advantages and disadvantages of each method at this place is almost impossible, but one can adhere to the statement that because of interviewers’ charges, face-to-face methodology is normally much more expensive than mailed questionnaires and, as shown above, money is often a limiting factor. Furthermore, studies showed that the problem of social desirable answering is reduced in mailed questionnaires (cf. Kury [1994]). Otherwise, an enormous problem of mailed surveys is the low response rate, in most cases it was less than 50 percent, sometimes even much lower. This is a limiting factor of the representativeness of a survey and complicates for example the estimation of dark figures or police effectiveness by means of victimisation survey data. In face-to-face surveys, as for example in some national German surveys, this problem is less distinctive, here the response rates are normally about 70 percent. Generally response rates have decreased in German surveys, for example Schwind et al (1975) got 95.6% by personal interviews, such response rates are nowadays utopian. As already mentioned, telephone interviews, and especially CATI technique, is extremely seldom in German research on victimisation, the reasons are discussed above. In fact, only the two ICVS surveys in 1989 (8) and 2005 (33) used this method. According to response rates the experiences were not better, sometimes even worse than other methodologies, the 1989 national ICS survey yielded 32% (Kury 1991) and in the 2005 ICVS it was 43.3% (EUICS Report 2005). Such response rates are even available with mailed questionnaires. A special form of administration is used in all student surveys (16, 23, 25, 27, 28, 32, 34, except the university student survey by Kreuzer et al., 4), here, written questionnaires were handed to the students in their classrooms and filled out during a lesson. Because of this special

112

Victimisation and Insecurity.book Page 113 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

methodology in a specific setting response rates are very high and about 90 percent. A couple of surveys combine methodologies, for example Schwind et al. (1998) (3) mixed face-to-face and telephone interviewing, while Kräupl and Ludwig (1993) (13) combined mail and face-to-face technique in their 1991 survey. But such a method mix is rather an exception than a rule in German victimisation surveys. Types of victimisations When calculating general victimisation rates, often the problem comes up that there is no consistent catalogue of victimisation incidents inserted into the surveys. But within the last years it has become a common practice in general German victimisation surveys using a list of victimisation questions, stemming from the 1989 ICVS (8) in order to harmonize the questionnaires. Further efforts were undertaken for example by establishing a standard inventory for community crime prevention surveys (Forschungsgruppe Kommunale Kriminalprävention Baden-Württemberg 2000). These lists mostly comprise all kinds of vehicle thefts (car, motorcycle, bicycle), vandalism (car and personal belongings), burglary (incl. attempt), theft of personal belongings, robbery, assault and harassment as well as molestation, sexual violence, and rape. In recent years some surveys included fraud and sometimes corruption. In some surveys reduced victimisation lists are used (e.g. 3, 6, 9). As the following figure 1 shows, the different types of theft, assault, vandalism and robbery are included in more than 90% of all victimisation surveys. Less frequent, but still the majority of all surveys asks for burglary and molestation. About half of all victimisation surveys ask for rape, fraud and harassment, while corruption or fleeing from the scene of an accident is rarely assessed. It is obvious that it has indeed to a large extent been possible to establish a basic victimisation questionnaire which is accepted among the German research institutes, although there are still some differences in measuring victimisation between the surveys. But still not all surveys use all questions from the virtual catalogues and this complicates comparability.

113

Victimisation and Insecurity.book Page 114 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Figure 1 :Types of victimisations in general victim surveys

Somewhat more heterogeneous is the reference time for prevalence rates. While in former studies (e.g. ICVS [1989]) one and five year periods were used, it has meanwhile become more common to limit the time to 12 months because of reducing telescoping effects. But many other reference times are also observable. The situation with thematic student surveys is different, unlike general victimisation surveys, almost no efforts were undertaken to standardize questionnaires. However, all student surveys ask for assault, the vast majority also for robbery and extortion. In about half of the surveys victimisation experiences in sexual violence are gathered. Less frequent are harassment, molestation, parental abuse and vandalism. Only a small minority of the surveys asks for vituperation, respectively slander – it is arguable whether these are real victimisations at all –, assault by a teacher or coercion. Often, victimisation incidents are asked for the situation in school, in the schoolyard, and on the way to school, some surveys also comprise victimisations at home (e.g. parental abuse) or during leisure time activities. Just to complete the picture on thematic surveys, the university students survey by Kreuzer et al. (1993) (4) comprised victimisation experiences according to child maltreatment, molestation and rape, while the national survey on victimisation experiences among women (Bundesministerium für Familie, Senioren, Frauen und Jugend [2004]) (31) asked for assault, sexual violence, and psychological violence in general as well as in close relationships. 114

Victimisation and Insecurity.book Page 115 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

Summing up, while general surveys are becoming more standardized according to types of victimisations, at least probably a consequence of the increased establishment of standardized victimisation inventories, all other surveys are still heterogeneous in this point. Fear of crime During the last decades there has been an extended discussion on fear of crime and its measurement. Almost all surveys depicted in table 1 comprise besides questions on victimisation the issue of fear of crime. Only six surveys (two of them general surveys 2, 6; student surveys 4, 16, 25, 27) do not consider fear of crime in their questionnaires. Meanwhile there exists a plenitude of different questions on this topic and a huge discussion arose within the last years which item would be most reliable and most valid. The prominent item of measuring fear of crime is still the so called “standard item” concerning feelings of insecurity in the home area after nightfall (“How safe do you feel in your home area walking alone after nightfall”). Because it has often been criticized measuring anything but not fear of crime it has been modified in several surveys, now it is often be amended by the expression “afraid of becoming a crime victim”. But recent methodological studies showed that the standard item is not as bad as its reputation (cf. Kury et al [2004]). The more or less “emotional” aspect of the “standard item” is often contrasted with cognitive aspects (3, 7, 11, 10, 14, 15, 18, 19, 20, 21, 22, 24, 26, 29, 33), which means asking for the assumed probability of becoming a crime victim (mostly different crimes are given) within a certain period of time, often one year. Other surveys ask for the avoidance of places because of fear or letting be escorted when going out after nightfall (e.g. 3, 8, 10, 13, 17, 18, 20, 21, 24, 26), such questions covers behavioral consequences of fear of crime. Further aspects are thinking about victimisation or concern about crime, items which are similar to the above mentioned assumed probability of victimisation but with more emotional aspects, fear at home after nightfall, again, similar to the so called standard item, or the limitation of leisure activities. This is only a selection of the most frequent questions in German surveys, there are many others, for example the description of fear inducing scenarios (cf. Kury et al. [2004]), the drawing of precautionary measures and so on. In local surveys, especially those with regard to community crime prevention, such questions are frequently supplemented by open questions concerning fear in particular city areas. A more in-depth analysis is given in figure 2. It shows the frequency of each item in the 28 analyzed surveys, which include questions on fear of crime.

115

Victimisation and Insecurity.book Page 116 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

As mentioned above, almost all surveys comprise the “standard item”, this means, despite the huge critique it is still most common and therefore important for comparative analyses. Meanwhile more than the half of the surveys request the subjects to assess the probability of becoming a crime victim, normally during the following 12 months, often according to different crimes. Half of the surveys ask for the avoidance of places about one third for thinking about victimisation. All other items are less frequent, e.g. the concern about various crimes, escort after nightfall, fear at home after nightfall, the limitation of leisure activities, the drawing of precautionary measures, feelings of safety, respectively unsafety in the home area, the drawing of self-defence measures, the expectation of victimisation, or staying at home because of fear. Despite the limited numbers, the figure shows the variety in fear of crime measurement, a harmonization is yet not in sight. It seems like each research institute and each survey would again reinvent the wheel. Hence, an informal agreement concerning the use of some standardized variables would be very helpful for comparative reasons.

Figure 2: Items measuring fear of crime in general victim surveys

In student surveys, questions on fear of crime do not play the same important role as in general victimisation surveys. Only 5 of 8 surveys comprised such items. Facing the small number, the data are hardly interpretable. Quite often used are questions concerning fear of violent crimes, feelings of safety or insecurity in various situations and the drawing 116

Victimisation and Insecurity.book Page 117 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

of precautionary measures. Other questions refer to the probability of victimisations or the fear of victimisation. These, only cautiously, interpretable data show that the situation among student surveys differ from general surveys, even the items measuring fear of crime are to some extent different. Since many years some thematic surveys in Germany considering fear of crime, respectively concern about crime, have been carried out. As already described above, these surveys do not have a criminological background, crime is only a small aspect within a broad collection of worries among Germans. One of these studies is the “Socio-Economic Panel (SOEP)” by the German Institute for Economic Research (www.diw.de/english/sop/ index.html). The first panel started in 1984, since then every year a large sample of private households has been surveyed by face-to-face interviews on household composition, occupational biographies, employment, earnings, health and satisfaction indicators. In 2005, nearly 12,000 households with more than 21,000 persons were sampled. Within the survey, personal worries are recorded, since 1994 one of these worries is the development of crime in Germany. Another survey is carried out by the R+V Versicherungen, one of the largest German insurance companies (www.ruv.de/de/presse/r_v_infocenter/ studien/index.jsp). This survey started for the first time in 1991 and since then it has been conducted every year. The questionnaire contains a rating of 15 major concerns, one of them the fear of becoming a crime victim. The sample consists of about 2,000 Germans 14 years and older surveyed by means of face-to-face interviews.

II.

Evaluation of the use of the surveys

1

Use and impact of the surveys

Political use and impact It would be illusionary to assume that surveys on victimisation and fear of crime are heavily demanded by public policy makers and that they will necessarily lead to political decisions. A study by Kasperzak (2000) showed that even after a short period of time, comparable local surveys were unknown or became forgotten among most local policy-makers, even among those who formerly launched the survey. An implementation of measures based on the results of these surveys did practically not occur. But, without doubt, there are some differences between the surveys: Those, covering actual and important political and societal topics have a much 117

Victimisation and Insecurity.book Page 118 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

higher chance of becoming noticed. For example, this was the case with the East-West comparative studies after the German reunification (10, 11, 14). Because no comparable data were available for the former GDR such results were widely noticed and discussed among national and regional public policy makers, but one can not really say, whether they had a real impact on policy making. For local surveys during the heyday of community crime prevention, the situation was somewhat similar. Here, local or regional policy makers were often involved in the planning of the surveys and therefore they were highly interested in the results (see below). For student surveys, the situation differs again. Their results are often then required when spectacular incidents, e.g. amok runs of students, occurred. In such situations policy makers ask for survey results and the chance that implications of the studies become realized increases. Furthermore, school authorities are often interested in such results because they get neutral indepth information on the prevailing situation in their schools and classrooms. Generally, the influence of surveys on policy making increases in two cases: * When they offer suitable results during election campaigns, then they are required by different political parties. * when the results und the drawn implications by the scientists provide an opportunity to save public money. Scientific use and impact The scientific use and impact of the surveys differs from their political or administrative use, sometimes it is nearly vice versa. Although their political impact was quite limited, there has been a substantial scientific use of national surveys, especially of those comparing East and West Germany. Their results led to intense discussions about the consequences of political change on crime and fear of crime development. Further they provide an insight into crime control in a formerly largely insulated country. With local surveys, the situation is quite opposite, while early local studies had a quite relevant scientific impact because of theory testing, e.g. social desorganization or defensible space, the use of later local surveys was often rather limited. In many cases they served for baseline determination or as more or less important advices for community crime prevention. Although, in this role, their political impact was often substantial. The scientific impact of student surveys is largely considerable, most times their results are widely discussed in psychological and pedagogical circles.

118

Victimisation and Insecurity.book Page 119 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

They serve as a state description what is going on in school as well as a tool for theory testing on crime development among youths.

2

Influences on policy making

Some influences on policy making are already described and discussed, most important for policy making were at least local surveys during the last two decades. Their results were highly demanded, not the least because the studies were commissioned by the communities themselves and introduced into the process of crime prevention development. But in fact, only few measures were directly drawn from the surveys, more often the survey results should pay for a belated justification of already designated activities. However, although often planned and discussed in crime prevention projects, a post intervention measurement by means of victimisation and insecurity surveys for evaluative purposes is extremely rare. The most common reason thereof is the lack of financial means, but one can also assume a more or less sceptical attitude towards this kind of surveys. They always bear the risk that the implemented measures are assessed as inefficient and such judgement would probably cause political consequences.

3

Dissemination and publication

Finally, it is necessary to discuss the dissemination and the publication of the surveys on victimisation and insecurity in Germany. The vast majority (26 of 34) of the surveys presented in this paper have been published or at least disseminated in different ways. Most larger surveys are published in journals, readers, and monographs, or at least in exhaustive research reports, sometimes electronically available. A few papers are only reported in more or less short articles, some few others are not disseminated. For these projects, full reports are surely existing, but as gray literature they are hardly accessible, because the results of the surveys are only disseminated among a particular subgroup which has been involved into the research process. Indeed, the number of surveys on victimisation and insecurity during the last decades is much larger than the 34 described, but it is only by chance getting knowledge of such unpublished surveys. Many of these latter surveys concern scientific qualification, e.g. Master theses, others might be local surveys for community crime prevention processes. Although in Germany there exists a databank called FORIS (www.gesis.org/ 119

Victimisation and Insecurity.book Page 120 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Information/FORIS/index.htm), which comprises studies in social sciences of the last 10 years, it is finally up to the author to enter his or her study to this databank. Unfortunately many surveys are not disseminated in any way and therefore became lost for the scientific community.

4

Future perspectives of surveys on victimisation and insecurity

A final paragraph should deal with some future perspectives on German research on victimisation and fear of crime. First of all, there is an urgent need for periodical national victimisation surveys and the latest news from the federal ministries generate optimism that this endeavor will – finally – become realized within the next few years. Further, there is a high probability that the number of large local surveys for community crime prevention purposes will decrease. Such surveys are quite expensive and their utility has to be doubted. Often, as Kasperzak pointed out (see above), the survey reports end up in a bookshelf and no further consequences are drawn. In recent years, the importance of methodological aspects has increased, especially concerning fear of crime, the question came up ‘what do we measure und what is the best way to do this’. Other methodological problems are the declining response rates, especially in mailed surveys, and the need for alternative sampling methods with regard to limited financial resources. Finally, thematic surveys, not only on students or women, have increased, and often such surveys, e.g. on selected subgroups (e.g. minorities, employees of special branches, older people etc.) or on single aspects like fear of crime, will be more promising than ordinary general victim surveys. One current example is a media notice from March 2007, when the Federal Home Ministry has commissioned a national student survey with a sample size of about 50,000 students in order to develop new strategies for preventing juvenile crime.

References ABEN R. 1992, Kriminologische Regionalanalyse Lübeck In Koch K.-F. (Hrsg), Kriminalitätslagebilder Wiesbaden, Bundeskriminalamt, 305-341. ARNOLD H., KORINEK L. 1991, Victimisation, attitudes towards crime and related issues: Comparative research results from Hungary. In KAISER G., KURY H., ALBRECHT H.-J. (Hrsg.), Victims and criminal justice Freiburg, Max-Planck-Institut für Strafrecht Vol.51, 99-121.

120

Victimisation and Insecurity.book Page 121 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

BAIER D., PFEIFFER C., WINDZIO M., RABOLD S. 2006, Schülerbefragung 2005: Gewalterfahrungen, Schulabsentismus und Medienkonsum von Kindern und Jugendlichen Hannover, Kriminologisches Forschungsinstitut Niedersachsen www.kfn.de/schuelerbefragung.shtml. BOERS K. 1991, Kriminalitätsfurcht. Pfaffenweiler, Centaurus. BOERS K., GUTSCHE G., SESSAR K. (Hrsg.), 1997, Sozialer Umbruch und Kriminalität in Deutschland, Opladen, Westdeutscher Verlag. BUNDESMINISTERIUM FÜR FAMILIE, SENIOREN, FRAUEN UND JUGEND 2004, Lebenssituation, Sicherheit und Gesundheit von Frauen in Deutschland Berlin, BMFSFJ. DÖLLING D., FELTES T., HEINZ W., KURY H. (Hrsg.) 2003, Kommunale Kriminalprävention – Analysen und Perspektiven. Holzkirchen, Felix. DREHER G., KURY H., OBERGFELL-FUCHS J. 2005, Bevölkerungsumfragen in ländlichen und städtischen Regionen – Kriminalitätsanalysen und gezielte proaktive Maßnahmen am Beispiel der Stadt Rottweil In BANNENBERG B., COESTER M., MARKS E. (Hrsg.), Kommunale Kriminalprävention, Mönchengladbach, Forum Verlag Godesberg, 169-188. EU ICS REPORT 2005, The burden of crime in the EU, a comparative analysis of the European Survey of Crime and Safety (EU ICS). www.europeansafetyobservatory.eu/ FELTES, T., GOLDBERG, B. 2006, Selbstberichtete Delinquenz, Viktimisierung und Verbrechensfurcht bei Schülern mit und ohne Migrationshintergrund. In OBERGFELL-FUCHS J., BRANDENSTEIN M. (Hrsg.), Nationale und internationale Entwicklungen in der Kriminologie. Festschrift für Helmut Kury zum 65. Geburtstag, Frankfurt/Main, Verlag für Polizeiwissenschaft, 203-237. FORSCHUNGSGRUPPE KOMMUNALE KRIMINALPRÄVENTION BADEN-WÜRTTEMBERG 1998, Viktimisierungen, Kriminalitätsfurcht und Bewertung der Polizei in Deutschland, Monatsschrift für Kriminologie und Strafrechtsreform, 81, 2, 67-82. FORSCHUNGSGRUPPE KOMMUNALE KRIMINALPRÄVENTION IN BADEN-WÜRTTEMBERG, 2000, Handbuch zur Planung und Durchführung von Bevölkerungsbefragungen im Rahmen der Kommunalen Kriminalprävention (2nd ed.), Stuttgart, Innenministerium Baden-Württemberg. FUNK W. (Hrsg.) 1995, Nürnberger Schüler-Studie 1994, Regensburg, S. Roderer. HERMANN D., DÖLLING, D. 2001, Kriminalprävention und Wertorientierungen in komplexen Gesellschaften, Mainz, Weisser Ring. HERMANN D., LAUE C. 2005, Wirkungen kommunaler Kriminalprävention – Ein Fallbeispiel. In BANNENBERG B., COESTER M., MARKS E. (Hrsg.), Kommunale Kriminalprävention. Ausgewählte Beiträge des 9. Deutschen Präventionstages, Mönchengladbach, Forum Verlag Godesberg, 197-208. 121

Victimisation and Insecurity.book Page 122 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

KASPERZAK T. 2000, Stadtstruktur, Kriminalitätsbelastung und Verbrechensfurcht, Holzkirchen, Felix. KRÄUPL G., LUDWIG H. 1993, Wandel kommunaler Lebenslagen, Kriminalität und Sanktionserwartungen. Bevölkerungsbefragung in einer städtischen Region Thüringens1991/92 (Jenaer Kriminalitätsbefragung), Freiburg, Max-Planck-Institut für Strafrecht. KRÄUPL G., LUDWIG H. 2000, Wahrnehmung von Kriminalität und Sanktionen im Kontext gesellschaftlicher Transformation Freiburg edition iuscrim. KREUZER A., GÖRGEN T., KRÜGER R., MÜNCH V., SCHNEIDER H. 1993, Jugenddelinquenz in Ost und West, Bonn, Forum Verlag Godesberg. KURY H. 1991, Victims of crime – Results of a representative telephone survey of 5,000 citizens of the former Federal Republic of Germany. In KAISER G., KURY H. ALBRECHT H.J. (Eds.), Victims and criminal justice, Freiburg, Max-Planck-Institut für Strafrecht, Vol. 50, 265-304.. KURY H. 1994, The influence of the specific formulation of questions on the results of victim studies European Journal on Criminal Policy and Research, 2, 48-68. KURY H., DÖRMANN U., RICHTER H., WÜRGER M. 1992, Opfererfahrungen und Meinungen zur Inneren Sicherheit in Deutschland,Wiesbaden, Bundeskriminalamt. KURY H., LICHTBLAU A., NEUMAIER A., OBERGFELL-FUCHS J. 2004, Zur Validität der Erfassung von Kriminalitätsfurcht, Soziale Probleme, 15, 141-165. KURY H., OBERGFELL-FUCHS J., BRAUN V. 1999a, Kriminologische Dunkelfeldanalyse der Stadt Reutlingen, Freiburg, unpublished. KURY H., OBERGFELL-FUCHS J., TELL S. 1999b, Kriminologische Dunkelfeldanalyse der Stadt Metzingen. Freiburg, unpublished. KURY H., OBERGFELL-FUCHS J., WÜRGER M. 2000, Gemeinde und Kriminalität. Eine Untersuchung in Ost- und Westdeutschland, Freiburg, edition iuscrim. LUDWIG H., KRÄUPL G. 2005, Viktimisierung, Sanktionen und Strafverfolgung. Jenaer Kriminalitätsbefragung über ein Jahrzehnt gesellschaftlicher Transformation, Mönchengladbach, Forum Verlag Godesberg. OBERWITTLER D., BLANK T., KÖLLISCH T., NAPLAVA T. 2001, Soziale Lebenslagen und Delinquenz von Jugendlichen. Arbeitsberichte 1/2001 Freiburg, Max-Planck-Institut für Strafrecht. OBERWITTLER D., KÖLLISCH T., NAPLAVA T., BLANK T. 2002, MPI-Schulbefragung Breisgau/Markgräfler Land 2000, Freiburg, Max-Planck-Institut für Strafrecht. PLATE M., SCHWINGES U., WEIß R. 1985, Strukturen der Kriminalität in Solingen, Wiesbaden, Bundeskriminalamt. 122

Victimisation and Insecurity.book Page 123 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

POSCH C., MAIER I., KURY H., OBERGFELL-FUCHS J. 2001, Kriminologische Dunkelfeldanalyse der Stadt Mannheim, Freiburg, unpublished. SCHNELL R. KREUTER F. 2000, Untersuchungen zur Ursache unterschiedlicher Ergebnisse sehr ähnlicher Viktimisierungssurveys Kölner Zeitschrift für Soziologie und Sozialpsychologie, 52, 1, 96-117. SCHWIND H. -D., AHLBORN W., EGER H.J., JANY U., PUDEL V., WEIß R. 1975, Dunkelfeldforschung in Göttingen 1973/74, Wiesbaden, Bundeskriminalamt. SCHWIND H.-D., FETCHENHAUER D., AHLBORN W., WEIß R. 2001, Kriminalitätsphänomene im Langzeitvergleich am Beispiel einer deutschen Großstad. Neuwied, Kriftel, Luchterhand. SCHWIND H.-D., ROITSCH K., AHLBORN W., GIELEN B. (Hrsg.) 1995, Gewalt in der Schule. Mainz, Weisser Ring. SESSAR, K. 1992, Wiedergutmachen oder strafen: Einstellungen in der Bevölkerung und der Justiz Pfaffenweiler, Centaurus. SESSAR K. 2006, Warum man abends nicht das Haus verlässt. Kriminologische Ergebnisse aus einem europäischen Forschungsprojekt zu Unsicherheiten in Großstädten In OBERGFELL-FUCHS J., BRANDENSTEIN M. (Hrsg.), Nationale und internationale Entwicklungen in der Kriminologie. Festschrift für Helmut Kury zum 65. Geburtstag Frankfurt/Main, Verlag für Polizeiwissenschaft, 265-293. STEPHAN E. 1976, Die Stuttgarter Opferbefragung. Wiesbaden, Bundeskriminalamt. TESKE R. H. C., ARNOLD H. 1991, A comparative victimisation study in the United States and the Federal Republic of Germany: A description of the principal results In KAISER G., KURY H. ALBRECHT H.J. (Eds.), Victims and criminal justice, Freiburg, Max Planck Institut für Strafrecht, Vol. 51, 3-44. WETZELS P., ENZMANN D., MECKLENBURG E., PFEIFFER C. 2001, Jugend und Gewalt Baden-Baden, Nomos. WETZELS P., GREVE W., MECKLENBURG E., BILSKY W., PFEIFFER C. 1995, Kriminalität im Leben alter Menschen Stuttgart et al., W. Kohlhammer. WILMERS N., ENZMANN D., SCHAEFER D., HERBERS K., GREVE W., WETZELS P. 2002, Jugendliche in Deutschland zur Jahrtausendwende: Gefährlich oder gefährdet? Baden-Baden, Nomos.

123

Victimisation and Insecurity.book Page 124 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Appendix No Author 1 Stephan

Year 1973

Territorial Scope local: Stuttgart

Focus general

Publication Stephan 1976

2 3

Schwind Schwind et al.

local: Goettingen local: Bochum

general general

Schwind et al. 1975 Schwind et al. 2001

4

Kreuzer et al.

1973/74 1975 – 1986 – 1998 1990

no area specified

university Kreuzer et al. 1993 students

5

Arnold et al.

1981

general

Teske & Arnold 1991; Arnold & Korinek 1991

6 7

Plate et al. Sessar, Boers

1982 1984

regional/ supranational: BadenWuerttemberg, Germany – Baranya, Hungary – Texas, USA local: Solingen local: Hamburg

general general

8 9 10

Kury Aben Kury et al 1990

1989 1990 1991

national: 1st ICVS general local: Luebeck general national: East and West general Germany

Plate et al. 1985 Boers 1991; Sessar 1992 Kury 1991 Aben 1992 Kury et al. 1992

11

Boers et al.

national: East and West general Germany

Boers et al. 1997

12

Kury et al.

1991 – 1993 – 1995 1991/92 – 1995-96

general

Kury et al. 2000

13

general

14

Kräupl, Ludwig 1991/92 – 1995/96 – 2001/02 Wetzels et al. 1992

local/regional: Freiburg – Emmendingen – Loeffingen local/regional: Jena – Kahla- Suhl

Kräupl & Ludwig 1993; 2000; Ludwig & Kräupl 2005 Wetzels et al. 1995

15

Schwind et al.

1993

16

Funk et al.

1994

17

Research Group 1994 Community Crime Prevention in BadenWuerttemberg Research Group 1995 Community Crime Prevention BadenWuerttemberg

18

124

national: East and West general Germany local: Bochum students 6–21y local: Nuremberg students 12–15y local/regional: Calw, general Freiburg, Ravensburg, Weingarten

National

general

Schwind et al. 1995 Funk 1995 Dölling et al. 2003

Forschungsgruppe Kommunale Kriminalprävention BadenWuerttemberg 1998

Victimisation and Insecurity.book Page 125 Thursday, December 4, 2008 12:33 PM

Crime victims and insecurity surveys in Germany

No Author 19 Heinz et al.

Year 1997

Territorial Scope National

Focus general

20

Hermann et al.

local: Schwetzingen

general

21 22 23

Kury et al. Kury et al. Wetzels et al.

1997 – 2004 1998 1998 1998

general general students 15–17y

24

28

Hermann, Dölling Oberwittler et al. Kury et al. Oberwittler et al. Wilmers et al.

29

Sessar et al.

2001

30 31

Dreher et al. Müller, Schröttle

2003 2003

local: Reutlingen local: Metzingen local: Munich, Kiel, Hamburg, Hanover, Wunstorf, Lilienthal, Leipzig, Stuttgart, Schwaebisch-Gmuend local: Freiburg, Heidelberg local: Freiburg, Cologne local: Mannheim regional: Markgraefler Land local/regional: Hamburg, Hanover, Leipzig, Munich, Friesland local/supranational: Hamburg, Amsterdam, Budapest, Cracow, Vienna local: Rottweil National

32

Feltes, Goldberg

2003/04

local: Bochum

33

Gallup/Europe, Kury et al. Pfeiffer et al.

2005

national: ICVS

2005

local/regional: Munich, students Stuttgart, Schwaebisch- 10–11 y Gmuend, Kassel, 15–17 y Dortmund, Oldenburg, Lehrte, Belm, Wallenhorst, Peine, Soltau-Fallingbostel, Thuringia

25 26 27

34

1998 1999 2000 2000 2000

Publication Schnell & Kreuter 2000 Hermann & Laue 2005 Kury et al. 1999a Kury et al. 1999b Wetzels et al. 2001

general 14–70y students 13–17y general students 13–18y students 15–17y

Hermann & Dölling 2001 Oberwittler et al. 2001 Posch et al. 2001 Oberwittler et al. 2002 Wilmers et al. 2002

general

Sessar 2006

general women

students 11–17y

Dreher et al. 2005 Bundesministerium für Familie, Senioren, Frauen und Jugend 2004 Feltes & Goldberg 2006

general

EUICS Report 2005 Baier et al. 2006

Table 1: Analyzed victim and insecurity surveys

125

Victimisation and Insecurity.book Page 126 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity.book Page 127 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy Giovanni Sacchini, Rossella Selmini

I.

A general overview of research about victimisation and insecurity

For a long period, at least until the first 1990’s, the only statistical sources of information about crime in Italy were the data registered by the police or the judiciary. In this framework, the most important quantitative information about crime activities are i) crimes reported to the police by citizens, or discovered by the police through its investigation tasks (the so-called statistiche della delittuosità, which started in 1955) and ii) crimes registered by the judiciary, and defined as statistiche della criminalità, gathered since the end of 1800. These statistics are differently gathered, they come from different sources and they offer different information, but both of them are used as indicators of the wide–spreading and evolution of crime. To the well-known traditional limits of these official statistics, we have to add, in the Italian case, some peculiarities which make them less reliable than it is in other countries. Among them, deserve to be mentioned the presence, here, of five police corps, whose competencies very often overlap, the traditional backwardness of the national statistic institutions and of the whole administrative apparatus, the delay in the development of quantitative research on crime, the professional culture of the Italian police – not used to tackle with ordinary crimes and rather distant from the civil society. It is only at the beginning of the 1990’s that the first signs of interest in developing more sophisticated methodologies of research and in improving the nature and quality of statistical information about crime emerged. The explosion of crime and security issues as national problems and objects of political competition and public debate, the increasing crime rates, the pressure of local governments (which were in search of new sources of information to support their emerging local security policies), together with the modernisation of some national agencies of research – especially the National Institute of Statistics (ISTAT)1 – are the main 1.

ISTAT (National Institute of Statistics) is a public research agency acting autonomously from the Government. It gathers and produces information on the Italian economy and society, and make these studies available for the research and the decision-making process.

127

Victimisation and Insecurity.book Page 128 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

reasons for exploring new sources of information. It is in this climate of more general renewal that the first surveys – local and national – and especially the first victimisation and insecurity surveys were carried out. Very sketchy, the types of surveys carried out in Italy during the past 15 years2 are the following ones: * Surveys and opinion polls on fear and concern about crime. These investigations are both national, but mainly local (interregional, regional, and municipal). * Surveys about the living condition of households that contain sections about both victimisation and feelings of insecurity (the so called Indagini multiscopo). These investigations are nationwide, but they allow a regional disaggregation of the results. * National victimisations surveys. * Local extension of the national victimisation surveys * Thematic victimisation surveys and polls (both national and local)

1.

Surveys and opinion polls on fear and concern about crime.

Even though this kind of investigations is not the main object of analysis in the WP7, it deserves to be mentioned because it is rather wide-spread, it sometimes containes modules about victimisation, and, in some cases, it is based on well-structured questionnaires, a large sample and often a historical series is available3. We’ll mention here only these last cases, without considering the great variety of smaller opinion polls, most of them of a very low quality. The first – and the only – national survey carried out at a national level based on a rather susbstantial sample of citizens was published by DOXA in 1991, and repeated in 1999. The main issues here investigated are: * perceptions about the spread of crime in the interviewed’s neighbourhood; * perceptions about the future trend of crime phenomena; * feeling of security outside, by night; * feelings of security at home4. 2. 3. 4.

For general overviews of the limits, features and development of statistical research about crime in Italy see Bandini et al. (2003, 1-66) and Savona (2004). For a general overview of features and critical aspects, see Nardi (2003). A similar investigation has been carried out also by ISPES (Istituto di Studi Politici ed Economici) in 1992. After the transformation of ISPES in Eurispes, many other opinion polls have been carried out by this institute, often with a thematic focus, but all of them are of a very low quality and rather oriented to satisfy the needs of the press. The same for what concerns the works by Censis, a public/private institute of research.

128

Victimisation and Insecurity.book Page 129 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

At the local level, on the contrary, these investigations are rather widespread. There are many reasons for this strong relying upon polls by local governments: the difficulty in obtaining from the national government data about registered crime disaggregated at a local level; the technical and economic difficult of carrying out a local victimisations survey (see later) ; the fact that they are less expensive than other kind of research, also because there is a great supply from private polling institutions. Moreover, it is supposed that they do not require a specific professional culture to be interpreted and analysed and they offer data which can be more useful for the development of local policies than the data about reported crime and victimisation5 (or many local policy makers think so). Many of these surveys focus mainly on ‘citizens’ perceptions, opinions and feelings, but often they also include some questions about victimisation. Most of them, at least those we are going to describe later on, are based on the established distinction between “fear of crime” and “concern about crime”6. The oldest and in-depth survey on this issue is that one developed continuously by Città sicure7 since 1995 and titled Devianza, sicurezza ed opinione pubblica. The investigation is carried out at the regional level (on a sample of 1200 individuals, with an over-sampling in some cities of the region) by means of the CATI system and it focuses on: – fear of and concern about crime; – experiences of victimisation; – perceptions of urban disorder, incivilities and “soft crimes”;

5.

6.

7.

Local governments, in fact, have no constitutional competencies in crime control, but only in crime prevention, a concept that acquires a different meaning – remaining nonetheless rather ambiguous – at the level of local safety policies. Having a knowledge of the citizens’ feelings and perceptions about crime allows local governments to develop social reassurance policies, which many think more consistent with the institutional aims of these institutions. There seems to be a sort of a bifurcation : the state maintains its powers in crime control and repression, while local governments develop crime prevention and social reassurance: but in the everyday practices of the safety policies these borders are not quite clear. In the sense firstly proposed by Furstenberg (1971). See Mosconi and Toller (1998) who recalls the international literature supporting this distinction and its use in the Italian case, and also Barbagli (1998). Città sicure is the name given to a pioneering project developed in the regional government of Emilia-Romagna in 1994, which is now a well structured Department of this local administration, and whose right name is “Service for safety policies and Local Police”. The Service has administrative, political and also research tasks. Given the fact that it is better known under the common name of Città sicure, we’ll use it in this paper.

129

Victimisation and Insecurity.book Page 130 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

– attitudes and opinions towards the criminal justice system (above all punishment and the police’s work); – attitudes and opinions about immigrants; – self-protective behaviours and avoiding strategies. There are also some experiences of interregional polls carried out by a private polling institution in the Veneto and Friuli Venezia Giulia regions (the so called Sondaggi dell’Osservatorio sul Nordest, Istituto Poster, Vicenza)8, but they are limited to the analysis of the general social concern about crime and the relationships between crime, deviance and immigration. At the level of Municipalities, deserve to be mentioned, for their continuity, careful research drawing and elaboration, the experiences of the cities of Modena and Bologna9. Modena is a town of 180,000 inhabitans in Emilia–Romagna, where a survey about fear of crime, feelings of insecurity and also victimisation is carried out since 1997. The size of the sample (1,200 people) allows also a disaggregation of the results in the four main areas of the city. A similar survey is carried out – always yearly – in the city of Bologna (370,000 inhabitans). In the year 2002, in order to have a focus on the historical centre, an over-sampling was implemented (600 people residents in this area – one which is considered in the public opinion and in the media as one of the most problematic in Italy) and a focus on incivilities and urban decay was included in the series of questions.10 The reliability of these polls and their use have always been an open question in the scientific debate in Italy. Nonetheless, also those scholars who were the most critical towards these kind of investigation [and, more generally, towards the issue of fear of crime – that was supposed to be in most cases “irrational” and inconsistent with “objective” crime rates – did not hesitate to exploring their use (Guarnieri, Mosconi, 1995; Mosconi 1996; Mosconi, Toller 1998)]. The scientific assumptions leading this use are those ones of the British Left Realism, which had a strong influence, at the beginning of the 1990’s, on some of the Italian criminologists and sociologists most engaged in understanding “what could to be done about 8.

Detailed information on this private institute, which works mainly for public administration and that is considered one of the best qualified in Italy, for its scientific leadership, see http://www.fondazionenordest.net 9. Detailed accounts of these local surveys, together with the analysis of the results are in Roversi (1996) and Nobili (2004) 10. The issue of urban disorder, without having – yet? – become a thematic and separated survey, is an emerging and increasingly relevant issue inside this kind of research. For a national secondary analysis of the related data resulting from different surveys, see especially Sartori (2003).

130

Victimisation and Insecurity.book Page 131 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

crime”. Fear of crime started to be considered also by some critical criminologists as an interesting issue, and in any case a political problem which could not be disregarded neither by reasearchers nor by politicians. The main controversial issues remained anyway open: firstly the way to formulate question about fear of crime and secondly the inclusion of victimisations questions inside the polls. The limited sample used in this polls, in fact, do not allow a rigorous reconstruction of the citizens’ victimisation experiences. Usually, data about victimisations – when included in these surveys – are considered not reliable by researchers and methodologists; but sometimes politicians do not take these warnings into consideration.

2.

Surveys about the living conditions of the households

These surveys differ from the previous one because here analysis of fear and concern about crime, or of victimisation (or both) represent a section of a general investigation about living conditions and attitudes, opinions, perceptions of citizens about different aspects of their everyday life. The most structured and established of these pieces of researches is the so called Indagine multiscopo sulle famiglie, carried out yearly since 1993 by ISTAT on a sample of 50,000 citizens (belonging to 20,000 families) (ISTAT 1993; 2000). The surveys focus mainly on insecurity with reference to the immediate context of the interviewed and on her/his concrete evaluation (Arcidiacono and Sacchini, 2004, 76-77). The focus is on: – citizens’ opinion about the area where they live with regards to crime and security; – how frequently citizens see, where they live, signs of (mainly) social disorder; – what citizens think about the trend of crime, always in the area where they live. The first of these indicators is considered important in order to understand fear of crime in its concrete dimension, and in the citizens’ personal context (the neighbourhood). The second focuses once again on the citizens’ immediate context, in order to understand what their perception of the socalled “soft crimes” is. The interviewed, in fact, are called to report11 how many times they see drug addicts and drug dealers, prostitutes and their clients, and homeless or other groups who live in the street in order to go deep into the – eventual – link between fear of crime and disorder. The 11. Both through a self-filled questionnaire and a face to face interview.

131

Victimisation and Insecurity.book Page 132 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

survey includes also questions about victimisation (about three crimes: pickpocketing, bag-snatching and burglary12) and on social concern about crime. The opportunity to analyse the results on these issues crossing them with the routinary activities of the individuals is what make this survey particularly interesting for the secondary exploitation of the results (Arcidiacono, Sacchini, 2004), and for a theoretical perspective which see crime as mainly the result of circumstances, attitudes, collective behaviour and routine activities. Deserves to be mentioned also the yearly opinion polls by ISPO, a private polling agency with a national relevance that, since 1997, keeps under control the concern about crime in the context of a wider research on citizens’ attitudes and opinions on different issues of social life13. Actually, the poll investigates only the social concern about crime, by means of a very general question14 and it is surprising enough that so limited a study has become the basis of a whole reasoning and understanding of the development of feelings of insecurity in Italy (Diamanti, Bordignon, 2001).

3.

National victimisation surveys

Accordingly to the main purposes of the WP7, we’ll analyse this kind of research in a more detailed way. As we have already mentioned, the use of victimisation surveys is rather recent in Italy. At the end of the 1980’s, ISTAT carried out two victimisation surveys, based on a very small sample and not rather sophisticated (ISTAT 1993). In 1991, and also later on, an investigation about victimisation was also carried out by UNICRI, with the support of the Ministry of the Interior. It involved a sample of 2,024 people older than 16 (Savona, 1993; UNICRI, 1993; 2004), and it has been used mainly for international comparative purposes.

12. A comment on these crimes and an analysis of the results in comparison with data coming from other sources is to be found in Muratore (2003). 13. In this investigation, interviews are carried out through a “telematic panel” – at the household, by means of a PC connected via modem to a central computer – to a sample representative of the Italian population, distributed in 450 municipalities and stratified according to geographic and demographic parameters. For an overview of the historical series 1997-2001 of this survey, see Diamanti and Bordignon (2001) 14. “According to you, which are the most important issues that the Government should tackle with in this moment?”

132

Victimisation and Insecurity.book Page 133 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

In 1994, the Istituto Cattaneo15 developed a local victimisation survey in the city of Bologna (Barbagli, Pisati, 1995). In this case, the sample was of 1,614 individuals older than 18 and, and the interviews were conducted face to face. Together with the DOXA, the same institute realized a national survey about 6,291 individuals older than 15 (Barbagli, 1995). The first national victimisation survey was carried out by ISTAT between 1997 and 1998, the second five years later (2002)16. For its methodological design, the wide sample, and the range of issues investigated, it represents a valid point of reference for the study about victimisation and crime. Through an over-sampling rather substantial, the investigation has been extended also to some regional areas (see below). The general infrastructure of the survey is based mainly on the most important established victimisation surveys, especially the National Crime Victimisation Survey carried on in the U.S. from 1972, and the British Crime Survey. It is also clear what is the theoretical framework of reference of the whole survey : it is mainly devoted to analyse the interaction between authors and victims, to reconstruct the routine activities and the life–styles of individuals, and the emergence of opportunities and presence of circumstances which favour the occurrence of crime. The questionnaire has been elaborated under the supervision of Marzio Barbagli in the middle of 1990’s and it has been repeated – a part from some minor changes – also in the 2002 survey. The aims and the content of the investigation According to ISTAT (2004) the general purposes of the investigation are : * to gather statistical information in order to estimate the “real” numbers of crimes ; * to gather information about the places of crime, the time, the behaviours of the authors and of the victims, the external circumstances, etc. ; * to single out the “groups at risk” and those groups exposed to the risk of multiple or repeated victimisation ; * to give a regional account of the results, through a careful sampling design.

15. A well established research centre in Bologna, specialised in socio-political issues and funded mainly by public bodies. 16. The third has been carried out in 2007.

133

Victimisation and Insecurity.book Page 134 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

As it is common in these investigations, only some crimes are the object of the analysis : – those characterized by objective parameters (namely, when there is no ambiguity about the experience of having been a victim) ; – those whose most likely victims are individuals and households, so following the distinction between “personal crimes” and “household crimes”, already used in the USA National Crime Surveys and also in the international survey by Unicri (1993) ; – those crimes for which there is not the risk that the people interviewed are somehow involved in the crime itself17. The perception and feelings of insecurity The survey includes a section about the perceptions of crime and the feelings of insecurity, in so recognizing the complexity of the experience of crime in post-modern societies, where “objective” victimisation is always related to subjective perception. This section is also aimed at giving relevance to the issue of the fear of crime, and legitimacy to citizens’ perception, according to the development of studies in this fields (Barbagli, 1998). The issues investigated and commented on are: – fear of crime and social groups; – fear of crime and physical context; – fear of crime and its relationships with objective victimisation (the “reason” or “unreason” of fear of crime); – the relationships between fear of crime, crime and incivilities; – the relationship with the police; – the self-protection techniques and protection of the household. The feeling of insecurity has been investigated by means of the very conventional and internationally wide spread questions about fear of crime (walking alone in the neighbourhood, fear of crime at night, at home, etc.). Concerns about the place where the interviewed live has been investigated by means of questions about the presence of drug addicts and prostitutes and the perceptions of such phenomena as vandalism, graffitis, social problems (presence of poor people and homeless). The relationship with the police is rather well investigated through a series of questions about the attitude of the police corps when one goes to report crime, how great is trust in the police’s work, and how great are the expectations about police’s ability in keeping the neighbourhood under control and in coping with crime activities. 17. For the kind of crimes investigated, see the Appendix I.

134

Victimisation and Insecurity.book Page 135 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

Technical characteristics Given the delicacy of the topics, ISTAT has choosen the technique of the telephonic surveys that is considered a better guarantee for what concerns privacy, confidentiality and anonymity. The option in favour of the CATI technique has implied the definition of working modalities that could assure the quality of the data. Specific instruments to know, monitor and estimate the process of surveying have been projected and implemented. A careful attention has been paid to the professional training18 of the staff who carried out the interviews, and all of them are women (given the great relevance of the section about sexual violence and harassments, as we’ll see see later on). The drawing of the questionnaire In the drawing of the questionnaire a great attention has been paid above all to these issues: * the period of reference of the events (12 month, 3 years and the course of life); * the memory and its different way of working according to the different events (telescoping effect) ; * the problems related to the psychological mechanisms that can lead someone to forget or even remove the experience of victimisation; * the formulation of questions, in order to avoid misunderstandings about the real meaning of the question itself. Offences have been translated in common language19. The questions around crimes are rather detailed and carefully elaborated, in order to guarantee a reliable response by most of the interviewed and to gather information about features and behaviours of victims, about how, when and where they have been victimized, their lifestyle and attitudes etc, and the author, when it is known by the victim.

18. A system of permament professional training has been implemented, in order to offer the interviewers all the neccessary support, also from an emotional and psychological point of view. 19. Deserve to be mentioned here the technical choices of the screening (a series of questions in order to discover if the respondents has been victims of a series of crime) of the wording (the attention to the formulation of the questions) and of the sequence of questions, putting the most intimate questions at the end of the questionnaire.

135

Victimisation and Insecurity.book Page 136 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

The sampling A final technical question concerns the sampling techniques. ISTAT interviewed 55,000 individuals in the first survey, and 60,000 in the second one (of 14 years and more). The sample is a two stages one, with stratification of the units at the first stage. The units at the first stage are the telephone addresses and the households to whom they correspond. The units of the second stage are the individuals: for each household selected at the first stage a random individual of 14 years or more has been selected among the members of the household.

4.

The local extension of the nationwide victimisation surveys

The first national victimisation survey has been extended also to a regional area (Emilia–Romagna) through an oversampling of 9,000 individuals (that add to the already available 2,000 interviews in the Region, for a total amount of 11,000 individuals interviewed, in an area with 4 millions inhabitans) and according to an agreement between the regional government and the ISTAT (Regione Emilia–Romagna, 2000; Sacchini, 1999). In 2002, this example has been followed by other Italian regions: Tuscany, Abruzzo, Campania, Lazio20 and by a city (Bologna). Nothwistanding the over-sampling, the local extension of the survey has not been much used by local governments, for reasons that will be better explained later on. We can advance here the fact that the problem of the scarcity of the criminal events remains anyway a very crucial one, and it impedes a disaggregation of the results at a more local level (in this case, not beyond the District border), also with such a large sample. At the same time, local governments cannot finance by themselves these kinds of surveys, and they need to rely upon the ISTAT framework. The issue is still, in Italy, an open question. In the case of the city of Bologna, the oversampling has been of 1,000 interviews, that, added to those already available by the national research with reference to the city, brings the total number of the individuals interviewed at 1,707.

20. A general overview of the researches in these five regions is now available on PDF format on Internet (Cisis 2006).

136

Victimisation and Insecurity.book Page 137 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

5.

Thematic surveys

Surveys about sexual violence, harassment and gender violence This issue is the most important, in the Italian case, for what concerns thematic surveys ; dedicated research on other phenomena are not at all so well structured and wide-spread. Both the first and the second national victim surveys by ISTAT include a section devoted to sexual harassment and violence against women. Moreover, in 2005 a thematic survey about gender violence has been carried out too21. We’ll focus firstly on the two sections of the national victimisation survey that are part of the general victimisations survey, which deserves to be noted for some reasons: they represents a relevant part of the whole survey; they are carried out with a particular attention towards both the issues investigated and the attitudes of the respondents, and they have been the object of a distinct analysis (Sabbadini, 1998 ; ISTAT, 2005). The aim of the researchers, as they explain in their comments to the results, has been that one to reconstruct a phenomenon which is strongly affected by the limit of the dark number, probably much more than in other offences, in order to give to the institutions and the public more information for the policy making process and the prevention of such behaviours (Sabbadini, 1998, 4). The authors of the research, however, and even though they consider the research successful (only the 4,2% of women gave a partial reply to the questions or refused to respond), do not conceal the limits and complexity of investigating these issues by means of a survey. In the first edition, the survey involved a sample of 20,064 women, with an age that goes from 14 to 59 years, and it was carried out through the same methodology of the whole research. The behaviours here investigated are : – sexual physical harassments; – exhibitionism ; – obscene calls; – sexual blackmails at work ; – sexual violence. Verbal harassment and similar bothers have been excluded in the first edition of the survey, for the extreme subjectivity in the perception of this form of behaviour (but they have been included in the second one). Also 21. A first, detailed analysis of the results in ISTAT (2007). For the methodological aspects see the website http://www.istat.it/giustizia/sicurezza/.

137

Victimisation and Insecurity.book Page 138 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

the domestic violence and other forms of psychological abuse have not been included in the first edition (but, as already mentioned, they have been the focus of an ad hoc investigation in 2006). Consequently, the survey does not allow to estimate the total number of victims of such behaviors. The incidents have been investigated both with reference to the last three years and the course of life; the questions are formulated in order to avoid both a generic reference to the behaviours and the strict legal definition of the offence22. As is well known, the use of victimisation surveys for what concern sexual violence, rape and harassments is still an open question, and the Italian survey, even though the researchers show an awareness of these limits and they tried to overcome them, can’t avoid some of them. Gender literature on this issue is rather critic about the use of those instruments, above all because they imply a selection of behaviours and a reduction of the complexity of the phenomenon which can severely affect the reliability of the results (Johnson and Sacco, 1995, 283 ; Stanko, 1994, 123). A connected issue concerns the definition of behaviours. As we have already mentioned, ISTAT opted for a broad definition and for the use of examples, and this methodological choice may have limited this problem. Certainly enough, however, also in this case it was impossible to investigate the whole range of aggressive behaviours against women23. A further problem is related to the different perceptions – a difference which depends upon cultural, anthropological and social contexts – of harassment by women: not necessarily all women perceived of such behaviours as a trouble: someone could justify them as a common male’s behaviour, or also as an appreciation. This difference is higher for what concerns verbal harassments, and it is lower for exhibitionism and obscene calls (Sabbadini, 1998).24 22. About this point, see the Appendix II. 23. As the feminist literature suggests, in order to carry out a correct investigation of these phenomena, an ad hoc investigation would be necessary, that would be able to reconstruct the whole range of aggressive, violent and bothering behaviour that women usually experience in their everyday life. This extension would probably show that women are much more frequently victimised than what result from conventional surveys, and this would also allow to better explain the so called “paradox of fear”. (Junger, 1987, 362 ; Stanko, 1994, 130). 24. There are also also other methodologal problems related to the individual response: memory is an important factor (as it is the case for all questions in victims surveys), and finally, there is the difficulty in talking about facts which are considered as related to our more intimate life. This last problem is particularly relevant for what concerns sexual violence, given the reluctance of women in reporting this experience not only to the police.

138

Victimisation and Insecurity.book Page 139 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

In the second national survey of 2002 (ISTAT 2005) this section of the questionnaire has been repeated and enlarged to the verbal harassments and to the shadowings. Also the sample has been enlardeg to the number of 22,700 women. The questions remain similar to those of the first survey, with the exception of that one about the obscene calls, which now includes also obscene messagges received by mail or e-mail. Finally, ISTAT – with the support of the Department of Equal Opportunities – has also carried out recently, in 2006, the first national dedicated survey about gender violence on a sample of 25,000 women. This work is rather interesting under many points of view, not least the methodogical ones25. We can’t give here a detailed account of the research, we’ll just mention the main issues that have been investigated. That are the following ones: – the everyday life, leisure time, social networks, health, etc. of women; – the experience of violence by non partners, previous partners and current partners; – characteristics of current partner and partner experience of violence; – characteristics of last partner and last partner experience of violence; – characteristics of previous partner and previous partner experience of violence; – mother and partners abuse history and childhood victimisation. The research – which takes into consideration physical, sexual and psychological violence – relies upon the more advanced studies in this field and it is rather influenced by foreigner experiences and by the feminist thought about male violence and gender relationships. Other thematic surveys, on different issues, are rather fragmented, not serialized and carried out only at a local level. The issues investigated – very different among them – vary from the citizens’ opinion about immigrants, violence at school (above all bullyism); the youth’s behaviour related to crime and disorder; the mafia and organized crime26.

6.

Funding agencies

National government funds the national surveys by ISTAT, and sometimes support research by other public/private centre, but in a very small

25. For instance, the criteria for selecting interviewers (half oh them coming from shelters for battered women). 26. Most of these latter investigations are methodologically rather lacking. One of the most accurate opinion poll about percpetion of Mafia and related crime is the opinion poll by SWG (1994).

139

Victimisation and Insecurity.book Page 140 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

percentage. It never funded dedicated national opinion polls about fear of crime and unsafety. The other investigations described in these pages have been economically supported mainly by local governments, which have given a strong impulse in the advance of statistical and empirical research in Italy, or by private research centres, often partially or totally supported by local governments. University does not fund research in this field, and the university centres that carry out statistical research (such as Transcrime, see the following note n. 27) usually rely upon existing data or are financed once again by local government or the European Union.

II.

The scientific and non-scientific use of the research

1.

The scientific users

The academic context, with a few exceptions, has remained rather distant from the development of the sources of information described in the previous pages. An overview of the research programmes funded by the Ministry of the scientific research shows that no of the researches developed in these years by the Italian universities relied upon the results of these surveys. Those sociologists and criminologists most involved in the analysis of the results – or in the elaborations of the general plan of the national victmization surveys, as it is the case of Marzio Barbagli – used them in different contexts, for instance when they work as consultants of local or national governments, or for other research centres. We could say that the advance in research on crime – if we want to consider in this way the emergence of surveys – does not affected deeply the academic context. On the contrary, the world of the research centres (both public and private) has made a larger use of the surveys, above all a secondary use of the victimisations results. The main goal of this use is to go deeper into the analysis of crime, of fear of crime and perceptions of security, often with reference to some areas of the country. Through an agreement between ISTAT and Regione Emilia–Romagna, for instance, ISTAT envoys directly the data base of the national victmization survey to the reserach centre established inside Città sicure, where data are elaborated and analysed in different ways: to understand better the relationships between recorded and survey statistics, to shed light on some specific issue [for instance, the group at risk (Arcidiacono, 2004), or victimisation of women (Selmini, 1999; Selmini, Arcidiacono, 2005) ; to develop the analysis of some offences, for instance assaults, pickpocketing and burglary (Arsani 2005 ;

140

Victimisation and Insecurity.book Page 141 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

Arcidiacono, 2005a and 2005b)]; to compare different regions (CISIS 2006), for attempts to integrate these data with other coming from different sources27. A great effort has been made to “localize” the result of the national survey (Regione Emilia–Romagna, 2000; Sacchini, 1999, 2004). The same with respect to the data from opinion polls about fear of crime and feelings of insecurity, where the local research centre as Città sicure are not only users, but also producers. In this case, the regional survey about insecurity is reported each year in the Annual Report by Città sicure, and many local surveys have been carried out by many municipalities, through the regional scientific and technical support. This “localization”, as is the case for the national victims’ survey, has brought to the wide- spreading of local analysis of fear of crime and insecurity, with the purpose to develop “local diagnosis” which not always have been successfull in orienting policies and strategies of local administrators, as we’ll explain better in the following point. Concluding, the most relevant use is the public diffusion and dissemination of the data coming from this research, eventually revisited or integrated with other data, for scientific development and discussion. A further example of this use is the abovementioned volume by the Forum Italiano per la sicurezza urbana (2002) where, under the scientific supervision by Giovanni Sacchini, an accurate cross national comparison of the data about insecurity of the national survey about the living condition of households has been carried out. A similar work of accurate analysis and exploitation of data (both from security surveys and victimisation) is to be found in the seven Annual Reports by Transcrime28 about the state of the security in the Trentino Region, under the general scientific supervision by Ernesto Savona29.

27. Among these attempts, deserve to be mentioned the interesting comparison between the judicial data and those from victimisation surveys with reference to sexual violence (Terragni, 1999). 28. Transcrime is a joint research centre on transnational crime between the Università degli Studi di Trento and Università Cattolica di Milano. Directed by Ernesto Savona, professor of criminology in Milan, the centre’s funding comes from contract with local administrations, national bodies, international corporations and European Union. 29. See above all the last Report with introduction and comments by Savona (2004).

141

Victimisation and Insecurity.book Page 142 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

2.

The non–scientific users

We have to distinguish between the use of the data by national government and its agencies and the use made by local governments (Regions and Municipalities). In the first case, apparently no use has been made of research neither by the Ministry of the Interior and its departments and nor by the national police. The two National Reports about crime published by the Ministry of the Interior in the last years do not mention the surveys by ISTAT and are based only on police data and official statistics on crime. Similarly in the legal and administrative official documents of the Ministry (guidelines and similar institutional documents). So, the national government, while financing the national surveys, apparently do not rely upon these information in order to develop its national strategies for policing, crime prevention, etc. It would be interesting to understand better this gap between research and national policies about safety, even if we can reasonably suppose that the lack of culture about crime prevention and unsafety, the distance – if not the suspicion – of this world towards information coming from sources other than their “own” numbers (more generally, towards research) could, at least partially, explain it. Regional and local governments, on the contrary, have made a larger use of all the results of these surveys. In the case of those local governments which have research centres, the use has been described in the previous section, and it is a rather “mixed” one (scientific and political). It is none the less difficult to understand how the results of these surveys are at the basis of political choices and strategies, also at the local level. Sometimes this use is hard for technical reasons, as it is the case of the data coming from the “localization” of the nationwide victim surveys, for the reason already mentioned. Giving these difficulties in using surveys’ data for implementing and evaluating policies, the main use made by local governments is rather similar to that one of the scientific users: data have been analysed and discussed, reintepreted and compared, integrated and enriched through some in-depth analysis, disseminated and published. Coming back to the “political“ use of the data, the results of the national victimisation surveys have been used sometimes (or are potentially useful for ) these political purposes: – Learning more about the needs of victims, and the relationship between victimisation and fear of crime, victims’ support being one of the most relevant field of local policy; – Discover when, where and how crimes take place and against who, in order to support the policy making process about preventive policies, 142

Victimisation and Insecurity.book Page 143 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

urban planning, or the use of local police (that in Italy is a regional competence). – Learn more about the relationship between fear of crime and victimisation (policies of social reassurance). – Learn more about disorder and urban decay, and its relationships with fear of crime and victimisation. – Learn more about social groups more at risk of victimisation, in order to develop preventive and social measures. – More generally, monitoring social change and dynamics, relating these results to other social data: immigration, demography, etc. Surveys about insecurity and opinion polls have been largerly used, especially for communication purposes and political debate. Concluding, even if local governments have made a larger use of these surveys, how they are interpreted by politicians and used as a basis for implementing local policies is not at all always clear.

References ARCIDIACONO E. 2004, I gruppi a rischio di vittimizzazione, Quaderni di Città sicure, 30, 76-113. ARCIDIACONO E. 2005a, I furti nelle abitazioni, Quaderni di Città sicure, 31, 23-58. ARCIDIACONO E. 2005b, I borseggi, Quaderni di Città sicure, 31, 111- 143. ARCIDIACONO E., SACCHINI G. 2004, Le fonti statistiche, in SELMINI R.(ed.) La sicurezza urbana, Bologna, il Mulino, 65-78. ARSANI S. 2005, Le aggressioni, Quaderni di Città sicure, 31, 93-110. BANDINI T. ET AL. 2003, Criminologia. Il contributo della ricerca alla conoscenza del crimine e della reazione sociale, Milano, Giuffrè. BARBAGLI M. 1995, L’occasione e l’uomo ladro. Furti e rapine in Italia, Bologna, il Mulino. BARBAGLI M. 1998, Reati, vittime, insicurezza dei cittadini, paper presented at the Conference “La sicurezza dei cittadini”, Roma, Istat, not published. BARBAGLI M., PISATI M. 1995, Rapporto sulla situazione sociale a Bologna, Bologna, il Mulino (in particolare le pagg. 215-272). CISIS 2006, La sicurezza dei cittadini in cinque regioni italiane, Roma, aprile 2006 (http://www.cisis.it/attivita/pubblicazioni/vol004.html) DIAMANTI I., BORDIGNON F. 2001, Sicurezza e opinione pubblica in Italia, Rassegna italiana di sociologia, XLII, 1, 115-135. DOXA 1992, La percezione della delinquenza, Bollettini Doxa, 19 –20.

143

Victimisation and Insecurity.book Page 144 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

DOXA 1999, La percezione della delinquenza, Inchieste Doxa, n. 11-13, in http://www.doxa.it FORUM ITALIANO PER LA SICUREZZA URBANA – FISU 2002, Dieci anni di delittuosità e percezione della sicurezza nelle regioni italiane: 1991- 2001, Imola, Datacomp. FURSTENBERG F. 1971, Public Reaction to Crime in the Streets, American Scholars, 40, 601-610. GUARNIERI F., MOSCONI, G. 1995, Devianza, sicurezza e opinione pubblica, Quaderni di Città sicure, 2, 47 -78. ISPES (ISTITUTO DI STUDI POLITICI ED ECONOMICI) 1992, La criminalità in Italia. Prima indagine sull’atteggiamento degli italiani, research report not published. ISTAT 1993, Indagini multiscopo sulle famiglie. Anni 1987-1991, vol. 5: I fatti delittuosi: caratteristiche degli eventi e delle vittime, Roma, Istat. ISTAT 2000, Famiglie, abitazioni e sicurezza dei cittadini, Roma, Istat. ISTAT 2004, La sicurezza dei cittadini. Reati, vittime, percezione della sicurezza e sistemi di protezione, Roma, Istat. ISTAT 2005, Molestie e violenze sessuali. Indagine multiscopo sulle famiglie "Sicurezza dei cittadini" – Anno 2002, Roma, Istat. ISTAT 2007, Violenza e maltrattamenti contro le donne dentro e fuori la famiglia, Nota del 21.02.2007, in http://www.istat.it/giustizia/ JOHNSON H., SACCO V. 1995, Researching Violence Against Women: Statistics Canada National Survey, Canadian Journal of Criminology, 37, 3, 281-304. JUNGER M. 1987, Womens’ Experience of Sexual Harassment, The British Journal of Criminology, 27, 1, 358 -383. MOSCONI G. 1996, Devianza, sicurezza e opinione pubblica, Quaderni di Città sicure, 4, 57 -94. MOSCONI G., TOLLER A. 1998, Criminalità, pena e opinione pubblica. La ricerca in Europa, Dei delitti e delle pene, 1, 149-201. MURATORE M.G. 2003, Le vittime: scippi, borseggi e furti in abitazione, in BARBAGLI M. (ed.), Rapporto sulla criminalità in Italia, Bologna, Il Mulino, 455-485. NARDI L. 2003, Il senso di insicurezza, in BARBAGLI M. (ed.) Rapporto sulla criminalità in Italia, Bologna, il Mulino, 525 – 555. NOBILI G.G. 2004, Bologna: trasformazione dei fenomeni e delle politiche, Quaderni di Città sicure, 30, 279-338. REGIONE EMILIA-ROMAGNA, LUCIANI M.E., SACCHINI G.(Eds.), 2000, La sicurezza dei cittadini in Emilia-Romagna. 1997-1998, Milano, Franco Angeli.

144

Victimisation and Insecurity.book Page 145 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

ROVERSI A. 1996, Criminalità, sicurezza e opinione pubblica, in AA.VV. Lo stato della sicurezza a Modena. Primo rapporto cittadino, Modena, Comune di Modena. SABBADINI L.L. 1998, Molestie e violenze sessuali, paper presented at the national conference “La sicurezza dei cittadini”, ISTAT, Roma, not published. SACCHINI G. 1999, Città e province dell’Emilia–Romagna nell’indagine di vittimizzazione, Quaderni di Città sicure, 18, 91 – 138. SACCHINI G. 2004, Reati denunciati e percezione della sicurezza: le variazioni territoriali dagli anni ’90 ad oggi, Quaderni di Città sicure, 30, 21-75. SARTORI L. 2003, Degrado e paura per la criminalità, in BARBAGLI M. (Ed.) Rapporto sulla criminalità in Italia, Bologna, il Mulino, 489 -524. SAVONA E. U. 1993, Experiences, Fear and Attitudes of Victims of Crime in Italy, in UNICRI 1993, 93-140. SAVONA E. U. 2004, Introduzione al Settimo Rapporto sulla sicurezza nel Trentino, Transcrime, Trento. SELMINI R. 1999, Molestie e violenze sessuali in Italia e in Emilia–Romagna, Quaderni di Città sicure, 14b, 123-149. SELMINI R., ARCIDIACONO E. 2005, Le aggressioni sessuali verso le donne. Una questione dimenticata?, Quaderni di Città sicure, 31, 143 -162. STANKO E. 1994, Typical Violence, Normal Precautions: Men, Women and Interpersonal Violence in England, Wales, Scotland and in the U.S.A., in HAMNER J., MAYNARD M. (eds.) Women, Violence and Social Control, Houndmills, Mcmillan, pp. 122-134. SWG 1994, Il vissuto e la percezione del fenomeno mafioso, research report, not published, Trieste. TERRAGNI L. 1999, Le violenze sessuali, Polis, 4, 255-269. UNICRI, ALVAZZI DEL FRATE A., VAN KESTEREN J. (Eds.), 2004, Criminal Victimisation in Urban Europe: Key Findings of the 2000 International Crime Victim Surveys, Turin, UNICRI. UNICRI, ALVAZZI DEL FRATE A., ZVEKIC U., VAN DIJK J.J.M. (Eds.), 1993, Understanding crime: Experiences of crime and crime control, Rome, UNICRI publication N. 49.

145

Victimisation and Insecurity.book Page 146 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

APPENDIX 1 Crimes investigated in the victimisation survey (Source: ISTAT, national victimisation surveys 1997-1998 and 2002-2003)

1.

Crimes against the individual’s property goods

Scippo [bag snatching] Borseggio [pickpoketing] Furto di oggetti personali [Theft of personal objects] Rapina [robbery]

2.

Crimes against the household

Furto in abitazione [Burglary] Furto di oggetti esterni all’abitazione [Theft of objects from family’s dwelling itself] Appropriazione indebita [Embezzlement] Furto di veicolo [Car or other vehicles theft] Furto di parti di veicolo [Theft of vehicule’s components] Furto di oggetti dal veicolo [Theft from inside cars or other vehicles] Vandalismi contro l’abitazione [Vandalism against the house] Violenze contro gli animali della famiglia [Violence against the family’s animals] Ingresso abusivo [Trespassing]

3.

Crimes against the person

Aggressioni [Assaults] Minacce [Threats]

146

Victimisation and Insecurity.book Page 147 Thursday, December 4, 2008 12:33 PM

Research on victimisation and insecurity in Italy

APPENDIX II Questions about sexual victimisation (Source: ISTAT, national victimisation survey 1997-1998) 1) Did someone attempt to touch you, to caress you, to kiss you, bothering you against your will, for instance at the cinema, on the bus, at work, at school, at home, etc? 2) Did someone ever try to show you his sexual organs, for instance exhibithing them in the street, in the elevator or in another place, in a provocative manner? 3) Did you receive obscene calls concerning sexual matters? 4) Have you been requested for a sexual relationship, with the promise to help you in obtaining a a job? 5) Have you been requested for a sexual relationship with the promise to help you in maintaining your job or in improving your professional career? 6) Did someone try – failing – to compel you to a sexual intercourse against your will, menacing or assaulting you? 7) Did someone force or compell you to a sexual intercourse menacing or assaulting you, in a way that it was impossible to resist?

147

Victimisation and Insecurity.book Page 148 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity.book Page 149 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal Amadeu Recasens i Brunet1

I.

Introduction. A vacuum: the fact and its causes

Concerning victimisation surveys, and actually surveys in general, as well as other tools aimed at developing knowledge about safety and unsafety2, the first finding for the Iberian Peninsula is that of a vacuum. True, an effort has been made to fill that vacuum during the last decades, but we cannot certify its success. Some surveys do exist, but they are still too young, or too fragile, or too irregular to provide reliable results and conclusions.3 An overview, or an inventory, cannot be confined to commenting a list of the few existing victimisation surveys. We must also, and perhaps essentially, look at the reasons, the causes of that vacuum, and the efforts, successful or not, to fill it usefully and coherently. Victimisation surveys are one of the tools at the disposal of investigators and politicians concerned with safety issues, through which they get a better understanding of these and possibly use that knowledge to attempt to change the situation. But there are some prerequisites. To succeed with those surveys, technical skill and an appropriate survey methodology are required, first of all, along with sufficiently trained, relatively independent scientists. Next, you must also have reliable responses, which is not the case, for example, if the respondent is frightened, feels that his name may be revealed or hates the institution for which the survey is implemented. Last, and very importantly, there must be a political resolve to conduct this sort of surveys, and therefore to possess those tools.

1.

2.

3.

The author wishes to thank J. Gondra, who wrote a preliminary report for the CRIMPREV seminar held in Barcelona on March 15, 2007, for his cooperation; a bit of that work has been used here. He also thanks Josefina Castro for her critical reading of the Portuguese part and Anabel Rodriguez Basanta, who was the moving spirit of the Seminar, and to whom the present article owes a great deal. The terms safety and unsafety are used here as different but not opposites, for they are not opposed. Whereas the former reflects and is the requisite for living together, the latter indicates a crisis which, precisely, cannot become permanent, beyond a definite level of tolerance for a given society (Recasens, 2007, 131ff.). Except, as will be seen, for the Barcelona survey.

149

Victimisation and Insecurity.book Page 150 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

But all that is merely the strict minimum required for surveys to be fairly good and reliable. There must also be a will and the ability to exploit them; that is, to use the findings, to make something of them. Otherwise surveys are just documents (good, less good, or bad), used to feed the news on the day the politicians present them, and then to evaporate into thin air. Given the above, the Iberian situation is unfortunately such that too many negative elements make it impossible for us to speak seriously of a tradition and a culture of surveys, and especially of victimisation surveys. To identify the main causes of that vacuum, the peculiar situation in the Iberian peninsula must be taken into consideration, with respect to the European context of the latter third of the 20th century; it has produced a series of features at variance with those to be found in the vast majority of western European states. Broadly speaking, economic, political, cultural and scientific circumstances have prevented the serious development of research in any field whatsoever. A further impediment is the specificity of the political regimes in these countries, which has made research on safety exceptional and difficult, if not to say impossible. Both Spain and Portugal were subjected to dictatorial political regimes (respectively from 1939 to 1975 and from 1926 to 1974), which deeply marked their development in every respect, and most particularly in everything pertaining to safety, regarded as the protection of the government against its own citizens. Until the mid-1990s, the problems of the political transition at the end of those dictatorships precluded the possibility of truly addressing crime and safety from a democratic viewpoint (Agra, Quintas, Fonseca, 2001; Recasens, 2001).4 At the end of the dictatorships, during the ‘80s, Spain and Portugal had to make tremendous efforts to “set their clock to European time”, and some ingredients are to be found then which went without saying in other countries but were real and sometimes traumatic novelties for the Iberian countries at the time. Much social and political energy was absorbed in developing a democratic culture, reforming governmental structures and powers, adjusting to the arrival of new social actors and being present in international forums and institutions. The economic situation, with the advent, triumph and crisis of the welfare state in the space of a single decade, left a bitter taste mingled with some frustration.

4.

Greece, we must remember, was also prey to a dictatorship, and its evolution and safety policies are quite similar to those of Spain and Portugal (Vidali, 2001).

150

Victimisation and Insecurity.book Page 151 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

In Spain, the safety situation was marked by the lack of change within the ministry of the Interior. The (not unjustified) obsession with terrorism5 greatly disrupted safety policies and imposed continuity within the ministry of the Interior (as opposed to what transpired in the ministry of Defense).6 That altered not only the operational capacity of the police forces, but also the reliability of their data and the transparency of their organizations. This all actually suited those forces: shaped by Francoism, they had gotten through the democratic transition practically intact and retained a chusquerismo7 mentality. Unsurprisingly, then, when the Iberian dictatorships ended, the administration’s statistics (actually the only ones in existence) were not only quite unreliable but also distorted and manipulated, and victimisation surveys were completely nonexistent. The latter would necessarily have contradicted the administration’s statistics and could not be controlled by police officials, so they were “dangerous”. The first victimisation surveys date back to the 1960s and began to spread around Europe (to France, Great Britain, Sweden and Denmark) at the end of the following decade. For Spain, the first survey may be dated to 1978. However, at a time when other European countries were beginning to develop those surveys systematically, Spain and Portugal were just beginning to slowly emerge from their nightmare, without succeeding in really participating in the trend. They never did catch up. When the era of change and overture did catch onto safety matters, the idea of developing victimisation surveys as well as research on the police and safety enjoyed some expansion, but there was no space available for that, everything had to be done almost from scratch. It took time, work and scientific commitment to create the conditions and tools for research, the researchers themselves and the centers. The universities, which had largely been the strongholds of resistance to the dictatorship, were strongly prejudiced against everything pertaining to the police forces, and by extension to safety, which left the subject in the hands of the administration, dominated by the policing structures.

5.

6. 7.

The Basque nationalist cause produced ETA, and on the government side, the Grapo (Grupos revolucionarios armados primero de octubre). It was at the young democracy’s most vulnerable moments that the latter struck. The real structural reform of the ministry of the Interior did not take place until 1994, and then only with mitigated success (Lopez, 1996). The word, derived from chusco – a bread roll characteristically served in the army – refers to attitudes that are both authoritarian and ritualistic, bureaucratic and strongly corporatist, within the armed forces.

151

Victimisation and Insecurity.book Page 152 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

II.

The state of victimisation surveys in Spain

Spain presently counts some twenty-odd survey operations containing questions about safety, but which do not correspond to an explicit will to establish periodical statistical series, nor to become a part of official statistics. What may be found, relatively frequently in spite of it all, are isolated questions about Justice system and National affairs problems within opinion polls, or sometimes in quality-of-life studies. The fact is that these tools are not a part of official statistics and cannot even be viewed as opinion polls on safety and/or policing issues. What they are, conversely, are tools aimed at obtaining information on the stances and views of citizens with respect to political debate on current events, or for forthcoming elections, and therefore highly biased, ideologically. If, given the lack of material, we attempt to define a field broader than that of victimisation surveys but not unrelated to them, we may look at surveys focusing, at least, on – the study of personal experiences of victimisation, – changes of behavior or of opinion induced by victimisation, – feelings of fear and perceived levels of safety, – citizens’ opinion on the causes of offending and on desirable measures, – the evaluation of public action, and especially of police action. Now even if we take this broader object, we find that there have been practically no studies of the sort, which is really a very unfortunate finding.

1.

The national level: surveys by the CIS8

The CIS is responsible for most of the surveys or other tools that introduce isolated questions on safety. It normally has two ways of working: using opinion polls, which may be specific monographs, but are generally monthly, and research (estudios), definitely monographic. This kind of operation has been conducted relatively frequently, but never has a periodic, normalised tool been created. The Centre has also conducted a few monographic studies on one or several problems pertaining to the Justice or Interior departments, without ever making a systematic effort to construct series or ensure comparability.

8.

The CIS (Centro de Investigaciones Sociologicas) is an independent agency, a legal entity within the administration, and attached to the ministry of the Presidency by the Act 4/1990 and the Real Decreto 1526/1990. Its objective is the scientific study of Spanish society.

152

Victimisation and Insecurity.book Page 153 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

In the early 1990s, an agreement between the CIS and the Spanish ministry of the Interior attempted to improve the periodicity and comparability of survey tools. However, the project was dropped. The ministry’s priorities, with terrorism in the forefront, relegated the rest, and especially research and the development of tools, viewed as superfluous by some police sectors, to the background. Of the CIS surveys dealing with public safety-related subjects, only two were the outcome of an agreement with the ministry of the Interior (in 1991 and 1995), and one with the Guardia Civil (1998), whereas most are polls which include a few isolated questions or occasional studies on a concrete subject, with varying samples and territories, and therefore are not comparable. They include: Year 1978 1978 1980 1980 1982 1985 1987 1988 1991 1995 1995 1998

Operation Estudio sobre criminalidad Estudio sobre victimización Estudio sobre victimización y droga Estudio sobre inseguridad ciudadana 1 Estudio sobre inseguridad ciudadana 2 Barómetro Terrorismo y seguridad ciudadana Barómetro Estudio sobre inseguridad ciudadana 3 Delincuencia, seguridad y policía Demanda de seguridad y victimización Seguridad ciudadana y victimización 1

Sample (N) 1,204 5,706 5,738

Territory Spain Spain Spain

CIS Ref. 1149 1152 1206

1,156

Madrid

1251

3,714

1313

2,493 2,495

Various municipalities Spain Spain

2,495 2,490

Spain Spain

3,919

Spain

1736 1974 Agreement Min. Int. 2152

14,994

Spain, (breakdown) Spain, (territory of Guardia Civil) Spain, (breakdown) Spain Spain Spain Spain Spain Spain

2,456

1999

Barómetro January

12,994

1999 2003 2003 2003 2004 2004

Barómetro June Barómetro January Barómetro June Barómetro December Barómetro January Barómetro March

2,496 2,480 2,497 2,496 2,489 2,494

1453 1714

2200 Agreement Min. Int. 2284 Agreement Guardia Civil 2315 2364 2477 2528 2548 2554 2558

153

Victimisation and Insecurity.book Page 154 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Year 2004 2005 2005

Operation Barómetro June Barómetro March Barómetro December

Sample (N) 2,497 2,488 2,491

Territory Spain Spain Spain

CIS Ref. 2568 2597 2630

Source : J. Gondra, preliminary study for CRIMPREV

Table 1 : CIS surveys

As may be seen at first glance, only the 1978, 1980, 1991 and 1995 surveys actually resemble victimisation surveys somewhat,9 or at least systematic surveys about safety and the police, but unfortunately they differ too much to reasonably be compared. The others may be viewed as opinion polls including, since 1995 and with almost no exception, two questions on the problems about which Spaniards – and concretely, the person interviewed – are most concerned.10 In fact, the problem of the inconsistency of questions is unfortunately a constant. Whereas the CIS has retained those on safety and sense of insecurity in many of its questionnaires and in several editions of its studies since the return of democracy, those questions have regrettably been overly affected by current events and political demands. They occasionally deal with the evolution of safety over a given period, for example, whereas at other times, they ask about feeling unsafe on the streets, for instance. The question may cover one, two or three years. While the area covered is almost always Spain as a whole (and may or may not include the African cities of Ceuta and Melilla), some are confined to cities (1982) or to specific cities (1980 and 1995), or again to areas connected with some policing functions.11 The first surveys, in 1978 and 1980 (but not the 1979 one), had comparable samples and questionnaires. The main difference is that the 9.

We might add a 1979 survey aimed mostly at two large Basque Country cities : Bilbao and San Sebastian (sample – people of both sexes and over age 18 – 520 respondents for those two places and 2,000 for all of Spain, or 2,520 in all) and excluding towns of less than 20,000 inhabitants. The questionnaire was precoded and addressed victimisation, sense of security, opinion on the death penalty, measures for coping with the rise in crime and finally policing (Alvira, Rubio, 1982, p. 34). 10. For an analysis of the CIS surveys and their relation with safety issues, see Rechea, Benitez, Fernandez (2005), Alvira, Rubio (1982). 11. Organic Act 2/1986 of March 13 on the Safety Forces and Corps stipulates, rather chaotically, in its articles 11 and 12, that the Cuerpo Nacional de Policía is deployed in the provincial capitals and other cities or urban centres designated by the government, whereas the Guardia Civil is deployed on the rest of the national territory. Functionally speaking, however, the latter has exclusive competence in several matters, such as traffic outside of cities, the protection of communication routes, nature conservation, border control and policing of arms and explosives.

154

Victimisation and Insecurity.book Page 155 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

second one included a battery of questions on beliefs about drugs and drug use. In 1991, following an agreement with the ministry of the Interior, the CIS did a survey of a sample of 2,490 respondents over age 18, covering the entire country. One of the themes was “degree of victimisation over the last two years”, mixed in with questions about the respondent’s opinion on and the level of safety, knowledge about the projected bill on safety, opinion about the Guardia Civil and about drugs. In 1995, two surveys were done. The first, conducted by the CIS in April, was called “Crime, Public Safety and the Image of the Police”. It attempted to determine citizens’ attitudes and experiences with respect to crime, public safety, the police, correctional policy and youth. In the midst of a great many questions, it includes two on “the probability of being the victim of an offense” and “the offenses you suffered over the last year, did you report them, and to what agency?”, but cannot be viewed as a victimisation survey. It was nationwide, to the exclusion of the African cities of Ceuta and Melilla, with a sample of 4,000 respondents of both sexes over age 18 (and 3,919 actually completed), divided into six subsamples of 500 each for the metropolitan areas of Madrid, Barcelona, Bilbao, Seville and Valencia, and 1,500 for the rest of the country (cities of over 30,000 inhabitants). The questionnaire was administered face-to-face, in people’s homes (Medina, 2003, 6). The second 1995 survey, called “Victimisation and Demand for Safety”, implemented at the request of the ministries of Justice and the Interior, was nationwide but broken down by province, with a sample of 14,994 respondents of both sexes over age 18. Within a much larger framework including questions on opinion about the police, the causes of crime and gender violence, it contains a series of questions that make it at least partially a victimisation survey. In 1998, at the request of the Directorate General of the Guardia Civil (functionally a part of the ministry of the Interior), the CIS conducted a survey called “Public Safety and Victimisation”, nationwide except for the Basque Country, with a sample of 2,456 respondents of both sexes over age 18 and residing in cities of less than 50,000 inhabitants; in other words, what may be viewed as the territory of the Guardia Civil.12 Few questions on victimisation were included, and they differ considerably from those in the 1995 survey. The 1999 survey on “Public Safety and Victimisation” was nationwide, including the African cities of Ceuta and Melilla this time, with a sample of 12,994 respondents of both sexes over age 18. It included – within a much broader framework giving priority to questions on opinions and attitudes 12. See note 11.

155

Victimisation and Insecurity.book Page 156 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

toward some misdemeanors or minor offenses – some questions directly or indirectly concerned with victimisation. In spite of its name, it cannot be viewed as a true victimisation survey however. The rest of the surveys shown in Table 1 include isolated questions within opinion polls, but are definitely outside the scope of our analysis. Moreover, Spain has participated in the International Crime Victims Survey, with a sample of 2,041 individuals (895 urban dwellers and 1,146 in rural areas) in 1989, and a sample of 2,034 respondents (840 for the capital and 1,194 for the rest of the country), in 2005.13 Its ongoing participation in that survey has not succeeded in eliciting the enthusiasm of either administration officials (Justice or the Interior) or academics and scientists. In short, we cannot speak of a true development of victimisation surveys at the central administration level, in Spain. The few incomplete, mutually incomparable attempts do not suffice to draw any valid conclusions.

2.

The autonomous regions level

The seventeen nationalities and autonomous regions of Spain all have their own status defining autonomy, establishing their specific areas of competence and those retained by the national government, in the framework of the constitution. This is an asymmetric or “a la carte” model, then, precluding comparisons. But where victimisation surveys are concerned the question is irrelevant, since Catalonia is the only region to conduct them with any degree of regularity. In 1999, the Generalitat de Catalunya began a pilot study in view of developing an annual survey on safety and the police throughout the autonomous community. The following year, the survey, although still experimental, became a part of the Statistics Plan for Catalonia. It partly overlapped with the one conducted by the Municipality of Barcelona since 1984, covering the city and its conurbation, grouped in the Mancomunitat de Municipis de l’Area Metropolitana de Barcelona. That situation was quite untenable.14 The only acceptable solution was an agreement between institutions, so that from 2002 on, the Generalitat, the Municipality of Barcelona and the Mancomunitat, with the support of the Statistical 13. http://www.unicri.it/wwd/analysis/icvs. For the 1989 survey see Van Dijk, Mayhew, M. Killias (1990). For the 2005 one, see Van Dijk, Manchin, Van Kesteren, Hideg (2005). 14. The city of Barcelona and the Mancomunitat cover a small portion of the Catalan territory, but about half of the Catalan population of over 7 million people resides there.

156

Victimisation and Insecurity.book Page 157 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

Institute of Catalonia, drew up an agreement, renewed annually, for the completing of a single survey on a same territory using a shared questionnaire, large samples15 and modules adjusted to the needs of each institution, which makes it very costly, both financially and in terms of bureaucracy. This Enquesta de Seguretat Publica de Catalunya (ESPC) may be consulted on the Internet.16 It is still an experiment part of the Statistical Plan for Catalonia, which at least distinguishes it clearly from opinion polls.17 Traditionally, the Municipality of Barcelona wanted to include questions on the perception of safety and anti-social behavior, whereas the Generalitat, which is in the process of deploying its autonomous police force (the Mossos d’Esquadra) throughout the territory, has shown specific interest in knowing citizens’ opinion on the police forces present in Catalonia. The mixture of questions on victimisation and of questions on the perception of safety and the image of the police obliges the survey to treat some statistically infrequent variables (victimisation, direct contact with the police) and other very or fairly common ones, such as opinions about safety, the police, traffic, and so on. For that reason, some questions are put to everyone in the sample whereas others (on opinions) are randomly distributed to a fraction of the sample (about 5,000 people). However, politicians have repeatedly expressed the desire to have this survey transformed into a sort of opinion poll, seemingly less expensive, more rapid and more to the taste of some political decision-makers who are in a hurry to provide immediate responses to concrete problems rather than to develop genuine public safety policies. Over the last year, the demise of the Centre d’Estudis de Seguretat, in charge of the ESPC and designed to become a consortium bringing together the Universities and the Generalitat, left the survey in the hands of the head of the political cabinet of the Secretaria de Seguretat of the Catalan Council of the Interior. This reinforces the idea, already in circulation for other reasons, of a reversion to a sort of neochusquerismo. There is therefore no assurance that the survey will continue in its present form – which is, moreover, cumbersome, costly, difficult and too frequent.18

15. 12,806 in 2000; 12,617 in 2001; 18,679 in 2002; 17,856 in 2003; 18,105 in 2004; 13,749 in 2005; 15,884 in 2006 and 13,992 in 2007. 16. http://www.gencat.net/interior/docs/int espc06.htm 17. For the way the ESPC was created, see Bas and Salarich (2002) and Bas and GarciaValdecasas (2003). 18. At the request of the Centre d’Estudis de Seguretat, Professor Philippe Robert did an audit (Audit sur les enquêtes de victimation en Catalogne), pointing out the survey’s main problems.

157

Victimisation and Insecurity.book Page 158 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Catalonia participated twice in the International Crime Victims Survey. In 1996, with a sample of over 6,000 respondents and in 2000 with a sample of 4,000, including 2,909 valid responses obtained by telephone without computer assistance.19

3.

The local level

The first local victimisation survey dates back to 1984. It was done by the Municipality of Barcelona, in an innovative context, when the municipality created a Consell de Seguretat Urbana with the intention of developing prevention policies. The city’s transformation, with a focus on the tourism industry, and its new immigration problems, relegated those well-meaning intentions somewhat into oblivion, and since 2000 situational prevention and administrative criminology have made something of a comeback, expressed by some highly repressive ordinances on safety. The Barcelona survey did resist, however. It continues, and is still conducted annually.20 In 1989 it was extended to encompass twenty-seven municipalities in the metropolitan conurbation; it is part of the ESPC since 2002. Other cities have followed Barcelona’s example, but have not succeeded in maintaining either the periodicity or the homogeneity. The case of Malaga is worth mentioning, however, for the Instituto Interuniversitario de Criminología (IAIC) implemented a victimisation survey in that city in 1994, under the auspices of the Crime Observatory for Andalusia, and repeated it in 2005, with a questionnaire and methodology they consider comparable, thus enabling its initiators to do a decade-long longitudinal study.21 The city of Malaga has also participated in the International Crime Victims Survey, in 1993 (with a 1,634-person sample) and 1994 (with 1,505 respondents). Other surveys have been done on a smaller scale in Cordoba, Huelva and Seville. Most of these were on the model of the International Crime Victims Survey.

19. The 1996 report was published by the Generalitat de Catalunya (Luque, 1999). For the 2000 study, see Van Kesteren, Mayhew and Nieuwbeerta (2000) and Alvazzi del Frate and Van Kesteren (2004). Luque (2001) also did research on the findings of that same survey. 20. For a discussion of the Barcelona survey and its development, see A. Alabart, J. M. Aragay and J. Sabaté (1991); J. Sabaté (2005); J. M. Lahosa (2002). 21. The results of these two surveys have been published: Díez Ripollés, Cerezo, Gonzalez-Torre, Stangeland (1996) and E. García España, F. Pérez Jiménez and M. J. Benítez Jiménez (2006).

158

Victimisation and Insecurity.book Page 159 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

4.

Thematic surveys

Several thematic surveys containing information on victimisation have been published in recent years. The major ones include: – Surveys on school violence, by the Defensor del Pueblo (the Spanish ombudsman). The first edition, done in cooperation with UNICEF, is dated 1999, and entitled Violencia escolar: el maltrato entre iguales en la Educacion secundaria obligatoria (ESO). This study was an attempt to get a better understanding of peer-group bullying in secondary schools. Another survey was done during the 2005-2006 school year, in 300 public and private secondary schools. A comparative study has been made of the results of the two surveys.22 – Surveys on sociability in schools and safety in Catalonia. There were two of these, bearing the same name. The first was conducted in 20002001, and was originally called the Survey on Youth and Safety. It covered secondary schools (approximately ages 12 to 18), with a sample of 7,416 students.23 The second, for the 2005-2006 school year, dealt with both elementary and secondary schools (that is, ages 8 to 18), with a sample of 10,414 students.24 Given the way they were designed, a periodic, comparable series may, hopefully, be obtained. These two surveys are the outcome of cooperation between the departments of the Interior and of Education of the Catalan Generalitat, with the support of the Statistical Institute of Catalonia. The second was almost entirely produced by the Centre d’Estudis de Seguretat. Two (computer-administered) questionnaires were prepared for this study: a shorter one about life inside their school for elementary school pupils, and a more complex one including problems outside the establishment for secondary school students. – A survey on tourism and security. This is a victimisation survey conducted in the autonomous region of Andalusia during the summer of 2002, addressed at visiting tourists, in order to analyze their exposure to the risk of victimisation.25 Meanwhile, the most considerable effort in surveying specific victimisation situations has been devoted to violence against women: 22. It may be consulted at http://www.defensordelpueblo.es/informes2.asp 23. Joventut i seguretat a Catalunya. http://www.gencat.net/interior/docs/text_integre.pdf 24. Enquesta de convivència escolar i seguretat a Catalunya Curs 2006-2006. http:// www.gencat.net/interior/departament/publicacions/estudis/ resum_informe_ECESC_2005-2006_cat.pdf 25. See Aebi, Mapelli Caffarena (2003). A previous study of criminality in tourist areas, enumerating victimisation surveys, may be found in Stangeland, Díez Ripollés, Duran (1998).

159

Victimisation and Insecurity.book Page 160 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Roughly, there have been three broad surveys aimed at quantifying violence occurring in their homes, for women aged 18 or over, residing in Spain. The idea is to determine the correlation between abuse and women’s sociodemographic characteristics. – The first, in 1999, was part of the First Plan of Action against Domestic Violence led by the Instituto de la Mujer. It studied a sample of 20,552 Spanish women over age 18, and included the cities of Ceuta and Melilla.26 – At the end of the Plan, in 2002, a second survey was conducted, with the same characteristics and a sample of 20,652 respondents. – The third, in 2006, attempted to measure the effects of the December 20, 2004 Organic Act on provisions for complete protection against gender violence. The sample was 32,426 women aged 18 or over, residing in Spain, including Ceuta and Melilla.27 Importantly, the Instituto de la Mujer emphasizes the fact that data on the periods from 1983 to 1996, from 1997 to 2001 and from 2002 on are not mutually comparable, since different criteria were used.28

III. Victimisation surveys in Portugal 1.

The national level. Surveys by the Gabinete de Estudos e Planeamiento do Ministerio da Justiça (GEPMJ)

As was the case in Spain, the Portuguese ministry of Justice, and more specifically the GEPMJ, did not begin to take interest in victimisation surveys until the early 1990s. It took off very cautiously: the first edition of the survey, in 1989, only covered the metropolitan Lisbon area, with its sixteen municipalities. That first attempt, with questions on personal and household victimisation, reporting, perception of crime and opinion on the police, also aimed at comparing the survey findings with the information and data provided by official institutional statistics.29 The second, also implemented by the GEPMJ, took place in 1992. With a sample of 11,363 individuals, it covered all of continental Portugal (exclusive of the Azores and Madeira), with a different questionnaire in 26. Macroencuesta sobre la violencia contra las mujeres en el ámbito doméstico. http:// www.mtas.es/mujer/mujeres/estud_inves/estud_anterior99.htm 27. Macroencuesta sobre la violencia contra las mujere. http://www.mtas.es/mujer/mujeres/ estud_inves/index.htm 28. http://www.mtas.es/mujer/mujeres/cifras/violencia/denuncias.htm 29. It was published by the ministry of Justice (GEPMJ, 1991).

160

Victimisation and Insecurity.book Page 161 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

that it also contained questions on support received by victims (from institutions, family, friends or other), fear of crime, the kind of punishment to be required for certain crimes and police attitudes toward the public.30 Only in 1994 was the survey extended to the entire Portuguese territory, including the autonomous regions of Madeira and the Azores, with a sample of 13,493 respondents. The survey, also implemented by the GEPMJ, used the same questionnaire as the previous one, so that findings could be compared.31 Portugal also participated in two editions of the International Crime Victims Survey. In 2000, with a sample of 3,569 respondents, including 2,000 valid responses obtained by CATI (Van Kesteren, Mayhew, Nieuwbeerta, 2000) and in 2005, with a sample of 2,011 respondents including 801 for the capital and 1,210 for the rest of the country (Van Dijk, Manchin, Van Kesteren, Hideg, 2005).

2.

The local level

At the local level, some activity is to be found in the two main Portuguese cities, Lisbon (Almeida, 1998) and Porto (Negreiros, 1999). Actually, the first survey, in 1989, may be viewed as a local one, since it did not cover much of the country, but the fact is that it was implemented by the ministry of Justice, and much of the Portuguese population resides in the Lisboan conurbation. In 2002, a project called Cibele, cofinanced by the European Commission’s Hippocrates program and conducted in partnership with several Lisboan agencies, the European Forum for Urban Safety, the city of Lübeck and Northumbria, produced two studies, where Portugal was concerned. The first pertained to the victimisation situation, the perception of crime and sense of security and insecurity in the metropolitan region of Lisbon, on the basis of a survey of respondents over age 15 in 1,190 randomly selected households. The second considered victimisation through the lenses of community agencies, prevention forums, victims’ support and other services.32 The city of Porto also has a long tradition of surveys. Interest in surveys of victimisation and self-reported delinquency took shape as early as the 1980s, with the University of Porto’s Center for the Study of Deviant Behavior (Faculty of Psychology) and from 1996 on, with the Permanent 30. Published by the ministry of the Justice (Almeida, 1993). 31. Published by the ministry of the Justice (Almeida and Alâo, 1995). 32. http://www.apav.pt/pdf/cibele_portugués.pdf An English-language available on the same site.

version

161

is

Victimisation and Insecurity.book Page 162 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

Security Observatory, an offshoot of the former. Between 1996 and 1998, that agency adopted the ICVS questionnaire for surveys on victimisation, and began its implementation in 1998 (Agra, Teixeira, Fonseca, Quintas [1999]; Negreiros [1999]; Agra, Quintas, Fonseca [2001]).

3.

Thematic surveys

In 1995, a sample of 1,000 women were questioned in a survey of violence suffered by women. It analyses the relations between psychological, sexual and physical violence (Lourenço, Lisboa, Pais [1997]).

IV. Technical characteristics In Spain, samples and techniques are variable. Some surveys are performed face-to-face, in homes, others (the largest ones) by telephone. The size of the samples, sometimes very small, raises problems for the confidence intervals, which may be wide open. The macrosurveys on violence against women are relatively coherent. The samples do not differ too much, except for the fact that the first survey is limited to Spanish women whereas the 2006 one covered women residing in Spain. The distribution of the sample is the same for the two, with a set figure of 300 respondents per autonomous community and the rest proportioned to the population, with the possibility, for the communities, to boost the sample, at a cost. Both used the CATI system, as well. For Catalonia, the ESPC has experienced several changes. Between 1984 and 1991 the sample was constructed via route sampling; starting in 1992, telephone data bases, a less expensive method, were used. In 2000 the CATI system was introduced. The survey therefore presently works with a randomly selected sample based on telephone data bases (for households), among which the individual to be questioned is selected, randomly as well, among the resident members over age 16. Interviews are conducted on landlines, which is increasingly problematic owing to the growing use of cell phones, sometimes to the exclusion of landlines. Concerning the sample, its size raises financial difficulties which will have to be solved either by a cutback in the sample or by changing the periodicity (every two years for example), but without losing the present confidence interval of less than 5% for cities of over 50,000 inhabitants.33 33. Cities of less than 50,000 inhabitants have the option of increasing the sample and paying extra. That possibility does not seem to be very attractive, since only four townships have subscribed to this formula so far.

162

Victimisation and Insecurity.book Page 163 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

There is a problem for non-European Union persons and for unregistered EU residents, which is not a minor one, given that Catalonia has a high rate of immigration, a great deal of tourism and many secondary homes belonging to EU citizens. School surveys are conducted within the premises, with the help of a computer and under the supervision of researchers. The national-level Portuguese surveys cover the resident population aged 16 or over. Questionnaires are mostly administered face-to-face by investigators habitually working for the National Institute of Statistics of Portugal, using a laptop computer, and cover very large samples (11,363 respondents in 1992; 13,493 in 1994). Broadly speaking, Portuguese and Spanish surveys may be said to suffer from the same ills and the same technical shortcomings, each on its own scale. In both countries the scientific and technical piloting is usually in the hands of institutions depending on the administration. These may set up research centers, more or less under their control depending on the period and political circumstances. Independent and academic research is scarce, probably because of corporatism and the secretive propensity of institutions in this sector, but further reinforced, in Spain and Portugal, by circumstances described above, relative to the transition from dictatorship to democracy.

V.

Conclusions

Fragmentation, superficiality and irregularity: that is how Castro and Agra (2007, 111) define the state of Portuguese research, and the expression may just as well be applied to the Iberian Peninsula as a whole. Victimisation surveys are nonexistent or poorly developed, as a rule. This desertion has several causes. The most important is the tremendous historical vacuum produced by the dictatorships. The transition to democracy left the police forces and their corporatist ideologies practically intact. This produced serious partitioning, and a corporatism that both refuses any scientific contribution and holds it in contempt. This chusquerismo culture seems to be making a comeback in the form of neochusquerismo, with very specific criminological and political models (situational prevention, left and right-wing neoconservatism), at least in Spain, following a promising parenthesis in the late 1980s and the first half of the 1990s, at the different levels of government. The lack of an academic research tradition, the refusal, by academics, to study some subjects – an aftermath of the dictatorships – makes things worse. In countries with a research tradition, strong research centers and trained 163

Victimisation and Insecurity.book Page 164 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

researchers, that trend, now apparently prevailing throughout Europe, may be mitigated. This is not the case in Spain and Portugal. The window opened during the second half of the 1980s and much of the 1990s, when relatively high-quality research could develop at the national and autonomous government levels and in some universities (especially in Catalonia and Andalusia), seems to be closing gradually. Moreover, there is an obvious lack of political interest in victimisation surveys, except as media hype and for political visibility, which creates a vicious circle. The lack of surveys does not create a perception that they are needed, whereas the lack of interest makes those surveys seem too expensive and useless.34 Surveys are practically not put to any scientific use, for lack of specialized centers and researchers. The non-scientific uses greatly depend on the will and ability to develop public safety policies. But that tradition is lacking in Spain and Portugal: even when the tools do exist, they are merely used for political purposes such as presentation to the media, presentation, in great pomp, to parliaments (central or regional) or, in the worst cases, for directly electoral purposes. That is why politicians tend to confuse surveys with opinion polls, and they definitely prefer the latter. There is little difference between the various levels of government in that respect. Technically speaking, the lack of a tradition and culture of research, along with the meager, biased interest of the political class, takes its toll, resulting in considerable oscillations in formulations, sampling, methods of implementation and confidence intervals. This prevents comparisons, the establishment of reliable series, and the elaboration of serious, valid reports. These technical difficulties undermine credibility and confidence in both the instruments and the methods, thus further removing them from the sphere of political options, in a sort of vicious circle which is apparently not at the breaking point for the time being.

References AGRA C., CASTRO J. 2007, Las violencias, las fiestas y sus actores. Un estudio empírico en Oporto in RECASENS A. (Ed.) La violencia entre jóvenes en espacios de ocio nocturno. Un estudio comparativo europeo, Atelier, Barcelona, 107-130. 34. One politician described the situation perfectly, in a conversation with the author of this paper. At a presentation of a survey, he told me that it was all very interesting, but such a pity that the study covered the previous year, so that it wasn’t of any use in politics, which functions on a day-to-day basis.

164

Victimisation and Insecurity.book Page 165 Thursday, December 4, 2008 12:33 PM

Victimisation surveys in Spain and Portugal

AEBI M.F., MAPELLI B. 2003, Turismo y seguridad en Andalucía, Consejería de Turismo y Deporte, Sevilla. AGRA C., QUINTAS, J. FONSECA E. 2001, De la sécurité démocratique à la démocratie de sécurité: le cas portugais”, Déviance et Societé, Les politiques de sécurité et de prévention en Europe, 25, 4, 499-513. AGRA C., TEIXEIRA J., FONSECA E., QUINTAS J. 1999, Crime e Delinqûencia auto-revelada, Observatorio Permanente de Segurança, Porto. ALABART A., ARAGAY J.M, SABATÉ J. 1991, Encuesta de victimización en área metropolitana de Barcelona (1990) Prevenció: Quaderns d’estudis i documentació, 7, 5-55. ALMEIDA M.R.C. 1993, Inquérito de vitimaçâo 1992, Gabinete de Estudos e Planeamiento do Ministerio da Justiça, Lisboa. ALMEIDA M.R.C. 1998, Vitimaçâo e insegurança no concelho de Lisboa, Gabinete de Estudos e Planeamiento do Ministerio da Justiça, Lisboa. ALMEIDA M.R.C., ALÂO A.P. 1995, Inquérito de vitimaçâo 1994, Gabinete de Estudos e Planeamiento do Ministerio da Justiça, Lisboa. ALVAZZI DEL FRATE, A., VAN KESTEREN J. 2004, Criminal victimisation in urban Europe, key findings of the 2000 International Crime Victim Surveys, UNICRI, Torino. ALVIRA F., RUBIO M. 1982, Victimización e inseguridad : la perspectiva de las encuestas de victimización en España, Revista Española de Investigaciones Sociológicas, 18, 29-50. BAS A., GARCÍA-VALDECASAS E. 2003, La enquesta de seguretat pública de Catalunya, VIII Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Panamá, 28-31 oct. 2003. BAS A., SALARICH A. 2002, Encuestas de Seguridad Pública de Catalunya, Revista Fuentes Estadísticas, 61, janvier, 17-19. DÍEZ RIPOLLÉS J.L., CEREZO A.I., GONZALEZ-TORRE F., STANGELAND P. 1996, Delincuencia y víctimas, Tirant lo Blanch, Valencia. GARCÍA ESPAÑA E., PÉREZ JIMÉNEZ F., BENÍTEZ JIMÉNEZ M.J. 2006, La delincuencia según la víctimas. Un enfoque integrado a partir de una encuesta de victimización, Instituto Andaluz Interuniversitario de Criminología et Fundación El Monte, Málaga. GEPMJ 1991, Inquérito de vitimaçâo 1990, Gabinete de Estudos e Planeamiento do Ministerio da Justiça, Lisboa. LAHOSA J. 2002, L’integrazione delle politiche locali: il caso di Barcellona, Dei delitti e delle pene, monographique Governare la sicurezza: attori, politiche e istituzioni in Europa, 1,2,3, 101-110. LÓPEZ F. 1996, En el laberinto. Diario de Interior 1994-1996, Plaza & Janés, Barcelona.

165

Victimisation and Insecurity.book Page 166 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

LOURENÇO N.; LISBOA M., PAIS E. 1997, Violênça contra as mulheres, Comissâo para a Igualdade e para os Direitos das Mulheres, Cadernos condiçâo feminina, 48, Lisboa. LUQUE E. 1999, Les victimes del delicte, Centre d’Estudis Jurídics i Formació Especialitzada, Generalitat de Catalunya. LUQUE E. 2001, La victimització a Catalunya l’any 1999. Recherche pour le Departament de Justicia de la Generalitat de Catalunya http:// www.gencat.net/justicia/temes/cejfe/recerca/cateleg/cronologic/2001/ sc106001/ MEDINA J. 2003, Inseguridad ciudadana, miedo al delito y policía en España, Revista Electrónica de Ciencia Penal y Criminología 05-03. http:/ /criminet.ugr.es/recpc NEGREIROS J. 1999, Vitimaçâo ciminal na cidade do Porto, Observatorio Permanente de Segurança, Vol.III, polycopié. RECASENS A. 2001, Politiques de sécurité et de prévention dans l’Espagne des années 1990, Déviance et Société, Les politiques de sécurité et de prévention en Europe, 25, 4, 479-497. RECASENS A. 2007, La seguridad y sus políticas, Atelier, Barcelona. RECHEA, C.; BENITEZ M.J., FERNANDEZ E. 2005, Evolucion de la seguridad ciudadana. Una valoracion de las encuestas del CIS http://www.uned.es/ investigacion/IUISI pub licaciones.Htm#septiembre04. SABATÉ J. 2005, L’enquesta de victimització de Barcelona i de l’Àrea Metropolitana, vint-i-dos anys, Institut d’Estudis Regionals i Metropolitans de Barcelona. STANGELAND P., DÍEZ RIPOLLÉS J.L., DURAN M.A. 1998, El blanco más fácil: la delincuencia en zonas turísticas, Tirant lo Blanch, Valencia. VAN DIJK J., MAYHEW P., KILLIAS M. 1990, Experiences of crime across the findings from the 1989 International Crime Survey, Kluwer, Deventer. VAN DIJK J., MANCHIN R., VAN KESTEREN J., NEVALA S., HIDEG G. 2005, The burden of Crime in the EU. Research Report: A Comparative Analysis of the European Survey of Crime and Safety (EU ICS) 2005 » http:// www.unicri.it/wwd/analysis/icvs/pdf_files/EUICS%20%20The%20Burden%20of%20Crime%20in%20the%20EU.pdf VAN KESTEREN J.N., MAYHEW P., NIEUWBEERTA P. 2000, Criminal Victimisation in Seventeen Industrialised Countries: Key-findings from the 2000 International Crime Victims Survey. Ministry of Justice, Den Haag. VIDALI S., 2001, “Politique de prévention et de sécurité en Grèce: le contrôle politique, passé et présent”, Déviance et Société, Les politiques de sécurité et de prévention en Europe, 25, 4, 515-531.

166

Victimisation and Insecurity.book Page 167 Thursday, December 4, 2008 12:33 PM

On the authors

Mike Hough is Director of the Institute for Criminal Policy Research in the School of Law at King’s College London. ICPR has a staff of 15, carrying out policy research for government, criminal justice agencies and for independent funders. Mike Hough has extensive experience in crime measurement and quantitative research methods, especially large-scale sample surveys such as the British Crime Survey. He has published extensively on a range of criminological topics including crime measurement, policing, crime prevention and community safety, anti-social behaviour, probation and drugs. His most recent book, Surveying Crime in the 21st Century’ (edited with Mike Maxfield) was published in 2007 by Willan Publishing. * Joachim Obergfell-Fuchs is a Psychologist, Head of the Criminological Service in the Baden-Wuerttemberg Department of Corrections. His main research topics are evaluation within the correctional system, sex offenders, sanctioning, crime prevention, forensics. His main publications are: Ansätze und Strategien kommunaler Kriminalprävention – Begleitforschung im Pilotprojekt Kommunale Kriminalprävention in Baden-Württemberg anhand der Stadt Freiburg im Breisgau, Freiburg, 2001, edition iuscrim. Nationale und internationale Entwicklungen in der Kriminologie. Festschrift für Helmut Kury zum 65. Geburtstag, Frankfurt/Main, Verlag für Polizeiwissenschaft 2006 (with M. Brandenstein as co-editor). Sexualkriminalität in: H. J. Schneider (Hrsg.), Internationales Handbuch der Kriminologie. Band 1: Grundlagen der Kriminologie (S. 613-666). Berlin, De Gruyter Recht, 2007 (with H. Kury). * Lieven Pauwels has a PhD in Criminology from Ghent University. He teaches courses on causes of offending, quantitative methodology and data analysis at the Department of Criminal Law and Criminology, Ghent University. He is senior researcher in the Research Group Social Analysis of Security. He has published on the social ecology of crime, causes of adolescent offending, the study of measurement error in surveys, and quantitative comparative research. 167

Victimisation and Insecurity.book Page 168 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

He has recently published: How Serious Is The Problem Of Item Nonresponse in Scale Constructs of Delinquency and Key Social Mechanisms? A Cross-National Inquiry of two Classroom PAPI-Self Report Studies in Antwerp And Halmstad, European Journal of Criminology 2008, 5 (3), in press (with R. Svensson). Geweld in groepsverband onder Antwerpse Jongeren: de rol van schoolcontext en leefstijl, Tijdschrift voor Criminologie, 2008, 50, 1, 3-16 * Stefaan Pleysier (drs.) is a lecturer in Research Methodology and Statistics, and senior researcher at the ‘Research Centre in Security, Safety and Society’ (KATHO University College) and the ‘Leuven Institute of Criminology’ (Katholieke Universiteit Leuven). He is a member of the Advisory Board and the Technical Working Group of the Belgian Safety Monitor, and has published on measurement issues and quality assessment in large surveys as the Safety Monitor He has recently published: Assessing cross-cultural validity of fear of crime measures through comparisons between linguistic communities in Belgium European Journal of Criminology 2005, 2, 2, 139-159 (with L. Pauwels). Cross-cultural Invariance and Gender Bias when Measuring ‘Fear of Crime’. International Review of Victimology, 2004, 10, 3, 245-260 (with G. Vervaeke and J. Goethals). * Amadeu Recasens I Brunet is an associate professor at the University of Barcelona. He also teaches at the Universities of Bologna, Modena-RegioEmilia and Porto. An associate researcher at the CESDIP and a member of the GERN, he is working on security, police, public policies and punishment. He has recently published La violencia entre jovenes en espacios de ocio nocturno; un studio compativo europeo, Barcelona, Atelier, 2007 La seguridad y sus politicas, Barcelona, Atelier, 2007. * Philippe Robert is a sociologist, directeur de recherches émérite at the Centre national de la recherche scientifique (CNRS). He is the founder and former head of the Centre de recherches sociologiques sur le droit et les institutions pénales (CESDIP), the main French research centre on crime. He has also founded the Groupe européen de recherches sur les normativités (GERN), an European scientific network. Within the Coordination Action CRIMPREV, 168

Victimisation and Insecurity.book Page 169 Thursday, December 4, 2008 12:33 PM

On the authors

he coordinated the workshop of which this volume is one of the outcomes. His current fields of research are criminal sociology, measuring crime, victimisation and insecurity surveys. Among his recent publications, Sociologia do crime, Petropolis, Vozes (www.editoravozes.com.br), 2007 Sociologie du crime, Paris, La Découverte (www.editionsladecouverte.fr), Coll. Repères, 2005. Bürger, Kriminalität und Staat, Wiesbaden, VS Verlag für Sozialwissenschaften/GmbH (www.vs-verlag.de), 2005. El ciudadano, el delito y el Estado, Barcelona, Atelier (www.atelierlibros.es), 2003. L’insécurité en France, Paris, La Découverte (www.editionsladecouverte.fr), Coll. Repères, 2002. O cidadão, o crime e o estado, Lisboa, Editorial Notícias (www.editorialnoticias.pt), 2002. * Giovanni Sacchini, sociologist, is responsible of the statistical activities in the Service of Urban Safety at the Emilia-Romagna Region and teaches Elements of Criminology at the University of Ferrara. His main works are in the field of quantitative analysis on sense of insecurity and crime diffusion (using both official statistics and victimisation surveys). Among his recent publications: Motivi che aiutano a violare le norme. Alcuni suggerimenti ricavati da unindagine tra gli studenti bolognesi, "Metronomie", 2007, XIV, 34-35. La sicurezza nelle opinioni dei cittadini, Quaderni di Città sicure, 2006, XII, 32, 29-56. Landamento della criminalità in Italia, in R. Selmini, ed., La sicurezza urbana, Bologna, il Mulino, 2004, 199-215. * Rossella Selmini is Professor of criminology at the University of Macerata and Director of the Service for local safety policies and local police of Regione Emilia-Romagna.She has recently published La sicurezza urbana, Bologna, Il Mulino, 2004 Towards Città Sicure? Political action and institutional conflict in contemporary preventive and safety policies in Italy, Theoretical Criminology, 2005, 9, 3, 307-323. * Renée Zauberman is a senior researcher at the Centre de recherches sociologiques sur le droit et les institutions pénales, a research centre affiliated 169

Victimisation and Insecurity.book Page 170 Thursday, December 4, 2008 12:33 PM

Victimisation and Insecurity in Europe

to the French Centre National de la Recherche scientifique. Her interest lies is norms and deviance, especially criminal justice and crime. While she has worked on the French Gendarmerie, her current focus is on victimisation surveys. She has published : Enquêtes de victimation et statistiques de police : les difficultés d’une comparaison, Déviance et Société, 2004, 3, 285-316 (with H. Lagrange, M.L. Pottier and Ph. Robert) Police, Minorities and the French Republican Ideal Criminology, 2003, 41, 4, 1065-1100 (with R. Lévy). La répression des infractions routières, le gendarme comme juge, Sociologie du travail, 1998, 40, 1, 43-64.

170