crime trends in the eu

37
CHRIS LEWIS, GORDON BARCLAY, BRUNO AUBUSSON DE CAVARLAY, MARIA JO ˜ AO MORGADO COSTA AND PAUL SMIT CRIME TRENDS IN THE EU ABSTRACT. The paper’s aim is to show to EU policy makers, academics, journalists and the general public what the available information tells us about crime levels, trends in crime and public opinion about crime among Member States. The paper centres on an analysis of current trends on crime levels and trends based on the data available both from victimisation surveys and police statistics. The victimisation survey source is the published data collected in the International Crime Victimisation Survey. A separate analysis based on the Eurobarometer was also carried out. Data on police statistics present two separate sources i.e. the Council of Europe Sourcebook and the crime data published annually by the UK Home Office. These two sources both add considerable value to the raw police statistics by their choice of data, their commentary and their technical explanations and definitions. The paper compares data on three crime types (robbery, domestic burglary and theft of a motor vehicle) across the 15 Member States of the European Union (as in 2003). These three types were selected in line with the priorities of the EU Commission and as types of crime that are a major concern for EU-citizens. The paper has been modified from a report produced by the European Crime prevention network for the EU Directorate of Justice and Home Affairs with the permission of the EU. The members of the network are listed in the appendix. KEY WORDS: domestic burglary, European Union, international comparisons, motor vehicle theft, robbery, statistics, victimisation INTRODUCTION:TWO PERSPECTIVES ON CRIME There are two methods of measuring crime: through victimisation surveys (e.g. the British Crime Survey) and through the measurement of crime as reported to and recorded by the police. Victimisation rates give an accurate picture of two distinct measures: the likelihood of being a victim and the fear of being a victim. However, such surveys only cover household crimes, exclude victimless crimes e.g. drugs offences and also crimes against businesses, and exclude crimes against children. Moreover, victimisation surveys are very expensive: few European coun- tries carry out their own victimisation surveys on a regular basis and the methodologies used in each country differ. The only major international comparative survey is the International Crime Victimisation survey (ICVS) which, as shown in this report, provides a basis for comparisons of crime through the use of standardised survey questions. However, the ICVS does not currently cover all EU countries and its sample size is 2000 per country. European Journal on Criminal Policy and Research 10: 187–223, 2004. C 2004 Kluwer Academic Publishers. Printed in the Netherlands.

Upload: chris-lewis

Post on 14-Jul-2016

224 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Crime Trends in the EU

CHRIS LEWIS, GORDON BARCLAY, BRUNO AUBUSSON DE CAVARLAY,MARIA JOAO MORGADO COSTA AND PAUL SMIT

CRIME TRENDS IN THE EU

ABSTRACT. The paper’s aim is to show to EU policy makers, academics, journalists andthe general public what the available information tells us about crime levels, trends incrime and public opinion about crime among Member States. The paper centres on ananalysis of current trends on crime levels and trends based on the data available both fromvictimisation surveys and police statistics. The victimisation survey source is the publisheddata collected in the International Crime Victimisation Survey. A separate analysis basedon the Eurobarometer was also carried out. Data on police statistics present two separatesources i.e. the Council of Europe Sourcebook and the crime data published annually by theUK Home Office. These two sources both add considerable value to the raw police statisticsby their choice of data, their commentary and their technical explanations and definitions.The paper compares data on three crime types (robbery, domestic burglary and theft of amotor vehicle) across the 15 Member States of the European Union (as in 2003). Thesethree types were selected in line with the priorities of the EU Commission and as types ofcrime that are a major concern for EU-citizens. The paper has been modified from a reportproduced by the European Crime prevention network for the EU Directorate of Justice andHome Affairs with the permission of the EU. The members of the network are listed in theappendix.

KEY WORDS: domestic burglary, European Union, international comparisons, motorvehicle theft, robbery, statistics, victimisation

INTRODUCTION: TWO PERSPECTIVES ON CRIME

There are two methods of measuring crime: through victimisation surveys(e.g. the British Crime Survey) and through the measurement of crime asreported to and recorded by the police.

Victimisation rates give an accurate picture of two distinct measures: thelikelihood of being a victim and the fear of being a victim. However, suchsurveys only cover household crimes, exclude victimless crimes e.g. drugsoffences and also crimes against businesses, and exclude crimes againstchildren.

Moreover, victimisation surveys are very expensive: few European coun-tries carry out their own victimisation surveys on a regular basis and themethodologies used in each country differ. The only major internationalcomparative survey is the International Crime Victimisation survey (ICVS)which, as shown in this report, provides a basis for comparisons of crimethrough the use of standardised survey questions. However, the ICVS doesnot currently cover all EU countries and its sample size is 2000 per country.

European Journal on Criminal Policy and Research 10: 187–223, 2004.C© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

Page 2: Crime Trends in the EU

188 CHRIS LEWIS ET AL.

The only other similar source is the European Commission funded Euro-barometer. This included a few questions on public attitudes to crime in itssurvey in the last quarter of 2002. It was carried out in all Member Statesof the EU and has a sample size of 1000 per country. The ICVS and theEurobarometer are usfeul for broad national comparisons only.

Data provided through police registered crime statistics are, in compari-son, collected and published at least annually by all EU countries, coveringa large range of crimes. However, the level of crime recorded reflects thelikelihood of a victim reporting the crime to the police and it being recorded.In looking at these figures it is also important to be aware of the followingdifferences in the way that countries collect such figures:

• Police forces record their crime statistics at different points in the crimi-nal justice system. For example, England and Wales record crimes whenthey are first reported to the police whereas France record crimes only ifand when the file is passed to the Prosecutor following an initial inves-tigation.

• Not all crimes are recorded in police statistics. In some countries, someoffences may not be recorded by the police but by other bodies, such asthe fiscal police or the prosecutor.

• Definitions of offences vary between countries. For example, countriesmay exclude less serious assaults from its statistics. Statistical count-ing rules may also vary as to whether incidents, victims or offencesare counted. It is important to be aware of both these findings, in theinterpretation of the data presented in this report, making use of trendcomparisons if more appropriate than absolute ones.

• Although problems have been identified with using both the survey andpolice approach, a good picture of crime can be obtained by combiningdata from both sources. The report on the last ICVS, published by theMinistry of Justice in The Netherlands showed that both victimisationdata and police data followed similar trends. This is, therefore, the basisused in the current report providing an important pointer as to how crimeis and could be measured in the future in both EU members and accessioncountries.

• International surveys results rest on similar understanding of the ques-tions asked in the inquiry. When covering countries with different lan-guages and legal framings the implementation of a survey must providestrict translation assuring, as far as possible, the same meaning and al-lowing respondents to provide equivalent responses.

• Another factor to bear in mind when looking at international surveys,depending on the question, is the influence of subjective judgments,cultural standards and social pressures, such as media presentations andpolitical or advertising campaigns.

Page 3: Crime Trends in the EU

CRIME TRENDS IN THE EU 189

INTERNATIONAL COMPARISONS OF NATIONAL AND OTHER DATA

Interpol

Since 1950 Interpol has compiled crime statistics from many countries (by2001, this had risen to over a hundred countries). The data collection formincludes broad categories of crime departing from legal definitions envis-aged by criminal codes. Interpol warns that such statistics are not meantfor international comparisons, because of different definitions, reportingand recording methods, and lack of quality control. They are quite useful,however, for comparing trends over time in different countries. Publishedin four languages (Arabic, English, French and Spanish), the statistics re-late to major categories of offences brought to the attention of the policein Interpol member countries, including:

• murder,• sex offences (including rape),• rape,• serious assault,• theft (all kinds of theft),• aggravated theft,• robbery and violent theft,• breaking and entering,• theft of motor cars,• other thefts,• fraud,• counterfeit currency offences,• drug offences,• total number of offences contained in national crime statistics.

Interpol statistics also contain information about the number of policein each country. All EU countries are included in the Interpol statistics. Incommon with other sources, the most recent data are not available for allEU countries.

The Interpol data can be accessed from the website: http://www.interpol.int/Public/Statistics/ICS/Default.asp.

UN CRIME SURVEY

The United Nations Survey of Crime Trends and Operations of CriminalJustice Systems (UN Crime Survey) has since 1976 provided informationon recorded crimes from many responding Member States. The major goalof the survey is to collect data on the incidence of reported crime and the

Page 4: Crime Trends in the EU

190 CHRIS LEWIS ET AL.

operations of criminal justice systems with a view to improving the analysisand dissemination of that information globally. The survey results providean overview of trends and interrelationships between various parts of thecriminal justice system to promote informed decision-making in adminis-tration, nationally and internationally. The Economic and Social Council,in its resolution 1984/1948 of 25 May 1984, requested that the Secretary-General maintain and develop the United Nations crime-related databaseby continuing to conduct surveys of crime trends and the operations ofcriminal justice systems.

The survey is now in its seventh sweep, although questionnaires forfurther sweeps have already been sent out. The seventh sweep asks fordata, primarily statistical, on the main components of the criminal justicesystem for 1998–2000. All EU countries have responded to the survey atleast once since 1976. Government agencies in each responding countryprovide official data on crimes adjusted to standard definitions, includingthe following:

• total recorded crimes, regardless of type,• intentional homicide and attempts (and committed with a firearm),• non-intentional homicide,• major assault,• total assault,• rape,• robbery,• major theft,• total theft,• automobile theft,• burglary,• fraud,• embezzlement,• drug-related crime,• bribery and/or corruption.

Data from the first to seventh UN Crime Surveys are avail-able on the Internet at the website http://www.odccp.org/odccp/crime cicp surveys.html. The questionnaire is available to down-load in English, French, Spanish, Arabic and Russian fromhttp://www.odccp.org/odccp/crime cicp survey seventh.html.

The UN, through its Centre for International Crime Prevention (CICP)in Vienna only reproduces the figures received on the questionnaire forms.The statistics cannot take into account the differences that exist between thelegal definitions of offences in various countries, of the different methods oftallying, etc. Consequently, these statistics must be interpreted with caution.

Page 5: Crime Trends in the EU

CRIME TRENDS IN THE EU 191

Broad judgements about trends are possible but not specific comparisonsacross countries.

However, where value and context is added to these data, they can bemore useful. An example is the analysis of the fifth survey results for thewhole world which is published in the Global Report on Crime and Justice(Newman, 1999). An update of that report is being planned, on the basis ofthe seventh survey results. Further useful analyses on the European levelhave been made by the Helsinki Institute, HEUNI (Kangaspunta et al. 1998;Aromaa et al, 2003.)

European Sourcebook on Crime and Criminal Justice Statistics

The Council of Europe is a wider organisation than the EU. It comprisessome 40 states in a looser relationship than the EU. All EU membersbelong to the Council of Europe and there has been co-operation in crimeand justice matters since the 1950s.

A group of statistical experts was convened by the Council at the be-ginning of the 1990s to develop a model for the collection of statis-tics on crime and criminal justice from all Council Member States.A first model was produced in 1995. A report containing 1990–1996data was published in 1999 with data from 34 European countries.(Council of Europe, 1999) A further report, published in 2003, in-cludes 1996–2000 data for 36 European countries. (European Sourcebook,2003).

The European Sourcebook makes use of official statistics from policeforces, prosecution services, courts and correctional systems. It also in-cludes some results from the ICVS.

In each country a national correspondent gathers these data accordingto a detailed questionnaire.

This questionnaire covers the following a set of offences:

• homicide (total and completed),• rape,• assault,• robbery,• burglary (total and domestic burglary),• motor vehicle theft,• other theft,• drug offences (total and drug trafficking).

For each of these offences a standard definition is provided and respon-dents are requested to indicate deviations from this definition in a system-atic way. Collected data are checked by a group of regional coordinators

Page 6: Crime Trends in the EU

192 CHRIS LEWIS ET AL.

and supplementary information on data may be requested from a nationalcorrespondent.

All sourcebook information can also be found on the European Source-book website: http://www.europeansourcebook.org.

The published sourcebook adds value to the raw data in a way that neitherthe Interpol statistics nor the Eurobarometer are able to do.

• It contains rates per 100,000 inhabitants.• It also includes proportions (e.g., of young offenders, female offenders,

aliens) for police statistics.• It also gives in a comparable way, the results of the technical parts of

the questionnaire (for example, deviations from the standard definitionof offences, counting units, changes in data collection). This enables thereader to know when he is comparing like with like.

• It contains a useful commentary that only draws conclusions that arejustified by the expert group’s reading of the statistics, for example clas-sifying changes in crime rates into broad bands: increase/decrease under10%; by 10–50%; more than 50%.

UK Home Office Annual Data Collection

The UK Home Office collects and publishes data on police statistics as partof their annual publication on International Comparison of Criminal Jus-tice Statistics. Information is collected through statistical contacts mainlyin Government Departments including all EU Member States, accedingcountries and the main OECD countries (e.g. the USA). Offences coveredare:

• homicide,• violent crimes (including robbery),• domestic burglary,• car theft,• drug trafficking.

The publication and associated data are available on the Home Officeresearch website www.homeoffice.gov.uk/rds.

In common with the European Sourcebook, this publication adds valueto the raw statistics: Except for homicides, only trends in time are shown inline with the findings of the European Sourcebook. It also enables policestatistics to be seen in the context of the other criminal justice informa-tion included in the annual publication (e.g. prison population, sentencingstatistics and police resources). It also provides more recent data than theother surveys mentioned above and gives a useful commentary with bullet

Page 7: Crime Trends in the EU

CRIME TRENDS IN THE EU 193

points. The 2003 publication contains data up to 2001 (Barclay and Tavares,2003).

International Crime Victimisation Survey

The First International Survey (1989) covered 15 developed/industrialisedcountries, one Eastern European and only one developing country. Startingfrom the second (1992) survey, a growing number of countries joined theproject. As of today approximately 70 countries have been involved atleast once. Due to the informal character of the organisation so far six ofthe 15 Member States did not participate in the last sweep, three of thesesix (Greece, Ireland and Luxembourg) did not participate in any sweep.The ICVS database contains a wealth of information.

The questionnaire includes sections on 13 types of crime, of which eachquestion provides a standard definition. Crimes included in the survey arethe following:

• theft of car,• theft from car,• car vandalism,• theft of motorcycle,• theft of bicycle,• burglary,• attempted burglary,• robbery,• theft of personal property,• sexual incidents (section addressed to women only),• assault/threat,• consumer fraud (cheating),• bribery/corruption.

Furthermore, the questionnaire explores whether crimes were reportedto the police, reasons for not reporting, attitudes toward the police, fearof crime and crime prevention measures. Analysis of the ICVS results forEuropean countries was published in (Van Dijk et al. 1990; Mayhew andVan Dijk 1997; Van Kesteren et al. 2000). There are plans for carryingout a further sweep of the ICVS for all existing states of the EU during2004/2005.

Eurobarometer

The standard Eurobarometer was established as far back as 1973. Eachsurvey consists of approximately 1000 interviews per Member State and

Page 8: Crime Trends in the EU

194 CHRIS LEWIS ET AL.

it is usually conducted between two and five times per year, with reportspublished twice yearly. These reports are available to download from thewebsite http://europa.eu.int/comm/public opinion.

Between 1996 and 2000 this survey included two questions aboutstreet safety after dark and contact with drug-related problems over thelast 12 months. The survey conducted in the last quarter of 2002 addedsome questions on fear of crime and perceptions of risk (risk of be-ing the victim of a series of crimes over the next 12 months) andpublic opinion on crime and crime prevention. Analysis of the publicopinion survey can be found in the report ‘Public Safety, Exposure toDrug-Related Problems and Crime’ (European Opinion Research Group,2003) The crimes covered by the perceptions of risk questions are thefollowing:

• theft of mobile phone,• theft of other personal property,• burglary or break-in at home,• mugging or robbery to steal a mobile phone,• mugging or robbery to steal something else,• assault or threat of assault.

The questions related to crime and crime prevention focused on:

• opinion on certain measures to reduce crime (burglar alarms, specialdoor locks, neighbourhood watch, better policing),

• opinion on police activity and responsibility for crime prevention,• opinion on ways to reduce youth delinquency,• opinion on organised crime.

As in other surveys, given the small sample used, there is a widesampling error on the estimates. When using and analysing the sur-vey results we must bear this in mind as well as the confidence limitsspecified.

OFFENCE DEFINITIONS USED IN THE EUROPEAN SOURCEBOOK

EU Member States have different legal systems and criminal codes. Itis, therefore, to be expected that variations exist in the definition of alloffences including those in this analysis. These variations can occur inboth the scope of the offence (e.g. is an out-building considered part ofa house?) and the seriousness of the offence included (e.g. the type ofthreat made in a robbery). The section below looks specifically at the dif-ferences for the three offences selected, i.e. robbery, domestic burglary

Page 9: Crime Trends in the EU

CRIME TRENDS IN THE EU 195

Figure 1. Robbery, recorded crime rate per 100,000 (2000). Sources: Van Kesteran et al.(2000), and WODC (2003).

and car theft. It should be noted that the definitions giving here are notnecessarily the legal definitions as used in each country, but the operationaldefinitions as used in the European Sourcebook. (European Sourcebook,2003)

Figure 2. Robbery, mean annual growth 1996–2001. Sources: Van Kesteran et al. (2000),and WODC (2003).

Page 10: Crime Trends in the EU

196 CHRIS LEWIS ET AL.

TABLE I

Robbery.

Victimisation rate Recorded crime rate(per 100 inhabitants) per 100,000 inhabitants

Mean annual growth

Country 1995 1999 Rates 2000 1990–1995 1996–2001

Austria 1.5 37 −2.1% 3.9%

(1996–1999)

Belgium 1.4 129 (1996)

Denmark 0.8 59 −2.7% 6.8%

Finland 0.8 0.7 50 −5.1% 2.7%

France 1.3 1.8 185 3.6% 10.6%

Germany 72 1.1% −3.9%

(1993–1995)

Greece 16 5.2% 2.1%

Ireland 37

Italy 68 (1999) −5.9% 8.6%

(1996–1999)

Luxembourg 70 −0.2% 4.2%

(1996–2000)

The Netherlands 0.8 1.5 117 5.3% 7.3%

Portugal 1.3 171 9.7% 7.9%

(1993–1995)

Spain 234 −1.6%

(1996–2000)

Sweden 0.5 1.5 101 −2.4% 9.2%

England and Wales 1.7 2.0 181 20.8%

(1998–2001)

Northern Ireland 0.7 0.1 108 17.5%

(1998–2001)

Scotland 1.0 0.9 85 −0.4% −3.4%

Sources: Van Kesteran et al. (2000) and WODC (2003).

Robbery

The European Sourcebook standard definition of robbery is stealing froma person with force or threat of force (Table I). In other words, robberypresupposes theft of property and that this theft occurs by means of violenceor the threat of violence.

The standard definition includes muggings (bag-snatching) and theftwith violence and excludes pick pocketing, extortion and blackmail. EU

Page 11: Crime Trends in the EU

CRIME TRENDS IN THE EU 197

Figure 3. Domestic burglary, recorded crime rate per 100,000 (2000). Sources: Van Kesteranet al. (2000), and WODC (2003).

Figure 4. Domestic burglary, mean annual growth 1996–2001. Sources: Van Kesteran et al.(2000), and WODC (2003).

Page 12: Crime Trends in the EU

198 CHRIS LEWIS ET AL.

TABLE II

Domestic burglary.

Victimisation rate Recorded crime rate(per 100 inhabitants) per 100,000 inhabitants

Mean annual growthRates

Country 1995 1999 2000 1990–1995 1996–2001

Austria 1.0 163 −6.0% −10.0%

(1996–1999)

Belgium 2.4 741 9.5%

(1998–2000)

Denmark 3.3 616 −1.1% −1.3%

Finland 0.8 0.5 179 1.2% −5.0%

France 2.9 1.0 314 2.2% −3.4%

Germany 170 −3.9% −7.8%

(1993–1995)

Greece 300 13.3% −7.7%

(1993–1995) (1996–2000)

Ireland 352 0.3% −8.1%

(1996–2000)

Italy 359 −0.5% −2.4%

(1996–2000)

Luxembourg 476 3.5% −10.4%

(1993–1995) (1996–2000)

The Netherlands 3.3 2.3 574 3.2% −2.9%

Portugal 1.8 211 6.3% −2.5%

(1993–1995)

Spain 562 −0.6%

(1996–2000)

Sweden 1.5 2.3 198 −6.7% −2.0%

England and Wales 3.4 3.4 765 −4.0%

(1998–2001)

Northern Ireland 1.8 1.7 514 5.5%

(1998–2001)

Scotland 1.6 1.5 628 −5.6%

Sources: Van Kesteran et al. (2000) and WODC (2003).

Page 13: Crime Trends in the EU

CRIME TRENDS IN THE EU 199

Figure 5. Motor vehicle theft, recorded crime rate per 100,000 (2000). Sources: VanKesteran et al. (2000), and WODC (2003).

countries that were not able to meet the standard definition in all respectswere Denmark and Italy where muggings and theft with violence wereexcluded, Sweden where only muggings were excluded and Scotland whereonly theft with violence was excluded.

Robbery offences may be divided into the categories of bank, post office,commercial and robbery of persons. The latter category is often termed‘street-robbery.’ Robbery against institutions such as banks or post officesis rare, whereas robberies against shops and other business establishmentsare more common.

Domestic Burglary

Definitions of domestic burglary, both in criminal law and in statistical prac-tice, vary across countries (Table II). According to the European Source-book standard definition, domestic burglary is gaining access to privatepremises by use of force with the intent to steal goods. Figures on domesticburglary should, where possible, include theft from an attic or from a base-ment (in multiple dwellings) as well as theft from a secondary residence(even if unoccupied); they should exclude, however, theft from a factory,shop or an office, as well as theft from a detached garage, shed, barn orstable, or from a fenced meadow or compound.

Not all EU countries could meet this standard European Sourcebookdefinition, for example:

Page 14: Crime Trends in the EU

200 CHRIS LEWIS ET AL.

Figure 6. Motor vehicle theft, mean annual growth 1996–2001. Sources: Van Kesteran etal. (2000), and WODC (2003).

• Theft from an attic or a basement was excluded in Austria, Finland,France and Sweden.

• Theft from a secondary residence was excluded in Denmark, Ireland andSweden.

• Theft from a factory, shop or office was included in the Netherlands.• Theft from a detached garage, shed, barn or stable was included in

Austria, Belgium, Greece and the Netherlands.• Theft from a fenced meadow or compound was included in Belgium and

the Netherlands.

Car Theft

The standard definition for the European Sourcebook for ‘theft of motorvehicle’ is ‘depriving a person/organisation of a motor vehicle without forcewith the intent to keep or to use it.’ The definition states that joyriding shouldbe included, but that theft of motorboats and receiving/handling a stolenvehicle excluded.

This definition attempts to avoid the two main sources of differencesbetween countries to collect data at the police level. Any theft of motorvehicle should be included but also the unauthorised use of another per-son’s vehicle (joyriding) in countries where such a legal distinction exists.‘Joyriding’ is included in all countries, with the exception of Luxembourg

Page 15: Crime Trends in the EU

CRIME TRENDS IN THE EU 201

and the Netherlands. This should result in lower figures at the police levelfor these two countries.

The situation is not quite clear regarding the type of stolen vehicles.Stolen motorboats are included in Finland, France, Italy, Spain, Swedenand England and Wales, but excluded in other countries. More importantdifferences could come from the restriction to cars that applies to Belgium,Greece and the Netherlands. The inclusion of motorcycles and lorries willaffect comparisons between countries.

Receiving/handling a stolen vehicle is excluded from motor vehicle theftfor all countries. In some countries, vehicles that are stolen with force(car jacking) or during a burglary may be excluded but included in othercountries. The situation is rather uncertain since it depends on the waythat police statistics record multiple offences. National practices may alsodiffer in relation to attempts. The general rule is to include attempts foreach offence index. For vehicle theft, some attempts could be recorded ascriminal damage to a vehicle or attempted theft from a vehicle.

TRENDS IN CRIME

This section covers analysis of the data readily available from the ICVSand Police Statistics for the three selected offences. Primary sources of dataare the individual returns to the ICVS and to the European Sourcebook.Secondary sources are the summary volumes of the ICVS and the EuropeanSourcebook. (Van Kesteren et al., 2000; WODC, 2003).

In all cases attention is given to any changes in offence definitions orstatistical rules in the period covered, which may affect trends. In addition,priority is given to analysis based on trends rather than absolute compar-isons that will be affected by definitional differences.

This section analyses three types of crime: robbery, domestic bur-glary and motor vehicle theft. The level of robbery is much lower thandomestic burglary and car theft in each country. However, it is in-cluded in the report, as robbery is a particularly significant crime atthe present moment, and in recent years has shown a general growthin most Member States. This has been at least partly due to the recentlarge increase in the number of people carrying around with them mod-ern small electronic devices such as walk-men, digital cameras and mobilephones.

Technical Remarks

To interpret the statistics properly, we need to make some technical remarkson the data.

Page 16: Crime Trends in the EU

202 CHRIS LEWIS ET AL.

• In Germany, data for 1990–1993 reflect the consequences of reunifi-cation. The data is for the whole of Germany from 1991 onwards, butorganisational problems related to the transition led to serious under-recording of offences in 1991 and 1992.

• For Portugal, data were not available before 1993.• The 2000 figures for Austria reflect a change in data collection as a

new electronic on-line system has led to a sharp increase in rates perinhabitants from 1999 to 2000.

• In Belgium, data collection had a lower geographical coverage in 1994and 1995 than it did in later years.

• In England and Wales and in Northern Ireland, a change in the offencecoverage and counting rules was introduced in 1998 with visible conse-quences for many offences. The rates displayed are based on these newcounting rules (e.g. 2000 = 1 April 2000 to 31 March 2001).

We need to remind readers that this paper reports work conductedbefore the enlargement of the EU in 2004. At that time the EU com-prised Austria, Belgium, Denmark, Finland, France, Germany, Greece,Ireland, Italy, Luxembourg, The Netherlands, Portugal, Spain, Swedenand the United Kingdom. The Czech Republic, Cyprus, Estonia, Hun-gary, Latvia, Lithuania, Malta, Poland, Slovakia, Slovenia joined the EU inMay 2004.

Robbery

DiscussionReporting rates for robbery depend heavily on who is the victim:

• Unrecorded offences of bank or post-office robbery, or robberies of largeorganisations, are practically non-existent.

• However, unrecorded cases of robbery against persons are probably ex-tremely high.

This is partly because most large organisations are insured and need toreport a robbery for insurance purposes. Many individuals are not insuredagainst street robbery, or are forced by their insurers to pay the first, say 100euros. Moreover, many robberies from individuals involve victims who areyoung people, or socially marginalised, who are less likely to report theseto the police.

Both the police recorded crime rates for Robbery and the ICVS victimi-sation rates for Robbery, where available, varied greatly between MemberStates of the EU. There was some correlation between the two measures,with both France and England and Wales having the highest ICVS and

Page 17: Crime Trends in the EU

CRIME TRENDS IN THE EU 203

police crime rates, but, in general the ranking of countries by police crimerates was not a good predictor of the ranking by ICVS rates.

It is possible that these wide variations in rates reflect the real situationin the different countries. However, the authors feel it unlikely that robberyrates would vary so much in what are quite similar countires. They considerthat it is also possible for there to be some problems with data collection,which cause these differences to be at least in part, statistical artefacts.Such problems could be connected with problems of definition in the po-lice data or of difficulties of translation with the ICVS data. The authorsjudge that more work is necessary on the validity of these Robbery mea-sures before they are used with confidence in comparing Member Statesof the EU.

However, more confidence can be placed in trends over the past 10 years,where the ICVS figures for 1996 and 2000, where available, generallyconfirm the trends shown in the police data. Although there was a smalldecrease in the police rate in several Member States during the first half ofthe 1990s, there was a clear increase in Robbery in most Member Statesfor the period 1996–2001.

Robbery is in general viewed as a serious crime, although it can rangefrom an extremely large theft with significant violence from, say, a bank,to the stealing of a few euros with minimal violence. Despite problemswith the statistics which make comparison of robbery levels unsatisfactory,comparison of trends in robbery across Member States can be regarded asrobust.

The clear increase during the second half of the 1990s is particularlydisturbing. It is probably heavily influenced by the large increase in theownership of small electronic devices, such as games, walk-men, digitalcameras, and mobile phones during this period, particular amongst youngpeople. Crime prevention measures in this area vary across the EU andsupport for robbery victims, in particular young victims, has recently beenimproved in some countries. An important policy question is to investigatefurther how the police are dealing with such robberies and what measuresare being taken by local crime prevention networks.

Domestic Burglary

DiscussionComparison of domestic burglary rates from the ICVS on victimisation,where available, and police crime levels need to be seen in the contextof differences in offence definitions and recording methods. For example,some countries may not distinguish between domestic burglary and otherburglaries and may or may not include unauthorised entering without loss.The levels found may also be affected by the large variations across Europe

Page 18: Crime Trends in the EU

204 CHRIS LEWIS ET AL.

in the numbers of second homes. Finally, reporting levels for domesticburglary will depend on whether there is a requirement to report suchcrimes to the police for insurance purposes as well as on the victims viewof the likely police response. However, most countries have a much higherreporting rate for domestic burglary than for robbery.

Furthermore, not all countries publish separate figures on domestic bur-glary, but instead, aggravated theft or similar categories derived from therespective criminal code are reported. Obviously, aggravated theft wouldunlikely be equivalent to domestic burglary in any jurisdiction. Also,many domestic burglaries may not be classified as aggravated thefts butas simple thefts. Another definitional problem is that where the categoryis recognised, domestic burglary may comprise a different scope of targetsin different jurisdictions. For example, in some countries, attics, basementsand even garages may be included in the category, and second homes orsummer cottages may or may not be counted.

In contrast with Robbery figures, these figures for Domestic Burglary canbe regarded as fairly robust measures, yielding valid comparisons betweenMember States, for both crime rates and trends.

England and Wales shows the highest levels of domestic burglary andAustria the lowest, for both ICVS and police crime rates. In general theranking of the countries is relatively similar for both sources, althoughScotland shows a victimisation rate well below that expected from policestatistics. France and Sweden showed wide variations between their 1996and 2000 ICVS results for domestic burglary.

Police statistics show that domestic burglary has fallen in most EU Mem-ber States between 1996 and 2000, with the highest falls in Luxembourg,Ireland and Germany.

Motor Vehicle Theft

DiscussionThe results for motor vehicle theft show large variations between countriesfor both levels and trends. In 2000, the highest recorded rate per 100 000inhabitants was about seven times the lowest rate.

The counting unit can influence statistical results. In some countries itmay be the stolen vehicle (as in France), or it may be the victim (as inEngland). Three cars belonging to the same company reported stolen arecounted as three offences in France and one offence in England. Compar-isons in level between countries for rates of motor vehicle theft need totake these technical problems into account.

Police figures can be used to explore trends of motor vehicle thefts inEuropean countries during the past 10 years. Since definitions and proce-dures for data collection may have changed in some countries, these time

Page 19: Crime Trends in the EU

CRIME TRENDS IN THE EU 205

Figure 7. Rate of motor vehicle theft per inhabitant and per vehicle in use (2000). Sources:Van Kesteran et al. (2000), WODC (2003), and Varly (1999).

series may not reflect an actual change in crime. The trends are summarisedin Table III. This shows great variations in trends for different countries.This implies the need for more research to understand what lies behindsuch trends.

Several countries show a decreasing rate during the second half of theperiod (namely Austria, Denmark, France, Germany, Italy, Luxembourg,The Netherlands, England and Wales, Scotland). Belgium has a rathersteady rate for this period. Spain and Sweden show a fall succeeded bya rise. For the remaining countries rates mainly increase steadily from1990 to 2000 but among them, several experienced high increases in thenumber of vehicles in use (Ireland, Greece, Portugal, Spain). For these fourcountries, the rate of motor vehicle theft per 1000 vehicles in use showssome significant differences in trend compared with those per 100,000inhabitants (see Table III, last column). In summary, despite the limitationsof the data, it is possible to state that the main trend is a decreasing one,at least since the mid-1990s. This is possibly due to increasing securitymeasures, suggesting that property crime in general could be decreasedsubstantially by these kind of measures.

To compare European countries in more detail some other rates may becalculated. One solution is to relate theft of cars to the opportunity for suchtheft. This can be done for each country by dividing the number of stolenvehicles by the total number of vehicles in use (car parc) for the same year.

Such a measure introduces other problems. The available figures for ve-hicles in use are from the ACEA (Association des Constructeurs Europeens

Page 20: Crime Trends in the EU

206 CHRIS LEWIS ET AL.TA

BL

EII

I

Mot

orve

hicl

eth

eft.

Vic

timiz

atio

nra

teR

ate

per

100,

000

inha

bita

nts

(rec

)R

ate

per

1000

vehi

cles

(rec

)(1

00in

habi

tant

s)M

ean

annu

algr

owth

Rat

eR

ate

Mea

nan

nual

grow

thC

ount

ry19

9519

9920

0019

90–1

995

1996

–200

120

0019

95–2

001

Aus

tria

0.1

130

−4.1

%−0

.4%

(199

6–19

99)

2.4

−4.1

%(1

996–

1999

)B

elgi

um0.

932

614

.4%

(199

4–19

95)

−0.2

%(1

996–

2000

)6.

4−3

.1%

(199

6–20

00)

Den

mar

k1.

263

21.

0%(1

993–

1995

)−7

.8%

15.1

−6.5

%Fi

nlan

d0.

50.

451

1−0

.1%

3.3%

10.9

1.9%

Fran

ce1.

82.

067

60.

3%−1

.7%

11.9

−4.0

%G

erm

any

155

−2.9

%(1

991–

1995

)−1

2.0%

2.7

−14.

0%G

reec

e15

610

.5%

5.9%

(199

6–20

00)

3.9

0.4%

(199

6–20

00)

Irel

and

420

−0.5

%1.

6%10

.5−1

.8%

Ital

y42

3−2

.3%

−5.5

%(1

996–

2000

)6.

7−5

.6%

(199

6–20

00)

Lux

embo

urg

124

15.7

%−9

.8%

(199

6–20

00)

Net

herl

ands

0.4

0.4

241

−4.2

%(1

993–

1995

)−1

.0%

5.1

−4.7

%Po

rtug

al1.

026

35.

4%(1

993–

1995

)5.

2%5.

60.

9%Sp

ain

336

−6.7

%3.

7%6.

31.

3%Sw

eden

1.5

1.4

978

−5.9

%3.

8%(1

996–

2000

)19

.81.

6%(1

996–

2000

)E

ngla

ndan

dW

ales

2.4

643

−6.4

%(1

998–

2001

)N

orth

ern

Irel

and

1.6

1.5

663

5.9%

(199

8–20

01)

Scot

land

2.0

0.9

504

−0.6

%−6

.2%

UK

12.3

−8.0

%(1

998–

2001

)

Sour

ces:

Van

Kes

tera

net

al.(

2000

)an

dW

OD

C(2

003)

.

Page 21: Crime Trends in the EU

CRIME TRENDS IN THE EU 207

Figure 8. Stolen vehicles, 2001. Cars only for Germany. Source: Europol Statistics.

d’Automobiles) and add all kinds of commercial vehicles to passenger cars.However, they do not include motorcycles that represent a large part of ve-hicle theft in some countries (25% in France in 2001 for instance). As thecontent of stolen vehicles and ‘car parc’ figures is not the same, this ratecould show some artificial differences between countries.

The variation among European countries is still important when thenumber of vehicles in use is taken into account (Figure 7). Such a result maybe a consequence of the variation in definitions regarding stolen vehiclesand vehicles in use, in particular in relation to motorcycles. Several authorshave emphasised how vehicle thefts recorded by the police may be linkedto insurance practice:

• Insurance against theft does not cover the same proportion of vehiclesamongst EU countries. For example, in 1996 for private cars, the pro-portion insured against theft varied from 23 to 90% (Varly, 1999). Sinceinsurance rules can be considered a reason to report vehicle theft to thepolice, this could explain some of the differences between countries.

• An unknown proportion of reported thefts are in fact insurance frauds,and insurance companies do not publish figures about the percentage ofthese cases.

Page 22: Crime Trends in the EU

208 CHRIS LEWIS ET AL.

TABLE IV

Comparison between ICVS and police statistics (motor vehicles)

1995 1999

ICVS incidence rates ICVS incidence ratesRecorded per 100 owners Recorded per 100 ownersrate per 100 rate per 100

Country vehicles Cars Motorcycles vehicles Cars Motorcycles

Austria 0.2 0.2 0.0 0.2Belgium 0.8 0.7Denmark 1.8 1.6 1.5 6.1Finland 0.9 0.6 1.2 1.2 0.8 0.5France 1.5 2.1 5.2 1.2 2.2 1.5Netherlands 0.6 0.4 5.0 0.5 0.5 3.0Portugal 0.5 0.6 1.3 1.4Sweden 1.8 1.8 3.1 1.8 1.8 1.6UK England 3.3 2.3 3.0 5.7and WalesUK Northern 1.9 0.0 1.8 0.0IrelandUK Scotland 2.6 1.3 1.2 4.0UK Total 2.0 1.4

Recorded rate: Cars only for Belgium and the Netherlands, Sources: Van Kesteran etal. (2000), and WODC (2003).

• Beyond the legal definitions of vehicle theft and the unauthorised use ofa vehicle, a common distinction is made about vehicle thefts. This refersto some stolen vehicles being recovered after a short use by the offenderwhereas others are permanently lost by the owner (e.g. because theyenter into some organised international vehicle trafficking). However,figures collected by Europol provide information on stolen vehicles,both recovered and not recovered. They are based on various databasesin use at national and European level.

It is important to comment on problems relating to definition and partic-ularly the effect on the comparability of the data. Based on the assumptionthat stolen vehicles that have not been recovered mainly supply the illegalmarket, Figure 8 shows that the risk of disappearance after a theft may beat comparable level in countries having very different levels for the risk ofrecorded theft. According to this data, the risk of disappearance is aboutthe same in Austria and Sweden, but Austria shows the lowest rate of theftper vehicle in use and Sweden the highest.

Table IV shows the results from ICVS (columns 2 and 3). Another wayto present these data is as an incident rate for vehicle owners, which is

Page 23: Crime Trends in the EU

CRIME TRENDS IN THE EU 209

the number of incidents reported in the survey for the preceding year per100 owners. These can be compared with the rate of thefts reported to thepolice per 100 vehicles in use. However, this definition is problematic asthe ICVS only records thefts from households and not from companiesand the ICVS results are given separately for theft of cars and for theftof motorcycles. Therefore, detailed data concerning motor vehicle ownersamong households would be necessary in order to avoid this difficulty.Table IV shows the results for 1995 and 1999, for the nine EU countriesinvolved in ICVS in 1996 (results for 1995) or 2000 (results for 1999).

These results confirm that the substantial differences in the rates of motorvehicle theft among EU countries cannot simply be attributed to differingdefinitions and data collections. However, the link between police dataand ICVS data is not uniform. In fact, ICVS data reveals that differencesin the numbers of motorcycle theft could be larger than numbers of cartheft where the risk for owners are compared. However, the small num-ber of cases (owners of motorcycle in the samples) leads to important butnon-significant variations in the estimations from one survey to the other.According to incident rates, motorcycle theft constitutes a large number ofall incidents of motor vehicle thefts (cars and motorcycles) in Denmark andthe Netherlands. Complementary information from the ICVS also confirmsthat in some countries the “recovery” rate is quite high, a result probablyrelated to the proportion of thefts for “joyriding” (Denmark, Portugal, Swe-den).

RESULTS FROM THE EUROBAROMETER

This section includes results from three Eurobarometer sweeps for 1996,2000 and 2002. Focusing on the offence types this report concentrates upon,it presents a selection of results on:

• fear of crime,• perception of risk,• opinion on crime prevention and measures.

Fear of Crime – Street Safety

How safe do you feel walking alone in the area you live after dark? Do you feel verysafe, fairly safe, a bit unsafe or very unsafe?

This question is used to measure ‘fear of crime’ and it is significantlydifferent from those questions that intend to measure perceptions of risksor concerns about specific types of crime. The same question has appeared

Page 24: Crime Trends in the EU

210 CHRIS LEWIS ET AL.

Figure 9. Percentage feeling a bit or very unsafe walking alone after dark in their area.Countries are sorted by percentage of feeling unsafe (2002). Source: Eurobarometer.

in the ICVS since 1992. In the report describing the findings from the lastsweep of the ICVS (Van Kesteren, 2000) it was stated that this question islikely to be differently interpreted in cross-country comparisons, “for somepeople the prospect of being out after dark evokes anxiety about a greatrange of mishaps (e.g., accidents as well as crime)” and for others who arerarely alone outside after dark “the question is also hypothetical.”

Trying to relate ‘fear of crime’ with risks or victimisation experience isdifficult. As it is stated in the ICVS (Van Kesteren, 2000) “this measure ofstreet safety is not consistently related to levels of contact crime (robbery,sexual incidents, and assaults and threats).” In that report it is acknowledged

Page 25: Crime Trends in the EU

CRIME TRENDS IN THE EU 211

Figure 10. Perceptions of risk of burglary or break-in at home. Countries are sorted bypercentage of ‘yes’. Source: Eurobarometer.

that “one implication of the lack of much relationship between anxiety andrisks is that fear of street crime may be influenced by specific ‘cultural’pressures, such as media presentations of violent crime”. As the Euro-barometer does not provide indicators about victimisation experience, thatrelationship cannot be evaluated here.

It has been shown in other surveys (Kershaw, 2001) that women and theelderly are the most fearful of being out alone after dark. In the three sweepsof the survey, feelings of insecurity are more common among the over-55age group than among the younger generation. However, the distinctionbetween male and female anxiety is clearer, as shown in Figure 9.

The anxiety about being out alone at night varies widely on a country-to-country basis. When looking at the 2002 Eurobarometer figures, themost anxious respondents were those in Greece (43%), Italy (42%) and theUK (42%). In contrast, the lowest figures for feeling a bit or very unsafewalking alone after dark in their area were observed in Denmark, where60% of respondents said they felt ‘very safe’.

When considering the results from the ICVS on the same questionand year (comparing the survey results of the 2000 sweep of ICVSand Eurobarometer concerning Belgium, Denmark, Finland, France, TheNetherlands, Portugal and Sweden), countries are as a general rule, sorted

Page 26: Crime Trends in the EU

212 CHRIS LEWIS ET AL.

Figure 11. Perceptions of risk of mugging or robbery in order to steal a mobile phone.

by perception of “unsafe walking alone after dark in their area” in the sameorder. The great exception is Belgium where respondents in the Eurobarom-eter survey show a relatively higher level of anxiety when compared to itsposition in the ranking of countries in the ICVS.

We observe, across the EU-15, a slight increase in the percentage ofthose feeling a bit or very unsafe, from 1996 to 2000 and from this year to2002. However, due to small sample sizes, it is unlikely that this increaseis sufficiently significant.

Above the median increase (Eurobarometer, 1996–2002) are Greece,UK, Italy, Luxembourg, Finland, France, The Netherlands and Denmark.The most pronounced change occurred in Greece, from 28% in 1996 to51% in 2000.

Perceptions of Risk

Over the next 12 months, do you think there is a risk that you will personally be the victimof one of the following:

Page 27: Crime Trends in the EU

CRIME TRENDS IN THE EU 213

Figure 12. Perceptions of risk of mugging or robbery in order to steal something else.Countries are sorted by percentage answering ‘yes’. Source: Eurobarometer.

Theft of mobile phone; theft of other personal property; mugging or robbery in order to steala mobile phone; mugging or robbery in order to steal something else; burglary or break-inat home; and assault or threat of assault?

Yes, no, don’t know?

The answers to this question in the Eurobarometer survey resulted in ameasure of concern about a series of types of crimes. It is important to notethat in some cases there were significant levels of ‘don’t know’ answers.These Eurobarometer results are a useful addition to an area where therelationship between fear of crime and some specific risk of victimisationfor the total population is unclear, despite the fact that research has shownthat the experience of victimisation does have a consequence on the victim’sperception of risk (Killias, 2001).

Risk of Burglary or Break-in at Home

About one-quarter of the respondents did consider there was a risk theywould fall victim to burglary or break-in at home over the coming year.

Page 28: Crime Trends in the EU

214 CHRIS LEWIS ET AL.

Figure 13. Measures such as burglar alarms and special door locks can reduce crime inmay area.

However, the scores are very different from country to country. The mostpessimistic ones were those in Greece (54%) and France (41%). The leastconcerned were those in Germany, (11%), and Austria (14%).

Across the EU-15, perceptions of risk are greater among the women(28%, when opposed to 25% among men) and elderly people (rising from21% for the 15–24 age group to 30% for the 55 and above). An increasedperception of risk is also found in urbanised areas and among those whohave access to a mobile phone in their household. The results show thatthe level of income is not related to a higher or lower expectation of therisk.

Although the Eurobarometer does not provide indicators to establish alink between the likelihood of burglary and the perceptions of risk, thatlink was found in the ICVS.

Page 29: Crime Trends in the EU

CRIME TRENDS IN THE EU 215

Figure 14. Measures such as a neighbourhood watch scheme can reduce crime in may area.

Risk of Mugging or Robbery

Among the respondents, 18% felt at risk over the next year of mugging orrobbery in order to steal a mobile phone (Figures 10–12). Again, scoresare very different from country to country. The countries with the highestproportions feeling at risk were Greece (37%), followed by France (30%),Luxembourg (28%) and Portugal (27%). The countries with the lowestproportions were Austria, Denmark (each 6%) and Germany (8%).

However, it is important to note that the question was asked bothto those who own and do not own a mobile phone. In another partof the survey people were asked if there was a mobile phone avail-able in their household. The risk of stealing something else was higherthan of stealing a mobile phone. The countries with the highest scoreswere Greece (49%), France (39%), Luxembourg and Portugal (35% each)and the UK (30%). The lowest scores were in Germany and Austria(both 10%).

Page 30: Crime Trends in the EU

216 CHRIS LEWIS ET AL.

Figure 15. Better policing would reduce crime in my area. Countries are sorted by percent-age of ‘agree’. Source: Eurobarometer.

In urbanised areas and among women there is a higher perception of riskboth for stealing a mobile phone or something else. Specifically for mobilephones, the age was noteworthy, with higher levels of ‘yes’ responsesregistered among younger people. For stealing something else, the incomewas important: both the poorest and the wealthiest tend to feel the mostvulnerable to this risk.

Crime and Crime Prevention

Respondents were asked in the Eurobarometer whether they tended to agreeor disagree with a series of statements relating to crime and crime preven-tion (Figures 13–15). From the set of statements, we have selected threeconcerning specific measures to reduce crime and one about the opinionon police work in the fight against crime.

Page 31: Crime Trends in the EU

CRIME TRENDS IN THE EU 217

Figure 16. Police in my area are doing a good job in the fight against crime. Countries aressorted by percentage of ‘agree’. Source: Eurobarometer.

In response to each of the three questions about measures to reducecrime, there was hardly any difference between men and women. Higherlevels of agreement were found amongst elder respondents and amongstthose living in small- or medium-sized towns, followed by those livingin large towns and, with significant lower percentages, by those in ruralareas/villages.

The responses in Figure 16 are concerned with whether police work inthe local area helped the fight against crime varied widely from country tocountry. For example, the following countries have 70% or more agreeingwith the statement “police in my area are doing a good job in the fightagainst crime” (Ireland 74%, Austria 72% and Finland 70%). In compari-son, the following countries had less than 50% agreeing with the statement(Luxembourg and Italy 49%, UK and the Netherlands 46% and Sweden45%). In addition, in some countries there were significant percentages ofthose who answered ‘don’t know’ (Sweden 23%, Germany 22% and theNetherlands 21%).

It was also evident that levels of agreement are higher among men andtend to rise with age.

Page 32: Crime Trends in the EU

218 CHRIS LEWIS ET AL.

CONCLUSIONS AND RECOMMENDATIONS

Conclusions

This paper illustrates that data currently exist from different sources toprovide a picture of crime trends in EU Member States. Criminal researchhas established that these different sources need to be used together toobtain an overall picture of crime and the examples in this report (robbery,domestic burglary and motor vehicle theft) show how to combine thesesources.

Absolute comparisons based on police statistics are limited by knowndifferences in the definitions of offences and the various national statisticalrules. However, data from the ICVS show that differences in level forthe rates of robbery, domestic burglary and vehicle theft are not a mereconsequence of technical statistical problems. This implies that throughoutthe EU improvements in police recorded crime statistics need to run inparallel with improvements in victimisation surveys.

Comparisons in trends remain problematic because over the last 10 yearsmany European countries have experienced changes in their police statisti-cal recording systems. However, it can be said with some certainty that thegeneral trend throughout the EU over the last 10 years is for a significantgrowth in robbery but a decline in domestic burglary and motor vehicletheft. This diversity in trend for the three selected offences types is a re-minder that it is not possible to describe changes in the crime situation witha single indicator. This again implies that improvements to European crimestatistics need to take place on a broad front.

Nevertheless, some countries do not follow these general trends. It canbe hypothesised that the opportunity to commit property offences factorsis an important driver of crime, so that some of this variation may beassociated with variations in such socio-economic factors as the proportionof mobile phone owners, of car or motorcycle owners, or of second homeownership in each Member State. But once again, more detailed, accurateand comparable data would be necessary to strengthen these explanationsand inform detailed crime prevention assessment.

Similarly, results from the Eurobarometer also show important differ-ences between EU countries. Within each country, many surveys show thatfear of crime and perception of risk of crime do not follow the same patternsince a high fear of crime is not shown to always result in a high perceptionof risk. The same could be true when comparing countries. Available datado not yet allow a direct comparison of countries ranked for the perceptionof risk and the actual risk for a specific offence.

Most EU countries are themselves made up of a number of regions andcities, which differ considerably from each other in social and economicbackground. Hence, we would expect considerable variation in crime rates

Page 33: Crime Trends in the EU

CRIME TRENDS IN THE EU 219

within countries, even with the same statistical methodology. Most coun-tries have police data on a regional and local basis. However, few coun-tries have victimisation and satisfaction data on anything but a nationalbasis, and the Eurobarometer has too small a sample to provide regionalanalyses.

If future criminal justice policies follow EU policies in other social areas,then future developments in EU criminal justice policies could be directedas much towards regions within countries as to the countries themselves.For example, crime prevention might be encouraged in capital cities, asthe most frequently visited areas: anti-terrorist activity might be concen-trated in centre of population rather than rural areas. If this were the case,then statistics would also be needed on a regional or capital city basis.This has implications for sample sizes of surveys, and for the degree ofdisaggregation of police data.

Recommendations for Future Developments

• Collecting good statistics costs money. Without some ongoing commit-ment at the EU level the recommendations in this report will not be ableto be carried through.

• There is a need to provide financial support to the ICVS and EuropeanSourcebook projects to guarantee that these projects will continue toprovide good crime and justice statistics. The sample size of the ICVSneeds to be large enough to provide a breakdown for regions withincountries.

• Statistics should be used both in both the drafting and the implementationof EU Directives on crime and justice. There is a growing series of Coun-cil Directives from the EU in this field. Many of could be better draftedif good statistics existed throughout the EU: they also need good dataif they are to be implemented properly. Examples of current proposeddirectives that would benefit from improved statistics are the two pro-posed directives (see EU web site http//ue.eu.int/uedocs/cmsUpload/EN-2004.Rd.pdf):◦ related to the compensation of crime victims, and◦ related to third-country nationals who have been victims of humantrafficking and co-operate with the authorities.

• The AEGIS fund should continue to be used to fund research towards bet-ter statistics on crime: e.g. The Italian research Institute TRANSCRIMEis directing a current study of the databases that are used in EUMember States to counter organised crime (see Transcrime web sitewww.jus.unitn.it/transcrime/).

• Statistics should be used as an objective measure to allocate funds toMember States. There is a likelihood of significant funding from the

Page 34: Crime Trends in the EU

220 CHRIS LEWIS ET AL.

EU in crime and justice area from 2007, especially in crime prevention.This allocation should be based on objective data rather than politicaljudgement. It would need statistics from all Member States to drive theallocation. Possible examples are:◦ regions or capital cities particularly targeted by terrorists,◦ areas particularly targeted by organised crime syndicates, such as drugor human trafficking.

• Current plans to increase the amount of statistical work shouldbe expedited. A group of experts in statistical data collection onorganised and volume crime was set up by DGJHA during 2003,following the Amsterdam summit and the Dublin declaration (seewww.tocpartnership.org/orgcrime 2003/). This group is now drawingup an action plan and it is hoped that this would enhance the decisionsof policy makers at national and EU level.

• Good statistical practice in criminal justice should be passed from onecountry to another using the authority of the EU Commission. Statisticalsystems in the current 15 Member States are generally, although notuniversally, more developed than in the 10 countries that joined in May2004. The challenges to data collectors and policy makers arising fromthis enlargement should be met through using the AEGIS fund and theexpert group, which contains representation all 25 EU Member States.

• Members of the EU statistical expert group should work with theEuropean Commission and other stakeholders, such as national institutesand other international bodies, to produce a longer-term programme forthe organisation, coverage and funding for work on the statistics of crimeand justice. Such a programme would need to bear in mind the followingpoints:◦ Differences in legal systems between EU Member States mean thatharmonization between police crime statistics, prosecution and courtstatistics is not possible in the short term.◦ Future work should link with the developing EU work program ofDGJAI, including existing framework agreements and potential fundingdevelopments.

• Suggested priorities for work on volume crime statistics are:• the extension of the collection of crime statistics to include all 25 Member

States, initially at a country level but eventually at a regional level withincountries;

• the production of an annual report on EU crime statistics;• the production annually of a short 4–6-page document for easy reading by

policy makers, listing key statistical facts about the situation in MemberStates;

• the development of a European Union Crime Victimisation survey aspart of the International Crime Victimisation Survey;

Page 35: Crime Trends in the EU

CRIME TRENDS IN THE EU 221

• in the longer term this should be expected to provide sub-national data,e.g. for major cities;

• consideration should be given to developing measures of crime againstcommercial properties and other sectors of the economy of concern tothe EU;

• repeat of the 2002 crime module in the Eurobarometer survey after areassessment of the accuracy of the previous results and the questionsasked;

• build up relationships with organisations such as regulatory bodies, non-government organisations, hospitals, insurance companies etc, to collectnon-official crime data;

• development of a research program to identify the causes of differencesin crime levels between countries/cities.

Web Site References

Recent developments in crime and justice in the European Union can befound in the web site of the Directorate of Justice and Home Affairs:http://ue.eu.int/comm/justice home/

A description of the Eurobarometer can be found on the web sitehttp://europa.eu.int/comm/public opinion/

Recent work on organised crime by TRANSCRIME can be found onthe TRANSCRIME website site www.jus.unitn.it/transcrime/

Recent developments on setting up partnerships in organised crime canbe found on the web site www.tocpartnership.org/orgcrime 2003/

APPENDIX: MEMBERS OF THE EUCPN SUBGROUP

(Chairman) – Mr Jaap De Waard(Austria) – Ms Erika Gramsjager(Belgium) – Mr Walter De Pauw(Belguim) – Mrs Marine Pattyn(Belgium) – Ms Ellen Tack(Denmark) – Mrs Britta Kyvsgaard(EUROPOL) – Mr Siegfried Kammhuber(Finland) – Mr Kauko Aromaa(France) – Mr Michel Allier Gayrard(France) – Mr Bruno Aubusson De Cavarlay(France) – Mr Patrick Delacour(France) – Ms Virginie Mora(Germany) – Mr Richard Blath(Germany) – Mr Uwe Doermann(Greece) – Mr. Emmanouil Daskalakis

Page 36: Crime Trends in the EU

222 CHRIS LEWIS ET AL.

(Greece) – Mr Marinos Stagakis(Ireland) – Mr Noel Carolan(Italy) – Mr Antonio Mannoni(Netherlands) – Mr Paul Smit(Portugal) – Ms Maria Joao Costa(Spain) – Mr Jesus Alonso Hernaiz(Spain) – Mr Jose Antonio Avila Morete(Spain) – Mr Javier Cirujano Gonzalez(Spain) – Ms Teresa Gonzalez Garcia(Spain) – Mr Mario Hernandez Lores(Sweden) – Mr Peter Lindstrom(UK) – Mr Gordon Barclay(UNICRI) – Ms Anna Del Frate(Secretary UK) – Ms Emma Wilby

REFERENCES

Aromaa, K., S. Leppa, S. Nevala and N. Ollus, Crime and Criminal Justice Systems inEurope and North America. Helsinki: European Institute for Crime Prevention and Con-trol affiliated with the United Nations, HEUNI, 2003.

Barclay, G. and C. Tavares, International Comparisons of Criminal JusticeStatistics 2001. Statistical Bulletin 12/2003. London: Home Office, 2003(www.homeoffice.gsi.gov.uk/rds).

Council of Europe, European Sourcebook of Crime and Criminal Justice Statistics. Stras-bourg: Council of Europe PC-S-ST (99) 8 DEF, 1999.

European Opinion Research Group, Public Safety, Exposure to Drug-Related Problemsand Crime – Public Opinion Survey. Report Prepared for The European Commission(unpublished). Brussels: EC, 2003.

European Sourcebook of Crime and Criminal Justice Statistics. The Hague: WODC/BJU, 2003.

Junger-Tas, J., G.-J. Terlouw and M.W. Klein (Eds.), Delinquent Behavior among YoungPeople in the Western World. Amsterdam: Kugler, 1994.

Kangaspunta, K., M. Joutsen and N. Ollus, Crime and Criminal Justice Systems inEurope and North America. Helsinki: European Institute for Crime Prevention and Con-trol affiliated with the United Nations, HEUNI, 1998.

Kerahaw, C., N. Chivite-Mathews, C. Thomas and R. Aust, The 2001 British Crime Survey,First Results, England and Wales. Statistical Bulletin 18/01. London: Home Office, 2001,(www.homeoffice.gsi.gov.uk/rds).

Killias, M., Precis de criminologie, pp. 400–416. Berne: Staempfli Editions, 2001.Maguire, M., Crime statistics: The data explosion and its implications. In M. Maguire (Ed.),

The Oxford Handbook of Criminology, 3rd ed., pp. 322–375. Oxford: Oxford UniversityPress, 2002.

Mayhew, P. and J.J.M. Van Dijk, Criminal Victimisation in Eleven Industrialised Countries.The Hague: WODC, Ministry of Justice of The Netherlands, 1997.

Newman, G. (Ed.), Global Report on Crime and Justice. Oxford and New York: UnitedNations, Office for Drug Control and Crime Prevention, Centre for International CrimePrevention, and Oxford University Press, 1999.

Page 37: Crime Trends in the EU

CRIME TRENDS IN THE EU 223

Van Dijk, J.J.M., P. Mayhew and M. Killias, Experiences of Crime across the World: KeyFindings of the 1989 International Crime Survey. Deventer: Kluwer, 1990.

Van Kesteren, J., P. Mayhew and P. Nieuwbeerta, Criminal Victimisation in SeventeenIndustrialised Countries. The Hague: WODC, Ministry of Justice of The Netherlands,2000.

Varly, P., Harmonisation des donnees relatives aux vols et trafics de vehicules dans l’UnionEuropeenne, rapport. Paris: IHESI (figures quoted from the CEA, Comite Europeen desAssurances), 1999.

CHRIS LEWIS

Institute of Criminal Justice Studies,University of Portsmouth,St. George’s House,141 High Street,Portsmouth, UKE-mail: [email protected]

GORDON BARCLAY

Home Office, UK

BRUNO AUBUSSON DE CAVARLAY

CESDIP, Paris, France

MARIA JOAO MORGADO COSTA

Legal Policy and Planning Office, Ministry of Justice, Portugal

PAUL SMIT

WODC, The Netherlands