Piecing Together

the Cross-National

Crime Puzzle

by Jan van Dijk and

Kristiina Kangaspunta

National Institute of Justice Journal ■ January 200035

Understanding national crimelevels has proven to be a dif-ficult task. In the United

States, for example, cities that haveexperienced dramatic declines intheir crime rates in recent years maylink those declines to increasedenforcement, a decrease in crackmarkets, and better job opportuni-ties. On the other hand, cities withsimilar social and economic devel-opments experience no commensu-rate decline in their crime rates.

Uncovering the mystery of why andhow crime varies cross-nationally iseven more difficult: Why, for exam-ple, is serious violent crime moreprevalent in certain Central andEastern European countries and theUnited States than in the EuropeanUnion countries and Canada?Understanding the variation incrime rates across countries canhelp policymakers put crime in theirown countries into perspective.

To develop a deeper understandingof how crime varies across coun-tries, a team of researchers atHEUNI, the European Institute forCrime Prevention and Control affil-iated with the United Nations,1 reg-ularly analyzes criminal justice datacollected by the UN from its mem-ber states.2

HEUNI’s latest analysis focuses onthe differences among European andNorth American countries and isbased on the Fifth United NationsSurvey on Crime Trends andOperations of Criminal JusticeSystems 1990–1994, as well as ondata from the International CrimeVictim Survey (ICVS).

The HEUNI study is the first everattempt to construct multisourcemeasures of crime that do not sufferfrom the well known difficulties ofmeasures based on the number ofcrimes reported to the police in var-ious countries. (See “The Difficultiesof Analyzing Crime Data AcrossCountries.”) The study may signify a

breakthrough in comparative crimestatistics.

This article focuses on three of theeight crimes HEUNI researchersstudied: Property crime, violence,and corruption in Central and

Eastern Europe, Western Europe,and North America.3 Before consid-ering these findings, however, it isimportant to be aware of the pitfallsrelated to statistical comparisons ofcrime in different countries.

about the authorsJan van Dijk, a former professor of criminology at the University of Leiden, TheNetherlands, is currently the officer in charge of the United Nations Centre forInternational Crime Prevention (UNCICP) in Vienna. Kristiina Kangaspunta, formerlyProgramme Officer at HEUNI, is currently a senior officer at UNCICP.

The Difficulties of Analyzing Crime Data Across CountriesResearchers face some well known and frequently rehearsed difficulties when analyzing the dataon reported crime for different countries, cultures, or jurisdictions. Cross-national comparisonspresent their own problems.

■One such problem is varying definitions. Different legal codes define crimes in different ways,so that the set of acts that constitute a given crime type in one country may not be identical tothe set of acts to which the same label is applied in another.

■Recording practices are another problem. Different police forces, in particular, have differentrules for when an event should be recorded as a crime and when it should not. For example, insome countries the police are said to be very careful about recording every theft of a bicycle,whereas in other countries, the police may not record every bicycle theft because the depart-ment has a higher workload of more serious crimes, has fewer resources, and may be lessorganized.

■A third difficulty is that of operating practices. In some countries, the main decisions regard-ing a case are made at the prosecution stage, so that many cases, especially trivial ones, donot appear in the records of those countries until that stage. Countries also vary with respect tocommon law and codified, civil law; therefore, comparisons between the raw numbers of differ-ent systems can be risky unless the person making the comparison is familiar with the detailsof the operations of the system and their implications for statistical recording.

■There also are large factual inequalities among countries as to their population size, popula-tion makeup (for example, percent urban and rural, and percent older than 60 and younger than25), and the size of the crime problem. Even in comparisons that take these differences intoaccount, hidden factors will affect the outcome.

■Finally, there are a set of problems specifically associated with recorded crime. Thenumbers provided by governments are regarded as indicators of the input into, and thereforethe workload of, the criminal justice system. They are not regarded as accurate statements asto the actual prevalence and incidence of a given crime type in a given jurisdiction, althoughthey may be that. Further information would be needed to validate the figures. It is generalcriminological wisdom that the less serious the crime type, the more questionable the officiallyrecorded figures. It also is widely accepted that victim surveys provide more valid data inregard to the incidence of most types of crime.

All of these issues are good reasons to construct measures of crime that are based on multiplesources of data.

Source: Burnham, R.W., “A First Analysis of the United Nations Data Set on Crime Trends and the Operationsof Criminal Justice Systems,” unpublished final report for NIJ grant 97–MU–CX–0002.

HEUNI’s full report also contains an in-depth discussion of these topics. See Kangaspunta K., Joutsen M., and Ollus N. (ed.), Crime and Criminal Justice in Europe and North America, 1990–1994, HEUNI PublicationNo. 32, (Helsinki, 1998).

Piecing Together the Cross-National Crime Puzzle36

The Pitfalls of Using InternationalStatisticsThe dangers of using official report-ed statistics as a reflection of crimewithin one’s own country are welldocumented. Reported crime is notthe same as actual crime, and statis-tics are collected for administrativepurposes, not to satisfy researchinterests. The vagaries of changinglaws and statistical practice and theidiosyncrasies of defining criminalincidents make it difficult to drawconclusions when comparing statis-tics from different areas or differenttimes. In addition, the “traditional”offenses (those usually noted in thestatistics) may not necessarily havethe greatest economic and socialconsequences for society.

International comparisons are evenmore rife with misunderstandings,as has been repeated throughoutdiscussions about the UN surveys.The major problems with compar-ing international crime data are dif-ferences in laws and in definitions oflegal terms, improper statistical clas-sifications, procedural differencesamong countries, ambiguous codingstructures, and differences in theunits of count used.

HEUNI researchers have attemptedto lessen any misinterpretation byusing data from different sources tosee if they point in the same direc-tion and by measuring differentdimensions of the same phenome-non. HEUNI’s cross-national analy-sis is based on more than the officialreported crime statistics of countriesresponding to the Fifth UN Survey.It also is based on results of theICVS—fully standardized victimiza-tion surveys on the general public’sexperiences with crime, which werecarried out in more than 60 coun-

tries. Added to these data are datafrom organizations such as theWorld Health Organization, theCenters for Disease Control andPrevention, and TransparencyInternational, an organization thatcollects data on corruption.4

The use of different data sourcesyielded a measure of crime that ismore scientifically robust than thosepreviously available. This measurecan be used reliably for comparativepurposes across countries in Europeand North America.

Sociological andBehavioral FactorsRelated to CrimeCrime levels can be interpreted asthe convergence of sufficient num-bers of motivated offenders, rela-tively weak mechanisms of socialcontrol, and the presence of suitabletargets of crime.5 For this study, theresearchers analyzed data related to:(1) motivational factors, includinglevel of affluence, alcohol consump-tion (both beer and strong alcohol),and a concept referred to as strain,or a person’s reaction to a shortfallin the achievement of goals (usuallysocioeconomic goals); and (2) op-portunity factors, including vehicleownership, handgun ownership,patterns of outdoor recreation,proportion of single-person households, and strength ofinformal social control.

Motivational Factors. Accord-ing to conventional criminologicaltheories, crime is related to econom-ic and/or social deprivation orinequalities. A key concept is strain,which the HEUNI team defined asthe number of people in a countryfor whom criminal activities mightbe economically rewarding and for

whom the involvement in criminalactivities is a viable option. As ameasure of strain, the team used therate of young people (16 to 29years) who are dissatisfied with theirincome and/or who are unemployed(per 100,000 inhabitants).

Opportunity Factors. Accordingto opportunity theory, the level ofcrime also is determined by thepresence of suitable targets of crimeand the extent of informal socialcontrol. A well documented exampleis the relationship between vehicleownership and vehicle-relatedcrime. Included in HEUNI’s recentanalyses were known risk factors,such as the frequency of outdoorvisits for recreational purposes,number of one-person households,composition of housing (apartmentbuildings or detached houses), andownership rates of motor vehicles,motorcycles, and bicycles. In addi-tion to these factors, the researchersstudied the prevalence of the posses-sion of handguns as a possible facili-tator of violent crime and the preva-lence of antiburglary device use.

Urbanization and modernizationhave been linked to high crime ratesdue to lower levels of social controlin urban areas.6 In previous analysesof ICVS data, the level of victimiza-tion by crime was strongly related tothe proportion of the populationliving in a large city. For its study,HEUNI used data on urbanizationtaken from the UN Compendiumon Human Settlement, which reflectthe proportion of the national pop-ulation living in settlements of20,000 inhabitants or more.

The Interplay of Motivationaland Opportunity Factors.Crime in societies is determined bythe interplay between motivationaland opportunity factors.

National Institute of Justice Journal ■ January 200037

Motivational factors can be seen asdeterminants of the “demand side”of national crime markets. To theextent that motivational factors aremore prevalent in a country, therewill be more potential offenderslooking for opportunities to offend.Structural characteristics that pro-vide viable opportunities for crimecan be seen as the “supply side” ofthe crime market. Owners of expen-sive cars and consumer goods arethe reluctant “suppliers” of opportu-nities for crime. In countries wheresuitable targets of crime are plentyand the level of social control isreduced, there are more potentialvictims of crime.

Affluence can be considered both amotivational and an opportunityfactor. It acts as both an importantinhibiting factor of certain forms ofcrime as well as a catalyst of others.In more affluent countries, there isless “demand” for crime—there areproportionately fewer individualswho are motivated to commitcrimes or who are looking for crim-inal opportunities. Important moti-vational factors, such as incomeinequalities, dissatisfaction withincome, and unemployment, tend tobe lower in more affluent countries.If levels of affluence rise—and if thenewly acquired wealth is evenlyspread—the pool of motivatedoffenders in a given society decreas-es. This trend will contribute to areduction in the level of crime. Atthe same time, affluence goestogether with the ownership ofcommodities that can be stolen withrelative ease, and also with a moreoutgoing lifestyle, which increasesexposure to criminal victimizationby strangers. Higher prosperity willinvite higher levels of opportunisticforms of crime.

However, there are no straightfor-ward, linear relationships betweenaffluence and crime. The dynamicsof crime in the big picture are fur-ther complicated by the increaseduse of sophisticated security mea-sures by potential victims in moreaffluent, high-crime nations. Thesemeasures reduce opportunities forcrime and, therefore, inhibit theoccurrence of certain types of prop-erty crime.

Comparing Crime inThree RegionsTo understand how and why crimevaries across countries, researchersexamined data from 49 countries inthree regions—Central and EasternEurope, Western Europe, and NorthAmerica. The researchers found thatthe most important predictor ofhigh crime rates was the percentageof the population who were youngmales, ages 16 to 29, who were dis-satisfied with their income or wereunemployed. The results of theHEUNI team’s analysis concerningthe three regions can be summed upas follows.

Crime in Central and EasternEurope. The motivation to offendappears greater in the countries ofCentral and Eastern Europe, whichare in transition to a market-basedeconomy, than in Western Europe.In Central and Eastern Europe,there are more male adolescentsexperiencing strain, and socioeco-nomic deprivation and alcoholabuse appear to help in the forma-tion of a breeding ground for differ-ent forms of violent crime. Assaults,homicides, and robberies appear tobe more prevalent in countrieswhere many young males experiencethe strain of unfulfilled aspirations.

Also, alcohol use is endemic in thesecountries.

Corruption also appears to be muchmore common in Central andEastern European countries than inNorth America and NorthernEurope. Corruption is related to alack of transparency and account-ability in the public domain, charac-teristics that are common amongthe developing countries of Centraland Eastern Europe. Corruption ingovernment circles appears to belower in the countries where eco-nomic and political restructuring isrelatively advanced—for example, inEstonia, Hungary, Poland, andSlovenia. In fact, the level of corrup-tion in some of these countries islower than in some Western coun-tries. These findings are encourag-ing. If the restructuring in othercountries in the region continues,the long-term prospects for decreas-ing levels of corruption seem fairlygood.

In the short term, the economic crisis in the Russian Federation and in some other countries in theregion may exacerbate existing economic and social problems. Inthe long term, the overall economicprospects might be better, but thisprobably will not reduce the preva-lence of strain among the lowersocial strata. Socioeconomicinequality is growing, and the ratesof unemployment will probablyremain high for many years tocome.

In most of the countries in transi-tion, people in urban areas typicallylive in apartments and car owner-ship is still relatively rare. These fac-tors may have so far inhibited fur-ther increases of property crimes.During the past 10 years, affluencehas increased in most Central and

Eastern European countries, partic-ularly in Hungary, Poland, Slovenia,and the Baltic states. If the grossnational product of these countries,which are among the first candi-dates for entry into the EuropeanUnion, continues to increase,vehicle-related crimes and someforms of petty crimes are likely toincrease. Household burglary ratesalso are likely to increase if house-holds start to possess more expen-sive commodities but do not investin antiburglary devices.

Eventually, investments in self-protection against car theft and burglary will increase and the ratesof property crimes will stabilize.

However, if strain among adoles-cents remains prevalent due to highunemployment, some of the offend-ers prevented by improved securityfrom committing less serious prop-erty crime may be desperate enoughto switch to more risky, violentforms of property crime (street rob-bery, car jacking, and householdrobbery). The easy availability offirearms in many countries in theregion will facilitate this. The crimeprofile of the countries in transitionmay start to resemble that of SouthAfrica or the more affluent LatinAmerican countries, where propertyis relatively well protected but rob-bery rates are high.

To sum up, the level of crime in thecountries in transition is relativelyhigh, and the overall criminologicaloutlook is fairly bleak. Even if theyovercome their current economicproblems, the rates of violent crimesprobably will remain high due tohigh levels of unemploymentamong young males and the highconsumption of strong alcohol.

Crime in Western Europe. Thecrime situation in the more indus-trialized and affluent nations ofWestern Europe must be understoodin terms of special opportunitystructures. Countries such asEngland and Wales that rely largelyon cars for transportation experi-ence high rates of vehicle-relatedcrimes. Countries such as TheNetherlands and Sweden, wherebicycles are common, experiencehigh rates of bicycle theft. Countrieswhere more people live in homesrather than apartments experiencehigh burglary rates. As protectionagainst car theft, theft from cars,bicycle theft, and household bur-glaries has increased, the overalllevel of property crimes has declined.Since the level of self-protectioncontinues to increase, crime ratesare likely to continue to fall.

Paradoxically, crimes of violence—particularly violent juvenile crime—show an upward trend in severalmember states of the EuropeanUnion. Street robberies might beincreasing as a result of displace-ment due to improved protection ofproperty. The emergence of an eth-nic underclass in the larger cities ofWestern Europe also might be con-tributing to violence because strainamong some ethnic parts of theurban population might be rising.The main challenge for WesternEuropean countries seems to be thesocial and economic integration ofyoung immigrants in the urbanareas.

Piecing Together the Cross-National Crime Puzzle38

National Institute of Justice Journal ■ January 200039

The outgoing lifestyle of young people and their use of alcohol anddrugs might also be a causal factorbehind juvenile violent crime. Oneof Western Europe’s main assets isthe relatively low level of handgunownership, which probably is animportant factor inhibiting homi-cides.

Relatively low levels of manifest cor-ruption by public officials appear tobe typical of affluent nations withstable democratic traditions. Thisrelationship also can be understoodin terms of criminal opportunitiesand motivations. In open democra-cies with relatively unregulated mar-kets, there are fewer opportunitiesfor public officials to require bribesfor their services. In such societies,public officials receive better salariesthan those in countries in transi-tion, and the norms against corrup-tion are stronger and more generallyshared.

Crime in North America. Since1988, the level of crime in theUnited States and Canada hasdeclined, according to both ICVSand police data. The level of self-protection against crime is high, andthe level of strain appears to be rela-tively low.

Crime in the United States differsless from Canada, the UnitedKingdom, and The Netherlandsthan is commonly assumed.7 Boththe United States and Canada haverelatively high levels of burglariesand car-related crimes in urban set-tings, but conventional crime andcorruption in the United States arenot exceptionally high. The mostimportant difference appears to bethe high level of homicides and rob-beries, which in the United Statesoften involve guns. The most proba-ble cause of this deviation from theEuropean pattern is the exceptional-ly high rate of gun ownership in theUnited States.

Framing theInternational Debateon Crime ReductionDespite the cautions needed whenworking with the available interna-tional data, the HEUNI team’sanalyses indicate that we are begin-ning to piece together some impor-tant parts of the intricate interna-tional puzzle of crime.

Crime indicators based on a combi-nation of police statistics and surveyfindings proved to be useful for ana-lyzing the determinants of crime.National crime profiles can beunderstood as the outcome of thedynamic interplay between motiva-tional and opportunity factors.

High levels of crime are found inboth poor and rich countries.However, the factors responsible forhigh crime rates differ acrossregions.

In Central and Eastern Europe,much crime is demand-driven—crimes are committed by youngmales as a means to survive in direeconomic situations.

In Western Europe and NorthAmerica, much crime is supply-driven—the prevalence and shape of crime are related to specialopportunity-structures (the avail-ability of targets and levels of socialcontrol and self-protection). Thisconclusion has important implica-tions for criminal justice policy.

Selected Country-Specific FindingsProperty Crimes. The United States, Canada, and the Czech Republic rankamong the highest in burglary, motor vehicle theft, and petty crimes.* Other coun-tries with relatively high levels of these types of crimes are Bulgaria, Estonia, andSlovakia.

Countries with relatively low levels of property crimes are Belarus, Norway,Switzerland, and the former Yugoslav Republic of Macedonia.

Violence. Serious violent crimes tend to be relatively more prevalent in the countries of the former Soviet Union, such as Estonia, Kazakhstan, Kyrgyzstan, and the Russian Federation.

The United States stands out with a high score on serious violence, which contrastswith much lower levels in Canada and the Western European countries.

Countries with low levels of violence tend to be found in Western Europe. Hungaryand the former Yugoslav Republic of Macedonia also have relatively low levels.

Corruption. High levels of corruption tend to be concentrated in Central, Eastern,and Southern Europe.

* Editor’s Note: Another study based on victim surveys and police statistics has concluded that

crime rates are much higher in England than in the United States. The authors of the study

report that in 1995, the results of a victim survey indicated that England’s burglary rate was

almost double America’s and its motor vehicle theft rate was more than double America’s. See

Langan, Patrick A., and Farrington, David P., Crime and Justice in the United States and in

England and Wales, 1981–96, Washington, D.C.: U.S. Department of Justice, Bureau of Justice

Statistics, 1998, NCJ 169284.

Piecing Together the Cross-National Crime Puzzle40

Policies that have worked in onecountry or region may be less usefulor even counterproductive in coun-tries where crime is determined byanother set of factors. In the inter-national debate on best practices incrime reduction, whether in theframework of the United Nationscrime program or elsewhere, differ-entiations in policies ought to bemade on the basis of criminologi-cally informed crime profiles.

NCJ 180081

Notes1. HEUNI is the European link

in the network of institutesoperating within the frameworkof the United Nations CrimePreven-tion and Criminal JusticeProgramme. HEUNI, which is based in Helsinki, Finland,was established through anagreement between the UnitedNations and the government of Finland, signed December 23,1981.

2. The UN has gathered informa-tion on crime and criminal jus-tice from its member states since1975. The first survey coveredthe period 1970 through 1975.In these surveys the informationon crime is based on nationalpolice statistics and reflects thenumbers of offenses recorded bythe police. HEUNI has carriedout the European and NorthAmerican analysis of the second,third, fourth, and fifth UnitedNations Survey on Crime Trendsand the Operation of CriminalJustice Systems.

3. For its most recent analysis,HEUNI used crime data on 40Western, Central, and EasternEuropean countries as well as on Canada and the UnitedStates. Researchers studied eightdifferent types of crime: nonfa-tal violence (assaults and rob-beries), homicides, serious vio-lence (a combination of nonfatalviolence and homicides), bur-glary, violence against women

(sexual violence and assaults),vehicle crimes (theft of andfrom cars), corruption, andpetty crimes (bicycle theft,motorcycle theft, noncontactpersonal theft, car vandalism,nonviolent sexual incidents,and threats).

4. Previous UN surveys reliedalmost exclusively on UN surveydata. The analyses of the FifthSurvey differ considerably fromprevious ones because addition-al sources of comparative datawere used. The HEUNI researchteam relied on a database calledthe Crime Guide, which consistsof data from various sources,including the Fifth UN Survey,the International Crime VictimSurvey, health and mortality statistics collected by the WorldHealth Organization and theCenters for Disease Control and Prevention, and data oncorruption from TransparencyInternational and the Inter-national Institute for Manage-ment Development, which col-lects information on improperpractices in the public sectorand in the workplace.

The Crime Guide databasedraws data from studies offirearm ownership, alcohol con-sumption, substance abuse, gen-der equality, urbanization, andemployment and compensation.

5. Cohen and Felson, “SocialChange and Crime Rate Trends:A Routine Activity Approach,”American Sociological Review,44(1979): 588–608; van Dijk,J.J.M., “Opportunities for Crime:A Test of the Rational-

For More InformationThis article is based on Crime and Criminal Justice in Europe and North America,1990–1994, edited by Kristiina Kangaspunta, Matti Joutsen, and Natalia Ollus(Helsinki, 1998). The report provides a full discussion of the HEUNI team’s researchand methodology, including background on the Fifth UN Survey and analyses of dif-ferent countries’ types of crime; a description of criminal justice system resources,case flow, and performance; and a review of sanctions.

HEUNI also has published a volume of criminal justice profiles of all European andNorth American countries with an independent criminal justice system. HEUNIresearch reports are available on its Web site at http://www.vn.fi/om/suomi/heuni/.

For more information on the UN Survey and the International Crime Victim Surveysee also: Graeme Newman (ed.), Global Report on Crime and Justice, United NationsCentre for International Crime Prevention, Oxford University Press, New York, Oxford,1999.

National Institute of Justice Journal ■ January 200041

Interactionist Model,” report atthe Eleventh CriminologicalColloquium of the Council ofEurope, Strasbourg, November28–30, 1994.

6. Shelley, Louise, Crime andModernisation: The Impact of Industrialisation andUrbanisation on Crime,Carbondale, IL: SouthernIllinois University Press, 1981.

7. Editor’s Note: Other researchbased on victim surveys andpolice statistics has concludedthat crime rates are much higherin England than in the UnitedStates. (See Langan, Patrick A.,

and Farrington, David P., Crimeand Justice in the United Statesand in England and Wales,1981–96, Washington, D.C.: U.S.Department of Justice, Bureauof Justice Statistics, 1998, NCJ169284.) These seemingly con-

tradictory findings provide atleast suggestive reinforcementfor the authors’ argument forthe use of multiple sources ofdata when making cross-countrycomparisons.

AcknowledgmentsHEUNI’s analyses of the Fifth UN Survey were carried out by an international expertworking group consisting of Jan van Dijk, Kristiina Kangaspunta, Carolyn Block of theUnited States, Matti Joutsen of HEUNI, Andre Kuhn of Switzerland, and Ineke HaenMarshall of The Netherlands and United States. Adam Bouloukos of the UN’s Centrefor International Crime Prevention and Ugljesa Zvekic of the UN’s Interregional Crimeand Justice Research Institute assisted the group in its work.

`