Compiling and comparing International Crime Statistics

It should be noted that the crime statistics reported to the United Nations in the context of its various surveys on crime levels and criminal justice trends are incidents of victimization that have been reported to the authorities in any given country. That means that this data is subject to the same problems of accuracy as all official crime data. The variety of potential problems with recorded crime statistics is illustrated in the diagram below.


A Actual events
Ae Deviations in whether an event is considered a crime or not, e.g.: different definitions used in different jurisdictions.
P Police recorded crime
Pe Errors that accumulate along the police recording process
e.g.: data entry errors at the source and at intermediate stages of the recording process, aggregation errors, interpretation errors, missing regions or double count, transmission errors (e.g. lost mail), etc.
V Victim reported crime
Ve Errors that accumulate along the victim reporting process
e.g.: victimless crimes, data entry at the source and at intermediate stages of the reporting process, interpretation errors, sampling errors, statistical errors, transmission errors, etc.
1 Crime is recorded by the police and reported by victim
2 Crime is recorded by the police but not reported by victim
3 Crime is reported by victim but not recorded by the police
4 A crime happens but is neither recorded by the police nor reported by the victim
5 Victim reports a crime that never happened (e.g for purposes of insurance fraud) -- in reality this area is much smaller than depicted in the schematic
6 Police records a crime that never happened (e.g. to attract a higher budget) -- in reality this area is much smaller than depicted in the schematic

The diagram shows that only some crimes are reported by victims to the police, who may only record a portion of reported incidents in the official data. Generally, more serious crimes, such as murder or serious armed robbery, are likely to be recorded. Although in the case of some serious crimes, such as rape, where reporting may be difficult for the victim given the context of the crime, the official data reflects only a portion of actual cases.

These problems do not, however, apply to data that reflects the processing of cases through the criminal justice system. Thus, any figures given with respect to the number of prosecutions and/or convictions as well as the number of prisoners in any jurisdiction should generally be regarded (although again there may be exceptions) as an accurate reflection of the reality in any society.

The presentation of large amounts of reported crime data in the same format, as is the case with the various country results reported on this web site, raises questions about the possibility of comparing reported crime levels in different countries. In fact, experts generally regard cross-country comparisons of crime statistics as being characterized by methodological difficulties for the following three reasons:

  • Different definitions for specific crime types in different countries: The category in which any incident of victimization is recorded relies on the legal definition of crime in any country. Should that definition be different, and indeed this is often the case, comparisons will not in fact be made of exactly the same crime type. This is particularly the case in crimes that require some discretion from a police officer or relevant authority when they are identified. For example, the definitional difference between serious or common assault in different legal jurisdictions may be different, and this will be reflected in the total number of incidents recorded.
  • Different levels of reporting and traditions of policing: Different societies across the globe have been shown to have different levels of reporting of criminal incidents. This relates closely to levels of development in any society, most clearly reflected in accessibility to the police. Thus factors such as the number of police stations or telephones in any society impact upon reporting levels. The level of insurance coverage in any community is also a key indicator of the likelihood of citizens approaching the police as their claim for compensation may require such notification. In addition, in societies where the police are or have been mistrusted by the population, most specifically during periods of authoritarian rule, reporting levels are likely to be lower than in cases where the police are regarded as important members of the community.
  • Different social, economic and political contexts: Comparing crime data from societies that are fundamentally different may ignore key issues present in the society that impact upon levels of reporting. For example, different social norms in some societies may make it almost impossible for women to report cases of rape or sexual abuse, while in others, women are encouraged to come forward.

These factors, while alerting the reader to the potential pitfalls of comparisons, apply more to some crimes than others. In selected cases, most notably homicide, country to country comparisons are safer, although may still be subject to the drawbacks outlined above. In the case of some categories of violent crime - such as rape or assault - country to country comparisons may simply be unreliable and misleading.

Given these problems, the question may be asked why there is a requirement to bring together a wealth of statistical data on criminal justice issues from a variety of jurisdictions. It should be emphasized here that the main purpose of the UN survey is not to measure the exact amount of crime that exists in the world or to compare countries, but rather to provide an accounting of crime and governmental responses to it. This aim of the survey is enhanced with an increasing number of sweeps that allow the emergence of a clearer picture of trends in individual societies.