There is not one single source of criminal justice data - they must be brought together from a variety of sources. The main crime and justice statistics come from two types of collections, administrative and survey, and we need to use both types of information to help inform our understanding of the level and effects of crime in the community.
Criminal justice agencies keep records of their workflow at different stages. For example police keep incident records, courts record the details of cases and their disposition, and corrections agencies have details of the offenders in their charge. Most basic information comes from these administrative collections, which have the advantage of covering the whole population that comes into contact with the criminal justice system, and remaining relatively stable in terms of collections and production over time.
There are limitations to these data, however, including comparability across agency and jurisdictional collections. Most of the data have been collated at a national level only relatively recently - recorded crime from police records since 1996, prisoners since 1983, and all criminal courts since 2001. The collections are not all based on the same unit of measurement; for example, police record details about victims, courts record cases, and corrections agencies record information about individual offenders.
Although there has been much improvement, definitions and collecting methods are not always uniform across jurisdictions, and recording quality may be an issue. It can take time to reach agreement at a national level on key issues including definitions of new and emerging offences. Often more detailed information about crime and justice is available at a jurisdictional level, even when it is not possible to produce national statistics.
Not all crimes are reported to police - this is believed to vary from a low of 20 percent for sexual assaults to a high of 95 percent for motor vehicle thefts. This is one of the main reasons that the other main type of data collection, surveys, is undertaken. Crime victimisation surveys have the advantage of asking the same questions in the same way across the whole of the sample population. These answers are then recorded in a similarly uniform way so that the information they provide is reliable and comparable.
Crime victimisation surveys are believed to provide a more comprehensive picture of actual crime rates in society. Surveys are expensive, however, so they tend to be either one-off or infrequent. It is not always valid to extrapolate from a sample to the whole population, however, and all sample surveys have a certain amount of error. Surveys used in this publication this year include the Australian component of the International Crime Victimisation Survey and a national farm crime survey, both conducted by the Australian Institute of Criminology.