Home › Guide for postgraduate students › Mastering the literature

Statistics

statistics

Learning objectives

At the end of this module, you will be able to:

  • explain the value of statistics in legal research
  • discover statistics relating to crime and justice in Australia
  • find relevant statistics from overseas jurisdictions
  • explain basic mathodological problems in the use of crime and justice statistics.

What are statistics?

The term statistics is used in two ways. In its broad sense, statistics is the science of collecting, analysing and presenting numerical data. The term statistics is also used more narrowly to refer to data sets, the actual numbers which form the basis of further study and analysis.

Why use statistics?

Statistical evidence enables you to draw conclusions regarding economic, social and political conditions in the past and present. Such evidence provides a means of evaluating competing claims made by different experts regarding the effects of past decisions, policies or events. Statistical information is usually necessary if a hypothesis or theory is to be subject to rigorous testing.

Where legal research intersects with other fields, such as criminology or sociology, an understanding of basic statistics is crucial.

Finding Australian statistics

The major sorce of statistical information in this country is the Australian Bureau of Statistics (ABS). Most recent ABS publications are freely available from the ABS site. Older publications are available at most university libraries.

What does the ABS hold?

The ABS National Centre for Crime and Justice Statistics collects data on recorded crime from the police, criminal courts and correctional agencies in each State and Territory. Links to much of this information is available from the Crime and Justice page on the ABS site.

Major publications in the area of crime and criminology by the ABS include:

The ABS also runs more general social surveys with crime and justice components.

The ABS and unpublished data

In some cases, the published summary data available from the ABS site is not enough. Fortunately, the ABS is able to provide researchers with access to more detailed data sets through its Information Consultancy Service. The Service operates on a cost-recovery basis. Reasonable charges apply for researching requests, data extraction and manipulation, and any analysis undertaken by ABS staff. Information on contacting the Service is available from the ABS site.

For small-scale studies, the microdata contained in a Confidentialised Unit Record File (CURF) is often extremely useful. A CURF is a file of local responses to an ABS survey which has been confidentialised through the removal of information which might identify individuals or businesses. An example are the CURF files available for the National Crime and Safety Survey conducted by the ABS in 2002. Most universities have access to CURF data. Your university's Contact Officer is listed on the CURF Responsible and Contact Officers page. You should consult this person when seeking access to CURF material.

html icon

The things that you're liable ...

ABS figures shold not be confused with the actual incidence of particular offences. They represent no more than the best available estimates. Why is this? Take the example of the published figures regarding the number of suicides which occur each year. There is good reason to assume that the official suicide statistics considerably under-report the actual number of deaths due to deliberate self-harm.

Stop for a moment and consider the legal, ethical and practical factors at work. Can you see why the official figures for the annual number of suicides are necessarily incomplete?

What about offences such as murder? Can we really assume that the reported incidence of deliberate homicide is a reliable guide to the number of deaths from violence? Improvements in police procedures and forensic science mean that we are better informed regarding the murder rate than ever before, but difficulties still remain. Thousands of people disappear every year. We can never be certain how many of the unexplained disappearances are the result of undetected murders. Perfect crimes happen in real life as well as in crime fiction.

Becoming familiar with the data

If you decide to make use of the statistical information provided by the ABS, you need to become familiar with the methodological notes published by the Bureau. You should also read widely to ensure that you are familiar with what academics and commentators outside the ABS are saying about the Bureau's figures.

The ABS goal is to provide a high-quality source of statistical information. Despite this aim, there are limitations on the statistics that the ABS can provide. Resources are insufficient to provide statistical data on every conceivable subject. There are also a range of sampling and methodological problems which impose technical limits on the accracy of ABS figures.

Fortunately, there are dozens of other official and semi-official bodies which collect, analyse and publish high-quality data on crime and the justice system in Australia. These bodies include government agencies, libraries, private firms, market research companies and research institutes. Much of this research output is now available on the Web at no cost.

Australia-wide agencies

State and Territory Agencies

Most State and Territory governments have specialised agencies which collect, analyse and publish information on crime and justice. The Web sites of these agencies provide access to research reports and comprehensive statistical information on the operation of the criminal justice system in the States and Territories.

Finding international statistics

A good place to start is the Statistical Section of the World Criminal Justice Library Network (WCJLN). This site provides a near-comprehensive set of links to crime statistics and other criminal justice reports published in many countries. Another starting point is the Crime and Criminal Justice Statistics page at the Australian Institute of Criminology (AIC) site.

Some other useful links include:

Bear in mind that cross-border comparisons of the official statistics on crimes rates are often problematic due to differences from country to country in the classification of crimes, reporting rates and the completeness of police statistics. Published statistics indicate countless anomalies which cast doubt on the usefulness of cross-border comparisons. For this reason, such comparisons are often resticted to official murder rates or the the use of victimisation surveys.

UN Statistics

The United Nations has created a search engine UNData, which is intended to provides quick access to international statistics. There are also specialised UN agencies with interests in criminology including:

  • UNICRI: The United Nations Interregional Crime and Justice Research Institute carries out applied research in the area of criminology, including the International Criminal Victim Surveys (ICVS) program.
  • UNDOC: The United Nations Office on Drugs and Crime assists member states in the fight against illicit drugs, crime and terrorism. It also provides a clearing-house for international crime statistics. The UNDOC site includes data and analysis on statics on illicit drugs, crime and corruption, international data on illicit drug markets (including the World Drug Report) and an online database of legislation adopted by states to put into effect international drug control conventions.

Warning: official statistics ahead

Crime statistics are not simple facts. Several factors influence published statistics on crime and justice:

  • undetected crime. Many crimes, such as minor frauds, go undetected, even by the victim.
  • underreporting of crime by members of the public. A surprising number of detected crimes go unreported. This applies not only to "victimless" crimes, but even to many serious offiences. NSW police estimate that 35 per cent of assaults in the State go unreported.
  • overreporting. In a few instances, crimes may be over-reported. Some reported car thefts, for example, are almost certainly attempts at insurance fraud.
  • informal police practices. The police do not report all the crime that comes to their notice. When on the beat, individual constables have a certain amount of discretion in their work. In addition, policing practices differ between police commands. Incidents that might be recorded as crimes in one command might not be regarded as worth reporting in another.
  • methodological factors. Differences in offence classifications and methods of counting crimes mean that the crime statistics reported by the ABS do not always agree with those produced by State and Territory authorities.
  • defintional changes. Definitions of offences change over time. Techical changes in the wording of legislation can have a sudden impact on the recorded incidence of certain classess of offences.
  • changes in government and police policies. Sudden fluctuations in the number of recorded offences in specific categories are often the product of increased policing or new government priorities. These can have spectacular impacts on recorded statistics. Between 1997 and 2005, appearances in the NSW Children's Court fell by over 40%. Did juvenile crime in NSW really decline so dramatically after 1997? In fact, much of the fall in the number of defendents before the Children's Court was largely due to the decision to change the manner in which the police dealt with young offenders: many were subject to informal or formal cautions or diverted from the courts to juvenile conferences.

For more information on methodological issues in interpreting published statistics, see the eSKILLS Plus module on Statistics.

Child Abuse in NSW

Is the incidence of child abuse in NSW rising or falling? Before reading on, make a judgement based on what you know from media reports. Are you correct?

The reported incidence of child abuse in NSW has been rising by 8 per cent a year since 2000. It is doubtful, however, that the abuse of children in NSW has actually increased at such a rapid rate. During the last few years, most other social indicators in NSW have trended upwards (many types of child abuse being very strongly linked to poverty). What happened was that the NSW government decided in the late 1990s to make the reduction of child abuse a major priority. The official reporting system in NSW was strengthened, legal definitions of child abuse and neglect were widened, and community attitudes towards what constituted reportable abuse changed. All these factors combined to bring to light a great deal of crime that went unreported earlier.

Despite these factors, many observers have argued for a rising incidence of child abuse in NSW. Such claims are impossible to verify or disprove. The actual prevalence of child abuse in the State remains unknown, but (based on economic and social indicators) there is a strong likelihood is that the incidence is on the decline.

The mass media as a source of statistics

Be wary of isolated statistics which appear in news media, particularly those without attribution or those attributed to non-government organisations (NGOs) or pressure groups. Figures are sometimes taken out of context or misunderstood. At other times, organisations issue alarmist or exaggerated figures in order to attract attention and support. Another problem is the demand from journalists for "good copy". The UN Human Security Report (2005) quotes an unnamed UN official, who complains:

"When it comes to statistics ... numbers take on a life of their own, gaining acceptance through repetition ... Journalists, bowing to the pressure of editors, demand numbers, any number. Organizations feel compelled to supply them, lending false precision and spurious authority to many reports."

This observation was made in the context of international human rights abuses, but applies equally well to domestic reporting. Figures quoted in the mass media, popular books or magazines are not necessarily incorrect, but they need to be verified. Where no hard evidence exists, such spurious statistics should be discarded.

Summary

This module examined:

  • the value of statistics in legal research
  • finding statistics relating to crime and justice in Australia
  • locating relevant statistics from overseas jurisdictions
  • basic mathodological problems in the use of crime and justice statistics.