Official statistics is pretty much a great favourite -certainly with teachers, if not students, mainly because there’s a fair amount of material and some good juicy issues to get stuck into. Of course it may not come up. But then again….

Official Statistics and Crime
Many sociologists use official statistics in their research.  There are many reasons why using official statistics can be useful: they are cheap, readily available, and provide detailed quantitative data which is reliable and often representative.  Official statistics also provide data for the whole country.  Given the scarcity of resources and the expense of funding research, sociologists would be unwise to disregard a cheap and easily available source of data.  However, some sociologists, such as Barry Hindess, have argued that official statistics on crime, do have serious deficiencies.

There are several reasons for these deficiencies. 

Firstly, there are many reasons why the public may not report all crimes to the police: they may not realise they’ve been a victim, embarrassment, they may implicate themselves in a criminal act, etc.

Secondly, there are also many reasons why the police may not take action against all offences which are known to them.  Sociological studies of the police, such as that conducted by Simon Holdaway in the 1980s, demonstrate that the police simply cannot take action against all offences which they identify, and therefore have to prioritize their activities.  Holdaway shows how the police develop an occupational culture which emphasises the notion of police discretion, and how officers are socialised into a particular set of norms and values.  The concept of police discretion implies that police officers have discretion – that is, they have the power to turn a ‘blind eye’ to offences when they feel that an offence is too minor to bother taking further action, or perhaps when they feel that the probable outcome will not warrant the effort that will be required on their part.  The notion of police discretion, Holdaway argues, is something which officers learn through the occupational culture.  According to Holdaway, the occupational culture of policing, puts great value on action and aggression.  This can lead police officers to focus their activities on particular types of offences, e.g. violent crime at the expense of traffic offences, armed robbery rather than shoplifting.

Dick Hobbs, in his study ‘Doing the Business’, argues that there was a reciprocal relationship between police and people in the east end of London.  Crime was a way of life for many of the people, accepted as an inevitable feature and a way of getting by.  So too, was getting arrested.  For police and people, crime was rather like a game.  However, both police and ‘criminals’ would bargain and make deals; perhaps a criminal would trade information with a police officer in return for the officer turning a blind eye towards some offences carried out by the criminal.  Hobbs argues that both police and criminals needed each other, and they evolved complex relationships which involved trading and bartering for information and favours.

The Dark Figure/Iceberg Approach
As a result of these criticisms, sociologists and social historians have talked of the ‘dark figure’ of crime, whilst others have used the metaphor of the iceberg to explain crime statistics. Measured levels of crime, are only levels of reported crime – there is always a ‘dark figure’ of unreported crime.  As with icebergs, a small proportion of crime is visible, but the bulk remains hidden from our view.

Valid Data?

These insights can lead to a comprehensive reassessment of the validity of official statistics on crime.  As Hindess argues, using official statistics can tend to push researchers further towards a positivist approach, and they can treat the statistical data as if it is revealing the social laws governing crime and deviance.  The criticisms examined above, suggest that official crime statistics need to be interpreted much more carefully.  The criticisms demonstrate another way in which we can say that crime is socially constructed; official statistics on crime are one of the ways in which crime is socially constructed.

Implications for theories of crime

The picture of crime that we gain from official statistics can now be exposed to closer scrutiny.  The picture of crime presented from official statistics suggests that the typical criminal is a young, working class male.  Many sociological theories have indeed taken this idea as an unexamined assumption.  However, if official statistics lack validity, and are not providing a true picture of crime, it may be that many of the sociological theories of crime require revision.  It may be that sociologists need to consider the view that crime is not mainly committed by young working class males.  Maybe males (and females?) of all ages, and from many different classes are involved?  Some of the theories examined previously can indeed help support this idea – labelling theory for instance, and some Marxist based approaches – but not all – would lend themselves to this sort of approach.

Other ways of measuring crime – Victim surveys and Self-Report Studies

The critique of official statistics has led sociologists to develop other methods to research crime.  Victim studies – the most famous of which have been the British Crime Surveys carried out by the Home Office – aim to identify the extent to which a representative sample of people have been victims of crimes.  Self-Report studies in contrast, aim to investigate the characteristics of offenders by getting a sample group to reveal details of offences committed.

Victim(ization) Studies – The British Crime Surveys (BCS)

The most well-known victim study in the UK is the BCS, carried out by the Home Office since 1981.  The BCS uses a very large sample –in 1996 it included 16,500 respondents, and aims to investigate how often individuals have been victims of crime within the last year prior to responding.  Respondents answer a detailed questionnaire.

The results of the BCS (there have been seven surveys to date) indicate that official crime statistics are unreliable.  Researchers can compare the BCS data to official crime statistics to check the validity of the latter.  BCS data may reveal either more or less offending in particular categories, implying that an offence is being either under reported or that it is being reported fairly accurately.  The BCS indicates that official statistics may vary in their accuracy; for instance, most vehicle thefts are reported (97%) but other offences vary much more.  BCS data indicates only about 57% of robberies and 26% of vandalism was reported in 1997, and overall it indicated that only about 44% of crime is reported.  Recent BCS data indicates that since 1981 levels of crime have increased overall, although more recently there have been reductions in some types of crime.  Home Office researchers also claim though, that the police statistics can exaggerate the extent of crime, in spite of the fact that overall the BCS data indicates an under reporting of crime.  This can be explained as being the outcome of police decisions to take action on particular types of offences.  All in all then, the BCS data seems to indicate that official statistics on crime do not provide a valid picture of the extent of crime, and overall they may underestimate the trend. 

Self-Report Studies

Self-report studies provide yet another insight into the extent and nature of crime.  Importantly, self-report studies indicate that offending is not confined to the lower socio-economic classes.  Various studies using self-reporting indicate that anything between 50% and 90% of the population have admitted to behaviour which if observed and acted upon, could see the respondent brought before a court of law.  Steven Box suggests that this data can be interpreted as implying that there is a selective approach to law enforcement, but in reality the working class are not more criminal than the middle class – to any significant extent.  Another researcher, Gold,  (1966, but still an interesting finding) has estimated that the difference between working class and middle class youth offending can be expressed as a ratio of 1.5:1 – hardly a huge difference, and similar patterns emerged when examining gender and ethnicity.  On this point though, researchers disagree – a study by Graham and Bowling in 1995, argued that working class youths did commit more crime than their middle class peers.  One conclusion to be made is that none of the methods discussed here will provide foolproof data.  However, this should not lead to despair – the data is useful, but it has to be used cautiously.

Nevertheless, there seems to be a range of good data confirming the notion of selective law enforcement.  Box however, does not suggest that we should jump to the conclusion that selective law enforcement is the result of a police conspiracy against the working class.  The processes involved are much more complex, and the biases arise from the constraints acting against the way the police force operates as well as police culture (see Holdaway, above) and ideologies – the last of which are not just confined to the police force!  A final point confirming this cautious approach is the recognition of what the police actually do.  As Moore and Sinclair (also supported by Sanderson 1994) point out, only 8% of arrests in the UK are initiated by the police – the rest come from complaints made by members of the public.  There is bias, but not just in the police force.


Official statistics have many pitfalls.  They have been demonstrated to lack validity and reliability.  Other methods, such as victim and self-report studies, whilst they may seem to boost validity, provide no guarantees of validity or reliability.  However, comparing the results of data from different sources – official statistics, victim studies, and self-report studies, can provide important insights into both the nature and the extent of crime, and does challenge preconceived ideas. In the area of crime, the study of official statistics has also led sociologists to reconsider theories of crime, for if there is selective law enforcement, crime is not to be seen as restricted to particular social groups, such as young, working class males. 

Job board

Wake up with today’s latest teaching resources & student support

Subscribe to our daily digest and get today’s content delivered fresh to your inbox every morning.

Or follow us

Explore tutor2u