Mainstreaming Methodology for Estimating Costs of Crime
The total volume of crime is never known for sure. There are three widely used data sources each of which has strengths and weaknesses.
Recorded crime
This measures the number of offences recorded by the police, or some other agency such as a statistics bureau or a court. Typically data from this source are plentiful and well-disaggregated by offence type, geographical area etc. But this source produces data that may substantially underestimate the true volume of offending. Reluctance of victims to report offences to the police and variations in the recording standards applied by the police are the two most obvious sources of inaccuracy.
Victim surveys
Responses to victim surveys are increasingly being used as a more reliable measure than recorded crime statistics of the incidence and prevalence of certain sorts of offences. But good, large-scale, nationally and locally representative surveys are very costly to conduct and are comparatively rare. Those which are conducted often exclude certain groups, such as the under-16’s or those living in institutions such as halls of residence and this can bias results to some degree.
International victim surveys
These use a harmonised methodology but often have a small sample size and are sometimes constrained by methodological weaknesses: for example they may contain a high proportion of interviews from the country’s capital city and may rely on telephone responses rather than personal interviews. They are very useful for certain purposes but need to be dealt with cautiously. For further discussion of the contribution data from these surveys can make see the note on: Using international victim surveys to estimate victim losses from crime
Use of Multipliers
A method that is employed quite widely in cost of crime studies is the use of ‘multipliers’ to apply to recorded crime as a means of accounting for unrecorded crime. The procedure involves making estimates from victim surveys of the total number of offences experienced by the whole population. By comparing these estimates with the number of recorded crimes it is possible to estimate the number of unrecorded crimes. For further discussion of the link between victim survey findings and police-recorded crime see page 24 of Crime in England and Wales 2006-07
A potential weakness in this ‘multiplier’ procedure is that it assumes that the cost of an unrecorded crime is the same as that of a corresponding recorded crime. This is not really a satisfactory approach because there is plenty of evidence that the decision about whether to report an incident to the police itself reflects the seriousness of the incident. [link to paper on recorded and unrecorded crime]
