Chapter 2 The development of OASys offending in individual cases. This type of information can be very valuable in informing the work undertaken to minimise risk to the public or self- harm. Actuarial approaches to risk assessment have developed in recognition of the limitations of clinical methods. This approach has its roots in the insurance industry’s attempts to maximise the accuracy of its calculations of the probability of certain negative events. Here the assessment is based on statistical data derived from sample groups. For example:
insurance companies predict the chances of someone having a motoring accident by considering such factors as age, gender, driving convictions, number of years driving and occupation. They use this information to set premiums for different types of customer. The Probation and the Prison Services have had experience for some time of scales which predict the risk of reconviction in this way for example OGRS (Home Office Offender Group Reconviction Scale) and the SPP (Prison Service Sentence Planning Risk Predictor).
These scales have been developed from reconviction studies which have demonstrated that a small number of offender characteristics differentiate between those who do and those who do not re-offend. As with most areas of human behaviour, past behaviour is found to be the best predictor of future behaviour. Thus these scales mainly contain items relating to criminal history, age at first conviction, number of custodial sentences, number of previous convictions, type of offence, and age at offence. These factors when weighted to a certain formula along with gender and age band can be highly successful in predicting reconviction rates for groups of offenders.
There are, of course, limitations to the value of actuarial scales. It must be remembered they are based on group information and cannot predict with certainty that a particular offender (even if in a high-risk group) will commit a further offence. Actuarial scales are not very useful in predicting rare events. For example, the frequency of serious offences (eg serious violent or sexual assaults) is very low even amongst offenders with a history of such offences, so the scope for using actuarial methods for predicting risk of serious harm is limited.
Actuarial scales for predicting reconviction tend to be mainly based on previous convictions, which means they do not work well for first time offenders and those with little criminal history. However, the greatest drawback of the currently used actuarial scales is that they are based on historical information, which does not allow for positive change. The only way an offender’s score on OGRS 2 can diminish is through ageing. The scale takes no account of changes in personal circumstances.
This means that an offender is judged to be at the same level of risk of reconviction at the end of the period of supervision or custodial sentence as at the beginning, despite any input from prison or probation personnel. OASys Manual V2 © Crown Copyright 2002 Page 6