In the United States significant expenditures of resources have been made for post-licensing driver improvement programs. The components of these programs vary and may include driving skill enhancement such as defensive driving courses, individual or group counselling aimed at attitude change, warning letters, and pre-suspension hearings. The driver improvement activity is designed to increase knowledge and modify attitudes and in turn positively change behaviour which may then be reflected in reduced crash experience. ¹

Struckman-Johnson, Lund, Williams and Osborne reviewed the comparative effects of non-alcohol related driver improvement programs on crashes and traffic offences after some reviewers of well-controlled evaluations had suggested that some programs result in measurable reductions in violations but not crashes. ³

Possible explanations for the lack of correlation between the effect on offences and that on crashes include:

• Crashes are relatively rare events in comparison to offences. (Any effect on crashes will therefore be more difficult to detect statistically);

• The behaviour noted in offence statistics is generally under the control of a driver whereas crash involvement involves a combination of factors including the behaviour of other drivers and other environmental circumstances. It is therefore possible to directly target the given behaviour(s) underlying the offences but not so directly for crashes;

• Reductions in offences could be the result of manipulation of the traffic safety system by program participants. One example would be drivers who repeatedly enrol in programs to prevent offences being entered on their driving records (in the USA); or

• Reductions in offences detected could reflect real changes in the behaviour being cited but the behaviour may be unrelated or minimally related to crashes. Gagliardi ⁴ in 1977 showed that penalty notices are issued less often for offences that correlate most with crash experience. ⁵

The major findings of the review were:

• That driver improvement activities generally do result in reductions in offences;

• There is an unpredictable and sometimes increased effect on crashes even in the presence of reduction of offences; and

• The lack of correlation between offences and crashes is not explained by the limitations of the statistical analysis methods used or by the types of offences affected.

Examination of driver improvement interventions revealed no strong evidence for different effects related to characteristics such as direct versus indirect participant contact and group versus individual contact.

In summary, it appeared that the success of programs in reducing subsequent offences does not imply similar success in reducing future crashes. ⁶

A review of the literature found that various road safety authorities and researchers have been unable to find evidence of a link between car driving skills and road trauma. Driving skill deficiencies have been found relevant in less than 5% of crashes.

Post-licence driver training courses may be divided into three categories. These are:

• Defensive driving courses which aim to help any involved driver or rider avoid getting into critical situations;

• Advanced driving courses which aim to help any involved driver or rider cope with critical situations; and

• Driver improvement programs which specifically target accident or violation involved drivers.

Reviews of evaluation studies have found no evidence that either advanced or defensive driving courses reduce the accident involvement of those who attend.

Even when targeted at drivers over-involved in accidents no significant reductions in accident involvement results. Such training has occasionally been found to reduce traffic offence involvement but there is no reliable evidence that this translates into accident reductions or even persists in the long term.

It may be concluded that the road safety value of these particular programs is questionable.

Victorian studies of both novice and advanced driving courses also found no evidence of accident reductions.

A study of a mixed defensive/ advanced driver training course for novice drivers in Adelaide found no accident reduction benefits. This is consistent with evaluations of driver training/ education programs in New Zealand. Indeed, it has been found that following voluntary attendance at a skill-oriented, advanced driver training course in Sydney those trained recorded increases in traffic convictions, particularly for speeding. ⁸ This is a disturbing though not unexpected finding and is consistent with the findings of another Australian study and the findings of the largest driver training study ever undertaken (De Kalb County, USA).

In 1987 the Traffic Authority of New South Wales published a research report which reviewed a Californian Driver Improvement Programme trial published in 1984 by Helander. ^{9 10} The trial was predominantly focused on drivers who had been involved in a fatal crash, those with three accidents within one year or those with one crash but a moderate number of traffic offences. The NSW report concluded that though the benefit-cost results appeared favourable the reduced number of crashes would be very small - of the order of 40 to 50 per annum. Further, even if the benefits were achieved it would be difficult to statistically verify the program from NSW crash data. It was estimated that it would require the two possible program variations to operate for three to four years before sufficient data would be available to detect any measurable effect with a reasonable degree of confidence.

An article published in 1993 describes a cost-effective driver improvement treatment in the State of Pennsylvania. ¹¹ The intervention was applied when offenders reached the six demerit point level (the equivalent of 12 demerit points in Victoria). A written examination was found to result in significant reductions in accident and traffic offence rates relative to the previous treatment of behind-the-wheel licence retesting.

The examination and accompanying handbook conveyed the message that unsafe driving behaviours resulting in accidents and convictions reflect inappropriate choices for particular traffic scenarios, and communicated the importance of individuals choosing to avoid specific behaviours that increase accident risk through real-world examples of familiar problem driving situations.

Use of this approach also produced savings of at least $US150 000 per annum through the replacement of the equivalent of at least eight driver licence examiners with three driver safety examiners. In the future automated scoring and examination are expected to result in additional savings.