It is apparent that we do not know enough about the causes of many motorcycle crashes. It is too easy to simply blame the rider without reviewing the other contributing factors. Any crash results from a combination of circumstances converging to a point when the rider does not have the skills or the options to avoid the crash. Systematic investigation is necessary to identify patterns of failure associated with driver/ rider behaviour, road conditions, vehicle features as well as rider skill.

Police crash investigation
Detailed crash investigation is carried out in all cases of fatalities and most serious injury crashes. However, the primary role of the Crash Investigation Officer is to determine cause and effect in order to establish criminal negligence. While the role does include responsibility to identify contributing factors that may indicate strategies to prevent similar crashes in the future, Police Crash Investigators are not trained research scientists nor forensic engineers.

For crash investigation purposes, a crash is determined to be 'serious' if there is a fatality or if injury could constitute 'grievous bodily harm'. An injury would generally be considered 'serious' if surgery is required. In such cases the first response police will notify the Crash Investigation Unit who may attend the scene, however even serious crashes are not always investigated by the Police Crash Investigators. The essential principle role for police is to determine fault for prosecution. If the injured party was at fault in the crash, then they are not generally considered to be 'a victim' in criminal justice terms. If there is no 'victim', there may be no criminal case of negligence for prosecution and therefore no need for investigation.

Single vehicle crashes will generally not be investigated unless they involve a fatality or if there are serious injuries and there are no witnesses to report on what caused the crash. This has particular relevance for motorcyclists as 42% of all motorcycle crashes are single vehicle crashes and we have very little consistent data about the causes.

The determination of 'fault' by attending Police therefore has significant implications for the investigation of motorcycle crashes. The assumption in any single vehicle crash is that the vehicle operator was at fault, essentially for losing control of the vehicle. This is generally assumed to mean they were going too fast for the conditions. The consequences can include the loss of insurance benefits and even their licence as they may be charged with negligent driving. However, as discussed earlier, changes in the road condition may occur with little warning but dramatic consequences for a motorcyclist even if they are riding well within the posted speed limit.

In the MCC survey, 68% of motorcyclists involved in single vehicle crashes claim road condition to be a contributing factor. However road condition is officially noted as a contributing factor in only 23% of single vehicle motorcycle crashes, and a quarter of motorcycle crashes are noted as involving excessive speed for conditions compared to 16% of all vehicle crashes (de Rome et al, 2001).

Simply blaming road condition to absolve rider responsibility is not the answer either, but a better understanding of the circumstances that trigger a critical incident would help to determine the means to avoid them. As discussed earlier, if the road surface is a problem, the incident may have been very difficult to avoid, even for the highly experienced. Where rider error is the cause, the consequences of even a minor error of judgment are far more severe on a motorcycle than in a car. While these are issues that need to be addressed in training, informed feedback and analysis of crashes in police reports could be an aid in developing strategies to be used in training.

The discrepancy may indicate that Police attending motorcycle crashes are ascribing fault to the motorcyclists simply because they do not have the experience to recognise other relevant contributing factors. If this is the case, this failure results in missed opportunities to remediate road conditions before they cause further injury. It also serves to perpetuate misunderstanding of the causes of motorcycle crashes. This is, at least in part, due to the involvement of general duties police who, in addition to the whole range of other crime prevention duties, are expected to attend crash scenes and make determinations on complex engineering matters without any specialist training. While it is not feasible to expect highly trained crash investigators to attend every crash, it would be more feasible to require crashes to be attended by the Highway Patrol, who could be provided with some training in the preliminary assessment of crash causes.

Crash data collection
The reporting system used by the NSW Police when attending a crash is the basis of all available data on the incidence and causes of crashes. However the system is designed to identify factors in relation to enforcement issues. It is not intended as the basis for research into the causes of crashes nor for the collation of data on crash trends. There are a number of specific disadvantages for motorcyclists that arise from the limitations of the crash reporting system. Perhaps most importantly, the system does not allow for a comprehensive assessment of the contribution of other factors such as road condition in crashes.

Police who attend a crash site make written notes from which they subsequently enter details into encoded fields in the COPS database. The data fields in the crash reporting system were designed primarily for car crashes and do not accommodate factors which would only be relevant to other vehicle crashes. While there are mandatory fields that must be completed, these relate to location, vehicle and casualty details. While provision is available for more detail to be entered into other optional fields, police tend to complete only the mandatory fields and provide the remaining detail in free form text.

The free text field is a weak point in the system as the level of detail provided varies from one officer and crash to another. In addition the text is then subject to interpretation at the point of data entry by the RTA. Of particular concern is the suggestion that incidents involving 'loss of traction' may be generally interpreted as 'speeding' and coded accordingly. This has no bearing on the outcomes for the controller in any particular crash, but it may obscure the causes of the crash and repeated over a large number of crashes create misinformation about the nature and causes of motorcycle crashes.

Problems arise from the design of the crash reporting system as well as from its implementation. The reliability of the information could be improved by police using more than the mandatory fields or by working from a check list of factors when completing the narrative text. The value of the system could also be enhanced by extending the fields to take account of factors more likely to be associated with other vehicle crashes such as motorcycles and trucks. Training in the assessment of crashes and completion of crash data reports is essential, however whereas the RTA used to work with the Police Academy to provide training to recruits on road safety and data collection, this practice has ceased in recent years.

The Police who attend a motorcycle crash will also determine whether a driving offence has been committed. In the case of single vehicle crashes, there is a common view that the rider lost control and therefore is charged with speeding and/ or negligent driving. There is reason to believe that this may explain under-reporting of motorcycle crashes by motorcyclists who believe reporting will be to their disadvantage. Research in England suggests that only 8% of non-injury crashes and 24% of injury crashes were reported. Severity of injury, involvement of other vehicles and degree of vehicle damage were all factors determining whether the crash was reported (Helen F. James, Under-reporting of road traffic accidents, 1991).