Austroads commissioned the authors to review the evidence on the effectiveness of different components of graduated licensing schemes for car drivers in order to inform licensing policy in Australia and New Zealand. The project scope was to review available literature on graduated driver licensing (GDL) and did not include the collation and analysis of data from each Australasian jurisdiction. The three main research questions addressed for each component were: 1. Is there research evidence that the GDL component addresses a contributing factor to young driver crashes and/or injuries? 2. Is there evaluation evidence that the GDL component is effective in reducing young driver crashes and/or injuries? 3. What is the potential (quantified) impact of the GDL component on young driver crashes and/or injuries? These questions were addressed and other considerations identified for policymakers to explore in relation to their own jurisdiction conditions and capabilities, rather than to make specific recommendations applicable to all jurisdictions. The authors initially drew from their multiple and extensive reviews of the GDL literature over the past decade with attention to identifying evaluation research with quantified outcomes. Electronic database searches were also conducted to identify more recent literature from Australasian jurisdictions. In addition, Austroads members were requested to provide any internal, unpublished evaluations. The scope was research arising from jurisdictions in Australasia, North America and high income countries in Europe. A summary of the evidence available in relation to each of the three main research questions for each GDL component was tabulated. The results are summarised in a table in the summary. Overall, evidence that each component addressed an important contributing factor to young driver crashes and/or injuries was found for all factors, but not all components had been evaluated as part of GDL models or in relation to crashes, or only limited evaluations had been conducted. The most well evaluated components, and therefore having the most examples of quantified benefits in terms of crash and/or injury reductions, were • a minimum learner age of 16 years • a minimum learner period of 12 months • minimum provisional age greater than 16 years (with increasing benefits with increasing age) • night driving restrictions, • peer passenger restrictions and • a zero blood alcohol concentration (BAC) limit. There was some evidence on the effectiveness of setting a high number of supervised driving hours for learners (80-120 hours), and including hazard perception tests and exit tests in GDL models, but not on when was the best stage to introduce the hazard perception tests. Emerging research suggested education programs to improve cognitive skill deficits, to build resilience and to involve parents had potential to reduce crashes but had not yet been adequately evaluated, while those focused on improving knowledge, awareness and/or attitudes were not effective and those reducing the length of the learner period or providing skid training were counterproductive. There was no evaluation available in order to determine the optimal requirements for supervisory drivers, optimal length of the provisional period, or impact of introducing restrictions on mobile phones or other technology; nor for a range of specific measures to address speeding, alcohol and other offences. Vehicle power restriction evaluation suggested the impact would be limited but primarily due to limited ownership of high performance vehicles by young drivers. It is important to distinguish in the evaluation of GDL components between negative, neutral, or inconclusive findings versus the absence of any evaluation or limited evaluation (second summary column in the table). The latter cannot be deemed to lack effectiveness or to suggest the component should not be introduced; particularly if there is clear evidence that the component addresses an important risk factor (first summary column). Therefore, a lack of evaluation and entry of ‘unknown’ impact should not be interpreted as a recommendation not to include that GDL component. Rather, more evaluation is needed to quantify the expected benefits. There are other limitations to the present review and additional potential for misinterpretation to be addressed. GDL operates as a system and therefore isolating components for analysis can be misleading when comparing across similar but nevertheless different systems. Current evaluation approaches might not be sensitive to these differences. Moreover, the effect of any one component will depend on what other components are already in place. Therefore, introducing a component into an otherwise weak GDL model could be expected to have a greater impact than when introduced into a stronger model. These sensitivities are masked in the result summaries. The examples of potential impact are also not directly comparable due to a lack of common age groups and outcomes for all components, with some including only crashes and others, importantly, fatalities and injuries. Therefore care must be taken when comparing the range of percentages reported. Further, estimating the range of benefits that might be expected in Australasia based on North American evaluations could also differ due to variations in compliance. Compliance can vary greatly depending on enforcement practices and intensity and these are likely less strong in North American jurisdictions, where most states do not mandate display of P plates and parents, rather than police, are considered the chief enforcers of GDL. Caution should also be exercised generalising results from those North American jurisdictions that allow progression through the GDL at younger ages than in Australia. A further limitation of the present research is that the majority of studies reviewed did not account for potential subgroup differences such as socioeconomic or other disadvantage. Such factors also contribute to young driver crash and injury risk and when considering introducing or strengthening a GDL component a balance is needed between ensuring the most promising GDL model for the majority of young drivers, but also ensuring that disadvantaged youth are not unduly further disadvantaged. This might need to be addressed by introducing additional support programs, or through exemptions or alternative requirements in certain circumstances that ensure that minimum standards are still met. This work has been completed based on existing evaluations, with many arising from the United States. This is despite Australasia offering the contrasting GDL models needed for comparative evaluations. While several evaluations are in progress, it is clear that there are several gaps that still need to be addressed. Each jurisdiction should continue to monitor GDL research, but also evaluate their own data to the extent possible to ensure the appropriateness of the recommendations to their jurisdiction. It is worth noting that a GDL model comprised of the strongest GDL components reported in the literature does not equate to a model that currently exists and has been tested as a unified model. Jurisdictions might choose to introduce or strengthen GDL components on a gradual basis rather than at once. Public discussion papers could also be released prior to significant changes in order to stimulate and canvass public debate to provide insights into community acceptance and potentially identify issues not yet considered. Another GDL initiative undertaken in the United States that could also be considered for Australasia is that taken by the Insurance Institute for Highway Safety. An on-line benchmarking initiative that rated states as ‘good’, ‘fair’, ‘marginal’ or ‘poor’ contributed to most states achieving ‘good’ status over time. The Institute now instead provides an on-line GDL calculator where potential reductions in crashes due to introducing or strengthening certain GDL components can be immediately calculated by jurisdiction. This provides a useful and likely persuasive, motivating tool for policymakers, advocates and the general public alike. Striving for high quality evaluations of contrasting initiatives that Australasia has to offer could allow such a tool to be developed in our region and therefore support future campaigns to help strengthen our GDL systems to continue to protect young lives from road trauma. (Author/publisher)
Abstract