category: literaturenote citekey: corbenstrengtheningroadsafetystrategy2010 title: Strengthening road safety strategy development ‘Towards Zero’ 2008–2020 – Western Australia’s experience scientific research on road safety management SWOV workshop 16 and 17 November 2009 authors: "Corben, Bruce F.; Logan, David B.; Fanciulli, Lisa; Farley, Roger; Cameron, Iain" year: 2010 date: 2010-00-00 2010 doi: 10.1016/j.ssci.2009.10.005 publication: Safety Science url: "https://doi.org/10.1016/j.ssci.2009.10.005" zotero_key: 4DRNCUW9 zotero_storage: 5H83CVWN collections: transpordi_planeerimine folder: 001_artiklid/Road Safety Indicators firstAuthor: "Corben, Bruce F."

status: converted

Contents lists available at ScienceDirect

Safety Science

Strengthening road safety strategy development 'Towards Zero' 2008–2020 – Western Australia's experience scientific research on road safety management SWOV workshop 16 and 17 November 2009

Bruce F. Corben a,*, David B. Logan ^a , Lisa Fanciulli ^b , Roger Farley ^b , Iain Cameron b,1

article info

Article history: Received 1 July 2009 Received in revised form 24 September 2009 Accepted 4 October 2009

Keywords: Safe System Vision Zero Sustainable Safety Computer modelling Prediction Serious casualties Strategy performance Road safety Community consultation

abstract

With a significant list of achievements in traffic safety over some 40 years, Australia's most successful initiatives have emerged from consideration of the scientific basis for achieving effective results. Over time, the increasing use of scientific methods has matured to an extent whereby road safety strategies for entire jurisdictions are now being formed and optimised on the basis of evidence-based mathematical modelling of predicted strategy performance over periods of ten or more years ahead. This paper describes the approach used in the State of Western Australia (WA) to support the development of its new strategy, 'Towards Zero', addressing the 12-year period from 2008 to 2020. Also described is the early development of the modelling approach in the State of Victoria, Australia, subsequently applied, refined and adapted to Western Australia. The 'Towards Zero' strategy aims to ensure that road safety policies continue to evolve within a strategic framework, based on Australasia's Safe System framework, which was derived from the best elements of the Swedish Vision Zero and the Dutch Sustainable Safety visions.

The process for the development of the strategy involved a greater degree of community and stakeholder engagement than has been the case previously, based on the fundamental belief that the community should be provided with the best evidence about what works no matter how controversial, so that it can debate and consider the options available to improve safety.

It is equally important to have a strategy based on the best possible evidence. The Monash University Accident Research Centre (MUARC) was appointed by the WA Office of Road Safety to identify, describe and recommend proven counter-measures within the Safe System framework and develop a series of scientifically sound directions and options.

This paper outlines the modelling technique used by MUARC along with results and summarises the community consultation process.

-2009 Elsevier Ltd. All rights reserved.

1. Background

The Western Australian government and Road Safety Council of Western Australia undertook to develop a new road safety strategy for the period 2008–2020 within Australasia's Safe System framework, with the addition of a bold, long term level of aspiration or vision. The Safe System vision had already been formulated from the best elements of the Swedish Vision Zero and the Dutch Sustainable Safety visions, however Western Australia had never before had a long term level of aspiration or vision. Such an approach has rarely been tried around the world. This project, therefore, should be recognised as an ambitious, highly-challenging initiative of the WA government. MUARC was commissioned by the government to provide the objective research basis for the strategy which, subject to successful implementation, can be expected to deliver large and lasting road safety benefits for Western Australia. A key aspect of the approach was to develop a model to estimate the projected benefits of a combination of best-practice Safe System countermeasures. The METS (Macro Estimates for Target Setting) model was developed for Victoria and, with the permission of Victoria's state road authority, VicRoads, the underlying model structure was used as the basis for substantial further development, culminating in the model now known as METS–WA (Corben et al., 2008).

Given the challenge of implementing such an approach, the WA Road Safety Council determined that the strategy should be developed in a consultative and transparent way to maximise

^a Monash University Accident Research Centre, Building 70 Monash University, Victoria 3800, Australia

^bOffice of Road Safety, Main Roads Western Australia, Government of Western Australia, Perth, Australia

^* Corresponding author. Tel.: +61 3 9905 4375; fax: +61 3 9905 4363. E-mail addresses: Bruce.Corben@muarc.monash.edu.au (B.F. Corben), Lisa. Fanciulli@mainroads.wa.gov.au (L. Fanciulli).

¹ Tel.: +61 8 9222 8781.

stakeholder and community acceptance. As such, the process for the development of the strategy for 2008–2020 involved a greater degree of community and stakeholder consultation than has been the case with previous strategies.

2. The Safe System

The adoption of the Safe System road safety philosophy represents a profound shift in thinking and insight in the Australian society's efforts to curb road trauma. Translating the Safe System philosophy and principles into real-world practice is believed to have the potential to deliver major advances in the safety performance of Australia's road-transport system over coming years and decades.

The Safe System framework was created specifically for Australia and New Zealand by combining the best elements of Sweden's Vision Zero and the Netherlands' Sustainable Safety road safety philosophies. Of the highly motorised countries, Sweden and the Netherlands are acknowledged world leaders in road safety performance and have each created and adopted during the 1990s bold, fundamentally new approaches to reducing road trauma. While these philosophies differ in their relative emphases, they are both founded on a strong ethical platform and both acknowledge the vital importance of explicitly recognising the limitations of humans using the road-transport system.

Like its Swedish and Dutch counterparts, Australasia's Safe System philosophy also aspires to eventually eliminating death and serious injury within its road-transport systems. Serious casualties are defined as the sum of deaths and serious injuries. While some will regard this goal as unrealistic, it is important to realise that:

- Designing and operating a system that tolerates consistently high-levels of serious trauma is both unethical and at risk of perpetuating the acceptance of failure by system designers and operators.
- New, fundamentally safe, designs will result from viewing the safety of the road-transport system in this new light. Intrinsically safe designs and protocols for operation can shift system safety a large step forward, in contrast with the incremental progress characterising traditional approaches.

The Safe System is underpinned by four key principles:

1. The limits of human performance Traditional approaches to road safety focus heavily on preventing human failures within the system. For example, there has been considerable success with programs targeting drink-driving, speeding, and restraint non-use. However, despite this partial success, these and other forms of human failure, such as poor gap selection at complex or simple intersections or while overtaking, failure to stay within travel lanes, even on high-speed high-standard roads, and driving while fatigued, have so far proven beyond our capabilities to eliminate through behaviour change programs. A guiding philosophy that explicitly acknowledges the limited intrinsic capabilities of humans is essential to future success.
2. The limits of human tolerance to violent forces Humans are limited in their biomechanical tolerance to the violent forces and energy exchanges that commonly occur in traffic crashes. When tolerance limits are exceeded serious injury or death results. The Safe System explicitly acknowledges these scientific facts by seeking to create a road-transport system that is forgiving of human failures within the system, such that any foreseeable collisions occur within the biomechanical limits of the

humans involved. To achieve this in practice, the ability of a vehicle to protect its occupants in common crash types and at typical impact speeds (resulting from legal travel speeds) must be known and taken into account in decisions about system design and operation. It is also vital to understand the biomechanical limits of unprotected road users (e.g., pedestrians, bicyclists and motorcyclists) in collisions across the full range of vehicle types and speed zones.

3. Safe road use While the ultimate responsibility for the safety performance of the road-transport system resides with the system designers and operators, it can only be possible to move progressively towards a situation of no deaths or serious injuries in the long-term provided road users comply with key rules such as speed limits, restraint-use, driving unimpaired and in vehicles with good safety features. Thus, under a shared responsibility for road safety, the system designers depend on compliance by road users for achieving a truly Safe System, while individual, complying users depend on inherently safe design standards and operation for their safety.
4. Creating a forgiving road-transport system The road-transport system is, fundamentally, a real-world illustration of the basic laws of nature governing the movement of objects, be they vehicles or humans, relative to the physical environment. How these objects interact is, ultimately, a matter of physics varied in time and space according to human intervention, which is particularly difficult to predict or control.

Safe outcomes are largely determined by the ability of individual road users to control kinetic energy, which is the physical energy gained by an object (most significantly a vehicle) as a result of its motion. Kinetic energy is a function of both the mass of an object, but more importantly, its speed at any instant in time. Unlike mass, which is linearly related to kinetic energy, speed has a squared (i.e. 2nd power) relationship with kinetic energy. A safe road-transport system depends, therefore, on the successful separation of sources of kinetic energy (crash avoidance) and, where this cannot be achieved, the controlled dissipation of kinetic energy to avoid exceeding human biomechanical limits (injury prevention or mitigation). Given the inherent limitations of human performance within the road-transport system, the Safe System seeks to create roads and roadsides that are tolerant of human error and reasonable levels of impairment. This is achieved by ensuring that sources of kinetic energy remain separated or, in worst cases, can be managed within the limits of human tolerance to serious injury.

A major challenge for the future is to identify and implement current initiatives that meet this goal, as well as to generate new initiatives consistent with these Safe System principles.

3. Approach to developing the optimal strategy

A key early challenge was to identify a way to apply the Safe System to a large geographic jurisdiction with a low population.

Western Australia has the largest land area of all the Australian States and Territories comprising 2.5 million sq km and 13,000 km of coastline. Of the population of 2.1 million people, some 1.6 million live in the capital city, Perth. The road network is 179,000 km long, of which 126,000 km is unsealed with 227 million vehicle kilometres travelled annually.

In 2007 the fatality rate was 112 per million people, of which about two-thirds were killed on regional roads at a rate of about 220 per million compared to a rate of 50 deaths per million people in Perth. A total of about 3000 people are seriously injured each year on Western Australian roads.

To facilitate the identification of the best mix of initiatives for Western Australia's next road safety strategy, the concept of a ''Safe System Matrix" was created. The matrix aimed to direct attention primarily to a relatively small number of the larger and most severe traffic injury problems, using proven high-impact solutions. This approach still permits other problems of lower priority to be addressed (e.g., where specific user groups are at especially high risk, even though their absolute numbers of injuries may be relatively low), but at a level commensurate with the potential savings in serious trauma.

The Safe System Matrix served two important purposes:

1. It provided a systematic, structured approach to identifying strategically important road safety initiatives, consistent with the principles of the Safe System. In particular, it ensured that each of the main components of the Safe System, namely, safe roads and roadsides, safe travel speeds, safe vehicles and safe road use, were comprehensively addressed within the strategy. These main components have become known as the Safe System Cornerstones.
2. The matrix was structured to ensure that the differing road safety problems and priorities between Western Australia's three main geographic areas – Metropolitan Perth, Regional WA and Remote WA – could be more precisely targeted within the strategy to meet these differing needs. 'Regional WA' represents the road network within a 300 km (approximately) radius of metropolitan Perth, while 'Remote WA' encompasses the rest of the state. These divisions were chosen to acknowledge the substantial differences in road, vehicle and other characteristics between each of these areas.

The Safe System Matrix is shown in generic form in Table 1. The matrix does not contain all of the proposed countermeasures, but rather those that are anticipated to have a significant effect upon serious casualties. There are a variety of other worthwhile initiatives, including additional bicycle paths, traffic calming, improved pedestrian separation and lighting (especially in remote areas of WA) included in the final strategy. A brief description of each of the major areas of the matrix follows.

3.1. Safe roads and roadsides

The Safe Infrastructure Cornerstone consists of countermeasures aimed at improving the safety of roads and roadsides. It al-

Table 1 The WA Safe System Matrix.

Safe roads and roadsides Safe speeds Safe vehicles Safe road use All of WA Statewide Accident 'Black Spot'^a program safer roads programs Enhanced speed enforcement Promotion of crash avoidance features Development of aggregate behaviour change programs Review and fine tune speed limits Promotion of advanced crashworthiness features Metro Perth Intersection countermeasures Adjustment of speed limits to complement infrastructure measures No metro-specific countermeasures Development of specific behaviour change for Metro issues Run-off-road countermeasures Regional WA Safe System transformation of strategically important routes radiating from metro boundary Rezone all limits downwards by 10 km/h Specific promotion of ESC in 4WDs Development of specific behaviour change programs for regional issues Promote ESC selection/fitment to heavy vehicles Remote WA Safe System transformation of strategically important routes around remote centres Rezone all limits downwards by 10 km/h Specific promotion of ESC in 4WDs Development of specific behaviour change programs for issues in remote areas Promote ESC selection/fitment to heavy vehicles

lows for the continuation of existing Accident Black Spot and Safer Roads programs across the whole of WA, provided the selected sites and routes do not overlap with the specific regional and remote treatments detailed below.

One of the key countermeasures for Metro Perth is a 'Safe System Transformation' of intersections. This would involve the implementation of such measures as roundabouts, full control of right-turns at signals, combination red light/speed cameras and others at key intersections. Sites would be selected on the basis of numbers of serious casualties, regardless of exposure, in order to maximise the potential savings in serious casualties as a result of countermeasures being implemented. The use of serious casualties, normalised for traffic volume, may redirect attention away from locations with the worst safety records in terms of absolute numbers of deaths and serious injuries. Run-off-road crashes are also a significant problem in urban areas, but few effective solutions currently exist. Ongoing research should be conducted into the development of new barrier solutions for urban applications, especially on higher speed roads such as those zoned at or above 70 km/h. For roads currently zoned at 60 km/h, reduced speed limits, coupled with enhanced enforcement, have the potential to address this problem without the need for barrier solutions.

In addition to the whole of WA programs, a Safe System Transformation was proposed for routes radiating from Perth. This aimed to address the run-off-road and potentially, head-on crash problem on key strategic routes from the Metro Perth boundary as far out into regional WA as practicable. A similar strategy could be adopted for key rural centres. Traditionally, specific road sections would be treated piecemeal on the basis of crash history. In the case of a Safe System Transformation, however, the goal is to start from the metropolitan end of each route (where crash densities tend to be higher) and work outwards 'injury-proofing' entire road lengths rather than individual sections.

3.2. Safe speeds

Reducing travel speeds across Western Australia is one area where significant serious casualty savings are possible at relatively low cost. This potential has been demonstrated clearly through Nilsson's power laws (Nilsson, 2004) reinforced with a variety of evaluations of reduced speeds conducted internationally (e.g. Aarts and van Schagen, 2006 and references therein). This is particularly true for a very large, sparsely populated jurisdiction like Western Australia where it will be practically impossible in the foreseeable future to provide best-practice infrastructure solutions across much of the state. For the whole of WA, a program of enhanced

^a In Australasia, 'Black Spot' has a similar meaning to high crash location/high risk location, commonly used internationally.

enforcement of speed limits was recommended, involving measures such as better-targeted police coverage and increased use of speed cameras. It was also recommended that a systematic program of 'fine tuning' of existing speed limits be pursued, particularly in metropolitan Perth as well as regional and remote centres. This involved redefining a change in speed limit so as to incorporate nearby known hazardous locations, such as intersections or high-risk curves, within the lower speed limit.

In addition to the above measures, two speed-related countermeasures were recommended for Metro Perth. The first of these was to reduce urban speed limits from 60 km/h to 50 km/h. Currently 27% of all serious casualties occur in 60 km/h speed zones within the metropolitan area and even small reductions in travel speed can be expected to lead to significant serious casualty savings. Another situation where lower speed limits can bring about significant trauma reductions is in areas of high pedestrian activity such as Perth's central business district and urban strip shopping centres. It was therefore recommended that these areas be rezoned either to 40 or 30 km/h to improve Safe System compatibility in both avoiding and mitigating pedestrian-vehicle conflicts.

In regional WA, it was recommended that speed limits be reduced by 10 km/h, with progressive reinstatement possible as the Safe System Transformation program is completed along the selected key routes. It was also deemed highly desirable for WA's open road speed limit to be reduced from 110 km/h to 100 km/h, where the vast majority of roads may never be practically able to be lifted to Safe System standards.

3.3. Safe vehicles

This Safe System Cornerstone aimed to capitalise on the proven safety performance of a number of vehicle safety features that, in general, are not currently-fitted as standard on many vehicles available on the Australian market. These features include electronic stability control (ESC) and Intelligent Speed Adaptation (ISA), which help to prevent crashes occurring and side-impact and curtain air bags, and active head restraints, which help to reduce occupant injury severity when a crash occurs. This initiative aimed to accelerate the introduction and rate of penetration of these features into Western Australia's vehicle population by requiring government fleet purchasers to specify selected features when renewing the government fleet and by encouraging corporate fleet purchasers to do likewise. Not only will this approach accelerate the take-up of these proven features in WA's vehicle fleets, it will also put much-needed pressure on vehicle manufacturers to provide these features as standard in new vehicles, or at least as options that can be specified as stand-alone features in their base models. These benefits will flow on to WA's private vehicle purchasers as government and fleet vehicles are sold into the general fleet.

Based on current and proposed levels of availability of these features, a timeline for their introduction to government and corporate fleet purchases was proposed as follows:

-ESC from Year 1.
-Curtain and side-impact airbags from Year 1.
-Intelligent Speed Adaptation from Year 3.
-Active head restraints from Year 5.

3.4. Safe road use

For a best-practice strategy for WA, the Safe Road Use Cornerstone would involve the creation of an integrated suite of campaigns that present and promote the Safe System road safety philosophy as well as targeting traditional and new problem areas. The main features would include:

- An emphasis on the Safe System being a new, ambitious approach capable of achieving major improvements in traffic safety over the next decade and beyond.
- The importance of shared responsibility, with road users to comply with key laws including drink-driving, drug-driving, seat belt use, fatigue-driving, etc. Correspondingly, the obligation of the system designers and operators to create a low-risk road-transport system would also be reinforced.
- A series of problem-specific education, promotion and publicity campaigns targeting key problem types amenable to improvement through behaviour change programs including drink-driving, drug-driving, speeding, restraint-use, distraction/inattention, fatigue-driving, etc. Furthermore, as part of each issue-specific campaign ongoing reinforcement of Safe System principles and the concept of shared responsibility.

4. Modelling strategy performance

Central to the overall approach to developing an optimal road safety strategy for Western Australia was the modelling of strategy performance using evidence-based estimates of the effectiveness of the individual road safety initiatives described in the previous section as well as combinations of initiatives. The modelling approach operates at a macro-level and so includes only those initiatives expected to have a sizeable impact on the reduction in severe road trauma. Using evidence-based estimates of effectiveness and actual crash data for Western Australia over recent years, the model used mathematical methods to forecast future savings in serious casualties in each year of the 12-year period from 2008 to 2020. These predicted savings in serious casualties were summed over the strategy life for each individual initiative, assuming it alone was implemented. These savings were also summed for combinations of initiatives implemented as part of a strategy, allowing differentiation between options in terms of their potential contribution to preventing road fatalities and serious injuries. The principal measure of worth of a strategy option is the total number of serious casualties prevented over the strategy life.

These savings are estimated relative to the level of serious casualties that could be expected to occur in the absence of a significant road safety strategy.

Forecasting future road trauma occurrence and trends is a challenging exercise, characterised by uncertainty and assumptions about the nature of the road-transport system, now and in the future. However, to plan for success, it is necessary to follow such a path, while reducing as far as practicable, the potential for error. The approach taken with modelling for the WA strategy has attempted to account for two main, whole-of-system, influences.

- Future growth in serious casualties as a result of increasing exposure, reflected by vehicle kilometres travelled (VKT). This figure was estimated at 2.6% per annum for the life of the strategy, based on growth between 2001 and 2005, and could be an underestimate depending upon the duration and size of the economic boom experienced by WA at the time of modelling.
- Future improvement in serious casualty rates due to the gradual effects of increasing motorisation (e.g. Jacobs et al., 2000). This value was estimated at 1.6% per annum for WA, derived from the trend in serious casualties per vehicle-kilometre travelled using annual police-reported data for the period 1990–1996.

These two influences, while tending to counteract each other, resulted in a forecast general upward trend of an average 1.0% per annum in serious casualties through to 2020 due to the higher rate of traffic growth than decline in serious casualties associated with motorisation. Together, the combined effect of traffic growth and serious casualty reduction with increasing motorisation was used to define a continuous point of reference against which potential savings in serious casualties due to the proposed strategy and its individual elements were estimated.

The model produces two principal outputs:

1. The estimated total number of serious casualties saved over the life of the strategy (relative to the forecast numbers of serious casualties in the absence of a significant strategy).
2. The percentage reduction in serious casualties in the final year of the strategy compared with a selected baseline period, which would be the most recent year, or an average of the most recent three years (if substantially different from the most recent year) for which full serious casualty data were available. The total number of serious casualties saved over the life of a strategy can be used to identify the most effective strategy, while the percentage reduction in the final year of the strategy (Year 12) provides a target for achievement by the end of the strategy (and at intermediate milestones throughout).

The model also produces a number of additional output measures for each initiative and each strategy option. These included the estimated cost (in present-day values), as provided by agencies responsible for implementation, allowing the average cost of implementation per serious casualty saved, predicted numbers of fatalities and serious injuries in the final year of the strategy, among others to be estimated.

5. Model structure

5.1. Description

The METS (Macro Estimates for Target Setting) model was originally developed over a period of 16 months from early 2006, funded by VicRoads as part of the development of the next Victorian road safety strategy. Although the Western Australian version was based on the latest version of METS, METS-WA has undergone extensive development and customisation with the aim of tailoring it to the Western Australian road safety environment, as well as incorporating significant improvements to the underlying model. The model is a numerical implementation of concepts developed in the late 1990s by MUARC researchers Peter Vulcan and Bruce Corben (e.g. Vulcan and Corben, 1998). It starts with a base number of annual serious casualties (killed and hospitalised) derived from police-reported data. The effectiveness of individual countermeasures in reducing the proportion of total serious casualties is used to predict serious casualty numbers from the effects of the countermeasure in isolation for each year for the duration of the strategy. Underlying trends are also taken into account as detailed above. Individual countermeasures are then combined into strategy 'packages', either additively or multiplicatively, the latter avoiding double-counting of savings when different initiatives act upon the same 'pool' of serious casualties.

5.2. Estimating scope for reduction in serious casualties

This section presents a brief overview of the mathematical basis for estimating the reductions in serious casualties possible from specific initiatives. In essence, this section summarises the work of Vulcan and Corben (1998), adapting it to the present application. As noted above, fatalities and serious injuries can be reduced by initiatives undertaken within the four Safe System Cornerstones. If no improvements are made in any of these areas, it is considered reasonable to assume that total fatalities (and serious injuries) would change according to an annual average serious casualty growth factor, g, defined as the percentage growth in traffic minus underlying rate of improvement due to the general effects of increased motorisation. As with Vulcan and Corben (1998), the effects of traffic congestion and differential growth on different parts of the road network were neglected.

$$SC_t = SC_0 \left( 1 + \frac{g}{100} \right)^t \tag{1}$$

where SC^t = serious casualties in year t, SC0 = serious casualties in base year, g = annual average serious casualty growth factor, as defined above.

To take account of the empirical finding that even in the absence of any additional road safety measures, fatality rate gradually decreases as motorisation (vehicles per head of population) increases and the road, vehicles and the road user population ''matures". While such effects are difficult to measure allowance has been made for this effect.

Then if the total percentage reduction in fatalities and serious injuries resulting from all safety measures combined is R percent

$$SC_R = SC_t \left( 1 - \frac{R}{100} \right) \tag{2}$$

where SC^R = number of serious casualties in the final year of the strategy (of duration t years) during which all the safety measures were applied.

To ensure that savings in serious casualties are not counted more than once, they are either added or multiplied together, depending upon the actual mechanisms for achieving the savings. As described earlier, some initiatives act on the same set of crashes as other initiatives and therefore the savings need to be combined multiplicatively, while other initiatives can act on separate crash sets and so have been combined additively. Vulcan and Corben (1998) present a number of examples to illustrate this point, namely:

$$1 - \frac{R}{100} = \left(1 - \frac{r_b}{100}\right) \left(1 - \frac{r_s}{100}\right) \tag{3}$$

where rb = percent reduction from better vehicle brakes, rs = percent reduction from increased seat belt wearing, R = total percent reduction from these two measures.

Similarly, if there are four sets of measures (r1–r4), the formula for calculating the total percentage reduction from all measures R is

$$1 - \frac{R}{100} = \left(1 - \frac{r_1}{100}\right) \left(1 - \frac{r_2}{100}\right) \left(1 - \frac{r_3}{100}\right) \left(1 - \frac{r_4}{100}\right) \tag{4a}$$

where r¹ = percentage reduction from all safer vehicle measures, r² = percentage reduction from all safer road measures, r³ = percentage reduction from all safer road user measures, r⁴ = percentage reduction from improved emergency medical services.

Or, more generally, for n road safety countermeasures, each resulting in a percentage reduction, rn:

$$1 - \frac{R}{100} = \prod_{i=1}^{n} \left( 1 - \frac{r_n}{100} \right) \tag{4b}$$

Interactions between these four categories of measures have been ignored.

Each of the above four percentage reductions represents the net effect of several measures some of which are additive, some are multiplicative and some of which overlap or interact (e.g. speeding and drink-driving enforcement).

It can be seen from Eq. (1) that the extent to which a reduction in fatality rate translates to a reduction in the number of fatalities also depends on traffic growth. Thus influencing the rate of growth or fall in road use also represents scope for further improvements in road safety as well as vehicle emissions.

A unique feature of METS-WA is its ability to allow different levels and combinations of initiatives to be relatively easily compared.

5.3. Key assumptions

METS uses a worksheet with more than 150 input variables, both user-defined and calculated. Variables are separated into general parameters as well as parameters relating to each of the Safe System Cornerstones, namely infrastructure (safe roads and roadsides), speed, vehicles and behaviour (safe road use). Where possible, input variables were derived from either reliable agency-supplied statistical data or the results of sound research. For a number of minor variables, however, neither of these sources was able to provide the specific data required. Where this was the case, conservative consensus estimates were used, based on the best evidence currently available.

5.4. Outputs

The two principal outputs of the model are: (a) cumulative numbers of serious casualties saved over the life of the strategy and (b) performance of the strategy in its final year, as well as selected milestone points. Strategy performance was measured as the percentage reduction in the number of serious casualties in the selected year compared with the chosen baseline period (as defined earlier), usually the most recent year for which full serious casualty data are available. The number of cumulative serious casualties saved is a relative figure calculated by projecting the future state of baseline serious casualties in the absence of a specific strategy. Although it is a somewhat arbitrary baseline, it allows different strategy options to be compared with one another. Moreover, cumulative savings are important to consider, as strategy performance in a single year may give little indication of overall strategy performance.

Where the costs of implementing initiatives were provided, the model also provides approximate indications of: (a) total strategy cost, (b) cost per serious casualty saved, (c) the monetary valuation by society resulting from the serious casualty savings, and (d) the ratio of this monetary valuation to strategy cost. Costs are all expressed in present-day values, with future cost increases not taken into account.

To illustrate the use of the equations presented in Section 5.2 (refer to Vulcan and Corben, 1998 for original explanation), consider the case of initiatives being assessed to improve road infrastructure safety, accelerate the take-up of proven vehicle safety features and strengthen speed management. Set out below are three charts which show the estimated annual savings in serious casualties from each of the above initiatives, as well as the yearby-year build-up in the cumulative savings in serious casualties over the period of the 12-year strategy developed for WA. These savings were calculated using the above equations, from knowledge of the:

average annual number of serious casualties in the chosen base period;
effectiveness of individual initiatives in reducing serious casualties; in some cases it is important to distinguish between the different levels of effectiveness of initiatives on fatalities and on serious injuries;
annual growth in traffic over the strategy life;
annual reduction in the rate of serious casualties per vehicle kilometre of travel over the strategy life;
extent of implementation of the initiative, either as a onceonly decision or in terms of the annual (or other periodic) level of implementation;
timing and the means by which the initiative has its effect, especially with respect to the cumulative effects over time (this is described more fully below).

5.5. Example 1 – Safe Roads and Roadsides Cornerstone

The effectiveness of black spot and other safety-based road investment undertaken in Australia over many years were used to estimate savings in serious casualties from improving road and roadside safety. Estimates of effectiveness of these types of program were obtained from local evaluations. Applying these findings to the WA trauma problem (Meuleners et al., 2005), and adjusting for the actual level of investment each year across a number of safe infrastructure investment options, it was estimated that 58 serious casualties would be prevented each year if \$135 million (in present-day values) is invested annually over the life of the strategy. The estimated annual savings of 58 serious casualties for an annual investment of \$135 million is calculated from a weighted average of the individual levels of effectiveness of different types of infrastructure investment (i.e., accident black spot program, safer roads program, Safe System Transformation of metropolitan intersections, and regional and remote routes).

Because infrastructure programs of this scale require a significant lead time to reach full implementation capacity, no savings were allowed in the first two years of the strategy and only half of annual savings were counted in the third year, before building to full delivery of annual savings from Year 4 for the life of the strategy. Such assumptions attempt to be both realistic about potential savings in serious casualties and conservative in nature, so that implementing agencies are confident they are not being asked to perform beyond their true capacity.

A major feature of investment in safe infrastructure is that savings in serious casualties gained in the first year continue to accrue in each of subsequent years until the theoretical, useful lives of these projects have expired. During this period the savings which accrued from investment in the Years 2, 3, 4, etc. also accrue in the same basic way so that the net effect is a rapidly building profile of savings in serious casualties that actually extend well beyond the 12-year life of the strategy. Because only the savings accrued during the life of the strategy are counted in the strategy assessment method, the model is quite conservative in its estimation of strategy benefits. It is, however, assumed that the effectiveness of safety measures implemented within this cornerstone remain fully effective throughout the life of the strategy (see Fig. 1).

5.6. Example 2 – Safe Vehicles Cornerstone

The individual estimates of effectiveness of proven vehicle safety features are used to estimate the rate of accumulation of savings in serious casualties resulting from these features finding their way into a jurisdiction's vehicle fleet. For example, it has been found that Electronic Stability Control (ESC) reduces single-vehicle serious casualties involving passenger vehicles by 35% (after Lie et al. (2006), and Scully and Newstead, 2008), while curtain airbags are effective in reducing all serious casualties from side-impacts and rollovers involving passenger vehicles by 20% (reference: MUARC estimate). Similarly, research has found that Intelligent Speed Adaptation (ISA) is effective in

Fig. 1. Estimated cumulative serious casualty savings by year compared with ''business as usual" for the modelled package of infrastructure investment.

reducing all serious casualties involving passenger vehicles by an estimated 10% (Regan et al., 2006) and that Active Head Restraints are effective in reducing all serious casualties from rear-impacts involving passenger vehicles by 15% (MUARC estimate). These estimates of effectiveness were then applied to actual Western Australian data on each target crash type to estimate savings in serious casualties for a given rate of introduction of the various technologies.

The rates of penetration of the technologies into the WA fleet was based on the assumption that from a specified date for each technology, government and corporate vehicle purchasing policies and practices will change to ensure that such features would always be specified from the assumed start date forward. It was therefore also necessary to obtain information on the current rate of purchase and sale of government and corporate fleet vehicles, and project these rates throughout the 12-year life of the strategy.

In order to maintain an acceptable degree of conservatism in the setting of an overall strategy target, the start dates for each vehicle safety feature were chosen to enable adequate time for implementing agencies to prepare for the required changes in fleet purchasing policy. For largely practical reasons, it was assumed that take-up of both ESC and side/curtain airbags would increase from actual 2008 levels to 100% by 2011 and 2012 respectively, ISA from Year 3 and active head restraints from Year 7, and that only 80% of purchased vehicles in first year of implementation would have the specified safety features, 90% in second year of implementation and 100% thereafter. It was also hoped that the collective actions of fleet purchasers would be sufficient to bring about a change in the policy of many vehicle manufacturers to make these features standard in base models so that from that time forward the occupants of all new vehicles would enjoy the safety benefits possible from these features. The added savings in serious casualties derivable, should such a decision be taken by the manufacturers of vehicles sold in Australia, have not been counted in the total benefits possible during the life of the strategy.

Likewise, in keeping with the conservative nature of this modelling approach, savings in serious casualties from private purchases have not been counted; such savings are regarded as a bonus to overall strategy performance. Notwithstanding, ongoing efforts should be made to promote the purchase by private buyers of vehicles with these proven features. Similarly, the benefits able to be derived from improved vehicle designs offering greater protection in impacts with pedestrians and cyclists, while strongly supported, have not been quantified (see Fig. 2).

5.7. Example 3 – Safe Speeds Cornerstone

Two main initiatives were considered under the Safe Speeds Cornerstone; first, achieving improved driver and rider compliance with existing speed limits and, secondly, reducing speed limits on selected types of road. The methods used to estimate savings in serious casualties by implementing the latter of these two initiatives are presented here to illustrate the approach taken.

The work of Nilsson (2004) and Elvik et al. (2004) resulted in the establishment and widespread use of scientifically robust relationships between changes in mean travel speeds and changes in the incidence of fatalities, serious injuries and other injuries. In essence, Nilsson found that a decrease in mean travel speeds resulted in decreases in the incidence of fatalities, serious injuries and other injuries according to power functions. The decrease in injuries is a function of the decrease in mean speed raised to the fourth, third and second powers, for fatalities, serious injuries and other injuries, respectively.

To estimate the potential savings in serious casualties as a result of reducing the speed limits in existing 60 km/h zones to 50 km/h, Nilsson's relationships were applied to actual injury data for WA, on the basis that this potential reduction in speed limit would reduce the mean speed in 60 km/h zones from 40 km/h to 37 km/h. This estimate was based on actual mean travel speed data and previous research by Haworth et al. (2001) which estimated that such a speed limit reduction would produce a 3 km/ h reduction in mean speeds. No additional enforcement is expected to be needed in order to achieve this reduction in mean speeds. These assumptions resulted in the calculation of an overall serious casualty reduction factor of 0.21, derived as a weighted average of the fourth power exponent applied to the actual average annual number of fatalities and the third power exponent applied to the actual average annual number of serious injuries occurring throughout WA's 60 km/h speed zones. It was also assumed that this change in speed limits would not take place until Year 5 of the strategy to enable agencies to prepare for the change, including supporting the somewhat controversial initiative (refer to the section on community consultation) with appropriate education and publicity in advance of the change. The model also included a reduction in 110 km/h posted speed limits to 100 km/h (Year 5) as well as the remainder of state speed limits being reduced by 10 km/h in Year 8 (see Fig. 3).

Fig. 2. Estimated cumulative serious casualty savings by year compared with ''business as usual" for the modelled package of vehicle safety countermeasures.

Fig. 3. Estimated cumulative serious casualty savings by year with respect to ''business as usual" for a reduction of 60 km/h speed limits to 50 km/h and 110 km/ h posted speed limits to 100 km/h in Year 5, with the remainder of state speed limits reduced by 10 km/h in Year 8.

5.8. Combining the effects of all initiatives

As noted earlier, if initiatives target the same crashes, then the total savings from the combined effects may be estimated by multiplying together the individual estimates, whereas when the initiatives target different crashes, then the savings may be added to estimate the overall effect.

In the case of the initiatives illustrated above, all vehicle safety features except for ISA target different crash types and therefore their combined effect can be estimated by adding the individual effects. ISA, on the other hand, targets all crash types and so the total savings in serious casualties from all vehicle safety features would be estimated by multiplying the sum of the other three vehicle safety initiatives by the reduction factor for serious casualties from ISA. The overall saving from the initiatives spanning the three Safe System Cornerstones would then be estimated by multiplying the total savings from vehicle safety features by the reduction factors for serious injuries for the infrastructure investment initiative and for the speed limit reduction initiative, as these latter initiatives act on the same crashes.

6. Results

6.1. MUARC's recommended Safe System strategy

Fig. 4 shows the relative performance of each of the groups of initiatives (as shown in the WA Safe System Matrix section) that were selected for consideration for inclusion in the strategy. From left to right, the bars represent the projected cumulative serious casualties estimated over the 12-year life of the strategy for infrastructure (safer roads and roadsides), enhanced enforcement, speed limit reductions, vehicle safety measures and an aggregate behaviour change program.

The above initiatives were combined to create what was called the 'Optimum' Safe System Option (OSSO). Here, 'optimum' is defined as a pragmatic optimum, derived from consultation with representatives from responsible agencies in WA. Defining this optimum had regard for what would be realistic levels of input from agencies within each cornerstone, while still recognising practical constraints on agency delivery, levels of maturity of new vehicle technology and the influences of political and community views. The OSSO was estimated to yield cumulative savings of 16,000 serious casualties over the 12-year duration of the strategy, reducing annual serious casualties from 2962 in

Fig. 4. Performance of individual initiatives, if implemented in isolation.

2006 to around 1420 serious casualties per annum by Year 12 in 2020. This represents a 52% reduction in serious casualties compared with 2006. It is estimated that by the conclusion of Year 12 the number of state-wide fatalities annually could reach as low as 100 (c.f. an average annual base level of 202 for the 3 year period from 2005 to 2007).

As traffic growth has a significant influence on model outputs, as well as being somewhat variable, a range of traffic growth figures were investigated. The results, shown in Table 2, demonstrate how the model is easily able to show how serious casualty outcomes might change with movements in other variables. The effect of lower traffic growth is to improve the performance of the strategy, while higher traffic growth results in greater cumulative serious casualty savings, but worse strategy performance overall, highlighting the potential risks in strategy implementation.

6.2. Strategy alternatives

Several variations to the Safe System strategy were modelled, to investigate the effects of some of the compromises that might be considered for each of the main Safe System initiative groups. Each change was modelled while keeping the others at levels specified by the OSSO. The following delays were examined:

- Infrastructure: reduce annual investment by 30% (found to result in a similar number of savings in serious casualties forgone as if the investment in infrastructure were delayed by 2 years).
- Speed: delay 60–50 km/h, 110–100 km/h until Year 5 (from Year 1); delay 10 km/h reduction until Year 7 (from Year 3).
- Enforcement: delay enhanced enforcement until Year 5 (from Year 1).
- Vehicles: delay ESC, side/curtain airbags until Year 3 (from Year 1); delay ISA until Year 7 (from Year 3); no change for Active Head Restraints (Year 5)

No changes were made to the aggregate behaviour change program. Fig. 5 shows the results of each change, based on the effects of delaying one initiative at a time. Should multiple delays occur, the cumulative effects on serious casualty savings would be considerably greater.

6.3. Limitations of modelling approach

With any form of forecasting, there will potentially be a degree of uncertainty about the direction, magnitude and sustainability of predicted change. The modelling method described

Table 2 Effects of varying traffic growth on model outputs.

Output estimate	Traffic growth
	2.6% (base)	2.0%	3.0%
Cumulative serious casualty savings over 12 years (rounded to nearest 500)	16,000	15,500	17,000
Percentage reduction in Year 12	52%	57%	49%
Estimated serious casualties in Year 12 (rounded to nearest 10)	1420	1290	1520
Estimated fatalities in Year 12 (rounded to nearest 10)	100	90	110

Fig. 5. Effects of varying initiative commitment on cumulative serious casualties saved.

here clearly has a number of limitations on its reliability and accuracy. Some of the main limitations are discussed briefly below, though a full discussion of the effects of these limitations is not practical here.

quality and consistency of reported crash data describing fatalities and serious injuries;
accuracy of estimates of effectiveness of initiatives and countermeasures being modelled, including the transferability of non-local research findings to other jurisdictions;
quality of other, non-safety, data such as traffic growth, vehicle sales and future investment in general road improvements;
expectations that historical trends in society in general and the transport system in particular will continue to be highly influential in defining future trends;
future trauma savings due to infrastructure investment are substantially underestimated as only the savings accrued during the life of the strategy (usually 10–12 years) are counted, whereas much of the infrastructure investment will continue to deliver trauma savings for up to 30 years. A similar argument applies to vehicle safety, though the average life of a passenger vehicle is about 10–12 years, thereby limiting this effect as it relates to vehicles.

To date, the model has not been developed to investigate the effects of shifts in travel mode, such as the growth in motorcycling, walking and public transport use, for example. New projects are currently exploring these possibilities.

To address these and other concerns, the modelling approach has been highly conservative in estimating savings in severe trauma. For example, the general philosophy of the modelling process assumes:

lower levels of effectiveness of initiatives;
slower implementation/take-up of initiatives;
higher costs to achieve the predicted savings;
sensitivity testing of predicted traffic growth, generally adopting conservative values.

The overall effect of this typically conservative approach has been to ensure governments are not unwittingly promising to deliver more than is realistically possible for an agreed level of investment. In jurisdictions with highly ambitious aspirations, modelling can readily be undertaken in a way that would genuinely 'stretch' the available resources to maximise the achievement of trauma reductions. There appears to be no single correct approach; instead, the modelling can be carried out to suit the aspirational preferences of a particular jurisdiction.

Finally, it is important to recognise that the greater value of this form of modelling lies in establishing the relative benefits of alternative strategy options rather their absolute values. The METS modelling approach is presented as a macro-level approach rather than creating the unjustified 'illusion of precision' (Sherwood, 2009). As a result, there are many road safety initiatives that were not modelled explicitly, primarily because, individually, the predicted savings are small and therefore incompatible with the 'macro' nature of the overall approach. Their aggregate effect, however, can be significant, yet is not counted here and instead is regarded as a 'bonus' contribution to the adopted strategy. In the case of continuing initiatives in place prior to the strategy, some of these benefits will be implicit in the adjustments made for the effects of increased motorisation.

7. The community consultation

To promote community, stakeholder and government acceptance of the ambitious approach taken to the development of the strategy, the Road Safety Council (RSC) undertook an extensive consultation process parallel to MUARC's modelling work. The consultation had as its basis the fundamental belief that the community should be provided with the best evidence about what works no matter how controversial that may be so that it can debate and consider the options available to improve safety. This approach was strongly supported by the Road Safety Minister.

7.1. Process

The RSC called on the expertise of community engagement specialists Estill and Associates and market research company Synovate to develop, undertake and analyse the community consultation. Over 4000 people directly participated in the consultation.

The consultation process involved three phases:

- Phase 1 gathered the community's views on road safety and introduced it to the Safe System philosophy and the concept of shared responsibility through a series of community forums hosted by local Members of Parliament in their constituencies and an online survey.
- Phase 2 presents the findings of the first phase of consultation and the recommended package of initiatives developed by MUARC to the community for comment through a discussion paper (with online and paper based feedback options), community forums and a survey of a scientifically representative sample of the broader population.
- Phase 3 involves the communication of the endorsed strategy to the public, stakeholders and the Parliament.

A bipartisan Parliamentary Reference Group with an independent chairperson was also established and consulted in each phase.

The three phases of the consultation process are further detailed in Appendix A.

7.2. Results

It was considered likely that the people who took the time to attend a forum or complete the online or paper-based questionnaire would have a stronger interest in the road safety initiatives proposed by MUARC than would the wider community. It was possible that they would be either more strongly in favour or more strongly against them. In contrast, the representative population sample was selected to be representative of the broader community's age, gender and geographical location and was considered to more accurately reflect the opinions held about road safety by the population at large. On the whole, as shown in Table 3, both groups supported the majority of the strategy proposals made by MUARC.

The research showed that the community was highly concerned about reducing death and serious injury on our roads and recognised that this is a shared responsibility of the entire community. The majority of the community supported MUARC's recommended road safety strategy; however, the level of support varied for individual initiatives and amongst different demographics. In particular there was strong support for initiatives aimed at maximising compliance with the road rules among road users, improving Western Australia's road network and enhancing the safety of vehicles travelling on our roads.

The representative sample indicates that the community was more divided on reducing speed limits in regional, urban and shopping areas, though it was far less resistant towards these changes than the group that completed the available, open survey. To add further complexity to the issue, approximately half of those who did not support speed limit reductions still believe these changes would be at least somewhat effective in reducing death and serious injury on the roads. Support for speed limit reductions is higher among women and those with dependents and lower among men, people without dependents and motorcyclists.

There is a statistically significant difference between the views of the representative population sample and the views of those with an existing interest in road safety who attended the forums or completed the open questionnaire. It appears that those with a passion for road safety are actually less open to speed reductions than the general population.

The community's main concerns surrounding speed limits changes were that irresponsible road users would continue to speed without better enforcement, while responsible road users would feel the impact of reduced travel times.

7.3. Implications of community consultation results for MUARC's recommended strategy

In light of the scientific evidence the RSC was convinced that the initiatives recommended by MUARC would effectively reduce the level of death and serious injuries on WA roads. With the high level of community support for MUARC's recommended initiatives to improve road user behaviour, the safety of roads and roadsides, the enforcement of speeding and vehicle safety, the RSC endorsed those initiatives as recommended by MUARC. In addition, the RSC endorsed additional research and evaluation projects for each of the cornerstones and undertaking to acknowledge and encourage some of the programs the community would like to see (such as, first aid training and attitude enhancement programs), provided they are not measures with proven negative impacts.

However, in responding to the divided community views on the desirability of speed limit reduction, the RSC did not adopt MUARC's recommendation to reduce most of the state's speed limits by 10 km/h. Instead, the RSC determined that it should undertake a range of demonstration projects to illustrate the effects of speed limit reductions; adjust speed limits to complement road infrastructure; introduce more consistent speed limits and develop a community education program to enhance community understanding of the benefits of speed limit reduction.

MUARC found that this amendment to its Optimal Safe System Option will reduce the strategy's potential to save people from being killed or seriously injured on WA roads from 12,600 to 11,000 over the 12 year life of the strategy. Nonetheless, by being able to clearly demonstrate the benefits and community views on this recommended initiative, the RSC has been able to give the Government and Parliament a high level of assurance that the strategy will be widely supported by the community which has led them to be more receptive and supportive of it.

Following a change of Government in September 2008, the Minister for Road Safety announced the new Government's endorsement of the policy direction established in Towards Zero in Parliament on 19 March 2009 and received bipartisan support. Implementation is now progressing with widespread community and political support.

Table 3 Snapshot of results of community consultation^a .

Safe road use	Safe roads and roadsides	Safe speeds – enforcement	Safe vehicles
94% supported or strongly supported education programs	95% supported or strongly supported improving intersections	74% supported or highly supported this initiative (with an additional 18% somewhat supporting the measure)	90% supported or strongly supported the ongoing promotion of vehicle safety features
86% supported or strongly supported improved policing of the road rules	91% supported or strongly supported improving lengths of road or roadsides and black spots 74% supported or strongly supported building roadside barriers 68% thought that improving the roads would be an effective or highly effective approach to improving road safety	Safe speeds – limit reductions 48% supported or strongly supported reducing speed limits near shopping centres 40% supported or strongly supported reducing regional speed limits from 110 km/h to 100 km/h 40% supported or strongly supported reducing urban speed limits from 60 km/h to 50 km/h	79% supported or strongly supported safety features on all Government vehicles
		31% believed speed limit reductions would be effective or highly effective (with an additional 27% thinking it would be somewhat effective).

^a Unless otherwise specified, the percentages shown represent the responses of the representative population group – who were randomly sampled by Synovate. The full report outlining the results of the community consultation can be found at: http://www.officeofroadsafety.wa.gov.au.

8. Conclusion

The Western Australian government decided to take an ambitious approach by developing a new road safety strategy in accordance with the principles of the Safe System, rarely attempted elsewhere. Based on the underlying structure of a mathematical model developed by the Monash University Accident Research Centre in Melbourne, Australia, the METS–WA model allowed comparisons in terms of cumulative serious casualties saved to be made between the various elements of the Safe System, including the Cornerstones incorporating safe roads and roadsides, safe speeds, safe vehicles and safe road use. METS–WA was also used to combine these initiatives to predict overall road safety strategy performance over its 12-year duration and investigate the effects of delays in implementation or changes in predicted traffic growth.

The strategy subsequently developed aimed to represent the best-practice solution achievable and has the potential to result in large reductions in severe trauma in Western Australia if adopted fully, with annual serious casualties estimated to be reduced by around 50% over 2006 levels (Office of Road Safety, 2008).

This is the first time Western Australia has conducted widespread community consultation during the development of a road safety strategy and it is regarded as crucial to establishing the foundations for endorsement and implementation.

To promote community acceptance of the Safe System philosophy and a bold, long term vision for road safety, it has been important to share the research with the community and listen to its views. The views of the community and the scientists on reducing road trauma were important considerations for Government in endorsing the recommended Towards Zero strategy.

The successful adoption of the Safe System framework depends upon all participants in the road-transport system accepting their share of the responsibility for its safe operation. This community and stakeholder consultation process increases the chance that those objectives will be met as Western Australia now begins to build its capacity to implement Towards Zero.

Acknowledgements

The development of the original METS model was funded by VicRoads. Its permission to use the underlying model in Western Australia is gratefully acknowledged, as is the permission of the Western Australia Office of Road Safety to publish this paper.

Appendix A. Summary of community consultation

What	Purpose	How	Where	Respondents	Note
Community forums (phase 1)	To inform the participants of road safety statistics in their local area and help participants gauge their own awareness of local road safety issues and driving behaviours In addition the workshops allowed for community debate on road safety and promoted community engagement in road safety	Hosted by members of state Parliament in their constituencies – all members of the community were invited to attend through media, advertising and direct invitations (from the MP)	36 forums were conducted in 19 locations	1326	Two workshops were held in most of the locations. The morning session was for community representatives (mainly students and apprentices). The afternoon session was for community leaders and 'influences', including those that, as employers, have some responsibility for the safety of their employees. Handheld electronic devices, which ensured anonymity, were used to record (and capture) participants' responses
Online survey (phase 1)	This was an internet based version of the questionnaire presented at the phase 1 community forums, allowing interested people who had not been able to attend a forum to still have input to the first phase of community consultation	Advertisements were placed in The West Australian and the details were emailed to an extensive database of people who had previously indicated an interest in road safety The survey also received extensive media coverage	Online	1188	Of the respondents 71% were male. The majority of the sample (43%) was aged between 40 and 59. A large proportion held a motorcycle licence (40%). This group cannot be considered a representative sample as there are too many males, too many respondents in full-time employment and the age groups do not represent the wider community However, the fact that the respondents took the trouble to complete the online survey suggests that they hold considered opinions and are concerned about road safety. Hence, while not a representative sample of the broader population, their views are being considered (continued on next page)

Appendix A (continued)

What	Purpose	How	Where	Respondents	Note
Community and stakeholder forums (phase 2)	This phase involved the community and stakeholders in providing feedback on the MUARC recommendations, which were presented as the Toward Zero draft	Advertisements and some articles were placed in local newspapers. Invites were sent to stakeholders and relevant	In eight locations across the state (metropolitan, regional and remote)	807	The attendees of these forums were largely stakeholders directly involved and concerned about road safety
Online feedback on 'Towards Zero' – MUARC's	discussion paper recommendations (phase 2)	database contacts The Towards Zero draft discussion paper was presented online, together with the feedback form allowing people who were unable to attend the community forums to submit their feedback online	Advertisements and some articles were placed in local newspapers. Details were sent to stakeholders and relevant database contacts. Promoted at the community forums	Online
This was an open 'survey'. Any interested person could complete it
Independent representative sample	The primary purpose of this research was to obtain feedback from the larger community and not just from those with an interest or with strong opinions on road safety	Recruited via a phone survey conducted through an independent social research company – Synovate	Respondents were randomly contacted by phone and were given the option to receive and then submit their feedback online or as a hard copy via post	649 (correct to ±3%)	The independent research company ensured this sample was representative of the community gender and age spread and included metro, regional and remote participants This representative community survey ensured a balanced view from across the community and all voices are heard – not only the loudest This is considered to be the most reliable representation of the thoughts and opinions of the general population
Additional specific forums (industry leaders, special interest groups (including cyclists and pedestrians) and Road Safety Council agency stakeholders)	These forums encouraged direct feedback from specific groups including vulnerable road users such as motorcyclists, the police and representatives of large companies	Invitations were sent to specific groups, companies, organisations, etc.	Perth convention and Exhibition centre	200

References

Aarts, L., Van Schagen, I., 2006. Driving speed and the risk of road crashes: a review. Accident Analysis and Prevention 38, 215–224.
Corben, B.F., Logan, D.B, Johnston, I.R., Vulcan, A.P., 2008. Development of a New Road Safety Strategy for Western Australia 2008–2020, Report Number 282, Monash University Accident Research Centre, June.
Elvik, R., Christensen, P., Amundsen, A., 2004 Speed and road accidents: an evaluation of the Power Model. Oslo.
Haworth, N., Ungers, R., Vulcan, A.P., Corben, B.F., 2001. Evaluation of a 50 km/h Default Urban Speed Limit for Australia. Monash University Accident Research Centre, Melbourne.
Jacobs, G., Aeron-Thomas, A., Astrop, A., 2000. Estimating global road fatalities. TRL Report 445. Transport Research Laboratory, Berkshire, UK.
Lie, A., Tingvall, C., Krafft, M., Kullgren, A., 2006. The effectiveness of electronic stability control (ESC) in reducing real life crashes and injuries. Traffic Injury Prevention 7, 38–43.
Meuleners, L., Hendrie, D., Legge, M., Cercarelli, L.R., 2005. An Evaluation of the Effectiveness of the Black Spot Programs in Western Australia, 2000–2002.
Injury Research Centre, School of Population Health, The University of Western Australia, RR155. Crawley, WA.
Nilsson, G., 2004. Traffic safety dimensions and the power model to describe the effect of speed on safety, Doctoral thesis, Department of Technology and Society, Traffic Engineering, Lund, Sweden. http://www.lub.lu.se/luft/diss/ tec\_733/tec\_733.pdf [30.6.09].
Office of Road Safety, W.A., 2008. Towards Zero Recommended Strategy. Perth, Western Australia.
Regan, M.A., Triggs, T.J., Young, K.L., Tomasevic, N., Mitsopoulos, E., Stephan, K., Tingvall, C., 2006 On-Road evaluation of Intelligent Speed Adaptation, Following Distance Warning and seatbelt reminder systems: final results of the TAC Safecar project. IN CENTRE, M.U.A.R. (Ed.) MUARC Report Series, Clayton, Monash University Accident Research Centre.
Scully, J., Newstead, S., 2008. Evaluation of electronic stability control effectiveness in Australasia. Accident Analysis & Prevention 40, 2050–2057.
Sherwood, O., 2009. ''The illusion of precision". Personal communication, 8 April.
Vulcan, A.P., Corben, B.F., 1998. Prediction of Australian road fatalities for the year 2010. Presented to: National Road Safety Summit, Canberra, ACT.

corbenstrengtheningroadsafetystrategy2010.md 85 KB Előzmények Nyers