Emeritus Professor Ann Williamson has provided her notes and thoughts in response to the Ministerial Roundtable on Road Safety, Adelaide, 11 July 2023.


Seat belts

Not wearing a seat belt is an issue because it only provides secondary protection and while they mitigate the effect of a crash after it occurs, the plays no role in the prevention of road crashes.  They are necessary but not sufficient solutions for road safety.

The first two dot points should be rephrased to say:

  • 22% of drivers were not wearing a seatbelt in fatal crashes. Drivers not wearing a seatbelt are 8.3 times more likely to sustain a fatal injury after a crash has occurred. 

These statistics should be interpreted in the context of very high rates of wearing seatbelts (self-reported) since rates are between 96 and 98% when sitting in the front and 95.2% when driving.  These have not changed much over recent years.

Change in seat belt use by drivers and passengers killed in crashes is comparatively small.

Remote areas have higher non-seatbelt use in fatal crashes (35-43%).  This indicates that fatal crashes are more likely in some areas and that there are some people in some circumstances (type of vehicle, geographic location, and different road type) who do not wear seatbelts who are more likely to be involved in these crashes.

Evidence that certain groups of professional drivers tend not to use seatbelts (taxi drivers, truck drivers), because they find them uncomfortable especially over long hours of driving.  Arguments used that seatbelts do not prevent injury, but they can trap a driver in a vehicle during a crash.

Seatbelt reminding technology may only be partially successful for people who already find them uncomfortable.  Many will find workarounds to disarm the technology, so it is not a solution for these people.

è Seat belt use is very high already, and we have already achieved this by actions so far.  Non-wearing is likely to involve some types of people, especially those who drive in some circumstances where crash risk is high.  We need to target these people.

We need to find out who doesn’t wear seatbelts, under what circumstances and target these. 

We need better data to understand who, where, when and why they do not wear seatbelts.

We need better designed seatbelts to suit the wide range and anthropometry of all drivers and ensure these are implemented through changing ADR’s to reflect these redesigned seatbelts.

Also, need studies of usability and effectiveness in use of seatbelt reminder technology.


  • What are the roles of government etc in improving seatbelt use?
    Ans – Better data and analysis of who, when, where and why seatbelts are not worn
  • What can be done to target seatbelt use in remote areas?
    Ans – Better data as above, but especially understanding reasons in remote areas
  • What are examples of successful interventions that could be adopted nationally?
    Ans – Depends on who you are targeting. Some workplaces have introduced more comfortable seatbelts.

Notes from meeting:

General agreement about the main issues in seatbelt non-use relate to specific driving groups and often relate to comfort issues.  The main take-away messages seem to be that we need better data on seatbelt usage in crashes and we need to work on issues of making seatbelts more usable/comfortable for road users/drivers who are at the extremes of the distribution.


Alcohol and drug driving

Alcohol trend for drivers looks like it has reduced somewhat in the decade, 2010 to 2021.  Motorcyclists have remained about the same.  Nevertheless, the involvement of illegal BAC in fatal crashes is still too high. 

We need to understand who is still driving with illegal BAC and why are they doing this.  Secondarily what kinds of crashes and conditions do these crashes occur.  This should be our target, not reducing the legal limit.

Enforcement of drink driving has produced a general deterrence effect which has produced a significant drop in crashes with illegal BAC since the 1970’s, but we got to the current level and this hasn’t really changed.  Reasons for this can include increased RBT activity so increasing driver’s perceptions that they are likely to be tested and penalised if they are over the limit. 

Drug driving

Primary issue is that drug testing for cannabis (THC) currently focusses on testing for presence of the drug in the person’s system.  This does not mean current state of intoxication due to the drug.  As a consequence, this testing is NOT about road safety, but the presence of the drug and since this test can be positive for days, if not weeks, a positive test does not reflect that the person is impaired and therefore increased risk of crashes. 

This means that studies relating crash risk for THC positive tests compared to sober/non-drug taking drivers is not actually reflecting anything to do with THC impairment.

The problems with this approach are that it does not reflect a road safety impact so may well have nothing to do with road safety and will not produce a benefit for road safety crashes.  Taking this approach means that road users have the right to point out that testing and penalties are about revenue raising and not about road safety.

Magistrate in NSW resigned based on this problem.


  • How can governments etc further reduce impaired driving?
    Ans – Understand who is using, when and where and why they are driving and target these causes where possible.
  • How can we better target enforcement and behaviour change at high risk groups/regions
    Ans – As above
  • How to separate drinking/drug use and driving?
    Ans – As above
  • Are new technologies needed for drug driving testing?
    Ans – Yes, must address impairment while driving not just drug use.


  • Are new rules needed to account for medicinal cannabis?
    Ans – Yes, as part of a review of how cannabis and THC are detected and whether it reflects impairment.  Review suggests medical cannabis does not impair driving.  This means you cannot rely on just presence of THC.


Notes from the meeting

Most of the discussion was around drug-driving and the issues of detecting cannabis and THC past usage but not current state which means many people are being penalised for having used the drug rather than for being impaired by the drug and so having little to do with making drivers safer at that time.




Risky road use

We need to clarify what we are talking about here.  Is it illegal road use, ie., speeding over the limit, breaking a road rule etc, or is it risky driving which includes actions that are not illegal but are or are potentially risky for the conditions and location at the time.

We need to understand much more about the involvement of these two types of risky road use and their involvement in crashes.  We also must be careful not to confuse them when talking to the public.

One important element here is the intentionality of the behaviour.  If I don’t know the legal behaviour (eg, the legal speed limit, the explicit rule at that place etc) then I am not intentionally being risky nor illegal.  This should be dealt with in a different way to the circumstance where the person deliberately breaks the road rule.

The point in the Discussion paper about people underestimating the riskiness of their own behaviour, because it isn’t as bad or risky as that of others is questionable.  What is the evidence for this?


Distracted driving

Defined as use of mobile phones.  This is justified as using mobiles is a risky behaviour when it requires the driver to take eyes off the road and attention from the driving task when they interact with their mobile phone.  Action to try to reduce this specific type of behaviour is also justified.  The evidence on effectiveness of bans and penalties for mobile use shows some decreases in use, but the effect is not as strong as hoped.  Seems that perhaps people perceive the likelihood of being detected as lower than it was.  Need more public awareness raising?

The even larger problem is that we are overlooking multiple other sources of distraction in modern vehicles which distract vision, attention and even require physical manipulation but which we allow.  Screen-based controls in vehicles are a feature in almost all new cars.  Why are we allowing this?


Speeding increases the risk of fatalities and injuries and also probably increases the risk of crashes occurring.  Consequently managing speed is a major concern for all aspects of road safety. 

Low-level speeding is stated in the Discussion paper to be a cause of most road fatalities, but this statement needs more evidence.  Statistically, a relationship is drawn between increases in speed and increases in risk of a casualty crash, but this does not acknowledge that the vehicle speed may not be the cause of the crash, but plays a role in the degree to which the person is injured.  This is not the same thing, For example, a crash caused when a driver merges into the right lane into the path of a passing vehicle may have little to do with the speed of the vehicle and more to do with the side-mirrors which provide inaccurate information about the distance of the passing vehicle when the driver first moved.  If this happens at a higher speed, there is a greater likelihood of injury. 

Our data on the causes of crashes simply gives us too little information about the causes of crashes to determine much more than a single cause of a crash.  We MUST do more to try to understand the cause of crashes especially those where we attribute risky behaviour to drivers/other road users.  When we understand the real causes of crashes, we can target countermeasures better.



  • Beyond enforcement, how to stop people using mobile’s while driving?
    – make enforcement much more obvious and likely.
    – use tech to stop mobile phone working while the vehicle is operating.
  • Changing risk perceptions in relation to low level speeding?
    – No, this isn’t the best approach, see above
    Reanalyse data on how low-level speeding is involved in crashes.  What other factors are involved and look at the interaction between those and speeding.
    2. Redesign speed limit use – make it very clear what is the speed limit in the particular location; is it 60, 70 or 80?  Frequent changes which are only signed at one point often leave drivers doing the previous speed which can be interpreted as low level speeding or higher.
  • What other approaches to minimise risky road use?

Help drivers to be safer and remove many of the risks in the road environment or help the driver to interpret them.  This includes:


Road design problems

  • Clear navigation signage and have multiple signs strategically located
  • Remove confusing lane markings
  • Reduce visual clutter and complexity in driving environments

Vehicle design problems

  • Problems of vision from vehicles
  • Side mirror problems
  • Distracting technology

Road rules that make it hard for drivers to comply

  • Speed management problems


Notes from meeting

This is a huge area but unfortunately the discussion was quite truncated due to lack of time.  There was some discussion around the use of technologies (cameras etc) to detect driver distraction and enforcement activities around them.  There was also some discussion on the effects of low-level speeding as a major area of concern.  I pointed out the problem with attributing low-level speeding as the cause of crashes, but found considerable opposition this this point on the basis that reducing speeding will mitigate the consequences of crashes by reducing the forces that increase the risk of fatalities and very serious injury.  This is clearly not an approach that focusses on prevention of crashes, which I also pointed out and that our approach should not just be reducing fatalities but allowing crashes and injury to continue to occur.  Time ran out before this discussion was complete.  It certainly highlights an area that HFESA and others can keep reminding decision-makers about the need to broaden thinking about what to target in road safety.

Skip to toolbar