6

In Australia, drivers earn "demerit points" for committing driving offenses. Being awarded more than a certain number of demerit points in a given period (typically 12 in any 3 year period) results in license suspension.

New South Wales, Western Australia and the Australian Capital Territory have "Double Demerit" periods - during long weekends and holiday periods when the death toll traditionally climbs, the point penalties for some driving offenses are doubled - e.g. speeding, seatbelt and motorcycle helmet offences.

The Australian Police Blog [a private blog] claims:

Records show that when double demerit points are in force, road fatalities are reduced. Since double demerit points were introduced in 1997, there has been a 20% reduction in fatal crashes over the relevant holiday periods.

ABC News claims:

Police say two decades of enforcing double demerits has resulted in at least 433 fewer deaths on New South Wales roads during holiday periods.

Does doubling the demerit point penalties cause drivers to be more safe on the roads?

Improve this question
6
  • 2
    "Does doubling the demerit point penalties cause drivers to be more safe on the roads?" This is only one possible mechanism. It could also be that licenses are getting suspended faster for bad drivers, before they can kill anyone over the next holiday.
    – Laurel
    May 31 at 11:07
  • @Laurel: .... Yeah, I guess that is true! Much tougher to measure though.
    – Oddthinking ♦
    May 31 at 11:46
  • @Laurel, probably, but they will surely count it as an intended part of the mechanism to support their claim. The whole point of demerit points (grr!) is exactly that: [threat of] removing drivers from the roads; by themselves, they cost nothing. Needless to say though, that the AP blog quote is insufficient to prove the point: fatalities have reduced across the board, and they must demonstrate that they have reduced more over long weekends with respect to ordinary periods of the same length (or ideally, of the same distance travelled).
    – Zeus
    Jun 1 at 2:10
  • @Oddthinking: a basic measure would be to compare the evolution of death rate in "double demerit" periods vs other periods.
    – Graffito
    Jun 1 at 22:06
  • According to wikipedia, the number of road fatalities in Australia was 1767 in 1997 and 1123 in 2021. i.e. a decrease of 36 %. So, what was the computation to assess the 20% reduction ?
    – Graffito
    Jun 1 at 22:26

1 Answer 1

Reset to default
8

There is a report evaluating the trial of the double demerits scheme in Western Australia in 2002/2003: Evaluation of the trial period for double demerits legislation in Western Australia. To compare the effectiveness of the measure they compared crash data between the holiday periods in which the scheme was active and similar periods in the previous years. As a control they did the same comparison for non-holiday periods (in which the scheme was not active). They found the following:

During the trial period for Double Demerits in 2002/2003, the total number of Reported crashes was down 11% during Double Demerit periods and the reduction was 3.4 times higher than that observed during non-Double Demerit periods (ie. relative degree of effect index = 3.4). Fatal crashes were 20% lower during Double Demerit periods, compared to a 9% increase during non-Double Demerit periods; Injury crashes were down 18% during Double Demerit periods and the reduction was 1.7 times higher than that observed during non-Double Demerit periods.

Total crashes where Speed was a factor were down 40% during Double Demerit periods and the reduction was 1.6 times higher than that observed during non-Double Demerit periods. Fatal crashes where Speed was a factor were down 52% during Double Demerit periods, compared to a 19% increase during non-Double Demerit periods; Injury crashes where Speed was a factor were down 43% during Double Demerit periods and the reduction was 1.5 times higher than that observed during non- Double Demerit periods.

Fatal crashes where Alcohol was a factor were down 54% during Double Demerit periods and the reduction was 10.3 times higher than that observed during non-Double Demerit periods.

While Fatal crashes where non-use of Restraints was a factor were down 7% during Double Demerit periods, they remain slightly higher during Double Demerit periods compared to non-Double Demerit periods. However, Injury crashes where non-use of Restraints was a factor were down 39% during Double Demerit periods and the reduction was 1.6 times higher than that observed during non-Double Demerit periods.

Based on this evidence they recommended making the scheme permanent. Some possible caveats to this report. There were other factors that could impact the result such as 8-fold increase in police enforcement activities during the trial periods. Of course the report does not say anything about the long term impact of the measure.

Improve this answer
4
  • 1
    Those results sound extremely good. So good infract that they sound a bit too good to be true.
    – pinegulf
    Jun 2 at 11:14
  • I've been thinking about self-answering this based on my research since. This article seems to be the key one, but I found it less than convincing, and I am trying to figure out if my biases are showing before responding.
    – Oddthinking ♦
    Jun 2 at 16:19
  • 2
    Yeah I see an obvious problem. Being the first attempt it likely was all over the news and talked about making drivers cautious. If you make this a standard policy it won't be in the news any more and most drivers will forget it's in effect, thus removing most of the power of the test. Unfortunately I don't have a good experimental way of better predicting the long term effects once such a policy is no longer novel.
    – dsollen
    Jun 2 at 17:33
  • While not exactly discussing double point demerit, more recent paper etrr.springeropen.com/articles/10.1007/s12544-010-0027-0 does not find correlation in implementation of single merit implementation.
    – pinegulf
    Jun 12 at 12:09

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .