Third Open Letter to Executive Director of PACTS

To Robert Gifford

Executive Director, The Parliamentary Advisory Council for Transport Safety

Dear Rob

Thank you for sending me a paper claiming that seat belts have saved 57000 lives in the UK between 1983 and 2007.  Is this the source of the 60000 claim posted on  your website in 2008?

I have discovered that the paper you have sent me (by DC Richards, R Hutchins, RE Cookson, P Massie and RW Cuerden) is accessible on line at http://bast.opus.hbz-nrw.de/volltexte/2012/527/pdf/ .  

The paper documents an increase in seat belt wearing rates for drivers of well over 100%. I have no reason to doubt the accuracy of the data represented in their Figure 6.


Figure 6. Effect of seat belt legislation on seat belt wearing rates for car drivers

It then estimates that seat belts are 61% effective in preventing fatalities in crashes. The fact that wearing a seat belt considerably improves one’s chance of surviving a crash is also not disputed.

It then calculates (their Figure 5) how many lives have been saved by seat belts since the UK law came into effect.


Figure 5. Number of car occupant casualties prevented by seat belts

It also presents a graph showing how many fatalities remain. (their Figure 4)


Figure 4. Fatal and serious car occupant casualties

And herein lies a mystery.

Richards et al have charted a large increase in the use of a safety measure that provides significant protection in crashes. This measure they claim saved over 2500 lives a year (and rising!) in the first 8 years of the law. With such a sudden large increase in the adoption of such an effective safety measure one should see a large downward step in the established downward trend. Where is it? One cannot see any effect in the graphs here and here . Richards et al don’t even look for it; their graphs don’t start until 1983. They appear remarkably uncurious about why it should be that in the first eight years after the law the more lives seat belts were saving (Figure 5), the more car occupants were being killed (Figure 4). And they display no interest in the possibility of risk compensation: “factors such as any change in driving behaviour for occupants who do / do not wear a seat belt were not taken into account” (p 229).

Further they make no comment on the fact that the reduction in driver fatalities that did occur in 1983 (the year in which evidential breath testing was introduced) consisted almost entirely of drivers with alcohol in their blood.


Great Britain driver deaths by place and alcohol level in dead driver. Source: Broughton and Stark 1986

Three questions:

1.Do you consider that the paper by Richards et al validates the 60000 claim still on your website?

2.If so can you explain the absence of the large downward step that should appear in the graphs here and here ?

3.Can you explain why seat belts were so extraordinarily selective in saving the lives of drivers who had been drinking?

Finally you say:

“I do not wish to be drawn into the debate about the benefits/disbenefits of cycle helmets. PACTS’ aim must be to improve the safety of all classes of road user and tackle the issue of the disproportionate risk placed upon those least able to protect themselves.”

It was not my intention to draw you into a debate about the benefits/disbenefits of cycle helmets; the PACTS reporton this subject seems to me to be admirably clear and balanced. But I am interested in PACTS’ aspiration to “tackle the issue of the disproportionate risk placed upon those least able to protect themselves.”

I am seeking confirmation of a point I raised in my last open letter. From your Significance article it appears that it Is PACTS’ position that a measure that saves the lives of motorists is justified so long as the number of vulnerable road userskilled as a result is smaller. Is this your position?


No ping yet

  1. Guy Chapman says:

    The facts, I think, speak for themselves, so why will the “road safety” establishment not acknowledge this? I believe they’ve already acknowledged that the effect ascribed to seat belts at the time was more likely to be a result of the contemporaneous introduction of stiffer drink-drive laws, so why is it hard to admit that the claims of lives saved are unsupportable?

    I believe much of the answer lies in cognitive dissonance. The road safety lobby worked hard to get the seat belt law passed, it was a major cultural shift, probably the first time since the debates over the speed limit in the first half of the 20th Century when the pendulum swung decisively in their favour. Now we have the Tories shaping up to remove fixed speed cameras, the pendulum is swinging back towards the motorist and against the road safety establishment (in truth the pendulum has rarely reached even the midpoint, libertarian motorists have always been on the winning side in any debate around road danger reduction).

    To come along right now and publicly slaughter one of their sacred cows is going to be unpopular; to ask them to display the carcass in their shop window is perhaps asking too much.

  2. Jim Tubman says:

    Guy Chapman’s point about cognitive dissonance is well made. The mental process (I hesitate to call it “thinking”) seems to be, “This measure ought to work, therefore it does work, so there is no need to compare data from before and after, or to consider other possible explanations.”

    Another thing to keep in mind is that the safety establishment tends to measure its success on the basis of compliance, not actual demonstrable reductions in deaths and injuries. This may follow from the point made in the previous paragraph, or, more cynically, it may be from a desire to avoid accountability for results.

    I think that the best one could hope for, assuming that PACTS has sufficient intellectual integrity to examine the data and accept what it says, would be for them to quietly drop the 60,000 figure. To have them declare loudly that they were grossly in error on this matter, and have been for years, is too much to expect.

  3. Guy Chapman says:

    Jim, you make a good point about the measurement of success. We see this in cycle helmet laws. The law is proposed based on the erroneous claim of 85% injury reductions, then when the injury reduction turns out to be 0% they claim success on the basis that more cyclists are wearing helmets, therefore they *must* be safer even if the figures show they aren’t.

  4. NCPLH says:

    Europe is moving towards integration of alcohol licences. This will be like a dream come true for responsible drinking!

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>