One of the curiosities of this subject is that the evidence that speed kills is much flimsier than you might expect. Government ministers sometimes say that the link is well-established, but rarely spell out evidence for it. For example, the road safety minister Jim Fitzpatrick in a speech last month simply said:
Tackling the problem of speeding will remain a key priority. Indeed, one of the biggest challenges of the next decade will be to make speeding as unacceptable to mainstream society, as drink-driving has become.
Although there are still far too many drivers who regularly break speed limits, we are slowly winning the battle against speed. That process must continue, backed by speed cameras where they are effective, by a police force focused on successful enforcement, and by THINK! advertising emphasising the message that speeding is not only dangerous, but also unacceptable.
The obvious place to look is the Transport Research Laboratory (free registration required to download reports, most .pdfs are free), but after browsing back through the last 30 years’ reports I could not find one specifically on the evidence that speed kills. There were some which mentioned it in passing, for example TRL296, published in 1998, included:
In 1987, 40 American states raised speed limits on rural interstate highways from 55 mph to 65 mph … Controlling for a wide variety of factors known to affect speed and accident data (such as vehicle miles travelled, seat belt laws, seasonal effects, long- term trends and economic conditions), several studies concluded that the increase in average speed of between 5 mph and 10 mph on those roads was the direct cause of the increase in accidents of between 19 and 30% … Earlier, in 1979, Denmark reduced its non-urban speed limits by 10 kph, which produced an overall reduction in average speed of around 2 kph and an 18% cut in rural injury accidents …
Yet the matter of speed-accident links is not actually as straightforward as this, and a range of other variables and facts complicate the picture. For instance, although absolute speeds on motorways are considerably higher than on other roads, accidents are least likely to occur there. Although some correlation has been found between individuals’ reports of their habitual speeds and their accident involvement, it is also the case that many faster- than-average drivers have avoided accidents altogether or have a lower than expected accident rate even when their mileage is taken into account. For some drivers, slower speeds seem to induce lower levels of attention and concentration and higher speeds (and speeding convictions) are associated with higher alertness and arousal. Lower speed limits may increase drivers’ perceptions of safety and corresponding risk of accident, through creating a false sense of security. Further, as traffic police officers often told us in this present study, it is not necessarily the case that fast speeds per se are dangerous. It is inappropriate speed for the circumstances that is risky. Thus very low as well as very high speeds on a dual carriageway or motorway can lead to accidents if they increase the speed variance on the road.
Moreover, re-analysis of US data on fatal accidents occurring after the raise in speed limits on rural interstates suggests it is incorrect to say that the change led directly to an increase. … when the statewide fatality rate was used as the dependent variable instead of the rural interstate fatality rate, there was an decline in state-by-state fatality rates of between 3.4 and 5.1%.
Hardly the ringing endorsement of lower speeds that politicians are fond of claiming.
There is one rather key distinction. Consider the following two propositions:
(A) drivers driving significantly faster than the average speed of drivers on a particular road are more likely to be involved in accidents;
(B) reducing average speeds (by a combination of enforcement action and lower limits) will save lives.
There has been substantial evidence for (A) for many decades. But the evidence also shows that driving significantly slower than the average also increases the risk of accidents. In other words, accidents are lowest when all vehicles are moving at roughly the same speed. It took quite a while for this to be generally accepted - people claimed, for example, that the slow speed effect was not real, but was due to accidents near junctions where traffic was slowing down to exit, or speeding up after entering. That turned out to false. There was still a substantial effect after eliminating that element. So slow driving is just as bad as fast driving. But attempts to use minimum speed limits have not been particularly successful, and it has proved much harder to whip up public feeling against slow drivers.
The evidence for (B) is trickier. Certainly, high-speed accidents tend to be worse than slow-speed accidents, so a higher proportion of high-speed accidents involve deaths than slow-speed accidents. But apart from that, the evidence in favour of lower speeds is not as strong as you might expect.
Post a Comment