That damn war on motorists, eh? Not only did they put a bus lane on the M4 (which actually improved traffic flow significantly), now they’re putting 20 mph speed limits on roads! Naturally, the Daily Mail is unhappy, and now they have the ammunition to take these speed limits down! M… maybe.
Why death rates INCREASED in 20mph zones… and getting rid of cameras reduces accidents says today’s Mail. Now, you might think that with the headline beginning with the word “why”, that mean the Daily Mail was going to tell us why death rates increased. Of course, they don’t. That might be because death rates didn’t really increase at all in any meaningful sense.
The report the Daily Mail refers to is the snappy-sounding Interim Evaluation of the Implementation of 20 mph Speed Limits in Portsmouth (PDF). Already, there’s one problem – the report is looking a particular type of 20 mph speed limit, where there are none of the usual speed bumps or chicanes, and comparing it to regular 20 mph speed limit schemes. You can’t draw grand conclusions based on a single outlier, as the Daily Mail tries to.
Secondly, here are the death rates that the Mail refers to: in the year before speed limits were introduced, 19 people died on the roads of Portsmouth; in the year after, 20 people died. That’s a change of just one person. Death rates are going to vary slightly from year to year anyway – car crashes are basically random, and some years will be slightly worse than others. A variation of just one person may be tragic, but it’s not statistically significant, as the report makes clear. There’s no way of telling whether that one death was caused by the speed limits or just chance.
Using the much better casualty data, also in the report, the casualty rates fell from 183 per year before traffic calming to 142 per year afterwards. That’s a change of 41 casualties, or 22% – a much more significant change, and better than the national average of 14%.
The Daily Mail also tries to point to the fact that “Motorists’ average speeds reduced by only 1.3mph to 18.5mph.” While this is true, this is because a lot of the roads the speed limit was applied to were narrow side roads where it wouldn’t be smart to drive above 20 mph anyway. However, on the worst roads, where before the average speed was 28.5 mph, the speed limit brought that down to a much safer 22.2 mph (table 3.3).
Finally, because the actual point of the report was to compare the scheme to other 20 mph zones across the country, the report includes data from Hull and London as well. If the journalist who wrote this article had read this far into the report (table 8.2), he’d have seen that London (which would give you a much larger sample than Portsmouth anyway) saw a 56% reduction in traffic deaths, while in Hull the number of deaths dropped by an astonishing 90%. The casualty rates fell in London and Hull fell more sharply too, 43% and 56% respectively versus Portsmouth’s 22%.
Of course, the difference between Portsmouth and these two cities is that London and Hull both used speed bumps whereas Portsmouth didn’t. The actual conclusion that the report came to was that a speed limit was mildly effective at making roads safer, but not as effective as speed bumps. Of course “Report proves speed bumps save lives” probably isn’t a headline you’ll see in the Daily Mail any time soon.
Edit: The cycling site road.cc has a very good post about the same article.