More cars, fewer pedestrian deaths

by on May 28, 2014 at 6:52 am in Books, History, Travel | Permalink

Michael Blastland and David Spiegelhalter have a new book about risk — The Norm Chronicles: Stories and Numbers About Dangers and Death — and it does actually have new material on what is by now a somewhat worn out topic.  Here is one example:

In 1951 there were fewer than 4 million registered vehicles on the roads in Britain.  They meandered the highways free of restrictions such as road markings, traffic calming, certificates for roadworthiness, or low-impact bumpers.  Children played in the streets and walked to school.  The result was that 907 children under 15 were killed on the roads in 1951, including 707 pedestrians and 130 cyclists.  Even this was less than the 1,400 a year killed before the war.

The carnage had dropped to 533 child deaths in 1995, to 124 in 2008, to 81 in 2009, and in 2010 to 55 — each a tragedy for the family, but still a staggering 90 percent fall over 60 years.

You can buy the book here.

1 Michael Foody May 28, 2014 at 7:04 am

My guess: Fewer Pedestrians

2 Ray Lopez May 28, 2014 at 7:19 am

No, you guess wrong. Better safety, including the Nordic “3-point harness”, and airbags, etc, despite the Freakonomics fact about consuming risk, this has the effect of lessening deaths, especially if you live in Sweden. Unless you are in China which still has 400 road deaths…a day. See:

3 Gordon Mohr (@gojomo) May 28, 2014 at 7:54 am

I’ve never seen the three-point-harness & airbags traditionally worn by British child pedestrians. Do you have photos?

4 anon May 28, 2014 at 8:10 am


trees and forest

5 Ray Lopez May 28, 2014 at 8:26 am

It is not clear whether the children were in the cars or not. If I had to guess, the statistics are probably “road deaths” meaning the kids were in the cars. So pedestrian deaths is not necessarily assumed.

6 anon May 28, 2014 at 8:31 am

“Children played in the streets and walked to school. The result was”

7 Ray Lopez @ anon May 28, 2014 at 10:16 am

Children playing in the streets proves nothing. It’s a narrative. ” The result was that 907 children under 15 were killed on the roads in 1951, including 707 pedestrians and 130 cyclists.” – So, by your logic, the 907 children are classified NOT differently from pedestrians but differently from cyclists? Ludicrous. The passage is ambiguous. To me the sentence reads: ‘907 children killed (probably in their parents cars), when the parents attempted to swerve to avoid hitting pedestrians, but unfortunately the swerving did not always work, and along with kids (both in the road and in the car), 707 pedestrians died, as well as, separately, 130 cyclists’.

8 sort_of_knowledgeable May 28, 2014 at 10:41 am

It’s not completely clear but it reads to me 907 deaths out of which 837 were not in the car (pedestrians and cyclists) and 70 were in the car. I assume playing in road or running after a ball that bounced into the road is considered a pedestrian for the statistics. Even without seat belts a passenger in a car hitting a pedestrian is much less likely to die than the pedestrian unless the pedestrian is Clark Kent.

9 Ray Lopez @ sort_of_Know... May 28, 2014 at 11:45 am

Aha! I sort of agree with you. But it means that in a collision between a car and a pedestrian, the car driver died about 70 times out of 907 (8%), meaning the car was out of control, possibly to avoid the pedestrian or simply going too fast. That’s quite a number of fatalities for a car driver hitting a pedestrian, though, if we use a modern analogy of a car hitting a deer, common in Northern VA, it’s understandable if the pedestrian goes through the windshield and interferes with the driver. Not willing to concede defeat, I say “Further research is needed on this pedestrian issue”

10 msgkings May 28, 2014 at 1:11 pm

No, Ray, the other 70 killed were children under 15, not drivers. The 907 deaths discussed were all kids. You haven’t acquitted yourself very well in this thread. Maybe take a few plays off there, Champ.

11 Donald A. Coffin May 28, 2014 at 2:38 pm

“…including 707 pedestrians and 130 cyclists…”

That leaves at most 70 car passengers.

12 careless May 29, 2014 at 1:08 pm

There’s that famous brain of his at work

13 Bill May 28, 2014 at 1:49 pm

Michael might be right.

Suburban living is less population dense and fewer pedestrian activities.
You take the car everywhere, and therefore are not a pedestrian.

14 Willitts May 29, 2014 at 2:58 am

One must not also dismiss better education over time. Or rather, natural selection.

15 Donald A. Coffin May 28, 2014 at 2:42 pm

Let me try this again:

“…907 children under 15 were killed on the roads in 1951, including 707 pedestrians and 130 cyclists …”

The syntax is clear–907 children were killed, of which 707 were pedestrians and 130 were cyclists. 907 children, which included 707 pedestrians and 130 cyclists. To read this any other way is. frankly, absurd. Had the sentence read “XXXX people died on the road, of which 907 were children, and including 707 pedestrians and 130 cyclists,” you might have a point.

16 ChrisA May 29, 2014 at 3:36 am
17 Z May 28, 2014 at 7:19 am

I wonder if he comes in for the John Lott treatment from anti-car groups.

18 T. Shaw May 28, 2014 at 8:38 am

Ban assault automobiles and high-capacity engines!!!

Maybe it’s the “Darwin Award” effect/natural selection. Children that didn’t know to get off the road didn’t procreate.

19 Thor May 28, 2014 at 2:52 pm

I think this is a pretty thoughtless comment. It’s hardly as if stupid kids are waltzing around blithely on roads. Have you seen the number of people texting and driving, or — if anecdotes don’t appeal to you — read about the citations and fines given for this?

20 steve May 28, 2014 at 7:33 am

I’ve always wondered whether there has been any measurable change in animal “roadkill”. Have there been enough generations of squirrels, racoons, etc., that there has been evolutionary progress towards the avoidance of roads or at least on-coming cars?

21 Axa May 28, 2014 at 7:35 am
22 dearieme May 28, 2014 at 8:49 am

You certainly see very few dead raccoons on British roadsides.

23 Mark Thorson May 28, 2014 at 11:10 am
24 XVO May 28, 2014 at 10:30 am

One would think that eventually natural selection would select for animals that are able to look both ways before they cross the road. Not yet apparently…..

25 XVO May 28, 2014 at 10:31 am

Although actually birds seem to do a pretty good job so long as you don’t accelerate with the intention of hitting them. There have been a few daredevils who didn’t make it but mostly they seem to get out of the way before they get hit, even if it’s a little jarring how long they wait to move.

26 JWatts May 28, 2014 at 10:36 am

“even if it’s a little jarring how long they wait to move.”

Agreed it is a little jarring. It’s probably because birds have a higher strength to weight ratio than mammals. From they’re perspective they have time to move, from a mammals perspective they are cutting it way too close.

27 Roy May 28, 2014 at 11:22 am

Many years ago I saw, on several occassions, a particular cat look both ways before crossing a street in Austin. In the past few years I have seen multiple coyotes do this in the Salt Lake area, and I have seen cats do this throughout the PNW. So maybe they are learning. It is probably learned behavior, though it is clear that animals have culture.

28 dan1111 May 29, 2014 at 7:05 am

There are many studies showing that animals can pass on learned behavior to each other, even apart from exposure to the original thing that caused the behavior. I recall studies involving monkeys, and also crows or ravens.

29 chuck martel May 28, 2014 at 11:25 am

The invention of the “deer whistle” , attached to the front of the car has had an enormous positive effect on animal highway deaths, I guess. Sure wish I’d have invented something that nobody but lower forms of life can hear and then been able to sell it to people concerned with wrecking their car when no proof that it actually works exists.

30 Alexei Sadeski May 28, 2014 at 3:49 pm

Someone should market a vehicle mounted “deer flute” which soothingly beckons the deer out into the middle of the road.

31 Chip May 28, 2014 at 7:43 am

Anti-lock brakes, better tires and improved medical treatment.

The latter has also affected gun deaths.

32 Gordon Mohr (@gojomo) May 28, 2014 at 7:47 am

Playing-in-the-street is a heritable trait, fighting strong negative-selection during the last 60 years.

33 XVO May 28, 2014 at 10:35 am

Also a cultural trait, mothers telling their children never to play in the road because they lost a friend or family member vs the lackadaisical and ignorant approach of previous generations. Also could be related to the design of infrastructure as well, where families with children choose to live, parks, road design (keeping the main roads away from residences)

34 mb May 28, 2014 at 7:49 am

As a cycling commuter, I can see this as being true. I would rather ride in heavy traffic than light for 2 main reasons, cars are moving slowly or not at all and people are paying better attention. Moderate traffic is the worst. The only draw back to heavy traffic are the frustrated idiots that try to beat the traffic – thankfully there aren’t as many of them in Boston as there once were.

35 Willitts May 29, 2014 at 3:00 am

Same is true for motorcycles.

36 careless May 29, 2014 at 3:39 pm

” The only draw back to heavy traffic are the frustrated idiots that try to beat the traffic ”

I usually call them “bikers

37 Eric May 28, 2014 at 7:57 am

This book came out about a year ago, BTW – it’s not new. I enjoyed it mostly because it enabled me to ask questions like: “is this hamburger worth living half an hour shorter”.

38 XVO May 28, 2014 at 10:37 am

The answer is always yes.

39 Willitts May 29, 2014 at 3:01 am

Is a cheeseburger worth an hour?

40 Curt F. May 28, 2014 at 12:10 pm

I’d call that question a great example of what I”ll call the fallacy of homoskedasticity. If you were starving, I’m pretty sure that hamburger consumption would result in increased lifespan.

There’s no way of knowing that an effect size for hamburger consumption on lifespan identified in studying a huge population applies to you.

41 Brandon Berg May 28, 2014 at 7:57 am

The decline since 1995 is the interesting one, I think. What could have changed so much in 15 years? Parents keeping their kids on a tighter leash?

42 The Engineer May 28, 2014 at 8:11 am

Better videogames. Seriously.

43 Steve Sailer May 28, 2014 at 7:43 pm

Sounds plausible.

44 anon May 28, 2014 at 8:11 am

Another unintended consequence of the the Roe effect?

45 Salem May 28, 2014 at 8:22 am

Video games and the internet are part of it, but I also think:

1. Newer cars. In the 80s it was normal to see 20-30 year old rustbuckets driving around (presumably with poor safety features/breaks). That stopped in the mid-90s.
2. The Jamie Bulger case. That marked a sea-change in parents’ attitudes to children playing on their own.

46 Z May 28, 2014 at 9:10 am

The average age of cars on the road has gone up, not down. In America, the average age is 11.4 year, the highest in history.

Fewer kids in the streets is probably the place to start. In my lifetime, the change has been dramatic. When I was a boy, unsupervised outdoor activity was the norm. You got home from school and went to play with your buddies until dinner. In the summer and weekends, you spent all day outside. Today, it is rare to see kids playing outside unsupervised. The kids are inside or hovered over by doting parents.

Whether or now that’s good will be determined downstream.

47 KLO May 28, 2014 at 9:35 am

I was not that old in 80s, but I certainly don’t remember lots of cars from the 50s being driven around. Cars made in the 50s were not very durable, and I don’t think many of them made it to 20, much less 30 while still being driven. More likely the cars of which speak only looked 30 years old. The paint and design of cars made in 70s dated them very quickly. The cars of today look contemporary for a much longer period.

48 anon2 May 28, 2014 at 11:54 am

Me too but…there was this one deLorean i could’ve sworn I’d seen before.

49 msgkings May 28, 2014 at 1:15 pm

Thread winner

50 Brenton May 28, 2014 at 3:48 pm

long term effects of lead paint and leaded gasoline bans. Same with the crime decline.

51 Kari May 28, 2014 at 8:26 am

One aspect is the Smeed’s law:'s_law

The number of fatalities per vehicle seems to come down almost “automatically” when there are more vehicles. This cannot be explained by technology since it happened in USA with “old” technology decades before Western Europe and there decades before developing countries.

The most probable explanation is that when the number of fatalities grow (as well as standard of living), society won’t tolerate fatalities any more, and different kind of policy actions are taken to hinder them.

52 Tom T. May 28, 2014 at 7:22 pm

Perhaps there comes a point where cars are now so plentiful that pedestrian attitudes toward them shift? I.e., once the roads are sufficiently crowded with cars, pedestrians look both ways?

53 Michael Savage May 28, 2014 at 8:32 am

Across many countries we observe declining rate of traffic fatalities over time, first identified by Smeed in 1949. A part of that is technology (safer cars, better medicine), but a big part is attitudinal. On one hand, pedestrian behaviour changes – when one car a day passes through your village you act as if there is no traffic. When ten an hour pass by, you look both ways. And if traffic reaches a certain point, you can’t use the street any more – there are communities cut off by busy roads. Another part of it is social expectation. In the Netherlands there was a concerted campaign against ‘child-killing’, which led to institutional changes in favour of cycling. In the US, pedestrians were criminalised (jaywalking laws). Shocking that rate of death by distance driven in the US is more than double that of the UK (8.5 v 3.6 per bn km), despite similar medical and vehicle technology. I also agree that the big change in child death has been caused by children staying inside – that becomes the obvious explanation when you look at how much more slowly overall deaths have fallen. Stats here:

54 dearieme May 28, 2014 at 8:56 am

“Shocking that rate of death by distance driven in the US is more than double that of the UK (8.5 v 3.6 per bn km)”: surely that’s just in line with the other evidence that Americans like killing each other more?

55 Mike May 28, 2014 at 9:38 am

Or that Americans often travel at higher rates of speed because stuff is farther apart.

56 Michael Savage May 28, 2014 at 10:10 am

I’m not convinced that it’s related to speed. The US actually has low speed limits by international standards, and road deaths tend to be concentrated in urban areas where there are more pedestrians. I can only cite anecdote, and it would be interesting to look at breakdown (e.g. speed at time of accident, type of road).

There is something to be said for dearieme’s comment, which I would express more neutrally as saying that Americans may have a higher risk tolerance. That’s not always a bad thing, although the distribution of costs and benefits is inequitable; drivers consume performance benefits, passing costs to pedestrians and would-be pedestrians.

57 JWatts May 28, 2014 at 10:39 am

Canada has a similar fatality rate per distance as the US’s rate.

58 Finch May 28, 2014 at 10:42 am

Pedestrian deaths are something like 1/8th of motor vehicle accident deaths, so they are not a significant factor in the relative national death rates.

59 Finch May 28, 2014 at 10:05 am

If you drive around London averaging 10 kph, it’s unlikely that you will die in a traffic accident even in typical European vehicles.

You’re usually quite good fun, dearieme, so I’ll write this one off as a joke that didn’t work so well.

60 dearieme May 28, 2014 at 10:50 am

Not entirely a joke; let me rephrase that it’s in line with evidence that Americans are unusually inclined to be careless of the interests of others. Thus: high murder rate, high divorce rate, high deaths in traffic rate, high proportion of lawyers in the population. Maybe even high proportion of undergraduates studying economics?

Perhaps they even dislike themselves, but international comparisons of suicide rates aren’t likely to be helpful – I’d guess they are likelier to measure the attitudes of the coroners’ courts, or the equivalents, than the underlying phenomenon. How about addiction to dangerous drugs rates? Alcoholism? Same problems of international comparisons I suppose.

Is there any interesting evidence about, say, industrial accident rates? Would it be valid to attempt to compare those internationally?

I wonder which of those proposed measures correlate with each other, at least if you restrict the study to advanced countries.

61 Finch May 28, 2014 at 11:05 am

Thank you, this is much better phrasing.

But I still think in the answer is obviously speed. Smaller cars are massively more dangerous than large cars, all else equal. Europeans drive smaller cars but don’t generally die at much higher rates. Therefore all else is not equal. The obvious difference is speed. The British are tootling around London in a Peugeot while Americans are driving from suburb to suburb on an interstate.

I don’t see a huge difference in alcoholism (and if anything it runs in the wrong direction). And I don’t think there’s a huge difference in seat-belt compliance. Savage’s pedestrian hypothesis is a red herring.

62 dearieme May 28, 2014 at 11:31 am

Speed doesn’t explain Belgium.

63 Finch May 28, 2014 at 11:48 am

Small sample size and somebody has to be the outlier? Only fully-lit motorway in the world? Lots of German and French drivers passing through and unfamiliar?

A little Googling finds this article that says it’s speed (in particular poor enforcement) and unusual right-of-way rules:

64 careless May 29, 2014 at 8:27 pm

And yet, my murder, divorce, etc rates are so low!

65 JWatts May 28, 2014 at 9:56 am

“Shocking that rate of death by distance driven in the US is more than double that of the UK (8.5 v 3.6 per bn km), despite similar medical and vehicle technology.”

Deaths per billion/km
Greece 17.4
Belgium 10.4
Spain 8.5

Greece’s rate must be terribly shocking to you then.

66 Michael Savage May 28, 2014 at 10:01 am

Greece is a much poorer country, so I’m not so shocked. Belgium is more shocking to me, and I don’t know why theirs is so high. The US puts far greater burden on pedestrians (e.g. criminalisation of jaywalking), but that strategy has failed.

67 Dan Hanson May 28, 2014 at 2:00 pm

I wonder if that isn’t part of the problem – In America drivers may assume that there will be no pedestrians between intersections and so they aren’t as vigilant?

Along the same lines, If the incidence of pedestrians is lower in the U.S., it could paradoxically result in higher pedestrian accidents because a pedestrian on the road is a more unexpected event, and therefore a driver is less likely to notice and stop.

68 Tom T. May 28, 2014 at 7:24 pm

Belgian children fleeing pedophiles by running into the street?

69 Brian Donohue May 28, 2014 at 3:27 pm

Heh. Good ole dearieme never misses a chance to troll Americans. He sure spends a lot of time on American websites though.

70 cfh May 28, 2014 at 8:34 am

This is an aspect of The Great Stagnation. The low-hanging fruit has already been run over.

71 HL May 28, 2014 at 9:46 am


72 dearieme May 28, 2014 at 8:54 am

“free of restrictions such as road markings”: in 1951, really? I’ll grant that I didn’t drive in ’51, but I’d be astonished if there were not at least central stripes, usually equipped with cat’s eyes.

73 Corvus May 28, 2014 at 10:28 am

No cat’s eyes – they weren’t invented yet. Many many more miles of roads with minimal or no marking. Lots less markings (things like fog lines, turn lanes). Lots less highway infrastructure “furniture” (curbs, controlled intersections, signage, etc.). Significantly less traffic planning involved in highway construction. No superhighways. Etc.

In short, very significantly different environment.

74 Spencer May 28, 2014 at 11:32 am

The big difference may be the interstate highway system that did not exist in 1951.

It is safer than unlimited access roads and it removes a lot of long distance travelers, especially trucks, off the local roads.

75 dearieme May 28, 2014 at 10:53 am

“No cat’s eyes – they weren’t invented yet.” Quite wrong I’m afraid. Maybe they hadn’t reached your neck of the woods?'s_eye_(road)

76 Tom May 28, 2014 at 1:17 pm

Is it possible that Europeans who have cars are simply better drivers and less likely to hit pedestrians? I had a French friend who was shocked at the behavior of American drivers and at the pathetic requirements of an American driver’s test. He considered passing the French test a badge of honor, as it was administered by people who were looking for an excuse to fail him.

In France, the mentality is that if you’re not good at driving, you should take mass transit. In most of the U.S., the mentality is that if you deprive someone of their car, you have effectively banished them from society.

77 JWatts May 28, 2014 at 2:04 pm

“Is it possible that Europeans who have cars are simply better drivers and less likely to hit pedestrians?”

At the very least European drivers tend to have a higher average amount of miles driven and European countries require much more exhaustive training. The European approach is more elitist, but it wouldn’t surprise me that the result is a higher average quality of driver. At the very least the much higher costs of driving in Europe would tend to decrease the amount of marginal drivers.

On the other hand, you have Belgium, Spain and Greece.

78 XVO May 28, 2014 at 4:12 pm

Are Russians Europeans? Because you might need to look up some Russian traffic videos and come up with a new hypothesis.

79 Steve Sailer May 28, 2014 at 7:48 pm

Chechen dashboard camera videos are the best.

80 dearieme May 29, 2014 at 5:44 am

Asia begins at the Landstrasse.

81 Steve Sailer May 28, 2014 at 7:48 pm

American driver’s tests have gotten significantly harder to pass over the last generation, at least generalizing from my high school classmates’ pass rates in the 1970s vs. my sons’ classmates’ pass rates at the same DMV facilities in this century.

82 careless May 29, 2014 at 8:33 pm

Not much change between mine nearly twenty years ago and my wife’s two weeks ago. Pretty easy.

83 Foobarista May 29, 2014 at 1:02 am

Better cars and brakes, better signage, lighting, and signals, better-designed roads. Also, more pedestrians on sidewalks rather than on shoulders or the road itself, as you still see outside cities all over the world.

84 ChrisA May 29, 2014 at 3:47 am

There is an explicit claim that reduced road deaths resulting from more safety regulations. But all we have is correlation, more regulation and lower road deaths are occurring at the same time, but it doesn’t mean that one causes the other. Seems to me that we need more proof that regulations are the cause and not other things, including the one mentioned here; less pedestrians and more careful pedestrians. It is entirely possible, indeed likely, that people will modify their behavior over the years to reduce their perceived risk. It is also possible that there is a secular decline in risk tolerance in more advanced countries. If we know the cause maybe we can get rid of some of the regulations, regulations should be tested against cost benefits. So how would you separate the regulation factors? Looking at countries that have not imposed such rigorous regulations and see if their rates have come down? But that would suffer from two issues which would confuse the results, 1) countries with higher tolerance of risk may also be the ones not changing their regulations 2) spill over from more regulated countries into less well regulated countries of some of the technologies.

Is is possible to do an analysis using US state data, after all some states have more road safety regulations than other similar states?

Comments on this entry are closed.

Previous post:

Next post: