Monday, January 11, 2010

Sun! Sand! Definitely no sea!

And to cheer you up, something from sunnier climes, Arizona, to be precise. It seems the state legislature failed to pass a statute that presumes delivery of anything that the State says it's posted. So unlike this great country, unless a speeding fine is served on you personally, it will expire. Seems quite sensible, as it means the police will concentrate on the more serious offences. The law of unintended consequences has swung into action, and the state has ninety MILLION dollars of fines outstanding. Outstanding!

Lest we all get teary-eyed at the prospect of getting out of second gear and above the six miles an hour we've all been doing recently, consider this: more than 40,000 people a year die on the road in America.

My strictly amateur analysis is as follows: deaths per distance unit travelled, the Yanks are way ahead because of the huge distances they drive. For example, in 2004, it was 1.46 deaths per 100 million vehicle miles travelled. In 2007, in the UK, this was 48 per 100 million vehicle kilometres. So per mile, it's above 60. Three years apart, yes, but that certainly doesn't account for the difference.
I can't find a decent comparison per journey or per driver.

Any statisticians out there want to volunteer an analysis of whether this is meaningful, or simply a big number I've added to a blog post to make a point?


  1. I'd have thought the relevant comparison was per capita.

    UK - 60m population, 2,500 deaths. One per 24,000 population.

    US - 300m population, 40,000 deaths. One per 7,500 population.

    Where it gets fun is if you assume that's consitent over an average 70 year life.

    US - 1 on 107 people will die on the road.
    UK - 1 in 343 people will die on the road.

    Round numbers, of course, but close 'nuff for these purposes.

  2. To Anonymous at 7.05am on 18 January 2010: The relevant comparison depends on what you're trying to analyse. If you want to show which population is more likely to die from traffic accidents, then traffic fatalities per 100,000 persons may be most relevant. If you want to show which population is more likely to be fatally reckless when they drive, then fatalities per 100,000,000 miles may be more relevant. (Alternatively, you can use fatalities per 100,000,000 hours spent behind the wheel. My understanding is that people tend to cover more distance with less risk per mile when on motorways.)

    Even if US citizens aren't particularly dangerous on the roads, I agree with Anonymous Prosecutor's assessment that they tend to drive far too much. I am yet to convince very many people from the US that a high tax on petrol would be better for national health and safety even before you consider the environmental benefits.

  3. The USA has an average 83 people for every square mile, the UK has 657. So, for every average mile you drive in the UK, you encounter eight times as many people to have accidents with. That's why a 'per mile' analysis isn't really comparable.

    Many more people in America (say in Arizona) might well pull out of the driveway, go in a perfectly straight line for five miles without seeing more than a couple of other vehicles, and reach the destination. When did that last happen to you?

  4. The last comment is somewhat misleading, while the UK has a much higher population density overall, there is a decidedly large amount of sod-all in the middle of the US that skews the numbers (Wyoming anyone?) and the major, denser population centres of the West and East Coast still have similar casualty rates.
    There are a number of reasons relating to this, ranging from things relating to driving style and 'cultural' values to more simple things such as the UK having a far better record of funding road safety improvements with a beneficial Cost-Benefit Analysis than the US. This might in part be due to the mixed levels of government in the US and a sort of deferall of responsibility.