09 July 2010
What’s wrong with this graph?

This came to me in an email from the Tax Payers’ Alliance referencing this report (the graphic is on page 11).

I do hate it when ideological friends make such elementary errors.

Mind you, it does beg the question: what would I do?  Of course, in Patrick Crozier’s nirvana all roads would be privately owned and whatever rules there were would be up to the owners of those roads.  But what would those rules be likely to be?  I have this awful feeling that they wouldn’t be all that different.  If I owned a road I would like it to be fast - happier customers - but what I would really like it to be is safe - more crashes, less use, lower revenue.

PermalinkFeedback (7)Roads


  1. Base rates. No account is given for any likely increase in overall traffic over the time period.

    I presume (perhaps at some risk of being wrong?) that the data came only from roads which were, after 1990 or so, monitored by speed cameras? Surely they haven’t lumped together casualties from all sorts of different roads - those with and those without speed cameras?

    And who was responsible for making the post-1990 projection? Surely if it had been advocates of speed cameras, then I would imagine they’d have expected the introduction of the cameras to cause a much steeper decline of the casualty rate.

    Posted by mike on 12 July 2010 at 01:02pm

  2. This is a style of public rhetoric that drives me nearly insane with rage, the one that asks a question and insists on you giving the wrong answer before they give you the right one.

    I actually once walked right out of a dinner party because of this.  Mein fucking Host did this trick on me, asking some utterly vague question with no definite answer, and he insisted that I had to supply an answer for him to contradict, at which point he could then lecture me from a position of lofty superiority.  I said: you are trying to make a point.  Please make it.  He refused and continued to insist that I answered his damn question.  I said if he did NOT make his point, I would leave at once.  He did not, and I did.  Wanker.

    So, to sum up, maybe you’ll tell us all what is wrong with this graph and maybe you won’t. Until you reveal the answer, I don’t care, and by the time you do, it will probably be, for me, too late.

    You are trying to make a point.  Please make it.

    Sorry, you just pushed one of my buttons.

    Posted by Brian Micklethwait on 12 July 2010 at 10:14pm

  3. Wow.  Wasn’t expecting that.

    My point is that the projection is a straight line and one that is headed towards zero and presumably beyond.  In other words, according to it at some point in the near future the death toll on the roads will no negative.  This is clearly absurd.

    Posted by Patrick Crozier on 14 July 2010 at 01:46am

  4. I was going to suggest that the straight line seemed wrong. It didn’t immediately occur to me about it crossing the axis but would have sooner or later. The green line looks like the sort of curve you would expect from the law of diminishing returns; speed cameras I would imagine have had no significant impact. Cars, meanwhile, get steadily safer and more reliable but there is only so much you can do. Expect a step-change in the graph when some new technology, like self-driving cars, is introduced.

    I think that in your nirvana, people would choose faster, more dangerous roads than they get to drive on on the real world. Road “safety” laws only ratchet in one direction; the most vocal road lobbyists are more risk averse than the general population. Most people in a free market would choose more convenient but riskier roads; ultra-safe roads would be a niche.

    Posted by Rob Fisher on 14 July 2010 at 03:12am

  5. Is there any evidence of an inverse correlation between road speed and road safety? I’m asking a genuine question, not making a point by asking a rhetorical one.

    German autobahns are the world’s fastest roads - some of them are still sufficiently uncongested that “no speed limit” isn’t just a formality. Are they among the world’s not dangerous? They don’t feel like it - although I freely admit that my personal experience of never having died on one is the anecdotal of evidence.

    Posted by Alan Little on 16 July 2010 at 12:05am

  6. Oops. For “not dangerous” read “most dangerous”

    Posted by Alan Little on 16 July 2010 at 12:40pm

  7. This is a style of public rhetoric that drives me nearly insane with rage, the one that asks a question and insists on you giving the wrong answer before they give you the right one.

    This is indeed a particularly annoyingly rhetorical trick. It is at its most annoying when the person who is aiming it at you has a particular answer in mind, and when you refuse to give it to him and instead give him another more sophisticated answer than includes subtleties he is unaware of he decides to “correct” you anyway. It is even more annoying when this happens in circumstances when you are being paid to listen to him and are unable to walk out. (I am bitter. This moron was my boss once, long ago).

    On the other hand, I think that Patrick’s point (that the relationship between time and casualty rate could not possibly be linear) is sufficiently obvious that he does not actually deserve much criticism. This would be my main problem with the Taxpayer’s Alliance, too. Although they are good at public relations and are generally on our side, their actual research is generally extremely sloppy.

    Posted by Michael Jennings on 18 July 2010 at 04:13am

Commenting is not available in this channel entry.