More unintended consequences in the safety theatre of the absurd


I guess we’re all aware of the wave of texting while driving legislation, as well as recent moves in a number of jurisdictions to make the penalties more draconian. And it seems like a reasonable supposition that such legislation would reduce the incidence of accidents doesn’t it?

Unfortunately, it appears that the laws have the exact opposite effect. A comparative study conducted by the Highway Loss Data Institute in the US compared accident rates between states that did and did not implement texting laws, and found that accident rates increased in the states which implemented the legislation.

…bans haven’t reduced crashes at all. In a perverse twist, crashes increased in 3 of the 4 states we studied after bans were enacted. It’s an indication that texting bans might even increase the risk of texting for drivers who continue to do so despite the laws.

Adrian Lund, president HLDI

What’s interesting about this example is that here we’re trying to control a human/technological system in a simplistic fashion that overlooks the adaptive nature of said system, resulting in completely contrary results to those intended.

An NRMA survey last month found 91 per cent of drivers admitted to texting or reading emails, half updated Facebook and played games, and 76 per cent took photos behind the wheel.

Sydney Morning Herald, 13 Oct 2013

Unfortunately the legislators in NSW, my home patch, still figure that if the original law isn’t working, then more of the same ought to, including further extensions to police powers of search and seizure. I look forward to the increased accident rate. When we legislate for safety it seems the law of unintended consequences is one that is not just inevitably passed but also inevitably compounded.

3 responses to More unintended consequences in the safety theatre of the absurd


    Matthew: of course, the loop is closed by means of the enforcement element (police) who are to observe the offending behavior and end it.

    One might fuss about delays and phasing and sample rate in the feedback, but those are merely details in the architecture

    John Goodpasture (Musings)


      Matthew Squair 27/11/2013 at 10:39 am

      Ideally yes that’s the role of policing, but in this case not so much.

      What appears to be happening is that users are putting the phones in their lap rather than on the wheel which means the police can’t see them. Which is a persuasive argument for both the subsequent increase in accidents observed, and also the failure of that particular control. Put humans in the loop (so to speak) and you have an adaptive system in this case it adapts in the direction of negating the intended effect, and increasing the risk.

      In a broader sense public policy should always be scrutinised for whether it works or not, and that should be on the basis of empirical evidence, something that’s missing from this and other debates about road safety. Now there’s a larger feedback loop that seems to be missing from policy debates.



    If you want to change a behavior, make the consequences of the behavior non-transferable. For example: pass legislation to allow insurance companies to void the coverage of any driver found to have been in an accident while diddling with their devices. Liability tends to focus the attention.