November 24, 2013

Blood alcohol change report

The detailed Ministry of Transport paper on changing the legal blood alcohol limit is now available.  There’s a story in Stuff, which is, if anything, unduly critical (an interesting change). It doesn’t mention the cost-benefit analysis, and implies a fines grab

Transport officials calculate nearly 20,000 people will be caught by the lower drink-driving limit – earning the Government $5 million extra in fines. 

which is a bit misleading since the report (paragraph 93) actually estimates a net increase in costs to the justice system of about $2 million in the first year and about half a million in subsequent years, ie, the fines don’t cover the costs of enforcing the change.

Basically, whether the change is a benefit or not depends on how much inconvenience and risk is caused to the average driver, the only major component that isn’t taken into account in the calculations.  If this is worth only 50c/month or so, the policy makes sense. If it’s worth a few dollars a month, not.

On the other hand, the policy is popular, and since most people should have a reasonable appreciation for how the change will affect them personally, that’s a more persuasive argument than it would ordinarily be.

avatar

Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »

Comments

  • avatar
    Martin Connelly

    I wonder if there is also not another major component that does not get taken into account – namely, how much do people value being on roads that are now safer because they are sharing the roads with less inebriated drivers. (Assuming the new limits are enforced). I think many parents watching their children drive off for the first time will be a lot happier knowing their is less chance they are will meet a tipsy driver.

    10 years ago

    • avatar
      Thomas Lumley

      The prediction is for this to reduce crashes by less than 2%, so the only way parents could rationally be happier is if they think a crash with a tipsy driver is much, much worse than the same crash with a sober driver. That’s not impossible — people can have surprising preferences — but I’d guess it’s more likely that a parent feeling a lot happier didn’t understand the actual risks.

      10 years ago

      • avatar
        megan pledger

        You’re assuming that teenagers and drunk drivers are driving uniformly over time whereas they are most likely to be on the roads together, at the most dangerous time of day for drink driving e.g. 3 am in the morning coming home from the pub/nightclub/friend’s house.

        And the amount of time teenagers put themselves at risk at this dangerous time is also way greater than for the typical driver. Teenagers are out multiple times per week, all through the year, socialising late into the night.

        There may be a 4% reduction in crashes but it’ll be interesting to see if this is uniform across age groups.

        10 years ago

      • avatar
        Martin Connelly

        That’s right – the thing that makes you happier is the perception not the actual knowledge. In the case of car crashes most people accept their is some risk, but accept that risk because of the benefits in having a car. But these day, people think an accident with an intoxicated driver is an unnecessary risk. We feel happier when we think something could not be helped that when we think it could have been avoided

        10 years ago

  • avatar

    Will have to OIA their full cost benefit analysis. They’re saying consumer surplus drops by $3.5m over ten years so $350k per year.

    Each driver in the .05 to .08 range has to have about one fewer drink. About 1.5% of drivers are in that range.

    Does it seem plausible that net enjoyment of the last drink by each of those 1.5% of drivers is small enough to get to $350k total? There are more than a million car trips per day. Say then 10k trips affected per day. So 3m or so affected trips per year lower bound? So they’re claiming about ten cents CS loss per affected trip?

    Does it seem at all plausible that somebody paying north of $5 at the bar for his last drink only values it at $5.10?

    10 years ago

    • avatar
      Martin Connelly

      The way I read the advice was that they did not think that bar owners would be much affected – meaning that they think many people will continue to drink pretty much as they do at present but find other ways to get home – the sober driver or the pub-provided min-van maybe

      10 years ago