The precautionary principle

January 11th, 2011 at 2:00 pm by David Farrar

blogs:

RadioNZ reports that Minister of Transport, Steven Joyce, says “more evidence is needed before the Government will consider lowering the general drink-driving limit”.  Others such have David Farrar have echoed the claim that specific evidence is needed that lowering the drink drive limit will have an instrumental effect on the number of road deaths and accidents.
Baloney.

Yes, changes to laws should be justified.  But, no, the justification need not be specific evidence.

Both Joyce and Farrar are ignoring the precautionary principle.  In general terms, this principle says that, in relation to risky activities where there is scientific or empirical doubt about the nature and extent of the risk, policy- and law-makers should favour the course of action which avoids the risk.  That is, in the face of uncertainty, the burden shifts to those undertaking the risky activity to demonstrate it is not harmful. 

I call baloney on Dean’s baloney.

First of all, transport policy is not about eliminating risk, it is about balancing it – a point he admits later. Without that balance, then one would easily conclude that driving at faster than 30 km/hr is dangerous (and it is) and should be banned.

So, we know that is a risky activity.  That’s why we prohibit driving above the current blood-alcohol level (80mg).  If there’s doubt about whether driving while above a reduced limit of 50mg (which I not sure there is doubt about when looking at international practice), then the precautionary principle would favour lowering the limit anyway and collecting data to demonstrate it is not risky or has no instrumental influence on road accidents and deaths – not the other way around!

Dean ignores one crucial point – if you lower the limit now, you will never ever be able to collect data on the prevalance of legally driving at a BAC of 0.05 to 0.08, and its associated risk.

Ultimately, these things involve a cost-benefit calculus. There are seldom king-hits in law- and policy-making. A balance must be drawn.

Joyce and Farrar et al are, however, underplaying the benefit (risks avoided) of lowering the limit by proclaiming uncertainty about their nature and extent.  In this context, though, we’re entitled to assume there is a risk associated with driving with a blood-alcohol of over 50mg, unless evidence shows otherwise.  This means, for the purpose of the cost-benefit calculus, we can assume lowering the limit is beneficial   

Yes there is a risk driving with a BAC between 0.05 and 0.05.  Lowering the limit would lower that risk. But there is also a risk in allowing people to drive at faster than 30 km/hr. There is a risk at allowing people to drive with passengers in their car, as they can be distractions.

Banning passengers and lowering the speed limit to 30 km/hr would also be “beneficial” if you only are focused on the road toll.

Of course, on the other side of the ledger, we can also say there is negligible cost associated with lowering the blood-alcohol limit. 

That is nonsense. There is significant cost associated with such a lowering. It could mean that many more people will be unable to legally drive, and in rural areas especially could even lead to the closure of pubs, where public transport is not a viable option. It may also impose extra costs on people who then take taxis home. Now you may think that is a good thing, but it is also a cost. And again I’d like to know the costs before a decision is made.

And here is the problem – we currently do not measure at all what the “cost” would be of lowering the limit, because we do not know how many people currently drive with a BAC between 0.05 and 0.08, and hence we can’t calculate what the cost will be of a law change.

I want the Government to be able to answer a few simple questions before they make a decision:

  1. What is the current prevalance of drivers with BAC between 0.05 and 0.08
  2. How many drivers with a BAC between 0.05 and 0.08 are involved in accidents
  3. What then are the accident rates for drivers with a BAC below 0.05, betwene 0.05 and 0.08 and above 0.08
  4. How many accidents and fatalities are caused by adult drivers with a BAC betwene 0.05 and 0.05

Once you have that data, then you can make a sensible decision about whether the benefits of reducing the BAC to 0.05 is worth the cost.

If you go down Dean’s path, then we will never ever be able to gain that knowledge. His precautionary principle plea, is in fact a cloak for making a decision that one could never later challenge.

Tags: ,

32 Responses to “The precautionary principle”

  1. Manolo (13,339 comments) says:

    And who on earth is Dean Knight?

    According to his blog he is: Senior Lecturer, Faculty of Law Associate Director, NZ Centre for Public Law Victoria University of Wellington. He forgot to add Do-gooder and Wowser Extraordinaire.

    Another pompous and arrogant academic trying to tell us how to live our lives. F..k off!

    [DPF: Attack the message, not the messenger]

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  2. RightNow (6,647 comments) says:

    The precautionary principle should also be applied to cyclists on the road. Some die, ban all.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  3. Sonny Blount (1,845 comments) says:

    There are few people I detest more than those calling for the BAC to be lowered from its current levels.

    He’s a contemptable toad.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  4. Sonny Blount (1,845 comments) says:

    Its people driving home oln Sunday morning, and going to the video store in the evening after a beer with lunch who will be done by these changes.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  5. MT_Tinman (2,985 comments) says:

    Not very often I agree with you completely Manolo but this is one.

    Typical city-based school teacher.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  6. ben (2,396 comments) says:

    Dean Knight wrote:

    Of course, on the other side of the ledger, we can also say there is negligible cost associated with lowering the blood-alcohol limit.

    What. A. Howler. How could he possibly think that? Thousands more people will be convicted of drink driving. Millions of meals will be consumed without the pleasure of even a single glass of wine out of precaution to avoid any risk of breaching the lower limit. Millions more who do have a glass will worry if they’re over the lower limit or not. Careers will be accidentally ruined by the lower limit. Now these costs may be justified by the benefits of a lower limit, or they may not, but it is a howler to say the costs are negligible.

    I guess this is the price of being a public intellectual – sooner or later you’re going to say something like “negligible cost” and be remembered for it.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  7. Adolf Fiinkensein (2,790 comments) says:

    Another brainless academic.

    David, for God’s sake will you proof read your bloody figures?

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  8. ben (2,396 comments) says:

    I think Dean misunderstands the precautionary principle as well. I’m quite sure Joyce and Farrar are not invoking the principle in their defense of a higher blood alcohol level. There is actually little to no uncertainty on the effect of drinking on driving: we know drinking is bad for driving and the relationship between blood alcohol and driving performance is well understood. The uncertainty is whether the benefits of a lower limit outweigh the costs – and the precautionary principle doesn’t tell us which way to go on this. The precautionary principle does not say: prefer saving human lives to human enjoyment. The principle says: avoid the really big though very unlikely losses (e.g. runaway global warming). A very large, very unlikely negative outcome isn’t in play here.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  9. GPT1 (2,087 comments) says:

    I will say it again, from what i see in court 1 the majority of the problem is recidivist drink drivers often as not without licences. How lowering the BAC is going to deal with this I do not know.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  10. georgebolwing (602 comments) says:

    ben is dead right.

    This is an issue of costs v. benefits. The benefits of a lower BAC are fewer accidents, the costs are (a) direct costs of enforcement and (b) the loss in enjoyment/behavioural change induced by the law change.

    The benefits can be quantified to reasonable degrees of uncertainity. The costs, however, are much more difficult to quantify. But those promoting the law change should carry the burden of proof.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  11. ciaron (1,314 comments) says:

    I once read somewhere (insurance company?) that 2/3 of accidents are not alcohol related…

    Really aren’t we better off investing in expanding the driver license requirements and improving the general aptitude of all drivers?

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  12. eszett (2,332 comments) says:

    Actually, there are some far simpler questions that should be answered first:

    - What evidence and rational was the recommendation by the MoT to lower the BAC to 0.05 based on?
    - Why is that evidence not sufficient so that further studies are necessary?

    All we hear about is the Joyce went against the advise from his own officials, but we don’t really have any idea what that advise was based on and why it is not sufficient.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  13. Pete George (22,781 comments) says:

    It is especially important to go far further the some “precautionary principle” if it means tightened restrictions on drivers – it is important that suitable studies are done first. If you wanted top keep invoking the precautionary principle then why stop at .05, why not go to 0.00? And why stop at 100kph, why not go to 0? Just in case.

    Or – Dean Knight could exercise his own precautionary principle instead and choose not to travel on the roads, just in case.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  14. BeaB (2,057 comments) says:

    NZ academics sadly have always had the habit of mixing polemic into their arguments. Remember all those biased lectures we had to sit through! One day we might hear from one of our academics some unalloyed fact and logic with an intelligent conclusion at the end. Fat chance.

    And why do people have such a credulous faith in these magic numbers? Most drunk drivers who cause crashes are thoroughly pissed. They haven’t just had a sip of sherry.

    Am I the only one who smells the whiff of wowserism in this current battle against the booze?

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  15. ben (2,396 comments) says:

    Am I the only one who smells the whiff of wowserism in this current battle against the booze?

    Ah, no.

    The problem with the drink driving debate is that it serves as a useful signalling device among authoritarians: what Mr Knight is doing is publicly announcing to those authoritarians already in power that he is one of them, and that he can be safely brought into the fold and given the money and power and awards of higher office. Dean’s argument has bad logic but is undoubtedly helpful for his career prospects. Whether he actually believes anything he is saying I don’t know.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  16. labrator (1,745 comments) says:

    Wikipedia states

    The precautionary principle states that if an action or policy has a suspected risk of causing harm to the public…

    I agree that Dean Knight seems to be misusing the pre-cautionary principle.

    There is a non-linear relationship between drink-driving and the harm caused. As a society we have decided a certain level that we think weighs the benefits of drink-driving against the costs. The burden then is on the people wanting to change this agreed level that it needs to change. This has nothing to do with precaution and it only would if we’d just discovered alcohol. As such, the proponents of the change have the burden of proving that the benefits of reducing the BAC level to 0.05 outweigh the costs. So far as I have seen, they have not done this to date.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  17. Crampton (214 comments) says:

    As a matter of straight logic, I’d thought the precautionary principle said that we have to stick with the status quo, no matter what the status quo is, absent somebody proving that a change is for the better. .08 is the status quo.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  18. RightNow (6,647 comments) says:

    IIRC Dean Knight is a long time Salvation Army member. Their ‘Articles of War’ (which they might want to rename given the current controversy around political rhetoric) include complete abstinence from alcohol.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  19. Simon Arnold (98 comments) says:

    I’m with Dean on this.

    Apply the precautionary principle to NZ becoming a totalitarian state say I. Every time a new law’s proposed.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  20. gravedodger (1,510 comments) says:

    We must employ all the tools at hand to lower the carnage and my god here is the complete TOOL.
    This member of academia’s idea of Happy Hour is probably a shandy in the university club bar then wobbling off to his nearby apartment for tea and tiny cakes on his bike. Me, my totally antisocial behavior includes meeting up with an assortment of chaps and chapesses of all walks of life in a pub, catch up on the world at large over a jug sometimes with a handle chaser on Friday nights, enjoying the food from the host of “The Tree Fellers Arms” and wobbling off home along almost deserted roads in my trusty old vw beetle.
    I agree the problem drivers are well over 0.08, often recidivists, speeding (to get home before anything untoward happens) and they do it totally pissed and often.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  21. Kimble (4,377 comments) says:

    When it comes to feeling good about doing good, there is nothing quite like solving a problem that people just assume exists.

    When people assume a problem exists you can roll out a quick and easy solution with the least amount of fuss. Of course, that means that you have to belittle or ignore people who say there is no problem. Those guys are just standing in the way of the easy solution.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  22. ben (2,396 comments) says:

    Apply the precautionary principle to NZ becoming a totalitarian state say I. Every time a new law’s proposed.

    Indeed. Same goes for every new application of socialist lunacy in the global warming/climate change/climate disruption nonsense. No amount of bad weather can kill, maim and starve like a government without constraint.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  23. DT (104 comments) says:

    David. You are correct that in deciding between uncertain outcomes (ie where there is risk) policy makers should balance the expected value of the competing options (ie probabilities of uncertain outcomes multiplied by payoffs/costs if they occur). But that is quite different from the precautionary principle, which in this case relates to uncertainty of the probabilities of the different outcomes. Thats a major slip-up.

    On the other hand, on the question of lowering the breathe alcohol limited, I’m not 100% convinced myself.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  24. DT (104 comments) says:

    And not wanting to rattle on (but I will anyway), where the uncertain risk is potentially large enough (and potentially large numbers of human lives seem to me to fit that), then it can be shown mathematically that a cost benefit analysis will support the precautionary principle on utility maximising grounds *even if this precludes proof being found*.

    Of course, I don’t believe that lowering the breath alcohol limit would do that. Policy have data on people pulled over and their breath alcohol limits, similarly from major crashes. The question of the extent to which breath alcohol readings between 0.05 and 0.08 would cause fatal or major crashes would be highly difficult to determine from empirical data anyway (ie, what if those same crashes would have occurred under 0.06 as under 0.2? What if under a limit of 0.05 the same crash would have occurred?) It seems to me that the best way to determine the link between breath alcohol and impairment, and impairment with likelihood of crashing, would be through laboratory trials. Which could be still be done under a limit of 0.08

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  25. ben (2,396 comments) says:

    DT, risk is a red herring here. Precautionary principle becomes relevant when a very small risk of a strong negative outcome exists e.g. the end of the world from some unexpectedly bad climate outcome. The principle is only helpful when that risk exists on one side of the argument and not the other, as it does in the climate debate (although that is debatable – the 20th century is a lesson in the danger of government run amok, and there has to be a small but non-zero chance that Kyoto-style policies create the opportunity for governments to do just that once again).

    In the drink driving case, depending on how you define negative outcomes, the negative outcome to be avoided either exists on both sides or neither side of the argument. Dean badly misunderstands the precautionary principle when he uses it to argue for taking the option with the lowest risk – that is not what the precautionary principle means. As others have pointed out, this mis-use of the principle reduces it to being trivial – simply pick the policy that produces least risk without any regard to trade offs. Pointed at anything else – vehicle speed, sport, heck drinking coffee and eating anything except mung beans and baked chicken (my choice for least-risky meal on the planet) – the absurdities of the rule quickly become obvious. Dean is surely smart enough to understand this, which makes me suspect, as I noted above, that something else is going on.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  26. PaulL (5,872 comments) says:

    DT: it should be possible to compare the rate of accidents for those between 0.05 and 0.08 with the rate of accidents for those between say 0.03 and 0.05. You’d have to correct for time of day (more accidents occur at night, and more people drink alcohol at night, so you need to be careful to not confuse causation and correlation), but it is certainly possible.

    Last time this discussion came up on kiwiblog I think the raw numbers suggested that only 1 or 2 people died each year in an accident caused by someone with blood alcohol between 0.05 and 0.08. And even if those people had been stone cold sober, at least some of those accidents would have happened anyway (if the earlier commenter who suggested only one third of road accidents have someone intoxicated associated with them). My guess is that it would save precisely zero lives, but I’m probably biased since I am totally against this change. But at most, it would save 1-2 lives, probably 1, per year.

    As a comparison point, I note the following opportunities to reduce the road toll:
    – at the time we required cyclists to wear helmets, it was well established the wearing helmets in cars significantly reduces the incidence of severe head injuries. Of course, Joe Average voter doesn’t ride a bike or a motor bike, and does drive car, so we’ve never made car helmets mandatory. I think the analysis suggested > 10 lives per annum would be saved
    – when we decided that young children shoudn’t ride in the front seat because it was more dangerous, we were notably silent on the fact that your spouse/significant other would also be safer riding in the back seat. We could pass a law that made it illegal to use the front passenger seat unless all the other seats in the car are occupied. Of course, your wife might not be happy, and she does get a vote, but if we care about road deaths, we’d save quite a few lives this way.

    In short, the wowsers who want to reduce the blood alcohol limit are suggesting it because it doesn’t impact them. And they don’t care that it does impact other people, they just claim that it has “negligible costs.” Well, how about they put their money where their mouth is? Forcing your spouse to ride in the back seat would have no costs at all (unless you count back seat driving, whinging, and looking stupid). And it clearly has a benefit. The precautionary principle would say that we should immediately legislate on this pressing matter. Or would that actually impact them, despite it having no costs?

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  27. Kimble (4,377 comments) says:

    “where the uncertain risk is potentially large enough (and potentially large numbers of human lives seem to me to fit that), then it can be shown mathematically that a cost benefit analysis will support the precautionary principle on utility maximising grounds *even if this precludes proof being found*.”

    DT, the world will explode and all who are alive now will die, and all who have lived shall rise from the grave and then die again, and the whole lot of us will burn in hell of all of eternity, UNLESS you do one of the following.

    1) switch the lights in your room on and off five times with the finger on your left hand that last touched an orange,
    2) kick the next dog you see,
    3) kick the next child you see,
    4) kick yourself in chest,
    5) only eat pork for a month.

    For the sake of the world, man, choose one! Quickly!

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  28. thedavincimode (6,530 comments) says:

    “[DPF: Attack the message, not the messenger]”

    The precautionary principle would suggest that the message should not be attacked where the messenger is a complete fuckwit who purports to build an argument on vaccuous rhetoric whilst completely ignoring relevant questions such as the those you you have posed.

    I call baloney on your call not to attack the messenger because the precautionary principle applies here. Dean is a fuckwit. I eagerly await his low powered call that stopping when pursued by police should be entirely voluntary. Or have I missed it?

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  29. wat dabney (3,655 comments) says:

    If this Senior Lecturer is not smacked about the head immediately with a giant dildo we can’t be sure the world won’t end next week…

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  30. Hagues (711 comments) says:

    There is not much worse than someone saying “I can’t prove my point, but you should just do what I say anyway.”

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  31. slijmbal (1,210 comments) says:

    There has been a lot of testing of physical and cognitive impairment caused through alcohol (and tiredness and being in a good mood and ….). These were pretty catagorical. Impairment is marginal and pretty linear at first and then increases sharply ie it becomes exponential. It also gets paradoxically harder to increase one’s blood alcohol levels once it starts to get high ie getting to double the legal limit requires one to drink a lot. If 3-4 drinks brings one to the legal limit then it can typically take 10-14 to get to twice the legal limit. We all know from personal experience that that amount of alcohol renders most people unable to properly control their mouth never mind a motor vehicle.

    If one wanted to insitute an alcohol measure that would be known to have an effect then vastly increasing the deterrence including removing the perpetrator’s ability to drive through incarceration when the blood alcohol levels are very high would be more likely to have an effect on accidents caused through alcohol.

    If we add to this one can be over the limit slightly through a miscalculation, one cannot miscalculate blood alcohol levels of .12 and above. This is deliberate and flagrant. There is no possibility of getting it wrong by having just that one extra drink. Based on what I have observed blood alcohol levels have to be insanely high for anything other than the standard 6 months suspension to be allocated as punishment and then it tends to be 12 months and a slightly larger fine.

    As an aside it is well arguable that texting while driving is at least as dangerous being on the blood alcohol limit. If the behaviour was based upon risky behaviour surely that should receive a 6 months license suspension. to be comparable.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  32. ross (1,454 comments) says:

    Dean Knight seems to have the precautionary principle arse about face. The principle refers to a course of action that an individual or policy-maker takes and the risk(s) associated with the action. The action here is a lowering of the BAC. So Dean needs to justify a lowering of the BAC and the associated risks (and costs).

    He says “we know that drink driving is a risky activity”. Actually we don’t know that at all. Certainly if a driver is drunk, the odds are greater that the driver will be involved in an accident…but Dean didn’t say that. He referred to “drink-driving”. So someone who has had one drink is presumably the target of Dean’s message.

    It should be noted that many accidents – the vast majority I suspect – involve drivers who have consumed ZERO alcohol. What should be make of that? Well, according to Dean’s logic, driving is a risky activity and every driver should have to demonstrate that they are competent to drive. Hmmmm, that’s not a bad idea :)

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote

Leave a Reply

You must be logged in to post a comment.