> >The number of lives saved in the US, even if the depressing effect of risk compensation is 5/6 of the potential, say 85%, would still be my lowest figure of 70. That is enough to make a moral case for action.
>
> 85%?
>
> That has been utterly discredited, to the extent that even the
>
> majority of the authors have disowned it, and it's use in advertising
>
> in some countries is actually illegal.
I really don't know what you're talking about, Phil. You have some other cat by its tail, with your thumb up its arse.
Pay attention, Phil. What I'm saying is NOT that there would be a 16% risk compensation. What I'm saying is that, even in the event of a ludicrous 85% risk compensation, my numbers of positive effects are so conservative, they will be hardly affected in the low and the middle ranges. (A point Matt himself volunteered before I reinforced it.) I'm not speculating that either the 85% or the 1/6 is correct, I'm merely using hyperbole to show that, regardless of what anyone's guesstimate for counterfactors like risk compensation may be, my solid number for lives saved by wearing helmets, calculated from really hard numbers (see my original post in this thread), stands with monumental firmness. Matt agreed instantly, and so will any statistician and psychologist with experience in field work. (My hard numbers apply only to the States, eh? Extension to even other anglophone countries will be bedevilled by all the usual objections to extending the Australian results, even if they were reliable, to the rest of the world.)
Andre Jute
Here's Phil's entire post for context. Nothing more from me below this point.
On Friday, May 17, 2013 8:27:07 PM UTC+1, Phil W Lee wrote:
> Andre Jute <
fiul...@yahoo.com> considered Fri, 17 May 2013 01:12:09
>
> -0700 (PDT) the perfect time to write:
>
>
>
> >Matt >
>
> >Andre >>
>
> >
>
> >>I would have no issue with mandatory helmet laws for minors.
>
> >
>
> >Okay, we'll leave that aside then, just noting that helmets for both adults and children could be improved.
>
> >
>
> >> >Would you care to cite evidence for the "two behaviors, with references to papers available on the net? Put your best foot forward because I don't want to do more analysis than is strictly necessary.
>
> >>
>
> >> Risk compensation in cyclists … with other references cited …
http://injuryprevention.bmj.com/content/7/2/89.full
>
> >
>
> >I've seen this Adams and Hillman report many times before. It's one long waffle. There are no numbers in it. Essentially Adams and Hillman want us to take the word of another academic, Hedlund, that four rules he has established have primacy. It's crap. I see absolutely no reason why Hedlund, Adams and Hillman cannot put
>
> probabilities on "rules" like "If I don't know it's there I won't compensate for a safety measure" -- except that they *intend* to waffle because the number doesn't support their case, or, worse, is unknown because the "rule" was not researched but pulled from thin air. This crap isn't very far removed from "scientists" telling
>
> governments or pressure groups what they want to hear on, for instance, global warming.
>
> >
>
> >> Motorists response to helmeted cyclists …
http://www.bath.ac.uk/news/articles/archive/overtaking110906.html
>
> >
>
> >Now you're cooking with gas, Matt. This is far more solid stuff, though technically it can probably be attacked on grounds of relevance ("confidence levels") as some of those "closer shaves" are really small changes. But I won't be making that case, as I think this study is already large enough to give use the sort of general
>
> trend I was talking about in the original post. It also accords well with common everyday experience, of which I cite only the striking example of the bastardy of White Man Van, which is quantified in this study.
>
>
>
> I think the biggest problem with the helmet portion of that research
>
> is that (speaking for myself, but I can't imagine many people arguing
>
> differently) I'm not in the least bit interested in the /average/
>
> passing distance.
>
> What I want to know about is what it does to the /minimum/ passing
>
> distance.
>
>
>
> So if the people who were passing slow and wide already do so to a
>
> lightly lesser extent, it doesn't bother me in the slightest.
>
> The exact same reduction is passing distance from those who are
>
> already uncomfortably close will turn those passes into scary (and
>
> likely dangerous) close passes, yet both have exactly the same effect
>
> on the average passing distance, assuming roughly equal numbers of
>
> each. If people who were passing slow and wide outnumber those who
>
> were passing uncomfortably close, the slightly closer passes by the
>
> slow and wide contingent will change the average /more/ than turning
>
> some of the close passes into impacts.
>
> >
>
> >The problem with this study is actually the other way, that it is too good, and it's results too. What it actually proves is that most passes are legal, in that the average (it is possible or even likely that Dr Walker said "median" and was reported as "average" -- it is not the same thing, statistically, whatever it may be in the
>
> vernacular) automobile and even van and truck pass is *measured* as giving the cyclist the legal (in some places) one meter, plus 10%. In short, this study tends to prove that with or without a helmet, cycling is safer than most people assume, a case I've been making for years.
>
> >
>
> >But we are still left with a problem, which you've already identified:
>
> >
>
> >> I would certainly concede that the effect here is likely to very much within the broad range of outcomes you’ve already postulated.
>
> >
>
> >We can't even discuss the Adams/Hillman/Hedlund "contribution" because it is currently worthless. If those guys want us to consider their point of view, they must put verifiable numbers on their "rules", like Dr Walker does. Until they have repeatable numbers, it is nothing but a theory. We have too many unverified theories
>
> presented to cyclists as "facts" already, and too many ignoramuses like Krygowski and Lee presenting whatever theory fits their political outlook as a "fact".
>
> >
>
> >I could probably stretch to putting a guestimate on the Walker study, but it would, as you suspect, make no change to my ultra-conservative analysis in the original post, which relates only to the United States.
>
> >
>
> >We are thus no forrarder, unless we want to give Adams and Hillman a weight they don't deserve.
>
> >
>
> >> >> Cycling (along with other aerobic exercise) is one of the most effective means to reduce the incidence of heart disease. It is particularly effective since cycling, if included as part of the transport network rather than simply for recreation, can easily become a daily exercise that doesn't require additional time to be
>
> found.
>
> >
>
> >> .... that’s why I think the point is valid; it is precisely the casual, travel user who is more likely to be deterred by a helmet law than the committed cyclist. To use me as a case (not statistically valid I’m well aware) … I ALWAYS wear a helmet when I ride my mountain bike for recreation … and, by time spent, this is the
>
> vast majority of my time on a bike. However, I also use the Barclays Bike Rental facilities in London to travel around the city. I never wear a helmet to do this because it’s not convenient to carry one around simply on the off-chance that I might take a bike. The vast majority of bike renters I observe behave in the same
>
> manner; very very few wear helmets.
>
> >
>
> >> >> Studies in Holland strongly suggested that mandatory helmet use would significantly deter the casual cyclist from riding by two mechanisms; it would prevent the casual user from using cycle schemes (such as London's Bicycle Hire Scheme) and it would also portray cycling as dangerous thereby discouraging the activity.
>
> >
>
> >MHL could exempt those riding the bikes of recognized community bikes.
>
> >
>
> >> I think, if I may, I will instead refer you to Australia. The problem is that the Dutch study was a literature review used to defeat a proposal whereas the Australian one, which is cited, is more pertinent since they DID introduce a ban and saw a concomitant reduction in cycling in their cities. There are many references to
>
> this but I would suggest, at the risk of opprobrium, the Wikipedia page
http://en.wikipedia.org/wiki/Bicycle_helmets_in_Australia since it fairly shows both sides of the argument and presents a good, balanced view.
>
> >
>
> >I have forced Wikipedia to remove all reference to me under threat of action against their Board. That demonstrates the contempt in which I hold Wikipedia. However, I did read that article. I'm familiar with most of those papers, and lived in Australia through part of the relevant period. All of those numbers are bedevilled by the
>
> general downward trend of cycling through the entire latter half of the last century.
>
> >
>
> >> However, what is relatively clear is that there has been a significant reduction in journeys undertaken by bike since the helmet law introduction. There has also been a significant increase in obesity and heart disease too … although that is, of course, not sufficient to show causation.
>
> >
>
> >No, this really won't do. You're giving us the same sort of generalities that Adams and Hillman do, when we need hard facts. For instance, how much of the obesity and heart disease would have been avoided if cycling had not, as you claim, been depressed by MHL. I agree, theres a link between lack of exercise and respiratory
>
> diseases, but it would be foolish to assume that 100% of overweight parties would have taken up cycling, or even 10% of them.
>
>
>
> You don't need to show that /any/ of the overweight population took up
>
> cycling, only that a certain percentage stopped cycling, and if you
>
> want to gild the Lilly, that they did not replace it with another
>
> aerobic activity (which seems rather likely - those put off are the
>
> transport cyclists, who mostly aren't cycling for exercise in the
>
> first place, it just happens to be a fringe benefit of their choice of
>
> transport).
>
> >
>
> >> >If indeed you can prove that mandatory helmet laws deter cycling.
>
> >
>
> >> Indeed, and of course you can’t. The best predictor however would be to consider where such a move has been made elsewhere. Whilst all factors are not equal it is still likely to be illuminating. If we consider the case in Australia, where helmets were made mandatory, the overall reduction in cycling (for transport rather
>
> than recreation) was between 25% - 40% (depending on the removal of certain sites where anomalies year on year pertained) in the years following the application of the rule.
>
> >
>
> >Even the 25% number is open to serious doubt from any experienced statistician. There's a problem which nobody mentions, which is that cycling was in a transitional state between poverty transport and a claim on discretionary incomes over much or even all of that period. There is absolutely no proof that cycling was growing, and
>
> serious suspicion that cycling was in decline anyway. It is the sort of consideration that any commercial researcher would first eliminate, but the academics seem unable to understand that usages are dynamic.
>
> >
>
> >> >> So, overall, the likely benefit is considerably lower than postulated and the behaviour modification it might engender could be sufficient to more than offset any gains overall. More study required.
>
> >
>
> >The number of lives saved in the US, even if the depressing effect of risk compensation is 5/6 of the potential, say 85%, would still be my lowest figure of 70. That is enough to make a moral case for action.
>
>
>
> 85%?
>
> That has been utterly discredited, to the extent that even the
>
> majority of the authors have disowned it, and it's use in advertising
>
> in some countries is actually illegal.
>
> If there is any effect overall, it's too small to measure at all, so
>
> what would be the effect of 1/6th of bugger all?
>
> >
>
> >> >...[my] present numbers prove beyond peradventure that nationally in the USA, making >cycling helmets compulsory would save between 70 and 400 plus avoidable deaths every >year.
>
> >
>
> >> ... Each year, approximately 2.5 million people die in the USA from all causes (source World Health Organisation) … a death rate of 8.39/thousand. So, the 200 lives you are discussing have to be put in that overall context … it’s a very large population. Essentially, what you are saying is that you could improve the death
>
> rate by 0.008%. It’s reasonable, I would suggest, to question whether there might not be other factors which should be considered for what is, overall, a very marginal improvement.
>
> >
>
> >Really? I'm talking about cutting the cycling death rate in the States in half, say, and you say that's not enough. Bit hard on the 357 cyclists you've just condemned to death because there aren't enough of them dying to make saving any worthwhile -- that is your argument, isn't it? It's the same argument for which I treat
>
> Krygowski like barbarian.
>
>
>
> You haven't saved any - see the 1/6th of bugger all, above.
>
> But you've condemned a lot to death from heart disease, diabetes,
>
> stroke. etc.
>
> >
>
> >> If we look at death rates by type then the annual figure for Ischaemic Heart Disease is 1.26 per thousand (15% of 8.39/thousand). The death rate for cycling is 0.0024/thousand. It is therefore reasonable to assume that were, say, 20% of the cycling population to cease doing so that this would impact their health. Their risk
>
> rate would, for a while, still be lower than 1.26 but it would certainly rise. Say it went up by just one tenth … that would still be 629 additional deaths per annum from heart disease.
>
> >
>
> >What makes you think cyclists, who voluntarily take exercise, would stop exercising when they stop riding? Perhaps they use a rowing machine, perhaps they take up running or swimming. Certainly, the people I know who've given up cycling because they think it is too dangerous have not put on weight.
>
>
>
> I know a lot who have.
>
> The most affected group are those who rode not /for/ exercise, but for
>
> transport, and for whom any exercise gained was a fringe benefit.
>
> In other words the group most likely to be put off my a MHL are
>
> exactly the same group as those least likely to take up an alternative
>
> form of exercise - mostly because they just don't have the time (and
>
> having given up cycling in favour of sitting in traffic queues, they
>
> now have even less of it, plus having to work all the overtime they
>
> can get to pay for running a car).
>
> >
>
> >> So, it’s reasonable to say I think that the statistics are not clear cut and that a 0.008% improvement in the death rate is not a massive incentive to make a potentially liberty curtailing measure.
>
> >
>
> >Anything is "reasonable" if you're willing to let 400-odd cyclists die preventable deaths, every year.
>
>
>
> So you should be mandating cycling?
>
> >
>
> >If seat belts were a good law, then bicycle helmets wlll be a good law too. Improved bicycle helmets would be even more welcome.
>
> >
>
> >Andre Jute
>
>
>
> Seat belts largely moved death and injury from inside the car to
>
> outside it.
>
> Maybe a good start would be to ban them for drivers, but still require
>
> them for any passengers?
>
> It's a more restrained (badoom, tishhh) way of working /with/ risk
>
> compensation (instead of constantly fighting against it) than the
>
> frequently cited 8" spike on the steering wheel boss.