On 22/12/2010 20:19, Matt B wrote:

On 22/12/2010 18:00, JNugent wrote:
On 22/12/2010 17:09, Matt B wrote:
On 22/12/2010 16:35, JNugent wrote:

As a parallel, only a small proportion of drivers over the drink-drive
limit on any one night are ever caught. And that's partly because most
acts of drink-driving cause no real danger to anyone and don't attract
the attentions of the police for various reasons. That doesn't mean that
drink-driving is harmless or worth ignoring, does it?

You tell me. Is there evidence that drink-drivers are over represented
in the accident statistics? I know that a few years ago some police
force somewhere did a Christmas campaign against drink-driving and
"randomly" tested drivers who were not involved in accidents or traffic
offences. They found a larger proportion were over the limit than for
those in the same area who were tested after being involved in an
accident or committing an offence. I've no idea though how
representative that was of reality or whether there have been studies
and there is real evidence that drink-drivers cause more harm than
sober ones.

There's so much non-sequitur there, it's hard to know where to start.

"Is there evidence that drink-drivers are over represented in the
accident statistics?"

That's not the issue (which is whether the majority of drink-drivers get
away with it - and of course they do).

You said: "That doesn't mean that drink-driving is harmless or worth
ignoring, does it?". I was exploring whether we actually /know/ if it is
harmful or not.

I left out the word "potentially" before "harmless". That was an error on my
part.

Merely driving with blood alcohol above a certain arbitrary value is clearly
not harmful to anyone. However, there is a statistical increase in the
potential to make mistakes which might lead to an accident.

The majority of drivers with alcohol in their blood do not have accidents and
that has always been the case, even before 1967, despite the hysteria
emanating from some, then and now.

The whole argument is a statistical one. There is nothing certain about
having an accident whilst "over the limit".

And to the extent that it ever
could be the issue, there is likely to be plenty of evidence showing
that cyclists are more, rather than less, likely to be involved in
collisions if they cycle along footways - whether with legitimate
footway users or at the interface between footway, footway-crossing and
carriageway (perm any two from three).

More likely to correlate, possibly, but what is the causal factor - is it
cycling on the footway per-se, or is it more to do with a reliance on and
faith in arbitrarily set priority conventions.

Both are part of the same phenomenon. Others are not having their (entirely
reasonable) expectations - even entitlement - met.

"there is real evidence that drink-drivers cause more harm than sober ones"

That's a, possibly unintentional, misrepresentation of what I meant

It's a verbatim quote from what you posted and is not unfairly truncated so
as to change meaning through precluding context.

- which
was: "I've no idea though how representative that was of reality or whether
there have been studies and [whether] there is real evidence [from those
studies] that drink-drivers cause more harm than sober ones.".

I don't know where you get that from. It's counter-intuitive at the
least, and smacks to a degree of "Doug's "two kinds of dead".

Why would a driver who's had four pints (say) "cause more harm than [a]
sober [one]"?

If driving safety was adversely affected by drink, then one would surely
expect drink-drivers to cause more harm (through having more accidents) that
sober ones.

That's a different issue. Most sober drivers driving home late on any given
night don't have accidents. Most drivers who are over the arbitrary alcohol
limit driving home late on any given night don't have accidents either.

Is there likely to be any evidence that for any accident which does occur,
one involving a driver with a blood alcohol reading of (say) 70 (perfectly
lawful) is likely to be "worse" than one involving a driver whose reading is
100? Or even 140? Let's leave aside those with grossly excessive readings
(too drunk to stand up, etc).

The answer is that he wouldn't, necessarily, or probably. He might be
more likely to be involved in a collision, but that's a different matter
and unless there's some super data somewhere out there which proves what
you say, it doesn't seem likely that accidents involving drivers with
illegal amounts of alcohol in their blood are any worse than accidents
involving teetotal drivers.

But are they more, or less, likely to have one (an accident)? If it isn't
"more", then what is the drink-drive law about?

That's the point: it's a pure statistical argument.

Drink-driving is banned (FCVO"DD") in order to reduce the number of
collisions, not to make collisions less eerious when they happen (though
that might be a side-effect in some cases).

Is there evidence that drink-drivers have more collisions than sober drivers?

Statistical evidence, certainly. But it is pretty raw. Some sober drivers are
"worse" than some "over the limit" drivers and that will never (be allowed
to) come out in the PR.

Blood alcohol enforcment is a rough justice measure because there's nothing
else available.

If something better (a test of spot-competence, perhaps?) were available,
it'd be better.

I think we've done full-circle now - that's where I started from.

The issue has at least been clarified.