🚀 go-pugleaf

> Thus we have established a correlation between a higher rate of lane
> changes and a higher accident rate.

Rahul, I think you misunderstand what it means to establish a
correlation.  A correlation is based on facts--measured data.  You can't
simply conduct a thought experiment ("imagine that you change lanes more
frequently...") and conclude, from that, that there is a correlation. 
Whether you realize it or not, to do so is circular.

Now, if you were to measure real-world driving situations, and you
weren't very particular about how you measured it, perhaps you would
find such a correlation, particularly if you used a coarse measure
(total lane changes, indicating more total vehicles), rather than a more
sophisticated measure, such as lane changes per vehicle-mile.  But I
would argue that you aren't measuring a correlation between lane changes
and accidents so much as a correlation between traffic density, or
vehicle miles, and accidents.  This is related to a semantic problem
with your assertion, which I want to discuss more later.  But first I
want to explain more of my point of view, so that my discussion of this
semantic problem will make more sense.

> What happens when you drive much faster than others?  You weave through
> traffic -- and thus have a higher rate of lane changes.  What happens
> when you drive much slower than others?  Everybody else changes lanes to
> get around you and thus experiences a higher rate of lane changes.

Clearly, you see the speed differential as requiring more lane changes
and, in your view, more accident potential.  Let me propose to you a
different way of seeing the problem.  

For the purposes of this explanation, consider a divided, limited-access
highway with two lanes travelling the same direction.  The convention
(and the law in many jurisdictions) on such a highway is to drive in the
right lane, and pass in the left lane.  (If you live in a country where
the direction of traffic is reversed, invert "left" and "right" in this
explanation.)  However, this convention is not well observed in many
places--everywhere in North America that I have driven, for example. 
Failure to observe lane discipline necessitates additional lane changes
for passing vehicles and, perhaps more seriously, passing on the right
rather than the left.  This creates a needlessly complex and, therefore,
more hazardous traffic situation.  Ironically, however, it is reluctance
to change lanes that is often the cause of this unfortunate situation. 
Very often, the driver who is using the wrong lane is doing so because,
having ended up in that lane, he is reluctant to leave it.  I come to
this conclusion after discussing this very question with hundreds of
drivers over the course of more than twenty years.  In my view,
reluctance to change lanes is one of the roots of the poor lane
discipline found on North American highways.  And this poor lane
discipline is what makes lane changing _appear_ to be part of the
problem.  The reality, though, is quite the reverse.  It is the failure
to change lanes on the part of some drivers that complicates the traffic
flow, resulting in more accidents than would otherwise happen.

Now the semantic problem.  Your assertion about a correlation treats all
lane changes as the same.  But they are not, as I have alluded to
above.  The sort of lane changes we often see on the highway today,
necessitated by circumstance and poor lane discipline, are very
different from the sort of lane changes we would see if proper lane
discipline were observed.  With proper lane discipline, we could have
the same speed variance we have today--maybe even more--with fewer
accidents.  That, to me, is a highly desireable situation.  But to get
there, we need to stop seeing lane changes as the boogey-man, and
recognize that they are an essential tool in an orderly traffic flow. 
We need to stop seeing lane changing as part of the problem, and learn
to see that it is actually part of the solution.  And we need to train
people how to make proper lane choices, and proper lane changes.

By the way, I really like your post script.  "Any idiot can avoid
automobile accidents that are his own fault.  It takes skill to avoid
accidents that are the *other* person's fault."  This really conveys the
true meaning of "defensive driving."

Tedd