Unfortunately, I will not have time to do this justice. Some links in logic will have to be filled in by the reader. My apologies to both John and HH.
I think that the data is lacking in a lot of dimensions that would be helpful in
directly answering the questions the repeatedly appear on this and other forums. In my post I briefly reference collection/selection issues which confound many population studies. So let me begin that I have not spent the time to fully understand the nitty-gritty details of these original studies which are so important in interpreting the final results. That written, my personal experience leads me to believe that there are always "glips and glops" in data collection that create noise and bias that are deeply hidden to the casual observer (me). The consequence is that I am far more agnostic than the typical researcher.
So while it may be the case that car-overtaking accidents are low risk compared with accidents in intersections--something that I believe for the reasons stated earlier--the data fails to directly address whether a bike lane makes the road safer or not. Particularly if you think that there are external effects that I alluded to earlier.
Briefly, the step that I believe John and HH skip is the acceptance, expectation of their presence, and recognition of cyclists on the road in the first place. In other words, I find that motorist acceptance and appropriate behavior toward cyclists tend to be better on roads that have a lot of cyclists or are designated a "bikeway". Around here, bikeway simply means a regular street with a green sign on the sign reading "bike route". By appropriate--did I use the word accordingly in some context earlier?--I mean passing at safe locations, speed, and distances. I also mean fewer aggressive (oblivious) maneuvers at intersections, allowing left turns, and so forth. My sense of this is purely anecdotal and not scientific. Furthermore, the roads with special bike status are different from other roads in structural ways. So what I report here is some sort of mental hedonic equation where I take several observations on multiple dimensions and try to project them onto the safety dimension. I don't mean to be abstract; but it is the fastest way for me to get my idea across. A good for instance is that it is my observation that motorists are more careful opening their doors on these streets with bike lanes/routes than alternative streets.
In other words, in a lot of circumstances, I think it is unrealistic to simply take the lane and expect motorists to remain law-abiding in the truest sense of the word. So you
do need to negotiate your space if you take the lane (MSF certainly teaches this to motorcyclists). Advocating that motorists simply follow the law and treat cyclists as an equal user of the road is very different from actually implementing that behavior in practice. Diane and John had an exchange that included something about human beings being emotional creatures. I think that there is quite a bit of literature that also talks about how people learn, cognitive limitations, and so on. I can envision many circumstances where some paint on the ground could be more effective in establishing cyclists' rights on the road than a bunch of speaches to city planners and police officers or a minority of cyclists simply demanding a change in behavior by taking lanes.
Regarding equal access to the road, I gather that there is an expectation for the cyclist to use said bike lane or to remain on the space to the right when reasonable. Bike lanes or other facilities can be done poorly. But assuming that we are not talking about the pathological cases--e.g., where the law expects the cyclist to make a left from the right lane, to the right of a right-turn-only lane, or other places the cyclist in precarious situations--then I don't find that the issues described in other posts manifest in any regularity. In other words, when I need to make a left, I make the appropriate signals and negotiate my way across. If there is construction ahead or debris, I take the lane for extended stretches. If I pick up speed to the point that a bike lane is unsafe, I take the lane for extended stretches. If there are a bunch of right turn only lanes, I take the through lane for extended stretches. That is, when appropriate, I ride VC. This is not a problem when there is motorist acceptance that a cyclist belongs in the road. It is a problem when motorists are aggressive.
Note that the cyclist inferiority notion is appropriate when we look at the ability to sustain damage. Otherwise the interaction between motorist and cyclist is not very different at all. That is, I imagine that many motorists in a car with a spike pointing out from the steering wheel and a bomb in the trunk would drive fairly cautiously ... avoiding hi-speed roads with lots of traffic, short lines of sight and so on. Someone might even make three right turns to make a left turn (my mental equivalent of going onto the sidewalk/crosswalk). Not very different from the cyclist behavior that I read on this forum or personally observe.
I just scanned through this ... I am beginning to ramble. My quick final statements ...
- I think that the discussion went past just bike lanes and into other facilities that I deem more effective. But I think either have a similar marginal effect regarding road space and general attitudes.
- I don't think that my observations are universal. At least, I have not made enough observations to believe that they are universal. So I can accept that in different localities that attitudes differ such that the prescription changes.
- What we really need are comprehensive surveys and observations regarding transportational/recreational cycling to talk with authority on best methods. Otherwise, I feel that much of the conversation--including mine--is fairly speculative.
- What I believe people fail to give John credit is actually writing down a bunch of concepts regarding cycling. People may not agree with them; but it gives a basis for a lot of discussion and makes it easier for a bunch of lazy bozos like myself.