Originally Posted by galen_52657
In the left photo, other than the cyclist's position too close to the curb, the car moving over into an unoccupied adjacent lane is perfectly acceptable and proper.
In a word: Why?
Keep in mind: when a driver changes lanes, he needs to clear himself. That takes his attention off the cyclist, or at least splits it in the worse possible way, forcing the driver to focus keen attention in two directions, 180 degrees apart: to the front and right (for the cyclist), and to the left and back (for the lane change). This is regardless of whether the lane adjacent to the motorist is clear of traffic or not. Lane changes are also precisely the time when a driver must physically turn his head 180 degrees away from the cyclist to clear himself and cannot trust his mirrors.
Also: If the motorist miscalculates or drifts during the time when his attention is split (the motorist will tend to drift left when paying attention to the right, and drift right when paying attention to the left; and keep in mind he has to do both), the cyclist comes out the loser either way. The motorist, if he drifts left, will overcorrect to the right to avoid a car; and if drifts right, will either hit the cyclist or over correct to the left into another car. If the driver sideswipes the car to the left, then there is a good chance he will lose control of the car and hit the cyclist as well. In every case, I'd estimate that it is better than a coin toss that the cyclist loses.
We take it for granted that we want cars to move left even if there is clear room for both car and bike to travel without changing course. In reality, this causes much stress to the driver and tends toward information overload and puts the cyclist in danger.
This study is in support of my theory (same as other theories here, supported by anecdotal evidence as well as logic, but by only a couple studies, this one being one of them) that the bike lane improves safety by making traffic more predictable. As far as intersections go, a skilled cyclist can use this predictability to his advantage because the presence of a strip of paint provides a clear distinction between the cyclist being in the bike lane, out of the direct path of a car, and the cyclist taking the lane, into the direct path of an approaching car. This way, destination lane positioning is crisper (for want of a better word) and there is no ambiguity about the intent of a skilled cyclist. The painted line has the effect of providing a reference point telling the motorist exactly where the cyclist is on the road and what direction (taking the lane, falling back into the bike lane, preparing for a turn or lane change, etc.) he is headed.
Close passing is only more than a nuisance when there is not sufficent predictability; i.e. on rural highways where people seem to (hopefully only figuratively) close their eyes and step on the pedal, hoping that nothing bad happens. When there is a high degree of predictability, which a bike lane facilitates, then close passing is not an issue; or at least it is only an issue inasmuch as riding vehicularly in traffic is an issue in and of itself.