In a series of posts regarding Bike to Work Day, Jan Heine first lamented the indiscriminate application of segregated bicycle facilities and followed by a summary that includes points made in the comments. For those unfamiliar, Jan is explaining the vehicular cycling position based on traditional traffic engineering principles and largely describes my perspective on bicycle facilities. While there is a lot to discuss, here I focus on two observations during these deliberations.
The first point is that with respect to risk, "bicycle drivers" -- those that ride their bike according to the traffic principles for all vehicles -- tend to look at the marginal effect of a facility whereas "facilitators" focus on the "total" effect. The total effect here is the direct effect of the facility on risk plus a safety-in-numbers effect or Smead's Law. There are two things that I consider pretty obvious. One, building on the argument Jan present and considering the literature on cognitive and visual awareness the effect on the margin increases risk to the cyclist. Two, there is a broad and strong correlation between more cyclists and lower risk.
The safety-in-numbers effect deserves more attention. There appears to be no scientific agreement regarding the underlying causality of safety-in-numbers. The linked authors discuss several mechanisms by which safety-in-numbers could be causal versus a spurious effect. Writing specifically with respect to cycling we can imagine that greater numbers of cyclists are simply more visible or result in changes in driver behavior. Alternatively, greater numbers of cyclists could result in slower motor vehicle traffic which results lower risk. Given the segregated facility question, the facility might attract more risk-adverse cyclists, dissuade drivers from using the street, slow down drivers, or slow down cyclists resulting in lower risk. Of course, these are just a few examples since one can imagine lots of reasonable mechanisms that result in the observed correlation. In the end, my take is that segregated cycling facilities should be used with caution to connect comfortable transportation grids with a "residential" pace. (Jan's conclusion.)
The second point is that many of these conversations end up being dominated by people speaking past each other in part due to people assuming the worst from "the other side". That is, bicycle drivers often assume that facilitators are willing to accept any cycling specific facility regardless of design or environment while facilitators assume that bicycle drivers reject any cycling specific facility. For what it's worth, I broadly find that there is a lot agreement even if the final conclusion is different. With the years of heated debate, it will be difficult to move people with a single conversation. But the first step is to start more civilized discussions without the polemics.