The City of Austin, Texas, working with researchers from the University of Texas, has prepared a report on shared-lane markings. This is available at
I understand that this report also has been published in the Journal of the Institute of Traffic Engineers.
My critique of the shared-lane marking report follows.
First, safe bicyclist behavior was defined by three factors: (1) riding in the lane position indicated by the sharrow, (2) not riding outside of the lane (on the sidewalk or in empty parking spaces), and (3) not riding alongside queues of stopped vehicles.
None of these indices defines safety in any consistent way. Riding in the indicated lane position may or may not be safe, depending on conditions. Riding in empty parking spaces may be safe if a long string of them is empty. Riding alongside queues of stopped vehicles may also be safe, depending on lateral clearance, speed etc.
Second, safe motorist behavior was defined by three factors: (1) motorists give adequate space to bicyclists when passing, (2) motorists did not encroach on adjacent lanes when passing, and (3) motorists make complete lane changes when passing.
The first of these criteria defines safety. I can’t make any sense of the second. Just as in their Bicyclists May Use Full Lane report, the authors describe merging partway into an adjacent lane as “encroaching” — but with a shared-lane marking in one lane, it is hard to imagine how a motorist would overtake without using the adjacent one. Criterion 2 and criterion 3 seem to contradict one another. Perhaps the authors are trying to deprecate “straddle” passes (a definition is here) in favor or complete lane changes, but the latter involve even more “encroachment.”
The study does show generally positive results for behavior of both bicyclists and motorists following installation of shared-lane markings. That highlights the importance of correct placement of the markings, an issue which has become significant since their approval. The markings did not reduce filtering forward, but that is hardly to be expected, because filtering forward occurs when a lane is blocked by of stopped traffic.
Figure 3 — note that Guadalupe Street is one-way with 4 lanes, and that SLMs have been placed in both the right and the left outside lanes. This is unusual and might affect the results.
E 51st Street is a four-lane arterial that connects the suburban neighborhoods of north-central and north-east Austin. The facility has bicycle lanes west of Airport Boulevard and east of IH-35, but the lane width between Airport Boulevard and IH-35 is narrow, forcing bicyclists and motorists to share the road.
It would be more accurate to say that they share a lane.
Figure 7 — shows rather tight clearance next to a bus passing an SLM 11 feet from the curb, with parking, but no bicyclist is present. The SLM distance from the curb is the minimum specified in the MUTCD. Only small cars are shown parked at the curb.
The same safety criteria are repeated here as in the Executive Summary. Some of the same inaccurate terminology is used as in the BMUFL report (e.g., “incomplete passing maneuver” instead of “straddle pass”. The term “avoidance maneuver” is used incorrectly, as it is in the bike box report as well:
Avoidance maneuver – An avoidance maneuver was recorded whenever a bicyclist rode outside of the lane (e.g. rode on the sidewalk or cut through a driveway to turn).
This is a choice of route, not an avoidance maneuver, which is an emergency maneuver to avoid a collision.
There are other confused definitions:
Incomplete passing event – An incomplete passing event was recorded when the motorist passed a bicyclist without changing lanes.
This is an in-lane pass, not an incomplete pass. An incomplete pass would occur if a motorist initiates a pass and then decides not to pass.
Encroachment – Encroachment was recorded when a passing motorist occupied two lanes while passing.
This is not encroachment. It is a straddle pass, as defined and named accurately by Dan Gutierrez and Brian DeSousa in their report already cited. The motorist must yield to traffic in the lane he/she merges into.
Substantial changes in bicyclist and motorist behavior were recorded on Guadalupe Street and Dean Keeton. Reduction in sidewalk riding was more significant than change in bicyclist lateral position when on the roadway.
Changes were more subtle on E. 51st Street.
I note that the increase in bicyclist filtering forward (“bypassing the queue”) can be explained easily enough in that fewer bicyclists were on the sidewalk, and so more were able to bypass.
Conclusions and Recommendations
I generally agree with the conclusions and recommendations — that is, I think that shared-lane markings are useful — though I am appalled with the sloppiness of the report’s methodology and use of terminology, and I note that greater improvements would require education, not only placement of markings.