Tuesday, June 03, 2008

Bike documents: An analysis

Take a look at the latest city documents on cycling in SF. The Bicycle Coalition and the bike people commenting on this blog think these documents bolster their case that cycling in the city is on the rise and that it's safe to ride a bike in the city. They seem to be only half right, at best.

The "2007 Citywide Bicycle Counts Report" shows an increase in cycling in spite of the injunction preventing the city from implementing the Bicycle Plan until the EIR on the plan is completed. Under "Methodology" the city explains that it is counting cyclists primarily during the afternoon/evening, from 5:00-6:30. Why not count in the morning? "The evening peak period was chosen as the focus, as there is a better mix of trips than in the morning, where a greater majority of trips are work-related." 

But starting the count at 5:00 p.m. probably gets many of the same people going home after work, so it's not clear why that time represents "a better mix." Wouldn't it be useful for the bike people to show that commuting by bike is up? In any event, wouldn't "a better mix of trips" include counting during non-commuting hours? 

The last serious number we have on commuting by bike is the 2000 Census, which told us that only 1.9% of residents commute by bike. The SFCTA put it at 1%, though that number is the product of a computer model.

(Since the Bicycle Coalition works so closely with city departments on bike issues, one wonders whether they were tipped off about the count so that their members could be on the streets to inflate the numbers. I don't have any evidence for that, but neither the SFBC nor the city is above that sort of thing, given the zeal with which they are pursuing the bike fantasy.)

The latest count does show that the number of cyclists on city streets is up: 

The 2007 counts showed a 15% overall increase in the number of cyclists compared to the 2006 counts. Some locations saw an increase of up to 300% from 2006 to 2007. This increase is especially significant when viewed in light of the injunction against the City's Bicycle Plan.

Actually, only one location showed a 300% increase: Randall and San Jose, which went from 28 in 2006 to 112 in 2007, a large percentage gain but not a large number per se. That can be said of all the numbers in the report, including the totals: The 2006 total was 5626, and the 2007 total is 6454. Since the number of cyclists on city streets went up in spite of the injunction, this tends to show that riding a bike in SF doesn't depend on bike lanes.

The city has done the count for only two years, and this is "the first count that had an established baseline for comparison purposes," and "it is difficult to make specific conclusions about bicycle use or patterns." Only when the Bicycle Plan is completely implemented a few years from now will we have a good understanding of the relationship between bike lanes and overall bike use in the city.

The "2005-2006 San Francisco Bicycle Injury Collision Report" was put out by MTA in February of this year. This document is weakened by its failure to clarify a basic analytical concept. The report is tracking injuries to cyclists in "collisions," but does that term include accidents that don't involve other vehicles, the "solo falls" that make up most cycling injuries? Nowhere in the report is that distinction discussed, leaving the false impression that motor vehicles are the main safety problem cyclists face on city streets. 

The report notes that "Approximately 31% of bicycle collisions involved drivers from outside the City in both 2005 and 2006," implying that all that's being discussed in the report are "collisions" with motor vehicles. (Besides, what difference does it make where a driver comes from? Is the report implying that city drivers are more conscientious and less likely to be involved in accidents with cyclists?)

One interesting distinction is discussed: there are injury collision totals for both intersections and non-intersections, or "street segments":

Motorized traffic, bicycle and pedestrian traffic all play significant roles in determining injury collision totals: the more people that use an intersection, the higher the likelihood of collisions at those locations. Similarly, streets with bike lanes typically have significantly more bicycle traffic than streets that have shared lanes or no bicycle facility at all, and are more likely to have bicycle-related injury collisions.

That is, more cyclists on city streets inevitably means more injuries to cyclists. Another problem in methodology is the single source of the report's numbers:

The source of the collision data is the Statewide Integrated Traffic Records Systems (SWITRS) maintained by the California Highway Patrol (CHP). The California Vehicle Code (CVC) requires that local governments send their police collision reports to the State. The CHP provides electronic summaries of these reported collisions, which are then processed by local jurisdictions. 

The data used in this report excludes collisions that occurred on San Francisco freeways or private property, but includes collisions on city streets that are classified as state highways (such as 19th or Van Ness Avenues). There is typically a 6 month delay between the end of any calendar year and the official release of the composite SWITRS database by the CHP.

Does this mean that all injury accidents involving cyclists are reported by the SFPD to the state? And presumably injury accident reports are filed for all city streets, not just "city streets that are classified as state highways," as that poorly written sentence suggests. 

The Bicycle Plan itself expresses concern about the city's system of reporting cycling accidents:

The collision data presented in this chapter, while useful in identifying the most crucial roadway behaviors that lead to bicyclist injuries, does not include the many unreported bicycle collisions believed to occur in San Francisco...Currently, San Francisco General Hospital is not obligated to report bicycle injuries to the SFPD. This is left up to the injured parties. EMS (ambulance services) is supposed to report bicycle injuries, but many are not reported. Comparing police collision reports with SFGH emergency room visits or hospital admissions shows that approximately 20 percent of pedestrian injuries (caused by a collision with a motor vehicle) did not show up in police collision reports in 2000 and 2001. The rate for bicycle injuries is probably similarly under-reported (Policy Framework, page 6-12).

Another interesting---and unsurprising---conclusion we can draw from the data in this report: cyclists are often at fault for their own accidents. In 2006 Cyclists were at fault in injury accidents caused by running red lights 77% of the time, and they were at fault 88% of the time in accidents caused by "unsafe speed" (Bicycle Collision Report, page 8).

My assumption is that riding a bike in the city is more dangerous than this report indicates, which a more efficient accident reporting system, as advocated in the Bicycle Plan, would demonstrate. And the city's own numbers, however fragmentary, show that many cyclists are injured because of their own behavior, not due to the lack of bike lanes or the negligence of motor vehicle drivers.

Once the Bicycle Plan is completely implemented and all the "improvements" are made to city streets, cyclists will no longer be able to use a lack of bike lanes as an explanation for injuries to cyclists in San Francisco. The discussion will necessarily shift to the intrinsic dangers of cycling---as reflected in a more efficient system of reporting accidents---and to the reckless behavior of many cyclists as an important cause of such injuries on city streets.

Labels: , , , , , , , ,