Cara Seiderman, Transportation Program Manager, City of Cambridge
Ron Van Houten, Professor, Mount Saint Vincent University, Halifax, Nova Scotia, CA
Bicycle lanes have been established on city streets throughout the United States as a way of improving conditions for cycling and ensuring that motorists understand that bicyclists belong on the street. Multiple surveys have shown that bicyclists strongly prefer marked bicycle lanes when traveling on urban streets (figure 1). Some people have raised a concern about whether bicycle lanes are more likely to put cyclists at risk of coming in conflict with motorists opening car doors into the path of the cyclist. Although motorists parking a car are responsible for not opening a car door unless it is safe to do so, the reality is that many motorists have not been well educated about this. Attention has thus focused on whether pavement markings have an impact on bicyclist safety by influencing whether bicyclists ride closer to parked cars.
The purpose of this study was to determine how pavement markings influence where bicyclists and motorists position themselves on the road, particularly with regard to how far bicyclists travel from parked cars. The research examined the effects of sequentially adding the component markings that constitute a bike lane on Hampshire Street in Cambridge, MA. Hampshire Street has on-street parking and a substantial number of cyclists who travel on it. The street had just been repaved, offering the ideal opportunity for testing a variety of pavement markings. The study looked at what impacts the various markings had on parked motor vehicles, traveling motor vehicles and traveling bicyclists.
Research on bicycle facilities has often focused on examining bicycle lanes installed on roads without on-street parking (Harkey & Stewart, 1997; Hunter, Stewart, & Stutts, 1999). Several studies have shown that drivers make fewer wide swerves or close passes when passing bicyclists on streets with bicycle lanes (Kroll & Ramey, 1977; McHenry & Wallace, 1985) and have found that bike lanes reduced the percentage of encroachments by motorists into the next lane and resulted in less variation in the wheel path for bicycles and motor vehicles (McHenry & Wallace, 1985). McHenry and Wallace (1985) also found that motorists swerved less when passing cyclists when there was a marked bike lane.
Harkey and Stewart (1997) found that bicycle lanes as narrow as 0.9 m (3 ft) provide sufficient space for bicycles and motor vehicles to interact safely and that lanes of 1.2 m (4 ft) worked best. They also found that a stripe separating motor vehicles and bicycles produced fewer erratic maneuvers by motorists. Hunter, Stewart and Stutts (1999) discovered that there was more wrong-way cycling and more sidewalk riding at wide curb lane sites than at bicycle lane sites and that more cyclists obeyed stop signs at locations with bicycle lane sites. These studies involved comparisons of existing sites and did not involve comparisons of cyclist and driver behavior before and after facilities were installed.
One recent study did look at streets with on-street parking. The San Francisco Department of Parking & Traffic engaged Alta Planning & Design to study the effects of “shared use” markings on cyclists’ and motorists’ road position, cyclists’ riding behavior, and bicycle/motorist conflicts. The report, San Francisco’s Shared Lane Pavement Markings: Improving Bicycle Safety, (February, 2004) concluded that the markings increased the distance of cyclists from parked cars as well as the distance between cyclists and passing vehicles. One of the marking types, the “bike and chevron,” significantly reduced the number of wrong-way riders.
Figure 2. Only center lane markings.
Figure 3. Edge lines installed.
Figure 4. Edge lines with bicycle symbols and arrows.
Figure 5. Inner lines added.
Hampshire Street in Cambridge was the chosen location for implementing the series of pavement markings. Hampshire Street is 13.4 m (44 ft) wide, with parking on both sides of the street, an average daily traffic (ADT) of about 15,000 and bicycle volumes of 120 to 150 in peak periods.
The pavement marking treatments were implemented sequentially. First, data was gathered when the street was newly repaved and the only markings were a center line and crosswalks. Then, edge lines were established 3.7 m (12 ft) out from the curbs, creating 3 m (10 ft) travel lanes, and data collected with this measure. Then, bicycle symbols and arrows were put to the right of those lines, and data collected. Finally, inner lines were established, creating 2.1 m (7 ft) parking lanes, 1.5 m (5 ft) bicycle lanes and 3 m (10 ft) travel lanes. Figures 2–5 show these treatments.
The work was done between April and October of 2003.
Data measured were the distance cars parked from the curb, the distance bicyclists rode from the curb, and the distance traveling motor vehicles drove from the curb. The data on bicyclists and moving motor vehicles were gathered by videotape. The data on parked cars were gathered in the field. Data were collected at each stage of the implementation, so there were four sets of data collected: baseline, line alone, line with symbol, and full bicycle lane.
Surveys of bicyclists and motorists also were administered. An intercept survey of bicyclists and motorists was conducted during the baseline and final treatment condition. All intercept surveys were conducted at traffic signals on Hampshire Street. After the signal turned red, the research assistant or volunteer approached the stopped cyclist or driver and said, “Good morning/afternoon. I am doing a survey for the City of Cambridge and have a few brief questions to ask you. It will take less than a minute. May I proceed?” If the potential respondent refused, the surveyor approached the next person. There were few refusals. Cyclists who agreed to participate were asked to stay against the curb, out of the line of traffic. The baseline bicyclist survey (n = 117) had participants rate their comfort level on a five-point scale; how often they cycled on a five-point scale; and what they would change to improve cycling on Hampshire Street (an open-ended question). During the after survey (n = 123; 115 were scored for the rankings), cyclists were again asked to rate their comfort level on a five-point scale; how often they cycled on a five-point scale; if they noticed street markings on Hampshire Street over the course of the past few months (yes/no); and to rank each of the four conditions with “1” being most preferred and “4” being least preferred.
The baseline survey was administered to 129 motorists, and 120 received the “after” survey. The motorist survey asked drivers whether they were aware of bicyclists while driving on Hampshire Street; what about the street made them aware of bicyclists (an open-ended question); and how often they drove on Hampshire Street (five-point scale).
The three pavement marking treatments — an edge line demarcating the travel lane, the edge line and bicycle symbols, and a full bike lane — were all effective at influencing bicyclists to ride farther away from parked cars than when no pavement markings were present. Here are some details.
With the installation of the lane line (treatment 1), motorists parked significantly farther from the curb in both directions. The motorists moved in with each additional marking and in the end, there was no statistically significant difference between where motorists parked in the baseline condition and the full bike lane condition.
When one looks simply at an average position, the cyclists did move further away from parked cars in all circumstances, but only by a couple of inches — not as significant as might be hoped. However, the critical evaluation is the effect of the treatments on the distribution of where cyclists rode. Under all test markings, the distributions narrowed so that there were fewer outliers on either side (which is why the average did not change dramatically) (Van Houten and Seiderman, 2005). Most importantly, cyclists who were riding the closest to parked cars in the baseline condition moved further away, so the percentage of people riding more than 0.6 or 0.9 m (2 or 3 ft) from parked cars went up significantly.
The data also needed to be adjusted to account for the placement of the parked cars. At first blush, it looked as though the “line only” marking had the most influence on cyclist position, with the highest percentage of people riding more than 2.7 or 3 m (9 or 10 ft) out from the curb. However, when the data were adjusted to account for the change in where cars were parked, the three interventions became more equal in their impact of how far cyclists were from the parked cars.
There was also a difference among the locations, particularly between the locations near the signalized intersection and those near unsignalized intersections. The influence of the markings was greater on the cyclists near the former, because they started out closer to the parked cars. At the end of the study, the locations were similar as to where cyclists were riding.
The data revealed that the treatments had little effect on driver wheel path. Because Hampshire Street is relatively narrow and is busy at rush hour, when the data was collected, there may not have always been room for drivers to move into the opposing lane. The data on the mean distance between bicyclists and through vehicles show that the distance between bicyclists and the nearest through vehicle was greatest during baseline and significantly less at three of the four sites during the lane line alone condition. Since bicyclists were moving toward the travel lane with successive treatments, this finding is consistent.
Because this is a commuter route and because data were collected during commuting periods, it is not surprising that the vast majority of riders rode their bikes on Hampshire on a daily basis, and virtually all respondents rode at least several times a week. It was therefore reasonable to expect them to be aware of the various interventions.
Rider comfort ratings, on a five-point scale, averaged 3.4 during baseline survey and 3.3 during the after study survey — not statistically significant. Ratings in this range fall between neutral and fairly comfortable. When respondents were asked (in an open-ended question) what they would change to improve bicycling on Hampshire Street, by far the most common response was to “add a bike lane.”
During the after study survey, 80 percent of cyclists indicated they had noticed the markings. When asked to rank the various conditions from 1 (most preferred) to 4 (least preferred), cyclists ranked the full bike lane the highest (average rank of 1.25), the lane line plus bike symbol next (average rank 1.97), followed by the lane line alone (average rank of 2.95), and then no markings at all (average rank 3.78).
Another way of looking at this is to summarize which of the options were chosen as riders’ first preferences. Eighty-two percent of the respondents chose the full bike lane, and 8 percent chose the line with bike symbol. Since the latter is also a bike lane, 90 percent of the respondents prefered a bicycle lane.
Most drivers in both surveys drove on Hampshire on a daily basis. A similar percentage of drivers in both surveys responded that they were aware of cyclists on Hampshire (86 percent of the baseline respondents and 84 percent of the end of study survey respondents — not statistically different).
When asked, “What about this street makes you aware of bicyclists?,” motorists during baseline responded most frequently “nothing” (68 percent). After all of the treatments had been introduced the most frequent response was “bike lanes” (42 percent) and the second most frequent response was “I see them (the cyclists).”
This study shows that all three pavement marking options encouraged cyclists to ride farther away from parked cars. The bicycle lane was the most effective at keeping cars parked closer to the curb and encouraging cyclists to ride in a consistent position at intersections. Given that cyclists prefer marked lanes and have indicated that they make them feel welcome on the street, and that motorists do notice them, bicycle lanes can be seen as a preferred and positive way of providing for bicyclists in the transportation network.
This research was funded by the city of Cambridge. The project cost approximately $25,000 for the research effort, plus staff time, including markings done by staff and most of an intern’s time for about six months.
The research project was designed and evaluated by Dr. Ron Van Houten, Mount Saint Vincent University. In the City of Cambridge, those who participated in the study include: Susanne Rasmussen, Director, Environmental & Transportation Planning Division (E&TP), Community Development Dept. (CDD); Juan Avendano, E&TP, CDD; Joshua Kraus, E&TP, CDD; Michael Young, E&TP, CDD; Wayne Amaral, Traffic Operations Manager, Traffic, Parking & Transportation Department; and members of the Cambridge Bicycle Committee.
Transportation Program Manager
Environmental & Transportation Planning, Community Development Dept.
Cambridge, MA 02139
Ron Van Houten
Mount Saint Vincent University
Halifax, Nova Scotia, Canada B3M 2J6