2. BACKGROUND

Two activities provided background information for the study. One consisted of a literature review on vehicle speeds to update an earlier review performed for NHTSA by Leaf and Preusser (1999). The second involved the selection of a panel of experts in traffic calming and speed management and conduct of a workshop to explore the speed problem, possible countermeasures, and possible evaluation measures. Each of these activities is described below.

2.1 Literature Review Approach and Relevant Findings

A review of the relevant literature was performed to help guide the study effort. There were two separate areas of focus. The first was an update of an extensive review previously performed for NHTSA (Leaf and Preusser, 1999). The second was a search for references that could relate speed reductions to specific pedestrian safety benefits.

2.1.1 Update of Previous Review

The purpose of this effort was to identify relevant materials that had been produced since NHTSA's previous report on the effects of speeding (Leaf and Preusser, 1999) was compiled. Searches were made of the Transportation Research Board Transportation Research Information Services (TRIS) database to identify studies published after 1998 on the topics of traffic calming, speed, enforcement, education, and pedestrian safety. In addition, contact was made with selected NHTSA and Federal Highway Administration (FHWA) representatives, with bicycle and pedestrian professionals, and with contributors to the vehicle speed report to request additional references and materials. Finally, Web sites of cities with known traffic calming programs were accessed to obtain additional reports and information. Appendix A contains the entire letter report of the literature review. Its highlights are presented below.

In all, over 175 documents were identified, and abstracts of each were read. Based on the abstracts, hard copies of 60 of these documents were obtained and reviewed. The subject matter of the reviewed documents ranged widely and included descriptions of specific traffic calming techniques, legal aspects of traffic calming, crime issues, property value issues, and others. Although all hard-copy documents that were received were read, the major interest was in identifying evaluative studies of methods of traffic calming, especially those that might have combined education and enforcement with engineering. Very few evaluative studies were located. However, reports that summarized and quoted results of evaluative studies were found and proved useful (see Appendix A).

Two major documents were produced since the publication of NHTSA's speed report (Leaf and Preusser, 1999). One was an Institute of Transportation Engineers (ITE) report on the state of the practice of traffic calming ( Ewing , 1999). Among other topics, it provides a brief history of traffic calming, a toolbox of traffic calming measures, engineering and aesthetic issues, impacts of traffic management measures, legal authority and liability issues, warrants, project selection procedures, public involvement, traffic calming on other than neighborhood and collector streets, and traffic calming in new developments. Twenty United States traffic calming programs are featured in the document. Specifications proposed by certain jurisdictions or professional groups for selected measures are included.

A second major document of specific interest to the present effort was a synthesis of safety research related to speed and speed management sponsored by the Federal Highway Administration (FHWA) ( Stuster, Coffman, and Warren, 1998) . This report covers speed-safety relationships, factors influencing speed, speed limits and speeds, speed limits and safety, enforcement and engineering measures.

The supplementary literature review revealed very little new information germane to the present study beyond what was contained in NHTSA's vehicle speed report (Leaf and Preusser, 1999). However, much of the new information served to confirm or amplify information contained in the NHTSA report. Across all of the literature examined, the following points of relevance to the current study were culled or inferred:

• Although some professional groups and jurisdictions have proposed standards, there are no national standards for design and use of traffic calming measures.
  
  
• The most effective traffic calming methods involve vertical treatments to the roadway. Humps have been proven to be very effective, and the public believes they are effective in reducing speeding. This was confirmed by the fact that the test sites for the present study chose vertical treatments for all of the physical calming measures employed.
  
  
• Traffic circles such as those employed in Seattle have been proven not to induce vehicle-to-vehicle crashes. They are, however, quite expensive and therefore are not widely used. Neither test site in the present study employed traffic circles in the studied neighborhoods.
  
  
• There are rubber humps that can be used as a temporary traffic calming measure. Other measures can be created for temporary use, for example, permitting parking on opposite sides of the street for each block to create a chicane and using planters as bulbouts. The expert panel and officials at the chosen test sites, however, were opposed to using these temporary approaches because they could mar the aesthetics of the test neighborhoods.
  
  
• Data on the effectiveness of roadway perceptual devices are inconclusive. Although street narrowing decreases speeding, there are data that suggest that it increases crashes. Some data show that street width must be reduced to 20 feet or less before speed reductions are noted. As it turned out, street narrowing was not an option in any of the test neighborhoods.
  
  
• No studies were found in which education alone was used as a traffic calming measure. Typically, education programs consist of citizen watches combined with enforcement. Success in reducing speeding has been reported anecdotally, but no rigorous evaluation studies were found.
  
  
• Enforcement is effective but may be impractical on low volume streets. Compliance is greatest in the vicinity of the police vehicle. This is also true of speed display boards. The public finds enforcement effective in reducing speeding. Again, although these concepts were promulgated by several authors, no definitive evaluation study was uncovered.
  
  
• Anecdotal reports and limited scale studies indicate that photo radar has been successful in reducing speeds when deployed in neighborhoods in Europe . Some success has also been reported in the United States but the technique still has problems with political acceptability. The public apparently believes that photo radar is effective in reducing speeds. Photo radar was not permitted to be used at either of the test sites. As discussed below, however, speed trailers, that tell drivers the speed at which they are traveling without taking any automated enforcement actions, were used.
  
  
• Measures of effectiveness for traffic calming are typically 85 th percentile speed, average speed, percent exceeding posted speed, percent “ x” miles per hour above posted speed and volume. Crashes are included sometimes. Only one study ( Cambridge , 2000) was located that used pedestrian-related measures – it used number of drivers yielding to pedestrians.
  
  
• Some jurisdictions use “hidden” measures for traffic calming, e.g., changes in traffic lights, prohibiting turns, making a “maze-like” path by creating a series of one-way streets. The two test communities chose not to employ these in the present study.

2.1.2 Pedestrian Safety Benefits of Speed Reduction

Relatively few references were located that provided a quantitative link between speed reduction and pedestrian safety benefits. Many of the references cited in Appendix A allude to the presumed benefits in terms of both injury reduction and crash avoidance, but there is a paucity of specific studies that confirm these links.

The case for the injury reduction benefits of lower speeds is perhaps more fully documented because both biomechanical analyses and epidemiological studies are relevant. For example, Leaf and Preusser (1999) report on the effects of vehicle speed on pedestrian fatalities and cite several studies that show that the risk of a fatality increases exponentially with striking vehicle speed. One study reported from England shows that 5 percent of pedestrians struck at 20 mph will die, compared with 45 percent at 30 mph, and 85 percent at 40 mph (Department of Transport, 1997).

Other methodological development studies have proposed complex formulas for calculating the crash reduction potential of diminished travel speeds (cf., Navin, Chow, and Kwan, 2001; Davis, 1998). While these studies clearly support the general notion that lower speeds are associated with reduced crash risk, they do not provide a specific method in the context of the present study to translate any speed reduction obtained into an estimate of crashes avoided.

Perhaps the most direct evidence for the pedestrian crash reduction potential of reduced speeds comes from a study by Tester et al. (2004). The study examined the protective effectiveness of speed humps in reducing child pedestrian injuries in residential neighborhoods. This case-control study showed that children living on streets where speed humps had been installed had lower odds of being injured within their neighborhoods and being struck by a motor vehicle in front of their homes. While this study provides excellent support for the crash-reducing potential of successful speed countermeasures, it does not provide a direct formula for estimating the benefits of any particular application.

A recent article (Lindenmann, 2004) examines crash quantification tools and aids in the context of all highway crashes, not just pedestrians. The overwhelming evidence reported is that lowering speed produces a reduction in crashes. Lindenmann (2004) cites a study from Finland (Kallberg, 1997) that concluded that an increase in the average speed of traffic by 1 km/h (0.62 mph) increases the number of injury crashes by approximately 3 percent. This equates to an increase of about 4.8 percent for a speed increase of 1 mph. The study also points out that crash costs increase by about twice as much since the higher speeds increase severity. Obviously, pedestrians are included in this overall estimate, although no way is reported to separate out the specific pedestrian crash or injury risks.

Thus, there is a widely held and partially proven theory that lowering speeds in residential neighborhoods will produce safety benefits both in terms of crashes avoided and by lessening injury severity when a crash does occur. This suggests that measures of both the mean or average speed and of those traveling at the highest speeds are needed in order to assess countermeasure programs such as those mounted in the current project.

2.2 Workshop on Speed-Reducing Countermeasures

A panel of experts was assembled for a workshop to discuss existing speed reduction approaches and brainstorm new ideas. In addition to the authors and the NHTSA task order manager, the panel included practitioners from a cross-section of jurisdictions and specialists on speed countermeasures, pedestrian safety, the state-of-the-practice of traffic calming, traffic enforcement, and education. The workshop was held on May 10-11, 2001, at the FHWA Learning Center in Arlington , Virginia . The complete letter report of the workshop is included as Appendix B to this report. A summary of the key results follows.

The workshop participants were challenged to enumerate candidate test conditions, evaluation paradigms, and possible test locations for the study. The aim was to identify a reasonable (in terms of cost and time) and sufficiently general test or tests that could answer the question posed by the main objective of the study, i.e., to determine if a speed-reduction benefit is obtained by adding education and/or enforcement to more traditional traffic calming approaches.

The workshop was initiated by asking participants to identify speed-reducing countermeasures in the three E's–engineering, enforcement, and education. Constraints and positive aspects were then noted for each. In an exercise in which participants attempted to identify a desirable set of countermeasures from those identified, no consensus was reached. However, the following principles, guidelines, and considerations that could be applicable to any project that attempts to reduce neighborhood speeds emerged from this activity:

• A countermeasure can trick the senses but it can't be deceitful (e.g., an artificial construction zone).
  
  
• A countermeasure cannot devalue the neighborhood; it should improve the neighborhood.
  
  
• Neighbors might not like to be the “bad guys” (e.g., participate in a trial by peers).
  
  
• There should be a valid punitive value in any sanctions employed (e.g., a trial by peers may not be a deterrent if there is no meaningful punishment possible).
  
  
• There must be places in the area to mount the countermeasure (e.g., there are few red lights in residential areas).
  
  
• A countermeasure must not be annoying or contribute to neighborhood litter (e.g., windshield wiper flyers).
  
  
• A countermeasure must be legal (e.g., mounting flyers on utility poles is illegal in many jurisdictions).
  
  
• A countermeasure must be easily targeted (e.g., in-car cameras are tough to target).
  
  
• The implementation and maintenance costs must be reasonable (e.g., cameras are expensive).
  
  
• A countermeasure should not generate any privacy issues (e.g., people don't want to feel watched).
  
  
• Providing a good model for children is desirable (e.g., sending materials home from school).
  
  
• Crime reduction can be an added benefit to some of the countermeasures used (e.g., added police patrols, increased lighting).
  
  
• Sufficient space must be available (e.g., a street must be sufficiently wide to install a median).
  
  
• It is preferable that residential countermeasures not limit access to homes and driveways (as would a median; however, short medians could be installed).
  
  
• Appropriate data must be available or collectible to evaluate the actions.
  
  
• Neighborhood involvement is advisable (e.g., a neighborhood speed watch program).

The guidance represented by these principles was helpful in selecting and implementing countermeasures in Peoria and Phoenix . All of them were considered and applied to the extent applicable in each of the participating test neighborhoods. This list can also be helpful to anyone attempting any type of neighborhood-based programs.

Although there was no consensus on specific countermeasure approaches, the workshop participants did agree that the project should work with permanent installations of engineering treatments whenever possible. It was noted that temporary installations, e.g., rubber speed humps, are often not aesthetically pleasing and can therefore engender negative reactions based on appearances alone. This can cause a backlash and make it difficult to generalize from temporary to permanent installations.

With regard to process, it was considered desirable that the project start from the beginning and look at transition from a non-calmed to a calmed neighborhood. Since it was acknowledged that this might not be possible under the constraints of the study, it was agreed that using an already calmed location with available “before” data was a viable alternative.

In a discussion of meaningful speed reduction, it was agreed that 85 th percentile and average are not the most representative measures of what people are concerned about. People want the excessive speeders eliminated. It was proposed that the project focus on overall comfort level and unacceptably high speeds. The primary aim would be to eliminate speed outliers. Thus, the project should focus on the high tail of the speed distribution as well as the mean. To follow this recommendation, the study would require accurate speed data on each vehicle, not class interval or “bin” data.

A brief discussion was held on possible test sites for the program. The sites mentioned included those where pre- and post-studies have not had optimum results as well as those that have good process and treatments pending. Although participants mentioned several possible test sites, there was no consensus on an ideal test site for the study. Therefore, an independent site selection effort was mounted to find an appropriate test location.

The literature review and the expert panel session suggested that the study needed test sites that have implemented (or are in the process of implementing) traffic calming in the form of permanent rather than temporary physical changes. It was also considered desirable to select locales that would be willing to attempt some innovative treatments that might provide compelling visual stimuli without constituting an impediment to the transit of emergency vehicles. As discussed in the next section, the study was fortunate in finding two contiguous sites that met these basic criteria and were interested in helping to develop and test the Heed the Speed program.