The core study question to be answered is, "What effect does RLC programs have on intersection safety, as measured by changes in crashes?" The evaluation design presented later has this question as its primary focus. However, even this question requires some further clarification and expansion, in that RLC installations at selected intersections in a jurisdiction are often part of a larger "Reduce Red-Light-Running" campaign that is jurisdiction-wide. The overall program will include a public information component (whether planned or just as media coverage of this new enforcement technique) that can clearly have an effect on driving behavior at other intersections in the jurisdiction. Thus, the core question is immediately expanded to, "What are both the local effect of RLCs at treated intersections, and the 'spillover' effect at nearby intersections or jurisdiction-wide?" In addition, there are other program components or factors that might make such programs more beneficial in terms of crash reduction. Examples could include yellow interval or phasing changes done with camera installation, public information programs and signage related to RLC programs, and the issue of whether to ticket vehicle owners or vehicle drivers.
Because not every conceivable question can be answered in one evaluation, particularly a retrospective evaluation in which programs and data have been determined by the local agency rather than the evaluator, there was a need to clearly establish a list of key study questions to focus on (if the data would allow it). The FHWA established an internal project oversight panel to make these decisions. Forrest M. Council of BMI-SG and Bhagwant Persaud of Ryerson University, the project co-principal investigators, met with this panel in early January, 2002. Panel members attending that meeting included the following.
Michael Griffith, Chair, FHWA Office of Safety Research and Development
Pam Crenshaw, FHWA Office of Operations, Travel Management
Pat Hasson, FHWA Resource Center
Hari Kalla, FHWA Office of Safety, Safety Design
John McFadden, FHWA Resource Center
Joe Peters, ITS Joint Program Office
Amy Polk, NASA Jet Propulsion Laboratory
Greg Hatcher, Mitretek Systems, Inc.
Rob Maccubbin, Mitretek Systems, Inc.
The panel and project team discussed items such as issues related to the literature review (e.g., defining "critical studies," proposed study list, current progress, example detailed reviews), a listing of "lessons learned/issues raised" from the literature reviewed to date, preliminary thoughts on the experimental plan, etc. A major part of the discussion centered on the listing of study questions to be answered. At that initial meeting, the Chair and panel defined a draft listing of questions. This was revised slightly into the following final listing based on findings concerning available data and the range or spread of the data among the jurisdictions.
First-Level Priority:
- What effect do RLCs have on intersection safety (i.e., intersection crashes) at monitored intersections versus intersection safety throughout the jurisdiction?
- What is the relationship of signal timing (i.e., length of the yellow interval, length of the all-red interval, and various combinations of the yellow interval and all-red interval) with safety at intersections with RLCs? Later discussion indicated that the key factors of interest are yellow interval, all-red interval, cycle length, and signal coordination. The basic issue related to yellow time is the nature of the yellow phase length-e.g., a standard length, length based on Institute of Transportation Engineers (ITE) recommendations related to approach speeds and other factors, or some variation of these. The basic question for all-red phases is whether or not there is one (i.e., presence or absence of all-red phase). Cycle length is needed both to provide some measure of the number of red phases (and thus the number of opportunities for red-light-running) in a given time period, but also because longer red phases might "induce" more red-light-running. With respect to signal coordination, the issue is whether the treated signal approach is part of a set of coordinated signals that lead to queuing of vehicles (but not any additional details of the level of coordination).
- Are there certain improvements (e.g., signal timing, signage, geometric changes, etc.) done in conjunction with RLC installation that make the automated enforcement program more or less effective? Later discussion of the signage issue indicated that the key question has to do with presence or absence of "warning" signage, whether the sign is located at the intersection or away from the intersection (e.g., at the edge of town or at the beginning of a corridor), and whether informational signs providing data to the public on the number of red-light-running violations that have been issued are installed (because such signs have been shown to increase the effect of seatbelt enforcement programs and perhaps RLC programs in some cities).
- If other improvements are made during the installation of RLCs, what portion of the change in intersection crashes is due to these improvements and what portion is due to the RLC?
- What effect does a "good" public information program have on safety at intersections with RLCs? Given the fact that public information has been shown to be an important part of other effective enforcement programs (e.g., belt-use programs and driving under the influence (DUI) roadblocks), there is a need for some measure of "public information program level" for the cities. As noted later, the city interview form did contain three such levels.
Second-Level Priority:
- What is the effectiveness of "fine only" (i.e., owner liability) program versus "fine and points" (i.e., driver liability) program?
- What is the relationship between the reduction in violations and the reduction in crashes? (Note that this issue is not covered in the proposed experimental design because it would require a somewhat different study approach; crash effects is the primary focus.)
- What is the relationship of the grace period (i.e., ticketing threshold-how far into the red phase before a ticket is issued) and location of the camera activation loops with safety at intersections with RLCs? (Note that this was later removed as a question of interest because the "lag time" is a function of how the camera system is installed along with the legal definition of a violation. Because the driver is not aware of the extent of the "grace period," it is no longer felt to be of significant interest.)
Previous | Table of Contents | Next
FHWA-HRT-05-048
