Illegal pedestrian crossing at signalised intersections: Incidence and relative risk

https://doi.org/10.1016/j.aap.2009.01.008Get rights and content

Abstract

Illegal pedestrian behaviour is common and is reported as a factor in many pedestrian crashes. Since walking is being promoted for its health and environmental benefits, minimisation of its associated risks is of interest. The risk associated with illegal road crossing is unclear, and better information would assist in setting a rationale for enforcement and priorities for public education. An observation survey of pedestrian behaviour was conducted at signalised intersections in the Brisbane CBD (Queensland, Australia) on typical workdays, using behavioural categories that were identifiable in police crash reports. The survey confirmed high levels of crossing against the lights, or close enough to the lights that they should legally have been used. Measures of exposure for crossing legally, against the lights, and close to the lights were generated by weighting the observation data. Relative risk ratios were calculated for these categories using crash data from the observation sites and adjacent midblocks. Crossing against the lights and crossing close to the lights both exhibited a crash risk per crossing event approximately eight times that of legal crossing at signalised intersections. The implications of these results for enforcement and education are discussed, along with the limitations of the study.

Introduction

Virtually all people travel as pedestrians on the road system. Compared with driving, walking is subject to fewer rules which are confined largely to crossing or travelling on the road, however there is widespread non-compliance with pedestrian legislation. Several studies conducted between 1940 and 1982 found that about 25% of pedestrians crossed illegally at intersections (Mullen et al., 1990). More recently, Keegan and O’Mahony (2003) reported that 35% of pedestrians entered illegally at a signalised crossing. Pedestrian crashes account for around 15% of fatalities each year in Queensland (Australia) and about 8% of hospitalised casualties (Queensland Transport, 2005), and illegal pedestrian movements are a factor in these crashes. A study of pedestrian crashes at crossing facilities in New South Wales and Victoria (Austroads, 2000a) found that illegal pedestrian movements featured in 32–44% of pedestrian crashes at signalised intersections and 45% at pedestrian operated signals (i.e. not at a signalised intersection). In a more recent study, violation of traffic laws by the victim was found to be one of the “predominant contributing factors” in all pedestrian categories examined in a study of pedestrian crashes in El Paso County, TX (Ashur et al., 2003).

Knowledge of pedestrian rules does not seem to be the issue; rather, pedestrians want to cross where it is convenient for them, and with as little delay as possible (Gårder, 1989, Hamed, 2001, Holló et al., 1995, Sisiopiku and Akin, 2003). Enforcement of the rules by police is infrequent, and considered by the public to be unwarranted (Schonfeld and Musumeci, 2003). Engineering measures also tend to be resisted, with measures such as overpasses and underpasses (Holló et al., 1995) and pedestrian barriers (Kopelias et al., 2002) having little effect on illegal crossing behaviour.

Walking is being promoted for its health and environmental benefits, which has influenced approaches to engineering practice (Austroads, 2001a, Austroads, 2001b). Consistent with this approach is the recognition that many pedestrian crossing facilities are poorly located, as judged by the propensity of pedestrians to cross elsewhere. This propensity for illegal crossing could also be interpreted as indicating the rules imposed on pedestrians may unnecessarily restrict their mobility for the sake of safety benefits which are only modest or intermittent. However, there is a lack of information about the risks involved in illegal crossing behaviour. Such information would provide a basis for both enforcement and education campaigns, and would assist in prioritising engineering interventions. From the outset it needs to be acknowledged that the risks associated with illegal crossing are almost certain to be highly dependent on the context, including legal definitions, engineering practices and site characteristics. However, the existence and scale of such risks in one place would be suggestive of similar results in other places. The primary purpose of this research was therefore to quantify the relative risks of illegal crossing at a particular group of sites, while the secondary purpose was to investigate and comment on the variation between sites and the implications this has for generalisability of the results.

Constraints on resources and concerns about the safety of observers at night ruled out both a geographically widespread study and observations at night or on weekends. Accordingly, a study was undertaken on the relative risk of illegal crossing at signalised intersections in the Brisbane central business district (CBD) on typical workdays. Brisbane is the capital of the Australian State of Queensland and has a population approaching 2 million. Approximately half of all pedestrian crashes in Queensland occur in metropolitan areas (almost half of which occur at intersections), with a further third occurring in other urban areas (Austroads, 2000b). Looking at crashes by day of week and hour of day, a disproportionate share of crashes occurs on weekdays and in daylight (8 a.m.–6 p.m.) (Austroads, 2000b). In the period 2000–2004, about 17% of all pedestrian crashes in Queensland took place at “operating traffic lights” (which excludes “pedestrian operated signals”). Of these, a quarter occurred in Brisbane City (essentially the CBD), accounting for just over half of all pedestrian crashes in Brisbane City.

In Queensland, pedestrian lanterns feature the static shape of a man walking which is illuminated in one of three phases: when the illuminating light is a steady green, it is legal to commence and complete crossing; when the illumination changes to a flashing red, pedestrians must not start crossing but may complete a crossing commenced during the green phase; and when the illuminating light changes to steady red, it is illegal to start crossing and pedestrians still on the crossing must clear it as quickly as possible. Pedestrians must also use the pedestrian crossing if they are within 20 m of it, and cross between the marked lines. Any crash involving injury must be reported to police, and the generally more serious nature of pedestrian crashes means that levels of under-reporting are comparatively low.

Section snippets

Identification of data

The calculation of relative risks requires information on both crashes and exposure. Exposure was measured by conducting observations of crossing behaviour. The behaviours which were observed were determined based on the crossing rules, the kinds of illegal behaviour which are readily observable, and the availability of information in the police-reported crash data from the Queensland Road Crash Database, accessed via the net-based program WebCrash 2 (Queensland Transport, 2004). Observations

Incidence of illegal crossing

Results of the observations (Table 1) showed that the majority of pedestrians waited for the green man to become illuminated before crossing. The most common illegal behaviour was crossing away from the signals but within 20 m, followed by crossing against the flashing red man, then crossing against the steady red man. During observations, one intersection (Albert and Elizabeth Streets) was identified as problematic given the restrictions on access to one of the legs of the intersection, which

Comparison with previous research

When all illegal crossing types are combined, they accounted for 20% of observed crossings in this study. This less than the 25% reported in Mullen et al.’s (1990) review, though the sites in this study had high pedestrian volumes and were surveyed at times when volumes were highest. In contrast, illegal pedestrian movements were found to be involved in over 58% of crashes. This is greater than the 32–44% reported by Austroads (2000a), probably because the present study included midblock crash

Conclusion

In spite of these limitations, it is considered that the methodology employed in this study has been successfully piloted and can be improved and applied more widely. The results provide evidence that illegal crossing behaviours are associated with an increased crash level of risk. Replication of these results at a more general level would constitute grounds for a pedestrian crash prevention strategy involving publicity, a change in penalties, and enforcement. There are also implications for

References (25)

  • Austroads, 2001a. Traffic Flow Models Allowing for Pedestrians and Cyclists. Report No. AP-R193, Austroads,...
  • Austroads, 2001b. Potential Improvements for Non-Motorised and Vulnerable (Unprotected) Users in the Road System....
  • Cited by (0)

    View full text