The impact of speed and other variables on pedestrian safety in Maine

https://doi.org/10.1016/S0001-4575(03)00059-9Get rights and content

Abstract

The aim of this paper is to analyze pedestrian crashes with an emphasis on how actual travel speeds and characteristics of the locations influence crash numbers. Statewide data from Maine was analyzed. Descriptive crash and behavioral statistics are presented in the paper. Pedestrian and vehicle volumes were gathered for 122 locations in varying environments throughout Maine and crash numbers were predicted and compared to outcomes. Prediction models from Sweden and the UK were used since US models are nonexistant. It was found that high speeds and wide roads lead to more crashes and that the focus of safety improvement should be on arterials and major collectors. A strong relationship was also found between crash severity and speed.

Introduction

There are clear benefits from having people walk more, not least for their own health. For example, Carré (1997) estimates that a person gains 1 h of expected life for every hour he/she is engaged in moderate exercise, such as walking. And walking is probably the form of exercise that has the fewest negative side effects in the form of injuries—as long as injuries caused by collisions with motor vehicles are avoided. The primary objective of this paper is to quantify how pedestrian safety is influenced by geometric layout and actual speeds. Data from the State of Maine is used for this analysis. However, it is not only actual safety that is of importance in shaping people’s habits. Also, perceived risks may keep parents from wanting their children to walk, for example, to and from school. Table 1, based on interviews with 308 parents, shows that in Maine most people perceive traffic as more of a threat to their children than the risk of assault.

The pedestrian safety situation in the US is worse than in most other ‘civilized’ countries. The annual number of pedestrian fatalities per hundred thousand people in the US is according to NHTSA (2000) around 1.80. Most European counties have lower crash rates based on data from a report by the European Transport Safety Council (1999). The safest countries are the Netherlands (with a rate of 0.70) and Sweden (0.84). The fatality-rate comparison is certainly not flattering for the US considering the fact that people walk much less in the US than in the Netherlands and Sweden (Gårder, 2001). Surprisingly, the situation in Maine is noticeably better than in the average state. Maine had (according to the analysis presented here) 81 pedestrians fatally injured in the 5-year period 1994–1998. That gives a rate of 1.30 fatalities per 100,000 people. For 1991–1993, there were 45 fatal pedestrian crashes giving an almost identical rate (Maine Department of Transportation, 1994). However, this comparatively low rate may not be indicative of good planning or safe behavior, but purely reflect the fact that people do not walk as much in Maine as in many other states. And even if pedestrian safety is less of a public health issue in Maine than in some states, the situation is still far from satisfactory.

Section snippets

Crashes

Pedestrian crash data provided by Maine Department of Transportation show that there were a total of 1589 reported pedestrian crashes in the years 1994–1998. Analysis shows that Saturdays are over represented (P=0.03) compared to other weekdays and an analysis of the time of day shows that the statewide crash rate is about 0.05 crashes an hour in the morning and early afternoon, but double that between 4 and 7 p.m. Then, there is an almost linear reduction in numbers from 7 p.m. to 4 a.m. when

Method and data collection

Irrespective of severity, all pedestrian crashes were analyzed in detail for a small number of Maine towns—Bangor, Paris, Norway, Camden, Rockport, Hallowell, and Brunswick—representative of different types of communities in the state. The purpose of the analysis presented here is to study the relationship between actual speed and crash frequency.

Pedestrian and vehicle volumes were recorded along randomly chosen streets in the selected communities, prior to any knowledge of the crash data for

Risk from the pedestrians perspective

Crash data without exposure does not say much about risk. Therefore, crash numbers should be correlated to volumes. That can be done as outlined above or just by calculating crash rates, i.e. number of pedestrian crashes per crossing pedestrian. This latter rate is what is of interest to the pedestrian. For example, an injured pedestrian will be no happier if he/she is hit by a car on a road carrying 10,000 vehicles a day compared to on a road with 10 vehicles a day. In other words, the

Discussion and conclusions

It is human to make mistakes, no matter if you are a pedestrian or a driver. The consequences of mistakes can be deadly when vulnerable human beings are mixed with cars and trucks. It may even be human to break rules at times, and only in an ideal world could we regulate away all problems. Enforcement of existing rules governing safe behavior has possibilities to improve pedestrian safety—but probably only marginally if conventional police surveillance is used. That is, both because intense

Acknowledgements

This study was partially funded by the Federal Department of Transportation funds through the New England University Transportation Center and partly by Maine Department of Transportation.

References (13)

  • Brűde, U., Larsson, J., Hedman, K., 1998. Design of Major Junctions—Accident Prediction Models and Empirical...
  • Carré, J.-R., 1997. Mobilité urbaine et déplacements non motorisés: situation actuelle, évolutions, pratiques et choix...
  • Ekman, L., 1996. On the Treatment of Flow in Traffic Safety Analysis: A Non-Parametric Approach Applied on Vulnerable...
  • European Transport Safety Council, 1999. Safety of Pedestrians and Cyclists in Urban Areas. Brussels,...
  • FHWA (The Federal Highway Administration), 1994. Motor Vehicle Accident Costs. US Department of Transportation, FHWA...
  • Gårder, P., Ivan J., Du, J., 2002. Traffic Calming of State Highways: Application New England. US Department of...
There are more references available in the full text version of this article.

Cited by (193)

  • Modelling yielding dilemma zone for drivers at non-signalised crosswalks

    2024, Transportation Research Part F: Traffic Psychology and Behaviour
  • Predicting perceived risk of traffic scenes using computer vision

    2023, Transportation Research Part F: Traffic Psychology and Behaviour
View all citing articles on Scopus
View full text