SUMMARY REPORT
Factors Contributing to Pedestrian and
Bicycle Crashes on Rural Highways
Approximately 25 percent of nationwide pedestrian and bicycle fatal and injury
accidents occur on rural highways. In contrast to urban highways, rural highways
have certain characteristics that can be more hazardous to pedestrians and bicyclists,
such as higher average vehicle speeds and a lack of sidewalk provisions. Limited
research has been conducted on rural highways, where crash types have been
defined with more detailed coding than exists on standard police forms and where
crash data could be linked with roadway characteristics and traffic counts.
The goals of this study were to examine the differences between pedestrian and
bicycle crashes in urban and rural settings in North Carolina and to identify
problem areas (specific crash types and crash locations) on rural highways that
are of high priority for safety treatment and treatment development.
Background
A 1996 study analyzed 5,000 pedestrian and 3,000 bicycle-vehicle crashes
from five States.(1) This study used a crash-typing method developed by
the National Highway Transportation Safety Administration to refine and
update crash type distributions. The most common crash types were
midblock dart/dash, other midblock, intersection-related, and vehicle
The Highway Safety Information System
turn/merge. The majority of bicycle-vehicle collisions occurred when the
(HSIS) is a multi-State safety database that
parties were crossing paths, and they were usually due to a failure to yield.
contains crash, roadway inventory, and traffic
The most common crossing-path crash types were drivers failing to
volume data for a select group of States. The
yield, bicyclists failing to yield at an intersection, and bicyclists failing
current participating States California, Illinois,
to yield midblock.
Maine, Minnesota, North Carolina, Ohio,
and Washington were selected based on A recent review by Campbell, Zegeer, Huang, and Cynecki considered
over 200 studies pertaining to pedestrian safety.(2) They found that
the quality of their data, the range of data available,
and their ability to merge the data from the various while pedestrian crashes predominantly occur in urban areas, ped-
files. The HSIS is used by FHWA staff, contractors, estrian crashes in rural areas more often lead to pedestrian deaths,
university researchers, and others to study current possibly due to higher vehicle speeds. Pedestrian groups that were
highway safety issues, direct research efforts, and overrepresented were young pedestrians, pedestrians who had con-
evaluate the effectiveness of accident countermeasures. sumed alcohol, and older pedestrians. The most common crash types
were dart-outs, intersection dashes, and turning-vehicle collisions.
Other studies of rural pedestrian crashes found that vehicle type,
alcohol involvement, pedestrian age, weather, light conditions,
nonintersection location, road curve and grade, and surrounding
development type were all prominent characteristics in crash
injury and frequency.(3 5)
While the previously mentioned studies examined the general
characteristics of rural crashes, there is a need to investigate
more detailed crash types and characteristics of rural ped-
estrian and bicycle crashes to determine potential treatments.
This study explored these issues and addressed the role of
countermeasures in rural environments.
Research, Development, and Technology
.
Turner-Fairbank Highway Research Center
6300 Georgetown Pike McLean,V 22101-2296
A
Data
This study used data from pedestrian and bicyclist crashes in North Carolina and linked
them with roadway data from the Highway Safety Information System (HSIS).(6) A
subset of the North Carolina crash data were linked to urban and medium-to-high
volume rural roadway data in HSIS to provide the dataset used in this study, which
spans from 1997 to 2002. The data comprise 1,849 total bicycle-vehicle crashes
(52 percent rural crashes and 48 percent urban crashes) and 3,598 pedestrian-vehicle
crashes (54 percent rural crashes and 46 percent urban crashes).
General Comparison of Rural and Urban Crashes
The first objective of the study was to compare general descriptive statistics of
rural versus urban crashes. This general comparison is useful for indicating
which factors are common to both localities as well as which factors are over-
represented in a rural environment. The rural to urban crash comparison can
be found in table 1.
The most common crash types for bicyclists differed in rural and urban areas.
The most common rural crashes included bicyclists turning/merging into the
path of the driver and drivers overtaking the bicyclist. The most common
urban crashes included drivers failing to yield, bicyclists failing to yield
midblock, and bicyclists failing to yield at the intersection. One noticeable
difference is that common rural crash types generally occurred at midblock
segments, while urban crash types generally occurred at intersections.
Both rural and urban areas had the same top four most common pedestrian
crash types: (1) pedestrians walking along the roadway, (2) pedestrians
failing to yield, (3) miscellaneous, and (4) pedestrians darting/dashing
midblock. Similar to the comparison of bicycle crash types, the most
common rural pedestrian crash type (pedestrians walking along the
roadway) was more common at midblock segments, whereas the most
common urban crash type (pedestrians failing to yield) was mostly
found at intersections.
Table 1. General comparison of rural and urban crashes.
PERCENT OF CRASHES
CRASH FACTORS CRASH TYPE RURAL URBAN
Pedestrian 18 10
Resulted in fatality
Bicyclist 6 2
Pedestrian 24 19
Pedestrian or bicyclist alcohol involvement
Bicyclist 8 6
Pedestrian 46 20
Vehicle speed 41 60 mi/h
Bicyclist 47 9
Pedestrian 57 11
Road speed limit 50 mi/h or higher
Bicyclist 54 3
Pedestrian 18 39
Intersection-related
Bicyclist 23 48
Pedestrian 71 18
Occurred on road with unpaved shoulders
Bicyclist 80 20
41 54
Pedestrian
Occurred in daylight
Bicyclist 66 73
2
Exploration of Factors in Rural Crashes
The second objective of this study was to identify specific crash types and crash locations on rural highways that are
of high priority for safety treatment and treatment development. This section examines rural crashes according to
road class and crash type to identify problem areas. It also examines prominent characteristics of those problem areas.
Analysis of Rural Crashes by Road Class
Given the availability of roadway inventory data in HSIS, it was possible to examine rural pedestrian and bicycle
crashes with respect to characteristics of the roads where they occurred. The roadway-miles and vehicle-miles in
table 2 and table 3 are a statewide reflection and do not simply represent roads where crashes occurred.
Roads in rural areas were divided into classes according to number of lanes, level of access control, and other
characteristics. It should be noted that 94 percent of North Carolina State-owned roads are rural two-lane roads.
For bicycle crashes, crash frequency and crashes per vehicle-mile indicate that rural two-lane roads had the highest
number of crashes, even after vehicle exposure was accounted for. The crashes per roadway-miles indicate that
rural multilane undivided nonfreeways would be the most cost-effective roadway class to treat on a per-mile basis.
For pedestrian crashes, crash frequency indicates that rural two-lane roads had the highest number of crashes.
Crashes per roadway-mile indicate that rural multilane undivided nonfreeways would be the most cost-effective
roadway class to treat on a per-mile basis. Crashes per vehicle-mile indicate that rural two-lane roads and rural
multilane undivided nonfreeways had the highest number of crashes.
Overall, the rural two-lane roads class had the greatest need for safety improvements due to the large number and
rate of crashes that occurred on these roads. While it is true that it would be costly to treat the extensive mileage
of rural two-lane roads, funds for safety research and treatment development would be better spent if focused
on this roadway class, and the treatments could be targeted to certain locations or segments within this class.
The lack of pedestrian and bicyclist volume data prevents researchers from knowing whether the crashes are
distributed evenly in accordance with where people walk and ride or if they are disproportionately represented on
certain road classes. However, given the available data on crashes and vehicle volumes, the North Carolina data
indicate that treatments for rural pedestrian and bicycle crashes should focus on two-lane roads.
Table 2. Rural bicycle crashes by road class.
CRASHES PER
6-YEAR CRASHES PER 100 MILLION VEHICLE-MI
RURAL ROADWAY CLASS CRASH FREQUENCY 1,000 ROADWAY-MI PER YEAR
Two-lane roads 725 23.9 0.51
Multilane divided 43 52.2 0.17
Multilane undivided 28 76.9 0.28
Freeways 3 3.3 0.01
Table 3. Rural pedestrian crashes by road class.
CRASHES PER
6-YEAR CRASHES PER 100 MILLION VEHICLE-MI
RURAL ROADWAY CLASS CRASH FREQUENCY 1,000 ROADWAY-MI PER YEAR
Two-lane roads 1,331 43.9 0.93
Multilane divided 110 133.5 0.45
Multilane undivided 71 195.1 0.72
Freeways 118 131.1 0.20
3
Analysis of Rural Crashes by Crash Type
One of the most prominent features of the North Carolina pedestrian and bicycle crash
database is the crash type data. Based on police crash report sketches and narratives,
each crash is categorized as a particular crash type, such as drivers overtaking a bicy-
clist or pedestrians walking along the roadway. This study explored crashes accord-
ing to type and road class. Frequently occurring combinations of crash type and road
class were identified as problem areas (e.g., pedestrians walking along the roadway
on rural two-lane roads or drivers overtaking a bicyclist on rural two-lane roads).
These problem areas were then explored to determine recurring crash characteristics.
Pedestrian, bicyclist, driver, environmental, and roadway characteristics of the prob-
lem area were compared to the characteristics of all crash types on that road class
(e.g., all crash types on rural two-lane roads). This comparison indicated which
characteristics were overrepresented in that problem area. Any overrepresentation
of a particular characteristic indicated potential treatment areas. For example, if
characteristics of a problem area indicated that younger pedestrians were more
involved than the general distribution, treatments specific to younger pedestrians
would be more appropriate for improving the problem area.
The characteristics of each problem area were compared to the characteristics
of all crash types on that road class to determine how they differed. This study
examined each of the 11 pedestrian problem areas and 5 bicycle problem
areas. An example problem area (walking along roadway crashes on rural two-
lane roads) has been included to show the results of the examination of the
characteristics of the pedestrian, driver, environment, and location.
Discussion of Countermeasures for Rural Problem Areas
After the problem areas were identified, potential countermeasures were
discussed to reduce those types of crashes. The Federal Highway
Administration (FHWA) has sponsored two projects, PEDSAFE and
Example Examination of Problem Area Characteristics
Crash Type: Pedestrians Comparison of pedestrian walking along the roadway to all
walking along the pedestrian crash types.
roadway.
CRASH TYPE (PERCENT)
Road Class: Rural PEDESTRIANS ALL PEDESTRIAN
WALKING ALONG CRASH TYPES
two-lane roads.
THE ROADWAY ON ON RURAL
RURAL TWO-LANE TWO-LANE
Number of Crashes:
CRASH CHARACTERISTICS ROADS ROADS
369 (27 percent of all
Pedestrians 25 44 years old 45 34
rural two-lane road
crashes). Pedestrian alcohol involvement 35 24
Estimated vehicle speeds of
39 31
Crash Type Definition: 45 55 mi/h
The pedestrian was Dark unlighted roadways 76 50
walking or running Road Speed limits of 50 mi/h or
68 60
along the roadway and higher
was struck from the Undeveloped area 64 54
front or from Unpaved shoulders 92 86
behind by a vehicle.
4
BIKESAFE, to provide comprehensive information on pedestrian and bicyclist safety,
specifically focusing on crash types and countermeasures.(7,8) This current study
examined crash types from pedestrian and bicyclist problem areas, considered the counter-
measures that PEDSAFE or BIKESAFE suggests for that crash type, and discussed
the suitability of the countermeasure with respect to the rural setting. Through careful
consideration and discussion with a limited group of pedestrian and bicycle experts
who were involved in the development of PEDSAFE and BIKESAFE, the suitability
of the countermeasure for rural settings was given a consensus-based rating for
two measures the potential safety effectiveness of the countermeasure and the
feasibility of its implementation in rural areas.
Summary of Countermeasure Discussion
The countermeasures that rated high for potential safety effectiveness and
feasibility in rural areas for each crash type are listed below. Pedestrian and
bicycle activity can be relatively low on some rural roads. Thus, certain coun-
termeasures (indicated in the following list with the word targeted ) may
only be suitable for rural areas where there are higher levels of pedestrian and
bicycle activity. Note that a full discussion of possible treatments, including those
not considered feasible, is in the final report (Transportation Research Board
Annual Meeting 2007 Paper #07-2457).
Rural pedestrian crash types and their solutions include the following:
Pedestrians walking along the roadway.
Add sidewalks (targeted).
Add paved shoulders.
Add roadway lighting (targeted).
Pedestrians failing to yield midblock.
Educate pedestrians.
Pedestrians darting/dashing midblock.
Improve signing (targeted).
Educate pedestrians.
Utilize traffic-calming measures (targeted).
Disabled vehicle-related.
Educate drivers.
Pedestrians failing to yield at the intersection.
Educate pedestrians.
Install pedestrian signal (targeted).
Improve roadway lighting (targeted).
Pedestrians crossing the expressway.
Improve roadway lighting (targeted).
Install fence or barrier (targeted).
Rural bicycle crash types and their solutions include the follow-
ing:
Bicyclists turning/merging into the path of the driver
midblock.
5
Provide marked pavement space for bicyclists (locations with suitable pavement
REFERENCES
width).
Add paved shoulder.
1. Hunter, W., Stutts, J., Pein, W., and
Cox, C. (1996). Pedestrian and
Drivers overtaking midblock. Bicycle Crash Types of the Early
1990s, Report No. FHWA-RD 95-
Provide marked pavement space for bicyclists (locations with suitable pave- 163, University of North Carolina
ment width). Highway Safety Research Center,
Federal Highway Administration,
Improve roadway lighting (targeted).
Washington, DC.
2. Campbell, B.J., Zegeer, C., Huang, H.,
Bicyclists failing to yield midblock. and Cynecki, M. (2004). A Review
of Pedestrian Safety Research in the
Reduce lane width to minimize crossing distance and slow vehicles United States and Abroad, Report No.
(targeted). FHWA-RD-03-042, University of North
Carolina Highway Safety Research
Bicyclists failing to yield at the intersection. Center, Federal Highway Admin-
istration, Washington, DC.
Improve sight distance.
3. Ivan, J.N., G rder, P.E., and Zajac, S.S.
Improve school zones. (2001). Finding Strategies to Improve
Pedestrian Safety in Rural Areas, Storrs CT:
University of Connecticut, Connecticut
Conclusions Transportation Institute. Obtained from:
http://ntl.bts.gov/lib/11000/11500/11542/
UCNR.pdf. Site last accessed June 4, 2010.
4. Hall, J.W., Brogan, J.D., and Kondreddi, M.
The goal of this study was to gain additional knowledge on rural pedestrian (2004). Pedestrian Safety on Rural High-
and bicycle crashes. A general comparison of rural and urban crashes in ways, Report No. FHWA-SA-04-008, Federal
Highway Administration, Washington, DC.
North Carolina found that rural crashes were typified by higher fatality rates,
Obtained from: http://www.walkinginfo.org/
higher vehicle speeds, less roadway lighting, unpaved shoulders, and more pdf/FHWA/Rural_Pedestrian_Safety.pdf. Site
last accessed January 7, 2005.
nonintersection locations than urban crashes.
5. Ossenbruggen P.J., Pendharkar, J., and Ivan, J.
(2001). Roadway Safety in Rural and Small
An examination of rural crashes by road class showed that rural two-lane Urbanized Areas, Accident Analysis and Preven-
roads had the greatest needs for safety improvements due to their high raw tion, 33(4), 485 498.
6. Federal Highway Administration. (2003). Highway
crash frequencies and crash rates per vehicle-mile.
Safety Information System (HSIS), An Analysis Tool
for Making Informed Safety Decisions, Department
Specific problem areas were identified and described in terms of char- of Transportation, Washington, DC.
7. Harkey, D.L. and Zegeer, C.V. (2004). PEDSAFE:
acteristics of the crash participants and crash location. The study identified
Pedestrian Safety Guide and Countermeasure Selection
11 pedestrian problem areas, such as walking along roadway crashes on System, Report No. FHWA-SA-04-003, Office of Safety,
rural two-lane roads, and 5 bicycle problem areas, such as bicyclists turning/ Federal Highway Administration, Washington, DC.
Obtained from: http://www.walkinginfo.org/pedsafe.
merging into the path of drivers midblock on rural two-lane roads. Site last accessed June 4, 2010.
8. Hunter, W.H., Thomas, L., and Stutts, J.C. (2005).
Potential countermeasures for these problem areas were discussed to BIKESAFE: Bicycle Countermeasure Selection System,
Report No. FHWA-SA-05-006, Office of Safety, Federal
determine their potential safety effectiveness and feasibility for rural Highway Administration, Washington, DC. Obtained
areas. Potential pedestrian crash countermeasures for rural areas from: http://www.bicyclinginfo.org/bikesafe. Site last
accessed June 4, 2010.
included improving roadway lighting, educating pedestrians and
drivers, and adding sidewalks and paved shoulders. Potential bicycle
F OR MORE INFORMATION
crash countermeasures for rural areas included providing marked
pavement space for bicyclists, adding paved shoulders, and improving
Additional information can be found in the full report that
roadway lighting. was presented on this study. It can be obtained as follows:
Carter, D. and Council, F. (2007). Factors Contributing to
Pedestrian and Bicycle Crashes on Rural Highways, Transportation
Acknowledgements Research Board 86th Annual Meeting Paper #07-2457, Trans-
portation Research Board, Washington, DC. Obtained from:
http://gulliver.trb.org/news/blurb_detail.asp?id=7286.
This study was sponsored by the FHWA under HSIS project
The full report can also be found at www.hsisinfo.org.
TFH61-00-C-00034. The authors appreciate the support of HSIS
For more information about HSIS, contact Carol Tan, HSIS
Manager Carol Tan, Office of Safety Research and Development program manager, HRDS, 202-***-****, abqmc5@r.postjobfree.com.
and the Office of Safety, in defining the scope of the study and
for reviewing preliminary versions. They also express their
Visit us on the Web
appreciation to Charlie Zegeer and Bill Hunter of the University
at www.tfhrc.gov
of North Carolina Highway Research Safety Center for their
input on pedestrian and bicycle countermeasures.
6 FHWA-HRT-10-052
HRDS-06/06-10(1M)E