BICYCLE CRASH DATA ANALYSIS FOR HILLBOROUGH COUNTY SEPTEMBER, 2000 Hillsborough County Metropolitan Planning Organization 601 E. Kennedy Blvd., 18th Floor P.O. Box 1110 Tampa, Florida 33601-1110 (813) 272-5940
Background The Florida Department of Highway Safety and Motor Vehicles (DHSMV) maintains a database of bicycle/motor vehicle crashes and was the source for the data analyzed in this report. This crash data includes only those reported crashes. Not included are crashes where there are no injuries or any damage to the motor vehicle. Also not included are crashes involving two cyclists, a pedestrian and bicyclist or single bicycle crashes; crashes in these categories tend to be less severe and are difficult to track when an accident report is not recorded. Therefore, the data obtained tends to represent the more serious injuries and is very useful when concentrating efforts to improve bicyclists safety. Methodology The bicycle/motor vehicle crash reports accumulated in 1992, 1995, and 1998 were classified using a system developed by the U.S. Department of Transportation National Highway Traffic Safety Administration (NHTSA). This system identifies 44 specific types of crashes which can be grouped into seven major crash classes. In order to classify the crashes, each report was individually reviewed to determine who was at fault in the crash. Many times, the crashes involved Hit and Run, were reported some time after the crash occurred, or the officers did not record enough information to determine the cause of the crash. This corresponds to the high number of Miscellaneous or Other/Insufficient Info categories. New to the analysis was the addition of bicycle and pedestrian counts that were collected at twenty locations across Hillsborough County. Recorded was the number of subjects, time, direction, age group, gender, and whether cyclists were wearing helmets. Among many uses, this exposure data has been recognized as vital to determining whether high crash locations equate to a poor roadway design feature, or if the high crashes were due to a good design that attracted a high number of non-motorized travelers. The count data will be repeated along with the crash analysis to help develop trends in the amount of non-motorized activity, the types of crashes, the ages of cyclists, and the time of day crashes are occurring in our county. Trends The trend of Florida having the worst bicycle/motor vehicle crash rates has continued. Similar to the previous years of analysis, Florida s bicyclist fatality rate in 1998 was 0.66 per 100,000 population. This rate is more than twice the national average of 0.28 per 100,000. Even more alarming, in Hillsborough County the bicycle fatality rate was 1.08 per 100,000 which is almost four times the national average. Table 1 and Figure 1 show the trend in bicycle crash rates over the past eight years. The data show that crash rates in Hillsborough County have stayed relatively flat. 2
TABLE 1 Hillsborough County Bicycle/Motor Vehicle Crash Trends YEAR TOTAL # POPULATION CRASH RATE CRASHES (Per 100,000 pop) 1990 463 834,054 55.5 1991 475 839,185 56.6 1992 550 846,721 64.9 1993 548 859,457 63.8 1994 465 870,094 53.4 1995 552 891,680 61.9 1996 538 910,855 59.1 1997 480 922,100 52.1 1998 449* 925,277 48.5 *The DHSMV has stated that the injury figures should be considered as minimums. 22% of the crashes in Florida had not been counted when DHSMV compiled their data. 100 Figure 1 - Hillsborough County Bicycle/Motor Vehicle Crash Trends 80 60 40 20 0 1990 1991 1992 Crash Rate/100,000 Population 1993 1994 1995 1996 1997 1998 Types of Crashes The forty-four types of crashes were grouped into seven classes. Those classes are described below. 3
Class A Bicycle Rideout/Non-Intersection The types of crash situations in this class are when the cyclists rides out from a sidewalk, residential/commercial driveway, the shoulder, or mid-block. Mid-block rideouts occurred most frequently and are depicted below. Class B Bicycle Rideout/Intersection In this class of crash, the cyclists has not yielded for the motorist at a stop sign or other intersection, or in only two cases, the cyclists did not clear the intersection before the light turned green for cross traffic. Class C Motorist Turn or Merge This class, the largest class of crashes in all study years, includes situations where the motorist has failed to yield to the cyclists at a driveway, has driven through a sign or signal, is backing, or turned right on red in front of cyclist. Lastly, and the cause of 60% of the crashes in Hillsborough Coutny in this class, is when the motorist has pulled out from a road or driveway in front of a cyclist riding against the flow of traffic on a sidewalk. Class D Motorist Overtaking In crashes under this class, the motorist and cyclist were traveling properly in the same direction and the motorist failed to see the cyclist, misjudged passing space, lost control, or the cyclists swerved slightly to avoid an obstruction in the roadway. Motorist overtaking was the basis for most fatalities in 1998. Class E Bicycle Unexpected Turn In this class, the cyclist has turned or merged into the path of the motorist. This type of crash happens when the cyclist turns left in front of the motorist going the same direction or facing each other, the cyclist loses control, or the cyclist turns right from the wrong side of the street. 4
Class F Motorist Unexpected Turn These crashes happen when a motorist has turned or merged into the path of a cyclist. The motorist turned left in the same or opposite direction, or in many cases, turned right in front of a cyclist after passing. Class G - Miscellaneous includes: Under miscellaneous crashes are: wrong way cyclist or motorist, cyclist hits a slow or stopped vehicle, and non-roadway crashes (parking lots). Class Z Other/Insufficient Information The crash did not involve a cyclist, motorist or cyclist swings out too wide or cuts a corner, or the officer did not provide enough data on the report to determine the cause of the crash. Table 2 shows the comparison of those the crash classes for the three study years. TABLE 2 Bicycle Crashes By Class PERCENT OF TOTAL CRASHES CRASH CLASS 1992 1995 1998 A Bicycle Rideout/Non-Intersection 10.9 13.2 15.8 B Bicycle Rideout/Intersection 16.4 19.2 19.1 C Motorist Turn-Merge 22.5 23.4 24.5 D Motorist Overtaking 12.5 9.8 10.5 E Bicycle Unexpected Turn 10.7 8.5 5.1 F Motorist Unexpected Turn 11.3 8.7 8.5 G Miscellaneous 5.3 11.2 6.5 Z Other/Insufficient Information 10.4 6.0 10.0 5
The largest class of bicycle/motor vehicle crashes in Hillsborough in 1992, 1995 and 1998, is clearly Class C - Motorist Turn/Merge. Of the crashes in this category, most 6
Figure 2-1998 Bicycle Crashes by Class 150 A Bicycle Rideout Non-Intersection B Bicycle Rideout Intersection 100 C Motorist Turn/ Merg e 1992 1995 1998 A B C D E F G Z 50 0 D Motorist Overtaking E Bicycle Unexpected Turn F Motorist Unexpected Turn G Miscellaneous Z Other/Insufficient Inf o 7
have involved a cyclist riding on the sidewalk against the flow of traffic. The second greatest cause of crashes is Bicyclist Rideout usually from riding through a stop sign. As Figure 2 depicts, the number of crashes in each class remained generally consistent for the years of analysis. The most significant decreases of crashes in 1998 was found in Class E, Bicyclist Unexpected Turn. This category decreased from 8.5 percent of total crashes in 1995 to 5.1 percent in 1998. One possible contributing factor to the reduction of these crashes is that the Hillsborough County Bicycle Safety Education Program began in 1993 and was two years into existence. The focus of that program was to improve the knowledge and skill of cyclists riding in traffic. The course work included education in the classroom on applicable traffic laws, how to signal turning movements, properly fit a helmet, and the importance of bicycle maintenance. For the latter part of the training, the students were assigned to a bicycle and went through an obstacle course that simulated traffic situations. An increase is seen in crashes resulting from the bicyclist riding out into a motor vehicles path. Class A - Bicycle Rideout/Non-Intersection increased from 13.2 percent to 15.8 percent. Another increase is seen in Class D - Motorist Overtaking. The motorist speed, inability to detect a cyclists and allow for passing space, and at night, the cyclists lack of bicycle lights and wearing dark clothing are all contributing factors to these types of crashes. The variation in the last two classes, Miscellaneous and Other/Insufficient Information, is a result of how thorough the crash report has been filled out. The reports vary greatly with some having very specific data on the crash events, and others containing very little information. The State and local Bicycle/Pedestrian Coordinators have recognized the need for more law enforcement training in completing the crash reports consistently. Also discussed is the apparent reluctance of law enforcement to cite the motorist in crashes involving bicyclists. Severity Figure 3 shows the fatality trend in Hillsborough County since 1990. Of the eleven fatalities in 1998, four are attributed to motorist overtaking, where the cyclist is traveling in the correct location on the roadway, but the motorist misjudges or does not see the cyclist when passing. All four of the overtaking fatalities occurred during darkness. In total, eight of the eleven fatalities (73%), occurred during dark hours, whereas only 24% the total crashes in 1998 occurred during dark hours. There were 75 crashes with incapacitating injuries, with the greatest number of those (21%) caused by a cyclist riding out from a stop sign, without stopping. For some time now, bicycle and pedestrian planners have questioned the part that alcohol plays in bicycle and pedestrian crashes. In most of the reported crashes, the law 8
enforcement officers did not conduct a Blood Alcohol Test, although 39 reports indicated that the cyclist was drunk or had been drinking. 3 Figure 3 - Hillsborough County Bicycle/M otor Vehicle Fatality Trends 2.5 2 1.5 1 0.5 Fatality Rate/100,000 Population 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 Lighting Conditions The frequency of crashes by lighting condition is shown below in Figure 4. Two-thirds of all the 1998 crashes in Hillsborough County occurred during daylight hours. This number appears to be reasonable given that the purpose of most on-road bicycle trips occur during the day: commuting to work, school, shopping and social trips. Figure 4 - Bicycle Crashes by Lighting Conditions 6% 2% Daylight 17% Dusk Dawn Dark(street light) 67% 1% 7% Dark(no street light) Unknown 9
The frequency of crashes occurring in each class during daylight and non-daylight conditions is show below in Figure 5. Figure 5 - Bicycle Crashes by Class and Lighting 80 70 60 50 40 30 20 10 DAYLIGHT 0 DARK Rideout Non-Intersection A Rideout Intersection B Motorist Driveout C Motorist Overtaking D Bicyclist Turn E Motorist Turn F Miscellaneous G Insufficient Info Z DUSK/ DAWN Class D - Motorist Overtaking crashes in 1998 occurred more during non-daylight hours (20 crashes) than during the day (18 crashes). In the previous years, this was even more evident. Further, of all the crashes during non-daylight hours, Class D crashes account for 23 percent of those during non-daylight, yet only 11 percent of the all crashes were from motorist overtaking. Having dedicated lanes for cyclists and street lighting would greatly reduce these crashes, as well as educating the motorist to be more alert of bicycles sharing the roadway. There is also an important need to educate the cyclist to their visibility with both light colored clothing and bicycle lights. Gender Of the reported 449 cyclists involved in crashes in 1998, 79 percent of those were male and 21 percent were female. As shown in Table 3, the number of male cyclists involved in crashes has remained consistently greater than females over the years. The disproportionate number of males is probably due to a higher number of males riding bicycles. The slight increase of female cyclists involved in crashes may also correlates to the increase of female automobile drivers. Table 3 Percentage of Males vs. Females Male Female 1992 82.5% 16.5% 10
1995 81.4% 18.6% 1998 78.7% 21.3% In addition, in 1998 the ratio of males to females cyclists injured or killed across the State of Florida is 79.3% males, and 20.7& for females, which closely resembles the numbers for Hillsborough County. Race In 1998, bicyclists classified as white accounted for 54.9 percent of those involved in bicycle/motor vehicle crashes, while those classified as African-American accounted for 34.9 percent and Hispanic/other accounted for 10.2 percent. The African-American population of Hillsborough County in 1998 was estimated at 15 percent which indicates that the African-American population is over represented in bicycle/motor vehicle crashes. The number of crashes with cyclists classified as Hispanic is likely underreported due to being classified in the police reports as either white or African- American instead of Hispanic. The disproportionate number of African-Americans probably reflects their lower auto ownership and greater dependence on alternative modes of transportation. There is an obvious need to focus roadway improvements and education programs in areas where there are a greater number of minorities. Age Table 4 provides information on the age of bicyclists involved in bicycle/motor vehicle crashes. In 1998, crashes involving children under 15 years of age remained consistent with those occurring in 1995. The age group with the greatest decrease in crashes from 1995 to 1998 is seen in the age group of 15-24 years. Where most age groups have slight variations, but relatively constant numbers, there does appear to be an upward trend in crashes among the 45-64 year age group. Table 4 Age of Cyclists Represented in Crashes PERCENT OF TOTAL CRASHES AGE 1992 1995 1998 0-14 31.1 25.2 25.0 15-24 25.3 27.5 21.8 25-44 30.5 30.1 31.6 45-64 7.6 9.6 13.3 65+ 1.8 2.2 3.4 Unknown 3.6 5.4 4.8 11
Figure 6 shows the percentages of crashes in the age categories. The age group from 25-44 years had a higher number of crashes than the other age groups. Although this age group makes up almost 32 percent of the population of Hillsborough County, it is also the more active years for driving automobiles. It is interesting to note that this age category also had the largest number of crashes involving alcohol/drugs. Presumably, some cyclists who have had their driver s license revoked from driving under the influence choose cycling for transportation and continue to bicycle under the influence of drugs and alcohol. Figure 6-1998 Bicycle Crashes by Age 0-14 13% 3% 5% 15-24 25% 25-44 45-64 65+ 32% 22% Unknown As noted earlier in the report, the decrease in crashes due to Bicyclist Unexpected Turn could possibly be attributed to the Bicycle Safety Education Program, which stressed the importance of the laws that apply to cyclists. Even more so, the decrease in crashes for the age group of 15-24 from 27.5 percent in 1995 to 21.8 percent of crashes in 1998, may likely be a result of the Bicycle Safety Education Program. As students who have been trained in bicycle safety reach driving age, the lessons taught will make them safer drivers as well. While special efforts must be made to improve bicycle safety in each of these age groups, it is important to understand the differences in the class of crashes in each age group. Figure 7 shows the class of crashes occurring in each age group. Bicyclists in the under fourteen age group have been highly represented in Class A and Class B crashes in all years of study. These types of crashes are due to the cyclists riding out into traffic either at an intersection or midblock, without yielding the right of way. (Class A crashes for this age group did reduce from 33 crashes in 1995 to 24 in 1998). Children fail to stop first and look for traffic, especially when they do not understand that traffic on the street has the right of way over those entering the street. Added to the problem is that children have limited peripheral vision, and are not able to clearly judge vehicle speeds. As the age and awareness of the cyclist increases, the 12
type of crash class changes. More crashes occur as a result of motorist action in the age groups over 24 years old. 13
FIGURE 7-1998 Bicycle Crashes by Class and Age A Bicycle Rideout/ Non Intersection B Bicycle Rideout/Intersection C Motorist Turn/Merge D Motorist Overtaking E Bicycle Unexpected Turn F Motorist Unexpected Turn G Miscellaneous Z Other/Insufficient Info. 45 40 35 30 25 20 15 10 5 0 0-14 15-24 25-44 45-64 65+ UNKNOWN A B C D E F G Z 14
The data continues to emphasize the need to educate elementary-aged children about coming to a complete stop and understanding and obeying traffic laws. In addition, implementing a program for adult motorists and cyclists who receive a traffic citation that requires them to attend a bicycle/traffic safety course versus paying a fine, will surely help reduce the number of motorist/bicycle crashes. Sidewalk Riding and Riding Against Traffic Flow When a cyclists is on the sidewalk, they forfeit their right as a vehicle and must behave as a pedestrian. This means moving at a speed that allows for stopping at every curb cut along the sidewalk if necessary. For skilled cyclists, the safest location is on the roadway moving with the flow of traffic. Understandably, for young children, elderly, and in extreme adverse roadway situations, a sidewalk may be the best choice. Research indicates that it can be more dangerous to ride on a sidewalk, adjacent to the road, than on the road itself. The reasons include that at intersections, motorists are often not looking for bicyclists traveling at a higher speed than pedestrians. This is true especially when a motorists is making a turn. Sight distance can be impaired by buildings, walls, fences, shrubs along sidewalks and driveways. In 1995 there were 127 instances (23% of all crashes) where the cyclists was riding on the sidewalk or crosswalk, of those, 76 percent were against the flow of traffic. Another 43 crashes occurred with the cyclist riding against the flow of traffic in the roadway. In 1998, there were 100 crashes (22.3% of all crashes) where a cyclist was riding on the sidewalk. In 81 percent of those, the cyclist was on the sidewalk and against the flow of traffic. An additional 29 crashes occurred when riding against the flow of traffic in the roadway, taken with the 81 crashes against the flow on the sidewalk, there were a total of 110 crashes (25% of all crashes) that involved a cyclist riding against the flow of traffic. Table 5 depicts the sidewalk/crosswalk and against the flow of traffic crashes for each age group in 1998. For comparison, there was a decrease in the percentage of sidewalk/crosswalk crashes in the 15-24 age group from 32.2 percent in 1995 to 24.5 percent in 1998. Yet, a large increase in sidewalk crashes occurred in the 25-44 age group from 20.5 percent in 1995 to 30.9 percent in 1998. Although young children and older adult cyclists often ride on the sidewalk, their relatively low speeds may allow them to avoid crashes with motor vehicles. Conversely, the higher speeds of teen and adult riders increase the chance of these cyclists being involved in crashes at sidewalks and crosswalk. Motorists are not typically alert to a vehicle moving quickly on a sidewalk. The hazard of sidewalk riding is heightened when the cyclist is also traveling against the flow of traffic. A large percentage of crashes in almost all the age groups involved riding against the flow of traffic, mostly at sidewalk or crosswalk intersections, with some crashes occurring in the roadway. In 1998, almost 30 percent of all crashes for cyclists between 15 and over 65 years old involved riding against the flow of traffic. Yet the under 15 age group had the lowest percentage of these crashes at 16 percent. This once again may be attributed to the number of students that received the bicycle safety education programs at their school. 15
TABLE 5 AGE OF CYCLISTS INVOLVED IN SIDEWALK & AGAINST FLOW CRASHES AGE ALL CRASHES SIDEWALK/ CROSSWALK With Flow SIDEWALK/ CROSSWALK Against Flow % of TOTAL CRASHES SIDE/CROSSWALK With & Against Flow AGAINST FLOW In Roadway % of TOTAL CRASHES AGAINST FLOW In road & On Sidewalk 0-14 107 4 12 15 5 15.9 15-24 94 4 19 24.5 10 30.9 25-44 136 10 32 30.9 10 30.9 45-64 57 1 15 28.1 2 29.8 65+ 13 0 3 23.1 1 30.8 Unknw 42 0 0 0 1 2.4 n TOTAL 449 19 81 22.3 29 24.5 16
Location of Bicycle/Motor Vehicle Crashes The location of crashes involving bicycles and motor vehicles in Hillsborough County in 1998 is shown on Map 1. The location of crashes by jurisdiction is summarized in Table 6. More than half (58.8 percent) of the crashes occurred within the City of Tampa followed by 35.5 percent occurring in incorporated Hillsborough County. Table 6 Location of Bicycle/Motor Vehicle Crashes by Jurisdiction Jurisdiction % TOTAL CRASHES CRASHES PER 100,000 Pop # FATAL % TOTAL FATAL 1992 1995 1998 1992 1995 1998 1992 1995 1998 1992 1995 1998 Unincorporated 40.7 40.9 35.5 42.6 40.3 24.1 2 6 7 20 50 70 Tampa 55.8 55.1 58.8 109.5 107.2 81.0 7 3 2 70 25 10 Plant City 2.2 3.1 3.9 50.5 66.2 59.3 1 3 1 10 25 20 Temple Terrace 1.3 0.9 1.7 41.4 26.2 34.6 0 0 0 0 0 0 TOTAL 100 100 100 244 240 199 10 12 10 100 100 100 Specific areas which appear to have a high incidence of bicycle crashes include: the neighborhoods extending to the north of Downtown Tampa along Florida and Nebraska, as well as along Busch Boulevard, Hillsborough Avenue, Waters Avenue and Fletcher Avenue. These high crash areas may be characterized as low income areas or dense commercial development. Bicycles may be a major mode of transportation for many residents in these lower income areas who are traveling to work, school, and other destinations. In addition, the high volumes and speed of traffic may contribute to a higher rate of crashes. These 1998 crash locations are similar to those analyzed in previous years. Crash densities were higher in areas with higher population densities. This observation is consistent with 1995 and 1992. Particularly with the high percentage of crashes which occur within the City of Tampa. A large percentage of crashes in all three of the analysis years occurred within 3 miles of Downtown Tampa, as well as the area immediately west of the University of South Florida. Crashes also appear to be concentrated along the Interstate 275 corridor, extending north of Downtown, and in West Tampa along Columbus Drive. The Center for Urban Transportation Research has completed a University of South Florida Bicycle Crash Study that analyzed the bicycle-car crash statistics and makes recommendations for the reduction of on-campus bicycle crashes. Nearly 70% of all crashes were associated with sidewalk-riding and use of crosswalks. This demonstrates the inherent dangers of sidewalk riding and the problem of negotiating right of way within the 17
crosswalk. In the 19 crashes that occurred between bicyclists and car in crosswalks, 10 (53%) were the motorists' fault, 5 (26%) were the bicyclists fault, 2 (11%) were deemed mutual fault and 2 (11%) were undetermined. Recommendations The distinction of sidewalk crashes is important to note when we stress the need for providing on-road bicycle facilities in lieu of recommending the cyclist ride on the sidewalk. The lack of on-road facilities contributes to sidewalk riding. Another effective tool in educating motorists that cyclists have a right to use the roadways is signing and striping bicycle facilities. This analysis reveals a tremendous need for improving the safety of bicycle travel in Hillsborough County. Many of the physical constraints to bicycling can be alleviated through design, engineering or maintenance activities. These constraints and corresponding improvements are addressed in the Comprehensive Bicycle Plan. Along with physical improvements, attitudes and associated behaviors must also be addressed if bicycle safety is to be appropriately considered. While attitudes and behaviors are more difficult to change than the physical environment, they can be positively influenced by properly designed education and enforcement programs. Most safety improvement programs are designed to be implemented at the local level, yet efforts to support programs at the state and national level will also benefit Hillsborough County. Many programs can address more than one of the safety issues, however a variety of programs are tailored to different implementing agencies/organizations and funding potential. Safety improvement programs include, but are not limited to: Bicycle/Traffic Safety Education Law Enforcement Training Seminars Public Relations Campaign Improve Driver Education Curriculum, Handbook, Exams Develop and Distribute Bicycle Safety Information Increase Law Enforcement and Crash Reporting Efforts Create Citizen Watch Groups/Bike Patrols Create Parent/Neighborhood Safety Programs Develop and Distribute Safe Ways to Schools Maps Research/Lobby for Increased Funding Opportunities Host Bicycle Rodeos and Events in the Community The distribution of the Hillsborough County Bicycle Suitability Map, updated in October 2000, is a helpful tool for bicyclist looking for the safest, on-road, routes. 18
In addition, the need for continued data collection is evident. Making inferences from the data and developing a safety program for crash prevention is hampered by the lack of exposure and demand data. Exposure data, such as bicycle miles traveled, is essential to normalize crash data to gain a clearer picture for targeting needs groups. In April of 2000, a bicycle and pedestrian data collection effort began. The data gathered included: the number of cyclists and pedestrians at 20 sites across Hillsborough County, their age group, gender, location in the roadway, and any obvious violations. This same data will continue to be collected at a regular interval (every few years) and at the same sites. By doing so, trends in bicycle travel can be observed. This is essential in determining the effectiveness of the safety improvement programs, in validating the efforts of the mode-shift program to convert more trips to bicycle, and calibrating the demand models of the on-road facility program which is used to determine the location of priority bicycle facilities. If this data is not collected, and as bicycle facilities and ridership increase, the erroneous conclusion that bicycle facilities are dangerous as determined by increases in cross crash data or population based statistics is likely to occur. 19