[Home Page]
[Up: Table of Contents]
[Previous: Section II]
[Next: Section IV]


SECTION III
MAGNITUDE OF THE PROBLEM

The purpose of this section is to discuss what is known and what is not known about the magnitude of the bicycle-accident problem. The first part of the section presents recent data on the incidence, consequences, and costs of bicycle/motor-vehicle accidents. The incidence and consequences of bicycle-related accidents that do not involve a motor vehicle are discussed in the second part of this section. The relative brevity of the second part of this section reflects the paucity of information about the kinds of bicycle-related accidents that are not the result of a conflict between a bicycle and a moving motor vehicle.

BICYCLE/MOTOR-VEHICLE ACCIDENTS

Bicycle/motor-vehicle accidents are defined here as accidents that result from an actual collision between a bicycle and a motor vehicle, or a collision with another vehicle or object (including the ground) that was the direct result of actions -- by one or both parties -- to avoid a collision between a bicycle and a motor vehicle.

INCIDENCE OF BICYCLE/MOTOR-VEHICLE ACCIDENTS

The only systematic data on the incidence of bicycle/motor-vehicle accidents come from record-keeping agencies that tabulate annually the number of traffic accidents that are reported to the police. Each year, the National Safety Council compiles data on the police-reported accidents that occur in a sample of states and uses the sample data to estimate the number of injury-producing accidents that occurred throughout the United States during that year. The National Safety Council reports that bicycle/motor-vehicle accidents have resulted in about 1,000 fatalities and about 40,000 disabling injuries [l] each year since 1972 (National Safety Council, 1977). Although the National Safety Council's estimates are the best available gauge of the incidence of bicycle/motor-vehicle accidents, the estimates are highly conservative because they are based only on police-reported accidents. The findings of several recent studies indicate that a substantial number of bicycle/motor-vehicle accidents that occur each year are not reported to the police. For instance:

  • A survey of 1,307 motorists in Santa Barbara County revealed that 4.2% of the motorists had been involved in a bicycle/motor-vehicle accident in the recent past, and that only 25% of the accidents were reported to the police (Cross & deMille, 1973).

  • In a nationwide survey of 23,699 elementary school children, students were required to describe their most serious accident during the past year or, if none, their most serious accident during the past five years. Of the 393 students who indicated that their most serious accident was a bicycle/motor-vehicle accident, only 37% indicated that their accident was reported to the police (Chlapecka et al., 1975).

  • In a study by Cross and Fisher (1977), a total of 525 bicyclists and 385 motorists who had been involved in a police-reported bicycle/motor-vehicle accident were asked if they had been involved in any other bicycle/motor-vehicle accidents during the past 24 months. It was found that the combined sample of 910 persons had been involved in a total of 47 other bicycle/motor-vehicle accidents and that only 27% of these accidents were reported to the police.

Based upon the above findings, it seems reasonable to assume that at least two-thirds of all bicycle/motor-vehicle accidents go unreported. One explanation for the large proportion of unreported accidents is that many bicycle/motor-vehicle accidents result in little or no injury, and it is these inconsequential accidents that are not being reported to the police. Although little is known about the consequences of unreported bicycle/ motor-vehicle accidents, some information on this issue was obtained from the data compiled by Chlapecka and his colleagues (1975). A special analysis of Chlapecka's data was performed to determine the consequences of unreported bicycle/motor-vehicle accidents in the sample. The results showed that more than 50% of the unreported accidents were severe enough to require some form of medical treatment (Schupack, 1975). Unfortunately, the data were not in a form that enabled a more precise assessment to be made of the degree of injury sustained by the bicyclists in the unreported accidents.

Although data on the incidence of bicycle/motor-vehicle accidents are meager, it is nevertheless possible to define the general bounds of the problem. Since the National Safety Council's estimates are based on police-reported accidents, it seems reasonable to assume that these estimates -- 1,000 fatalities and 40,000 disabling injuries -- represent the lower limit of the problem. But, what about the upper bounds? First, consider the number of fatalities that occur each year. Because nearly all fatal accidents are reported to the police, the National Safety Council's estimate of 1,000 fatalities per year should be quite accurate. This view is reinforced by the fact that the National Safety Council's estimate of fatalities corresponds closely with estimates of the National Highway Traffic Safety Administration, who publishes a monthly running total of all types of fatal traffic accidents. Next, consider non-fatal but injury-producing accidents. If the survey data cited above are assumed to be representative of the nation, it can be estimated that about one-third of all bicycle/motor-vehicle accidents are reported to the police, and that about one-half of the unreported accidents are injury producing. Using 40,000 as the estimated number of police-reported accidents, it can be estimated that a total of about 80,000 injury-producing bicycle/motor-vehicle accidents occur each year.

BICYCLE/MOTOR-VEHICLE FATALITY RATE

Although there has been a substantial increase in the total number of persons killed each year in bicycle/motor-vehicle accidents, data reported by the National Safety Council indicate that the increase in fatalities has been proportionately less than the increase in bicycles over the last four decades. Table 5 shows the fatality rate per 100,000 bicycles in use from 1935 through 1976 (National Safety Council, 1977). There is a high correlation between the number of bicyclists killed and the number injured each year. So, it can be assumed that the rate for non-fatal accidents would show the same trends reflected by the data in Table 5.


TABLE 5

FATALITY RATE PER 100,000 BICYCLES IN USE
FROM 1935 THROUGH 1976 (NATIONAL SAFETY COUNCIL. 1977)

YEAR BICYCLES
IN USE*
(MILLIONS)
DEATHS DEATH
RATE
1935 3.5 450 12.80
1940 7.8 750 9.59
1945 9.0 500 5.55
1950 13.8 440 3.48
1955 23.1 410 1.78
1960 28.2 460 1.63
1965 38.8 680 1.75
1970 56.5 780 1.38
1971 -- -- --
1972 71.4 1000 1.40
1973 80.0 1000 1.25
1974 90.0 1000 1.11
1975 95.0 1000 1.05
1976 95.0 900 .94

*Bicycles in use for a given year is the ten-year total
of domestic production plus imports less exports.


It can be seen in Table 5 that the fatality rate has decreased from a high of nearly 13 in 1935 to less than one in 1976. Since 1955, the annual decrease in the fatality rate has been small in comparison to the previous period. Even so, if the fatality rate had remained the same since 1955, there would have been nearly 1,700 bicyclists killed in 1976 rather than the 900 that were reported. Assuming the number of bicycles in use remains the same, reducing the fatality rate by only 0.1 will result in about 100 fewer deaths each year. Thus, the reduction in fatality rate since 1955 cannot be considered inconsequential.

CONSEQUENCES OF BICYCLE/MOTOR-VEHICLE ACCIDENTS

Although most experts agree that bicycle/motor-vehicle accidents are an important problem, few attempts have been made to assess the consequences of such accidents. A small amount of data is available on the severity of personal injuries and the extent of property damage resulting from bicycle/motor-vehicle accidents. However, the consequences of such accidents go far beyond personal injuries and property damage. Bicycle/motor-vehicle accidents also may result in undesirable consequences for the non-injured party and for persons not directly involved in the accident. Some of these other undesirable consequences are discussed below, following a description of the data on severity of injuries and the extent of property damage.

Personal Injuries

Although it is usually the bicycle operator who is injured, some bicycle/motor-vehicle accidents result in injuries to the motor-vehicle operator or to a passenger of one of the vehicles. Table 6 shows the total number of operators and passengers killed and injured in a sample of 166 fatal and 753 non-fatal accident cases (Cross & Fisher, 1977). It can be seen that the 166 fatal cases resulted in a total of 172 fatalities; the 753 non-fatal cases resulted in 765 persons injured. One case was found in which two bicyclists, who were riding separate bicycles, were killed in the same accident. Also killed were one motorist, one motor-vehicle passenger, and three bicycle passengers: The fatally-injured motorist and the fatally-injured motor-vehicle passenger were riding a motorcycle at the time of the accident and were killed in separate accidents. Study of the 753 non-fatal cases revealed that 3.3% of the accidents resulted in injuries to the motor-vehicle operator, 2.1% of the accidents resulted in injuries to a bicycle passenger, and .5% of the cases resulted in injuries to a motor-vehicle passenger.


TABLE 6

PERSONS KILLED AND INJURED IN A SAMPLE
OF 166 FATAL AND 753 NON-FATAL ACCIDENTS

  KILLED INJURED
VEHICLE
OPERATORS
BICYCLISTS 167 720
MOTORISTS 1 25
VEHICLE
PASSENGERS
BICYCLE 3 16
MOTOR VEHICLE 1 4
COMBINED OPERATORS
AND PASSENGERS
172 765

In the Cross and Fisher study (1977), information on injury severity was obtained only for the injured bicyclists; injury severity was assessed during face-to-face interviews with a total of 525 bicyclists. It was found that 92% of the bicyclists suffered injuries severe enough to cause them pain and discomfort for at least one day following the accident. The injuries sustained by 55% of the bicyclists were severe enough to prevent them from going to work or school for at least one day; 18% of the bicyclists were hospitalized for one or more days. Based upon the injury data compiled on the sample of 525 bicyclists, a bicyclist who is involved in a non-fatal bicycle/motor-vehicle accident, on the average, can be expected to suffer the following consequences:
  • 1.4 days in the hospital;
  • 1.4 days in bed at home;
  • 4.3 days missed work or school;
  • 23.6 days suffering pain or discomfort.

Analysis of the non-fatal injuries revealed that 76.4% of the injuries were body-surface injuries, 17% were skeletal injuries, and six percent were internal non-skeletal injuries. Considering the body-surface injuries first, it was found that abrasions and bruises together accounted for nearly two-thirds of the injuries, and about 11% of the injuries were lacerations. Considering next the skeletal injuries, it was found that 7.5% of the injuries were fractures, 5.6% were sprains, 2.7% were concussions, .9% were dislocations, and .6% were broken teeth. Nearly five percent of the injuries were aches and pains in the muscles and joints, arid slightly over one percent were ruptures of subcutaneous tissue, arteries, vessels, or organs.

The distribution of injuries for fatal accidents would certainly be different from the distribution of injuries for non-fatal accidents. Although Cross and Fisher (1977) did not investigate injury type for the fatal cases, other research indicates that the relative frequency of head injuries and internal injuries would be much greater for fatal than for non-fatal accidents. For instance, autopsies performed on 181 bicyclists killed in traffic accidents during the period 1935-1963 (Tonge, O'Reilly, Davison, & Derrick, 1964) showed that brain damage was evidenced in over 80% of the fatalities with an associated skull fracture occurring in 71% of the cases. Injury to abdominal organs was found in over 50% of the victims. Similar findings are reported by Bowen (1970) and by Gissane, Bull, and Roberts (1970).

For each injury identified in the Cross and Fisher study, the bicyclist was asked to define what caused the injury. It was found that 60.4% of the injuries were the result of the bicyclist's impact with the roadway and 24.1% of the injuries resulted from impact with the motor vehicle. It was surprising to find that only 6.2% of the injuries resulted from the bicyclist's impact with the bicycle he was riding. The finding that most injuries are caused by the bicyclist's impact with the roadway suggests that one potentially effective at-crash countermeasure may be training the bicyclist in how to abandon his bicycle or fall in order to minimize injuries.

Property Damage

In the study by Cross and Fisher, the bicyclists and motorists who were interviewed were asked to estimate the cost of repairing their vehicle to its pre-crash condition or, if damaged beyond repair, the replacement cost of the vehicle. For the fatal accidents, estimates of the cost of vehicle damage was obtained from traffic accident reports.

On the average, the cost of the combined damage to vehicles involved in a non-fatal accident was $120: $65 for the bicycle damage and $55 for the motor-vehicle damage. The extent of damage was considerably greater for fatal than for non-fatal accidents. The average cost of the damage to the two vehicles involved in fatal accidents was about $325; the average cost was $100 for the bicycle damage and $225 for the motor-vehicle damage. There were some instances in which the motor vehicle collided with another motor vehicle after having collided with the bicycle. However, no data were obtained on the cost of the damage to other objects or vehicles that were struck by the motor vehicle involved in the bicycle/motor-vehicle accident.

Other Undesirable Consequences

The undesirable consequences of bicycle/motor-vehicle accidents extend far beyond operator injuries and property damage. Whether or not an accident-involved operator is culpable, he or she is nearly always grief-stricken at the sight of another person who was injured in the accident. Culpable motor-vehicle operators are often involved in litigation which is both costly and emotionally painful. Relatives and friends of an injured operator also suffer. In addition to the emotional stress associated with an injured loved one, relatives and friends often suffer substantial losses of money and time in caring for the injured operator.

Bicycle/motor-vehicle accidents often are publicized widely -- particularly those that result in fatal injuries. It has been suggested that many of the persons who read or hear of bicycle/motor-vehicle accidents conclude that riding a bicycle is simply too dangerous. As a consequence, the occurrence of accidents serves to curtail bicycling in favor of driving. The resulting societal losses due to increased fuel consumption, pollution, and traffic congestion are as real and important as the losses due to injuries and property damage.

COST OF BICYCLE/MOTOR-VEHICLE ACCIDENTS

Implicit in every non-arbitrary decision about safety-education programs is a tradeoff of societal costs and societal benefits. In principle, the implementation of a safety-education program can be justified only if the societal benefits resulting from the program outweigh the cost of developing and implementing the program. Thus, if judicial decisions are to be made about safety-education programs to curtail bicycle/motor-vehicle accidents, it is necessary to consider the total societal costs associated with accidents of this type.

Table 7 lists estimates of the cost of societal losses resulting from fatal and non-fatal bicycle/motor-vehicle accidents that are reported to the police. Most of the cost estimates presented in Table 7 were derived from cost data contained in a recent report on the cost of motor-vehicle accidents (Faigin, 1976). Cost estimates for most losses resulting from traffic accidents differ as a function of the age and sex distributions of the accident population, the average severity of injuries sustained in the accident, and the types of vehicles involved. Therefore, these factors were taken into consideration when estimating the cost of losses resulting from bicycle/ motor-vehicle accidents. Information about the age, sex, and injury distributions were taken from the Cross and Fisher (1977) study of bicycle/motor-vehicle accidents. The data and assumptions underlying the cost estimates shown in Table 7 are described and discussed in Appendix A.


TABLE 7

COST OF SOCIETAL LOSSES RESULTING FROM FATAL AND NON-FATAL
BICYCLE/MOTOR-VEHICLE ACCIDENTS (POLICE REPORTED)

  AVERAGE
COST PER
ACCIDENT
(DOLLARS)
COST CATEGORY FATAL NON-
FATAL
MARKET AND MARKET-PROXY PRODUCTION LOSSES $171,817 $ 73
HOME, FAMILY, AND COMMUNITY SERVICES PRODUCTION LOSSES 48,281 21
EMERGENCY TRANSPORTATION 55 9
EMERGENCY ROOM TREATMENT 156 85
HOSPITAL CARE 1,030 201
PHYSICIAN'S CARE 328 66
FUNERAL COSTS 987 --
CORONER/MEDICAL EXAMINER 130 --
LOSSES TO OTHERS 3,832 138
LEGAL AND COURT COSTS 4,096 315
INSURANCE ADMINISTRATION COSTS 250 25
ACCIDENT INVESTIGATION COSTS 70 35
VEHICLE DAMAGE 325 130
TOTAL COSTS $231,357 $1,098

It can be seen in Table 7 that the average cost of the societal losses resulting from a fatal accident totals $231,357. Assuming 1,000 fatal accidents each year, the total cost of societal losses resulting from fatal accidents exceeds 231 million dollars. The average cost of a non-fatal accident is estimated at $1,098. Although the average cost of nonfatal accidents is far less than for fatal accidents, the total cost of the 40,000 police-reported accidents that occur each year is nearly 44 million dollars. Thus, according to these estimates, the combined cost of fatal and non-fatal accidents exceeds 275 million dollars each year; this estimate does not include the cost of unreported accidents.

No attempt was made to establish a monetary value for such loss as pain and suffering, grief, loss of personal relationships, and so on. Although emotional trauma represents a real and important societal loss, no satisfactory technique has been established for placing a monetary value on such losses.

OTHER BICYCLE-RELATED ACCIDENTS

There are many kinds of bicycle-related accidents other than bicycle/motor-vehicle accidents. Bicycles collide with other bicycles, with pedestrians, and with fixed objects. In addition, bicyclists lose control of their bicycles and fall for a great variety of reasons. For ease of exposition, the class of bicycle accidents that do not involve a collision with a motor vehicle will be referred to hereafter as non-motor-vehicle accidents and, for obvious reasons, will be abbreviated as "NMV accidents."

Although it is generally recognized that NMV accidents occur with far greater frequency than bicycle/motor-vehicle accidents, there is surprisingly little data on their incidence, consequences, and causes. This lack of information is the result of at least two factors. First, the limited resources available for research into bicycle accidents have been devoted mainly to the study of bicycle/motor-vehicle accidents rather than NMV accidents. Bicycle/motor-vehicle accidents have been given greater emphasis because, on the average, accidents that involve a motor vehicle result in more severe injuries than NMV accidents. Secondly, the study of NMV accidents is inherently difficult because such accidents are not routinely reported to any record-keeping agency. Medical records maintained by hospitals and physicians contain a great deal of useful information about NMV accidents, but the records are difficult to locate and even more difficult to obtain. Moreover, the study of medical records includes only the accident cases that resulted in injuries severe enough to require professional medical care. Perhaps the only way to obtain information on the full range of NMV accidents that occur is to conduct a comprehensive survey of the general population of bicyclists. Such a survey would be expensive and time-consuming, but is sorely needed.

ESTIMATE OF THE INCIDENCE OF NMV ACCIDENTS

Although there is much to be learned about NMV accidents, there are sufficient data available to enable one to confidently conclude that NMV accidents represent a severe problem in the United States. The best data on NMV accidents come from the National Electronic Injury Surveillance System (NEISS). This computerized system was developed by the Consumer Products Safety Commission to continuously monitor product-related injuries treated in the emergency rooms of a selected sample of 119 hospitals at diverse locations throughout the United States.

An analysis of NEISS data for calendar year 1975 revealed that 18% of all bicycle-related fatalities and 94.5% of all bicycle-related injuries were the result of NMV accidents. The remaining 82% of fatalities and 5.5% of injuries were the result of bicycle/ motor-vehicle accidents. Since the NEISS data include only the accidents that were treated in a hospital emergency room, it is necessary to somehow extrapolate these data in order to estimate the total number of NMV accidents that occur each year.

One extrapolation method involves the use of data on the annual number of bicycle/ motor-vehicle accidents as a basis for estimating the total number of NMV accidents that occur each year. This method assumes that the ratio of bicycle/motor-vehicle accidents versus NMV accidents found in the NEISS data is the same as the ratio for the total population of accidents -- whether or not they were treated in an emergency room. The assumptions are as follows: (a) bicycle/motor-vehicle accidents account for 82% of all bicycle-related fatalities, (b) bicycle/motor-vehicle accidents account for 5.5% of all bicycle-related injuries, and (c) about 1,000 fatalities and 80,000 injuries result from bicycle/motor-vehicle accidents each year. With these assumed values, it is possible to set up the following equations:

1,000 = .82 (x)
80,000 = .055 (y)
where: x = total number of fatalities
y = total number of disabling injuries

It is a simple matter to solve for the unknowns and arrive at an estimate of about 1,220 for total fatalities and 1,454,000 for total injuries. Subtracting the number of deaths and injuries resulting from bicycle/motor-vehicle accidents from these totals yields an estimate of 220 fatalities and 1,374,000 serious injuries as the annual toll for NMV accidents.

A different extrapolation approach has been used by the National Safety Council. The National Safety Council used the NEISS data and a set of prediction equations to estimate the total number of bicycle-related accidents that were treated in all the hospital emergency rooms in the nation during 1976. Then, the total number of bicycle-related accidents was computed with the assumption that accidents treated in emergency rooms account for 38% of all disabling injury accidents. This method yielded an estimate of 1,100 fatalities and 460,000 serious injuries. Subtracting the number of deaths and injuries resulting from bicycle/motor-vehicle accidents in 1976 from these totals yields an estimate of 200 fatalities and 420,000 serious injuries from NMV accidents.

These two approaches yield quite different estimates for both fatalities and disabling injuries. Although it is not known which extrapolation approach is best, it seems reasonable to estimate that NMV accidents account for no fewer than 100 fatalities and one-half million serious injuries each year.

TYPES OF NMV ACCIDENTS

Table 8 shows the relative frequency of three general types of NMV accidents: bicycle-bicycle accidents, bicycle-pedestrian accidents, and collisions with fixed objects or falling. The data shown in Table 8 are from four studies conducted in different geographical areas and covering different bicycling populations. The two studies by Barton-Aschman are probably the most representative because they sampled the full bicycling population within the sampling area. Although the size of the accident sample is small, the consistency in the percentage values shown for the two different states tend to support the reliability of the data. The study by Chlapecka et al. (1975) was limited to school-age children between the ages of seven and 13 years. The accident data refer to the "most serious" accident a child had experienced in the recent past. Chung's data come from a study of medical records for all accidents treated in the student health facility on the campus of the University of California, Santa Barbara (Chung, 1976). Chung examined records for all the accidents that occurred between September 1971 and March of 1976.


TABLE 8

RELATIVE FREQUENCY OF TYPES OF NMV ACCIDENTS

  TYPES OF NMV ACCIDENTS
DESCRIPTION OF SAMPLE AND SOURCE NUM-
BER
OF ACCI-
DENTS
IN SAM-
PLE
BICYCLE-
BICYCLE
BICYCLE-
PEDES-
TRIAN
COLLI-
SION
WITH
FIXED OBJECT
OR
FALLING
SURVEY OF GENERAL POPULATION IN THE STATE OF TENNESSEE (Barton-Aschman Associates, 1974b) 47 11%- 0% 89%
SURVEY OF GENERAL POPULATION IN THE STATE OF PENNSYLVANIA (Barton-Aschman Associates, 1975) 98 9% 1% 90%
SURVEY OF A SAMPLE OF GRADE-SCHOOL CHILDREN [AGES 7-13] IN 170 SCHOOLS IN 110 CITIES IN 37 STATES (Chlapecka et al., 1975) 5601 11% 1% 88%
ALL ACCIDENTS TREATED IN THE STUDENT HEALTH FACILITY AT THE UNIVERSITY OF CALIFORNIA, SANTA BARBARA, DURING THE PERIOD BETWEEN 1971 AND 1976 (Chung, 1976) 794 42% 6% 52%

It can be seen that the data on the survey of the general population and the survey of grade-school children are quite consistent. For these populations, it can be seen that:
  • Bicycle-bicycle accidents account for between nine percent and 11% of all NMV accidents.

  • Bicycle-pedestrian accidents account for no more than one percent of all NMV accidents.

  • Between 88% and 90% of all NMV accidents result from the bicyclist colliding with a fixed object or falling.

The accidents reported by Chung involved university students and occurred on a university campus that had an excellent system of bikeways during the entire reporting period. It can be seen that the relative frequency of bicycle-bicycle and bicycle-pedestrian accidents in Chung's sample was far greater than for the other three samples; the relative frequency of collisions with fixed objects and falling accidents was less.

Chung reports an NMV accident rate of 20.4 accidents per 1,000 students for the 74-75 school year. If this accident rate is representative of other college and university campuses throughout the nation, it can be estimated that about 136,000 college/university students per year are involved in an NMV accident that results in injuries severe enough to require professional medical treatment. This estimate does not include persons treated in medical facilities other than student health facilities and does not include accidents that result only in minor injuries and/or bicycle damage. Since student health facilities are not included in the sample of NEISS hospitals, the NEISS data may underestimate the total number of serious NMV accidents by as much as 27%.

The NMV accident problem on college and university campuses may be indicative of the problems that may arise in other areas if the volume of bicycle traffic continues to rise. It appears that college and university campuses would provide a fertile research environment for both bicycle-safety education specialists and the traffic engineers who are attempting to develop design standards for future bicycle facilities. Clearly, college and university students must be considered one of the most important target groups for bicycle-safety education programs.


[Home Page]
[Up: Table of Contents]
[Previous: Section II]
[Next: Section IV]

Contents copyright 1978,
AAA Safety and Educational Foundation
Republished with permission
Internet edition prepared by John S. Allen