United States Geeeral ,&counting Office GAO Report to the Chairman, Subcommittee on Investigations and Oversight, Committee on Public Works and Transportation, House of Representatives March 1990 HIGHWAY SAFETY Trends in Highway Fatalities 1975-87 ’ ’ GAO/PEMD-9wO United States General Accounting Office Washington, D.C. 20548 Program Evaluation and Methodology Division P-237223 March 9, 1990 The Honorable Glenn M. Anderson Chairman, Subcommittee on Investigations and Oversight Committee on Public Works and Transportation House of Representatives Dear Mr. Chairman: On October 11, 1988. the Subcommittee Chairman requested that we undertake a study of fatal traffic accidents in the IJnited States over a 1%year period for which there are data in the Fatal Accident Reporting System (Y.us). The request asked that we focus on motor vehicle safety policies as they relate to the vehicle, the driver, and the roadway envi- ronment in 1975 through 1987. The request also asked that we give par- ticular attention to several highway environment issues-namely, (1) narrow bridges, (2) operational deficiencies (for example, the absence of traffic controls)! (3) wet weather performance, (4) st,udded tires, (5) freeway signs and related highway geometry, and (6) roadside hazards. Most of the information in this letter and its appendixes is derived from data for 1975 through 1987 in the FARS data base. developed and main- taincd by the National Ilighway Traffic Safety Administration (NHTSA). I~~WSinchrdes data on about 4 1,000 fatal accidems per year involving about fiO.000 vehicles and about 110,000 persons who may be vehicle occupants, pedestrians, or cyclists. The National Safety Council reports that since 1948, there have been Background about lOO,OOOaccidental deaths per year and, on the average, almost half of these deaths resulted from motor vehicle accidents. Motor vehi- cle accidents are the leading cause of accidental death overall and the leading cause of accidental death for persons age 1 to 74. For persons 75 and older, motor vehicle accidents are the second leading cause of death from accidents. exceeded only by deaths resulting from falls, Deaths from motor vehicle accidents are a special problem for the youth of our cBountry. In 1984, almost three fourths of all accidental deaths for per- sons age 15-24 resulted from motor vehicle accidents, and these deaths accounted for almost 40 percent of all deaths for that age group. The National Safety Council has gathered statistics on deaths from motor vehicle accidents since 1913. Over this extended time, the number Page 1 GAO, PF.MD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 R-237223 compared to the low year of 1982-83 associated with the upturn in the overall trend. Driver fatalities account for more than half of all motor-vehicle-related fatalities, and male fatalities dominate the fatality statistics, whether viewed as simple counts or adjusted for exposure in terms of fatality rates per million population. We also compared fatalities to other cxpo- sure measures--such as miles driven, drivers, and registered vehicles The fact that these fat,ality rates have steadily declined suggests that, the apparent increase in various fatal accident statistics since 1983 is most likely a function of increased motor vehicle activity rather than a decline in general motor vehicle safety. However, more sophisticated analyses of disaggregattxd statistics, which we plan to undertake in sub- sequent work, may indicate that some types of vehicles are, in fact, unambiguously safer than others. The female fatality rate for I& and 17.year-olds has st.rongly influenced the overall rate for this age group since 1983. The fatality rate for females of this age group increased from a little over 170 per million population in 1982 to over 240 per million in 1987, an increase of about 40 percent. The 1987 rate was exceeded only by the 1980 rate, but there was not a great disparity between the two. The c>xpcrience for males of this age group is not nearly as dramatic. Table 1 highlights other general fatality statistics that, by 1987. showed inc.reases of 20 percent or more, either from t,hc 1975 base year or from the low year associated with the upturn in the overall trend in 1982 or 1983. Additional information on general fatal accident trends is containc‘tl in appendix II Page 3 QAO;PEMD-90-10 Highway Safety: Trends in Highway Fatalitirs 1975-X7 Table 2 highlights the driver-related statistics that, by 1987, showed increases of 20 percent or more, either from the 1975 base year or from the low year associated with the upturn in the overall trend in 1982 or 1983. See appendix III for an in-depth discussion of trends in driver- related statistics. Table 2: Highlights of Driver-Related Fatal Accident Statistic@ Percent increase in 1997 Variable --Over 1975 Over 198243 drover involvement rate By gender Female 26 09 22 23 By age group Age 16 17 27 40 t&e byage group 16~17 .20 55 Female bv aae -16-17 50 78 59 87 18-20 35 71 26 03 ~21~25 27 79 26 59 Over 65 47 20 31 81 Speed of vekes in mph 36-45 29 54 46-55 25 77 22 07 56-65 68 96 49 12 Over 65 24 42 Dwers not wng safety restraint 26 42 22 10 ‘Blank cells indicate that the rate of change did not exceed 20 percent The types of vehicles involved in fatal accidents have changed over the years. The number of small cars involved in fatal motor vehicle acci- dents has increased more than 100 percent from 1975 through 1987; the number of light trucks and vans in fatal accidents has increased more t,han 50 percent in the same time. The numbers of fatalities in these types of vehicles show similar increases. However, both the rate of vehi- cle involvement in fatal accidents per number of registered vehicles and the number of fatalities per number of registered vehicles are still gener- ally declining for these types of vehicles. Exceptions to the general decline are the rates for subcompact automobiles and conventional pickup trucks, which have been increasing since 1983. Even though acci- dent involvement rates and fatality rates for small automobiles have been declining, they arc still a matter of concern, since the rates are con- siderably higher than t,hose of larger automobiles. While medium and Page 5 (iA0 ‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-W B-237223 Table 3: Highlights of Vehicle-Related Fatal Accident Statistics’ Percent increase in 1987 Over 1975 Over 199243 Accident ~-~~ rwolvement by automoblle swe Mrislze 150.73 Subcompact 193 13 54 12 Compact 751 II 228.37 All small automoblles 257 15 69 57 IntermedIate automobIles 11672 Fatalltles by automobile SIE Mlnwze 13436 Subcompact ~~ 17061 46 30 Compact 672 64 222 94 All small automoblles 224 13 59 03 IntermedIate automoblles 10230 Accident involvement bv truck tvoe All light trucks 69 57 31 00 All trucks 51 58 23 15 Number of fatalltles by truck tbpe All light trucks 63 87 32 59 All trucks 47 05 25 85 Other vehicles Involved Motorcycles 24 41 Buses 22 49 Fatalltles r other vehicles Motorcycles .~ 26 40 Buses 28 57 Other 31 85 Deaths by rltlal Impact of accident vehicles NoncollIsIon 29 49 Side 20 70 Rearend 59 73 Other 46 90 Deaths bv DrinciDal imoact of accident vehicles NoncollIsIon 29 45 Rearend 65.23 41 97 'Blank cells indicate that the rate of change did not exceed 20 percent While the numbers of fatal accidents under various environmental con- ditions reflect the effects of those conditions within any particular year. in most cases we did not find that they caused patterns to deviate from the overall trend. Exceptions include some of the specific areas of the Subcommittee’s concern-namely, traffic controls, freeway accidents Page 7 GAO;PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 R-237223 Over the years, tires are being reported less and less as a contributing factor in fatal motor vehicle accidents. The number of vehicles in fatal accidents with tires as a contributing factor declined more than 40 per- cent from 1977 through 1987. The use of studded tires is not specifically reported in the FANS data base. In most years, more than 90 percent of freeway accidents occurred where no special signs or other traffic controls existed. While we could not obtain data indicating the relative mileage for freeway locations, with and without, signs, the increasing trend of freeway accidents is steeper when no freeway traffic controls exist. FAKSreports roadside hazards as a problem for very few fatal accidents, never totaling more than 400 a year, and the total number of such acci- dents decreased rather steadily from 1976 through 1981. As a conse- quence, the E:-\KSsystem stopped collecting special data on hazards after 1981. We provided drafts of this report to the National Center for Statistics Agency Comments and Analysis of NITSA and met with an official of the center to discuss the study results. He expressed general agreement with the study results, making a few editorial suggestions that improve the clarity of the presentation. We incorporated these suggestions in the report where appropriate. As agreed with your office, this report is being issued on an unrestricted basis. We are sending copies to the Administrator of the National High- way Traffic Safety Administration, to other organizations interested in highways and highwa\, safety issues, and to others upon request. Page 9 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Page 11 GAO,‘PEMI)-90-10 Highway Safety: Trends in Highway Fatalitirs .~75-U7 Appendix V 86 Statistics Related to Fatal Accidents by Type of Roadway 86 Roadway Conditions 87 the Driving Roadside and Traffic Conditions 88 Environment Fatal Accidents and Weather Conditions 93 Fatal Accidents by Day of t,he Week 93 Accidents by the Time of’ Day 95 Accidents by Season of the Year 96 Conchlsions 98 Appendix VI 99 Major Contributors to This Report Bibliography 100 - Tables Table 1: Highlights of General Fatal Accident Statistics 4 Table 2: Highlights of Driver-Related Fatal Accident 5 Statistics Table 3: Highlights of L’ehicle-Related Fatal Accident 7 Statistics Table 4: Highlights of Environment-Related Fatal 8 Accident Statistics Table 1.1: Relationship of Accidental Deaths to Total 18 Deaths in 1984 Table 1.2: Summary of Legislative Safety Concerns 20 Table II. 1: Fatal Accidents by Number of Fatalities 31 Involved Table 11.2:Fatalities by Role 32 Table 11.3:Fatal Accidents by Number of Vehicles 32 Involved Table 11.4:Vehicles Involved in One-Vehicle Fatal 33 Accidents Table 11.5:Fatalities by Age and Gender 36 Table 11.6: Pedestrian Fatalities by Age and Gender 44 Table III. 1: Driver Involvement in Fatal Accidents Per 49 Million Population in 1987 Table 111.2:Drivers Involved in Fatal Accidents by Age 53 and Gender Page 13 GAO, PEMD-9010 Highway Safety: Trends in Highway Fatalities 1975.87 Figure III. 10: Motor Vehicle Occupants in Fatal Accidents 59 Whose Reported lJsc of Restraints Is lJnknown Figure III.1 1: Motor Vehicle Occupants in Fatal Accidents 60 Reported Not Using Safety Restraints Figure III.12: Occupants Killed Who Were Not Using 61 Safety Restraints Figure III. 13: Occupants Killed Who Were ITsing Safety 62 Rest,raints Figure 111.14:Occupants Killed Whose LTseof Restraints 63 Was [inknown Figure IV. 1: Fatal Accident Rate by Size of Automobile 65 Figure IV.2: Fatal Accident Rate for Small Automobiles 66 Figure 1\‘.3: Fat,al Accident Rate by Type of Truck 67 Figure IV.4: Fatal Accident Rate by Type of Light Truck 68 Figure IV.5 Fatal A(,cident Rate for Medium and Heavy 69 Trucks Figure IV.6: Fatality Races by Size of Automobile 70 Figure IV.7: Fat,ality Rates for Small Automobiles 71 Figure IV.8: Truck Fatality Rates by Type of Truck 72 Figure IV.9: Overall Truck Fatality Rate 73 Figure IV.10: Fatality Rates for Light Trucks 73 Figure IV. 11: The A\~ragc~ Age of All Automobiles and of 74 Those in Fatal r\ccidents Figure IV. 12: The Av(xtg(l .4ge of Trucks and Those in 75 Fatal Accidents Figure IV. 13: Automobiles in Fatal Accidents by Age 76 Figure IV. 14: Trucks in Fatal Accidents by Age 77 Figure IV.15: Tires as iI Contributing Factor in Fatal 81 Accidents Figure IV. 16: Vehi&~ Fatalities by Direction of Initial 82 Impact Figure IV.17: Vehicle Fatalities by Direction of Principal 83 Impact Figure IV.18: Vehicle Fatalities With Principal Rearend 84 Impact Figure IV.19: Vehicle Fat,alities From Xoncollision 84 Accidents Figure V. 1: Fatal Accidents by Type of Roadway 87 Figure V.2: Fatal Arcident,s on Freeways 90 Figure V.3: Freeway Fatal Accidents With Some Traffic 90 Controls Figure V.4: Fatal Accidents Involving Roadside Hazards 91 Figure V.5: Fatal Acxidents Involving Bridges 92 Page 15 GAO, I’EMI)-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Page 17 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975%’ Appendix I Rackgmund The Department of Transportation Act (Public Law 89-670), dated Octo- Legislative History ber 15,1966, established the Department of Transportation (LXX) and gave to it the responsibilities under the National Traffic and Motor Vehi- cle Safety Act of 1966 and the Highway Safety Act of 1966. The High- way Safety Act of 1970 (Public Law 91-605) created the National Highway Traffic Safety Administration within the DOT and assigned to it the responsibilities for the National Traffic and Motor Vehicle Safety Act of 1966 and the portions of the Highway Safety Act of 1966 related to highway safety programs not otherwise assigned to the Federal High- way Administration. Since 1966, the Congress has passed several other laws that relate, either directly or indirectly, to highway safety. All the legislation related to highway safety addressed, to varying degrees, three basic areas related to highway safety-the motor vehicle, the vehicle driver, and the highway environment. Some of the specific con- cerns of this legislation are summarized in the table 1.2. Page 19 GAO :PEMD-YO-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix I Background This study would seem to call for directing more future safety efforts toward preparing better drivers and improving the driving environment. The study also reinforces a bifurcation in highway safety research between crash avoidance measures and occupant protection measures; the former focus on vehicle control while the latter focus on crash energy management. While the total number of highway deaths has been declining over the Objectives, Scope, and last 15 years, the aggregate statistic hides upward trends of selected Methodology components. The objectives of this report are to identify and describe the changing composition of the nation’s highway fatality toll. We focus on fatality trends over time and how these trends compare or contrast with safety policy as it. relates to the driver, the vehicle, and the road- way environment. WC describe only the trends that are derivable from the FAKS data base maintained by NIITSA and related measures of expo- sure to fatal accidents. ’ We do not attempt to explain causes for trends or to determine the interaction of various elements included in the FARS dat,a base. We also do not consider the effects of such developments as helicopter evacuation and hospital trauma units on fatalities. In developing this report, we used the annual computerized ~4~s data base maintained by NHTSA.We present results developed from the ~44~s data base for 1976 through 1987, using the three basic FAKS subfiles- the accident file, the person file, and the vehicle-driver file. These files include data on about 41,000 fatal accidents per year, about 60,000 vehicles per year involved in those accidents, and about 110,000 persons per year involved as vehicle occupants, as pedestrians, or in other roles. We recoded some of the data to meet our needs (for example, age catego- ries). and we recoded other data (for example, vehicle size) from input from NIITSA. Data to compute exposure rates came from various sources. Driver registration data were not readily available to satisfy the break- downs necessary for our analysis. Instead, we used population d&a, which we obtained from Bureau of the Census publications, that include only the IT.S. resident population. Vehicle age data came from Motor ‘The twm “exposure,” as used m traffic safety research, is the measure of th? total magnitude of vat-IOUS cate@xs of mteresf (for example. all dnvers by age and gender, or the age, type. and size of all registered vchrles) that vould be mvolred in fatal traffic accidents. Exposure rates cornpan’ the actual fatal traffic awldent StatiStics that occur to these various universe sizes Some common cxpo- sure meawres XP popnlat~~m driwl- registration data. and vehlclc registration data. Page 21 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix I Background vehicle or a nonmotorist-within 30 days of the accident. NHTSA adopted the 30.day requirement because studies show that 98 percent of all motor-vehicle-related fatalities occur within 30 days of the accident and because this allows expeditious reporting. Most other countries use the 30-day reporting period. FARS collects data at three levels: (1) the accident level, containing data on accident characteristics such as location, time, day of week, number of vehicles involved, and descriptions of the road conditions; (2) the vehicle-driver level, containing data on each vehicle and driver involved in the accident such as the vehicle’s description and how it was damaged and variables describing the driving history of the drivers involved; and (3) the person level. containing data on each person involved in the acci- dent, such as age, degree of injury, use of safety restraints, alcohol involvement, and role (driver. passenger, pedestrian, and so on). FARS data are collected by state employees. KHTSA has contracts with all 50 states, Puerto Rico. and the District of Columbia to provide the neces- sary information. KHTS.~furnishes standardized data collection instru- ments, and state IQWSanalysts use sources such as state vehicle registrations, driver licensing and highway department files, and vital statistics and death ct,rtificates to gather the necessary information. As state FAKS analysts CWer the data into NHTSA'S computerized central data file, the data are aut,omatically checked on-line for range and consis- tency as part of P,alisquality control. While we have identified numerous citations of automobile safety The Contribution of research-including numerous studies performed by NHTSA using FAHS- This Report very little of thal rcsc~ch discussed the changes in the characteristics of fatal accident statistics over time. Moreover, while studies on specific fatal accident characteristics--such as trucks versus cars or male versus female drivers-have, been performed by others, these studies have tended toward a narrow focus. In addition, because of the difficulty in obtaining accident t’xposure information, little information is available t,hat compares accidc,nt fatality statistics to various measures of acci- dent exposure such iIs vehicle miles traveled, number of registered vehi- cles, or number ol‘ drlvcrs. This report attempts to fill some of these gaps by presenting int’ormat,ion that is (1) trend-based, (2) extensive in accident charactt%tics discussed, and (3) related, where possible, to measures of cxposut’(’ to fatal accidents. Page 23 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-W Appendix II General Fatal Accident Statistics Figure 11.1: Number of Accidents, Vehicles, and Deaths in Fatal Accidents 70000 Number ,976 1976 19n 1976 1979 1980 1961 1982 1983 1994 1966 1966 1967 - Number of Accidents ---- Number of Vehicles m Number of Deaths The comparisons of the three trends indicate the obvious-on the one hand, fatalities do not occur in all vehicles involved in fatal accidents and, on the other hand, some fatal accidents have more than one fatal- ity. The comparisons would also seem to indicate that despite a gener- ally better record than in the late 1970’s, the occurrence of fatal accidents and related fatalities appears to be on the rise again. The apparent increase since 1983 is somewhat tempered when increase in exposure to motor vehicle accidents is considered. Since our analyses showed that the overall trend tended to be predominant, the following sections of this report concentrate on patterns that depart from the overall trend. To determine how much of the change in numbers of accidents is simply Fatal Accident Rate a reflection of a larger number of motor vehicles being on the road, we Trends compared the numbtars of fatal motor vehicle accidents and fatalities to three generally acccptcd units of exposure to such accidents-namely, miles driven, the number of registered vehicles, and the number of regis- tered drivers. ThtstL comparisons show mixed results. While the fatal Page 25 GAOIPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix II General Fatal Accident Statistics Figure 11.3: Fatality Rate Trends 4.0 Rate 2.5 - Rate per 10,ooO Vehides ---- Rateper10,oooDrivers m Rate per 100 Million Miles Driven Since we were not able to obtain data to group the nation’s drivers by Fatality Rates Per age and gender, we focused on the fatality rates per one million popula- Million Population tion to display any differences. The pattern of this fatality rate trend is basically the same as that of the overall trend. (See figure 11.4.) F’atali- ties reached a high of almost 230 per million population in 1979, fell to a low of a little over 180 per million population in 1983, and increased to a rate of about 190 per million population in 1987. The fatality rate for males has been two and a half to three times as large as that for females. (See figure 11.5.) In recent years, the fatality rate for females has shown a higher rate of growth than that of males. Despite this more rapid growth, however, the fatality rate for females was still less than half the rate for males in 1987. Since 1983, the overall fatality rate has increased about 6 pcrc,cnt; the rate for males, however, has increased less than 3 percent white the rate for females has increased 10 percent. Page 27 tiAO/PEMIXlO-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Stat istics Figure 11.5: Fatalities Per Million Population by Gender 360 Fatally Rale - Male ---I Female There is also a wide divergence between age groups in fatality rates; however, fatality rate trends for all age groups tend to follow the over- all trend.’ The lowest fatality rate is that for people younger than 16-a rate that has never been over 100 per million population and that has decreased more than 20 percent since 1975. The highest fatality rate has consistently been that for ages 18 through 20, ranging from 414 to 542 per million population. The only aspect of the trends themselves that appear worthy of special comment is the er;pcWmce of 16. and 17.year-olds and those over 65 since 1983. Since 1983, the fatality rate for 16. and 17-year-olds has increased from 30 1 per million population to 352 per million population, an increase of about 17 prrcent. At the same time, the rate for those Page 29 (GAO ‘PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975.87 -- Appendix II Genrral Fatal Accident Statistics accidents had only one fatality, while accidents with two and three fatalities accounted for about 8 and 1 percent of the accidents. Less than 1 percent of the accidents involved four or more fatalities. (See table 11.1.) Table 11.1: Fatal Accidents by Number of Fatalities Involved Number of fatalities 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 Olle 35,019 35,451 37,819 39,870 40,608 40,747 39,853 35,356 34.382 36,000 35,562 37,207 37,526 TWO 3,260 3,323 3,498 3,620 3,708 3,638 3,315 2980 2,888 2.928 2,927 3,127 3171 Three 633 668 637 657 667 654 620 530 499 504 513 525 534 Four - 166 202 180 207 176 174 137 146 147 143 132 161 136 FIW -~ 44 61 -48 58 37 44 53 48 37 34 36 47 45 More thank 30 41 -29 20 27 27 21 32 23 22 25 23 23 ~~ ~__ ___.~~ Total accidentsa 39,161 39,747 42,211 44,433 45,223 45,284 44,000 39,092 37,976 39,631 39,196 41,090 41,436 *Even though FARS IS supposed to include only fatal accidents the data flies do Include a few acc~ dents for which zero fatalltles wre recorded Therefore, these totals columns do not all add uu Accidents with one. two, and three fatalities follow the overall trend. Fatal accidents with four or more fatalities show very erratic patterns and are generally small in number, the highest being 208 for accidents with four fatalities. Who gets killed in fatal motor vehicle accidents? Have the trends in Fatalities by Person’s fatalities differed for various roles (drivers, passengers, pedestrians, Role and others)‘? Drivers constitute the majority of motor vehicle fatalities (about 58 percent in 1987) followed by motor vehicle passengers (about 25 percent), pedestrians (about 15 percent), and others (about 2 per- cent). (See table 11.2.) Fatalities among drivers follow the overall trend. Passenger fatalities show a similar trend, although it is not as pro- nounced. Pedestrian fatalities show trends different from either driver or passenger fatalities. 0ther fatalities consist mostly of pedalcyclists and fewer than 100 other nonoccupant fatalities per year. Page RI GAO ‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 -~ -- Appendix II General Fatal Accident Statistics vehicles increased so much from 1983 onward that 1987 surpassed all previous years. However, since fatal accidents involving three or more vehicles are not large in number, the peak of 2,367 in 1987 is only about 360 more fatal accident,s than the previous peak of 1978. One-Vehicle Accidents Since one-vehicle fatal accidents arc by far the most frequent, we show the composition of such accidents in detail. Table II.4 quantifies the extent to which particular kinds of vehicles are involved in one-vehicle accidents. Table 11.4:Vehicles Involved in One-Vehicle Fatal Accidents Type of vehicle 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 AutomobIle Small 2,235 2,468 2,755 3,648 3.612 4,409 4.729 4,644 4,932 5,581- 5,817 6,968 7,196 lntermedlate 1,274 1395 1,401 1,710 2,173 2.509 2,497 2,228 2,149 2.312 2 105 2.185 2,154 Full~slzed 7,646 7,691 7,737 8,043 7,956 7.940 7.150 6 I%0 5.475 5,138- -i,313 4,411 3,887 S~zeunknown 5,505 4,948 4549 ~.. . 115 1,966 2,153 1,816 1,588 1,490 1.248 1,391 Total 16,660 16,502 16,442 16,667 16,747 17,273 16.342 ~~15.045 14.372 14.619 13.725 14.812 14.628 Trucks Van~based light 566 518 644 783 886 91 1 837 740 669 684 695 ~~705 845 Conventional ~- light 3,223 3,639 3,925 4,389 4,646 5,110 4769 4,077 3,993 4 203 4,320 4,733 5,039 MedIumand heavy 1,163 1,350 1,443 1,573 1,574 1,423 1.336 1,081 1,205 1,259 1.160 1,162 1,084 Total --4,952 5,507 6,012 6,745-. 7,706 7,444 6,942 5,898 5,867 6,146 6,175 6,600 6,968 Motorcycles 1,276 1,357 1,721 1,876 2,047 2,238 2102 1,965 1,927 2,052 2,049 2,041 1,751 Buses 158 142 150 147 153 156 155 118 130- 125 124 115 110 Other vehicles 605-- i92 565 333 321 313 357 825-~ -752 755 802 707 702 Total vehicles 23,651 24,100 24,890 i5,768- 26,374 27,424 25,898 23,851 23,048 23,69?- 22,875 24,275 24,156 Automobiles and Trucks As might be expect ~1, automobiles are by far the most likely vehicles to be involved in one-vehicle fatal accidents. More automobiles have been involved in one-vehicle fatal accidents than all other vehicle types. How- ever, the differencca in such accidents between automobiles and trucks has been narrowing somewhat in recent years. In 1976, automobiles (about 70 percent of the total) were involved in about three and a half times as many one-vehicle fatal accidents as trucks (about 20 percent); by 1987, automobiles (about 60 percent of the total vehicles) were less involved in such accidents, and trucks (almost 30 percent) were much more involved. so that the number of automobiles involved in one-vehi- cle accidents was only about twice the number of trucks. Page 33 GAO ‘PEMD-90-10 Highway Safety: Tlpnds in Highway Fatalities 197587 Appendix II General Fatal Accident Statistics Figure 11.6:One-Vehicle Fatal Automobile Accidents by Automobile Size 20000 Number of Accidenlo UnknownSize Intermediate Buses One-vehicle fatal bus accidents have never been large in number; there were fewer than 160 in the peak year of 1975. From 1975 through 1981, fatal one-vehicle bus accidents varied up and down within a range of 20 of the 1975 peak year and then dropped almost 25 percent in 1982. After a slight imrease in 1983 one-vehicle fatal bus accidents fell to 110 in 1987, the lowest total on record. In 1987, buses accounted for less than 1 percent of the vehicles involved in onc-vehicle fatal accidents. Motorcycles One-vehicle fatal motorcycle accidents show a trend different from any other type of one-vehicle accident. The first year of our study, 1975, had the fewest one-vehicle fatal motorcycle accidents. One-vehicle motorcy- cle accidents increased dramatically, however, from 1975 through 1980, increasing about 75 percent, With the exception of 1984, however, one- vehicle fatal motorcycle accidents have been on the decline since 1980, but the number is still considerably higher than in 1975. In 1987, motorcycles accounted for about 7 percent of the vehicles involved in one-vehicle fatal accidents Page 35 GAO PEMD-‘JO-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Statistics Fatalities by Gender Clearly more males than females are killed in motor vehicle accidents. In fact, almost three times as many males as females die in motor vehicle accidents each year. The principal difference in trends is that, while trends for both are similar to the overall trend, the number of female fatalities has increased much more rapidly than the number of male fatalities since 1983; 1987 was the peak year for female fatalities. By 1987, females accounted for about 30 percent of motor vehicle fatalities while males accounted for about 70 percent. Fatalities by Age Just as there were differences in the fatality trends for the genders irre- spective of age, there were’ some differences in the trends by age irre- spective of gender. The 16-17, 21-25, and 51-65 age groups tend to follow the overall trend. The other age groups are worthy of some dis- cussion, however, because of certain departures from the overall trend. The under-16 age group did not follow the overall trend at all until 1983. Fatalities for this age group continually declined from 1975 through 1983 and then increased slightly through 1987. The 1987 level of fatali- ties is still about 26 percent below the peak level of 1975, however. The 18-20 group, while showing patterns similar to the overall trend in the early years, has not shown the general tendency to increase since 1983. Fatalities for the 2B-50 group tended to follow the overall trend through 1983. However, the decline from the peak year in 1981 was not as dra- matic as in other cases. and the number of fatalities for this age group in 1987-the peak year-is almost 30-percent higher than in 1975. Fatali- ties for the over-65 group show perhaps the most disturbing pattern. After showing slight declines in fatalities from 1975 through 1981, fatalities for this age group dropped about 5 percent in 1982. IJnfortu- nately, since 1982, fatalities for this age group have increased more than 20 percent. reaching a peak in 1987. Fatalities by Age and It is interesting to SW how age and gender, considered together, illus- Gender trate departures from the overall trend while also pointing out any dif- ferences between male and female for each age group. The various age groups show the following differences, either from the overall trend or between genders. The group 16-17 shows a greater percentage increase in female fatalities in recent years t.han male fatalities. (Sec.figure 11.7.) Page 37 GAO;PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix II Gwwral Fatal Accident Stalktics Figure 11.8: Fatalities for Ages 18-20 by Gender” 24 Percent Deviation From Xi-Year Average 20 16 12 6 4 0 4 .6 -12 -16 -20 -24 -Percentages are normallred as percentage devlatlons from the 13 year average of fatalltles The group 21-25 shows significant trend differences since the 1980 peak year. The percentage decline in female fatalities from 1980 through 1987 was only about half the percentage decline in male fatalities for the same period. Moreover, while male fatalities in 1987 for this age group were only about 100 more than the previous low year of 1975, female fatalities for 1987 were still more than 25 percent higher than thr 1975 low. (SW f’igllrc 11.9.) Page 39 GAO PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix II General Fatal Accident Statistics Figure 11.10: Fatalities for Ages 26-50 by Gender” 20 Percent Deviation Fmm 13-Year Average 16 6 4 0 -4 a -12 -16 -20 ‘Percentages are normallred as percentage dewatlons from the 13.year average of annual fatalltles The 51-65 group shows differences in fatality trends between males and females in recent years, after somewhat similar experiences in the early years. (See figure II. 11.) Page 41 GAO/PEMD-SO.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix Il General Fatal Accident Statistics Figure 11.12: Fatalities for Ages 65 and Older by Gender” 21 Percent DavYIon From W-Year Avenge 16 15 12 9 6 3 0 4 4 a -12 - Female dPercentages are normallred ds percentage dwallons from the 13.year average of annual fatalkes Pedestrian fatalities show trends different from either driver or passen- Pedestrian Fatalities ger fatalities. As with fatalities in general, we look at (1) fatalities by gender irrespective of age, (2) fatalities by age irrespective of gender, and (3) fatalities by age and gender taken together. Table II.6 shows the distribution of pedestrian fatalities. Page 43 GAO ‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 -. Appendix I1 C;meral Fatal Accident Statistics those older than age 50 have been generally declining, 1987 fatalities being about 15 percent less than in 1975. Fatalities by Age and There is very little difference in trends for pedestrian fatalities between males and females under age 21. Pedestrian fatalities for both have been Gender declining rather steadily, and the number in 1987 is only about 60 per- cent of the 1975 pcbaktotal for both. However, fatalities for pedestrians age 21 through 50 showed steady increases for both genders through 1980. Thereafter, female fatalities in this age group tended to follow the overall trend while male fatalities tended to decline. For the 50 and older age group, fatalities for both males and females have declined since 1975. However, while fatalities for females tend to follow the overall trend, fatalities for males again have generally continued to decline. The overall trend-increases from 1975 through 1980, decreases Conclusions through 1983, and then increases through 1987-applies to many, but not all, of the general fatality statistics discussed. Drivers are the greater part of motor vehicle fatalities, and male fatalities dominate the fatality statistics, whether viewed as simple counts or as fatality rates per million population. Relating fatalities to other exposure measures such as miles driven and numbers of drivers and registered vehicles sug- gests that not all but much of the apparent increase in various fatal acci- dent statistics since 1983 is a function of increased motor vehicle activity rather than a decline in general motor vehicle safety. However, more sophisticated analyses of disaggregated statistics that we plan to use in subsequent reports may indicate that some types of vehicles are safer than others. Page 45 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 - - Appendix III Driver-Related Statistics Figure 111.1:Driver Fatal Accident Rates by Gender 700 Rate per Million Population 666 600 1 400 300 zoo ,w _11~1~11---------------------------------~~~~~~~~~ l_lll______l___l_____--I----I- 0 1976 1976 19n 1976 1979 1960 1961 1962 1963 1964 1966 1966 1967 Page 47 GAO.‘PEMD~YO-10 Highway Safety: Trends in Highway Fatalities 19’75-87 Appendix III Driver-Related Statistics Table 111.1:Driver Involvement in Fatal Accidents Per Million Population in 1987 Age group Male Female Overall 16-17 622 243 437 18~20 922 266 596 21-25 847 218 533 26-50 501 135 317 51-65 319 89 198 Old&r-thi;65 ~-~ 300 83 170 Overall 468 127 296 Figure 111.3:Fatal Accident Rate for Drivers 16 and 17 666 Rate per Ylllion Population 476 425 375 325 1 1976 1976 1977 1976 1979 1980 1961 1962 1983 1964 19SJ 1986 1967 Page 49 GAO/PEMD-90-10 Highway Safety: Trends in Hiway Fatalities 1975-87 Appendix III Driver-Related Statistics Figure 111.5:Male Driver Fatal Accident Rate by Age Group 800 Rate per Million Population 750 --------0-9 ---.--. 600 1975 197s 1977 1979 1979 1960 1981 1962 1963 1964 1965 1966 1987 - Under 21 -9-9 Age2160 - Age51-65 . . . . OlderThan Page 51 GAOIPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix III Driver-Related Statistics Table 111.2:Drivers Involved in Fatal Accidents by Age and Gender Gender Age group 1975 1976 1977 1976 1979 1960 1981 1982 1983 1984 1985 1986 1987 Male Under16 444 449 481 460 460 424 388 332 328 366 384 382 365 16~17 2,922 2,985 3,222 3,300 3,140 2,%- 2,601 2,092 1,967 2,059 2,037 2,422 ~2.426 18-20 6,858 6,996 7,580 8,049 7,980 7,595 6,824 5,852 5,348 5.655 5,227 5,456 5;i31 21~25 8,949 9,023 9,949 10 565 10,976 lo@ 10,469 9,112 8,429 8,859 8,882 9,051 8,655 26-50 18,341 18,008 19,566 21,174 21,784 21,531 21,731 19,404 19,175 19,955 20,293 21,218 21,822 51-65 5,289 5,368 5,456 5,705 5,527 5,428 5,430 4,744 4,622 4,805 4822 4,697 4361 Older than 65 2,728 2,711 2,775 2,872 2,801 2,701 2,783 2,673 2,788 2,880 3,029 3,239 3,336 Unknown 120 93 105 110 115 y-1 8 110 161 155 144 172 188 186 Total 45.651 45.633 49.134 52.235 52.783 51.463 50.336 44.370 42.812 44.723 44.846 46.653 46.882 Female Under16 74 81 ~- 102 102 84 98 93 80 88 80 95 122 105 16-17 676 732 SOS 793 851 801 708 578 660 687 753 854 900 18-20 1,221 1.407 1.534 1,619 1,526 l-i67 1,502 1,336 1,358 1,402 1,368 1431 1.454 21-25 i,6<2 1,694 -I,%?0 2.058 2,021 2,028 2,008 1,856 1,886 2.020 2,135 2,131 cil8 26-50 3,855 3,941 4,226 4,427 4.700 4.792 4,801 4,482 4,605 5,117 5,131 5,420 6002 - 51-65 ~ --- 1,301 ~~ 1,365 ~~ ~~ 1,501 ~~__ 1,436 ~.~ _ 1,399 1:438- 1,482 1,360 1,383 1.473 1,502 1,472 1,534 Older than 65 701 814 788 888 809 827 895 967 962 1107 1,142 1,289 1373 Unknown 27 10 15 15 17 15 20 16 16 21 16 19 16 Total 9,457 10,044 10,892 11,338 11,409 11,466 11,509 10,675 10,958 11,907 12,142 12,744 13,604 Unknown 34 20 23 27 39 28 309 984 886 882 895 938 948 Total drivers 55,142 55,697 60,049 63,600 64,231 62,957 62,154 56,029 54,656 57,512 57,883 60,335 61,434 For drivers 16 and 17, the number of male drivers again tends to follow the overall trend, while the number of female drivers shows substantial swings in the trend. The involvement of male drivers 16 and 17 years old reached a peak in 1978 but then decreased about 40 percent through 1983. Even though the number of male drivers of this age involved in fatal accidents increased from 1983 through 1987, the number in 1987 was still about. 25 percent less than the peak year of 1978. Female driv- ers in this age group involved in fatal accidents increased about 26 pcr- cent from 1975 through 1979 and then fell dramatically by about 32 percent to a low in 1982. From 1982 through 1987, however, female drivers of this age in fatal accidents increased about 56 percent, to reach the peak of 900 in 1987. While the number of male drivers 18 through 20 followed the overall trend through 1983. this group did not show the increase since 1983 that is characteristic, of the overall trend. While there were both Page 53 GAO/PEMD-99.10 Highway Safety: Trends in Highway Fatalities 1975437 Appendix III Driver-Related Statistics vehicles involved in fatal accidents was reported as unknown. Keverthe- less, the analysis of known speeds yields some interesting insights. For vehicles for which speed is known, the 36 to 55 miles per hour range is the most frequent speed encountered. (See figure 111.7.)Further break- down of this speed bracket shows that vehicles with speeds of 46-55 mph account for about 60 percent of the vehicles in this bracket. These breakdowns also show that vehicles with speeds of less than 55 mph are more involved in fatal accidents than vehicles with higher speeds. IIow- ever, since data on speed are available for only about 40 to 45 percent of the vehicles in fatal accidents, the issue of the relationship of speed to fatal accidents still necsdsto be investigated further. Figure 111.7:Speed of Vehicles in Fatal Accident9 14ooo Number of Vehicles 2000 1975 1976 l9i7 1976 1979 1990 1991 1962 1963 1964 1965 1966 1967 - 35 mph or less -1-1 36-55mph - 56-65mph n m.. Morethan65mph ‘Data on speed are unavaiwat,lt for 1980 and 1981 The prcsenrc of drinking drivers has been a matter of concern in high- Drinking Drivers way safety for many years. ITnfortunately, not all states have been dili- gent in determining whether drivers in fatal accidents had been Page 6.5 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities lY75-87 Appendix III Drivrr-Rrlated St&i&c’s Figure 111.8: Drinking Drivers Reported in FARS Compared to Single-Vehicle Nighttime Accident9 45 Percent of Total Aocldents 40 c-------c 35 . r* z 30 I .O r' 26 : 2 20 -11----2 15 1975 1976 1977 1973 1979 1980 1991 1982 1953 1964 1965 1966 1967 - Single Vehicle Nighttime Surrogate I--- FARS Drinking Percent 'For lhls analysts night time IS 6 00 p m to 6 00 am Perhaps the most interesting insight about the use of safety restraints is The Use of Safety the relatively small. though increasing, percentage of drivers and pas- Restraints sengers in fatal accidents who use them and the increasing ability of accident investigators to determine whether they were used, as evi- denced by the continuing decline in the percentage of both drivers and passengers whose IW of them was labeled unknown. (See figures III.9 and III. 10.) Since both the percentage not using safety restraints and the percentage whose usage is unknown have been declining at the same time in recent years, the increasing percentages shown for the use of safety restraints sincr 1981 are, indeed, real increases. Pagr.57 GAO ‘PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975.97 Appendix III Driver-Related Statistics Figure 111.10:Motor Vehicle Occupants in Fatal Accidents Whose Reported Use of Restraints Is Unknown 35 Perem of occupants 1975 1975 1977 1979 1979 1990 1981 1932 1983 1994 1985 1986 1987 - Drivers ---- Passengers For both drivers and passengers, the percentage reported not using safety restraints continued to rise from 1975 until 1981, but both have been on the decline since then. (See figure III. 11.) Nonuse by drivers reached a high of about 73 percent in 1981 and fell after that to about 54 percent in 1987. Nonuse by passengers fell from about 78 percent to about 64 percent in the same period. The reported use of safety restraints increased from about 6 to about 30 percent over this period, while the reported use for passengers increased from about 3 to about 23 percent. Over this same period, the percentage whose use was unknown fell from about 22 to about 17 for drivers and from about 19 to about 13 for passengers. The experience of recent years. therefore, shows either an increased interest by the driving public in prot,ecting themselves in motor vehicle accidents or the effectiveness of recently enacted mandatory st’at belt laws, or perhaps both, Page 59 GAO /PEMD-SO-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix 111 Drkrr-Related Statistic!. Figure 111.12:Occupants Killed Who Were Not Using Safety Restraints 60 Percent Killed When Restraints Were Not Used - Driver - - -- Passenger Pagr 61 GAO PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975.87 - Appendix III Driver-Related Statistics Figure 111.14:Occupants Killed Whose Use of Restraints Was Unknown 45 Percent Killed Whore Restraint Usage Was Unknown Increased use of motor vehicle safety restraints since 1979 or 1980 appears to have sawd the lives of many mot,or vehicle drivers and pas- sengers. More and mow wcupants are reported using safety restraints. Moreover. the fatality trend for occupants who arc reported as using restraints has been steadily declining whik t,hc trend for those reported as not using restraint s has been steadily increasing. While high driving speeds arc likely to bc ;I problem. FAKSdata art’ so limited that they arc’ of litt,le help in firmly c,stablishing trends rclatcd to speed. Page 63 GAO PF:M1XNlO Highway Safet.y: Trends in Highway Fatalitirs 1975.87 Appendix lV Vehicle-Related Statistics Figure IV.l: Fatal Accident Rate by Size of Automobilea 500 Rate per Million Automobiles 250 2m 1978 1979 1980 1981 1992 1993 1994 1985 1985 - Small ---- Intermediate m Full ‘The number of registered dulomoblles by we was not wallable for years prior to 1978 The involvement rate for small automobiles is also not as clear as figure IV.1 might indicate. Disaggregating the total shows not only differing trends for the different sizes of small automobiles but also very differ- ent fatal accident involvement rates. (See figure IV.2.) Even though it is on the decline, the involvement rate for minisize automobiles is still the highest by far. Interestingly, the rate for compact automobiles has con- sistently been higher than that for subcompacts. However, the rate for subcompacts has increased substantially since 1983. The rate for com- pact automobiles, however, after increasing in the early years, declined rapidly after 19Sl and has been rather steady since 1983. Page 65 GAO/PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-X7 Appendix IV Vehicle-Related Statistic?, Figure IV.3: Fatal Accident Rate by Type of Trucka 600 Involwmem Rate per Million Trucks 400 3M) -----1-....- -------....ll ---........... 200 -------..........l im \ 1990 1961 1982 1963 1984 1965 1966 - All Trucks ---- Van-basedTrucks m Cnnvenhnal Pickups n nnn Medium and Heavy Trucks The number of registered lnrkb by see was not available for years prior to 1979 Page 67 GAWPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-W Appendix IV Vehicle-Rrlatfxl Statistics Figure IV.5 Fatal Accident Rate for Medium and Heavy Truck?? 700 Invo(vsnnti Rate per Mllllon Trucks 600 1980 1981 1992 1993 1984 1985 1966 ‘The number of reQlSiered trucks by type was not wallabe for years prior to 1979 We also analyzed f&al motor vehicle accidents to find the differences, if Fatality Rate by Type any, in the rates of fatalities by type or size of vehicle and in the trends and Size of Vehicle for these fatalities. Automobiles Generally, the relationships for automobile occupant fatalities are simi- lar to the fatal accident involvement rates. (See figure IV.6.) In 1975, the fatality rate for intermediate automobiles was the highest of all automo- biles, but by 1986 it was the lowest. The rate for small automobiles has been consistently high, while the rate for full-size automobiles was the lowest in all years except 1986. The rate for full-size automobiles had been declining but, in 1986 it increased almost 14 percent. The fatality rate for the occupanm of intermediate automobiles has been steadily declining, having decreased about 50 percent from 1978 through 1986. The fatality rate for minisize automobiles was higher than the rate for other small cars, the rate for subcompact cars being the lowest. (See fig- ure IV.7.) The fatality rates for the various types of small automobiles show the same trends as the fatal accident involvement rates for those automobiles. Page 69 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 197587 Appendix IV Vehicle-Related Statistics Figure IV.7: Fatality Rates for Small Automobilesa 600 Faalnles per Milllon Automobiles 450 400 360 200 160 1978 1979 1960 1961 1962 1963 1964 1966 1986 - All Small ---- Minisize m Subcompact nnnn Compact “The number of registered autamob!les was not available by size for years prior to 1978 Trucks Analysis of fatality ratths for truck occupants shows how important con- ventional pickup trucks arc in the overall fatality rate for trucks. (See figure IV.8.)2 The fatality rate for conventional pickup trucks tends to raise the overall trurk rate as well as the rate for all light trucks. The rates for van-based light trucks and medium and heavy trucks are small by comparison. The fatality rate for all truck occupants irrespective of siztl or type of truck has been declining rather steadily since 1978; it has dcclincd over 30 percent since that year. (See figure IV.9.) The number of c*onventional pickup t,ruc:ks has cxert,cd a substantial influence on the light truck fatality ratc‘ and has shown a general pattern of increase since 1983. While the fatality rate for occupants of van-based light truc*ks has declined stc,adily since 1980, decreasing about 45 percent, the rat t’ for oc*cupants of cclnvcntional pickups declined only through Page 71 GAO ‘PEMD-WI0 Highway Safety: Trends in Highway Fatalities IW’R-X7 Appendix IV Vehicle-Related Statistics Figure IV.9: Overall Truck Fatality Rate 280 Fstality Rate per Million Trucks 270 260 im Figure IV.10: Fatality Rates for Light Trucks 350 Fatality Rate per Million Trucks 300 250 2M) 150 -----1-...- -0.. -. -9.. --0.. 100 --.....-----..~~~ --.......----= 50 1980 19sl 1992 1983 1994 1995 1995 - All Light Trucks - - - - Van-based Trucks m Conventional Pi&ups Page 73 GAO, PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1976.87 Appendix IV Vehicle-Related Statistics Figure IV.12: The Average Age of Trucks and Those in Fatal Accidents 9 Average Aqs in Years - Trucks in Accidents ---- Trudtson the Roads Another difference between automobiles and trucks is that the accident rates per 100,000 registered vehicles for various age groupings of aut,o- mobiles have tended to converge over the years toward a similar acci- dent rate while the rates for trucks have tended to maintain differing rates for different ages. (See figures IV.13 and V.14.) Moreover, trucks less than 5 years old are more involved in fatal accidents than any other age group for trucks while automobiles 11 to 15 years old tend to have t,hc highest involvemc,nt rate. In both instances, however, the rate of involvement has bc~n 5tcadily declining sixe 1978. Some of the trend dlrcctions for vehicles of various age can be seen when WCanalyze each age group of vehicles separately. The accident rates per 100,000 rcgistrred vehicles for automobiles 15 years old and less have been generally declining since about 1978 and the involvement rates for automobiles 5 years old and less, 6 to 10 years old, and 11 to 15 years old were all beltcvcen 30 and 33 per 100,000 registered automobiles in 1987. The rate per 100.000 registered automobiles more than 15 years old was about 27 in 1087. close to the rate for automobiles of other ages. The fatal-accidt~nt-in~( Ilvement rates for trucks are different from those of automobiles. Similat, to automobiles, the rate for trucaks 10 years old Page 75 GAO PF:MD-90-10 Highway Safety: Trends in Highway Fatalities 1975.R7 Appendix IV Vrhiclr-Related Statistics Figure IV.14 Trucks in Fatal Accidents by Age 90 Accidents per 100,000 Registered Trucks m ,,,,......====.....,......=~~8 l mmm •,~mmmmmmmmm~ l -m-m................ ¤mmmmmmmmmmmm~mmmm 20 1975 1976 19i7 1979 1979 1980 1981 1992 1933 1994 1995 1986 1987 - 5 years and under ---- 6loyears - II-15years . . . . 16yearsandolder Automobiles are clearly more involved than trucks in fatal accidents. The Types of Vehicles Two to three times as many automobiles are involved. Other types of Involved in Fatal vehicles such as buses and motorcycles are involved even less often than Accidents trucks. (See table II’. 1.) Page77 GAO ‘PKMD-90-10 Highway Safety: Trends in Highway Fatalities 197587 Appendix N Vehicle-Related Statistics accidents has increased almost ‘70 percent since 1975. The number of medium and heavy trucks in fatal accidents, however, increased about 40 percent from 1975 through 1979 but has since fallen off. so that the number for 1987 is only about 20 percent higher than that in 1975. Other Vehicles Other types of vehicles show varying trends of involvement in fatal motor vehicle accidents. The number of motorcycles increased almost 60 percent from 1975 through 1980 but has declined since then, so that the number in 1987 was only about 25 percent more than in 1975. The number of buses has nc’ver hew very high, never reaching as many as 400 in a year. Trends in fatal motor vehicle accidents by type of vehicle are generally Fatalities by Type of reflected in the number of fatalities in those vehicles. There arc clearly Vehicle Involved more fatalities in automobiles than in other types of vehicle. There are three to four times as many automobile fatalities as truck fatalities. Fatalities in other types of vehicles such as buses and motorcycles occur even less often. (See table IV.2.) Table IV.2: Fatalities by Type of Vehicle Type of vehicle 1975 1976 1977 1976 1979 1960 1961 1962 1963 -~ 1964 1965 1966 1967 Automobiles Small 4,289 4.903 5,606 6351 7 354 8,348 9,220 8.742 9,336~ 10,662 11,<02 i3jO99 13,902 IntermedIate 1,911 2147 2,265 2,843 --3,486.- 3 999 4.108 3,778 3743 3,807 3,663 3,902 3 866 Full~slzed 10,758 11,074-- 11,426 12,179 11,797 11 315 9,948 -~a,ls< 7,701 7189 6,339 6,631 5,856 Size unknown 8,992 8,062 77499 6,793 5,183 3.793 3375 3,061 2,612 2,322 2070 1.682 1,850 Total 25,950 26,166 26,796 26,166 27,620 27,455 26,651 23,735 2w92 23,960 23,574 25,314.25,474 Trucks Van~hased llqht 643 624-~ 745 926 1,019 1000 958 828 729 775 797 885 1049 ConventIonal plckup 4,029 4,706 5,104 5,710 6,102 6.461 6050 5,110 5,045 5,328 5477 6.007 6,607 MedIumand heavvtrucks 1,185 1303 1,481 1.601 1569 1.347 1.279 1.041 1070 1 188 1.120 1041 957 Total 5,657 6,633 7,330 6,237 6,690 6,606 6,267 6,979 6,644 7,291 7,394 7,933 6,613 Motorcycles 3,189 3,312 4,104 4,577 4 894 5 144 4,906 4,453 4,265 4 608 4.564 4,588 4031 Buses 53 73.- 42 40 39 46 57 35 51- 46 57 18 45 Other vehicles 876 1,335 1,461 ~1,100 1,153 1,369 1.395. 473~ 314- 397 499 400 414 Total fatalities 35,925 37,539 39,733 42,120 42,956 42,622 41,296 35,675 34,666 36,;22 36,066 36,253 36,577 lbgr 79 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix IV VehiclrRrlated Statistics IV.15.) Even in the peak year of 1977. only about 1,400 vehicles in fatal accidents were reported as having problems with tires. This number declined more than 40 percent until, in 1987. only 850 vehicles were reported with tires as a contributing factor. In none of these instances was the use of studded tires specifically reported. Figure IV.15: Tires as a Contributina Factor in Fatal Accidents 1500 Number of Vehicles imo Most fatalities in vehicles occur from head-on impacts, whether the Vehicle Fatalities and head-on collision is the initial or principal impact.,’ (See figures IV. 16 and Collisions IV. 17.) Fatalities from collision with the side, whether passengers’ or drivers’ side, taken together were consistently less than 50 percent of the fatalities from head-on impacts. Fatalities in head-on impacts tend to follow the overall trend. as do fatalities from drivers’ and passengers’ side impacts. Fatalities from rearend impacts, however, are on the mcreasc. (See figure IV. 18 on page 84.) The trend when the impact was classified as top, undercarriage, override, or underride is not clear. When only the initial impact is considered, fatalities follow the overall Page 81 GAO, PEMD-SO-10 Highway Safety: Trends in Highway Fatalitirs 1975-87 Appendix IV Vehicle-Related Statistics Figure IV.17: Vehicle Fatalities by Direction of Principal Impacta 22500 Number of Fatalities 1975 1976 i9n 1978 1979 1980 1981 1982 1982 1984 1985 1986 1987 - Head-on ---- Rearend - Side . . . . Other ‘Other Includes accldenls r which the prlnc~pal Impact IS described as noncoll~s~on, top, undercarri age overrfde or unknown Page RL GAO:PEMD-SO-10 Highway Safety: Trends in Highway Fatalities 1976.87 Appendix IV Vehicle-Related Statistics accidents, and so are more and more light trucks and vans. Fatalities in these vehicles are also on the increase. However, the fatal accident involvement rates and the fatality rates per number of registered vehi- cles are still generally declining. Exceptions to the general decline in these rates are the rates for subcompact automobiles and conventional pickup trucks, which have been increasing since 1983. Even though acci- dent involvement rates and fatality rates for small automobiles have been declining, they arc still a matter of concern, since they are consid- crably higher than those of larger automobiles. While medium and heavy trucks have one of the lowest occupant fatality rates, they have one’ of the highest fatal accident involvement rates. Age of vehicle does not appear to be as important a factor for automobiles as for trucks, since automobiles of all ages have very similar accident involvement rates. The only types of accidents that showed trends different from the overall trend are rcarend collisions and noncollision accidents, which have been increasing rather steadily. Page 85 GAO JPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix V Statistics Related to the Driving Environment Figure V.l: Fatal Accidents by Type of Roadway 25000 Number of FaInI Accldsnls IIIL----- ---------~-. ,e-- *c** zoooo I-II.I-c*- 15000 10900 5000 0 1975 1976 1977 1978 19-n 1980 1981 1992 1999 19s.l 1985 1995 1997 - Limited Access -1-1 Major Roads - Other Roads Roadway surface conditions are not a major factor in most fatal acci- Roadway Conditions dents, since over 80 percent of all fatal accidents occur on dry roads. (See table V. 1.) Accidents on both wet and dry roads tend to follow the overall trend. Fatal accidents under other road surface conditions such as snow and ice have always been few in number. Table V.l: Fatal Accidents by Road Surface Condition Condition 1975 1976 1977 1976 1979 1980 1981 1982 1963 1984 1985 1986 1987 Dry 31,630 32,848 34,170 ~36,312 36,201~ 38,062 36666 31,515 30,618 32,233 31,818 33,909 34,411 Wet 5,752 5,133 5,942 5,958 6,929 5272 5496 5823 5,734 5,673 5,439 5,801 5,625 Snow or slush 657 549 748 778 F333- 843 779 77i 694 685 902 497 566 Ice 722 833 956 963 846 727 506 667 620 788 802 604 587 Other 400 384 395 422 414 380 553 312 310 252 235 279 240 Total accidents 39,161 39,747 42,211 44,433 45,223 45,284 44,000 $9,092 37,976 39,631 39,196 41,090 41,435 Paye 87 GAO /PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix 1’ Statist.ics Related to the Driving Environment Table V.2: Fatal Accidents by Type of Traffic Control Control 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 None 31638 31,920 33774 35,399 35,752 34,841 34353 31,514 30,916 31,862 31462 33,094 33.203 stop signal Color SIgnal 1913 1937 2097 2,216 2,389 2,382 2346 1,894 1,925 2048 2,069 2,202 2,209 stop sign 2,952 3.073 3287 3633 3542 3,339 3,386 2,979 2,627 2 930 3,023 3,179 3,349 Total 4,865 5,010 5,384 5,849 5,931 5,721 5,732 4,873 4,552 4,978 5,092 5,381 5,558 Yield slgnal Flashing slgnal 278 243 255 273 289 260 292 280 259 281 -251 269 299 Yield sign 167 155 174 162 184 187 144 161 114 133 140~ 158 121 Sct~ool zone s,gn 11 12 15 5 8 7 6 6 6 15 9 9 9 Pedestmn sIgnal 0 0 0 32 43 53 52 240 169 217 195 174 225 Total 456 410 444 472 524 507 494 687 548 646 595 610 654 Railroad crowng Physlcalcontrol 185 188 200 230 203 223 237 78 70 86 79 89 74 stop s,gn 156 155 148 145 129 109 96 51 46 60 29 27 29 Other 275 331 302 308 302 275 192 301 282 366 305 316 314 Total 616 674 650 683 634 607 525 430 398 512 413 432 417 Traffic control not functIonlng 50 37 42 32 40 53 40 94 76 65 68 76 32 Other 1,279 1,452 1,696 1,819 2,214 3,396 2,500 1390 1,419 1,519 1,482 1456 1,523 Unknown 257 244 221 179 128 159 356 104 67 49 84 41 48 Total accidents 39.161 39.747 42,211 44,433 45,223 45,284 44,000 39,092 37,976 39,631 39,196 41,090 41,435 Freeway Signs IW.KShas recorded data specifically on freeway accidents only since 1981. Since then. frtx(l\Vay accidents have accounted for less than 15 per- cent of all fatal accidents. IIowever, after a slight drop in 1982. the tots1 number of fatal accidents on frerways has increased about 18 percent. (SW figure V.2.) Thtl brllk of this increase occurred in 1987. In most, years, over 90 pcrccnt of these accidents occurred where no special signs or other traffic c~ontrols existed. The pattern of accident increase on freeways tends to b(l more intense when no freeway traffic controls Mst (SW figure i’ :i i Page 89 GAO, PEMD-90.10 Highway Safety: Trends in Highway Fatalitirs 1975-87 Apprndix V Statistics Related to the Driving Environment Roadside Hazards FAKS reports roadside hazards as a problem for very few fatal accidents; they never total more than 400 accidents a year, and the total of such hazards has been decreasing rather steadily. (See figure V.4.) As a con- sequence. the E’XRSsystem stopped collecting special data on hazards after 198 1. Over half the hazards reported were trees or plants; build- ings and billboards accounted for fewer than 60 fatal accidents per year. Other roadside hazards were present at fewer than 120 fatal accidents per year. Because of the small numbers involved and the short time, not much can be said about individual types of hazards. Figure V.4: Fatal Accidents Involving Roadside Hazard9 400 Numkr of Fatal Accidents 350 3m 250 200 150 im 50 0 E-l Other Obsbwtions Trees, Plants Buildings, Billboards, and U-e Like Narrow-Bridge Accidents The Subcommittee txprc,ssed interest in bridge accidents, especially nar- row-bridge accidents. Ikidge accidents in total have always been fewer than 1.OOOper year, and acx,idents involving narrow bridges have been Pagr 91 GAO PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix V StaLktics Related to thr Drilin(! Environment Figure V.6: Vehicles Involved in Bridge Accidentsa 600 Number of Vehicles 550 500 1979 1990 1961 1962 Dab not routinely collected prior tu 1979 Weather is not a significant factor in most fatal accidents, over SOper- Fatal Accidents and cent of the accidents occurring under what was classified as normal Weather Conditions weather conditions. (See table V.3.) The number of normal weather acci- dents tends to follow the overall trend, while accidents in adverse weather show inconsisWnt trends. Table V.3: Fatal Accidents by Type of Weather Condition Condition 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 Normal 32,847 34,126 36 545 38,526 38,677 39,759 38377 33,374 32,381 34,197 33,647 35,748 36,159 Rar 3,953 3,514 3,841 3.965 4565 3597 3,723 3,939 4,088 3,645 3,733 3,851 ~3777 Sleet 80 70 77 137 109 107 86 108 107 110 114 91 119 %lOW 742 649 755 644 672 808 632 603 681 -636 805 475 622 Fog or other 728 543 623 799 864 797 809 876 562 910~- 770 784 645 Unknown 811 845 370 362 336 216 373 192 157 133 127 141 113 Total accidents 39,161 39,747 42,211 44,433 45,223 45,284 44,000 39,092 37,976 39,631 39,196 41,090 41,345 _________-. ~~~~~ Most fatal accidents oc‘cur on weekends, a pattern that is consistent Fatal Accidents by throughout the years. (SW figure V.7.) About 20 percent of all fatal acci- Day of the Week dents occur on Saturdays, Saturday accidents accounting for almost 2.000 more accidents caachyear than for Fridays and Sundays, the next PNgr 93 GAO PEMI)-90.10 Highway Safety: Trends in Highway Fatalities 1975-W Appendix V Statistics Related to the Driving Environment Table V.4: Vehicle Trips and Miles in 1983 by Day of the Week Percent of vehicle miles Percent of Percer$; Percent of trips traveled per Time trips per day day Weekday day 57 7% 54 4% 115% 10 9% Weekday night 17 7 175 44 44 Weekend day 173 19 7 a 7- ~- 99 Weekend mght 68 -78 23 26 Unknown 05 06 Most fatal traffic accidents occur during the nighttime hours of 7:00 Accidents by the Time p.m. through 1 a.m. (See figure V.8.) Fatal accidents in this time period of Day tend to follow the overall trend, but the total in 1987 was more than 15 percent below the peak year of 1980. Accidents in daytime and rush hours show some tendency to follow the overall t,rend, but they increased after 1982, and 1987 was the peak year for accidents in both time periods. Early morning accidents declined more than 20 percent from 1980 to 1983 and have remained rather steady since. Comparing data from the 1983 Nationwide Personal Transportation Study to fatal accidents for that year shows that the percentage of fatal accidents in the early morning hours is about four times the percentage of vehicle miles traveled for that time period. (See table V.5.) The percentage of fatal accidents during nighttime hours is over twice the percentage of vehicle miles traveled for t,hat time period. However, the percentage of fatal accidents in rush hour or other dayt,ime periods is far less than the percentage of vehicle miles traveled for those periods. As for the day of accidents, it would seem that aspects of driver behavior are more responsible for fatal accidents than volume of travel is. Page 95 GAO, I’EMD-SO-IO Highway Safety: Trends in Highway Fatalities 1975-87 Appendix V Statistics Related to the Driving Environment seasons is strongly correlated to the vehicle miles traveled for those sea- sons. (See table V.G.) While the fatal accident number for each season tends to follow the overall trend, each showed a different peak year, ranging from 1978 through 1981. Season appears to affect when acci- dents occur within a year. while changes across years appear more to reflect the overall trend. Figure V.9: Fatal Accidents by Season of the Year 14000 Number of Fatal Accidents 19500 19000 12500 12000 - wmter mm-1 Spring - Summer . . . . Fail Table V.6: Relationship of 1983 Accidents to Vehicle Miles Traveled by Percent of vehicle miles Season Season Percent of accidents traveled Summer 28 2% 24 1% Spring 23 3 28 1 Winter 21 4 22 8 Fall 27 1 25 0 Page 97 GAO/PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix VI Major Contributors to This Report Richard ‘I’. ISarms, Assistant Director Program Evaluation Roy Ii. .Jones, Project Manager and Methodology Dale IV. IIarrison, Opcrirt ions Research Analyst Division Page 99 GAO. P&MI)-90-10 Highway Safety: Trends in Highway Fatalities IYW87 Bibliography Griffin, Lindsay I., et al. Fatal Case Study Analysis. Chapel Hill, N.C.: Highway Safety Research Center, IJniversity of North Carolina, Septem- ber 1975. IIeckard, R. F.. .J. A. Fachuta, and F. A. Haight. Safety Aspects of the National 55 MPH Speed Limit. [Jniversity Park, Penn.: Pennsylvania Transportation Reseurrh Institute, The Pennsylvania State liniversity, November 1976. lIedlund, James, et al. “An Assessment of the 1982 Traffic Fatality J)ecrcase.” Accident. Analysis and Prevention, 164 (1984), 247-61. .Jatras, Kathy Pappas. Motorcycles. Fatal Accident Reporting System. Special Report on LMoto?cycles. Washington, D.C.: National Highway Traffic Safety Administration. 1979. .Joksch, Hans C. “The Relationship Between Motor Vehicle Accident I)caths and Economic, Activity.” Accident Analysis and Prevention, 16:3 (.June 1984). 207-l 0. Joksch, Hans C., and Stephen Thoren. Car Size and Occupant Fatality Risk, Adjusted for Differences in Drivers and Driving Conditions. Hart- ford, Conn.: The Center for the Environment and Man, Inc., January 1984. Kleit, Andrew N. The Impact of Automobile Fuel Economy Standards. Working paper 160. Washington, D.C.: Bureau of Economics, Federal Trade Commission, February 1988. Kurucz, Charles N., and Bertan W. Morrow. “A Causal Model for Single Vehicle Accidents.” Presented at the Motor Vehicle Collision Investiga- t,ion Symposium, Buffalo. New York, October 1975. McDonald, Gary C. “Nonparametric Selection Procedures Applied to State Traffic Fatality Rates.” Technometrics, 21:4 (November 1979), 515-23. McaKelvey, Francis X.. et al. “Highway Accidents and the Older Driver.” Presented at the 67th annual meeting of the Transportation Research Board, Washington, D.(‘., .January 1988. Page 101 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Bibliography National Highway Traffic Safety Administration. Fatal Accident Report- ing System 1986: A Review of Information on Fatal Traffic Accidents in the 1J.S.in 1986. Washington, D.C.: U.S. Department of Transportation, March 1988. National Highway Traffic Safety Administration. Fatal Accident Report- ing System 1987: A Review of Information on Fatal Traffic Accidents in the ITS. in 1987. Washington. D.C.: U.S. Department of Transportation, December 1988. National Highway Traffic Safety Administration. National Accident Sampling System 1985: 11 Report on Traffic Accidents and Injuries in the I.nited States. Washinzn. D.C.: L1.S.Department of Transportation, February 1987. National Safety Council. Accident Facts 1988 Edition. Chicago, 111.:1988. O’Day, .James, et al. Data Sources to Support the NHTSA Defects Investi- gation System. Ann Arbor. Mich.: Highway Safety Research Institute, The I’niversity of Michigan. March 1978. O’Day, *James, et al. Combination Vehicles: Five-Year Accident Experi- ence. Ann Arbor. Mi&.: Ilighway Safety Research Institute, The Univer- sity of Michigan, .July 1980. O-Day, .James, and Richard Kaplan. “The FAKS Data and Side-Impact Col- lisions.” HSRI Research Review, 9:5 (March-April 1979), 9-17. Partyka, Susan C. “SimJ)lc Xlodels of Fatality Trends Using Employment and Population Data.” Accident Analysis and Prevention, 16:3 (1984), 21 l-22. Rana, Riaz H., and Roger I’. Quene. Review of Motorcycle Exposure Data. Columbia, Md.: Statistira, Inc.. July 1982. Robertson, Leon S. “I’attc>rns of Teenaged Driver Involvement in Fatal Motor Vehicle Crasht,s: implications for Policy Choice.” Journal of Ilealth Politics, Policy d anti ~~ Law. ~~~ 6:2 (Summer 1981), 303-14. Semans, Thomas R. Problems in the Establishment of an Equitable Exposure Denominator in i2ccident. Injury and Fatality Rates. Boise, aaho: Bureau of Highnay Safety, Idaho Transportation Department, Soptcmb(~r 1977. Pagr 103 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 ____-.----_ ., ,,, ,. “. ,. ., Solomon, Kenneth A.. et al. Improving Automotive Safety: The Role of Industry, the Government, and the Driver. Santa Monica, Calif.: The Rand Corporation, i&y 1985. Treat J. R.. et al. Tri-Level Study of the Causes of Traffic Accidents. Bloomington, Ind.: Institute for Research and Safety, Indiana lTniversity, 1979). Treat, .J. R., and David Shinar. “A Methodology for Assessing and Classi- fying Traffic Accident Causes.” I’rcsented at the Motor Vehicle Collision Investigation Symposium, IJuffalo, New York, October 1975. I..S. Department of’ Transportation, Office of the Secretary. Personal Travel in the I1.S. Vol. I, A Report on the Findings from the 1983-1984 Nationwide Personal Transportation Study. Washington, D.C.: August 19%. Waller, Patricia F. I’lugging the Gaps in Data Collection Systems. Chapel Hill, N.C.: Highway Safety Research Center, ITnivcrsity of North Caro- lina. October 1983. Waller, Patricia I’.. ct al. Methods for Measuring Exposure to Automobile Accidents. Chapel lhll. N.C.: Highway Safety Research Ccntcr, IJniver- sity of North Carolina. November 1974. WHO Ad Hoc Technicxl Group on Road Traffic Accident Statist,&. Road Traffic Accident Stat istics. Copenhagen, Denmark: World IIcalth Organi- zation, 1979. -- Williams, Allan P’. “highttime Driving and Fat,al Crash Involvement of Teenagers.” Accident Analysis and Prevention. 17: 1 (IS%), 16. Yaksich, Sam .Jr. ilnalysis of HSRI Study of Car/Truck Crashes in the Ilnited States. Falls C’hurch, Va.: American Automobile Association Foundation for ‘l’raff~c Safety. 1982. Page 104 (;A0 ‘PEMD-SO-10 Highway Safety: Trends in Highway Falalities 1975.87 Bibliography Mak, King K., et, al. Assessment of Existing General Purpose Data Bases for Highway Safety Analysis. College Station, Texas: Texas Transporta- tion Institute, The Tc>sasA&M University System, January 1988. Mak, King K.. and Lindsay I. Griffin. Assessment of Existing Data Bases for Highway Safety Analysis. College Station, Texas: Texas Transporta- tion Institute, The Texas A&M University System, November 1985. Mcla, Donald F. “l+:xposure Data Needs.” Presented at the Motor Vehicle Collision Investigation Symposium, Buffalo, New York, October 1975. Morganstcin, David 1~.“Fatal Accident Reporting System and National Accident Reporting System.” Presented at the Motor Vehicle Collision Investigation Symposium, Buffalo, New York, October 1975. Motor Vehicle Manul’a~turcrs Association of the United States, Inc. MVMA Motor V~chiclc Facts & Figures ‘88. Detroit, Mich., and Washing- ton, D.C.: 1988. National Center for Statistics and Analysis. Fatal Accident Reporting System: 1984 Coding and Validation Manual. Washington, D.C.: National Highway Traffic Safety Administration, ITS. Department of Transporta- tion, 1984. National Center for Statistics and Analysis. Fatal Accident Reporting System: 1985 Coding and Validation Manual. Washington, D.C.: National Highway Traffic Safety Administration, ITS. Department of Transporta- tion, 1985. National Highway Traffic Safety Administration. Alcohol and Highway Safety 1984: A Rcvicw of the State of the Knowledge. Washington, D.C.: ITS. Department of Transportation, February 1985. National Ihghway Traffic: Safety Administration. Fatal Accident Report ing System (I+.I~s)Quality Control: A Description of the Quality Control Systems I-sed in F’-\I~sWashington, D.C.: ITS. Department of Transporta- tion October 1983. National Highway Traffic Safety Administration. Fatal Accident Report- ing System 1985: I1 Review of Information on Fatal Traffic Accidents in the IIS. in 1985,. Washington, D.C.: I!.S. Department of Transportation. February 1987. Page 102 GAO ‘PEMD-90.10 Highway Safety: Trends in Higlway Fatalities 197587 Bibliography Benguerel, Yvett,e. “Mandatory Seat Belt Legislation: Panacea for High- way Traffic Fatalities?” Syracuse Law Review, 36 (198fi), 1341-71. Bureau of the Census. Preliminary Estimates of the Population of the I-nited States by Age, Sex. and Race: 1970 to 1981. Current Population Reports, Population Estimates and Projections, Series P-26, No. 917. Washington, D.C.: I’.% Department of Commerce, .July 1982. Bureau of the Census. I rnited States Population Estimate by Age, Sex, and Race: 1980 to I X3’?.-Current Population Reports, Population Esti- mates and Projections, Series P-25. Ko. 1022. Washington, D.C.: ITS. Department of Commt’rce. March 1988. Carsten, Oliver. “llsc, of the Nationwide Personal Transportation Study to Calculate Exposure.” HSRI Research Review, 115 (May-June 1981), l-8. Carsten. Oliver. and Lc,slie C. Pcttis. Trucks Involved in Fatal Accidents 1980-84 by PO&r I ‘nit Type. Ann Arbor, Mich.: The IJniversity of Mich- igan Transportation Research Institute, August 1987. Council, Forrest M.. et al. Accident Research Manual. Chapel Hill, N.C.: IIighway Safety R~~sc~srch Center, IJniversity of North Carolina, *January 1980. Crandall, Robert W.. and .John D. Graham. “The Effect of Fuel Economy Standards on Automobile Safety.” The *Journal of Law and Economics, 32:l (1989). 97-l 18 Crew, Alexander R. Time Series Forecasting of Highway Accident Fatali- ties. Washington, DC.: Institute for Applied Technology, Technical Anal- ysis Division, Nat ion al I3ureau of Standards: Department of Commerce, March 1973. Eldridge, Marie D. “Status of Present Accident Data Systems.” Pre- sented at the Motor Vehicle Collision Investigation Symposium, Buffalo, New York, Ortobcr 1975. Garrett, .John W. “St atus of Present Accident Data Systems: A Recent IIistory of i\ccidrnt Invc>stigation and Data Analysis.” Presented at the Motor Vehicle Collision Investigation Symposium, Buffalo, New York, October 1975. Page 100 GAO,‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1976-87 Appendix V Statistics Related to the Driving Environment While the numbers of fatal accidents under various environmental con- Conclusions ditions reflect the effects of those conditions within any particular year, they tend not to cause trend patterns to deviate from the overall trend. Exceptions include some of the specific areas of the Subcommittee’s con- cern-namely, traffic controls, freeway accidents and freeway signs, roadside hazards, and narrow bridges. The number of accidents related to roadside hazards and narrow bridges appears to be steadily declining. Accidents where only “yield” traffic controls exist are on the increase. Freeway accidents are increasing the most where no traffic controls are present. However, accidents have increased the most on county and other local roads. Page 98 GAO/PEMB90-10 Highway Safety: Tends in Highway Fatalities 1975.87 Appendix \’ Statistics Related to the Driving Environment Figure V.8: Fatal Accidents by the Time of Daya 20000 Number of Fatal Accidents Morning Hours (1 a.m. to 6 a.m.) Rush Hours (6 to 9 a.m. and 4 to 7 p.m.) Daytime Hours (9 a.m. to 4 p.m.) Nighttime Hours (7 p.m. to 1 a.m.) ‘Times are based on those reported r the 1983 Natwwde Personal Transportation Survey Table V.5: Relationship of 1993 Accidents to Vehicle Miles Traveled by Percent of vehicle miles Time of Day Time Percent of accidents traveled Morning 19.1% 4 ax Rush hours 23 7 39 i DaytIme 22 9 40 7 NighttIme 33 7 148 Unknown 06 cl6 Marc fatal accidents occur during the summer than during the other sea- Accidents by Season sons of the year. ’ However. the numbers for spring and fall are not far of the Year behind. (Set figure 1’9 ) The number of fatal accidents in the various Pay? 96 G4O,‘PEMD-90-10 Highway Safety: l’rends in Highway Fatalities 1975.87 Appendix V Statistics Related to the Dri\ iag Envirmmmt highest days. The pattern for all days, however, tends to follow the overall trend. The occurrence of most fatal accidents on weekends appears not to bc related to the existence of more motor vehicle activity on weekends. Data from the 1983 Nationwide Personal Transportation Survey indicate that not only are more total miles driven on weekdays than on weekends but also the average per weekend day is less than the average per weekday. These same relationships exist for the number of vehicle trips on weekends and weekdays. It would seem, therefore, that some aspects of driving behavior-perhaps drinking and driving-are more responsible I’or fatal accidents than the volume of travel. (See table \‘.4.) Figure V.7: Fatal Accidents by Day of the Week loo00 Number of Fatal Accidents - Sunday --mm Average per Weekday - Friday . . . . Saturday Page 94 GAO/PEMI>-90.10 Highway Safety: Trends in Highway Fatalities 197587 Appendix V Statistics Related to the Driving Environment - even fewer. (See figure V.5.)’ Moreover, these accidents have been steadily on the decline. The number of vehicles actually striking bridges is even smaller than the number of accidents involving bridges, and this number has been steadily declining. (See figure V.6.) Only 450 such acci- dents occurred in 1987. Figure V.5: Fatal Accidents Involving Bridge9 000 Number of Fatal Accidents 850 800 750 700 659 690 550 500 450 400 1979 1980 1981 1962 1983 1984 1985 1986 1987 Page 92 GAO PEMD-YO-10 Highway Safety: Trends in Highway Fatalities 197587 Appendix V Statistics Related to the Ikibing Ewironmrnt Figure V.2: Fatal Accidents on Freeway9 6000 Number of Fatal Accidents 5wo 4000 3000 2000 1000 0 1981 1932 1983 1984 1985 1986 1987 Data spec~i~cally related 11 frecwajs not avaIlable prior to 1981 Figure ‘4.3: Freeway Fatal Accidents With Some Traffic Control9 1000 Number of Fatal Accidents 600 0 1981 1982 1983 1984 1985 1986 1987 ‘Ddta speoflcallp related 1~ frr”~avs not avakablc pmr to 1981 Paye 90 GAO PEMD-90-10 Highway Safely: Trends in Highway Fatalities 197547 Appendix V Statistics Related to the Driving Environmmt The Subcommittee expressed particular interest in how such elements as Roadside and Traffic roadside hazards, narrow bridges, traffic controls, and the like affect Conditions fatal accidents. IJnfortunately, FAKS does not routinely collect data on many of these elements, and no specific provision is made for them on the FARS data collection instrument. For some elements, information is available only when accident investigators specifically report them as “contributing factors.” Therefore, we can report some information, but we do not have a good sense of the completeness of the data. Accidents and Traffic Most fatal accidents-about 89 percent-occur where there are no traf- Controls fit controls. (See table \‘.2.) Whether controls are present or not, fatal accidents tend to follow the overall trend. However, when individual types of traffic controls are considered, trend differences do appear. Accidents occurring where there are stop signals-either lights or stop signs-follow the overall trend. However, accidents where railroad sig- nals exist are on the decline while accidents where only yield signals exist are ,just as clearly increasing. Accidents where other unidentified types of traffic controls existed increased steadily from 1975 through 1980 and then decreased rapidly through 1982 to the earlier level, and they have been rather constant since. Accidents where existing traffic controls were not functioning have always been few. Page 88 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix V Statistics Related to the Driving Environment Elements of the driving environment-such as weather, time of day, and type of roadway-can also contribute to fatal motor vehicle acci- dents. In this appendix, we discuss various aspects of the driving envi- ronment and their relationship to fatality trends. Since most legislation addressing environmental issues has addressed roadway conditions and roadway hazards, WCaddress the roadways first. Finally, we discuss elements that legislat,ion cannot address directly-such as weather and time of day-but t,hat , like speed limits, can be addressed through local law enforcement. Since exposure information specifically related to environmental conditions was not readily available and since many of the phenomena discussed occur infrequently, we discuss only the basic accident frequencies. Most fatal motor vehicle accidents occur on major roads such as LJS- Fatal Accidents by numbered and stat<>-numbered routes and similar major arteries. (See Type of Roadway figure V. 1.) Local, c.ount.y, and other roads are the locations for the next highest number. Thts f(wcst accidents occur on limited-access highways or freeways. In nearly all the years covered by our analysis, over five times as many fatal accaidcnts occurred on major roads as on limited- access highways, and over four times as many occurred on local roads. These ratios have remained fairly constant over the years, although the ratio of major highway to limited-access highway accidents is declining somewhat. This suggests that speed-limit legislation addressing only lim- ited-access highways is not necessarily the optimal method of cutting the number of fatal accidents. Fatal accidents on hmitcd-access highways are becoming more of a prob- lem in recent years, however. Although fatal accidents on both limited- access highways and major roads tend to follow the overall trend, acci- dents on limited-access highways have been increasing at a faster rate since 1983. The number of’ fatal accidents on limited-access highways has increased slightly over 15 percent since 1983, while the number on major roads has inc~rcascd only a little over 5 percent. However, recent NIWSA reports indicate that a large part of the fatality increase stems from increased travc,l on limited-access highways and that the fatality rate per miles travc>kd shows a much smaller rate of increase. Fatal accidents on local roads have increased over 10 percent since 1983, the 1987 total of 18,200 accidc,nts being only about 700 less than for the 1981 peak year for s11c.11 accidents. Page 80 GAO PEMD-YO-IO Highway Safety: Trends in Highway Fatalities 1975.87 Appendix N Vrhicle.Rc=latrd Statistics Figure IV.18: Vehicle Fatalities With Principal Rearend Impact 1600 1500 1400 1300 1200 1100 1000 900 1975 1978 1977 1979 1979 1980 1981 1992 1983 1984 1985 19aa 1987 Figure IV.19: Vehicle Fatalities From Noncollision Accidents 5590 Number of Faialiiles 5090 4500 4000 3500 3oao 2500 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1995 1988 1987 The types of vehicles involved in fatal accidents has been changing over Conclusions the years. More and more small cars are involved in fatal motor vehicle Page 84 GAO ‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 197587 Appendix IV Vehicle-Related Statistics trend. When only principal impact is considered, however, such fatali- ties have been generally on the decline since 1979. Fatalities from non- collision accidents, after an initial decline from 1975 through 1978, have been steadily increasing since then, an increase of more than 70 percent. (See figure IV.19 on page 84.) Figure IV.16 Vehicle Fatalities by Direction of Initial Impact’ 30000 Number of Deaths 25000 20000 - l ===.... l =..=.=............gg,,*m~ l m.m.mm.....m...mmm,..mmm..mm.mmm..m.mmm l mmmmmmmmmmm woo m-1-1--1-- -111--1------1---11-IIIIIL-ll-l---l-ll------------------------------------------------ 0 1975 197s 19Tr 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 - Head-on -1-1 Rearend - Side mm.. other Page 82 GAO/PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix IV Vehicle-Related Statistics Automobiles The number of fatalities in automobiles tends to follow the overall trend in fatal automobile accidents. However, like the number of accidents. there are differences according to size of automobile. The number of fatalities in full-size automobiles has decreased more than 50 percent since the peak year of 1978. Meanwhile, the number of fatalities in intermediate cars more than doubled from 1975 through 1981 and then declined slightly. The number of fatalities in small automobiles has increased steadily since 1975, only 1 year showing any decrease. This trend toward increase is apparent regardless of the type of small automobile. Trucks The number of fatalities in trucks has been rather steadily increasing over the years but the increase is dominated by fatalities in van-based light trucks and conventional pickup trucks. Fatalities in trucks approached their highest level in 1987, reflecting an increase of about 47 percent since 1975. Fatalities in light trucks account for a substantial proportion of these fatalities and, in turn, fatalities in conventional pick- ups account for most of the light truck fatalities. The number of all light truck fatalities in 1987 was almost 65 percent higher than in 1975; the number of fatalities in van-based trucks was also almost 65 percent higher than in 1975, as were the number in conventional pickups. How- ever, the number of fatalities in medium and heavy trucks increased from 1975 through 1978, but they have since declined about 40 percent to a low in 1987 that was almost 20 percent less than in 1975. Other Vehicles Motorcycle fatalities follow a trend similar to that for medium and heavy trucks. After increasing more than 60 percent from 1975 through 1980, motorcycle fatalities declined more than 20 percent through 1987. The number of fatalities on buses has never been large, generally accounting for only :3Clto 40 fatalities per year. - ~~ One of the Subcommittee’s requests was that we inquire into the use of Vehicle Tires and studded tires and their effects, if any, on fatal accidents. The FARS sys- Fatal Accidents tem, unfortunately, does not routinely collect information on the effects of tires on fatal accidrnts: there are no specified elements on the F'AKS data collection instrllmcnts to collect data on tires. Some data on tires is collected but only whcsn accident investigators use a miscellaneous cate- gory called “related factors.” We analyzed the “related factors” vari- ables and found that, over the years, tires are being reported less and less as a contributing factor in fatal motor vehicle accidents. (See figure Page SO (GAO PEMD-90-10 Highway Safely: Trends in Highway Fatalities 147587 Appendix IV Vehicle-Related Statistics Table IV.1: Vehicles Involved in Fatal Accidents Type 01 vehicle 1975 1976 1977 1976 1979 1960 1981 1982 1983 1984 1985 1986 1987 AutomobIles Small 5,046 5,627 6,506 7,451 8,603 9795 11041 10,628 11,572 13,272 14,571 16,806 18,022 lntermedlate 2,638 2,810 3,133 3,964 4,945 5597 6036 5,359 5,291 5,664 5,488 5,718 5,717 Full-sized 17,942 17,830 18,967 19,983 19.!%0 18.501 17 193. 147368 13,092 12,752 11,512 11,532 10,403 Sueunknown 12,299 10,819 10,321 9,035 6,763 5065 4451. 4,332 3 686 3,295 2,972 2,532 2,802 Total 37,925 37,066 38,927 40,433 39,901 38,958 38,721 34,687f33,641 34,983 34,543 36,588 36,944 Trucks Van-based light 1,208 1,186 1446 1,831 2,037 2041 1989 1,775 1,663 1,779 1,853 2,001 2,305 ConventIonal pIckup 6,916 7,710 8.548 9,668 10.331 10.566 10,105 8,970 8,853 9,497 9,850 10,601 11,471 Medlumorheavy truck 4,570 4,958 5,724 6,333 6,421 5589 5,603 4.880 5,159 5.479 5,565 5,468 5,466 Total 12,694 13,854 15,718 17,832 18,789 18,196 17,697 15,625 15,675 16,755 17,268 18,070 19,242 Motorcycles 3,265 3 343 4,164 4,643 4,916 5,194 4,963 4,495 4,302 4,659 4,608 4,571 4,062 Buses 327 319 321 372 347 330 342 289 307 320 337 286 354 Other vehicles 1,323 1,333 1246 725 682 693 816 1 227. 1,059 1,144 1,323 1218 1,147 Total vehicles 55,534 56,084 60,516 64,144 64,762 %3,485 62,699 56,455 55,106 57,972 58,271 60,792 61,825 Automobiles While the fact that aut,omobiles have been the most frequent vehicle type involved in Fatal accidents is important in itself, breaking down automobiles by size shows even more interesting results. After increas- ing slightly from 1975 through 1978, the number of full-size automobiles in fatal accidents declined almost 50 percent from 1978 through 1987. The number of intermediate automobiles involved in fatal accidents increased steadily from 1975 through 1981, increasing over 125 percent. Since 1981. the number of intermediate automobiles in accidents has fluctuated; the number in 1987 was slightly below that in the 1981 peak year. The number of small automobiles in fatal accidents has been stead- ily on the increascl. The number of minisize automobiles has increased about 150 percent o\‘tr 1975, the number of subcompact cars has increased almost 200 percent, and the number of compact cars has increased about 750 pcrccnt. Trucks Trend patterns for trucks involved in fatal motor vehicle accidents are related to truck t ypt’. The number of medium and heavy trucks tends to follow the overall trend. while the number of light. trucks and vans is on the increase to such a degree that their numbers tend to dominate the overall total tru(.k t rr,nd. The number of light, trucks and vans in fatal Page 78 GAO,PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appmdix IV VehicleRelated Statistics and less has been steadily declining since 1978. However, while trucks 11 to 15 years old underwent a similar decline in the rate from 1978 through 1982. the rate increased significantly from 1982 to 1987. Figure IV.13: Automobiles in Fatal Accidents by Age 50 Accidmls per 109,099 Regislered Automobiles 30 1975 1976 1977 197s 1979 1980 1981 1982 1983 1984 1985 1986 1987 - 5 years and under ---- 6-10yearS m 11-15 years nnnn 16yearsandolder Page 76 GAO, PRMD~!JClO Highway Safety: Trends in Highway Fatalities 1975-87 Appendix IV Vehicle-Related Statistics The age of vehicles has always been a matter of concern because older The Age of Vehicles vehicles tend not to be as well maintained as newer vehicles. Moreover, Involved in Fatal older vehicles do not have all the safety devices mandated by changes in Accidents motor vehicle safety standards over the years. Since data on age for vehicles on the nation’s highways were readily available only for auto- mobiles and trucks, we have restricted this analysis to these types of vehicles. The average age of both automobiles and trucks has been steadily increasing since 1975, although it has leveled off in recent years. During this time, the average age of trucks has been consistently older than that of automobiles. However! the average age of automobiles involved in fatal motor vehicle accidents tends to be older than the average age of trucks in such accidents. For automobiles, the average ages of vehicles on the road and of the vehicles involved in fatal accidents are almost the same, but since 1983 there has been a slight divergence. (See figure IV.11.) The same pattern is not prevalent for trucks; the average age has been consistently older than the average age of trucks involved in acci- dents. (See figure IV. 12 ) Figure IV.1 1: The Average Age of All Automobiles and of Those in Fatal Accidents 6.0 Average Age In Years - Automobiles in Atidents ---- Automobiles on the Roads Page 74 GAOIPEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix IV Vehicle-Related Statistics 1983-a decline of over 25 percent-and then increased about 10 per- cent from 1983 through 1986. (See figure IV.10.) Although the fatality rate for occupants of medium and heavy trucks did increase slightly in 1984, the rate’s overall trend declines, having declined almost 40 per- cent since 1980. Figure IV.8: Truck Fatality Rates by Type of Trucka 350 Fstali(y Rate pr Million Trudm 300 2w 50 IWO 1981 1982 1983 1334 1985 1924 - AllTrudts ---- Van-basedTrucks m Conventional Pickups nnnn Mediumand Heavy Trucks The number of registered trucks was not available by type for years prior to 1979 Page 72 GAO,‘PEMB90-10 Highway Safety: Trends in Highway Fatalities 1975.87 ~_~ ~~__ .4ppcndii IV Vehicle-Related Statistics Figure IV.6: Fatality Rates by Size of Automobilea 350 FataWes per Milliin Automobiles 300 250 200 150 - Small -I-- Intermediate m Full ‘The nlrmber of regstered autornobks was not available by we for years prior to 1978 Page 70 (iAO/PEMD-9010 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix IV Vehicle-Related Statistics Figure IV.4: Fatal Accident Rate by Type of Light Trucka 550 Involvsmenl Rale pr Million Trucks 500 4!50 wwd.lw.----- -.mlll.---- 400 350 300 250 200 150 1983 1904 1995 ,986 1980 1991 1982 - van-based Trucks ---- Cawntional Pi&ups - All Light Trucks ‘The number 01 reglslered trucks thy type was not available for years prior to 1979 Page 68 GAO,‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix IV V&irk-Related Statistics Figure IV.2 Fatal Accident Rate for Small Automobile9 650 Rate per Million Automobiles 300 250 - 1976 1979 1990 1981 1962 1983 1984 1995 1985 - All Small Automobiles -w-w Minisize ‘The number of reglsiered aut~,mob~l~es by size was not available for years pr,or to 1978 Trucks Fatal motor vehicle accident involvement rates for trucks show interest- ing patterns. (See figure IV.3.) The trucks with the highest involvement rate in fatal motor vehicle accidents are medium and heavy trucks. Van- based light trucks have the lowest involvement rate. Overall truck involvement in fatal accidents has been steadily declining since 1980, only 1 year showing an increase in the rate of involvement. While the involvement, of light trucks overall is also declining, the involvement of conventional pickups has been increasing since 1983. (See figure IV.4 on page 68.) The involvement rate for conventional pickup trucks is sub- stantially higher t ban that for van-based light trucks, and conventional pickup trucks cxctrt the greatest influence on the overall involvement ratL>for light truc,ks While the involvement rate for medium and heavy trucks is the highest of all truck sizes. this rate also has declined since 1980. (See figurcb I\?? on page 69.) However, the bulk of this decline occurred in only 1 year: otherwise, the involvement has been rather con- stant, especially from 1982 through 1986. Page 66 (&W fPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix I\ Vehicle-RelatedStatistics In this appendix, we address several issues related to motor vehicles, including accident involvement rates and fatality rates by type of vehi- cle. the effects of t,htkc,hanging combination of vehicle sizes and the aging of vehicles. and how fatalities depend on the type of collision. The discussion on involvement rates and fatality rates is limited to automo- biles and trucks, since registration information is not readily available for other types of vc+iclcs. The number of full-sizcl automobiles in fatal accidents declined at the Fatal Accident same time that there were significant increases in the number of small Involvement Rates by and int.ermcdiatc cars involved in Fatal accidents. The question arises, Type and Size of therefore, as to wh(%hc)r t,his indicates inherent safety differences by size of automobile or \vhtt her the accident numbers simply reflect the Vehicle changing composition of vehicle types the public drives. Relating the numbers of accidents 1o the number of registered vehicles helps answer this question.’ In some cases, the trends in involvement rates and fatal- ity rates are generally the same as the absolute numbers (for example, full-size automobil( whereas in others the rates show a trend com- pletely the reversr of t,hc absolut,e numbers (for example, small and intermediate automc)bilrs). Automobiles The increases notcbd in the number of small and intermediate automo- biles involved in fatal accidents are not apparent when t,he increase in the number of these automobiles on the roads is taken into account. (See figure IV. 1.) It is also interesting that while the accident involvement rate for intermediate automobiles was the highest by far in 1978, it has declined so rapidly-almost a 50.percent decline since 1978-that by 1986 the rate for thc,sclcars was the lowest. This figure also clearly shows that the fat al accident involvement rate for small automobiles is now higher than for c,ither intermediate or full-size cars. The fatal acci- dent involvement rat t‘ for full-size automobiles \vas the lowest in most years but did not show the same continued decline as the number of such aut,omobilcs. As a consequence, full-sizr aut,omobiles no longer have the lowest fatality rate. Paye 64 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 - Appendix III Driver-Related Statistics Figure 111.13:Occupants Killed Who Were Using Safety Restraints 60 Percent Killed When ibtmints Were Used 25 - Drivers 1-11 Passengers The overall trend-increases through 1980, decreases through 1983, Conclusions and then increases through 1987-applies not to all but to many driver- related statistics. Drivers clearly are the greater part of motor vehicle fatalities, and male drivers dominate in involvement in fatal accidents, whether viewed in simple numbers or as rates of involvement in fatal accidents. However, more females are becoming involved in fatal accidents. Prinking drivers are still a very serious traffic safety problem, but FARS data do not disclose any trends that diverge much from the overall trend. Our analysis of the number of fatalities related to single-vehicle nighttime accidents, a common surrogate measure for drinking drivers, shows that the reported number of drinking drivers may be more accu- rate than is generally believed. Page 62 GAO:PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix III Driver-Related Statistics Figure 111.11:Motor Vehicle Occupants in Fatal Accidents Reported Not Using Safety Restraints 90 Percent of occupanto - Drivers -9-- Passengers Has the greater use of safety restraints been reflected in fatality pat- terns‘? Comparing fatality data to use patterns shows that, while the percentage of drivers not using safety restraints who are killed has been steadily increasing, the percentage of drivers who used restraints and were killed anyway has been steadily declining. (See figures III.12- 111.14.)Moreover, while the percentage of passengers not, using safety restraints who are killed has tended to follow the overall trend, the per- centage of drivers who used restraints and wcrc killed has also been steadily declining. Since the percentage of both drivers and passengers killed whose safety restraint usage is unknown has also been declining, a case can be made that many whose rest rain1 nsagc is unknown proba- bly are using them. Page 60 GAO/PEMl)-90-10 Highsvay Safely: Trends in Highway Fatalities 1975-87 Figure 111.9:Motor Vehicle Occupants in Fatal Accidents Reported Using Safety Restraints 30 Percent of ocCupants 25 - Dwers I - - - b.SSenQerS Page 58 GAO PEMD-SO-10 Highway Safety: Trends in Highway Fatalitirs 1975-W - Appendix III Driver-Related Statistics drinking, so the FAKS data are not as useful as one would like. Data reporting on drinking drivers is improving, however, and NHWA has recently used analytical t.echniques to obtain measures that offset inade- quate reporting.’ Nevertheless, some insights into drinking as a factor in fatal accidents are available. Since 1977, more than 35 percent of the fatal accidents reported involved the presence of at least one drinking driver. (See figure 111.8.) Between 2 and 3 percent of those accidents involved two or more drink- ing drivers. The number of fatal accidents involving drinking drivers- whether one or two or more such drivers-tends to follow the overall trend, especially since 1977. In our opinion: 1975 and 1976 reflect data collection problems morr‘ than a better drinking-driver record for those years. Because of the reporting problems connected with drinking driv- ers, we determined the trend in the number of fatalities related to single- vehicle nightt,ime accidents, a common surrogate measure for drinking drivers. This analysis shows that the rates were not very different. Therefore, we believe the reported number of drinking drivers may be more accurate than is g:c>ncrallybelieved. Page 56 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix III Driver-Related Statistics increases and decreases in the number of male drivers in fatal accidents after 1983, the 1987 total was the lowest for the period of our study. The number of female drivers in this age group in fatal accidents tends to follow the overall trend. The number of female drivers involved in fatal accidents in 1987 is almost 20 percent higher than the low of 1975 but about 10 percent lower than the high of 1978. There are also significant differences between male and female drivers in the 21. through 25year-old age group. While the involvement of both groups tended to follow the overall trend, the involvement of female drivers increased dramatically after 1982. The involvement of male drivers showed modest increases after 1983, and the total for 1987 was the second lowest number of male drivers in this age group involved in fatal accidents. The number of female drivers for this age group involved in fatal accidents in 1987 was the highest so far, and it was almost 40 percent higher than the low of 1975. The male total for 1987 was about 3 percent lower than 1975 and more than 20 percent below the peak year of 1979. For both male and female drivers in the 26. through 50.year-old age group, the number involved in fatal accidents in 1987 was the highest on record. However, the number of female drivers in this age group increased over 50 percent since 1975, while the number of male drivers increased only about 20 percent. Moreover, the number of male drivers in this age group involved in fatal accidents decreased in 4 different years, but the numb(xr for female drivers decreased only once. Male and female drivers in the 5 l- to 65-year-old age group also show differences. Male involvement for this age group has tended to follow the overall trend; female involvement, while erratic. has shown a gen- eral tendency to increase since 1975. The number of male drivers of this age involved in fatal accidents has decreased about 6 percent since 1975, while the nurnb(>r of female drivers has increased about 18 per- caent.Only for the ox cr-65 age group does the involvement, of male and female drivers show similar trends, and both are on the increase. The involvement of fcmalc drivers in this age group has almost doubled since 1975, while ttrc number of male drivers has increased over 20 percent. _____~- ~~~~ The speed of vehicles involved in fatal traffic accidents is a measure of Speed of Vehicles driver behavior. IJnl’ortunately, speed is also difficult to measure, and for all years of our analysis, the speed of about 55 to 60 percent of the Page 54 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix III Drivrr-Related Statistics Figure 111.6:Female Driver Fatal Accident Rate by Age Group 175 Rate per Million Population 1975 1976 1977 1976 1979 1990 1961 1962 1963 1984 1966 1986 1967 - Under21 mm-. Age2160 - Age5165 ,‘.a OklerThm65 While the fatal accident rate for male drivers of nearly all age groups tends to follow the overall trend, the rate for females has increased rather steadily across all age groups, especially since 1982. The year 1987 showed the highest rate of involvement per million population for all female age groups. AMaledrivers are involved in fatal motor vehicle accidents about four Age and Gender of t,imes as often as femak drivers. (See table 111.2.)However, while the Driver number of male drivers involved in fatal accidents tends to follow the overall trend, the number of female drivers has been increasing since 1975. Nonetheless, the increase for female drivers since 1975 is not spread evenly over all age groups. Moreover, analyzing male drivers by age group shows that the overall trend also is not applicable to all age groups. The number of female drivers under age 16 involved in fatal accidents in 1987 was over 40 percent more than in 1975, while the number of malt dri\-crs this age was about 18 percent less than in 1976. In neither case were’ many drivers involved, however, so this trend exerted little influencx~ on the difference in the overall trend. Page 52 GAO PEMD-W-10 Highway Safety: Trends in Highway Fatalities 1975-87 Figure 111.4:Fatal Accident Rate for Drivers Older Than 65 175 Rate per Million Population im 146 140 1975 1976 1977 1978 1979 1960 1961 1962 1963 1964 1985 1966 1967 Involvement Rates by Age The distribution of fat a1 xcidertt involvement across age groups is dif- and Gender ferwt for male and fcSm;rlcdrivers. (Set figures III.5 and 111.6.)The rates for malts range from slightly over 300 fatal accidents per million popu- lation for those older t hatt 6.5 to a high of 900 to 1.100 per million for tlrc 1% to 20.year-old agesgroup. Females have the same high and low age groups, but the drtr,cit, involvement rates are less than 80 per million for the older femaks and from about 200 to 300 per million for the younger group. Page 50 GAO PEMI)-90 10 Highway Safety: Trends in Highway Fatalities 1975-X7 Appendix III Driver-Related Statistics Figure 111.2:Driver Fatal Accident Rates Related to 1975 Base Year by Gender’ 130 Percent of 1975 aas%Year Rate - Male ---- Female “Percentages are the relation 01 Ihe number of drivers involved in fatal accldents each year to the number of drivers wwolved in 1975 Involvement Rates by Age There is a wide divergence in driver involvement in fatal accidents across age groups. In 1987. fatal accidents ranged from fewer than 200 per million population for those older than 65 to a high of generally almost 600 per million for the 1% to 20-year-old age group. The involve- ment rate generally declines as drivers get older. (See table III. 1.) With some variations, trends for all age groups tend to follow the overall trend, the under- 16, 16 and 17.year-old, and over-M-year-old drivers being the principal exceptions. The interesting difference about the pat- tern for 16- and 17-year-olds is the increase in the rate of involvement since 1982. (See figure 111.3.)The rnte increased from about, 340 per mil- lion population in that year to about 440 per million in 1987, an increase of over 25 percent. After generally declining from 1975 through 1982, the rate of driver involvement in fatal accidents for those older than 65 increased about 16 percent from 1982 through 1987, 1987 having the highest involvement rate cm record. (See figure 111.4.) Page 48 GAO/PEMI)-W-10 Highway Safety: Treuds in Highway Fatalities 1975.87 Appendix III briver-Related Statistics As discussed in appendix I, motor vehicle safety legislation has been directed at three principal targets-the driver, the vehicle, and the driv- ing environment. The F.%RS system was designed to collect data on these three areas. In this appendix, we discuss information obtained from FARS that pertains to drivers. We will discuss such items as driver-involve- ment rates, age and gender of drivers, vehicle speed (where known), drinking drivers, and the use of safety restraints. - Analysis of driver involvement per million population gives a reasonable Driver-Involvement measure of whether there are real differences in fatal accident experi- Rates ence for various driver age and gender groups or whether they are merely a reflection of differences in population growth patterns.’ The trend for the overall driver-involvement rate follows the overall trend. Involvement Rates by The involvement of malt3 drivers per million population is generally four Gender to five times the female rate. (See figure 111.1.)The involvement of males overall tends to follow the overall trend, while the involvement of females has been on the increase, especially since 1982. The involve- ment rate of females scclmcd to be following the overall trend through 1982. but since then, the female driver involvement rate has increased over 20 percent, and 1987 is the highest involvement rate for females on record. (See figure III.‘?. ) Page 46 GAO, I’EMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix 11 General Fatal Accident Statistics Table 11.6: Pedestrian Fatalities by Age and Gender Gender Age group 1975 1976 1977 1970 1979 1980 1981 1982 1983 1984 1985 1986 1987 Male Under21 1,694 1,633 1,549 1.621 1621 1,535 1,339 1,272 1,163 1,149 1,070 1,097- 1,029 21~50 1.661 1.646 1760 1933 2138 2169 2291 2238 2.077 2.175 2.022 2.167 2141 Older than 50 1,831 1,786 1,930 1,841 1824 1,786 1689 1,534 1,452 1,597 1,484 1,418 1,542 Aaeunknown 60 55 92 99 118 123 140 100 83 95 95 a9 66 Total 5,196 5,120 5,3j; 5,494 5,701 5,613 5,459 5,144 4,775 5,016 4,671 4,771 4,778 Female Under21 877 874 852 821 777 725 684 651 578 560 571 521 495 21-50 498 515 532 589 640 689 668 723 618 582 629 619 636 Older than 50 921 892 976 855 927 989 923 777 834 829 904 a34 ais Aaeunknown 23 26 40 31 47 51 57 34 20 34 29 30 19 Total 2,319 2,307 2,400 2,296 2,391 2,454 2,332 2,185 2,050 2,005 2,133 2,004 1,966 Unknown 1 0 1 5 4 3 46 2 1 4 4 4 2 Total fatalities 7,516 7,427 7,732 7,795 8,096 8,070 7,837 7,331 6,826 7,025 6,808 6,779 6.746 Fatalities by Gender Like motor vehicle fatalities in general, pedestrian fatalities for males are consistently mart’ than twice the number for females. In 1987. malts accounted for 71 percent of these fatalities, while females accounted for only 29 percent. When age considerations are ignored, both male and female pedestrian fatalities have showed substantial declines since their peak years of 1979 and 1980. Fatalities by Age *Just as there were differences in the pedestrian fatality trends for male and female irrespective of age, there are also some differences in the trends by age irrespective of gender. We analyzed fatalities by three age categories-namely. ( 1 1under 21 years old, (2) ages 21-50, and (3) over age 50. In 1987, thcsc age categories accounted for 23, 41, and 35 per- cent of pedestrian fatalities, respectively. None of the age groups for pedestrian fatalities follows the overall t,rend very closely, but the pat- terns of divergence vary considerably. Pedestrian fatalities for persons younger than under age 2 1 have been steadily declining since 1975, only 1978 showing any imrease at all. Unlike for the younger age group, the year 1987 for persons age 21 through 50 is not the lowest fatality year. Pedestrian fatalities fog the 21. through SO-year-old age group show a rather steady pattern of increase from 1975 through 1981 but have decreased slightly since then. The 1987 total of almost 2,800 still exceeds the 1975 low y’ear-by over 30 percent. Pedestrian fatalities for Page 44 GAO ‘PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Statistics Figure 11.11: Fatalities for Ages 51-65 by Gendera 12 Percent ihviathm Fmm lOYear Average 10 8 a 4 2 0 -2 -4 4 -0 -10 12 -14 n Male Female ‘Percentages are normalized as percentage devlahons from the 13.year average of annual fatalltles The group older than 65 shows the greatest departure of all from the overall trend, especially for female fatalities. After varying little from 1975 through 1983, these increased about 25 percent from 1983 to 1987 and to a level over 30-percent higher than 1975. The percentage increase in male fatalities since 1982 was not nearly as large. (See figure 11.12.) Page 42 GAO:PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix II General Fatal Accident Statistics Figure 11.9: Fatalities for Ages 21-25 by GendeP 16 Percent Deviation From 13-Year Average 12 -16 ‘Percentages are normalbed as percentage dwatlons from the 13 year average of annual fatalltles The 26-50 group showed similar experiences for both male and female until 1980 or 1981. Thereafter, the difference was such that female fatalities reached a peak in 1987 that was about 13-percent higher than the previous peak of 1980; male fatalities in 1987 were still below the previous peak year. (See figure 11.10.) Page 40 GAOjPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendix II General Fatal Accident Statistics Figure 11.7:Fatalities for Ages 16-17 by Gendera 30 Petcent Devlallon From X&Year Average 25 20 15 10 5 0 -5 -10 -15 -20 -25 ‘Percentages are normalized as percentage dwatlons from the 13~year average of annual fatalltles The group 18-20 shows similar trends for both genders but does not show the marked increases since 1983 that are characteristic of the overall trend. The principal difference between male and female is that while female fatalities continued to decline after 1985, male fatalities increased slightly after 1985. (See figure 11.8.) Page 38 GAO, PEMD-9010 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II Genrrzd Fatal Accident Statistics Other Vehicles One-vehicle fatal accidents involving other vehicles declined steadily from 1975 through 1981.” In 1982, one-vehicle accidents involving other vehicles more than doubled, principally because of a dramatic increase in the number of vehicles reported with unknown body types. Since 1982, the number of other vehicles in one-vehicle accidents declined, reaching a low in 1987. In 1987, other vehicles accounted for about 3 percent of the vehicles involved in one-vehicle fatal accidents. Who is being killed in motor vehicle accidents and are there differing The Age and Gender trends when fatalities are analyzed by age and gender? Table II.5 shows of Fatalities the distribution of motor vehicle accident fatalities by age and gender. Table 11.5: Fatalities by Age and Gender Gender Age group 1975 1976 1977 1976 1979 1960 1981 1982 1983 1984 1985 1986 1987 Male Under16 3,336 3,183 3 065 3,123 2,943 2,802 2,465 2,269 2,251 2,257 2,274 2,391 2,460 16-17 2,161 2,299 2 386 2,504 2,452 2,258 2,030 1,722 1,567 1,580 1.532 1,861 1,780 18-20 4,781 4,895 5221 5,387 5,546 5430 4,775 4,268 3,901 3,990 3,592 3,918 3,690 21-25 5,513 5,730 6251 6,765 7,089 7 176 6,883 6,134 5,700 5,894 5,876 6,033 5,608 26-50 9,941 9,932 10566 11,730 12,352 12,646 12,808 11,291 11,011 11,425 11.370 12,233 12.449 51-65 3,673 3,786 3865 3,858 3,876 3,623 3,689 3,232 3,099 3,200 3,141 3,078 3,272 Older than 65 3,135 3,087 3 082 3,161 3,097 - 3,033 3,039 2,790 2,831 3,012 2,997 3,175 3.262 Unknownage 135 126 191 203 233 224 241 173 139 183 177 167 107 Total 32,675 33,038 34,627 36,731 37,588 37,392 35,930 31,879 30,499 31,541 30,959 32,856 32,628 Female Under16 1,899 1,921 1966 1,893 1,789 1,700 1,515 1,438 1,336 1,367 1,438 1,448 1,444 16~17 790 8% 980 973 999 996 807 658 680 715 716 856 896 18-20 1.240 1,437 1514 1,608 1,536 1,496 1,461 1,334 1,264 1,242 1,237 1,233 1,199 21~25 1,311 1,387 1,663 1.817 1,748 ~~ 1,844 1,801 1,582 1,560 1,650 1,623 1,623 1,643 26-50 3,065 3,101 3.309 3,545 3,750 3.876 3,819 3,538 3,633 3,872 3,827 3,933 4,384 51-65 1,606 1,688 1765 1,752 1,677 1,712 1,706 1.502 1,587 1,633 1,657 1,634 1,652 Older than 65 1,865 1,984 1,951 1,917 1,913 1,965 1,954 1,942 1,977 2,171 2,289 2,399 2,489 Unknownage 68 64 100 86 86 - 100 98 68 47 59 69 65 36 Total 11,844 12,481 13,248 13,591 13,498 13,689 13,161 12,062 12,084 12,709 12,856 13,191 13,743 Unknown 6 4 3 9 7 10 210 4 6 7 10 40 15 Total fatalities 44,525 45,523 47,878 50,331 51,093 51,091 49,301 43,945 42,589 44,257 43,825 46,087 46,386 ‘Includes such vehicles as snl,wmobilfs,nontruck farm equipment, all-terrain vehicles, nontruck con- struction rquipment, go carts, fork lifts, city strut sweepers, and unknown other vehicles Page 36 GAO/P&MD4@10 Highway Safety: Trends in Highway Fatalities 1975-37 Appendix II General Fatal Accident Statistics It is very apparent that the principal reason for the increase in truck involvement in one-vehicle fatal accidents was the increasing involve- ment of light trucks and vans.:’ In fact, except for 1983 and 1984, the involvement of medium and heavy trucks in one-vehicle fatal accidents has been on the decline since 1979. One-vehicle fatal accidents involving automobiles have also been generally declining since 1980, with the exception of increases in 1984 and 1986. One-Vehicle Accidents and While there are no unusual departures from the overall trend for total Automobile Size automobile involvement. since 1983, automobiles in one-car fatal acci- dents by size of automobile give a different picture.’ While the number of small cars involved in such accidents is on the increase, the involve- ment of full-size cars is on the decrease. (See figure 11.6.) The involve- ment of intermediate cars in one-vehicle fatal accidents increased almost 100 percent between 1976 and 1980, but the number has declined more than 10 percent since then. In 1987, almost 50 percent of the automo- biles involved in onevchicle fatal accidents were small cars; about 15 percent were intermediate cars, and about 27 percent were full-sized cars. Page 34 GAO, PEMD-SO-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Statistics Table 11.2: Fatalities bv Role Role -1975 1976 1977 1976 1979 1960 1961 1962 1963 1964 1965 1966 1967 DWer 23,652 24,500 26,170 28,283 28,863 28.816 28,200 24,690 24,138 25,589 25,337 26,630 26,831 Passenger 12,169 12,497 12,873 13,108 12,964 12972 12,055 10,867 10,595 10,586 10,619 11,498 11,618 Pedestrian 7,516 7,427 7,732 7,795 8,096 8,070 7,037 7,331 6,826 7,025 6,808 6779 6,746 All others 1,188 1,099 1,103 1,145 1,170 1.233 1,209 1,057 1,030 1,057 1,061 1,180 1,191 Total fatalities 44,525 45,523 47,676 50,331 51,093 51,091 49,301 43,945 42,569 44,527 43,625 46,067 46,366 Just as most fatal motor vehicle accidents result in only one fatality, Fatal Accidents by most fatal motor vehicle accidents also involve only one vehicle. (See Number of Vehicles table 11.3.) In 1987, one-vehicle accidents accounted for 58 percent of the Involved accidents involving fatalities. Two-vehicle accidents accounted for 36 percent, accidents with three or more vehicles about 6 percent. One- vehicle fatal accidents have consistently exceeded the second largest number-two-vehicle fatal accidents-by over 60 percent and have con- sistently exceeded all multivehicle accidents combined by over 40 percent. Table 11.3: Fatal Accidents by Number of Vehicles Involved -- Number of vehicles 1975 1976 1977 1976 1979 1980 1981 1982 1983 1984 1985 1986 1987 None’ 567 618 668 752 886 862 934 0 0 0 0 0 0 Ge- 23,651 24,100 24,890 25.768 26,374 27 424 25,898 23,851 23,048 23,697 22,875 24,275 24,159 Two 13,375 13,465 14,866 15,894 16,013 15,301 15,283 13,573 13,420 14,066 14,249 14,568 14,909 Three 1,290 1,322 1,445 1,681 1,604 1 392 1,533 1,365 1,371 1,520 1.675 1,816 1913 Four 210 186 241 248 254 215 263 226 223 243 287 311 317 Five 45 33 72 60 61 49 53 55 61 68 71 82 93 More than fwe 23 23 29 30 31 21- 36 22 33 37 39 38 44 Total accidents 39,161 39,747 42,211 44,433 45,223 45,284 44,000 39,092 37,976 39,631 39,196 41,090 41,435 "From 1975 through 1981 the FARS data flies vx3uded cases I” which no record of the number of vehicles was reported These ::ases show zero vehicles in accidents and number fewer than 1,000 per year However, there are differences in the trends for one-vehicle versus mul- tivehicle accidents. One- and two-vehicle fatal accidents show similar trend directions. but two-vehicle accidents have higher percentage swings. Accidents with three or more vehicles increased about 50 per- cent from 1975 though 1987. What is interesting about these accidents is that while they showed trend patterns similar to one- and two-vehicle fatal accidents in the early years, fatal accidents involving three or more Page 32 GAO!PEMD-W-10 Highway Safety: Trends in Highway Fatalities 197547 Appendix I1 General Fatal Accident Statistics older than 65 increased from 189 to 208 per million population, an increase of about 10 percent. Fatality rates by age and gender taken together show that for all age groups, the rate for males is significantly higher than that for females. In fact, the fatality rate for males is a minimum of about 50 percent higher than that for females and, in many cases, is close to three times the rate for females. The fatality rates for males 16 through 20 are of special concern. The rate for 16- and l7-year-old males has never been lower than 410 per million population, and in 1978, it exceeded 570. The rat,e for 1% to 20-year-old males has never been below 600 per million population and in 1979 it, was almost 840. The rates for females in the same age group never exceeded 250 per million population. The fatality rates for both males and females younger than 16 have declined dramat- ically over the years, even though they have increased slightly since 1983. Nearly all the other age and gender fatality rate trends approximate the overall trend, but thrrc exceptions related to females deserve mention. The female fatality rate for 16- and 17-year-olds has strongly influenced the overall rate for this age group since 1983. The fatality rate for females of this age group increased from a little over 170 per million population in 1982 to over 240 per million in 1987, an increase of almost 40 percent. The 1987 rate was exceeded only by the 1980 rate, but 1980 and 1987 were essentially the same. The experience for males of this age group is not nearly as dramatic. After reaching a low rate of 411 per million in 1983, thtl rate for males increased only a little over 10 percent through 1987. The fatality rate for females 51 through 65, while small in comparison to others, has increased almost 13 percent since 1982, and the rate for females older than 65 has increased about 17 percent since 1983. Increases for males of these age groups were only about 7 percent and 5 percent. respcbctively. As mentioned earlier. the number of motor vehicle fatalities exceeds the Fatalities Per Fatal number of fatal motor vehicle accidents. To determine the effect that Accident multifatality accidents exert on the overall number of motor vehicle fatalities, we analyzed both the average numbers of deaths per accident and the trends for various numbers of deaths per accident. Although there has been a genciral decline in the average number of fatalities per accident since 1976. r.hr decline is somewhat misleading, since the rate varies so little in amount, from the smallest to the largest. Most fatal accidents clearly ha\~ only one fatality. In 1987, over 90 percent of the Page 30 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 Appendii II General Fatal Accident Statistics Figure 11.4: Fatalities Per Million Population 230 Fatality Rate 220 210 290 199 180 1975 1976 1977 1979 1979 1980 1981 1992 1983 1984 1985 1996 1987 Page 28 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Statistics accident rate per 100 million miles driven shows a continuing decline since 1980, fatal accidents per 10,000 vehicles and per 10,000 drivers show mixed patterns of increase and decrease since 1983. (See figure 11.2.) The rate per 10,000 vehicles increased in 1984, decreased in 1985, increased again in 1986, and then decreased again in 1987. While the 1985 rate was lower than that for 1983, both the 1986 and 1987 rates were higher. The fatal accident rate per 10,000 drivers has generally been on the increase since 1983, with the exception of slight decreases in 1985 and 1987. Figure 11.2:Fatal Accident Rate Trends 4.0 Rate - Rate per 10,000 Vehides ---- Rate per 10,000 Driven m Rate per 100 Million Miles Driven Source Park data lrom Nai ondl Safety Council, Acadent Facts 1988 EdItIon (Chicago, III 1988) When only fatalities in fatal accidents are considered, similar patterns Fatality Rate Trends occur. (See figure 11.3.) The fatality rate per 100 million miles driven has continued to decline, w?th the exception of a small increase in 1986, while the rates per 10.000 vehicles and per 10,000 drivers show mixed increase-decrease l)att(rns after 1983. Page 26 GAO PEMD-90.10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix II General Fatal Accident Statistics Much has been written about the number of motor-vehicle-related fatali- Introduction ties over the past several years. The FAKS system enables us to look at these fatalities and to study them in some depth. Data are available for 1975 through 1987. In this appendix, we discuss general fatal accident statistics that are not necessarily related to the specific elements of motor vehicle safety discussed in appendix I-namely, the driver, the vehicle, and the highway environment. We discuss the overall trend in motor-vehicle-related fatalities, some special aspects of the overall trend such as the roles of the persons killed (driver, passenger, pedestrian, and so on), and certain fatality rate information related to general meas- ures of exposure to motor vehicle accidents. Since 1975, the number of fatal motor vehicle accidents, the number of fatalities, and the number of motor vehicles involved in fatal accidents have reflected very similar patterns. All three trends show steady growth from 1975 through 1978 or 1979, followed by substantial declines through 1983 and then growth again through 1987. (See figure II. 1.) The number of fatal motor vehicle accidents grew from about 39,000 in 1975 to about 45,000 in 1980-81 and then fell to about 38,000 accidents in 1983. Sincr, 1983, the number of fatal accidents has again been on the increase. growing to about 41,000 in 1987. The number of fatalities grew from about 44,500 in 1975 to a high of about 51,000 in 1980 and 1981 before falling to about 43,000 in 1983. After 1983, the number of fatalities rose to about 46,500 in 1987. The number of vehi- cles involved in fatal accidents increased from about 56,000 in 1975 to about 65,000 in 1978 and 1979, fell to about 55,000 in 1983, and then rose again to about Ci2.000 in 1987.’ Page 24 GAO PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Appendix I Background Vehicle Manufacturers Association publications, and vehicle registration data came from computer files maintained by NITSA.' Our review was made in accordance with generally accepted govern- ment auditing standards and included the tests we considered necessary to assure ourselves of the reliability of the FARS computer-based data. ____~ Fundamental to the purposes of NH'I‘SA is the collection of accident data Accident Data Sources that can identify safety problems, suggest solutions, and provide an ob.jective basis for evaluating the effectiveness of motor vehicle safety standards and highway safety countermeasures. To this end, NHTSA has developed and used various systems to collect data on motor vehicle accidents. The principal data collection systems NHTSA uses are FARS and the National Accident Sampling System (NASS). In this report, we use FARS data-created to analyze fatal accidents-for our analysis, since the sampling errors associated with NASS data are often too large to dis- ccrn whether apparent trends over time are legitimate. FAKS was conceived, designed, and developed by the National Center for Statistics and Analysis of NFITSA to provide, through data on fatalities, an overall measure of highway safety. The system was also intended to identify traffic safety problems, suggest solutions, and help provide an otjjective basis for evaluating the effectiveness of motor vehicle safety standards and highway safety programs. An agency official commenting on a draft of this report stressed that FARS is an outcome data base, since a fatality must have occurred before an accident qualifies for inclusion. This official also stressed that because FARS is an outcome data base, it, must be used with care to avoid overgeneralizing to all accidents from data biased toward fatalities. FEARSis a census of all fatal motor vehicle accidents occurring throughout the nation. After 3 years of system development, FAKS became operational for calen- dar year 1975, when the 50 states, Puerto Rico. and the District of Columbia began collecting and assimilating fatal highway accident data. 13-4~ data differ from some other fatal accident statistics such as those of the National Safety C’ouncil because of the definitions used for fatal accidents. To qualify for inclusion in FARS, an accident must involve a motor vehicle traveling on a road customarily open to the public and it must have resulted in the death of a person-whether an occupant of a Appendix I Background Table 1.2: Summary of Legislative Safety Concerns Area Statute (Public Law) Year Vehicle PromotIon of safety in manufacturing vehicles 84-627 1956 Brake fltld standards ~~ 87-637 1962 Seat belt standards 88-201 1963 l&or vehicle safety standards 89-563 1966 Rzatlonship between equpment performance 89-563 1966 and accidents and fnju y Vehicle reglstratlon, operatlon and&spectlon 89-564 1966 Fuel economy standards 94-163 1975 Child safety seats 68-263 1984 Length and width of Irrxks 98-554 1984 Inspection of trucks 98-554 1984 Driver study of speed llmlt enforcement and need for i4-627 ~1956 uniform safety and speed laws Driver registry 86-660 -1960 Expanded driver registry 89-563 1966 Study of relatIonshIp betaeer alcohol 89-564 ~- Ii66 consumption and highflay safety Improvement of driver performance 89-564 1966 NatIonal speed Ilrnit 93-239 and 93-643 1974 NatIonal minimum drlnklrig age 98-363 1984 Environment Study of need for uniform speed llrnlt 84-627 1956 Study of design and characterlshcs of hlrjhiiy 84-627 1956 HIghway design and maintenance 89-564 1966 Surveillance to find ihigh accident locatlons 89-564 1966 Harardousmater~als 93-633 and 98-559 1975, 1984 While safe vehicks are important to motor vehicle safety, research con- ducted after the 1%X primary safety legislation on motor vehicles sug- gests that other factors may be more important. In a 1979 Indiana tTniversity study. pc>rformed under a contract from KHTSA, the research- ers concluded that human factors were the leading causes of automobile accidents (93 pclrcc’nt ), followed by highway environment (34 percent) and vehicle factors ( 13 percent).’ Most of the vehicle-related factors involved vchiclc deterioration rather than vehicle design. and they could have been avoidcsd with proper inspection and maintenance procedures, Page 20 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalitirs 197647 Appendix I Background The Kational Safety Council reports that, since 1948, there have been almost 100,000 accidental deaths per year and, on the average, almost half of those deaths resulted from motor vehicle accidents.’ Motor vehi- cle accidents are the leading cause of accidental death overall and the leading cause of accidental death for persons age 1 to 74. For persons 75 and older, motor vehicle accidents are exceeded only by deaths resulting from falls. Deaths from motor vehicle accidents are a special problem for youths. In 1984, for persons 15-24, almost 40 percent of deaths from all causes resulted from motor vehicle accidents, and motor vehicle acci- dents accounted for almost three fourths of all accidental deaths for that age group (see table 1.1). Table 1.1: Relationship of Accidental Deaths to Total Deaths in 1984 Total deaths Motor vehicle deaths as from motor a percent of Total deaths from ~ Percent of total vehicle Accidental Total Age group All causes All accidents deaths accidents deaths deaths Under1 year 39.580 838 2 12% 161 1921% 041% 1-4 7372 2814 38 17 977 3472 1325 514 9.076 4 198 46 25 2,263 5391 2493 15-24 38,817 19,801 51 01 14738 7443 3797 2544 112,484 25,498 2267 15.036 58 97 1337 45-64 404,568 15,273 378 6.954 4553 172 6574 476570 8,424 177 3,020 3585 063 75 and older 950.902 16065 1 69 3,114 1938 033 Total deaths 2,039,369 92,911 4.56% 46,263 49.79% 2.27% The National Safety Council has gathered statistics on deaths from motor vehicle accidents as far back as 1913. There has been a rather steady climb in the number of such deaths since then, the years since World War II showing especially large losses. At the same time, how- ever, there has been a steady increase in the numbers of drivers and motor vehicles on the, nation’s highways and a like increase in the number of miles those drivers and vehicles travel each year. Conse- quently, the rate of motor vehicle deaths-whether related to drivers, vehicles, or miles travr%led-has generally been declining. Nevertheless, the rate of decline has not kept pace with the rate of decline for other types of accidental d&h, and the sheer number of deaths from motor vchiclc accidents r,;rc,hyear is still a matter for national concern. Page 1R GAO PEED-90.10 Highway Safely: Trrnds in Highway Fatalities 1975.87 Figure V.6: Vehicles Involved in Bridge Accidents 93 Figure V.7: Fatal Accidents by Day of the Week 94 Figure V.8: Fatal Accidents by the Time of Day 96 Figure V.9: Fatal Accidents by Season of the Year 97 Abbreviations DUT Department of Transportation FA’AKS Fatal Accident, Keporting System w.0 L1.S.General Accounting Office h-ASS National Accident Sampling System UHTSA National Highway Traffic Safety Administration Page 16 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 Table IV. 1: Vehicles Involved in Fatal Accidents 78 Table IV.2: Fatalities by Type of Vehicle 79 Table V. 1: Fatal Accidents by Road Surface Condition 87 Table V.2: Fatal Accidents by Type of Traffic Control 89 Table V.3: Fatal Accidents by Type of Weather Condition 93 Table V.4: Vehicle Trips and Miles in 1983 by Day of the 95 Week Table V.5: Relationship of 1983 Accidents to Vehicle Miles 96 Traveled by Time of Day Table V.6: Relationship of 1983 Accidents to Vehicle Miles 97 Traveled by Swson Figures Figure II. 1:Kumber of Awidents, Vehicles, and Deaths in 25 Fatal Accidents Figure 11.2: Fatal Awident Kate Trends 26 Figure 11.3: Fatality Rate Trends 27 Figure 11.4:Fatalities Per Million Population 28 Figure 11.5:Fatalities Per Million Population by Gender 29 Figure 11.6:One-Vc3hic.k Fatal Automobile Accidents by 35 Automobile Siw Figure 11.7:Fatalities for Ages 16-17 by Gender 38 Figure 11.8:Fatalit iw for Ages 18-20 by Gender 39 Figure 11.9:Fatalities t’or Ages 21-25 by Gender 40 Figure II. 10: Fatali t its for Ages 26-50 by Gender 41 Figure II. 11: Fatalities for Ages 51-65 by Gender 42 Figure 11.12: Fatalit,ios for Ages 65 and Older by Gender 43 Figure III. 1: Drivw Fatal Accident Rates by Gender 47 Figure 111.2:Driver Fatal Accident Rates Related to 1975 48 Base Year by Gctndcr Figure 111.3:Fatal I1c’cidtwt Rate for Drivers 16 and 17 49 Figure 111.4:Fatal .icxidwt Rate for Drivers Older Than 50 65 Figure 111.5:Male IJriver Fatal Accident Rate by Age 51 Group Figure 111.6:Fern&s Driver Fatal Accident Rate by Age 52 Group Figure 111.7:Speed of’ Vehicles in Fatal Accidents 55 Figure 111.8:Drinking Drivers Reported in FARS 57 Compared to Single-Vehicle Nighttime Accidents Figure 111.9:,Motor \‘&iclc Occupants in Fatal Accidents 58 Reported 1Tsing S;lt’cty Restraints Page 14 GAO PEMWSO-10 Highway Safely: Trends in Highway Fatalities 1975.H7 Contents Letter 1 Appendix I 18 Background Legislative History 19 Objectives, Scope, and Methodology 21 Accident Data Sources 22 The Contribution of This Report 23 Appendix II 24 General Fatal Accident Introduction 24 Fatal Accident Rate Trends 25 Statistics Fatality Rate Trends 26 Fatality Rates Per Million Population 27 Fatalities Per Fatal Accident 30 Fatalities by Person’s Role 31 Fatal Accidents by Number of Vehicles Involved 32 The Age and Gender of Fatalities 36 Pedestrian Fatalities 43 Conclusions 45 Appendix III 46 Driver-Related Driver-Involvement Rates 46 Age and Gender of Driver 52 Statistics Speed of Vehicles 54 Drinking Drivers 55 The ITse of Safety Restraint,s 57 Conclusions 62 - Appendix IV 64 Vehicle-Related Fatal Accident Involvement Rates by Type and Size of 64 Vehicle Statistics Fatality Rate by Type and Size of Vehicle 69 The Age of Vehicles Involved in Fatal Accidents 74 The Types of Vehicles Involved in Fatal Accidents 77 Fatalities by Type of Vehicle Involved 79 Vehicle Tires and Fatal Accidents 80 Vehicle Fatalities and Collisions 81 Conclusions 84 Page 12 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 B237223 If you have any questions or would like additional information, please call me at (202) 2751854 or Dr. Michael J. Wargo, Director of Program Evaluation in Physical Systems Areas, at (202) 275-3092. Other major contributors to this report are listed in appendix VI. Sincerely yours, Eleanor Chelimskl Assistant Comptroller General Page 10 GAO/PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-87 R-237223 and freeway signs, roadside hazards, and narrow bridges. The number of fatal accidents related to roadside hazards and narrow bridges appears to be steadily declining. Accidents where only “yield’‘-type traf- fic controls exist are on the increase. Freeway accidents have increased, especially where no traffic controls are present (more than a 15percent increase since 1982). but fatal accidents have increased most on local, county, and other roads. We were not, however, able to adjust these data by exposure measures. because we could not obtain annual data reflect- ing any changes in travel patterns on these roads. Table 4 highlights the environment-related statistics that show increases of 20 percent or more, either from the 1975 base year or from the low year associated with the upturn in the overall trend in 1982 or 1983. Additional envi- ronment-related trend statistics are contained in appendix V. Table 4: Highlights of Environment- Related Fatal Accident Statistic& Percent increase in 1987 Variable Over 1975 Over 1982-03 LImIted access roadway 25 40 Sto~&G 22 10 Yield traffvc controls 43 42 Some freeway sign controls 30 38 SIG 48 75 DaYtIme ~- 20 10 ‘Blank cells ~nd~calethat the rate of change did not exceed 20 percent ’ Base year IS 1981 With regard to specific Subcommittee concerns, our review disclosed the following. Narrow bridges have not been a factor in many fatal accidents, and the number of such accidents has been steadily decreasing. Accidents tend to follow the overall trend, whether or not traffic con- trols are present; only roadways involving yield signs show increases in fatal accidents. Most accidents occur where no traffic controls are pre- sent. Accidents where existing traffic controls were not functioning have always been small in number. Over 80 percent of all fatal accidents occur on dry roads. Accidents on both wet and dry roads tend to follow the overall trend. Page 8 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975.87 R-237223 heavy trucks have one of the lowest fatality rates for vehicle occupams. they have one of the highest fatal accident involvement rates. The average age of both automobiles and trucks has been increasing as it applies to both t,hc number on the road and the age of vehicles involved in fatal accidents. However, in 1976 through 1987, the period of our study, vehicle age appears to have become less a factor in fatal accidents for automobiles than for trucks. The only accidents that show trends different from the overall trend are rear-end collisions and noncollision accidents (such as overturns and fire, where no impact to the vehicle occurs throughout an accident). Both have increased more than 25 percent since 1982. Table 3 highlights the vehicle-related statistics that, by 1987, showed increases of 20 percent or more, either from the 1975 base year or from the low year associated with the upturn in the overall trend in 1982 or 1983. Appendix I\’ c*ontains additional information on vehicle-related t,rends. Page I3 GAOIPEMD-90-10 Highway Safety: Trends in Highway Fatalities 1975-W R-237223 Table 1: Highlights of General Fatal Accident Statisticsa Percent increase in 1987 Variable Over 1975 Over 1982-83 Number of vehicles involved Three 48 29 40 15 Four 50 95 42 15 Five 10667 69 09 More than fl;e 91 30 100 00 One-vehicle accidents Mlnlslze automoblles 13845 1555 Subcompact automoblles 171 79 39 37 Compact automoblles 61737 19507 T&al small automoblles 221 97 54 95 lntermedlate automobIles -69 07 V&based light trucks 49 29 2631 Conventional light trucks 56 35 26 20 Total light trucks 55 29 26 21 Total trucks 40 68 Motorcycles 37 23 ‘Blank cells Indicate that the pwcentage of change did not exceed 20 percent In addition to our analysis of overall fatality trends, we examined trends by various accident factors. We looked at factors associated with drivers (by age, gender, and use of safety restraints), vehicles (by type and size), and the roadway driving environment (by time of day and weather conditions). We found that the overall trend applies to many of the driver-related statistics discussed. One of the most revealing of these trends is the changing relationship between fatalities and safety restraint usage. Increased motor vehicle safety restraint use since 1979- 80 appears to have saved t,he lives of many drivers and passengers. Fur- ther, the percentage of occupants not - using safety restraints who were killed has continued to inc,rease. Our analysis also shows that the rate of involvement of women drivers in fatal accidents has increased more than 20 percent since 1975 and that this rate of incrclas;e applies to nearly all female age groups. Drinking drivers arc’ a very serious traffic safety problem. However, a NITSA official pointed out that when analytical procedures are used to adjust for the large amount of missing data regarding the presence of drinking drivers, the, t~sults indicate that the percentage of drinking drivers in fatal accidents has been declining. Page 4 GAO. PEMD-90-10 Highway Safety: Trends in Highway Fatalities 197587 B-237223 of such deaths has climbed steadily, the years since World War II show- ing especially large losses. At the same time, however, there has been a steady increase in the number of drivers and motor vehicles on the nation’s highways and a similar increase in the number of miles those drivers and vehicles travel each year. Consequently, the rate of motor vehicle deaths-whether related to drivers, vehicles, or miles trav- eled-has generally been declining. Nevertheless, the decline has not kept pace with the decline for other types of accidental death, and the absolute number of deaths from motor vehicle accidents each year is still a matter for national concern. Although we identified many instances of automobile safety research- including numerous SKI% studies using FARS-Very little of that research discussed the changes in the characteristics of fatal accident statistics. Moreover. studies have compared specific fatal accident char- acteristics-such as trucks versus cars or male versus female drivcrs- but these studies have tended to have a narrow focus. In addition, because of the difficult,y in obtaining accident-exposure information. only small amounts of information are available that compare accident fatality statistics to various measures of accident exposure, such as vehicle miles traveled, number of registered vehicles, or number of driv- ers. A NIITSA official’s comments on a draft of this report stressed that the lack of good exposure data is one of the ma,jor impediments to high way safety analysis. Our report attempts to fill some of these gaps by presenting information that is (1) based on fatal accident trends, (2) extensive in the accident characteristics discussed, and (3) related, where possible, to measures of accident cxposurc. A complete discussion of the legislative history, accident data sources, and our scope and meth- odology is contained in appendix I. Our analysis of fatal accident data from 1975 through 1987 indicated an Our Analysis increase in such accidents from 1975 to 1980, a decrease through 1982- 83, and then an increase again through 1987. We refer throughout to this general increasing-decreasing-increasing pattern of fatalities as the overall trend. This pattern applies to many of the general fatality statis- tics we present, and, in all cases, it serves as a convenient yardstick for comparison, so that our discussion concentrates principally on patterns that depart from t 1~~overall trend, with particular attention to depar- tures that indicattx a worsening situation. In particular, we highlight trends that, by 1987, had resulted in (1) a greater than ZO-percent increase compared to 197.5 or (2) a greater than 20-percent increase Paye 2 GAO, PEMD-90-10 Highway Safety: Trends in Highway Fatalities 197547
Highway Safety: Trends in Highway Fatalities 1975-1987
Published by the Government Accountability Office on 1990-03-09.
Below is a raw (and likely hideous) rendition of the original report. (PDF)