o 1 *!^ NOVEMBER NUMBER 25 ACCIDENT DRIVERS OVER A FIVE YEAR CHANGING CHARACTERISTICS OF HIGH PERIOD

Transcription

1 CHANGING CHARACTERISTICS OF HIGH ACCIDENT DRIVERS OVER A FIVE YEAR PERIOD NOVEMBER NUMBER 25 1 *!^ JOHN E. GOODSON HAROLD L MICHAEL JOINT HIGHWAY RESEARCH PROJECT PURDUE UNIVERSITY AND INDIANA STATE HIGHWAY COMMISSION

2

3 Technical Paper CHANGING CHARACTERISTICS OF HIGH ACCIDENT DRIVERS OVER A FIVE YEAR PERIOD TO: J. F. McLaughlin, Directr Jint Highway Research Prject FROM; H. L. Michael, Assciate Directr Jint Highway Research Prject Nvember 1, 1973 Prject: C File: The attached Technical Paper "Changing Characteristics f High Accident Drivers Over a Five Year Perid" has been authred by Mr. Jhn E. Gdsn, Graduate Assistant in Research n ur staff and Prfessr H. L. Michael, Assciate Directr. The Paper was presented at the 1973 Annual Purdue Rad Schl and will be published in the Prceedings. The Paper is a summary f the research reprt f the same title presented t the Bard previusly as a Final Reprt n a JHRP Study. It is presented t the Bard nw fr apprval f publicatin. Respectfully submitted, Harld L. Assciate Michael Directr HLM:ms cc: W. L. Dlch R. L. Eskew G. D. Gibsn W. H. Getz M. J. Gutzwiller G. K. Hallck R. H. Harrell M. L. Hayes C. W. Lvell G. W. Marks R. D. Miles G. T. Satterly C. F. Schler M. B. Sctt J. A. Spner N. W. Steinkamp H. R. J. Walsh E. J. Yder

4 Technical Paper CHANGING CHARACTERISTICS OF HIGH ACCIDENT DRIVERS OVER A FIVE YEAR PERIOD by Jhn E. Gdsn Graduate Assistant in Research and Harld L. Michael Assciate Directr Jint Highway Research Prject Prject N. : C File N. : Cnducted by Jint Highway Research Prject Engineering Experiment Statin Purdue University In Cperatin with Indiana State Highway Cmmissin Purdue University West Lafayette, Indiana Nvember 1, 1973

5 ABSTRACT The purpse f this research was t determine the number and characteristics f drivers wh had a high accident invlvement during a particular year. Then, using infrmatin available in the accident and driver recrd files, the changes in their characteristics and accident invlvement during the fllwing fur year perid were determined. Accident recrd summaries were cllected fr the years 1967 thrugh 1971 fr all drivers wh were identified as having been invlved in three r mre accidents during the year Driver recrd summaries were cllected fr drivers invlved in three r mre accidents during the year 1967 and a ttal f five r mre accidents during the entire five year study perid. Included in the final analysis were 776 drivers fr which bth driver and accident recrd summaries had been btained and an additinal 1288 drivers fr which nly accident recrds were cllected. It was fund that the accident recrds f mst f the drivers studied exhibited much imprvement during the fur year perid fllwing the base year. Apprximately frty percent f the drivers wh were invlved in three r mre accidents during the base year were identified as having n additinal reprted accidents during the fllwing fur year perid.

6 Using nly infrmatin available in the accident and driver recrd files, it was fund that it is difficult t predict, with any degree f accuracy, which drivers wh were high accident drivers during a particular year wuld cntinue t be excessive accident repeaters during fllwing years. Tw characteristics studied that appeared t have the greatest value as accident predictrs were the number f accidents during a year and the type f license. Fr the average driver f a high accident grup, it was fund that as the number f accidents in which he was invlved during the base year increased, the prbability f his being invlved in additinal accidents during the fllwing years als increased, Persns with chauffeur's licenses and particularly public passenger chauffeur's licenses were fund t ccupy a much greater prprtin f the high accident grups than their prprtins as licensed drivers wuld indicate. In general, the accident recrd f a driver during ne year, even when cmbined with infrmatin available in the Indiana Drivers Recrds, des nt enable a reasnable predictin f the drivers accident recrd fr future years.

7 Digitized by the Internet Archive in 2011 with funding frm LYRASIS members and Slan Fundatin; Indiana Department f Transprtatin

8 CHANGING CHARACTERISTICS OF HIGH ACCIDENT DRIVERS OVER A FIVE YEAR PERIOD Intrductin In rder t develp an effective accident preventin prgram it is necessary t knw the characteristics f all f the system cmpnents invlved: the envirnment, the vehicle, and the driver. The characteristics that are least understd and the mst difficult t predict are thse f the individual driver, and it is the driver wh is held respnsible fr the great majrity f traffic accidents. Varius research reprts in this field have presented cntradictry statements regarding the characteristics f drivers invlved in accidents. Fr example, sme shw that traffic accidents happen t almst everyne, while thers indicate that mst accidents are the result f vluntary vilatin f traffic laws, and are subject t crrelatin with the issuance f citatins. Evidence given by Crancer (3,4,5) indicates that it is the nn-average driver that is invlved in fatal accidents, while Gldstein (6) states that mst accidents ccur t nrmal, average peple. Each year a small percentage f drivers are invlved in an excessive number f accidents. Very little is knwn abut the characteristics f these drivers r hw their accident histries vary ver perids f time. In a 1946 reprt by Jhnsn (7), it is stated that "Operatrs wh are accidentrepeaters in ne perid tend t regress tward the average f the grup in anther perid." This temprary nature f

9 accident prneness is further reinfrced by mre recent reprts f Schuster and Guilfrd (10) and Campbell (1), but is cntradicted by Shaw and Sichel (12) wh say, "...The man whse past recrd has been unsatisfactry is unlikely t imprve, punishing him will have very little effect, and retraining him means thrwing gd mney after bad. The nly practical way f insuring that he des nt have accidents is t prevent him frm driving." It is recgnized that ne methd f pssibly reducing accidents wuld be t restrict the driving liberty f thse wh culd be classified as accident repeaters. But t prvide such restrictins, gd evidence f the benefits t sciety will be necessary. It wuld be desirable t knw the answer t several questins, fr example; Why d sme individuals have mre accidents than ther drivers? Des an individual's relative accident invlvement rate in ne year remain the same in succeeding years r des it increase r decrease? If imprvements r deteriratins are nted, what are the causative factrs f these changes? In 1967 there were abut 2.7 millin drivers licensed in the State f Indiana. During that year 301,921 drivers were invlved in 175,886 reprted mtr vehicle accidents within the State. During that same year, 2064 drivers were invlved in three r mre accidents each fr a ttal f 6573 reprted accidents. Thus less than ne tenth f ne percent f the ttal licensed drivers in Indiana were invlved in ver 3.7 percent f the accidents that were reprted during the year 1667.

10 While 3.7 percent is nt a very large prprtin f the ttal accidents, it is a substantial number. If it culd be determined that these high accident drivers were ging t cntinue t becme invlved in such a large number f accidents during the fllwing years, it wuld be flish t allw them t cntinue as a threat t sciety. Therefre, this study was directed tward thse 2064 individuals wh were invlved in three r mre accidents during the year Data were cllected fr tw classes f drivers fr the purpse f this study. The first class f drivers included all persns wh were identified as being invlved in three r mre accidents during the year Cmplete accident recrd summaries were cllected fr each member f this class fr the years 1967 thrugh In additin, driver recrd summaries were cllected fr all in this class wh had a ttal f five r mre accidents during the five year study perid. The secnd class f drivers was selected randmly frm the entire ppulatin f Indiana drivers. This class f drivers was selected t serve as a cntrl t which the multiple accident driver culd be cmpared. Cmplete accident recrd summaries and driver recrd summaries fr the years 1967 thrugh 1971 were cllected fr all drivers in the cntrl class. It must be reprted, hwever, that the infrmatin recrded n the driver and accident recrds cnveys nly a small amunt f infrmatin abut that persn's driving perfrmance. Furthermre, a driver may cntinuusly vilate traffic laws

11 but will receive a citatin which will be recrded n his recrd nly when ticketed by a law enfrcement fficial r perhaps when invlved in an accident. A driver may have been invlved in many ptential accidents (near misses) but was recrded ( ) as having an accident nly when an accident actually ccurred and was reprted - required by law if damage was ver fifty dllars and/r an injury r fatality ccurred. Even then it is prbable that sme drivers may have had ne r mre accidents that were unreprted. Final Data Grups The final 1967 data fr which bth accident and driver recrds were available were rganized int fur grups as fllws: Number f Drivers Grup in Grup Cntrl Grup Accident Ttal Accident Ttal 288 >8 Accident Ttal 127 In additin t the drivers in these fur grups, infrmatin cncerning accident recrds and age and residence f drivers, but nt ther driver characteristics was btained fr drivers wh had three r fur accidents in 1967.

12 RESULTS The Accident Picture As nted, the search f the 1967 accident recrds prduced 2064 drivers wh had three r mre accidents in Indiana during that year and wh are believed t have been Indiana drivers fr the five year study perid. The number f accidents in which these drivers were invlved during that single year ranged frm three t nine accidents as shwn in Table 1. The grup as a whle were invlved in a ttal f 6573 accidents during the year. Lking at the accident recrd f this same grup f drivers fr the entire five year study perid, 1967 thrugh 1971, it was fund that the number f accidents in which the individual drivers were invlved ranged frm a lw f three, all in 1967, t a high f twenty-nine accidents (see Table 2). The grup as a whle were invlved in 9269 reprted accidents during the entire five year study perid. Taken as a grup, it is clear that the accident recrd f these 2064 drivers underwent much imprvement during the fur year perid fllwing the 1967 base year. In 1967, the grup had an average f 3.18 accidents per driver while the average per year fr the fllwing fur year perid was 1.30 accidents per driver. In cmparisn, the cntrl driver was fund t have an average f 0.09 accidents per year. This latter figure is shwn in Table 3, a summary f imprtant driver characteristics. Taken as a grup, the drivers wh had three r mre accidents in 1967 shwed cnsiderable imprvement in their accident recrd during the fllwing

13 Table Multiple Accident Drivers Number f Accidents Per Driver During The Year 1967 Number f Drivers Ttal 1967 Accidents Ttal Additinal Accidents During Fllwing Fur Year Perid

14 Table 2. Ttal Accident Invlvement During Five Year Study Perid By Drivers Invlved in Three r Mre Accidents in Ttal Accidents Per Driver Number f Drivers Ttal Accidents Ttal Accidents Per Driver Cumulative Number f Drivers > > > > > > > > > > > > > > > >

15 10 Table 3. Summary Statistics. Numbers in Parentheses refer t number f drivers, suspensins r vilatins upn which percentage r mean is based. Cntrl Grup 326 Drivers 5 Accident Grup* 361 Drivers 6-7 Accident Grup* 288 Drivers > 8 Accident Grup* 127 Drivers Percent Male Mean Age Median Age License Type Percent Operatr Percent Chauffeur Percent Public Passenger Chauffeur Percent f Drivers With One Or Mre Vilatin Citatins in 1967 Mean Number f Vilatin Citatins Per Driver in 1967 Percent f Drivers With One r Mre Serius** Vilatin Citatins in 1967 Mean Number f Serius Vilatin Citatins Per Driver in 1967 Percent f Drivers With One r Mre License Suspensin in 1967 Mean Number f License Suspensins Per Driver in (193) 95.6 (345) 94.8 (273) 97.6 (124) (295) 7.7 ( 25) 70.9 (256) 24.7 ( 89) 67.0 (193) 26.0 ( 75) 58.3 ( 74) 23.6 ( 30) ( 6) ( 16) ( 20) ( 23) ( 16) ( 18) 1.2 ( 4) ( 4) 0.3 ( 1) ( 1) (152) (224) 12.5 ( 45) ( 58) 3.0 ( ID ( 18) (118) (193) 13.5 ( 39) ( 49) 2.4 ( 7) ( ID ( 58) ( 83) 11.8 ( 15) ( 18) 3.1 ( 4) ( 4)

16 11 Table 3. (Cntinued) Cntrl Grup 326 Drivers 5 Accident Grup* 361 Drivers 6-7 Accident Grup* 288 Drivers 1 8 Accident Grup* 127 Drivers Percent f Drivers With One r Mre Suspensins fr Driving While Intxicated in ( 1) 1.7 ( 6) 1.0 ( 3) 2.4 ( 3) Mean Number f Suspensins Fr Driving While Intxicated in 1967 Percent f Drivers With One r Mre Suspensins Fr Failing t Appear in 1967 Mean Number f Suspensins Fr Failing t Appear in 1967 Percent f Drivers With One r Mre Suspensins fr Leaving the Scene f an Accident in ( 1) ( 0) ( 0) ( 0) ( 8) 0.6 ( 2) ( 2) 1.1 ( 4) ( 3) 0.7 ( 2) ( 2) ( 3) 0.8 ( 1) ( 1) ( 2) ( 0) Mean Number f Suspensins fr Leaving the Scene f an Accident in 1967 ( 0) ( 5) ( 3) ( 0) Mean Number f Accidents in ( 26) (1161) (980) (489) Mean Number f Accidents During Entire Study Perid (148) (1805) (1821) (1219) * Ttal accidents during the five year study perid. ** Serius vilatin refers t a vilatin with a pint cunt f five r mre r a vilatin that requires mandatry license suspensin.

17 12 fur year perid, but they still had mre than ten times the number f accidents as the average driver. Table 4 presents sme interesting statistics n the accident picture ver the 5-year perid f the drivers in the study grup. Of the grup f 1763 drivers wh had three accidents in 1967, 728 r 41.3 percent f the grup had n additinal accidents during the fllwing fur year perid. Of that same grup, 27.3 percent had nly ne additinal accident and 16.5 percent had tw additinal accidents during the same fur year perid. If all f the 1763 drivers in this study wh had three accidents during the year 1967 were prevented frm driving during the fllwing fur year perid, they as a grup wuld nt have becme invlved in 2102 additinal accidents that they had during this perid. Of the grup f 241 drivers wh had fur accidents in 1967, 32.8 percent had n additinal accidents during the fllwing fur year perid. Of that same grup, 25.7 percent had nly ne additinal accident and 18.7 percent had tw additinal accidents during the fllwing fur year perid. Of the grup f frty-six drivers wh had five accidents during the year 1967, nly eight drivers, r 17.4 percent f the grup remained accident free during the fllwing fur year perid. Of that same grup, 21.7 percent had nly ne additinal accident, 26.1 percent had tw additinal accidents, and 15.2 percent had three additinal accidents, all during the same fur year perid.

18 13 a in m IN 1/1 IN in X u a Cm A 5UOOJ3J 8 J SAT Jfl j ascjuirin cm C Im (1 5 c H in a c a r- v cm i-h C H u a H a) u > u H ^ U IM u <t> A M H) Oj Ul *J 3 a 7. Q) "O H u (1 c < H <W 4-J <0 r- m a: M V fj c 3 H Z qujj 9A tie]nujn quojjsj SJDAT j 30 JuqujnN 00JOJ ^v [nuin 3UOOJOJ BJOAf an j j.-irpunn 3UOOJ3J OATqefnujri qusjnj SJ3AT JO j jaquinn IN CM IN U3 CM 00 vc CM m IN O CM 00 1 '1 r-l <D f r^ r-l IN 1- T in P- H in in CO «r IX' r-l IN IN -^ CT U1 u C # 01 i T) r* H CM u M (1 1 <* OD <0 >M CM <-l u 11 0) O q U d 01 z a. J3 fl) H }uaj;ij sat;b tnuinn r- "* CM 00 in c ^ *r r- i-i IN.-1 in n t-i qu00j3 c (N r-k m CO 1 cm IN n in r- n H IN SJOAT 3Q CD,-1 ~\ v f> "* r-j CO in IN *T j JsquiriN I 3 '-i it).-! (U r-i r-l >< T IN r-i im id m <-i c *-> tu u r- C 3 cm h w a) cm n m w ct- r-i i r-l IN m *» in f> n -H i-l H -H CD E= T) U M CM lj u> A 3D US CJJi z < << a -i a h

19 14 Of the furteen drivers wh were fund that had six r mre accidents during the year 1967, all had at least ne accident during the next fur years. Only ne, in fact had nly ne accident. Almst half f the drivers in this grup had fur r mre accidents during this same fur year perid. One pint is clear frm this analysis - the greater number f accidents that a driver has during a particular year increases the prbability that he will have additinal accidents during the fllwing years. Anther pint, hwever, that is als clear is that ne cannt with much accuracy predict wh the high-accident drivers f future years are ging t be frm the high accident drivers f tday. Almst 40 percent f thse drivers wh had three r mre accidents in 1967 did nt have anther accident during the next fur years. Denying these 815 drivers the right t drive wuld have prvided n recgnizable benefits t sciety. An additinal 551 drivers r 27 percent f the grup had nly ne additinal accident during the fllwing fur year perid. The average number f accidents per driver ver the five year perid fr the cntrl grup was 0.45, in ther wrds the average accident rate f an Indiana driver is abut ne accident per driver every eleven years. Perhaps ne accident during a certain fur year perid is nt t bad a recrd. Even tw accidents in a certain fur years might nt be cnsidered t unusual. By denying licenses t all drivers wh had a ttal f three r mre accidents during the year individuals wh had nne r nly ne accident during the

20 15 fllwing fur year perid wuld have suffered in rder t prevent 698 persns wh cntinued t have a number f accidents frm driving and having these accidents. Similar results wuld have ccurred if 4 r mre accidents in 1967 had been chsen as a decisin criteria fr driver license cntinuance. Mre drivers wuld have been prevented frm driving wh had n accidents r nly ne accident during the next fur years as drivers wh had tw r mre accidents. If the criteria f 5 r mre accidents in a year wuld have been used in 1967 nly 60 drivers wuld have lst their license and nly 201 accidents wuld have been prevented ver the next 4 years - nt a very significant number - but it wuld have been helpful. But even here, a n-license penalty might be t severe fr the 19 f the 60 wh had n mre than ne accident during the fllwing fur year perid. In additin t the number f accidents, it might be enlightening t lk at the characteristics f high accident drivers. Perhaps then ne culd better predict wh the high accident driver is likely t be and direct safety effrts at him. The drivers recrds f each f the high accident drivers and f the cntrl grup, therefre, were btained and available characteristics cmpared. Characteristics f the accidents f the high accident drivers were als studied. First a few wrds abut these accident characteristics.

21 16 Seriusness f the Accidents State Plice recrds classified accidents int three categries based n the type f accident. These categries in were prperty damage nly fr accidents in which fifty dllars r mre damage resulted and n ne was killed r injured; injury accidents in which smene sustained a persnal injury; and fatality accidents in which a death resulted. Using these three categries as a criterin n which t base the seriusness f an accident, the data indicated (Figure 1) that, in general, the mre accidents that a driver has in a year the less serius are the accidents. Abut twenty-seven percent f the reprted accidents in which the three-accident driver grup were invlved were either injury r fatality accidents. In cmparisn, less than ten percent f the accidents in which the eight r mre accident grup were invlved were injury r fatality accidents. Where the Accidents Occurred A definite relatinship was als fund between the number f accidents that a driver had and his residence and the lcatin f the accidents (see Figure 2). Abut thirteen percent f the drivers in the cntrl grup were fund t reside in Indianaplis while mre than twenty percent f the drivers in the five r mre accident grup resided in Indianaplis. Abut furteen percent f the accidents incurred by the drivers f the cntrl grup tk place in Indianaplis. Mre than twenty percent f the accidents f the five r mre accident driver grup ccurred in Indianaplis. Over frtysix percent, r nearly half f all accidents in which the

22 17 % CJ t/> SE g < t I- z UJ Q u< % UJ -s vf 8 CJ uj x 5 i l t- u: p z UJ r * r«52 * CJ * CO H z UJ H- z UJ a 2 < UJ > u _J >z * m S f- z Z UJ z z a Z UJ UJ UJ UJ a a a Q < 1 < * 8 < ft < I < a. 3 r« cc r in <3 < 3 t CO UJ z CO ID H Z Ul a r UJ CO < UJ r O

23 18 C/> _I e < QZ H z z UJ mmm :e UJ UJ > O u z u > cc z : 3 O \- z UJ Q h- O z u LjJ Q < _J O <r < u H z < O H UJ z UJ Q Z UJ UJ r a: CM UJ O

24 19 drivers f the grup wh had eight r mre accidents were invlved ccurred either in Frt Wayne r Indianaplis. These findings seem t suggest that the high accident driver is mst likely t be a resident f a large urbanized, mre densely ppulated area. This seems reasnable since there are mre vehicles and pedestrians in these areas, mre cngestin, and thus many mre cnflicts and pprtunities fr accidents t ccur. Driver Characteristics Several persnal characteristics f the drivers were available frm the driver and accident recrds. These characteristics included age, sex, and license restrictins. Age The high accident driver is, in general, yunger than the average driver (see Table 5). The cntrl driver was fund t have an average age f years. The mean age f the high accident driver ranged frm years fr the five accident grup t years fr the three accident grup. Many studies have nted that certain age grups, mst ntably the yunger drivers, have a higher prbability f being invlved in accidents. The data frm this study agree with this finding. Hwever, this study indicates that age has little value as a predictr f which drivers wh are high accident drivers during a particular year will cntinue t have accidents during the fllwing years.

25 ' 1 * ' ' 20 M C CO-H O, H a UU < as in c in r^m as vdvd rh CN CN ih - c ^ ^,h ^- cr,^-' cm * cm - H rh CN rh m Oi O CM ** r- CN r^ ^ c *-- O r-l O ~ Os <N CM c t u -P c 0) ftl 13 rh U rf r-l rsi in CO «* as n -I 00 rh 00 r- in en -*r m c rh S > T rh O 00 r-l CN ^r in in «* CM rh TT CM " O ~ O CM m OO '-' rh *-* (N CN in rh * r-l \ CTs CM <u c H CO -p c CU ft T3 3 H O U ^ D O < CO CO <T CO rh CN in kd as cm c m c r~ Os CM CO CO >* in 00 CO O rh CM CN in r^ i in r rh. \ ) *-* VD -^ V > ^ in * as 00 * c < rh ^ O CO rh rh CM rh «00 in as CN CD > H HQ in O P c <u P c cu -a H O < a CN rh r- O OS 00 O CM CO r-~ OS rh 1-^ c r» O CN C > CN cn r- in r- in TT 00 O O rh CM «r-l V > m CO in *r tj* ^- r~ c r~- O <-{ ""J 1»-* s- 00 «rh rh CN»-* r-{ rh *- rh CO CN CO M 01 -p c d) Q T3 H P O O V-i h r- i <h m<n cc r^ in r-t ** ^ s cm «in»»c «rr»» -^r «<h O^ m-^ ' r~~ TfrH CO H <y\ r- 1 ^rh r-i * l )^-' n"-- 00 CO O 00 CN r-~ * (N c H 4J A H UP CO -H Q O ft U 3 -P O C U u m m rh cn CO - CM rh CN O U3 CO CO l > r^ CN CO "tf O CN v ko rh. ^ ^ O CM CO ^O *~* rh -~- CO CO > CM --' <>D "~f CM rh rh rh rj- ys CM O CO *- in CO 0) < m rh CU t V-I (0 CU <: i 10 m as IJ 1 TT»* *r ** ** rh CN CO in I i 1 i 00 in in in in m rh CN CM c ^T m \ A in CU < c i

26 21 The 728 drivers wh had three accidents in 1967 and n additinal accidents during the fllwing fur year perid had a mean age f years and a median age f 27.5 years as indicated in Table 5. In cntrast, the 127 drivers wh had eight r mre accidents during the five year study perid had a mean age f years and a median age f 27.9 years. There is less than half a years difference between either the means r medians f the ages f these tw driving grups. Clearly the age f the driver wh had already had three r mre accidents during a particular year will, in mst cases, be f little use in predicting whether he will have additinal accidents during fllwing years. While it is well established that male drivers have a greater accident invlvement than females, the results f this study are striking. Less than sixty percent f the licensed drivers in Indiana are male; hwever, male drivers accunted fr ver ninety-fur percent f the drivers with three r mre accidents during the year 1967 and a ttal f five r mre accidents during the entire five year study perid, Of the 776 drivers in this grup f drivers, nly thirty-fur were female. License Type It is unfrtunate that n infrmatin was available cncerning the expsure rate f the drivers under study. The number f miles that a persn drives and the circumstances under which they are driven shuld undubtedly be an imprtant

27 22 cnsideratin in this type f an accident study. It is prbable that a persn wh hlds a chauffeur r public passenger chauffeur's license is driving mre than the average persn with just an peratr's license. Hwever, it is als prbable that the high accident drivers in this study are nt average drivers. It is certainly pssible that many f these high accident drivers als travelled many mre miles in 1967 and the years fllwing than the average driver with an peratr's license. The statistics f this study indicate that as the number f accidents in which a driver is invlved ges up, the prbability that he is a hlder f either a chauffeur r public passenger chauffeur's license als ges up. While nly 7.7 percent f the cntrl grup were drivers with chauffeur's licenses, 23.6 percent f the drivers with eight r mre accidents during the five year study perid were drivers with chauffeur's licenses. The accident invlvement f drivers hlding the public passenger chauffeur's license is even mre stunning. While they cnstitute nly 1.8 percent f the cntrl grup, 18.1 percent f the drivers in the eight r mre accident grup were public passenger chauffeur's license hlders. This evidence that a high percentage f the drivers wh had a high number f accidents during the five year study perid hld chauffeur's r public passenger chauffeur's licenses prbably prvides sme explanatin f sme f the ther characteristics f this grup. The finding that a high number

28 23 f the drivers f this grup resided in large urban areas and had accidents there is undubtedly highly crrelated with the license type. Similarly the finding that as the number f accidents increased the seriusness f the accidents decreased prbably results frm the fact that many f these accidents are in urban areas where speeds are lw and cngestin great and that thse pssessing chauffeur's r public passenger chauffeur's licenses are likely driving during these cngested cnditins. In fact a pattern f wh the very high accident driver is mst likely t be, emerges. He is likely t be a yung male with a chauffeurs r public passenger chauffeurs license wh resides in and drives a lt in the larger urban areas f Indiana and has a large number f prperty damage accidents each year. Sunds like a taxi driver desn't it - and it prbably is. The Vilatin Recrd The 1967 vilatin recrd f drivers with three r mre accidents during the year 1967 and a ttal f five r mre accidents during the entire five year study perid appears t be f little use in predicting thse drivers that will cntinue t have accidents. Drivers wh had n vilatin citatins in 1967 cnstituted 54.3 percent f the drivers wh had eight r mre accidents during the five year study perid and 57.9 percent f the drivers wh had five accidents during the same perid (see Table 6). Thse drivers wh had tw r mre vilatin citatins during the year 1967 cnstituted 13.9 percent f the drivers wh had five ttal

29 24 * W -P r-h v rh c qj A -H V ^ueasa CO in in >-{ en (N rh CM VO j aequirin VO c c CM CM ^ueje< aat^ex nuin D * r-l *r r-{ O c VD tr r*- rh <x> VO (N U) -p C O [^ r-- CN r~ r^ 1 1 t a; 1 1 i n ^usaaj C\ ^ cri ^r <-\ VO -t-\ in CM O u < SJ9AT JO -H CO CM in CM r^ r-~ CM rh 50 aaqumn r-\ CO CO CM q.uaaed SAT^Exniun * f-t ih (T\ cm <T\ v c r c rh *r O OOO H UVa {/) -p c O) <N r- CO c CO CO in d O H -P (0 P H U a H -p (0 rh -H > VO 0) m ^U9DJ9d t) r- 00 ttl c O -H m CM u < SJ8ATJCa (T\ <N in eg r-{ rh O rh aj >H p -P u j aaquirin CM r-l ;U90I9(J SAT^eX nuln 3 +U9DI9d SI9ATJ0 j asquinn iw t O c rh cn vd *r rh CO v m «* en 1 1 O "» CM >-\ rh r VO CO VD CM CO 4J W U m >i > VM 0) -p tp c w P c H CJ O <0 vd 0) H XI a) H c r- U -H VD <D -P CT. «H 3 -H C Z U -H CM c in VO 10 p

30 25 accidents and 14.2 percent f the drivers wh had eight r mre ttal accidents. The percentage f drivers having varius numbers f citatins frm each f the high accident grups is rughly the same, thus making vilatin citatins f little use as a predictr t determine whether a driver wh has a number f accidents during a particular year will cntinue t have accidents during the fllwing years. The same situatin seems t be true f serius vilatins The fact that a driver had ne r mre serius vilatins during the year 1967 did nt have any significant relatinship t the number f accidents in which the driver was invlved during the fllwing years. While in many cases the vilatin recrd f a driver may be highly crrelated t his accident recrd (2,4,8,11), it appears t be f little use fr predicting whether drivers wh are high accident drivers during ne perid will cntinue t be high accident drivers in the fllwing perid. Case Histries Finally, the case histries f thse drivers wh cntinued t have- accidents during the perid were studied in detail, especially thse with large numbers f accidents. Tw f these cases fllw:

31 26 Case Histry A RESIDENCE: St. Anthny LICENSE: Operatr AGE: 4 5 SEX: RESTRICTIONS: Male Nne This driver had a recrd f fur license suspensins, tw f which were fr driving while intxicated, during the years 1960 thrugh His last suspensin was still in effect during the first part f the year DATE ACTION VIOLATION POINTS 5/29/67 Prperty damage accident 6/10/67 Reckless driving vilatin 6/10/67 6/10/67 Leaving the scene f an accident vilatin License suspended (1) -leaving the scene f an accident 6/10/67) 7/01/67 Prperty damage accident 7/10/67 Failed t reprt accident vilatin 7/10/67 Driving while vilatin suspended 9/02/67 Injury accident 10/09/67 Reckless speed vilatin 10/09/67 10/09/67 Driving while vilatin Driving while vilatin intxicated suspended 12/10/67 6/10/68 License suspended (2) -driving while suspended (7/10/67) Suspensin (2) clsed

32 27 VIOLATION DATE ACTION POINTS 6/10/68 License suspended (3) -driving while suspended (10/09/67) 10/21/68 Driving while intxicated vilatin 10/21/68 Driving while suspended vilatin 12/10/68 License suspended (4) -driving while intxicated (10/09/67) 11/01/69 Prperty damage accident 11/01/69 Prperty damage accident 11/03/69 Reckless driving vilatin 6 11/03/69 Driving while suspended vilatin 12/10/69 License suspended (5) -driving while intxicated (10/21/68) 3/15/70 Driving while intxicated vilatin 3/15/70 Driving while suspended vilatin 3/22/70 License suspended (6) -respnsible fr accident (11/01/69) 5/21/70 Suspensin (6) clsed 12/10/70 License suspended (7) -driving while suspended (10/21/68) 12/10/71 License suspended (8) -driving while suspended (11/03/69) 12/14/71 Suspensin (7) clsed 12/10/72* License suspended (9) -driving while intxicated (3/15/70) 12/10/73* License suspended (10) -driving while suspended (3/15/70) 2/15/72 Current status-suspended until 6/08/74 *These are dates that the suspensins will autmatically g int effect fr the past vilatins listed.

33 28 The fact that this driver's license was suspended during the entire five year study perid gives emphasis t the need fr better cntrls t prevent unlicensed individuals frm driving. Five year summary: Ttal accidents 5 Ttal vilatins 13 Ttal vilatin pints 25 Ttal suspensins 10 Case Histry B RESIDENCE: LICENSE: Indianaplis Public passenger chauffeur AGE: 29 SEX: RESTRICTIONS: Male Nne This driver had ne vilatin fr speeding fr which his license was suspended fr ne mnth in DATE ACTION VIOLATION POINTS 1/20/67 Prperty damage accident 6/22/67 Prperty damage accident 9/05/67 Prperty damage accident 10/26/67 Injury accident 10/29/67 Prperty damage accident 11/11/67 Prperty damage accident 3/29/68 Chauffeur license issued 7/06/68 Prperty damage accident 8/29/68 Injury accident 11/06/68 Prperty damage accident

34 29 DATE ACTION VIOLATION POINTS 1/11/69 Injury accident 5/30/69 Injury accident 6/05/69 Prperty damage accident 6/17/69 Prperty damage accident 7/11/69 Speeding vilatin-44 mph in 30 mph zne 10/25/69 Prperty damage accident 12/24/69 Prperty damage accident 12/30/69 Prperty damage accident 1/02/70 Injury accident 1/07/70 Prperty damage accident 1/21/70 Signal vilatin 1/16/70 Injury accident 4/14/70 Injury accident 5/04/70 Prperty damage accident 5/18/70 Signal vilatin-lcal rdinance 6/12/70 Prperty damage accident 8/15/70 Injury accident 9/03/70 Injury accident 9/18/70 Prperty damage accident 11/08/70 Prperty damage accident 2/12/71 Prperty damage accident 2/18/71 Injury accident 5/10/71 Operatr license issued 5/11/71 Prperty damage accident 6/24/71 Signal vilatin-lcal rdinance 6/24/71 Signal vilatin-lcal rdinance 2 2 7/09/71 Public passenger chauffeur license issued 2/10/72 Current status-6 active vilatin pints

35 30 The lngest perid that this driver was able t g during the five year study perid withut having an accident was less than six mnths. During the same perid he averaged apprximately ne accident every tw mnths. Five year summary: Ttal accidents 29 Ttal vilatins 5 Ttal vilatin pints 10 Ttal suspensins Driver A has clearly been identified as a prblem driver. His license was suspended fr the entire five year study perid. This, hwever, did nt prevent him frm driving as he was invlved in five reprted accidents during this same perid. Driver B was invlved in six reprted accidents during the year 1967 and a ttal f 29 during the five year study perid. While his vilatin recrd is relatively mild cmpared t ther study drivers, he was invlved in the mst reprted accidents f the study. While these tw drivers are extreme examples f prblem drivers, there were a few ther drivers with very similar recrds. It is clear that mre wrk is needed bth in the identificatin f such prblem drivers and either rehabilitatin and effective driving preventin methds fr these drivers. There is sme indicatin that a study f driver and accident recrds ver a perid f several years wuld permit develpment f a few case histries which wuld prvide a

36 31 reasnably accurate predictin f a high future accident invlvement. The evidence, hwever, must be s great that nly an extremely small number f drivers culd be s rated. Unfrtunately, it appears this can be predicted nly after they have had numerus accidents and vilatins. Even if these drivers were remved frm the rad - which driver license suspensin as currently managed des nt d - nly an extremely small reductin in the accidents in any ne year wuld result. CONCLUSIONS The fllwing cnclusins cncerning high accident drivers summarize the findings f this research. 1. As a grup, the accident recrd f drivers wh are invlved in three r mre accidents during a particular year tends t shw much imprvement during fllwing years. 2. A large prprtin f thse drivers wh are invlved in three r mre accidents during a particular year remain accident free during the fllwing fur years. 3. Nearly all high accident drivers are male. 4. Drivers with chauffeur's licenses and particularly public passenger chauffeur's licenses accunt fr a large prprtin f the individuals wh are high accident drivers during ne year and cntinue t be high accident drivers during the fllwing years.

37 High accident drivers are likely t be residents f large urban areas. 6. The vilatin recrd f a driver wh has a high accident invlvement during a particular year has little relatinship t his accident recrd in the fllwing years. 7. In general, the accident recrd f a driver during ne year, even when cmbined with infrmatin available in the Indiana drivers recrds, des nt enable a reasnable predictin f the drivers accident recrd fr future years. In a very few cases, where a histry f numerus vilatins and accidents are recrded, a reasnable predictin f cntinued high accident invlvement can pssibly be made

38 33 REFERENCES 1. Campbell, B. J., "Wh Really Causes the Accidents?", Signal 99, Nrth Carlina Gvernr's Highway Safety Prgram, Vl. 1, N. 2, Spring Cppin, R. S., Peck, R. C, The 1964 Califrnia Driver Recrd Study, Part 8, The Predictin f Accident Invlvement Using Cncurrent Driver Recrd Data, Reprt N. 20, Research and Statistics Sectin, Califrnia Department f Mtr Vehicles, Sacrament, January Crancer, Alfred, Jr., "Invlvement f the Prblem Driver in Fatal Mtr Vehicle Accidents", Traffic Quarterly, Octber 1967, pp Crancer, Alfred, Jr., Accident and Vilatin Rates fr Washingtn Drivers, Washingtn Department f Mtr Vehicles, Olympia, Reprt N. 003, May Crancer, Alfred, Jr., McMurray, Lucille, Credit Ratings as a Predictr f Driving Behavir and Imprvement, Washingtn Department f Mtr Vehicles, Olympia, Reprt N. 010, May Gldstein, Len G., "Psychlgical Aspects f Traffic Accidents", Traffic Digest and Review, Vl. 12, N. 6, July 1964, pp , Jhnsn, H. M., "The Detectin and Treatment f Accident Prne Drivers", Psychlgical Bulletin, N. 43, 1946, pp O'Neall, P. A., Relatinship f Accident Invlvement and Number f Citatins: 1966 Data, Washingtn Department f Mtr Vehicles, Olympia, Reprt N. 006, Penn, H. S., Causes and Characteristics f Single Car Accidents, Parts 3 $ 4, Califrnia Highway Patrl, Sacrament, January Schuster, D. H., Guilfrd, J.P., An Analysis f Accident Repeater and Chrnic Vilater Drivers, Traffic Prject Reprt N. 1, U.S. Public Health Grant M-2353, Ls Angeles, Califrnia, Schuster, D. H., "Predictin f Fllw-up Driving Accidents and Vilatins", Research Review, March 1968, pp Shaw, L., Sichel, H. S., "The Reductin f Accidents in a Transprt Cmpany by the Determinatin f the Accident Liability f Individual Drivers", Research Review, December Waller, Julian A., "Traffic Accidents, A Medical Viewpint, Part 1", Autmtive Industrie s, Vl. 136, N. 8, April 15, 1967, pp

39

40