ICADTS T '9 2 REPEAT DHI OFFENDERS: THEIR INVOLVEMENT IN FATAL CRASHES

Transcription

1 ICADTS T '9 2 REPEAT DHI OFFENDERS: THEIR INVOLVEMENT IN FATAL CRASHES by James C. Fel 1 Chief, Scientific & Technical Affairs Office f Alchl and State Prgrams (NTS-20) Natinal Highway Traffic Safety Administratin Washingtn, DC USA ABSTRACT Drivers with p ri r DHI cnvictins are verrepresented amng drivers Invlved 1n fa ta l crashes and the re la tiv e risk, f fa ta l crash Invlvement 1s fa r greater f r these repeat ffenders. Only abut 3 percent f a ll licensed drivers have a p ri r DHI arrest within the past three years, yet clse t 12 percent f Intxicated drivers in fa ta l crashes have had at least ne p ri r DHI cnvictin w ithin three years f th e ir crash. Hwever, while repeat DHI ffenders are at a substantially higher ris k f fa ta l crash Invlvement, the vast m ajrity f Intxicated drivers In fa ta l crashes are nt repeat ffenders (7 ut f 8) accrding t available natinal recrds. I t 1s recmmended that lcal ju risd icti n s use the methds described 1n th is paper t determine the magnitude f the repeat ffender prblem in th e ir area s th at apprpriate decisins can be made with regard t assessments, treatm ent, and sanctins 1n dealing with these drivers. BACKGROUND Accrding t the U.S. Department f Justice, law enfrcement agencies made an estimated 14.2 millin arrests in 1990 fr all criminal Infractins excluding traffic vilatins. Mre peple were arrested fr driving under the influence (DUI) r driving while intxicated (DWI) than any ther reprted criminal Infractin. Over 1.8 millin drivers were arrested fr DUI r DWI cmpared t 1.6 millin peple fr larceny r theft and 1.1 millin peple fr drug abuse vilatins [1]. Ignring repeat ffenders, this means that apprximately 1 ut f every 100 licensed drivers 1n the U.S. 1s arrested fr DUI/DWI In any given year. But we knw that we cannt ignre repeat ffenders. There is much evidence that repeat ffenders, as a grup, are high risk prblem drinker drivers [2]. Are these drivers with prir DWI's verrepresented as drivers invlved in fatal crashes accrding t natinal data? Is the relative risk f fatal crash Invlvement greater fr these repeat'ffenders? These questins will be addressed in the next sectins. METHOD AND DATA AVAILABLE The U.S. Department f Justice reprted that an estimated 1,810,800 drivers were arrested fr DUI/DWI in 1990 [1]. DWI arrests fr 1989 were 1,736,200 and fr 1988 they were 1,792,500. S, ver the three year perid frm , apprximately 5,339,500 drivers were arrested fr DWI Alchl, Drugs and Traffic Safety - T92 Ed. by Utzelmann / Berghaus / Krj Verlag TÜV Rheinland GmbH, Köln -1993

2 The Federal Highway Administratin reprted that there were 167,015,250 licensed drivers In the U.S. In 1990 [31. The Natinal Highway Traffic Safety Administratin's Fatal Accident Reprting System (FARS) Indicated that 58,796 drivers were Invlved In fatal crashes 1n 1990 [4]. A ttal f 14,558 f these crash Invlved drivers (24.8 percent) were estimated t have been Intxicated (I.e. their bld alchl cncentratin (BAC) was greater than r equal t.10 g/dl) and anther 4373 drivers had been drinking alchl (BAC's were.01 t.09 g/dl). The remaining 39,865 drivers did nt have any estimated alchl Invlvement (BAC's were.00 g/dl). The 1990 FARS data als Indicated that 1721 (11.8%) f the 14,558 Intxicated drivers In fatal crashes had a prir DWI c nvictin within the past three years. A ttal f 244 drivers (5.6%) f the 4373 drinking drivers had prir DWI cnvictins within the past three years, while 687 (1.7%) f the 39,865 zer BAC drivers were reprted as having a prir DWI cnvictin 1n the past three years. All f these data are summarized 1n TABLE 1. TABLE Drivers Arrested f r Driving Under the Influence (DUI-DHI) 1,810,800 Licensed Drivers 167,015,250 Drivers Invlved In Fatal Crashes 58,796 Drivers Invlved 1n Fatal Crashes With Prir DUI-DHI Cnvictin In Past Three Years 2,652 Drivers Invlved 1n Fatal Crashes: Hlth BAC s >.10 Hith P rir DHI's Hith BAC's = H1th Prir DHI's Hith BAC's =.00 Hith P rir DHI's 14,558 1,721 4, ,

3 In rder t address the questin f verrepresentatin f drivers with prir DWI1s 1n fatal crashes, several prprtins will be cmpared. First, a cmparisn will be made f the percent f all licensed drivers with a prir DWI 1n the past three years with the percent f all drivers Invlved In fatal crashes (drinking and nn-drlnk1ng) with prir DWI's. If drivers invlved 1n fatal crashes are n different frm all licensed drivers, ne wuld expect the same percentage f prir DWI's fr these tw grups. If drivers invlved in fatal crashes, and especially drunk drivers, have a larger percent f drivers with prir DWI's in the past three years,, there Is an verrepresentatin. Secnd, given certain assumptins and limitatins f the data, we can calculate the risk f a driver with a prir DWI being invlved in a fatal crash relative t the risk f a driver with n prir DWI's being invlved in a fatal crash. This Is called the "relative risk" f fatal crash invlvement. This Is calculated as fllws: P (Invlvement. pr )r DWI) # Drivers In Fatal Crashes with Prir DWI In = past 3 years # Drivers Arrested fr DWI 1n past 3 years P (Invlvement. N pr ir DWI) # Drivers In Fatal Crashes with N Prir DWIs» In past 3 years # Licensed Drivers with N Prir DWIs in past 3 years Relative Risk f Fatal Crash Invlvement = P (Invlvement: Prir DWI) P (Invlvement: N Prir DWI) These calculatins were made using 1990 data. RESULTS The prprtin f all licensed drivers in 1990 (167,015,250) wh had a prir DWI arrest within the past three years (5,339,500) was 3.2%. In 1990, the prprtin f all drivers invlved 1n fatal crashes (58,796) wh had a prir DWI cnvictin within the past three years (2652) was 4.5X. The prprtin f all drivers Invlved in fatal crashes in 1990 with a BAC = /. (4373) wh had a prir DWI cnvictin In the past three years (244) was 5.6X. The prprtin f all drivers Invlved In fatal crashes in 1990 wh were intxicated (BAC =.10+)(14,558) and had a prir DWI cnvictin In the past three years (1721) was 11,8t. 1046

4 The prprtin f all drivers invlved In fatal crashes in 1990 whse BAC was zer (39,865) wh had a prir DWI cnvlctin 1n the past three years (687) was 1.7X. The relative risk f being Invlved 1n a fatal crash fr drivers with prir DWI cnvictins within the past three years was the fllwing: R elative Risk f Fatal- p (Invlvement: Crash Invlvement with - Prir DWI) « = Prir DWI In 1990 p (Invlvement: N Prir DWI) 161,675, Relative Risk f Intxicated Driver Fatal Crash Invlvement with Prir DWI 1n 1990 CONCLUSIONS ** (Intx.Inv. = with Prir DWI) P (Intx.Inv. N Prir DWI) = = ,675,750 At a minimum, drivers with prir DWI cnvictins are verrepresented as drivers in fatal crashes. While 3.2% f all licensed drivers had a prir DWI arrest within the past three years In 1990, 4.5% f drivers 1n fatal crashes had a prir DWI cnvictin within the past three years. S drivers with prir DWI's are verrepresented by a factr f 1.4 (I.e. 4.5%/3.2%). Drivers with prir DWI's are als verrepresented as drinking drivers in fatal crashes with lw BAC's ( ) by a factr f 1.8 (5.6%/3.2%). Drivers with prir DWI's are verrepresented as Intxicated drivers in fatal crashes by a factr f 3.7 (11.8%/3.2%). Drivers with prir DWI'S are underrepresented as drivers In fatal crashes with zer BAC'S by a factr f 0.5 (1.7%/3.2%). Drivers with prir DWI cnvictins within the past three years have 1.4 times the risk f being invlved In a fatal crash as drivers withut prir DWI's. Intxicated drivers with prir DWI cnvictins have 4.1 times the risk f being In a fatal crash as Intxicated drivers withut prir DWI's. 1047

5 DISCUSSION The limitatins In the data shuld be nted. Fr the ppulatin-at-rlsk data, we nly have Infrmatin n prir DWI arrests within the past three years. Fr fatal crash invlved drivers, we nly have prir DWI cnvictins within the past three years frm the State where the driver was licensed. This excludes mst f the drivers wh had prir DWIs in ther States. This renders the verrepresentatin and.relative risk calculatins as cnservative since all DWI arrests d nt result In DWI cnvictins. Sme jurisdictins in the U.S. reprt that nly abut 20-30% f DWI arrests result in cnvictin. Other jurisdictins reprt an cnvictin rate. Cnsequently, these verrepresentatins and relative risks shuld be qualified by the statement: "At a minimum-----" Fr example, 1f the natinwide cnvictin rate fr DWI arrests was 50%, then the relative risk f being Invlved in a fatal crash fr drivers with a prir DWI cnvictin 1n the past three years wuld be: p Invlvement 2652 Prir DWI Cnvictin (x.5Q) P Invlvement» N Prir DWI 161,675,750-5,339,500(.5) Cnvictin , ,006,000 = = This is twice the relative risk when using DWI arrest data nly. The fact that nly 1.7% f the zer BAC drivers in fatal crashes had a prir DWI cnvictin prbably reflects the natinwide cnvictin rate in the U.S. (I.e. 1.7% cnvicted/3.2% arrested -.5 t.6) Finally, while It is true that drivers with prir DWI' s mst certainly are verrepresented as drinking drivers in fatal crashes and have a much greater risk f fatal crash Invlvement than drivers withut prir DWI's, It is als clear that the vast majrity f Intxicated drivers Invlved in fatal crashes d nt have prir DWI cnvictins, at least within the past three years. The data indicate that abut 7 ut f 8 intxicated drivers In fatal crashes have n reprted DWI cnvictins within the past three years. The prprtin f intxicated drivers Invlved In fatal crashes with prir DWI's wuld certainly Increase if ne lked at recrds fr the past 5 r 10 years, if all ut f State recrds were checked fr each driver, and 1f ne lked at DWI arrests rather the cnvictins. Unfrtunately, thse data are nt available n a natinal basis. Studies cnducted by individual States Indicate that abut 20% t 30% f drivers arrested each year fr DWI are repeat ffenders. 1048

6 RECOMMENDATIONS It 1s recmmended that lcal highway safety fficials determine these prprtins and relative risks assciated with their wn jurisdictins at Issue. Better decisins can then be made regarding the magnitude f the repeat ffender prblem In that area and apprpriate assessment, treatment and sanctins can then be instituted. NHTSA believes that effective actins shuld be taken against repeat ffenders' vehicle registratins, license plates, and the vehicles themselves. A series f Increasingly severe sanctins shuld be levied against cnvicted drunk drivers wh either drive while their license Is suspended r are caught again fr DWI. Fr example, the vehicle registratin culd be revked fr driving under suspensin; license plates culd be cnfiscated fr a secnd ffense f DWI; the vehicle Itself culd be frfeited fr a third ffense f DWI, r fr a DWI cmmitted while driving under suspensin. NHTSA als recmmends ja il r cmmunity service fr repeat ffenders. REFERENCES tl] U.S. Department f Justice, "Crime In the United States ", Unifrm Crime Reprts, Federal Bureau f Investigatin, Washingtn, DC, August, [2] Natinal Highway Traffic Safety Administratin, "Alchl and Highway Safety 1989: A Review f the State f Knwledge", U.S. Department f Transprtatin, Washingtn, DC, DOT-HS , March E33 Federal Highway Administratin, "Highway Statistics ," Office f Highway Infrmatin Management, FHWA-PL , [4] Natinal Highway Traffic Safety Administratin, "Fatal Accident Reprting System ," Natinal Center fr Statistics and Analysis, Washingtn, DC, DOT-HS , December