Preventing Impaired Driving Ralph W - PDF document

Preventing Impaired Driving Ralph W. Hingson, Sc.D., Timothy Heeren, Ph.D., and Michael R. Winter, M.P.H. Although moderate drinking does not necessarily increase a personÕs blood alcoholconcentration (BAC) to the level at which driving is legally prohibited in the United States, EYWORDSminimum drinking age; traffic accident; AODR (alcohol and other drug related) mortality; Vol. 23, No. 1, 199931 a below the legal BAC limit for driversin most States. Eight percent of thesedrivers had BACs between 0.01 and0.04 percent. This article explores the relationshipbetween alcohol consumption andimpairment and examines legislativeapproaches for reducing alcohol- Moderate Drinking and Impairment Even at low BAC levels, alcohol impairsdriving performance by reducing thedrivers reaction time and slowing his or her decisionmaking process ALPHW. H consumption, age, gender, and theproportion of body mass that is fattytissue. Typically, a 170-pound manwould need to consume five drinks in1 hour on an empty stomach to reacha BAC of 0.10 percent. To reach a BACof 0.08 percent, the legal limit in 17States, a 170-pound man would needan empty stomach (see figure 1).Women absorb and metabolizealcohol differently than do men. Ingeneral, compared with men, womenwater to absorb each drink. Womenalso exhibit lower activity levels of thedehydrogenase (ADH) in the stomach,causing a larger portion of ingestedalcohol to reach the blood (NIAAAneed to consume three drinks in 1hour on an empty stomach to reach a0.08-percent BAC and four drinks in2 hours to reach a 0.10-percent BAC(see figure 1). Drinking over a longerrequired to reach these BAC levels(NHTSA 1997).s BAC to exceed theing increases the risk of being involvedin a fatal crash. Compared with driverswho have not consumed alcohol,drivers with BACs between 0.02 and0.04 percent are 1.4 times as likely tobe involved in a single-vehicle fatalcrash. Furthermore, this risk increasesdrivers with BACs between 0.05 and0.09 percent, 48 times higher fordrivers with BACs between 0.10 and0.14 percent, and 380 times higherfor drivers with BACs at or above 0.15percent (Zador 1991). For drivers under age 21, fatal crashrisk increases more with each 0.02-percent increase in BAC than it doesfor older drivers (Zador 1991). At allBAC levels including zero, the fatalcrash risk for female drivers ages 16 todrivers age 25 and over, and the riskfor male drivers ages 16 to 20 is triplethe risk for male drivers age 25 andover (Zador 1991). Young driversgenerally have less driving experiencethan older drivers and are more likelydisobeying traffic signals, and not wear-ing safety belts (Hingson and Howland1993). Because alcohol consumptionfurther increases the risk of crashinvolvement for young drivers, all Stateshave adopted zero-tolerance laws fordrivers under age 21, prohibiting drivingRecognizing the threat to the publicsafety associated with even moderatedrinking and driving by transporta-tion workers, the Federal governmentprohibits commercial truck drivers,railroad and mass transit workers,marine employees, and aircraft pilotsfrom operating their vehicles with aBAC at or greater than 0.04 percent.To reach this relatively low BAClimit, however, most people wouldhave to drink above the level of mod-erate drinking. The American MedicalAssociation (1986) has endorsed low-ering the legal BAC limit to 0.05 per-cent for all drivers; however, no Statehas yet adopted this standard.Several other countries legal BAClimits for drivers come closer than theUnited States to restricting drivers toCanada, Austria, and the UnitedKingdom is 0.08 percent. Legal limitsin Australia range from 0.05 to 0.08percent. The Netherlands, Finland,France, Germany, and Japan all have0.05-percent legal limits. Sweden haslowered the legal BAC limit for driversto 0.02 percent (NHTSA 1997). Legislative Measures Society has implemented a variety oflegislative measures to reduce alcohol-impaired driving. To date, U.S. lawshave not attempted to restrict adultdrivers to moderate drinking. None-theless, research indicates that lawsadopted in the United States havereduced alcohol-related traffic deathsthose who have very high blood alcohollevels.Current U.S. laws that limitdrivers BACs to 0.10 and 0.08 percenthave been found effective for reducingalcohol-related crashes, as have license 32 Figure 1 Average blood alcohol concentration (BAC) after different levels of alcoholconsumption. In this diagram, one drink equals 0.54 oz alcohol; this is the = one drink; lb = pound; oz = ounce. .10.09 in 1 Hour of Drinking .09 Female 137 lbMale 170 lb in 2 Hours of Drinking revocation/suspension and otherpenalties for exceeding those limits aswell as laws restricting access to alcohol. Legal BAC Limits for Drivers Reducing the BAC level at whichpeople can legally drive can effectivelyreduce alcohol-related traffic crashes(e.g., Johnson and Fell 1995; Hingsonet al. 1996). Drivers under age 21 aresubject to lower BAC limits than olderdrivers, and a few States have set lowerlimits for drivers previously convictedof driving while intoxicated (DWI). Lowering the Legal BAC Limits to0.08 Percent for Drivers Over Age 21. Seventeen States have lowered thecriminal per se legal BAC limit from0.10 to 0.08 percent for noncommercialdrivers age 21 and older (lower limitshave been established for drivers youngerthan age 21 and are discussed later inthis article). According to the criminalper se provision, prosecutors are notrequired to introduce evidence otherthan a BAC of 0.08 percent or higher todemonstrate impairment, thereby mak-laws also have administrative licenserevocation (ALR) provisions, whichseize the license of any driver with aBAC of 0.08 percent or higher. TheSeveral studies have examined theeffects of 0.08 laws on fatal crash trends.In California, the largest State to adopta 0.08 law, researchers found a 12-percent decline in alcohol-related fatalBecause California also adopted anper se law, the separate effects of eachlaw are difficult to determine (NHTSA1991). According to one study, mostof the effects occurred after the ALRprovisions were added (Rogers 1995).Johnson and Fell (1995) monitoredsix measures of driver alcohol involve-ment in the first five States to adopt0.08 laws (Utah, Oregon, Maine,California, and Vermont) and identi-fied several statistically significant pre-to post-law decreases. Because the studydid not compare States with the 0.08law with States that did not have thelaw, researchers could not determinewhether the changes were indepen-dent of general regional trends. Theresearchers concluded that the effectsof the law were independent ofnational trends.In a subsequent analysis, Hingsonand colleagues (1996) paired theaforementioned first five States toadopt a 0.08 law with five nearbyStates that retained the 0.10-percentlegal BAC limit (Idaho, Washington,Massachusetts, Texas, and NewHampshire). Relative to the compari-son States, the States that adopted thereduction (i.e., 16 percent) in the pro-portion of fatal crashesinvolvingfatally injured drivers with BACs of0.08 percent or higher (see figure 2).These results resembled those reportedin both the United Kingdom andFrance when those countries firstlicense revocation provisions (Ross1973; Ross 1982).The majority of drivers killed inpositive for alcohol had BACs of 0.15percent or higher. The first five Statesa significant decline (i.e., 18 percent)relative to the comparison States inthe proportion of fatal crashes involv-ing fatally injured drivers with BACsof 0.15 percent or higher (Hingson etal. 1996). Compared with 0.10-percentStates, the 0.08-percent States mayhave been more concerned aboutalcohol-impaired driving and more Vol. 23, No. 1, 199933 Preventing Impaired Driving BeforeAfterBeforeAfterComparison States.08% Law States 05101520 21.923,68621,04927,28723,672 ( )( )( ) �Injured Driver With BAC .08%Figure 2 alcohol concentrations (BACs) of .08% or higher before and after the SOURCE: Hingson, Heeren, and Winter 1996. responsive to legislative initiatives toreduce the problem. For example, allfive 0.08 States had ALR laws duringthe study period, and three of the Statesimplemented ALR laws within 1 yearof the 0.08 law. This proximity restrictedof the 0.08 laws from that of the ALRlaws. ALR laws alone have been asso-ciated with a 5-percent decline in bothalcohol- and non-alcohol-relatedcrashes and as much as a 9-percentdecline in alcohol-related fatal crashes(Zador et al. 1989). In another study, researchers pairedsix additional States that had adoptedNorth Carolina, Florida, New Hamp-shire, New Mexico, and Virginia)with six similar contiguous or nearbyStatesthat retained a 0.10-percentlegal limit (Oklahoma, Tennessee,Maryland, Georgia, Colorado, andConnecticut) (Hingson et al. in press).As a group, the 0.08 States experi-enced a 16-percent reduction in theproportion of alcohol-related fataldriver or pedestrian had a BAC of0.10 percent or higher), which wassignificantly greater than the 11-percent decline identified in comparisonStates duringthe same time period.Similarly, the 0.08 States experienceda 20-percent decline in the propor-tion of drivers with BACs of 0.10 per-who were involved incantly greater than the 14-percentdecline in the comparison States.Declines experienced by comparisonStates were not significantly differentfrom those experiencedby other 0.10States across the country during thesame time period. Four of the 0.08States (i.e., Kansas, North Carolina,Florida, and New Mexico) had ALRlaws in place before the studyperiod.Therefore, ALR laws in those Statesings. Relative to the comparisonstates,the four 0.08 States with ALRgreater decline in the proportion ofalcohol-related fatal crashes (17 versus13 percent) and in the proportion ofdrivers involved in fatal crashes withBACs of 0.10 percent or higher (21versus 16 percent). Thus, indepen-dent 0.08 law effects were seen inthese States, although the effects wereof States that adopted both 0.08 anda close timeframe (Hingson 1996;Rogers 1995).In the most comprehensive studyconducted over a 16-year period iden-tified an 8-percent decline in the pro-portion of drivers with positive BACsinvolved in fatal crashes in States thathad adopted 0.08 laws. This reductionDWI laws, such as the 0.10-percentBAC limit and ALR laws. The studyalso controlled for the effects of safetytrends in demographic, economic, andtion. The reduction was observedamong both drivers with BACs from0.01 to 0.09 percent and drivers withBACs of 0.10 percent and higher (Voasand Tippets 1999). In addition, an11-State study found that 0.08 legis-declines in alcohol-related fatalities inseven States, and significant declineswere associated specifically with 0.08laws in five States (Apsler et al. 1999).It appears that 0.08 percent laws havethat their greatest impact is in combi-For legal BAC limits to most effec-tively deter alcohol-impaired driving,A national survey of more than 4,000drivers (NHTSA 1996) found that 34 0103040 BeforeAfterBeforeAfterComparison StatesZero-Tolerance States 2,851 ( ) ( )( ) Vehicles at Night (%)Figure 3 and after zero-tolerance laws for youth. Researchers compared States SOURCE:Hingson, Heeren, and Winter 1994. only 54 percent of residents in Stateswith 0.08 laws and only 38 percent ofresidents in States with 0.10 laws couldcorrectly identify their States BAClimit. When the legal BAC limits fordrivers are lowered, the need to educateent.Research indicates that public educa-tion to promote awareness of a Statenew legal BAC limit can enhance thes effects (Blomberg 1992).Public education also is needed toincrease drivers knowledge about theimpairments associated with differentlevels of alcohol consumption. Amongdrivers surveyed, 75 percent thoughtthat at least one-half of all driverswould be dangerous if they drove afterconsuming five drinks in 2 hours, butonly 28 percent of the respondentsthought all drivers would be unsafeunder those conditions (NHTSA 1996). Zero-Tolerance Laws for Drivers UnderAge 21. Although all States prohibitpeople under age 21 from purchasing,drinking remains prevalent amongteenagers (Johnston et al. 1998).Zero-tolerance laws are designed toreduce drinking and driving amongyoung people by making it illegal forpersons under 21 to drive after anyBAC limit for drivers under age 21 at0.00 or 0.02 percent. In the fall of1995, the U.S. Congress mandatedthat Federal highway funds be with-held from States that did not adoptzero-tolerance laws. At that time, onlyone-half of the States had such laws. A recent study (Hingson et al. 1994)found that the first eight States toadopt zero-tolerance laws experienceda 20-percent greater decline in theproportion of nighttime single-vehiclecrashes among 15- to 20-year-olddriverscompared with eight nearbyStates without zero-tolerance laws (seefigure 3). Single-vehicle nighttimecrashes are the type of fatal crash mostlikely to involve alcohol.States that did not adopt zero-tolerance laws but lowered their BAClimits for drivers under age 21 to either0.04 or 0.06 percent experienceddeclines of 6 to 7 percent relative toStates with no BAC limit specific todrivers under age 21 (Hingson et al.1994). Setting BAC limits for youngdrivers at 0.04 or 0.06 percent allowsyouth to speculate about how muchthey can drink and still drive legally.Conversely, zero-tolerance laws send aclear message to young drivers that itis illegal to drive after engaging in any drinking. By the summer of 1998, all50 States had passed zero-tolerance laws.Unfortunately, some States havebroad public awareness of zero-tolerancelaws. Studies in both California andMassachusetts found that 45 to 50percent of young drivers were unawareof the law (Martin et al. 1996). Ob-viously, increasing awareness of thezero-tolerance law can enhance itseffects.Blomberg (1992) found aone-third greater decline in alcohol-involved crashes among young driversin Maryland counties where publicservice announcements about theStates zero-tolerance law were airedcompared with drivers in other counties.As a result of all States raising theto 21 and adopting zero-tolerance lawsfor persons under age 21, the greatestreductions in alcohol-related trafficcrashes in recent years have occurredamong drivers under 21. Since NHTSAmates of alcohol involvement in fatalcrashes, alcohol-related fatalities among15- to 20-year-olds have declined 59percent, from 5,380 in 1982 to 2,209in 1997. By comparison, alcohol-related fatalities among all other agegroups declined 29 percent, from19,785 to 13,980 (NHTSA 1998 a) (see figure 4). The proportion of allfatal crashes among 15- to 20-year-olds that involved alcohol declined 44percent compared with 30 percent forall other age groups from 1982 to 1997. Lower Legal BAC Limits for DriversConvicted of Driving While Intoxicated. Similar to drivers under age 21, driverswith prior DWI convictions are dis-proportionately at risk for alcohol-related fatal crash involvement. Onestudy found that drivers killed inalcohol-related crashes were eighttimes more likely to have had a DWIconviction in the previous 5 years thandrivers randomly selected from thegeneral population of licensed drivers(Brewer et al. 1994). Vol. 23, No. 1, 199935 01020 Ages 15Ñ2097969594939291908988878685848382 63%56%39%35% Figure 4 SOURCE:National Highway Traffic Safety Administration 1998 Preventing Impaired Driving In August 1988 Maine lowered thelegal BAC limit for drivers previouslyconvicted of DWI to 0.05 percent. Inillegal for these drivers to drive after any drinking.The lawallows first-time DWI offendersto be reinstated after a mandatory sus-that they not drive with a positive BACfor 1 year. Licenses of second-timeoffenders are reinstated on the condi-tion that the offenders not drive witha positive BAC for 10 years. ConvictedDWI offenders who are apprehendedwith positive BACs receive a 1-yearadministrative license suspension orrevocation and may receive court-One method for measuring theeffectiveness of laws lowering BAClimits for DWI-convicted drivers is todetermine any pre- to post-law changesin the extent to which DWI offendersare involved in fatal crashes. Duringthe first 6 years after Maine adoptedits 0.05-percent BAC limit for driverswith prior DWI convictions, the pro-portion of fatal crashes involving driverswith prior DWI convictions declined25 percent, compared with the 6 yearsbefore the law took effect. The pro-portion of fatal crashes involving driverswith prior DWI convictions and BACsof 0.05 percent or higher declined 31percent (Hingson et al. 1998).Opponents of lowering the legalBAC limits for drivers have arguedthat these measures target and have no effect ondrivers with high BACs or prior DWIconvictions. However, during the 6years after Maine adopted its 0.05-percent law for drivers with priorDWI convictions, the proportion offatal crashes involving drivers withprior DWI convictions and BACs of0.15 percent or higher declined 35percent. During those 6 years, com-pared with the previous 6 years, therest of the United States experiencedminimal change in the proportion offatal crashes involving drivers withprior DWI convictions and positiveBACs. In the rest of New England, theproportion of fatal crashes involvingdrivers with prior DWI convictionsand positive BACs increased over 40percent (Hingson et al. 1998).In 1995 Maine became the firstState to adopt a zero-tolerance law for convicted DWI offenders. Mainewas also the first State to adopt a zero-tolerance law for drivers under age 21in 1983. Because of the benefits of thatlaw, by the end of 1998, all States hadadopted a zero-tolerance law for driversunder 21. Further research is neededto evaluate the effects of Maineond zero-tolerance law. Sanctions for Drivers Convicted Other measures to reduce recidivismamong persons convicted of DWImandatory alcoholism treatment, andlicense plate or vehicle impoundment.In a meta-analysis of over 200 studies,Wells-Parker and colleagues (1995)reported that alcoholism treatment wasassociated with a 7- to 9-percent reduc-tion in alcohol-related crashes comparedwith standard sanctions (e.g., jail orfines). Sanctions combining punish-monitoring and aftercare were moreeffective for both first-time and repeatoffenders than any single approach. Various studies also have found thatimpounding of DWI offenders vehi-cles or license plates reduces recidivism(Voas et al. 1997, 1998; Rodgers 1994),devices that prevent vehicle operationwhen a drivers breath alcohol exceedsa designated limit (Beck et al. in press). Restricting Access to Alcohol Another approach to reducing drinkingand driving is to lower the accessibilityof alcohol. Decreased accessibility canbe accomplished by raising the priceof alcohol through increased taxes,restricting both alcohol outlet densityan area) and hours of operation, main-taining State control of alcohol sales,and implementing laws to restrictpersons under age 21 (Kenkel andManning 1996; Gruenewald et al.1996; DeJong and Hingson 1998).Research shows that alcohol-relatedtraffic fatalities can be reduced byincreasing taxes on alcohol (Cook1981; Saffer and Grossman 1987 a ;1987 b ; Grossman et al. 1991) andenforcing laws that hold alcohol serversresponsible for actions taken by under-agepersons or intoxicated patrons 36 01020304050100 5+ drinks in one session in the past 2 weeksUsed past year97969594939291908988878685848382818088%41%75%31% of alcohol has steadily declined. SOURCE:Johnston et al. 1998. who were sold alcohol (McKnight 1993;McKnight and Streff1994; Wagenaarand Holder 1991 b ; Sloan et al. 1994).alcohol outlets byprivate owners ratherthan govern-with increased alcohol consumption(Gruenewald et al. 1993; Gruenewaldand Ponicki 1995; Wagenaar andHolder 1991 a ; Holder and Wagenaar1990), and increased outlet densityhas been associated with increasedalcohol-related traffic fatalities (Scribner Minimum Legal Drinking Age. the MLDA to 21 was designed to re-duce impaired driving by restrictingthe accessibility of alcohol to everyoneeven moderate drinking by adolescentdrivers. In 1984, when the FederalGovernment passed the NationalMinimum Drinking Age Act, 25 Stateshad MLDAs of 21. By 1988 all 50States had adopted MLDAs of 21.Survey results show a decline indrinking among young people follow-ing the increase in the MLDA. Theproportion of high school seniors whoreported drinking during the yearbefore being surveyed declined from88 percent in 1980 to 75 percent in1997. The proportion of high-schoolseniors who reported consuming fiveor more drinks on at least one occasionin the past 2 weeks declined from 41to 31 percent (Johnston et al. 1998)(see figure 5). In addition to the decreasein drinking observed among personsunder age 21 following the increase inthe MLDA, research also suggests thatraising the MLDA resulted in reduceddrinking among 21- to 25-year-oldswho grew up in States with a MLDAof 21 compared with those who grewup in other States (OMalley andWagenaar 1991). Numerous studies have indicatedthat raising the MLDA to 21 reducesalcohol-related fatal crash involvementamong drivers under age 21 (UnitedStates General Accounting Office1987). Of the 29 studies completedsince the early 1980s that evaluatedincreases in the MLDA, 20 studiesshowed significant decreases in trafficunder age 21. Only three studies foundno change in traffic crashes involvingyouth. The remaining six studies hadequivocal results (Toomey et al. 1996).States adopting MLDAs of 21 in thepercent decline in alcohol-related trafficdeaths among young drivers comparedwith States that did not adopt suchlaws. NHTSA (1998 b has prevented more than 17,300 traf-21 since 1975, approximately 700 todecade (see figure 6). Research hashave also reduced alcohol-relatedcrash involvement among 21- to 25-year-olds who grew up in States witha MLDA of 21 relative to those whogrew up in other States.Despite the declines in teenagewith prohibiting the purchase andpossession of alcohol by persons underage 21, underage youth throughoutthe United States still can obtain alcohol easily from many sources.Heightened enforcement of MLDAlaws can reduce youth access to alco-hol, however. Preusser and colleagues(1994) reported that alcohol sales tounderage youth declined dramaticallyfollowing an enforcement campaigntargeted at retail alcohol vendors. Thecampaign involved four over 10 months in which under-agemale police cadets attempted topurchase alcohol at liquor, grocery,and drug stores. Store owners receivedwarnings if cadets were able to pur-Stores that sold alcohol to the cadetsduring subsequent attempts werepenalized accordingly. Over the 10-cessful purchase declined from 59 to26 percent. Conclusion impair driving performance, this levels BAC to reach the legallimit for most drivers in the UnitedStates. In most States, the legal BAClimit for drivers age 21 and older is0.10 percent. Based on this standard, What Is Moderate Drinking? 37 0510 96959493929190898875Ñ87 Number of Lives (thousands)Figure 6 laws, 1975 to 1997. The minimum drinking age laws are credited with SOURCE:National Highway Traffic Safety Administration 1998 Preventing Impaired Driving the public may incorrectly assume thata driver is not impaired at BACs lowerthan 0.10 percent. In one study, only28 percent of the drivers surveyedreportedly believed that all drivers wouldalcohol that would increase a personBAC to 0.08 percent. However, virtu-ally all drivers are seriously impairedat this BAC level. Drivers must becomemore informed about alcohol consump-s effects on BAC levels and thevarious impairments a person experi-ences as a result of increasing BAClevels. Educational programs aboutmust be developed, implemented, and evaluated in an effort to reducealcohol-related traffic crashes. Research has demonstrated theeffectiveness of lowering the legalBAC limit to 0.08 percent for driversover age 21 and to 0.05 percent fordrivers with prior DWI convictions,as well as implementing zero-toler-ance laws for drivers under age 21.Further research is needed to deter-mine whether Maines zero-tolerancelaw for drivers with prior DWI con-victions will further reduce theirinvolvement in fatal crashes just aszero-tolerance laws did for driversrelated fatal crashes associated withMaines 0.05 law for convicted DWIStates as well. Increased knowledge ofcan reasonably consume beforebecoming impaired would help raises sense of responsibility, bothas drinkers and as drivers, ultimatelysaving thousands of lives.  Acknowledgment This article is dedicated to AshleyFrazier, who on December 22, 1995,was struck and fatally injured at 8:20a.m. by a driver with a BAC ofapproximately 0.08 percent while shewas waiting for a school bus in frontof her Westminster, Maryland, home. References Affairs.Alcohol and the driver. Journal of theAmerican Medical Association , E.Effects of Alcohol, Drugs and Traffic Safety , in press., R.D. Lower BAC Limits for Youth:Evaluation of the Maryland .02 Law . DOT HS806Transportation, 1992., R.D.; MOLEWATETTAThe risk of dying in alcohol-related automobile NewEngland Journal of Medicine cirrhosis, and auto accidents.In: Moore, M.H., Alcohol and Public Policy .Ð285.DEJ, R. Strategies toreduce driving under the influence. AnnualReview of Public Health 19:359Ð378, 1998.F, R. The effectsof alcohol on performance. In: Smith, A.P., and ance: Volume 2.Health and Performance . London:Academic Press, 1992. pp. 73, F. Alcohol Regulation and Motor Vehicle Mortality: Bethesda, Accident Analysis and Prevention 27(2):249Ð259,1995.GA.J. Alcohol availability and the ecology ofdrinking behavior. Alcohol Health & Research 17(1):39Ð45, 1993.GP. Access to alcohol: Geography and preventionfor local communities. Alcohol Health & Research 20(4):244Ð251, 1996. H Human Psychopharmacology: Clinical andExperimental 7(2):105Ð114, 1992.H, R. Prevention of drinking and driving. Alcohol Health & Research World 20(4):219Ð226,1996.H, J. Promotingsafety in adolescents. In: Millstein, S.; Peterson, Promoting the Health ofAdolescents: New Directions for the 21 New York, NY: Oxford Press, 1993. pp.Ð327. H, M.Lower legal blood alcohol limits for young Public Health Reports 109(6):738Ð744,1994.H, M.Lowering state legal blood alcohol limits to 0.08 American Journal of Public Health 86(9):1297Ð1299, 1996.H, M.Effects of Mainealcohol level for drivers with DWI convictions. Public Health Reports 113(5):440Ð446, 1998.H, M.Legislative strategies to reduce alcohol impaired Risk Taking and TrafficSafety . Washington, DC: National HighwayTraffic Safety Administration, in press., A.C. Effects ofthe elimination of a state monopoly on distilled British Journal of Addiction 85(12):1615Ð1625,1990. H, P.; SLEET, I. Alcoholand driving: Is the 0.05 percent blood alcohol Drug and AlcoholReview 10(2):151Ð166, 1991.J States in the U.S.The 39 Annual Proceedings, Oct . Des Plains, IL: Associa-Ð20.J, L.D.; OM National Survey Results on Drug Usefrom the Monitoring the Future Study, 1975-1997: . NIHPublication No. 98National Institute on Drug Abuse, 1998., W. Perspectives onalcohol taxation. Alcohol Health & Research World 20(4):230Ð238, 1996. M, R. Zero tolerance laws: Effectivepublic policy? Alcoholism: Clinical andExperimental Research 20(8):147AÐ150A, 1996.MA.F. Skills performance at low blood alcohol levels. Journal of Studies on Alcohol 46(6):482 The Effects Following Implementation of an 0.08BAC Limit and Administrative Per Se Law in 38 What Is Moderate Drinking? California . DOT HS No. 807National Highway Traffic Safety Administration. National Survey on Drinking and DrivingAttitudes and Behavior 1995 . DOT HS No.808Transportation, 1996.National Highway Traffic Safety Administration. Setting Limits, Saving Lives: The Case for 0.08BAC Laws . DOT HS No. 808Ð524, 1997. Traffic Safety Facts 1997: A Compilation of MotorVehicle Crash Data from the Fatality Analysis 806.Washington, DC: a . Traffic Safety Facts 1997: Alcohol. National b .Alcoholism. The PhysiciansPatients With Alcohol Problems . Bethesda, MD:the Institute, 1995.National Institute on Alcohol Abuse andAlcoholism. Alcohol Alert No. 35: Metabolism.CK, A.J. Server intervention: Accom-plishments and needs. Alcohol Health & Research 17(1):76Ð83, 1993. MCK, F.M. Effect ofenforcement upon service of alcohol to intoxi- AccidentAnalysis and Prevention 26(1):79Ð88, 1994. OÕM, A.C. Effectsof minimum drinking age laws on alcohol use, Journal ofStudies on Alcohol 52(5):478Ð491, 1991.PH.B. Policing underage alcohol sales. Journal ofSafety Research 25(3):127Ð133, 1994.R, A. Effect of MinnesotaÕimpoundment law on recidivism of multiple Alcohol Drugs and Driving 10(2):127Ð134, 1994.R The General Deterrent Impact ofCaliforniaLimit and Administrative Per se License Suspension Sacramento, CA: California Department ofMotor Vehicles, 1995., H.L. Law, science, and accidents: TheBritish Road Safety Act of 1967. Journal of LegalStudies 2(1):1Ð79,1973.R Confronting Drunk Driving: SocialPolicy for Saving Lives . New Haven, CT: YaleUniversity Press, 1982., M. Beer taxes, thelegal drinking age, and youth motor vehicle fatali- Journal of Legal Studies 16(2):351Ð374,1987 a. S, M. Drinking agelaws and highway mortality rates: Cause and Economic Inquiry 25(July):4031987 b. S, R.A.; M, D.P.; , J.H. Alcohol outlet density and motorvehicle crashes in Los Angeles County cities. Journal of Studies on Alcohol 55(4):447, F.A.; REILLY, C.Effects of prices, civil and criminal sanctions, and Journal of Studies on Alcohol 55(4):454, G.A. Effects of low to moderate dosesof ethanol on human driving-related perfor- Human Metabolism of Alcohol: Vol. I. Pharmaco- .The minimum legal drinking age: History, effec- Alcohol Health &Research World 20(4):213Ð218, 1996. Drinking . GAO/PEMD No. 87 The Relationship ofAlcohol Safety Laws to Drinking Drivers in Fatal Washington, DC: National HighwayTraffic Safety Administration, 1999. Accident Analysis andPrevention 29(5):635Ð642, 1997.V Accident Analysisand Prevention 30(5):651Ð655, 1998.W, H.D. Changefrom public to private sale of wine: Results from Journal of Studies on Alcohol 52(2):162Ð173,1991 a. W, H.D. Effects ofalcoholic beverage server liability on traffic crash Alcoholism: Clinical and ExperimentalResearch 15(6):942Ð947, 1991 b . W, R.;CM, M. Final resultsfrom a meta-analysis of remedial interventions Addiction 90(7):907Ð926, 1995. Z, P.L. Alcohol related relative risk of fataldriver injuries in relation to driver age and sex. Journal of Studies on Alcohol 52(4):302Ð310,1991.Z, K. Fatal crash involvement and laws Vol. 23, No. 1, 199939 Preventing Impaired Driving