This article was downloaded by: [Prevention Research Center], [Mr James C. Fell]On: 20 August 2013, At: 13:58Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Traffic Injury PreventionPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/gcpi20

Strengthening Impaired-Driving Enforcement in theUnited StatesRobert B. Voas a & James C. Fell aa Impaired Driving Center , Pacific Institute for Research and Evaluation , Calverton ,MarylandAccepted author version posted online: 10 Jan 2013.Published online: 14 Aug 2013.

To cite this article: Robert B. Voas & James C. Fell (2013) Strengthening Impaired-Driving Enforcement in the United States,Traffic Injury Prevention, 14:7, 661-670, DOI: 10.1080/15389588.2012.754095

To link to this article: http://dx.doi.org/10.1080/15389588.2012.754095

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Traffic Injury Prevention (2013) 14, 661–670Copyright C© Taylor & Francis Group, LLCISSN: 1538-9588 print / 1538-957X onlineDOI: 10.1080/15389588.2012.754095

Strengthening Impaired-Driving Enforcementin the United States

ROBERT B. VOAS and JAMES C. FELL

Impaired Driving Center, Pacific Institute for Research and Evaluation, Calverton, Maryland

Received 18 September 2012, Accepted 26 November 2012

Objective: Progress in reducing alcohol-impaired driving crash fatalities in the United States has stagnated over the last 15 years. Thisarticle reviews 2 current U.S. driving-while-intoxicated (DWI) laws and their enforcement with an aim toward generating opportunitiesto improve their enforcement approaches.

Methods: Impaired-driving enforcement methods in Europe and Australia are compared with those in the United States, and thelegal basis for current DWI criminal procedures is examined.

Results: An examination of relevant U.S. Supreme Court decisions and current legal practices indicates that the requirements foruse of breath test technology to determine blood alcohol concentrations of drivers on public roads are not entirely clear. Severalpotential methods for using field breath test technology to improve the detection of impaired drivers are suggested. These include (a)breath testing all drivers stopped for certain violations that have a high probability of involving an impaired driver, (b) breath testingall drivers at sobriety checkpoints, and (c) breath testing all drivers involved in fatal and serious injury crashes.

Conclusions: Breath test technology has enabled other countries around the world to adopt and implement enforcement strategiesthat serve as both general and specific deterrents to alcohol-impaired driving. Many of these enforcement strategies have been shownto be effective. If any one of these strategies can be adopted in the United States, further progress in reducing impaired driving isprobable. It may be necessary to provide the U.S. Supreme Court with a test case of breath testing all drivers at a sobriety checkpoint,depending upon whether or not a police agency is willing to use that strategy.

Keywords: impaired driving, blood alcohol concentration (BAC), breath test technology, per se laws, enforcement procedures, U.S.Supreme Court decisions

Introduction

The Problem

Motor vehicle crashes are the leading cause of death for Ameri-cans age 4 and ages 11 through 27 (Subramanian 2012). In theUnited States, alcohol-impaired driving was responsible for31 percent of fatal crashes in 2010 and resulted in more than10,000 deaths (National Center for Statistics and Analysis[NCSA] 2012a). Alcohol-related crashes cost the United Statesan estimated $129 billion in 2006 (Zaloshnja and Miller 2009).A national roadside survey of nighttime weekend drivers in2007 indicated that 2 percent of drivers on the road have ille-gal blood alcohol concentrations (BACs; Lacey et al. 2009).Zador et al. (2000) estimated that only one of 88 drivers withillegal BACs is arrested for driving while intoxicated (DWI),whereas Beitel et al. (2000) estimated one in 200 using a differ-ent method. No comparable probabilities could be found forother countries around the world. A national telephone surveyof more than 10,000 drivers showed that U.S. drivers admitted

Address correspondence to Robert B. Voas, Pacific Institutefor Research and Evaluation, 11720 Beltsville Drive, Suite 900,Calverton, MD 20705-3111. E-mail: voas@pire.org

to 85.5 million drinking-and-driving trips in the past 30 daysduring 2008 (Moulton et al. 2010). Although the United Statesmade progress in reducing impaired driving between 1982 and1997, little has been achieved since that time (Dang 2008; Fellet al. 2009).

Background

A recent report from an expert committee established by theNational Research Council (NRC 2010) to compare highwaysafety programs in high-income “benchmark” nations to thosein the United States concluded that

the gap between traffic safety progress in the United Statesand other high-income countries deserves the attention oftransportation administrators and the public because it in-dicates that the United States may be missing importantopportunities to reduce traffic deaths and injuries. (p. 9)

One area where the United States may be missing an opportu-nity, which is now a common practice in other nations, is theuse of sobriety checkpoints with field breath tests in the en-forcement of impaired-driving laws. As the NRC report noted,the United States has made little progress in reducing traffic

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

662 Voas and Fell

Fig. 1. Proportion of all drivers involved in fatal crashes estimatedto have been legally intoxicated (BAC ≥ 0.08), 1982–2010. Source:Fatality Analysis Reporting System (NHTSA 2012).

fatalities since the mid-1990s (excluding the recent U.S. re-cession years of 2008 and 2009). This has been particularlytrue for impaired-driving–related fatalities because the per-centage of all fatal crashes involving an impaired driver hasnot changed from 1997 to 2008. When the recession in 2008and 2009 reduced total fatalities by 9.3 and 9.7 percent (NCSA2011), respectively, the percentage that involved an impaireddriver actually increased slightly (NCSA 2010). The lack ofprogress in reducing impaired-driving–related fatal crashes(see Figure 1) has resulted in calls by both citizen activist or-ganizations, such as Mothers Against Drunk Driving (2006),and researchers (e.g., Robertson et al. 2009) for increased at-tention to the impaired-driving problem and increased ex-penditures on impaired-driving enforcement. Although greatprogress was made between 1982 and 1997, Figure 1 shows thatsince 1997, the proportion of drivers involved in fatal crasheswho were intoxicated has remained at 20 to 22 percent. Moth-ers Against Drunk Driving is currently supporting legislationcalling for mandatory alcohol ignition interlock devices for alloffenders convicted of DWI. It remains to be seen how manystates adopt such legislation and whether it will have a gen-eral deterrent effect similar to what per se and administrativelicense revocation laws provided (Wagenaar and Maldonado-Molina 2007; Wagenaar et al. 2007). Unfortunately, both theability to focus public attention on impaired driving and thecapability to increase funding for DWI enforcement is limitedby the current U.S. recession and the growing concern withterrorism. This suggests the importance of finding methodsto improve the effectiveness of current laws and policies with-out requiring additional funding for DWI enforcement andadjudication. This study examines the gap described by theNRC (2010) between impaired-driving enforcement policiesin the United States and other similar industrialized nations.The objective is to improve the effectiveness of current U.S.enforcement programs without broad new legislation or sub-stantial increases in funding.

Methods

Procedures for stopping and arresting drivers for DWI inthe United States and other countries were examined. Two

Table 1. The 3 steps in the DWI apprehension and arrest process

Australia Sweden Britain United States

Step 1. Select vehiclea. At random Xb. At checkpoints Xc. Based on impaired driving X Xd. Based on traffic offense X

Step 2. Roadside interviewa. Mandatory breath test X X Xb. Field sobriety test X

Step 3. At the police stationa. Mandatory breath test X X Xb. Voluntary breath test/license

penaltyX

key elements in the impaired-driving enforcement processare

• the policies for stopping vehicles to check for impaireddrivers and

• the procedures for breath testing drivers to determine theirBACs.

Currently, apprehension of impaired drivers in the UnitedStates proceeds in 3 steps (see Table 1, which compares theU.S. process with Australia, Sweden, and Great Britain):

1. a vehicle in the traffic flow is selected for investigation;2. the driver is interviewed and, if impairment is suspected,

that driver may be required to perform sobriety tests; and3. if there is probable cause to believe that the driver is im-

paired, the driver is arrested and transported to the policestation for evidential testing and booking.

When the first impaired-driving laws were adopted early inthe 20th century, the initial enforcement procedures requiredpolice officers to observe the driver’s behavior and present suf-ficient evidence in court to support the ambiguous definitionof “intoxicated driving.” During that period, DWI laws in Eu-ropean nations typically required that the suspect be examinedby a physician (National Safety Council 1976).

In Sweden in the 1920s and 1930s, Widmark (1932) pub-lished a series of articles that described the relationship ofalcohol consumption to BACs. This led to the developmentof highly accurate blood and urine alcohol test procedures foruse in the third phase of the apprehension process in the po-lice station (Jones 2000). Initially, these tests were used only ascorroborating evidence to augment the officer’s or physician’stestimony. In 1936, however, Norway became the first nationto enact a per se law establishing a BAC of 0.05 g/dL as an of-fense for drivers. This per se law provided authority to sanctiondrivers for DWI based upon the result of the BAC test—noother evidence (e.g., behavioral, field tests) was needed. If adriver was over the illegal BAC limit (.05 g/dL in this in-stance), that driver was guilty per se of DWI and appropriatesanctions were administered. Over time, most European na-tions have followed Norway’s lead (Jones 2000). The UnitedStates was slow to adopt BAC per se laws. By the time theNational Highway Traffic Safety Administration (NHTSA)recommended such laws (NHTSA 1972), only 2 states had ac-tually adopted BAC per se laws. However, all 50 states and the

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

Impaired-Driving Enforcement in the United States 663

District of Columbia have such laws today set at 0.08 g/dLas the illegal limit (Mercer et al. 2010). A BAC per se lawpotentially streamlines the apprehension and prosecution ofimpaired drivers, because in concept at least, it is based ononly 2 findings: the accused was in charge of the vehicle (wasthe driver) and had a BAC at or higher than the illegal limit.If both findings are proven, the driver is guilty, per se, ofDWI.

Borkenstein’s invention of the breathalyzer provided a low-intrusion device for measuring BAC that officers could op-erate accurately in the police station (Borkenstein and Smith1961). This new technology avoided the need to take offend-ers to a hospital or phlebotomist and wait for laboratory re-sults of the blood analysis. This stimulated the passage ofper se laws in the United States. In the late 1960s, progressin breath testing technology led to the development of fieldBAC breath test devices. Early handheld devices that thepolice could use at the roadside were crude; they requiredthe suspect to inflate a balloon through a tube of reagentthat changed color on contact with ethanol (Dubowski 1974;Jones 2000; Voas 1988). These crude devices were supersededby the development of handheld electronic sensors based onfuel cell technology in the early 1970s (Jones 2000). This pro-vided the police with a method for rapidly conducting ac-curate BAC tests at the roadside in step 2 of the apprehen-sion process, rather than having to wait until the suspect hadbeen transported to the police station for step 3 processing(Table 1).

Application of Field Breath Tests Under BAC Per se Laws inSweden and Australia

In Sweden and Australia, the availability of portable fieldbreath test devices encouraged the use of random stop-and-test procedures. Sweden instituted a checkpoint procedure thatempowered police to establish roadblocks at which all cars, ora random selection of cars, could be stopped and their driversinterviewed and breath tested (Ross 1984). Australia extendedthis random breath test (RBT) enforcement technique to itslogical limit by providing police with the authority to stop anyvehicle at any time and demand a breath test. An over-the-limitBAC result immediately led to arrest and a second evidentialtest at the police station. Test refusal was criminalized and car-ried the same sanctions as a conviction for impaired driving(Homel 1990, 1993). Because any vehicle could be stopped atany time, this RBT procedure essentially moved breath test-ing to the front end of the apprehension process. Thus, RBTrepresents a full actuation of the BAC per se concept in the en-forcement of impaired driving because only the 2 fundamentalfindings of the BAC per se law (being in charge of the vehicleand having a BAC greater than or equal to the limit) are in-volved in apprehension under the impaired-driving system (seeTable 1). This can be characterized as a “chemistry-based” en-forcement system (Voas and Lacey 1990). The RBT procedurebecame the principal method of impaired-driving enforcementin Australia, with some provinces testing a number of driverseach year equal to the total number of licensed drivers in theprovince (Homel 1993).

The American Experience with Checkpoints

In 1967, when the U.S. Department of Transportation was es-tablished, there was a limited tradition in some states of con-ducting checkpoints to verify that drivers had valid licenses.Strengthened by the Swedish example, this tradition was usedas a foundation for the first national DWI enforcement pro-gram in 1969, which was conducted by the newly formedNational Highway Safety Bureau, the predecessor to today’sNHTSA. In that program, 35 Alcohol Safety Action Projectswere implemented in communities across the United States(Levy et al. 1977). Although more than one enforcement pro-cedure was used in those projects, sobriety checkpoints wereemphasized. The success of this strategy throughout the yearshas resulted in checkpoints becoming one of the NHTSA’s pri-mary recommended DWI enforcement procedures (Compton1983; NHTSA 1990, 1993, 1999, 2000).

In the United States, however, sobriety checkpoints arequite different from those in Sweden or the RBT programs inAustralia because handheld preliminary breath testers (PBTs),the central feature of checkpoints outside the United States,are not often used. Rather, at checkpoints in the United States,police engage in brief interviews with the motorist in an ef-fort to identify impaired drivers. Although officers are trainedto detect alcohol impairment, the interview time provided atcheckpoints is too short to permit high accuracy of detec-tion in identifying high-BAC drivers (Fell et al. 2004). Thisinterview procedure has been shown to be significantly less ef-ficient than using breath tests. It results in half of the over-the-BAC-limit drivers passing through the checkpoint undetected(Ferguson et al. 1995).

Henstridge et al. (1997), in a time-series analysis of datafrom 4 Australian states, found that RBT was twice as ef-fective as selective checkpoints. Sherman (1990) found thatin Queensland, Australia, the RBT program resulted in a 35percent reduction in fatal crashes compared to a 15 percentreduction due to selective checkpoints. Sherman (1990) esti-mated that every increase of 1000 drivers in the daily RBT ratecorresponded to a decline of 6 percent in all serious crashesand 19 percent in single-vehicle nighttime crashes. Peek-Asa(1999) and Shults et al. (2001) found small differences in RBT’sinfluence on crashes in Australia versus the effectiveness of so-briety checkpoints in the United States. However, in a reviewby Solomon et al. (2011), RBT in Australia and Europe wasdeemed substantially more effective than sobriety checkpointsas they are now conducted in Canada and the United States.Lund and Jones (1987) noted, however, that the efficiency ofU.S. checkpoints could be increased when a passive alcoholsensing device was used on all motorists passing through acheckpoint (see later discussion of passive alcohol sensors).

The limited use of field breath testers in the U.S. law en-forcement system stems from the current interpretation ofthe Fourth Amendment to the U.S. Constitution, which pro-hibits unreasonable searches and seizures (meaning withoutindividualized suspicion). Checkpoints involve both a seizure(stopping the car) and a search (breath testing if a field breathsensor is used). U.S. courts have concluded that random stop-ping of a vehicle is an unreasonable “seizure” under theFourth Amendment (Delaware v Prouse, 440 U.S. 648, 1979).

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

664 Voas and Fell

However, an exception was made for stops in special “sobrietycheckpoint” operations if they met specific conditions estab-lished by the U.S. Supreme Court based on a balancing test(Michigan Dept. of State Police v Sitz, 496 U.S. 444, 1990;hereinafter referred to as Sitz). The balancing test consideredthe need for the state to protect the public weighed againstthe small intrusion involved in stopping a driver for a briefinterview.

Use of Preliminary Breath Tests at Checkpoints

Unfortunately, the other key element of the sobriety check-point procedure—the use of a field breath test device (a“search”)—was not used to detect the drinking of all driversstopped at the checkpoint in the U.S. Supreme Court Sitz case.Rather, the officers were required to detect impaired motoriststhough the observations they could make during a brief inter-view, which averaged 25 seconds (Sitz 1990). Thus, the decisiondid not specifically deal with breath testing. However, ChiefJustice Rehnquist, delivering the majority opinion, concludedthat:

Experts in police science might disagree over which of sev-eral methods of apprehending drunken drivers is preferableas an ideal. But for purposes of Fourth Amendment analy-sis, the choice among such reasonable alternatives remainswith the government officials who have a unique under-standing of and responsibility for, limited public resources,including a finite number of police officers. (Sitz, 496 U.S.at 453–454)

This appeared to provide discretion for police departmentsto require a breath test of all drivers stopped at checkpoints.However, the decision was based on balancing the state’s inter-est in preventing drunk driving against the degree of intrusionupon the motorist in which the degree of intrusion was judgedto be “slight” (Sitz, 496 U.S. at 451). Whether the PBT testwould be a slight intrusion was not considered. The absenceof a mandatory breath test for all motorists in the Sitz caseleaves the issue unresolved as to whether such a requirement ata checkpoint would increase the level of intrusion sufficientlyto tip the balance in favor of motorist privacy concerns. Thestatement in the Rehnquist opinion, that detection methodsshould be left to the discretion of police experts, appears to in-vite police agencies to create a test case for this issue by breathtesting all motorists stopped at sobriety checkpoints.

To date, the requisite case has not occurred, so in mostjurisdictions, all drivers are not required to provide a prelimi-nary breath test at the beginning of the apprehension processas they would be in Sweden or Australia or in 44 other coun-tries around the world (Solomon et al. 2011). It will likely takesubstantial persuasive powers to convince a police agency (oragencies) to breath test every driver at a checkpoint. It couldresult in a public backlash that most agencies want to avoid.However, if the local community is educated on the balancebetween the need to protect the public and the small intrusionof breath testing at a sobriety checkpoint, and they buy intoit, that public support could influence police action.

In the United States, however, PBTs may be used in somelocations if the officer has reason to believe that the driver

Table 2. Police detection of drivers with BACs ≥ 0.10 g/dL: withand without the use of PAS

Percentage detected

Without PAS With PAS

Sobriety checkpoints:Charlottesville, VA 45 68Fairfax, VA 55 71

Routine patrol:Columbus, OH 69 77

Special DUI patrol:Chattanooga, TN 88 94

Source: Ferguson et al. (1995); Wells et al. (1996).

is impaired. Generally, this occurs after completion of thebehavioral standardized field sobriety tests (SFSTs) to confirmthe officer’s judgment that the suspect is over the BAC limit.

Results

Efforts to Overcome Opposition to Preliminary Breath Testsand Sobriety Checkpoints

In an effort to overcome the current limitations on preliminarybreath testing by making the test less intrusive, a fuel-cellpassive alcohol sensor (PAS) was developed over 2 decadesago that, when held 4 to 6 inches in front of the suspect’s face,can detect with reasonable accuracy the alcohol in the expiredair of drivers who have been drinking (Voas et al. 2006).Evaluations of the use of PAS devices have demonstratedthat, without the PAS units, police officers miss up to half ofthe drivers who are over the BAC limit when passing throughcheckpoints (Ferguson et al. 1995). The use of PAS devices,which are typically built into the officers’ flashlights, has beenshown to identify half of those over-the-limit drivers whocurrently escape detection at checkpoints (see Table 2) whiledecreasing the number of drivers unnecessarily detained atcheckpoints because of ambiguous initial screening (Fergusonet al. 1995; Wells et al. 1996). Despite this evidence for theireffectiveness, it has been difficult to persuade officers to usePAS devices at checkpoints (Fell et al. 2008; Voas et al. 2005).The reasons for this are not entirely clear, but apparently,officers find them awkward to handle and are not fullyconvinced that, given their own ability to detect impaireddrivers, the PAS devices are needed.

Furthermore, despite the evidence that checkpoints are themethod of choice for high-visibility enforcement (Elder et al.2002), many police departments currently resist the use ofcheckpoints. There are apparently 2 primary reasons: the be-lief that sobriety checkpoints require a substantial number ofofficers (and therefore are expensive) and that checkpoints re-sult in few arrests (and therefore appear ineffective; Fell et al.2004). Research has identified 2 procedures—low-staff check-points and PAS devices—that deal with those issues. Stud-ies have shown that sobriety checkpoints conducted by rela-tively few officers are as effective in reducing alcohol-relatedcrashes as those involving large groups of officers (Lacey et al.2006; Stuster and Blowers 1995), and the use of PAS devices

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

Impaired-Driving Enforcement in the United States 665

results in apprehension rates per officer-hour as high as thosein traditional dedicated patrol operations (Voas 2008). These2 concepts have been combined in a “PASpoint EnforcementSystem” that can be integrated with typical police patrol op-erations without employing officers on overtime (Voas et al.2005). This procedure will permit conducting of checkpointsregularly by medium- to small-sized police departments.However, this system remains to be widely demonstratedand adopted by police departments, although the Fresno,California, police department has expressed a willingness totest the PASpoint concept over a one-year period.

Current Administrative Search Programs

A new factor since the U.S. Supreme Court establishedthe constitutional basis for sobriety checkpoints is home-land security, under which the public has come to expecthighly enhanced security procedures: (1) going through mag-netic detectors and having purses and briefcases inspectedin order to enter government buildings; and (2) under the2001 Aviation and Transportation Security Act (Aviationand Transportation Security Act, S. 1447–107th Congress,Pub. L. 107–71; 2001), having luggage, shoes, and coats X-rayed, and going through body scanners and pat-downs atairports. The concern for transportation safety and securityhas also justified legislation requiring random testing of em-ployees in travel-related industries (Brady et al. 2009). Thisnew security environment has reduced the public expectationof privacy and presumably increased the public’s acceptance ofintrusive searches compared to the period when the SupremeCourt handed down its opinion on Sitz. The current accep-tance of the intrusive air travel screening is based on the tragicevents of September 11, 2001, in which almost 3000 peopledied (2752 at the World Trade Center, 184 at the Pentagon,and 45 in the airplane that crashed in Pennsylvania). Since thattime, no terrorism deaths have occurred on U.S. airlines. Muchless understood or communicated to the public is the fact thatimpaired driving has resulted in 3000 deaths every 3 months onaverage since 9/11. In a recent national poll of U.S. drivers, 87percent said that people driving after drinking alcohol was avery serious threat to their personal safety, and 98 percent saidthat they personally considered drinking and driving as unac-ceptable (AAA Foundation for Traffic Safety 2010). Therefore,a better understanding by the public of the extent of the lossescaused by alcohol-impaired drivers should result in strongersupport for impaired-driving laws, similar to the support beinggiven to airline and homeland security. Because the key basisfor the Sitz decision was a balancing test of the need for thestate to protect the public versus the extent of the intrusionon the individual driver, the public’s view and the SupremeCourt’s evaluation of the extent of the intrusion involved in afield breath test should be lower in the current context.

Justice Alito’s 2007 opinion, while serving on the 9th CircuitCourt of Appeals, upholding airport checkpoints is an indica-tion that the current U.S. Supreme Court might accept BACbreath tests at roadside checkpoints. An interesting statementin his opinion is that searches are reasonable if they escalateinvasiveness after a lower level of screening discloses a reason

to conduct a more probing search (Rosen 2010). This raisesthe issue of the relative intrusiveness of the BAC breath testversus SFSTs. Administering a PBT requires 30 s or less andcan be conducted in the privacy of the suspect’s car, whereasthe SFSTs require 10 to 20 min and are conducted on theroadside usually in full view of the public. It could be arguedthat the first action of an officer who has a suspicion that adriver is impaired should be the less intrusive PBT test, thenmoving on to the SFSTs only after having obtained an in-dication from a PBT that the suspect is over the BAC limit.Currently, this more efficient procedure is not being used be-cause of concern that the officer will overlook the possibility ofdrug impairment, but the absence of an indication that alco-hol is accounting for the driver’s behavior should immediatelyalert the officer to the possibility of drug impairment.

The importance of the public’s reaction to being stoppedwas recognized in the dissent to Sitz written by Justice Stevenswith concurrence by Justices Brennan and Marshall. Theprimary focus of the dissent was based on the intrusionproduced by surprise when unexpectedly confronted bya sobriety checkpoint. The dissent distinguished betweenmobile checkpoints that “. . . depend upon surprise . . .” and“. . . a host of other investigatory procedures, . . .” includingstationary checkpoints that do not depend on surprise (Sitz,496 U.S. at 473). The dissent notes the “. . . common practiceto require every prospective airline passenger or every visitorto a public building to pass through a metal detector . . .”(Sitz, 496 U.S. at 473) and includes the suggestion that “. . .

a state could condition access to its toll roads upon not onlypaying a toll but also taking a uniformly administered breath-alyzer test” (Sitz, 496 U.S. at 474). This position appears tominimize the concern over BAC breath testing in relation tothe significance of the conditions of the vehicle stop (seizure).

Use of Field Breath Tests in Traditional Impaired-DrivingEnforcement

The movement of the biological testing from the police stationor hospital to the roadside in the second half of the 20th cen-tury was dubbed the “Scandinavian System” by Ross (1975)because of its origin in Sweden. It was adopted in modifiedform by most industrialized nations. The Netherlands, France(Ross 1984), and Australia (Homel 1990, 1993, 1994) emulatedthe Swedish checkpoint system; other countries provided forbreath testing devices for officers to use in their traditional en-forcement operations not related to checkpoint stops. Britaininitiated an innovative procedure in the Road Safety Act of1967 (Ross 1973). That legislation provided British police withthe authority to require a breath test of any driver they stoppedif any one or more of 3 conditions was met: (1) the officer hadreason to believe that the driver was impaired, (2) the driverwas involved in a crash, or (3) the driver was guilty of commit-ting another traffic offense. When implemented, the BritishRoad Safety Act had a dramatic effect, not only markedly re-ducing alcohol-related crashes but also reducing non-alcohol-related crashes apparently by persuading even sober drivers toexert greater care when driving to avoid the possibility of beingbreath tested for BAC. Unfortunately, this success diminished

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

666 Voas and Fell

because the additional resources required for implementationof this law, including police overtime and pay, led to the strat-egy being abandoned. It is very likely that the large numberof moving violations cited by officers at traffic stops discour-aged the continued implementation of the method. Tellingly,a short revival of the system by one sheriff produced a rapiddrop in alcohol-related crashes, demonstrating anew that thisapproach could have continued to be effective (Ross 1977).Nevertheless, despite this demonstration, enforcement of thelaw was halted once again.

The British Road Safety Act points to some possible op-tions for strengthening DWI enforcement in the United States.That law provided for BAC breath testing under 3 conditions:(1) the officer had reason to believe that the driver was im-paired (current practice in the United States), (2) the drivercommitted a traffic offense, and (3) the driver was involvedin a crash. The first condition—stopping based on the policeofficer’s suspicion—is one already used in the United States.Stuster (1997) conducted research on the development of cuesavailable to police officers on patrol that indicate that the driveris impaired. The guide that resulted from the research is shownin Figure 2. On the left are the vehicle maneuvers that indicatethat the driver may be impaired; on the right are the driver be-haviors that the officer can observe once the vehicle has beenstopped. Note the relatively high probability values of a driverhaving a BAC ≥ 0.08 g/dL associated with the vehicle maneu-vers ranging from a probability of 0.35 to 0.90. These have beenaccepted by the courts as sufficient cause for stopping (seizing)the vehicle under the Fourth Amendment. Presumably, suchsigns provide the individualized suspicion required to performa search (a BAC breath test) under the Fourth Amendment;however, PBTs are rarely used until standardized field sobrietytests are performed, which requires considerably more time.

Testing Drivers Stopped and Cited for Traffic Violations

The second provision of the British Road Safety Act providedthat a BAC breath test could be required if the driver was guiltyof a traffic law violation. This may have been overreaching,

Fig. 2. DWI driver detection guide. Source: Stuster (1997).

which contributed to the failure of the police to continue toenforce the law fully. Because there are so many opportunitiesto ticket drivers for relatively minor moving violations when,in most cases, there is no evidence of impairment, officersmay have been reluctant to take the time to breath test driversqualifying under the traffic violation provision.

Requiring a breath test for drivers cited for a moving trafficoffense avoids the issue of random stopping, because it pro-vides individualized suspicion that an offense has been com-mitted. Though it seems unlikely that the U.S. public wouldsupport the testing of every driver receiving a traffic citation,there are offenses (e.g., reckless driving) that are a clear andserious threat to safety. These offenses are so far outside thelimits of normal driving behavior that they suggest the possi-bility of alcohol impairment, which could make BAC breathtesting reasonable and acceptable. Perhaps speeding 20 milesan hour over the speed limit and driving with a suspendedlicense would also meet this criterion. Perhaps the 4 behaviorsshown in Figure 2 would suffice. States could adopt legisla-tion requiring a screening BAC breath test for such offensesbecause the traffic stops would be for cause and the behav-ior suggestive of intoxication. In this procedure, the drivingbehavior that justified the stop would also provide the basisfor conducting a PBT at roadside, and an over-the-limit resultwould provide probable cause to make a DWI arrest and torequire an evidential breath test.

Testing Drivers in Crashes for BAC

The final basis for field breath testing in the British RoadSafety Act was involvement in a crash (Ross 1973). Cer-tainly, the fact that a third of all fatal crashes in the UnitedStates involve a driver at or higher than the 0.08 BAC ille-gal limit should clearly make it “reasonable” to require alldrivers in such crashes to be tested. However, as the Fatal-ity Analysis Reporting System (FARS) record demonstrates,the lack of a mandatory testing law has made this difficultto achieve (NCSA 2008). Because the rate of alcohol involve-ment is higher in serious injury and fatal crashes comparedto property-damage–only crashes, it should be possible to es-tablish public support for identifying involvement in a seriousinjury crash as a basis for further investigation. This investi-gation would include a mandatory PBT screening test that, ifover the BAC limit, provides probable cause for an arrest andan evidential test. Currently, the BAC testing rate for driverskilled in crashes is 71 percent, but only 30 percent of survivingdrivers in fatal crashes are tested for a BAC (NCSA 2012b).A recent NHTSA report recommended that police adopt apolicy to test all surviving drivers involved in fatal crashes fora BAC (Casanova et al. 2012). The National TransportationSafety Board (2012) recently recommended that a goal of atleast 60 percent of surviving drivers in fatal crashes be testedfor a BAC.

In sum, there are abundant and compelling opportunitiesfor strengthening the DWI enforcement system in the UnitedStates if society accepts moving toward a more aggressive useof field alcohol screening. There are 2 limitations on the useof screening devices in the United States not encountered in

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

Impaired-Driving Enforcement in the United States 667

similar countries abroad. First, the opportunities for the appli-cation of BAC breath testing are limited by requiring suspicionof impairment as a prerequisite for using PBTs. Second, stateshave not made either BAC screening test refusal or an over-the-limit BAC test screening result sufficient evidence to makean arrest and require an evidential breath test, as is commonpractice in other countries. This failure to take advantage ofthe PBT technology results in a failure to detect a significantnumber of over-the-limit drivers (Ferguson et al. 1995). Inaddition, the reliance on the observation of impairment re-quires extensive field observations and extended paperworkby the police, often requiring 3 or more hours to make a DWIarrest, discourages officers from becoming involved in DWIenforcement (Simpson and Robertson 2001).

Racial Profiling and BAC Per se Laws

A key feature of BAC per se laws is that they define the offensein objective terms. Short of videotaping the field arrest activity,behavioral observations are necessarily somewhat subjective.This can open the arrest process to the possibility of bias inthe judgments that officers must make in apprehending animpaired driver. Much of the time, there are no witnesses tothe interviews or the field sobriety tests of the offender. Often,the offender is in a weak position to challenge the officer’sstatements in court. Therefore, Justice Stevens, in his dissentfrom the Sitz decision expressed concern with the officer’sdiscretion in examining an impaired-driving suspect:

A Michigan officer who questions a motorist at a sobri-ety checkpoint has virtually unlimited discretion to de-tain the driver on the basis of the slightest suspicion. Aruddy complexion, an unbuttoned shirt, bloodshot eyes, ora speech impediment may suffice to prolong the detention.Any driver who had just consumed a glass of beer, or evena sip of wine, would almost certainly have the burden ofdemonstrating to the officer that his or her driving abilitywas not impaired. (Sitz, 496 U.S. at 464–465)

The PBT can quickly (within 30 s) provide the officer withan objective, highly accurate indication of the true extent towhich the motorist has been drinking. The PBT result couldbe the basis for inviting the driver out of the car for the SF-STs. That decision is important because the SFSTs are usuallyperformed in public with the potential to cause considerableembarrassment to an innocent, unimpaired driver, as well astaking up the officer’s time unnecessarily. Where breath mea-surement has been used at the front end of the checkpointinterviews, the number of DWI arrests has increased by 50percent, and the number of drivers unnecessarily detained forSFSTs has been reduced (Ferguson et al. 1995). A study byZador et al. (2000) using arrest information and survey dataindicated that police conduct about 6 DWI investigations foreach DWI arrest they make. Using the PBT more aggressivelywould reduce the number of fruitless investigations for the of-ficers, as well as reduce the number of impaired drivers whoescape detection and the extent of unnecessary intrusion onthe public. A well-managed preliminary and evidential chemi-cal test program can provide the best assurance for the driving

public that the data used in the criminal justice system to pros-ecute impaired drivers are accurate and unbiased. However,in the current U.S. system, where conducting a preliminarybreath test and an evidential test must be based upon suspicionof alcohol impairment and where both tests can be refusedwith minimal or no consequences, the detection of behavioralimpairment by the officer becomes necessary (Simpson andRobertson 2001; Zwicker et al. 2005).

When field breath tests for BAC measures were first intro-duced, there was some speculation that many suspects wouldwelcome a chemical test as a method of proving their inno-cence. But that was overshadowed by the strong advice comingfrom many defense attorneys to refuse the chemical test be-cause an over-the-limit result would make avoiding convictionmore difficult. Nevertheless, the PBT could be used in thecase cited by Justice Stevens to shortcut the investigation andavoid subjection to the highly intrusive SFSTs. It could evenimpose on the officer the responsibility to offer the test beforeproceeding further in the investigation. This potential utilityfor the innocent driver is prevented, however, by the presenceof the simultaneous historic criminal impaired-driving law,which allows the arrest of a driver with a low or zero BACif there are signs of driving impairment. As noted, the policyin most jurisdictions is not to use the PBT at the beginningof an investigation to prevent officers from overreliance onthe device and missing drug-impaired drivers. The trade-offbetween missing the substantial number of impaired driverswhose tolerance to alcohol makes them difficult to detect withthe SFSTs versus the number of drug-impaired drivers ar-rested based on their performance on the SFSTs has yet to bedetermined, but it raises the issue of the overall efficiency ofcurrent DWI apprehension policies.

Discussion

Progress in reducing alcohol-impaired driving in the UnitedStates has stagnated. In 2006, 34 percent of fatally injureddrivers in the United States had BACs ≥ 0.05 g/dL comparedto 29 percent in Australia (Watson 2010) and only 17 percentin Sweden (Jones 2010) (see Figure 3).

Fig. 3. Proportion of fatally injured drivers with BACs >

0.05 g/dL in 2006. Sources: United States, FARS (NHTSA 2012);Australia, Watson (2010); Sweden, Jones (2010). Note: The illegalBAC limit for driving is 0.08 g/dL in the United States, 0.05 g/dLin Australia, and 0.02 g/dL in Sweden.

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

668 Voas and Fell

Although the police in the United States are as wellequipped and well trained for impaired-driving enforcementas any in the world, the United States has fallen behind otherindustrialized nations in the efficiency of enforcement opera-tions. This is due to the failure to exploit fully the programsand technologies based on BAC per se laws. This failure can betraced to the lack of clarity in judicial decisions regarding theuse of BAC-sensing technologies and resulting police policiesthat minimize the use of field breath test devices. There is aneed to examine current enforcement procedures and revisitlegal precedents that govern the DWI criminal justice system.Current enforcement procedures in Sweden and Australia arenot considered overly intrusive by the citizens of those coun-tries and, in fact, are considered acceptable for public safetyreasons.

Several opportunities exist for clarifying the legal issuesthat are influencing current enforcement practices via carefullyplanned demonstration programs. Among them are

1. requiring a preliminary breath test of all motorists stoppedat a sobriety checkpoint;

2. requiring a preliminary breath test of all motorists stoppedfor exhibiting driving behaviors that have been shown to beindicative of driver impairment (see Figure 2); and

3. requiring a preliminary breath test of all motorists involvedin a serious injury or fatal crash.

Testing the efficacy of these concepts in demonstrationprojects could clarify current laws and lead to enforcementpolicy changes that could be much more effective and efficientthan those currently in practice. It is likely that these strategieswill pass the balancing test of minimal intrusion for the publicsafety interest if they are challenged. The main barrier is topersuade a police agency to test these concepts and be willingto accept the possible public backlash and court challenges.Developing public support for these strategies will be a key tosuccessful implementation.

To achieve the needed strengthening of the DWI enforce-ment system in the United States, it will be necessary to harnessthe public’s concern with the problem through an improvedunderstanding of the risk that all road users face because ofalcohol-impaired drivers. With better recognition of the emo-tional and societal costs of impaired driving to the nation, thepublic may demand greater protection from impaired driversand support DWI enforcement programs that are more effec-tive. Despite the clear evidence that alcohol-related crashesare a major source of death, injury, and societal costs, theUnited States has failed to take full advantage of establishedtechnologies for identifying drivers impaired by alcohol. If theU.S. DWI enforcement system remains almost completely de-pendent on demonstrating behavior impairment, it is unlikelyto either increase the apprehension of impaired drivers or re-duce the number of road users fatally injured by drivers withillegal BACs.

Acknowledgments

Funding for this article was provided partially through a grantfrom the National Institute on Alcohol Abuse and Alcoholism

(NIAAA, Grant Nos. R01 AA018352 and P20 AA017831).The authors thank the following individuals for their initial re-view of this article and the important comments and guidancethey provided: Erika Chamberlain, Heidi Coleman, RichardCompton, Michael Greene, J. T. Griffin, Ralph Hingson, AndyMurie, Robert Solomon, Stephen Talpins, Alexander Wage-naar, and Michele Fields.

References

AAA Foundation for Traffic Safety. 2010 Traffic Safety Culture Index.Washington, DC: AAA Foundation for Traffic Safety; 2010.

Adoption of 0.08 BAC per se law. 23 CFR 1225.5 - Code of Federal Regu-lations (annual edition). Vol Title 23: Highways. Part 1225: Operationof Motor Vehicles by Intoxicated Persons; 2000.

Beitel GA, Sharp MC, Glauz WD. Probability of arrest while drivingunder the influence of alcohol. Inj Prev. 2000;6:158–161.

Borkenstein RF, Smith HW. The breathalyzer and its application. MedSci Law. 1961;1:13.

Brady JE, Baker SP, DiMaggio C, McCarthy ML, Rebok GW, Li G.Effectiveness of mandatory alcohol testing programs in reducing al-cohol involvement in fatal motor carrier crashes. Am J Epidemiol.2009;170:775–782.

Casanova T, Hedlund J, Tison J. State Blood Alcohol Concentation(BAC) Testing and Reporting for Drivers Involved in Fatal Crashes:Current Practices, Results and Strategies, 1997–2009. Washington, DC:US Department of Transportation, National Highway Traffic SafetyAdministration; 2012. DOT HS 811 661.

Compton R. The Use of Safety Checkpoints for DWI Enforcement.Washington, DC: US Department of Transportation, NationalHighway Traffic Safety Administration; 1983. DOT HS 806 476.Available at: http://ntl.bts.gov/lib/25000/25700/25713/DOT-HS-806-476.pdf. Accessed April 15, 2012.

Dang JN. Statistical Analysis of Alcohol-Related Driving Trends,1982–2005. Washington, DC: National Highway Traffic Safety Ad-ministration; 2008. DOT HS 810 942. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/810942.pdf. Accessed April 15, 2012.

Delaware v Prouse. 440 U.S. 648: 1979.Dubowski KM. Recent developments in breath-alcohol analysis. In: Is-

raelstam S, Lambert S, eds. Alcohol Drugs and Traffic Safety, Proceed-ings of the Sixth Inernational Conference on Alcohol, Drugs and TrafficSafety. Toronto, Canada: Addiction Research Foundation of Ontario;1974:483–494.

Elder RW, Shults RA, Sleet DA, Nichols JL, Zaza S, Thompson RS.Effectiveness of sobriety checkpoints for reducing alcohol-involvedcrashes. Traffic Inj Prev. 2002;3:266–274.

Fell JC, Compton C, Voas RB. A note on the use of passive alcoholsensors during routine traffic stops. Traffic Inj Prev. 2008;9:534–538.

Fell JC, Lacey JH, Voas RB. Sobriety checkpoints: evidence of effective-ness is strong, but use is limited. Traffic Inj Prev. 2004;5:220–227.

Fell JC, Tippetts AS, Voas RB. Fatal traffic crashes involving drinkingdrivers: What have we learned? Ann Adv Automot Med. 2009;53:63–76.

Ferguson SA, Wells JK, Lund AK. The role of passive alcohol sensorsin detecting alcohol-impaired drivers at sobriety checkpoints. AlcoholDrugs Driving. 1995;11:23–30.

Henstridge J, Homel R, Mackay P. The Long-Term Effects of RandomBreath Testing in Four Australian States: A Time Series Analysis.Canberra, Australia: Federal Office of Road Safety; 1997. No. CR162.

Homel R. Random breath testing and random stopping programs inAustralia. In: Wilson RJ, Mann RE, eds. Drinking and Driving: Ad-vances in Research and Prevention. New York, NY: The Guilford Press,1990:159–202.

Homel R. Random breath testing in Australia: getting it to work accord-ing to specifications. Addiction. 1993;88:S27–S33.

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

Impaired-Driving Enforcement in the United States 669

Homel R. Drink-driving law enforcement and the legal blood alcohollimit in New South Wales. Accid Anal Prev. 1994;26:147–155.

Jones AW. Measuring alcohol in blood and breath for forensicpurposes—a historical review. Forensic Sci Rev. 2000;12:151–182.

Jones AW. Occurrence of alcohol and other drugs in femoral bloodsamples from drivers killed in road-traffic crashes in Sweden. Paperpresented at: T2010 Conference; August 22–26, 2010; Oslo, Norway.

Lacey JH, Kelley-Baker T, Ferguson SA, Rider RP. Low-manpowercheckpoints: can they provide effective DUI enforcement for smallcommunities? Traffic Inj Prev. 2006;7:213–218.

Lacey J, Kelley-Baker T, Furr-Holden CDM, et al. 2007 National Road-side Survey of Alcohol and Drug Use by Drivers: Alcohol Results. Wash-ington, DC: US Department of Transportation, National HighwayTraffic Safety Administration; 2009. DOT HS 811 248. Available at:http://www.nhtsa.gov/Driving+Safety/Research+&+Evaluation/2007+National+Roadside+Survey+of+Alcohol+and+Drug+Use+by+Drivers. Accessed April 18, 2012.

Levy P, Voas RB, Johnson P, Klein T. Evaluation of the ASAPs. J SafRes. 1977;10:162–176.

Lund AF, Jones IS. Detection of impaired drivers with a passive alcoholsensor. In: Noordzij PC, Roszbach R, eds. Alcohol, Drugs and TrafficSafety ‘T86. New York, NY: Excerpta Medica, 1987:379–382.

Mercer SL, Sleet DA, Elder RW, Cole KH, Shults RA, Nichols JL.Translating evidence into policy: lessons learned from the case oflowering the legal blood alcohol limit for drivers. Ann Epidemiol.2010;20:412–420.

Michigan Dept. of State Police v. Sitz, 496 U.S. 444, 1990.Mothers Against Drunk Driving. Alcohol-Related Laws. 2006. Available

at: http://www.madd.org/laws. Accessed April 18, 2012.Moulton BE, Peterson A, Haddix D, Drew L. National Survey of Drink-

ing and Driving Attitudes and Behaviors: 2008: Vol 2. Findings Report.Washington, DC: US Department of Transportation, National High-way Traffic Safety Administration; 2010. DOT HS 811 343.

National Center for Statistics and Analysis. Traffic Safety Facts:2007 Data: State Alcohol-Impaired Driving Estimates. Washington,DC: US Department of Transportation, National Highway Traf-fic Safety Administration; 2008. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811167.pdf. Accessed April 18, 2012.

National Center for Statistics and Analysis. Alcohol-impaired driving.In: Traffic Safety Facts: 2009 Data. Washington, DC: US Depart-ment of Transportation, National Highway Traffic Safety Admin-istration; 2010. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811385.PDF. Accessed April 18, 2012.

National Center for Statistics and Analysis. Early estimate of mo-tor vehicle fatalities in 2012. In: Traffic Safety Facts: Crash Stats.Washington, DC: US Department of Transportation, National High-way Traffic Safety Administration; 2011. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811451.pdf. Accessed April 18, 2012.

National Center for Statistics and Analysis. Traffic Safety Facts:2010 Data: Alcohol-impaired Driving. Washington, DC: US De-partment of Transportation, National Highway Traffic Safety Ad-ministration; 2012a. DOT HS 811 606. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811606.pdf. Accessed April 18, 2012.

National Center for Statistics and Analysis. Traffic Safety Facts, 2010Data: State Alcohol-Impaired Driving Estimates. Washington, DC:US Department of Transportation, National Highway Traffic SafetyAdministration; 2012b. DOT HS 811 612. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811612.pdf. Accessed April 18, 2012.

National Highway Traffic Safety Administration. Fatality analysis report-ing system data files, 1982-2010. Washington, DC: U.S. Departmentof Transportation, National Highway Traffic Safety Administration;2012. Available at: ftp://ftp.nhtsa.dot.gov/fars/. Accessed April 12,2012.

National Highway Traffic Safety Administration. Highway safety pro-gram standards. Federal Register. 1972;37:15602–15621.

National Highway Traffic Safety Administration. The Use of Sobri-ety Checkpoints for Impaired Driving Enforcement. Washington, DC:

US Department of Transportation, National Highway Traffic SafetyAdministration, Office of Enforcement and Emergency Services;1990. DOT HS 807 656. Available at: http://stopdruggeddriving.org/pdfs/SobrietyCheckpoints.pdf. Accessed April 18, 2012.

National Highway Traffic Safety Administration. Sobriety Checkpoints:Point/Counterpoint. Washington, DC: US Department of Transporta-tion, National Highway Traffic Safety Administration; 1993. DOT HS807 916.

National Highway Traffic Safety Administration. Law EnforcementTraining Video on Sobriety Checkpoints. Washington, DC: US De-partment of Transportation, National Highway Traffic Safety Admin-istration; 1999. DOT HS 808 990.

National Highway Traffic Safety Administration. Saturation Patrols andSobriety Checkpoints: A How-to Guide for Planning and Publicizing Im-paired Driving Enforcement Efforts. Washington, DC: US Departmentof Transportation, National Highway Traffic Safety Administration;2000. DOT HS 809 063. Available at: http://www.nhtsa.gov/people/injury/alcohol/saturation patrols/index.html. Accessed April 22,2012.

National Research Council. Achieving Traffic Safety Goals in theUnited States: Lessons for Other Nations. Washington, DC: NationalAcademy of Sciences; 2010. TRB Special Report 300.

National Safety Council. Alcohol and the Impaired Driver. Chicago, IL:National Safety Council; 1976.

National Transportation Safety Board. Safety Recommendation. Wash-ington, DC: National Transportation Safety Board; 2012. H-12-32and -33.

Peek-Asa C. The effect of random alcohol screening in reducing motorvehicle crash injuries. Am J Prev Med. 1999;16:57–67.

Robertson RD, Holmes E, Simpson H. A criminal justice perspective.In: Impaired Driving Priorities. Ottawa, ON, Canada: Traffic InjuryResearch Foundation; 2009.

Rosen J. Why the TSA pat-downs and body scans are uncon-stitutional [Opinions]. The Washington Post, 2010, Novem-ber 28. Available at http://www.washingtonpost.com/wp-dyn/content/article/2010/11/24/AR2010112404510.html AccessedApril 12, 2010.

Ross HL. Law, science and accidents: the British Road Safety Act of1967. J Legal Stud. 1973;2:1–78.

Ross HL. The Scandinavian myth: the effectiveness of drinking-and-driving legislation in Sweden and Norway. J Legal Stud.1975;4:285–310.

Ross HL. Deterrence regained: the Cheshire constabulary’s breathalyzerblitz. J Legal Stud. 1977;6:241–249.

Ross HL. Deterring the Drinking Driver: Legal Policy and Social Control.2nd ed. Lexington, MA: Lexington Books; 1984.

Sherman LW. Police Crackdowns: Initial and Residual Deterrence. Crimeand Justice: An Annual Review of Research. Chicago, IL: University ofChicago Press; 1990.

Shults RA, Elder RW, Sleet DA, et al. Reviews of evidence regardinginterventions to reduce alcohol-impaired driving. Am J Prev Med.2001;21:66–88.

Simpson HM, Robertson RD. Enforcement. In: DWI System Improve-ments for Dealing With Hard Core Drinking Drivers. Ottawa, ON,Canada: Traffic Injury Research Foundation; 2001:17.

Solomon R, Chamberlain E, Abdoullea M, Tinholt B. Random breathtesting: a Canadian perspective. Traffic Inj Prev. 2011;12:111–119.

Stuster JW. The Detection of DWI at BACs Below 0.10. Washington,DC: US Department of Transportation, National Highway TrafficSafety Administration; 1997. Final Report. Available at: http://www.popcenter.org/problems/drunk driving/PDFs/NHTSA 1997.pdf. Accessed April 12, 2012.

Stuster JW, Blowers MA. Experimental Evaluation of Sobriety Check-point Programs. Washington, DC: US Department of Transporta-tion, National Highway Safety Administration; 1995. DOT HS 808287. Available at: http://ntl.bts.gov/lib/25000/25900/25932/DOT-HS-808-287.pdf. Accessed April 10, 2012.

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013

670 Voas and Fell

Subramanian R. Motor vehicle traffic crashes as a leading cause of deathin the United States, 2008 and 2009. In: Traffic Safety Facts: Re-search Notes. Washington, DC: US Department of Transportation,National Highway of Traffic Safety; 2012. Available at: http://www-nrd.nhtsa.dot.gov/Pubs/811620.pdf. Accessed April 14, 2012.

Voas RB. Emerging technologies for controlling the drunk driver.In: Lawrence MD, Nortum JR, Zimring FE, eds. Social Controlof the Drunk Driver. Chicago, IL: University of Chicago Press;1988:321–370.

Voas RB. A new look at NHTSA’s evaluation of the 1984 CharlottesvilleSobriety Checkpoint Program: implications for current checkpointIssues. Traffic Inj Prev. 2008;9:22–30.

Voas RB, Lacey JH. Drunk driving enforcement, adjudication, and sanc-tions in the United States. In: Wilson RJ, Mann RE, eds. Drinking andDriving: Advances in Research and Prevention. New York, NY: TheGuilford Press; 1990:116–158.

Voas RB, Lacey JH, Fell JC. The PASpoint system—passive sensors atmini-checkpoints: bringing Australia’s random breath-test system tothe United States. Transp Res Circ. 2005;E-C072:45–53.

Voas RB, Romano E, Peck R. Validity of the passive sensor for identifyinghigh BAC drivers at the crash scene. J Stud Alcohol. 2006;67:714–721.

Wagenaar AC, Maldonado-Molina MM. Effects of drivers’ licensesuspension policies on alcohol-related crash involvement: long-term

follow-up in forty-six states. Alcohol Clin Exp Res. 2007;31:1399–1406.

Wagenaar A, Maldonado-Molina M, Ma L, Tobler A, Komro K. Effectsof legal BAC limits on fatal crash involvement: Analyses of 28 statesfrom 1976 through 2002. J Saf Res. 2007;38:493–499.

Watson B. Trends in impaired driving in Australia in 2010. Paper pre-sented at: T2010 Conference; August 22–26, 2010; Oslo, Norway.

Wells JK, Greene MA, Foss RD, Ferguson SA, Williams AF. DriversWith High BACs Missed at Sobriety Checkpoints. Arlington, VA: In-surance Institute for Highway Safety; 1996.

Widmark EMP. Principles and Applications of Medicolegal Alcohol De-termination. Davis, CA: Biomedical Publications; 1932.

Zador P, Krawchuk S, Moore B. Drinking and Driving Trips, Stops byPolice, and Arrests: Analyses of the 1995 National Survey of Drinkingand Driving Attitudes and Behavior. Washington, DC: US Departmentof Transportation, National Highway Traffic Safety Administration;2000. DOT HS 809 184.

Zaloshnja E, Miller TR. Cost of crashes related to road conditions,United States, 2006. Ann Adv Automot Med. 2009;53:141–153.

Zwicker TJ, Hedlund J, Northrup VS. Breath Test Refusals in DWIEnforcement: An Interim Report. Washington, DC: US Departmentof Transportation, National Highway Traffic Safety Administration;2005. DOT HS 809 876.

Dow

nloa

ded

by [

Prev

entio

n R

esea

rch

Cen

ter]

, [M

r Ja

mes

C. F

ell]

at 1

3:58

20

Aug

ust 2

013