Reducing Alcohol-impaired Driving: Can We Regain the Momentum? International Conference on Urban Traffic Safety Edmonton, Alberta April 29, 2015 Anne T. McCartt, Ph.D. iihs.org
IIHS is an independent, nonprofit scientific and educational organization dedicated to reducing the losses deaths, injuries and property damage from crashes on the nation s roads. HLDI shares this mission by analyzing insurance data representing human and economic losses from crashes and other events related to vehicle ownership. Both organizations are wholly supported by auto insurers.
Motor vehicle crash deaths and deaths per billion vehicle miles traveled 1950-2013 60,000 80 55,000 50,000 Motor vehicle crash deaths 70 60 50 45,000 40 40,000 30 35,000 Crash deaths per billion vehicle miles traveled 20 10 2013 11.0 per billion 32,719 deaths 30,000 1950 1960 1970 1980 1990 2000 2010 0
What has contributed to declining crash death rates?
1959 Bel Air vs. 2009 Malibu crash test 40 mph moderate frontal offset crash test
IIHS announces new smalloverlap crash test program August 14, 2012
Small overlap front crash is latest structural test
Small overlap ratings by model year January 2015 120 100 poor marginal acceptable 80 good 60 40 20 0 2012 2013 2014 2015
August 18, 2011
Safety belts saved an estimated 12,174 lives in 2012; belt use has increased largely because of primary belt use laws and publicized enforcement 100 Percent of front seat occupants using belts, 1983-2014 80 60 40 20 0 1983 85 87 89 91 93 95 97 99 01 03 05 07 09 11 13
Child passenger vehicle occupant deaths per million children By age,1975-2013 60 <1 year 1-3 years 4-8 years 9-12 years 40 20 0 1975 1980 1985 1990 1995 2000 2005 2010
Passenger vehicle fatal crash involvements per 100,000 people By driver age, 1996-2013 50 40 30 16 17 18 19 30-59 20 10 0 1996 2000 2005 2010
Almost one-quarter of crash deaths occur at intersections. Conversion of stop sign and traffic signal intersections to roundabouts: 40 percent reduction in all crashes 80 percent reduction in injury crashes 90 percent reduction in fatal & incapacitating injury crashes
Progress in building roundabouts April 2015 WA OR CA NV ID MT UT AZ WY CO NM ND SD NE KS OK TX MN IA MO AR LA WI IL MS MI IN OH KY TN AL GA NH VT ME NY MA PA NJ RI WV CT DE VA MD NC DC SC 50 or more 10-49 AK FL fewer than 10 HI
Progress has been more difficult in other areas
Speeding as factor in vehicle crash deaths in 2004-13, by percent 40 30 20 10 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Maximum authorized speed limits June 2010 WA OR CA NV ID MT UT AZ WY CO NM ND SD NE KS OK TX MN IA MO AR LA WI IL MS MI IN OH KY TN AL GA PA WV VA NC SC NY NJ VT NH DC CT DE MD ME MA RI 55 mph (D.C. only) 60 mph AK FL 65 mph 70 mph 75 mph HI
Maximum authorized speed limits April 2015 WA OR CA NV ID UT AZ MT WY CO NM ND SD NE KS OK TX MN IA MO AR LA WI IL MS MI IN OH KY TN AL GA PA WV VA NC SC ME VT NY NH MA RI CT NJ DE MD DC AK FL 70 mph 75 mph HI 80 mph 85 mph
U.S. communities with speed cameras 1995-2014 140 120 100 80 60 40 20 0 1995 2003 2007 2008 2009 2010 2011 2012 2013 2014
Alcohol-impaired driving in the U.S.
Relative risk of fatal crash involvement at various BACs compared to zero BAC Passenger vehicle drivers by age (Voas et al., 2012) 80 60 ages 16-20 ages 21-34 ages 35+ 40 20 0 0.00 0.05 0.10 0.15 0.20 0.25 BAC (g/dl)
Police-reported crash risk estimates by BAC Adjusted by age and gender 40 30 Virginia Beach study, 2015 Ft. Lauderdale and Long Beach, 2005 20 10 0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 BAC (g/dl)
May 22, 1972
December 6, 1982
November 27, 1993
February 8, 2003
April 2, 2005
March 6, 2012
Percent of crash deaths involving drivers with BACs of 0.08 g/dl or higher By calendar year, 1982-2013 60 50 40 30 20 10 0 at least one driver with BAC 0.08g/dl at least one driver, pedestrian, or bicyclist with BAC 0.08 g/dl 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
Percent of weekend nighttime drivers with positive alcohol test in national roadside surveys 40 positive BAC 30 BAC 0.08 g/dl 20 10 0 1973 1986 1996 2007 2013-14
December 30, 2013
Percent of weekend nighttime drivers with positive drug test in national roadside surveys 20 15 any illegal drug marijuana (THC) medications 10 5 0 2007 2013-14
Percent change in crash risk associated with alcohol and drugs, adjusted for demographic variables 700 600 500 400 300 200 100 0-100 no alcohol, positive drug 0 < BAC < 0.05%, no drug 0 < BAC < 0.05%, positive drug BAC 0.05%, no drug BAC 0.05%, positive drug
Profile of alcohol-impaired drivers
Percentage of drivers and pedestrians in fatal crashes with BACs 0.08 g/dl, 1982-2013 60 passenger vehicle drivers large truck drivers motorcycle drivers 40 pedestrians (age 16 and older) 20 0 1982 1986 1990 1994 1998 2002 2006 2010
Percentage of drivers in fatal crashes with BACs 0.08 g/dl, by age group, 1982-2013 50 40 30 20 ages16-20 10 ages 21-24 ages 25-34 ages 35+ 0 1982 1986 1990 1994 1998 2002 2006 2010
Percentage of drivers in fatal crashes with BACs 0.08 g/dl, 1996 and 2013 Gender, age, restraint use 1996 2013 males 25 23 females 13 15 age 16-20 17 17 age 21-30 31 31 age 31-40 28 26 age 41-60 18 20 age 61+ 7 8 restrained drivers 11 12 unrestrained drivers 37 41
Percentage of drivers in fatal crashes with BACs 0.08 g/dl, 1996 and 2013 Urban vs. rural, nighttime vs. daytime 1996 2013 urban roadway 20 20 rural roadway 23 22 nighttime (9pm-6am) 44 42 other hours 12 12
Percent of drivers with BACs 0.08 g/dl in national roadside surveys, by age Voas et al., 1998; Lacey et al., 2009 12 10 8 <21 21 to 34 35 to 44 45 6 4 2 0 1973 1986 1996 2007
Percent of drivers with BACs 0.08 g/dl in national roadside surveys, by gender Voas et al., 1998; Lacey et al., 2009 10 8 male female 6 4 2 0 1973 1986 1996 2007
Should we focus primarily on the hard-core DUI offender?
Deaths in 2013 involving passenger vehicle drivers with various BACs 4,000 3,000 2,000 1,000 0 0.08-0.14 g/dl 0.15-0.19 g/dl 0.20-0.24 g/dl 0.25 g/dl or more
Percentage of passenger vehicle drivers in fatal crashes with DUI convictions within 3 years 2013 no prior offenses 95.7 1 prior offense 2.1 2 or more prior offenses 0.4 unknown 1.8
Percentage of passenger vehicle drivers in fatal crashes with various BACs, 1982-2013 15 0.08-0.14 g/dl 0.15-0.19 g/dl 10 0.20-0.24 g/dl 0.25 g/dl or more 5 0 1982 1986 1990 1994 1998 2002 2006 2010
Counts of DUI-related convictions in Washington By quarter of arrest, January 1999-December 2009 8,000 6,000 4,000 1st offense repeat offenses 2,000 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
What works to reduce alcoholimpaired driving?
Laws requiring administrative license suspension (ALS) for first DUI offense April 2015 WA OR CA NV ID UT AZ MT WY CO NM ND SD NE KS OK MN IA MO AR WI IL MI IN KY TN OH PA WV VA NC SC VT NY NH NJ DE MD DC ME CT MA RI TX LA MS AL GA AK FL HI
Administrative license suspension (ALS) laws ALS laws remove impaired drivers from the road quickly and ensure that penalties are applied Effectiveness of ALS laws Reduce alcohol-impaired driving among the general population Reduce recidivism among DUI offenders Reduce the number of fatal crashes likely to involve alcohol and the number of drivers with positive BACs involved in fatal crashes (Zador et al., 1989; Klein, 1989; Wagenaar and Maldonado-Molina, 2007)
Strong laws strongly enforced are a proven countermeasure for reducing impaired driving Median decrease in alcohol-involved fatal crashes of 9 percent associated with well-publicized sobriety checkpoint programs (Bergen, 2014) Other types of targeted DUI enforcement such as saturation patrols also can be effective, especially when publicized
Traditional large-scale sobriety checkpoint
Low manpower checkpoint program in West Virginia reduced percent of drivers with BACs 0.05 g/dl (Lacey et al., 2006)
Use of passive alcohol sensors at sobriety checkpoints improves detection of alcoholimpaired drivers
Percent of high BAC drivers detected by officers, with and without passive alcohol sensors IIHS, 1986-1995 without sensors with sensors sobriety checkpoints 45-55 68-71 routine patrol 69 77 special DUI patrol 88 94
Sobriety checkpoints work, but they aren t used often June 30, 2001
Annual number of sobriety checkpoints conducted in states per 100,000 residents Survey of state highway safety offices, 2012 WA OR CA NV ID MT WY UT CO AZ NM ND SD NE KS OK TX MN IA MO AR LA WI IL MS MI PA IN OH WV KY VA TN NC SC AL GA NY ME VT NH MA RI NJ CT DE MD DC > 20 HI AK FL 3 to 20 < 3 not conducted unknown
Frequency of sobriety checkpoints conducted by local and state agencies National survey of enforcement agencies, 2012-14 100% 80% 60% at least monthly quarterly less than quarterly none other or unknown 40% 20% 0% county municipal state national estimate
How sobriety checkpoints are conducted Few agencies conduct low manpower checkpoints 17 percent typically use fewer than 7 officers 48 percent use 7-15 officers 27 percent use more than 15 officers 63 percent typically pool resources with other law enforcement agencies Almost all agencies (87 percent) publicize their checkpoints
Frequency of dedicated DUI patrols by state and local law enforcement agencies 100% 80% 60% at least monthly quarterly less than quarterly none other or unknown 40% 20% 0% county municipal state national estimate
Use of passive alcohol sensors 20 percent of agencies used passive alcohol sensors Among agencies using passive alcohol sensors, most used them infrequently (54 percent)
Percent of agencies with officers trained in alcohol and/or drug impairment detection 100 80 60 all or most officers half of officers few officers none 40 20 0 Standardized Field Sobriety Test Advanced Roadside Impaired Driving Enforcement Drug Recognition Expert
Biggest challenge agencies face in conducting DUI enforcement percent staffing 56 lack of funding 24 excessive paperwork 6
College community campaign to reduce drinking and driving with strong enforcement emphasis McCartt et al., 2009
Percent reduction in odds of given BAC in program community, by age group 0% -20% -40% -60% -80% 16-20 years old 21-24 years old 25 years old significant at 0.05-100% BAC 0.02 g/dl BAC 0.05 g/dl BAC 0.08 g/dl
Alcohol ignition interlocks
Alcohol ignition interlocks Driver blows into a tube to test breath for alcohol Vehicle won t start if certain amount of alcohol detected Periodic retests while vehicle is in motion BAC 0.025 g/dl is typical threshold for DUI offenders
Laws mandating alcohol ignition interlock orders April 2015 WA OR CA NV ID MT UT AZ WY CO NM ND SD NE KS OK TX MN IA MO AR LA WI MS IL MI IN OH KY TN AL GA PA WV VA NC SC NH VT ME NY MA NJ RI CT DE MD DC judicial discretion only repeat offenders AK FL high-bac offenders high-bac first & repeat offenders all offenders HI
Installed interlocks in United States 1986-2014 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 1986 1990 1994 1997 2002 2006 2010
The Los Angeles Times, January 31, 2012 Drunken drivers: Congress gets behind breath-test ignition devices FOXNEWS.com, February 3, 2012 Restaurant lobby seeks to put ignition interlock provision on ice TampaBayOnline.com, February 4, 2012 The uncompromising campaign to end social drinking one law at a time The New York Times, March 11, 2013 Alcohol-Detection Measures Receive Support in Transportation Bill
Prior studies of effects of interlocks on recidivism Most studies compared recidivism rates of offenders who installed interlocks vs. those who did not Reductions in recidivism of 60-80 percent while interlocks installed Reductions found for both repeat and first offenders Effects dissipate after interlocks removed Difficult to adequately account for potential differences between offenders who elected to get interlocks and those who did not No prior study examined the effects of an all-offender law requiring interlock orders on recidivism among all offenders affected by the law Little prior rigorous research on the effects of interlock programs on alcohol-related crashes
March 6, 2012
Timeline of Washington state interlock laws Courts permitted to order interlocks Issuance of interlock orders moves from courts to Department of Licensing Interlock available immediately after arrest in lieu of administrative license suspension When reasonably available in area, interlocks must photograph person giving breath sample July 23, 2003 Jan. 1, 2009 Jan. 1, 2013 1987 Jan. 1, 1999 June 10, 2004 Jan. 1, 2011 Courts must order interlocks for repeat offenders and first offenders with BACs 0.15% or alcohol test refusal Interlocks Interlock required orders for required first offenders for first offenders with BACs with < 0.15% BACs < (simple 0.15% (simple DUI) DUI) Minimum 4-month interlock installment required for license reinstatement
Timeline of Washington state interlock laws Courts permitted to order interlocks Issuance of interlock orders moves from courts to Department of Licensing Interlock available immediately after arrest in lieu of administrative license suspension When reasonably available in area, interlocks must photograph person giving breath sample. July 23, 2003 Jan. 1, 2009 Jan. 1, 2013 1987 Jan. 1, 1999 June 10, 2004 Jan. 1, 2011 Courts must order interlocks for repeat offenders and first offenders with BACs 0.15% or alcohol test refusal Interlocks required for first offenders with BACs < 0.15% (simple DUI) Minimum 4-month interlock installment required for license reinstatement
Distribution of 1 st DUI-related convictions by type By quarter of arrest, January 1999-December 2009 60% 40% simple DUI high BAC test refusal negligent driving deferred interlock available after arrest 20% court to Dept. of Licensing all offender law 0% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Percent of 1 st DUI offenders who installed interlocks By quarter of arrest and conviction type, January 1999-December 2009 court to Dept. of Licensing all offender law 60% 40% 20% simple DUI high BAC test refusal negligent driving deferral interlock available after arrest 0% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Effects of interlock law changes on recidivism
Cumulative percent of recidivism among 1 st simple DUI offenders arrested after 2004 law change By status of interlock, arrested between June 2004 and June 2006 12% installed interlock ordered to install interlock but did not 8% 4% 0% 1 3 5 7 9 11 13 15 17 19 21 23 months after arrest
Predicted cumulative 2-year recidivism rate for 1st simple DUI convictions with & without 2004 law change By quarter of arrest, January 1999-June 2006 12% 10% -12% 8% 6% 4% 2% without 2004 law change with 2004 law change all offender law 0% 1999 2000 2001 2002 2003 2004 2005 2006
Cumulative 2-year recidivism rate and projected rate with 100 percent interlock use Arrests April-June 2006 10 8 recidivism with actual use recidivism with 100% use 6 4 2 0 1st simple DUI offenders (34% actual interlock use rate) all 1st DUI offenders (24% actual interlock use rate)
Effects of interlock law changes on crashes
Percent decreases in single-vehicle, late-night crashes in Washington State Relative to trends in Oregon and California percent p value 2003 law change: issuance of interlock orders moved to Department of Licensing -6.4 0.0792 2004 law change: interlock order requirement extended to all DUI convictions -8.3 0.0183
Conclusions Installation rates increased somewhat after administration was moved to Department of Licensing in 2003 and after the interlock requirement was extended to all DUI offenders in 2004 Preliminary data suggest that allowing immediate interlock installation after arrest, in lieu of an administrative license suspension, led to more installations and earlier installations Extending interlock requirement to all first-time DUI convictions in Washington reduced recidivism by 12 percent Additional gains achievable with higher installation rates The all-offender law was associated with a significant reduction in risk of single-vehicle late-night crashes, suggesting a general deterrent effect
Attitudes toward requiring alcohol ignition interlocks for convicted DWI offenders National telephone survey, 2009 2% 7% 8% very good idea good idea 45% not a good idea bad idea 39% don't know
Implications for states? Expand use of alcohol ignition interlocks by DUI offenders Extend laws to all DUI convictions Seek ways to increase interlock use rates Reconsider loopholes in laws that allow reductions in DUI charges to traffic offenses without interlock requirements or other DUI penalties Publicize interlock laws to deter all drivers from driving impaired
There is little research on many specific implementation issues Is an interlock program more effectively administered by courts or licensing agencies? What are the effects on safety of allowing an interlock immediately after arrest and in lieu of an administrative license suspension? Should offenders be penalized for failing the interlock start test? Is so, how? Will this discourage interlock installations? Which law is better one that allows offenders to serve the suspension rather than getting an interlock and then to reinstate their license, or one that does not permit offenders to reinstate until they show proof they installed the interlock?
Advanced in-vehicle alcohol detection technology
September 17, 2009
Potential lives saved in 2013 if all drivers with specific BACs were kept off the roads BAC 0.08 g/dl positive BAC drivers with multiple DUI convictions within 3 years drivers with at least one prior DUI conviction within 3 years 67 97 428 614 all drivers 6,904 10,494
Broader installation of alcohol detection technology Current ignition interlocks are not suitable for use in all vehicles Inconvenient, expensive, require calibration People who obey alcohol-impaired driving laws People who don t drink alcohol Need technology that is virtually invisible to driver without sacrificing precision Accurately detects BAC 0.08 g/dl (without false positives or false negatives) Must differentiate between driver and passenger
Driver Alcohol Detection System for Safety Partnership between federal government and automakers Research, develop, test advanced alcohol detection technology suitable for all vehicles Build public support for vehicle-based approach Two technology firms developing systems that reliably determine driver s BAC in 1/3 second and are reliable and durable enough to install in vehicles Phase three 5 years beginning fall 2013 Refine devices, test human subjects, show commercial feasibility Gauge public response to specific technologies
Two technologies being developed Breath-based system Using infrared light beams, measures concentrations of alcohol and carbon dioxide from driver s exhaled breath in vehicle cabin Amount of carbon dioxide in breath indicates degree of dilution of alcohol concentration in expired air Takes instantaneous, unobtrusive readings as driver breaths normally and accurately and reliably identify breath of driver vs. passengers Touch-based system Uses near-infrared tissue spectrometry and measures light absorption to detect level of alcohol in blood Goal is to integrate system into current vehicle controls, e.g., start button, and take multiple, accurate readings in less than a second
Attitudes toward advanced alcohol test technology in all vehicles, if technology shown to be reliable IIHS national telephone survey, 2009 19% 11% 6% very good idea good idea 36% not a good idea bad idea don't know 28%
Percent of respondents who think advanced alcohol test technology in all vehicles is good or very good idea 80 60 40 20 0 all drive regularly drink drive after drinking don't drink
Would survey respondents want advanced alcohol test technology in next vehicle, if available as option and cost is reasonable? Respondents who drive regularly 4% 42% yes no 54% don't know/not sure
Percent of drivers who would want advanced alcohol test technology in next vehicle 60 50 40 30 20 10 0 all drivers who drink drive after drinking may have driven over legal limit
Crash avoidance is the new frontier
The Google vision Autonomous vehicles delivering occupants to pre-set destinations
Vehicles without and with ESC
Effects on crash risk Percent change in crash rates for vehicles with standard ESC vs. optional or no ESC, updated May 2010 20 all crashes fatal injury 0-20 -40-60 -80 all crashes multiple-vehicle single-vehicle
Driver assistance features Radar, LIDAR, ultrasonic, infrared, cameras, GPS
Front crash prevention technologies can mitigate or prevent crashes because drivers are distracted, impaired, or make errors
Annual crashes relevant to 4 crash avoidance systems Based on 2004-08 crash totals all injury fatal forward collision warning 1,165,000 66,000 879 lane departure warning 179,000 37,000 7,529 side view assist 395,000 20,000 393 adaptive headlights 142,000 29,000 2,484 total unique crashes 1,866,000 149,000 10,238 percent of crashes 32% 21% 31%
Changes in physical damage insurance claim rates for front crash prevention systems 20% low speed warning only warning with autobrake 10% 0% -10% -20% PDL -30% Volvo City Safety Honda Accord camera (with LDW) Honda Accord radar (with LDW + ACC) collision Mercedes Volvo Acura Mercedes Subaru (with LDW) Volvo (with LDW)
Rating front crash prevention systems Automatic braking tests at 12 mph and 25 mph
Estimated 80 million viewers
BMW comparison video 12 mph tests
Front crash prevention ratings 2014 and 2015 models 120 117 2014 2015 80 70 40 38 46 24 33 19 27 0 not qualified/ not available basic advanced superior
Vehicle Research Center expansion Estimated $30 million cost
Covered test track rendering
Vehicle to pedestrian testing
Vehicle to moving target testing
Vehicle to slower moving target testing
Robotic target platform Crossing path collision
Robotic test driver Full steering, accelerator and braking control
Forward collision prevention systems are working to reduce insurance collision claims Adaptive headlights are working The benefits of some systems are less clear Lane departure warning Blind spot warning Rearview cameras Parking proximity sensors
percent decrease Percent change in rates of collision and property damage liability (PDL) claims associated with adaptive headlights 12 8 4 0-4 -8-12 -16-20 PDL PDL collision collision statistically significant Acura Mazda Mercedes Volvo
It takes a long time for vehicle safety features to penetrate the fleet
New vehicle series with electronic stability control U.S., by model year 100% 80% 60% 40% 20% 0% 1995 1997 1999 2001 2003 2005 2007 2009 standard optional not available
Registered vehicles with electronic stability control U.S., by calendar year 100% 80% 60% 40% 20% 0% 1995 1997 1999 2001 2003 2005 2007 2009 standard optional not available
Registered vehicles with electronic stability control, actual and predicted U.S., by calendar year 100% 80% 60% 40% actual predicted 20% 0% 1995 2000 2010 2020 2030 2040
Registered vehicles with forward collision warning, actual and predicted By calendar year 100% 80% 60% 40% 20% 0% actual predicted predicted if all 2013 models equipped 2000 2010 2020 2030 2040
Summary We are entering an exciting period in advanced automotive technology Reaping the rewards of this technology will be a gradual, slow process; unclear how it will affect alcohol-impaired driving crashes A crash-free environment is not on the immediate horizon; putting people in packages that protect them in crashes remains a top priority In the near term over the next ten years and likely beyond our success in reducing crash deaths and injuries will also depend on other proven countermeasures in highway safety This includes strong DUI laws strongly enforced and meaningful penalties, including alcohol ignition interlocks
Summary Automated driving will help prevent and mitigate crashes Actual effectiveness of partial automation has been documented Ideal automated systems cannot be distracted as drivers can be Automated driving systems are not created equal Not all are living up to their promise yet Even those with similar functional descriptions may act differently Consumer information testing can help identify important differences Widespread automated driving will take time Current partial automated systems are evolving quickly Older vehicles are replaced by state-of-the-art vehicles slowly
USA Today, July 15, 2005
The New Yorker, April 23, 2007
More information and links to our YouTube channel and Twitter feed at iihs.org Anne McCartt, Ph.D. Senior Vice President, Research amccartt@iihs.org iihs.org
Research objectives Examine the effects of 2003 law change moving administration to the Department of Licensing and 2004 all-offender law requiring interlock orders on conviction types, interlock installation rates, and the recidivism rate among all offenders affected by the law Examine the preliminary effects of the 2009 law change making interlocks available immediately after arrest on conviction types and interlock installation rate Examine the general deterrent effect of 2003 and 2004 law changes on single-vehicle late-night crashes
Effects of extending interlock requirement to all first-time DUI convictions in Washington state Recidivism among affected offenders reduced by 12 percent, even with low interlock installation rate (about one-third) Additional gains achievable with higher installation rates predicted recidivism about half at 100 percent vs. 34 percent installation rate Law change associated with significant 8.3 percent reduction in risk of single-vehicle nighttime crashes, suggesting a general deterrent, population effect Implications for states? Extend interlock requirements to all DUI convictions Publicize interlock laws to deter all drivers from driving impaired