SURGEON GENERAL'S WORKSHOP ON DRUNK DRIVING Background Papers Washington, D.C. o December 14-16, 1988 U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Offke of the Surgeon General 5600 Fishers Ime Rockville, Maryland 20857 ACKNOWLEDGMENTS The Surgeon General's Workshop on Drunk Driving was supported by the follow- ing organizations: U.S. Department of Defense U.S. Department of Education U.S. Department of Health and Human Services U.S. Public Health Service Alcohol, Drug Abuse, and Mental Health Administration National Institute on Alcohol Abuse and Alcoholism Office for Substance Abuse Prevention Centers for Disease Control Health Resources and Services Administration Bureau of Health Care Delivery and Assistance Bureau of Maternal and Child Health and Resources Development Indian Health Service U.S. Department of Justice Office of Juvenile Justice and Delinquency Prevention U.S. Department of Transportation National Highway Traffic Safety Administration Editorial support and camera-readycopy were provided by Janus Associates under contract 85080401. AN material in this volume is in the public domain and may be used or reproduced without permission from the Office ojthe Surgeon General or the aurhors. Citation of the source is appreciated. The opinions apressed herein are the views of tire authors and participants and do JlOf necessarily reflect the official position of tlze Ofice of the Surgeon General or any other sponsoriltg agency of the U.S. Goventr7latt. Printed 1989 FOREWORD The combination of drinking and driving claims one life every 20 minutes and injures hundreds every day. Tragically, the major casualties are the youth of our Nation. Daily- as a society and as individuals-we make decisions about drinking, about serving alcoholic beverages, about driving, and about "having fun." These decisions affect our own lives as well as those of our families, friends, and neighbors. Choices about drinking and driving can protect our lives or destroy them. Thii serious health and safety problem is preventable. Many individuals and concerned groups have campaigned to end this devastation, but it cannot be reduced without national leadership. The Surgeon General's Workshop on Drunk Driving was held in December 1988 to devise a combination of strategies for addressing the problem at the local, State, and National levels. Experts in diverse fields related to drinking and driving were invited to the workshop to share their expertise and experiences. Their recommendations, made after 3 days of intense discussion, can be found in the Proceedings. The background papers in this volume were commissioned to provide a foundation for and launch the discussion of the 11 expert panels of the workshop. The authors presented the state of the art in the different fields and describe the various attempts throughout the country and the world to prevent alcohol-impaired driving- the trials, the errors, the successes, and the obstacles. With up-to-date knowledge about the problem and proposed solutions, and a coordinated comprehensive plan, we are ready to enlist our country's help in preventing this major threat to our health and well-being. C. Everett Koop, M.D., Sc.D. Surgeon General . . . III Workshop' Planning Com,mittee Jerald H. Anderson Chairperson Intoxicated Driving Prevention Task Force Office of the Assistant Secretary of Defense Force Management and Personnel Safety and Occupational Health Policy U.S. Department of Defense Loran Archer Deputy Director National Institute on Alcohol Abuse and Alcoholism Alcohol, Drug Abuse, and Mental Health Administration U.S. Department of Health and Human Services Amy Barkin Coordinator for the Drunk Driving Workshop Office of the Surgeon General U.S. Department of Health and Human Services Maura Daly Senior Policy Analyst Office of the Secretary U.S. Department of Education Roberta Dorn Director Concentration of Federal Effort Program Office of Juvenile Justice and Delinquency Prevention U.S. Department of Justice Robert W. Denniston Director Division of Communication Programs Office for Substance Abuse Prevention Alcohol, Drug Abuse, and Mental Health Administration U.S. Department of Health and Human Services Arthur S. Funke, Ph.D. Chief Psychologist Child and Adolescent Branch Bureau of Maternal and Child Health and Resources Development Health Resources and Services Administration U.S. Department of Health and Human Services V Mary L. Ganikos, Ph.D. Program Director Community Prevention and Late Life Alcohol Abuse National Institute on Alcohol Abuse and Alcoholism Alcohol, Drug Abuse, and Mental Health Administration U.S. Department of Health and Human Services Jan Howard, Ph.D. Chief Prevention Research Branch National Institute on Alcohol Abuse and Alcoholism Alcohol, Drug Abuse, and Mental Health Administration U.S. Department of Health and Human Services Michael T. Impellizzeri Chief Special Programs Division Office of Alcohol and State Programs National Highway Traffic Safety Administration U.S. Department of Transportation Steve Moore Associate Chief of Staff Office of the Surgeon General U.S. Department of Health and Human Services James L. Nichols, Ph.D. Deputy Director, Scientific and Technical Affairs National Highway Traffic Safety Administration U.S. Department of Transportation Joan White Quinlan Prevention Program Coordinator Special Programs Division Office of Alcohol and State Programs National Highway Traffic Safety Administration U.S. Department of Transportation Richard J. Smith III Manager Injury Prevention Program Indian Health Service U.S Department of Health and Human Services vi Federal Advisory Group on Followup Activities for the Workshop U.S. Department of Defense Jerald H. Anderson Chairperson, DOD Intoxicated Driving Prevention Task Force Office of the Assistant Secretary of Defense Force Management and Personnel Safety and Occupational Health Policy Room 3A272 The Pentagon, Washington, DC 203014000 U.S. Department of Education Maura Daly Senior Policy Analyst Office of the Secretary, 400 Maryland Avenue, S.W. Room 4145, Washington, DC 20202 U.S. Department of Health and Human Services U.S. Public Health Service Robert W. Denniston Director, Division of Communication Programs Oflice for Substance'Abuse Prevention Alcohol, Drug Abuse and Mental Health Administration Room 13A-54 Parklawn Building, 5600 Fishers Lane Rockville, MD 20857 Arthur S. Funke, Ph.D. Chief Psychologist, Child and Adolescent Branch Bureau of Maternal and Child Health and Resources Development Health Resources and Services Administration Room 9-21 Parklawn Building, 5600 Fishers Lane Rockville, MD 20857 Mary L. Ganikos, Ph.D. Program Director Community Prevention and Late Life Alcohol Abuse National Institute on Alcohol Abuse and Alcoholism Alcohol, Drug Abuse and Mental Health Administration Room 16C-03 Parklawn Building, 5600 Fishers Lane Rockville, MD 20857 vii U.S. Department of Health and Human Services US. Public Health Service (continued) Susan J. Lockhart Program Specialist Office of the Surgeon General Office of the Assistant Secretary for Health Room 18-67 Parklawn Building, 5600 Fishers Lane Rockville, MD 20857 Richard J. Smith III Manager, Injury Prevention Program Indian Health Service 5600 Fishers Lane, Rockville, MD 20857 Bob Vollinger Presidential Management Intern Office of the Surgeon General Office of the Assistant Secretary for Health Room 18-67 Parklawn Building, 5600 Fishers Lane Rockville, MD 20857 U.S. Department of Justice John Dawson Program Manager Office of Juvenile Justice and Delinquency Prevention Room 758,633 Indiana Avenue, N.W. Washington, DC 28531 Donni Hassler, Ed.D. Social Science Analyst Office of Juvenile Justice and Delinquency Prevention Room 742,633 Indiana Avenue, N.W. Washington, DC 20531 U.S. Department of Transportation James L. Nichols, Ph.D. Deputy Director Scientific and Technical Affairs National Highway Traffic Safety Administration 400 5th Street, S.W. Room 5130, Washington, D.C 20590 Joan White Quinlan Prevention Program Coordinator, Special Programs Division Office of Alcohol and State Programs National Highway Traffic Safety Administration 400 7th Street, S.W. Room 5130, Washington, DC 20590 . . . VIII Abbreviations 402 Programs 408 Programs 410 Programs ABC ADAMHA ASAP ASTHO BAC BIA CDC CPL DDP DHHS DOJ DOT DUI DWI EMS FARS FBI FCC IACP IHS Illegal per se MADD MCD NAB NAGHSR NCADD 23 U.S.C. 402 Highway Safety Programs. These formula grants to the States for highway safety programs are administered by the Governors' Representatives for Highway Safety. 23 U.S.C. 408-I incentive grants to States for alcohol traffic safety progr- 23 U.S.C. 410-I incentive grants to States for drunk driving prevention programs Alcohol Beverage Control Alcohol, Drug Abuse, and Mental Health Administration Alcohol Safety Action Projects Association of State and Territorial Health Officers Blood Alcohol Concentration Bureau of Indian Affairs Centers for Disease Control Certified products list Designated driver programs Department of Health and Human Services Department of Justice ' Department of Transportation Driving Under the Influence Driving While Intoxicated The terms DWI and DUI are synonymous, meaning either driving while intoxicated or driving while under the influence. These are general terms referring to the criminal action of driving a motor vehicle either (1) while "illegal per se" or (2) while impaired, under the influence, or intoxicated by alcohol or other drugs. Emergency Medical Systems Fatal Accident Reporting System Federal Bureau of Investigation Federal Communications Commission International Association of Chiefs of Police Indian Health Service Refers to State laws that make it an offense to operate a motor vehicle whiie at or above a specified blood alcohol level. Mothers Against Drunk Driving Multiple Cause of Death file taken from death certificate data by National Center for Health Statistics National Association of Broadcasters National Association of Governor's Highway Safety Representatives National Commission Against Drunk Driving NCADI NCIC NHTSA NIDA NSA OJJDP OSAP PBT PHS PI&E PSA RADAR RFP RID RWlD SRP TEAM TIPS National Clearinghouse for Alcohol and Drug Information National Crime Information Center National Highway Traffic Safety Administration National Institute on Alcohol Abuse and Alcoholism National Institute on Drug Abuse National Sheriffs Association Office of Juvenile Justice and Delinquency Prevention Office for Substance Abuse Prevention Preliminary breath tester Public Health Service Public information and education Public service announcement Regional Alcohol and Drug Awareness Resource Network Request for proposals Remove Intoxicated Drivers Riding With Intoxicated Driver Safe ride programs Techniques of Effective Alcohol Management Training for Intervention Procedures by Servers of Alcohol List of Participants Ritchie Aanderud. President, Remove Intoxicated Drivers, Washington State, Puyallup, WA. Doris Aiken. President, Remove Intoxicated Drivers, Schenectady, NY. John Allen, PhD. Chief, Treatment Research Branch, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. Gloria Ames, CDR. Immigration and Naturalization Service, Health Care Program, Washington, DC. David Anderson, PhD. Substance Abuse Specialist, Arlington, VA. Robert Apsler, PhD. Harold Russell Associates, Inc., Winchester, MA. Loran Archer. Deputy Director, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. Katherine Armstrong, PhD. Coordinator, Children and School Programs, Oflice of Disease Prevention and Health Promotion, U.S. Public Health Service, Washington, DC. Charles K. Atkin, PhD. Professor, Department of Communications, Michigan State University, East Lansing, MI. Darryl Bertolucci. Mathematical Statistician, Division of Biometry and Epidemology, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, Rockville, MD. Harold Brandt, MD. Mothers Against Drunk Driving, Baton Rouge, LA. Rebecca Brown. National Public Policy Committee Chairman, Mothers Against Drunk Driving, New Port Richey, FL. Johnny Mack Brown. Sheriff, Greenville County Sheriffs Office, Greenville, SC. Raul Caetano, MD, PhD. Alcohol Research Group, Berkeley, CA. Beverly Campbell. Church of Jesus Christ of Latter Day Saints, Washington, DC. Kay Chopard. Staff Attorney, Prosecuting Attorneys' Training Council, Des Moines, IA. Marlene Cole, LCDR. Veterinary Resources Branch, Division of Research Services, National Institutes of Health, U.S. Public Health Service, Bethesda, MD. Ricky Davidson. Chief, Emergency Medical Services Committee, International Association of Firechiefs, Shreveport, LA. Galen Davis. Special Assistant to the Governor, Topeka, KS. Delores Delaney. Member, National PTA, Committee on Health, Virginia Beach, VA. Robert W. Denniston. Director, Division of Communication Programs, Office for Substance Abuse Prevention, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. Ted Doege, MD. American Medical Association, Chicago, IL. John Donovan, PhD. Research Associate, Institute of Behavioral Sciences, University of Colorado, Boulder, CO. xi James P. Donovan. Chief, Division of Law Enforcement Services, Bureau of Indian Affairs, Washington, DC. Martin R. Eichelberger, MD. Director, Emergency Trauma Service, Children's Hospital, National Medical Center, Washington, DC. Vernon Ellingstad, PhD. Department of Psychology, University of South Dakota, Vermillion, SD. Ann Esch. North Carolinians Against Intoxicated Drivers, Winston-Salem, NC. James A. Far-row, MD. Director, Division of Adolescent Medicine, Director, Reduce Adolescent Drinking and Driving Project, Department of Pediatrics, University of Washington, Seattle, WA. Howard Filston, MD. Duke University Medical Center, Durham, NC. Mary L. Ganikos, PhD. Program Director, Late Life Alcohol Abuse, Prevention Research Branch, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, Mental Health Administration, U.S. Public Health,Service, Rockville, MD. Dean Gerstein. Study Director, National Academy of Sciences, Washington, DC. Mark Goldman, PhD. Professor of Psychology, Director, Clinical Program, University of South Florida, Tampa, FL. Michael Goodstadt, PhD. Director, Division of Prevention, Rutgers Center for Alcohol Studies, Rutgers University, Piscataway, NJ. Lawrence'A. Greenfeld. Deputy Associate Director, Bureau of Justice Statistics, Department of Justice, Washington, DC. Joseph Gusfield, PhD. Professor, Department of Sociology, University of California at San Diego, La Jolla, CA. Frank Hamilton, MD, CAPT. National Institutes of Health, U.S. Public Health Service, Bethesda, MD. Maury Hannigan. Deputy Commissioner, California Highway Patrol, Sacramento, CA. John Harvey. Elementary and Traffic Safety Education Consultant, State Department of Education;Montpelier, VT. William Hayes. Manager, New Jersey Highway Safety Office, Trenton, NJ. Anthony J. Heckemeyer. Presiding Circuit Judge, 33rd Judicial District, Benton, MS. Sandy Heverly. Mothers Against Drunk Driving, Las Vegas, NV. Ralph Hingson, ScD. Professor, Chief of Social and Behavioral Sciences, School of Public Health/College of Medicine, Boston University, Boston, MA. Harold Holder, PhD. Director, Prevention Research Center, Berkeley, CA. Jan Howard, PhD. Chief, Research Prevention Branch, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. Michael Impellizzeri. Chief, Special Programs Division, National Highway Traffic Safety Administration, Washington, DC. Michael Jacobson, PhD. Director, Center for Science in the Public Interest, Washington, DC. Paul Kamenar. Executive Legal Director, Washington Legal Foundation, Washington, DC. Jean Kilbourne, EdD. Board Member, National Council on Alcoholism, West Newton, MA. Michael Klitzner, PhD. Pacific Institute of Research and Evaluation, Bethesda, MD. John H. Lacy, PhD. Director for Alcohol Studies, The University of North Carolina, Highway Safety Research Center, Chapel Hill, NC. Ray Larson, Commonwealth's Attorney, Lexington, KY. xii Patricia D. Mail, CAPT. Division of AIDS Programs, U.S. Public Health Service, &dviile, MD. Philip May, PhD. Professor, Department of Sociology, University of New Mexico, Albuquerque, NM. Michael Mazis, PhD; Professor and Chair, Department of Marketing, American University, Washington, DC. Katherine McCarter. Associate Executive Director, American Public Health Association, Washington, DC. John McCarthy, PhD. Department of Sociology, Catholic University, Washington, DC. George McCarthy. Chairman, Alcoholic Beverage Control Commission, Boston, MA. Susan McLaughlin, MSN, RN. Senior Staff Specialist, Center for Nursing Practices, American Nurses Association, Kansas City, MO. Allan Meyers, PhD. Associate Dean, College of Liberal Arts, Boston University, Boston, MA. John Moulden. National Transportation Safety Board, Washington, DC. Joyanne Murphy, LCDR. National Institutes of Health, U.S. Public Health Service, Bethesda, MD. Dennis Nalty, PhD. Director, Oftice of Policy Analysis and Development, South Carolina Commission on Alcohol and Drug Abuse, Columbia, SC. Carl E. Nash. Chief, Accident Investigations Division, National Center for Statistics and Analysis, National Highway Traffic Safety Administration, Washington, DC. James Nichols, PhD. Deputy Director, Scientific and Technical Affairs, National Highway Traffic Safety Administration, Department of Transportation, Washington, DC. Pcrla Niguidula. Club Operations Specialist, Recreation Services Department, Naval Military Personnel Command, Arlington, VA. Robert Niven, MD. Harper Hospital, Department of Psychiatry, Detroit, MI. Pctcr O'Rourke. Director, Office of Traffic Safety, Business and Transportation Agency, Sacramento, CA. . Howard P. Patinkin. Assistant Deputy Superintendent, Chicago Police Department, Traffic Division Administration, Chicago, IL. Terry Pence. Traffic Safety Section, State Department of Highways and Public Transportation, Austin, TX. M.W. Perrine, PhD. Director, Vermont Alcohol Research Center, Burlington, VT. James Peters. Responsible Hospitality Institute, Springfield, MA. Charles Phelps, PhD. Professor of Economic and Political Science, Director, Public Policy Analysis Program, University of Rochester, Rochester, NY. Joan White Quinlan. Coordinator, Prevention Program, Special Programs Division, National Highway Traffic Safety Administration, Department of Transportation, Washington, DC. George L. Reagle. Associate Administrator for Traffic Safety Programs, National Highway Traffic Safety Administration, Washington, DC. Nancy Ricci. President, Remove Intoxicated Drivers, Connecticut, Wallingford, CT. Mary Beth Robinson. Director of Public Relations, National Automobile Dealers' Association, McLean, VA. James D. Rogers, Judge. Fourth Judicial District Court, Minneapolis, MN. H- Laurence Ross, PhD. National Institute of Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. . . . XIII Wes Roy. Representative, City of Radcliff, Radcliff, KY. Sue Rusche. National Drug Information Center of Families in Action, Atlanta, GA. Micky Sadoff. National President, Mothers Against Drunk Driving, Hurst, TX. Robert Sal& PhD. Assistant Director, Prevention Research Center, Berkeley, CA. Lawrence Schneider. Health Services and Facilities Consultant, Office of Emergency Medical Services, Department of Health and Rehabilitative Services, Tallahassee, FL. William Scott. Director, Office of Alcohol and State Programs, National Highway Traffic Safety Administration, Washington, DC. Jerri Shaw. Johnson, Bassin and Shaw, Inc., Silver Spring, MD. Harvey Siegel, PhD. Professor and Director, Wright State Substance Abuse Intervention Programs, School of Medicine, Wright State University, Dayton, OH. Bruce Simon+Morton, EdD, MPH. Associate Professor, Center forHealth Promotion Research and Development, University of Texas, Health Science Center at Houston, Houston, TX. David A. Sleet, PhD. Professor, Department of Health Science, Graduate School of Public Health, San Diego State University, San Diego, CA. Phillip Smith, MD, CDR. Indian Health Service, U.S. Public Health Service, Rockville, MD. Richard J. Smith III, CDR. Manager, Injury Prevention Program, Indian Health Service, U.S. Public Health Service, Rockville, MD. Carl Soderstrom, MD. Assistant Professor of Surgery, Associate Director of Physician Education, University of Maryland, Maryland Institute for Emergency Medical Services Systems, Baltimore, MD. Dorothy Stephens, LTJG. Health Resources and Services Administration, U.S. Public Health Service, Rockville, MD. Alvera Stern, PhD. Administrator, Division of Prevention and Education, Illinois Department of Alcoholism and Substance Abuse, Chicago, IL, Kathryn Stewart. Pacific Institute for Research and Evaluation, Vienna, VA. James W. Swinehart, PhD. President, Public Communication Resources, Inc., Pelham Manor, NY. Elsie Taylor. Public Health Advisor, Prevention Research Branch, National Institute on Alcohol Abuse and Alcoholism, Alcohol, Drug Abuse, and Mental Health Administration, U.S. Public Health Service, Rockville, MD. John M. Templeton, Jr., MD. Childrens Hospital of Philadelphia, Philadelphia, PA. Stephen Teret, JD, MPH. Director, Johns Hopkins Injury Prevention Research Center, Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD. Rae Tyson. USA Today, Arlington, VA. Chauncey Veatch, III. President, National Association of State Alcohol and Drug Abuse Directors, Department of Alcohol and Drug Programs, Sacramento, CA. Robert B. Voas, PhD. Pyramid Planning, Bethesda, MD. Alexander Wagenaar, PhD. Associate Research Scientist, University of Michigan, Transportation Research Institute, Ann Arbor, MI. Lawrence Wallack, Dr.P.H. Associate Professor, School of Public Health, University of California, Berkeley, CA. Julian Waller, MD, MPH. Professor, Department of Medicine, University of Vermont, Burlington, VT. Patricia Waller, PhD. Associate Director for Driver Studies, Highway Safety Research Center, University of North Carolina, Chapel Hill, NC. xiv J& Watne. Special Assistant, Attorney General, State of Minnesota, Rosevihc, MN. Clark Watts, MD, MPH. Professor of Neurosurgery, University of Missouri, Arlington, VA. Richard Waxweiller, PhD. Chief, Epidemiology Branch, Division of Injury Epidemiology and Control, Center for Environmental Health, Centers for Disease Control, U.S. Public Health Service, Atlanta, GA., Elizabeth A. Weaver. Director of Education, Motorcycle Safety Foundation, Irx-ine, CA. Allan F. Williams, PhD. Senior Researcher, Insurance Institute for Highway Safety, Washington, DC. LIarsha Woodward. President, Remove Intoxicated Drivers, Allen County, Scottsvik, KY. Judy Zundel. Special Assistant to the Deputy Assistant Secretary, Indian Affairs (Trust and Economic Development), Bureau of Indian Affairs, Washington, DC. xv Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Workshop Planning Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fcdcral Advisory Group onFohowup Activities for the Workshop . . . . . . . . Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of P&icipliIltS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . pricing and Availability Alcohol Beverage Control Policies: Their Role in Preventing Alcohol-Impaired Driving . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander C. Wagenaar and Susan Farrell Advertising and Marketing Mass Communication Effects on Drinkiig and Driving . . . . . . . . . . . . Charles K Atkin Epidemiology and Data Management Epidemiologic Perspectives on Drunk Driving . . . . . . . . . . . . . . . . M. W. Penine, Raymond C. Peck, and James C. Fell Education Controlling Injuries Due to Drinking and Driving The Context and Functions of Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bruce G. Simon+Morton and Denise G. Simons-Motion Judicial and Administrative Processes The Effectiveness of Legal Sanctions in Dealing With Drinking Drivers . . James L. Nichols and H. Laurence Ross Prosecution, Adjudication, and Sanctioning: A Process Evaluation of Post-1980 Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James L. Nichols and Kevin E. Quinlan Law Enforcement Issues in the Enforcement of Impaired Driving Laws in the United States . Robert B. Voas and John H. Lacey TransPortation and Alcohol Service Policies Transportation Alternatives for Drinkers . . . . . . . . . . . . . . . . . . . . Robert Apsler Server Intervention and Responsible Beverage Service Programs . . . . . . Robert F. Salti . . . 111 V vii ix xi 1 15 35 77 93 113 136 157 169 xvii Injury Control Injury and Disability Prevention and Alcohol-Related Crashes . . . . . . . . 180 Julian A. Wailer Youth and Other Special Populations Youth Impaired Driving: Causes and Countermeasures . . . . . . . . . . . . 192 Michael Klitzner Motor Vehicle Crashes and Alcohol Among American Indians and Alaska Natives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Philip A. May Drunk Driving Among Blacks and Hispanics . . . . . . . . . . . . . . . . . . 224 Jan M. Howara', Elsie D. Taylor, H. Laurence Ross, ,, and Mary L. Ganikos Treatment Rehabilitation Countermeasures for Drinking Drivers . . . . . . . . . . . . . 234 Kizth~n Stewart and Vernon S. Ellingstad Citizen Advocacy Independent Citizen Advocacy: The Past and the Prospects . . . . . . . . . 247 John D. McCarthy and Debra S. Harvey . . . XVIII Pricing and Availability Alcohol Beverage Control Policies: Their Role in Preventing Alcohol-Impaired Driving ' Alexander C. Wagenaar, Ph.D. 77~ University of Michigan Susan Farrell National Institute on Alcohol Abuse and Akoholism The phrase "alcohol control policies " refers to the entire constellation of laws and regulations at the Federal, State, county, and city levels that affect how alcoholic ~,cvcragcs are manufactured, packaged, distributed, sold, and consumed. Control politics are central in any comprehensive discussion of the prevention of impaired (Iriving because the availability of alcoholic beverages is a necessary condition for inlpaircd driving. Furthermore, alcohol control policies, interacting with private market rncchanisms, directly determine the degree to which beverage alcohol is available to consumers. Concern with alcohol control policies has grown over the past two decades. Scientists xnti professionals in the alcohol studies field increasingly recognize that alcohol is a risk lxtor for a number of health problems, including traffic crashes, at both the individual :ind societal levels. That is, the more alcohol a given individual drinks, the higher the risk Ior health problems associated with that drinking, including automobile crash involve- rncnt (NHTSA 1985). 2 Perhaps more important for public policy, the relationship also holds true at the :ygrcgatc level. As a society consumes more alcohol, rates of alcohol-related problems xc likely to increase (Moore and Gerstein 1981). Clearly, the relationship is not ctnc-to-one, since hundreds of factors contribute to each health problem, including motor vehicle injuries. For example, an increase in injury risk associated with higher :Jcohol consumption could be offset by a decrease in risk resuking from other actions, \uch as increased safety belt use. The important point is that alcohol consumption and xsociated problems such as traffic crashes are viewed as public health problems, with ;I large population at risk of involvement in alcohol-related crashes. To be most effective, prevention strategies should reduce risks across the population, rather than focus on the rclalivcly small segment of society that at any given time exhibits extensive problems with alcohol (i.e., addicted drinkers). Since customs and patterns of alcohol consumption I Warm thanks are expressed to several individuals who provided helpful comments On an earlier draft: f larold Holder, James Mosher, Joan Quinlan, and Fredrick Streff. 2 Obviously, the relationship bemen alcohol consumption and alcohol problems is not deterministic, but probabilistic. Increased consumption of alcohol increases the probability of associated problems, such as `nfficcrashes. IMany individual differences and situation-specific factors affect the Outcome in any given case. BACKGROUND PAPERS apparently spread through the population by social diffusion (Skog 1980,1985), alcohol control measures are likely to affect all consumers of alcohol, including both those with low-risk drinking patterns and those with high-risk drinking patterns. Another consequence of the public health view is recognition that very small changes in behavior by huge populations can result in substantial net benefits to society in terms of reduced alcohol-related problems. For example, a small reduction in an individual's alcohol consumption is not likely to have an immediately observable effect on that person's health. However, the same proportionate decrease in alcohol consumption across the entire society is much more likely to have demonstrable benefits in terms of reduced rates of alcohol-related problems. Therefore, the relevant consideration is not whether a specific alcohol control policy has an observable effect on given individuals, but whether changes in behavior (perhaps undetectable at the individual level) cause demonstrable changes in rates of health problems in the aggregate. Alcohol control policies might affect impaired driving by two mechanisms. First, such policies encourage or restrain the total amount of alcohol consumed, and amount is a risk factor for impaired driving and the injuries that result. Second, specific control policies alter thepattern of alcohol consumption (i.e., how a given quantity of alcohol is consumed across time and across situations). For example, it is sometimes suggested that policies that encourage drinking in one's own home rather than in a bar or tavern be adopted to reduce the likelihood of impaired driving. Obviously, such policies might reduce traffic crashes but exacerbate other problems associated with alcohol, such as household injuries or spouse or child abuse. This chapter has three objectives. Fist, we describe the types of laws and regulations included under the broad rubric of Alcohol Beverage Control (ABC) policy. We briefly discuss many dimensions of ABC policy to encourage a broader consideration of the mechanisms already available that may be useful in efforts to reduce alcohol-impaired driving and its damaging consequences. We do not include a lengthy discussion and analysis of the research evidence for the efficacy of each of these many policy dimensions in reducing alcohol consumption and alcohol-related problems. For most dimensions of ABC policy, evaluation research is scarce, and many of the studies that are available have major research design or implementation problems that limit confidence in the results. Therefore, the second objective is more modest. We identify ABC policies that have a significant body of research available, specifically those for which there is a scientific basis for assessing their utility in reducing impaired driving. The third objective is to present recommendations for changes in ABC policy and its application, acknowledging that both scientific and political considerations necessarily influence both the develop- ment and implementation of public policy. Alcohol Control Policies To structure the discussion of the wide variety of alcohol control policies, we have grouped them into eight categories: o Economic control policies o Marketing control policies 3 This is not to minimize the benefits of relatively small changes in consumption in certain situations. For example, reducing a driver's blood alcohol concentration from 0.08 g/100 ml to 0.01 g/100 ml by consumption of hvo rather than four drinks in an hour reduces the risk of involvement in a traffic crash by more than SO percent (Jones and Joscelyn 1978). . Structure of the distribution system 0 Beg&&on of individual outlets o Sehmg/serving control policies . Controls on product contents and packaging 0 Legal availability control policies . Social availability control policies Some regulations span these categories; we have placed them in the category with shich they are most closely identified. The major function of the ategori&on is heuristic-to show the breadth of policies that fall under the term "alcohol control policy" and to show how specific policies are conceptually related. 3 Economic Control Policies ne most significant influence of ABC policy on the price of alcohol is the level of excise and sales taxes on those beverages. Some jurisdictions have special tax rates for ,pcciBc products (e.g., alcoholic beverages containing local citrus products are treated favorably in Florida). Federal excise tax rates on alcohol vary across beverage types (e.g., hccr, tine, distilled spirits). They are levied on the quantity sold (e.g., bottle, barrel, gallon) rather than on price, and are not adjusted for inflation. Except for a small increase in the tax on distilled spirits, Federal excise taxes have remained constant since 1951. As ;1 result of this and other factors, the real price of alcoholic beverages has fallen aut,stantially over the last several decades. State excise taxes on alcohol have been iucreascd periodically, but also tend to fall behind inflation. In addition, the effective price of alcohol to consumers is influenced by levels of disposable income available, with ;rlc(~hol becoming less expensive when macroeconomic conditions are favorable and incomes rise, unless retail alcohol prices rise accordingly. A number of other economic control policies affect the (nominal) price of alcohol to consumers. In some States, for some beverage classes, public policy determines the exact lcvcl of retail price charged to consumers, and prices are uniform throughout the jurisdiction. In some cases, price levels and variability are controlled, short of specifica- tion of exact retail prices of alcohol. Minimum/maximum prices can be established directly, or minimum/maximum markups over wholesale prices can be authorized. Rcb~tcs of purchase price after the sale arc prohibited in some areas, as are special price promotions such as "happy hour" discounts. Other inducements to purchase alcohol- such as coupons, gifts, and prizes-may be regulated or prohibited. Provisions under which credit can be extended to retailers and consumers for the purchase of alcohol also idhmcc the cost and accessibility of alcohol. Finally, the price of alcohol to some consumers is significantly affected by whether alcohol purchases are tax deductible (Mosher 1983). For those in higher income categories (i.e., with higher marginal tax rates), tax deductions for alcohol consumed in the course of business activity effectively reduce the price by one-third. Marketing Control Policies Most discussion of ABC policies concerning marketing focuses on restrictions on the advertising of alcoholic beverages. Advertising may be prohibited outright for some beverages in some media. More commonly, the content of advertising is regulated. Content issues include whether prices may be listed, whether the alcohol content of the advertised beverage may be stated, whether actual consumption of alcohol may be depicted, and minimum age for models that may be used. Current policies frequently include limits on more subjective characteristics, such as content that appeals to "Prurient interests"; is "offensive, gaudy, or blatant"; "' tllustrates women sensuously'; BACKGROUND PAPERS uses "religious signs or symbols"; or uses words like "booze" or "saloon." Current Federal regulations include language regarding limits on misleading or deceptive adver- tisements, although these limits have not been consistently enforced (Mosher and Wallack 1981). Prohibitions on lifestyle advertising bave been suggested. Lifestyle advertising closely ties alcohol consumption to personal, financial, athletic, and sexual satisfaction and success, in contrast to advertising that focuses on specific characteristics or descriptions of the beverage. Which media are appropriate for alcoholic beverage advertising is an issue in ABC policy. Should such advertising be permitted on billboards and in the broadcast media, where a substantial part of the audience is under the legal drinking age? A similar question holds for magazines having most of their readers under the legal age. The role of advertising revenues in influencing media coverage of health and social consequences of alcohol use is also relevant. The extent of such influence regarding alcohol is currently unknown. However, research has shown a clear relationship between amount of revenues received from tobacco advertisers and editorial content on the hazards of smoking. Publications with large numbers of cigarette advertisements rarely mention the hazards of smoking in their articles on health (Warner 1985,1986). In addition to advertising, many other dimensions of the promotion of alcoholic beverages are susceptible to regulation. Displays and posters promote alcoholic beverages at the point of sale. T-shirts, jackets, and other clothing reinforce messages of advertising campaigns. Other products with beverage alcohol names and images are frequently marketed (e.g., Bud Light Spuds MacKenzie dolls are sold in toy and novelty stores). Sponsorship of sporting matches, music concerts oriented toward teenagers (rock concerts), and other events also promotes alcoholic beverages. Alcoholic beverages are distributed free of charge at special promotions. Fees are paid to movie producers in exchange for depicting on-screen, integrated into the plot, the use of a specific brand of alcoholic beverages. This practice constitutes advertising even though viewers may not perceive it as such. In addition to controls on alcoholic beverage advertising, requirements for counter- advertising have been proposed. Requiring advertisements on the hazards of alcohol ("equal-time" policies) and specifying that alcoholic beverage containers have warning labels regarding those hazards are frequently mentioned as means to partially balance advertising claims that encourage alcohol use with information on the risks of such use. (Rarely do proposals for counter-advertisements literally specify "equal time." Typical- ly, a lower ratio of advertisements to counter-advertisements is proposed, for example, one counter-advertisement for every four or five advertisements.) Other proposals include compulsory warning messages in all alcohol advertising (similar to the warnings in cigarette advertisements) and required warning posters where alcohol is sold or consumed. Finally, allowing or limiting the tax deductibility of advertising and other promotional efforts is another dimension of ABC policy that affects the marketing of alcoholic beverages. Structure of the Distribution System The most commonly noted characteristic of the alcoholic beverage distribution system in the United States is whether a given State has a monopoly or license system. States are frequently dichotomized as to whether they have a monopoly on alcohol sales or whether they license private enterprises to distribute alcoholic beverages. In reality, the monopoly-license dimension is a continuum, with States distributed at varying points according to the degree to which they control alcohol sales. Monopolies are frequently limited to a single class of beverage; for example, distilled spirits may be monopolized, while beer and wine are not. Monopolies may be limited to the wholesale level, or may i~,,Tr both &oiesalc and retail sales. Conceptual an! empirjc.? development of scales , , n,c.sure where each State is on the control contmuum 1~ m the very early stages ;;r,,l~cr and Jams 1987). Such development should be encouraged to help move the rcsc.rch and policy diSCUSSiOns away from the simplistic tendency to dichotomize J&&ulion systems. ~~~~~~~~~ to rhe degree to which the distribution system is a public monopoly is the .Iruz(ult' and power of the agencies responsible for alcoholic beverage control. The :lu,T,t,c'f and characteristics of the people on the governing board, the nature of the ,,l,~~,~in[ing authority, and the grounds for removal of board members and the agency ,j;rcc[nr affcct how.responsive the control agency IS to local commumty concerns about .,j<-irhnl outkts. (;c~vcrnmcnt regulation affects many other dimensions of the distribution system ,[ruclurc, and these dimensions warrant attention regarding their effects on alcohol iL,nsumption and associated problems such as alcohol-impaired driving. Regulation of franchi.cc alcohol outlets, amount of competition permitted, degree to which private ,~~~,n~lpolics or oligopolies are permitted, provisions allowing localized prohibition of ' ,,l~~~h~>l salts, and extent of local government or community review of alcohol outlets are ,M]V ;I few of the dimensions of ABC policy that directly affect the structure of the .IIc;,hoIic hcxrage distribution system (Roth et al. 1987). l. Cpnfmdtlzatlon mvolves htmg the market to a small numkr of products and dlfferentlatmg them based on .pr$ct content, rather than on [hc image of the product and the market-segment for which d 1s positioned. For example, !hr kc. market might be restricted to hght, medmm, and heavy beer, based on alcohol il,nrcnt. With the exception of specifying the type, all packaging.and labeling would be i,lcnlicd across all brands. The implications of such a move toward generic alcoholic 1Yvcr3ges are complex and would represent a dramatic change from the current market rwcwre. - Incrcscd understanding of the nature of beverage alcohol markets and the potential rc,lC `,fABC policy in structuring those markets to minim& risks associated with alcohol m3v help identify less dramatic (and more feasible) regulatory changes that nevertheless miiht minimize adverse effects on public health. For example, wine coolers are new pr;Klucts that have been sumssftiy marketed in recent years. They are &signed to ~l~pca] to a different population from traditional wine drinkers. Even more recently, w&e Iation should be enacted making it illegal per se for a person with an alcohol concentration of 0.10 or higher within 3 hours of arrest to drive or be in actual physical control of a motor vehicle. As of January 1988,44 States plus Puerto Rico had illegal per se laws. Appellate Action Prosecutors should initiate appropriate appellate actions to ensure judicial compliance with statutory mandates governing DUK cases. . . . courts frequently &ore mandatory sentencing requirements in DUI cases. UXess the prosecutor is willing to seek an appellate remedy, the practice will continue unchecked. Adjudication Mandatory Sentencing llle sentences recommended upon conviction of driving under the influence should be mandatory and not subject to suspension orprobation. Specifically, the recommendations are that: Ail States establish substantial mandatory minimum fines for DUI offenders, with correspondingly higher mandatory fines for repeat offenders. As of January 1988,16 States had mandatory minimum fines for first offenders. Any person convicted of afirst violation of driving under the influence should receive a mandatory license suspension for a period of not less than 90 days, plus assignment of 100 hours of communi@ service or a minimum jail sentence of 48 consecutive hours. As of January 1988, 24 States had mandatory minimum license suspension for first-time offenders; 7 States mandated jail or community service for first offenders, and an additional 7 States mandated jail without any provisions for community service. Anyperson convicted of a second violation of driving under tile inji'uence within 5 years should receive a mandatory minimum jail sentence of 10 days and license revocation for not less than 1 year. As of January 1988,14 States mandated some period of jail or community service, and an additional 28 States mandated jail without any provision for community service for second-time offenders; 43 States and the District of Columbia mandated license suspen- sions or revocations for second offenders. Any person convicted of a third or subsequent DUI violation within 5 years should receive a mandatory minimum jail sentence of 120 days and license revocation for not less than 3 years. As of January 1988, 39 States mandated jail for a third offense. Nearly all States mandated license suspension for a third offense. JUDICIAL AND ADMINISTRATIVE PROCESSES 131 Sentencing of Suspended Drivers Who Continue to Drive Stales should enact a statute requiring a man&tory sentence of al least 30 days for any person convicted of driving with a suspended or revoked license or in violation of a restriction due lo a DUI conviction. Few States have effectively implemented such sanctions, alth,ough some appear interested. Some States have also been considering attacking the problem of driving while suspended by confacating the license plates or the vehicles of such offenders. The license plate confiication approach is intended to make driving while suspended a more visible offense. It has the added advantage of little or no cost. Felony DUI Causing death or serious injury to others while driving under the influence should be classified as a felony. In 1988, 44 States and the District of Columbia had death-related offenses, often called vehicular homicide. In 38 States and the District of Columbia, this constituted a felony charge, and in 6 States it was a misdemeanor. Court Administration Spee& trial: DUI cases al the trial levelshould be concluded within 60 days of arrest. Sentencing should be accomplished within 30 days. The appellate process should be expedited and concluded within 90 days. From information reviewed, it appears that few courts have achieved this. Preconviction Diversion Preconviclion diversion to alcohol education or alcohol rreatmenr programs should be eliminated. AJindingon the charge should be rendered andparticipa- tion in education or treatment programs should then become a condition of sentencing. Although most States have eliminated statewide diversion programs, a few States and courts in several States still regularly divert offenders from sanctions into education or treatment. Often in such systems, no conviction and no record of an alcohol-related offense exists. Presentence Investigation Before sentencing, a court should obtain and consider upresentence investiga- tion report detailing the defendant's driving and criminal record and, where possible, an alcohol problem assessment report. In all cases, an alcohol problem assessment report should be completed by qualifiedpersonnelprior lo rhe determination of an education or treatmentplan. NCADD estimated that 23 States complied as of 1988. Victim Programs Any person cdnvicted of driving under the influence who causes personal injur>, orproperty damage should pay res&ution. As of 1988, NCADD estimated that 42 States complied. The U.S. Congress should enact legislarion that eliminates the possibility rhat a dnrnk aver, judged civilly liable, will be able to escape the penalties of civil action by filing for bankmp~q. 132 No additional information available. State and local governments or courts by rule should require victim impact statements (including oral or written statements by victims or survivors) prior to sentencing in all cases where death or serious injury results from a DUI offense. Licensing Administration Administrative Per Se License Suspension States should enact legislation to require prompt suspension of the license of drivers charged with DUI upon a finding that the driver had a BAC of 0.10 in a legally requested andproperly administered test. 77zeprompt suspension should also extend to those who refuse the test (i.e., implied consent), as ~$1 as those who are tiving in violation of a restricted license. As of the end of 19&3,24 States plus the District of Columbia had administrative per se laws, with many variations in the provisions of these laws. Restricted Licenses Each State driver's licensing authority should review irs practice of issuing occupational hardship dn'ver's licenses following suspension or revocah*on and establislt strict uniform standards relative to issuance and control of such limited driving privileges. i%ese licenses should be issued only in exceptional cases. In no event should this be done for repeat offenders. In fact, most States with mandatory license revocation, whether by the court or by the administrative process, make extensive use of restricted, occupational, and probationary licenses. Only about 21 States (i.e., those qualified for 408 incentive grant funds) have mandatory, minimum hard 1iCense periods during which restricted licenses are not to be issued. Provisional License for Young Drivers States sllould adopt laws providing a provisional license for young beginner drivers which would be withdrawn for a DUI conviction or an implied consent refusal. Education and lkeatment Assignment Process Rekabilitation and education programs for individuals convicted of DUZ should beprovided as a supplement to othersanctions andnot as a replacement for those sanctions. Although most States appear to have moved away from diversion programs, some still make extensive use of them and allow sentencing to treatment programs in lieu of a conviction or a license or jail sanction. Presentence investigations, including alcoJzo1 assessments conducted by qualified personnel, sltould be available to all courts in order to appropriately classify the defendant's problem with alcohol. Repeat offenders should be required to undergo medicalscreeningfor alcoholism by a physician trained in alcoholism, an alcoholism counselor, or by an approved treatment facility, NCADD estimates that only 23 States require a presentence of a postsentence investigation. Fewer specify the type of personnel required to administer such tests. JUDICIAL AND ADMINISTRATIVE PROCESSES AlcoJtol education programs sltould be used only for those first offenders WJIO Jlave had no previous exposure to suclt programs. Problem drinkers and repeat offenders should be referred to more intensive rehabilitation programs. No additional information available. AlcoJ~ol treatment and rehabilitalion programs should be available for in- dividuals judged to need such services. The programs should be tailored to the individual's needs and the individual sltould be assigned to such programs for a length of time determined by treatment personnel and enforced by court probation. Most States use existing treatment facilities. It is not known how many programs include treatment tailored to the specific needs of the offenders. Compliance When assignments are not complied with, the COWS or the administrative licensing agency must take steps to impose furtrter restrictions on driving privileges or to assessjiutJlerpenalties as spelled out in tlte origirlal sentence. Evidence from the States suggests that this remains an important problem to be resolved. A records reporting system sltould be available to assure lhar individual of- fenders assigned to education or treacmenf services do in fact comply witJl tJle assignments and fo make information on compliance available to motor vellicle administration officials at tlte time of appearance for relicensing. Tracking of individual offenders from arrest through completion of sanctions remains a goal to be achieved. Software has been developed to aid community-level tracking systems. REFERENCES Automotive Transportation Center. "Do Laws Make a Difference? An Analysis of Indiana's 1983 Anti-Drunk Driving Law." Report to,the Governor's Task Force to Reduce Drunk Driving, Purdue University, West Lafayette, IN, 1986. Ilruce, T., and Bruce P. The legislative response to the drunk driving dilemma: An empirical analysis of its success and failure (in Missouri). St. Louis University Law Journal, in press. I3loch, S. One year later: A preliminary assessment of the effectiveness of California's new drinking and driving laws. Abstracts and Reviews in Alcohol and Driving 4(2):9-20, 1983. filoch, S., and Aizenberg, R. The effects of tough DUI laws: California's first twenty months. Journal of Trafic Safety Education 3 l(2): 1984. Bloch S., and Aizenberg, R. "The Effects of Tough Drinking and Driving Laws: An Examination of the Judicial and Traffic Safety Experience of California. Paper presented at the 1985 Annual Meeting of the Academy of Criminal Justice Sciences, Las Vegas, NV, March 31, 1985. Blomberg, R.; Preusser, D.; and Ulmer, R. Deterrent Effects of Mandatory License Sqxmion for DWZ Convictions. DOT HS 807 138, Washington, DC: National Highway Traffic Safety Administration, 1987. Clcary, J., and Rodgers, A.Analysis of the Effects of Recent Changes in Minnesota's DWILaws. St. Paul, MN: House of Representatives Research Department, 1986. Colorado Division of Highway Safety "A Decade of Progress." Denver, CO: the Division, 1984. Fatkowski, C. "The Impact of Two-Day Jail Sentences for Drunk Drivers in Hennepin County, Minnesota." Final Report on Contract No. DTNH-22-82-05110, National Highway Traf- fic Safety Administration, Washington, DC, 1984. Federal Bureau of Investigation. Crime in the United States. Uniform Crime Reports, Washington, DC: U.S. Department ofJustice, 1987. Foley, J.; Doherty, M.; and Habegger, J. Dnmk Driver Recidivist Penalties in Indiana for 1985. Report to the Governor's Task Force to Reduce Drunk Driving. West Lafayette, IN: Automotive Transportation Center, Purdue University, 1986. 133 134 BACKGROUND PAPERS Hagen, R. Effectiveness of License Suspe?uion for Drivers Convicted of Multiple Driving-Under- i'he-Injluence offenses. Report No. 59, Sacramento, CA: Department of Motor Vehicles, 1977. Helander, C. lXe California DUI Countermeasure SystemAn Evaluation of System Processingand Deficiencies. Sacramento, CA: California Department of Motor Vehicles, 1986. Hepperle, W., and Klein, D. DUIAdjudication Evaluation Project. Alameda County, CA: Office of Court Services, 1985. Hatos, S. DWI legislation. In: I?ze 1987-88 Alcohol Highway Safety Workhop Series Reference Manual, Washington, DC: National Highway Traffic Safety Administration, 1987. Hatos, S. .!qi.&ive Digest. Washington, DC: National Highway Traffic Safety Administration, 1988. Hilton, M. The Effectiveness of Recent Changes in California Law ar Drinking Driving Counter- meauues: An Interrupted Time-Series Analysir. Berkeley, CA: Alcohol Research Group, 1983. Hingson, R.; Heeren, T.; Kovenock, D.; Mangione, T.; Meyers, A; Morelock, S.; Lederman, R.; and Scotch, N. Effects of Maine's 1981 and Massachusetts' 1982driving-under-the-influ- ence legislation. American Journal of Public Health 77(5):593-597, 1987. Homel, R. Penalties and the drunk/driver: Astudy of one thousand offenders.Au.wczZi# and New Zealand Journal of Criminology 14:225-241,1981. Insurance Institute for Highway Safety. Drinking and Driving Down Sharply in U.S. During Last 13 Years. Washington, DC: the Institute, 1987. Indiana Governor's Task Force. The Governor's Task Force to Reduce Drunk Driving 1987 Progress Report. Indianapolis, IN: the Task Force, 1987. Johnson, D. The Effect of Administrative License Revocation on Employment. Washington, DC: Office of Alcohol and State Programs, National Highway Traffic Safety Administration, 1986. Jones, R.; Joksch, H.; Laccy, J.; and Schmidt, H. "Field Evaluation of Jail Sanctions for DWI." Final Report, Mid-America Research Institute, for the National Highway Traffic Safety Administration, Washington, DC, 1987. Katz, L., and Sweeney, R. Ohio's new drunk driving law: A halfhearted experiment in deterrence. Case Western Reserve Law Review 34(2/3), 1984. Kiingberg, C.; O'Connell, P.; Saizberg, P.; Chadwick, J.; and Paulsrude, P. An Evaluation of Washington State's 1979 Driving While Intoxicated (Dwr) Laws. NHTSA Report DOT HS 806 838, Washington, DC: NHTSA, 1984. Lacey, J.; Popkin, C.; Stewart, R.; and Rodgman, E. Beliminq Evaluation of the North Carolina Safe Roads Act of 1983. Chapel Hill, NC: Highway Safety Research Center, University of North Carolina, 1984. Lacey, J. Safe RoadrAcf Updare. Chapel Hill, NC: Highway Safety Research Center, University of North Carolina, 1987. Lacey, J. Safe Roads Act Update. Chapel Hill, NC: Highway Safety Research Center, University of North Carolina, 1988. Lang, K. Santa Clara County Driving Under the injluence Arrest Data Analysb. Santa Clara, CA: Santa Clara County Alcohol Bureau, 1986. Levy, P. "Analysis of Proportion of Drivers Intoxicated in 408 and non-408 States." Unpublished report. Office of Alcohol and State Programs, National Highway Traffic Safety Admin- istration, Washington, DC, 1987. Lewis, R. Estimates of DWI Recidivism in Minnesota Fatal Crashes. St. Paul, MN: Minnesota Criminal Justice System Task Force, 1985. Massachusetts Senate Committee of Post Audit and Oversight Drunk Driving in Massachusetts: 1984 Stafzis Report. Boston, MAz the Committee, 1984. Massachusetts Senate Committee of Post Audit and Oversight. The Stafe's Drank Driving Law: AppraisingPegomzance. Boston, MAI the Committee, 1986. Minnesota Department of Public Safety. Cost of DWI Control; Cost of Lack of ControL..Deterrence of Drinking Dn'vers; License Revocation and Other Penaities. St. Paul, MN: MDPS, 1987. National Highway Traffic Safety Administration. Dm Charge Reduction Study. Washington, DC: the Administration, 1986. New Jersey Department of Public Safety. State Comnubion on Drunk Driving: Second Annual Report. Trenton, NJ: Department of Law and Public Safety, 1986. New York State Department of Motor Vehicles. An Impact Evaluation of the New York State JUD[CW AND ADMINISTRATIVE PROCESSES 135 STOP-DWI Program. Albany, NY: The Institute for Traffic Safety Management and Research, 1985. NW York State Department of Motor Vehicles. STOP-DWI: The First Four Years, An Evaluation Update. Albany, NY: The Institute for Traffic Safety Management and Research, 1986. Nichols, J. Recent Trena!s in the Alcohol-Related Crash lkblem in the United States. Washington, DC Office of Alcohol and State Programs, National Highway Traffic Safety Administra- tion, 1988u. Nichols, J. "A Review of Trends of Alcohol-Related and Nighttime Fatal Crash Data From 1982 Through 1987." An internal, unpublished analysis, National Highway Traffic Safety Ad- ministration, Washington, DC, 1988b. Palmer, J., and Tii P. Minnesota Alcohol Roadside Survey. St. Paul, MN: Department of Public Safety, 1986. Peck, R. The Trajic Safery Impact of Cahfomia `s New Dmnk DrivingLaw (AB541):An Evaluation of the First Nine Months of Er@ence. Sacramento, CA: California Department of Motor Vehicles, 1983. Peck, R. "Effects of California's New DUI Laws." Paper presented at the California Association for Safety Education Conference, Stateline, NV, 1984. Peck, R.; Sadler, D.; and Pet-tine, M. The comparative effectiveness of alcohol rehabilitation and licensing control actions for drunk driving offenders: A review of the literature. Alcohol, Drugs and Driving: Abstracts and Reviews l( 14):15-39,1985. Perrin, M. Analysis of DUI Processing From Arrest Through Post Commun ication Counter- measures. Sacramento, CA State of California, Department of Motor Vehicles, 1984. Presidential Commission on Drunk Driving. Final Reprt. Available from National Commission Against Drunk Driving, Washington, DC, 1983. Quinlan, K. DWI prosecution, adjudication and sanctioning. In: The Z987-88 Alcohol Highway Safety Workshop Series Reference Manual. Washington, DC: National Highway Traffic Safety Administration, 1987. Rodgers, A, and Cleary, J. Analysis of the Effects of Recent Changes in Minnesota's DWI Laws, Part ZZ: The Perceptions of Minnesota's Drivers. Report. St. Paul, MN: Research Depart- ment, House of Representatives, 1983. Siegal, H. Impact of Driver Intervention Program on DWI Recidivism and Problem Drinking. Washington, DC: National Highway Traffic Safety Administration, 1985. Sotter, R. Impact of DUI Legistion in the State of Florida. Tallahassee, FL: Bureau of Public Safety Management, 1986. Sterling-Smith, R. PsychologicalIdentijication ojDrivers ResponsibleforFatal VehicularAccidents in Boston. Washington DC National Highway Trafftc Safety Administration, 1976. Stewart, K, and Laurence, S. Senate Concurrenf Resolution Number 27 Report: A Review of the Implementation and Effectiveness of Drinking and Driving Legislation. Sacramento, CA: Office of Traffic Safety,1987. Surla, L; Voas, R.; Koons, S.; and Reiner, G.An Evaluation of the Elimination of Plea Bargaining for DWI wen&rs. Washington, DC: National Highway Traffic Safety Administration, 1986. Tashima, H., and Peck, R.An Evakation ofthe Califomiakutk Driving Countermeasure System: Vohme 3, An Evaluation of the Specific Deterrent Effects ofAlternative Sanctions for First and Repeat DUI Offenders. Sacramento, CA: Department of Motor Vehicles, 1986. Ti, P., and Palmer, J. Minnesota Alcohol Roadside Swvey: Summary of Reports. St. Paul, MN: Department of Public Safety, 1987. Votey, H.L., and Shapiro, P. Cost effectiveness of alternative sanctions for control of drunken driving: The Swedish case. In: Kaye, S., and Meier, G-W., eds. Proceedings of the Ninth International Conference on Akoho,! Drugs, and Trafic Safety, San Juan, Puerto Rico, DOT HS-806-814. Washington, DC: National Highway Traffic Safety Administration, 1985. Wells-Parker, E., and Co&y, P. Impact of driver's license suspension on employment stability of drunken drivers. Social Research Report Series, 87(3); Social Science Research Center, Mississippi State University, 1987. ador, P.; Lund, A; Fields, M.; and Weinberg, K. Fatal Crash Involvement and Laws Against Alcohol-Zmpaired Driving. Washington, DC: Insurance Institute For Highway Safety, 1988. 136 Law Enforcement Issues in the Enforcement of Impaired Driving Laws in the United States Robert B. Voas, Ph.D. Environmental Research Institute of Michigan (ERkf) John H. Lacey, Ph.D. The Universiy of North Carolina Highway Safety Research Center In our best year, 1983, 1.9 million drivers were arrested for driving while impaired (DWI) in the United States. This number represented approximately 1 percent of the Nation's total licensed drivers. This was a significant increase over the 197Os, when only about one-half of 1 percent of licensed drivers were arrested for DWI each year. Still, it is not enough. Speaking a decade ago, Borkenstein (1975) noted that Roadside surveys of the occurrence of alcohol in the driving public have shown that when enforcement is at the current level of 2 arrests per officer per year, and with automobile density what it is in the average congested city today, there are about 2,000 violations for each arrest. A "violation" is a trip from one point to another with a blood alcohol concentration of .lO percent or higher; thus, in a typical community of 1 million population, with 1,000 patrol officers making two arrests per man per year, there will be 2,ooO arrests and 4 million violations. Since Borkenstein made that statement, the percentage of licensed drivers arrested for DWI has doubled and, therefore, the ratio of violations to arrests may now be down to 1,000 to one. Indeed, two studies suggested that where intensive enforcement is applied, the violation-to-arrest ratio can be reduced to approximately 300 to one (Beitel et al. 1975; Hause et al. 1982). These higher arrest rates, which are not typical of the enforcement level of the country as a whole, have been shown to produce small reduc- tions in alcohol-related accidents (Voas and Hause 1987). DWI arrests nationally rose significantly from 1979 to 1983; the proportion of highway fatalities that were alcohol-related dropped 10 to 15 percent from 1982 to 1986. The extent to which this increase in arrests contributed to the subsequent decrease in alcohol-related fatalities is difficult to determine. The increase probably contributed as one element in a larger complex of factors that included citizen activist programs, new alcohol legislation, and increased public interest in health and safety (Howland 1988). Regardless, a doubling of the total number of arrests has had, at best, a modest effect on the alcohol-related casualty rate. Luckily, deterrence of drunk driving is not determined by the absolute number of arrests but by the public's perception of the probability of being arrested (Ross 1984). While it may be generally true that the more arrests made, the more the public will be deterred, there is no precise relationship between the number of arrests and the extent of deterrence. In some cases, highly publicized programs result in a higher perceived LAW ENFORCEMENT 137 level of enforcement than is produced by simply raising the number of arrests without publicity. An example of this phenomenon was reported by Williams and Lund (1984). These researchers conducted a telephone survey of drivers in Fairfax County, Virginia and Montgomery County, Maryland. Fairfax County consistently had the highest arrest rate per licensed driver, but Montgomery Country police regularly used roadside sobriety checkpoints. Citizens of both Fairfax and Montgomery counties stated that they were more likely to be arrested in Montgomery County, apparently because of the higher visible use of checkpoints. The extent to which an enforcement program succeeds in convincing potential drinking drivers that their probability of apprehension is high is important. Highly visible enforcement offers the possibility that programs can be implemented that, while not greatly increasing the total number of DWI arrests, will reduce alcohol-related crashes. Development of the Traditional Behavioral Enforcement System in the United States The drunk driving problem was first recognized in scientific literature in 1904, approximately 5 years after the first highway safety fatality in the United States. The United States and Norway were among the first industrialized nations to make impaired driving a criminal offense. In 1910, New York adopted an impaired driving law, and in 1911, the State of California followed suit. This early criminalization of drunk driving set it apart from other traffic infractions. For example, higher penalties were provided for the offense, including incarceration and substantial periods of license suspension. By 1924, the State of Connecticut was jailing 254 drivers per year for DWI. Thus, from the early years of this century, the United States has treated this offense as seriously as any nation in the world. The system of enforcement that emerged can be described by the four-step process outlined in figure 1. ' The first step in this process is to identify vehicles in the traffic flow that are being driven by impaired operators. This is done based on either the vehicle being involved in a crash or the officer on patrol observing unusual, aberrant, or illegal behavior. Once stopped, the second step is performed. The driver is interviewed to determine whether he or she has been drinking and shows signs of intoxication. Common symptoms used for this purpose are bloodshot eyes, flushed appearance, slurred speech, odor of alcohol, and so forth. If this initial interview indicates that the individual may be impaired, the officer normally takes a third step, which is to invite the driver out of the vehicle to perform a set of sobriety tests (e.g., walking a straight line, touching the nose with eyes closed) which, alongwith the aberrant driving, become the basisof the officer's testimony to support the charge of "driving while impaired". The term "drunk driving" presented considerable problems in adjudicating the DWI offense because of its lack of objective definition. The popular conception of a drunk individual involved highly aberrant behavior (e.g., staggering gait, incoherence). How- ever, it soon became evident that individuals could be at increased risk of crash involvement without displaying such symptoms. Efforts were made to strengthen initial legislation by substituting such terms as "under the influence of alcohol" and, more recently, "impaired by alcohol." However, with no objective measure of driving skill available for testing individuals charged with drunk driving, much was left to the interpretation of the jury, which was prone to find that the behaviors described were similar to their own party behaviors and were not commonly accepted as being risky. Just before the Second World War, a new factor was added to the enforcement k. -. / FIGURE 1 STAGES IN THE DWI ENFORCEMENT PROCESS A. TRADITIONAL U.S. "BEHAVIORAL BASED" ENFORCEMENT Vehicle Selection Determination Of Alcohol Use o Accident o Behavioral Cues o Moving Violations (odor of alcohol o Erratic Driving slurred speech, etc.) 0 Passive sensor o Field or Behavioral sobriety tests 0 Preliminary breath tests B. SCANDINAVIAN "CHEMISTRY BASED" ENFORCEMENT Vehicle Selection Determination of Alcohol Use Test for Impairment o Checkpoints 0 Passive Sensors 0 Preliminary Breath Tests o Breath Test o Blood Test o Urine Test I I fines (22 states LAW ENFORCEMENT 139 process-the use of chemical tests for alcohol to determine impairment. Initially, these test results were added to the total evidence presented to support the testimony of the police officer. Once the courts began to accept thii new scientific evidence, State legislatures moved to enact laws that specifically provided for chemical testing. In 1939, Indiana became the frost State to provide for a chemical test; Maine, New York, and Oregon soon followed. - This legislation was significant in that it established the principal that chemical test data provided competent evidence of impairment. In addition, these laws established specific alcohol concentrations (AC) as presumptive evidence of intoxication. Estab- lishing such a presumption required the defense to provide other competent evidence to rebut the chemical test data or lose its case. The initially prescribed levels followed the recommendations of the American Medical Association (AMA), which proposed that an individual with an AC of 0.15 or greater was presumed to be under the influence, while an AC between 0.05 and 0.15 was competent evidence of impairment when supported by other, verbal testimony. Finally, the AMA recommendation held that an AC below 0.05 was presumptive evidence that the individual was not under the influence. Since the Second World War, most States have lowered this presumptive level to 0.10, and several States have lowered it to 0.08 Recently, the National Safety Council Committee on Alcohol and Drugs recommended that the presumption that an individual is not impaired when the AC is below 0.05 be stricken from DWI legislation, since recent evidence shows that the performance of a substantial number of individuals is impaired at ACs below 0.05 (Moskowitz and Robinson 1988). Thus, with the passage of these laws, a fourth step was added to the enforcement system illustrated in figure 1. Once police officers obtain sufficient evidence from sobriety tests to convince them that individuals are impaired under the State's DWI law, they will charge the drivers with the offense, take the individuals into custody, and transport them to the police station for a chemical test. Robert Borkenstein's development of an inexpensive breath test device, the Breathalyzer", provided a means for police departments to rapidly test individuals for their AC. The use of breath testing in the United States avoided many of the problems experienced in the foreign countries that continued to rely on blood tests that required a police surgeon to come to the station and draw blood. The Borkenstein Breathalyrer" and those breath test instruments that succeeded it have provided a reliable means of collecting highly accurate breath test data. States have established control systems for approving and calibrating these units and for training and supervising breath test operators in each police department. The success of the chemical test in achieving convictions for impaired driving raised the issue of whether the State could require drivers to submit to this test. In a landmark decision, the Supreme Court decided in Schmetier v California that the police had the authority to take a blood sample forcibly, under limited circumstances. The Court held with respect to the Fifth Amendment that this did not constitute self-incrimination, since the evidence gathered was not testimonial but physical in character. Secondly, the Court determined that the forcible taking of a blood sample did not violate the Fourth Amendment prohibition against unreasonable searches and seizures since there was full probable cause to suspect the driver of driving under the influence (see Laurence 1988 for a discussion of constitutional issues related to DWI enforcement and adjudication.) This decision opened the way for States to pass laws providing for the forcible taking of blood samples from arrested drivers. However, neither police departments nor legislatures wanted a system in which people would be held down and needles inserted in their arms as part of the arrest process. Therefore, acompromise was developed under which the State passed legislation providing that operating a vehicle on the State's highways implied consent for giviug a sample for a chemical test in the event of a DWI arrest. If the driver, having been arrested, refused to provide a sample, then the Motor 140 BACKGROUND PAPERS Vehicle Administrator was empowered to suspend the driver's permit for a some stated period. It required almost two decades for all States to adopt this implied consent procedure. To achieve adoption by the final hold-out States, it was necessary to increase the safeguards in the breath testing process. As the breath test became a more important element in the drunk driving litigation process and as implied consent statutes gave less opportunity for the driver to refuse testing, States added legislation to require that breath test devices be equipped with safeguards that would prevent the operator from making errors in the testing process. As a result, units were developed that automatically stepped through the process of calibrating and checking the instrument, collecting a breath sample, and providing printed output so that the possibility of error was minimized. The four-step process provides a reasonably effective enforcement system. However, the seriousness of the drunk driving offense, with its potential for a jail sentence and a lengthy driving suspension, resulted in a number of pressures being applied to the lower courts that reduced their overall effectiveness with DWI offenses. Those charged with DWI hired lawyers to argue their cases, increasing the procedural paperwork for police officers. Each step of the enforcement process had to be documented to demonstrate probable cause for the stop and the DWI charge and to show that the breath test was conducted according to State regulations by qualified personnel. The bureaucratic procedures became very onerous for the police, frequently requiring 2 to 4 hours for each arrest and thereby discouraging DWI enforcement. Because of the serious penalties, many offenders insisted upon full legal recourse, and court dockets frequently became overloaded. Significant backlogs were created, particularly when defendants demanded jury trials. The court and the prosecution were often motivated to seek plea bargains in which the individual charged with DWI pleaded to a lesser offense in return for having the drunk driving charge dropped. When the police saw this, some were discouraged from making DWI arrests. The traditional behavioral enforcement system provides wide discretion to the in- dividual oflicer in determining which vehicles to stop and, once the vehicle is stopped, whether to proceed with the investigation of the DWI offense. Thus, the officers' attitudes, detection skills, and motivation are extremely important to effective enforce- ment. In studies of police officers' attitudes toward DWI enforcement, most officers admitted to occasions when they did not pursue investigations where they were fairly sure the driver was impaired. One of the primary reasons given for failure to follow through was the length and bureaucratic nature of the paperwork involved. Arrests were less likely to be made toward the end of an officer's daily tour because completing the arrest would require staying overtime. Officers were also likely to consider the significance of drunk driving compared with the fairness of penalties for thii offense in making their arrest decision. Where they believed the penalties were inappropriately severe, they were more prone not to pursue arrests of marginally impaired drivers. Arrests were frequently avoided by allowing a passenger to take over the driving or, in the case of teenaged drivers, driving the individual home (Oates 1974). With the founding of the Department of Transportation, the new Highway Safety Bureau (soon to be the National Highway Traffic Safety Administration) attempted to overcome some of these problems by establishing 35 demonstration programs called Alcohol Safety Action Projects (ASAPS). These projects were designed to provide an integrated approach to the drinking/driving problem (NHTSA 1979). Courts, prosecutors, and the police received additional funds and participated in a coordinated program to increase DWl arrests by simplifying police paperwork and by increasing the speed of prosecution and adjudication. These projects generally succeeded in increasing (usually doubling) the number of IAW ENFORCEMENT 141 arrests for DWI (Levy et al. 1978). Within the enforcement activity, the arrest increases were primarily achieved through special DWI-emphasis patrols that operated on weekend evenings. These patrols usually consisted of 2 to 10 vehicles, depending on the size of the community. They normally made as many arrests in a year, on the two or three weekend evenings when they were active, as the full police force had made annually prior to the ASAP programs. While the ASAP programs came to an end by 1975, this dedicated patrol procedure has continued to be a feature of most communities in which DWI enforcement is emphasized. Aside from sponsoring the ASAP demonstrations, the Federal Government at- tempted to assist DWI law enforcement by developing more scientific and objective procedures for identifying drinking drivers. A program to determine which vehicle maneuvers were most likely to indicate an intoxicated driver was funded by NHTSA and resulted in a set of driving "symptoms" graded by the probability that the driver would be at 0.10 AC or greater (Harris et al. 1979). A second research effort was directed at developing a standardized set of field sobriety tests for use by police officers. The sobriety tests commonly in use, up until the last decade, were highly influenced by individual officers' preferences. NHTSA sponsored a review of the literature and the development of a standardized set of three tests: lateral gaze nystagmus, body sway, and divided attention. The availability of these tests, particularly the gaze nystagmus test, has increased the capability of police officers to estimate the probable alcohol content of the suspected driver (Tharp et al. 1981). By the latter part of the 197Os, the traditional behavioral system for detecting and apprehending drinking drivers had been significantly improved. This was evidenced by the fact that close to 1 million drivers were being arrested each year for this offense. It is probable that even more arrests would have been made had it been possible for the courts to handle the increase in case load. The Chemistry-Based Enforcement System The behavioral system of enforcement just described, which developed in the United States, was fairly typical of most industrialized nations. The U.S. system had some advantage in that it was based on breath rather than blood alcohol measurement. This simplified the enforcement process by not requiring the presence of a physician to collect blood. While this system was developing and maturing in the United States, the Scan- dinavian countries developed a significantly different approach to the enforcement of DWI laws. In 1936, Norway passed legislation that provided that being in charge of a vehicle and having a blood alcohol concentration in excess of 0.50 was an offense. This was the first of the so-called "illegal per se" laws. Similar laws were later adopted by the other Scandinavian countries. The significance of the illegal per se approach is that it circum- vents the issue of behavioral interpretation, since the offense has only two relevant criteria- being in charge of a vehicle and having an AC over a given limit. Once these laws were in place, the police departments in the Scandinavian countries began to use field breath test devices. These consisted of tubes through which the suspect blew. Any alcohol in the breath would cause a chemical reaction in the dichromate crystals in the tube and produce a color change, from yellow to green. The length of the stain provided a rough measure of alcohol concentration. These legal and chemical test changes, when combined with the traditionally severe sanctions provided in Scan- dinavian laws, became known as the "Scandinavian model" (Ross 1975; Andenaes 1988). The British Government implemented elements of this system in the Road Safety Act of 1967. Because of the wide publicity elicited in the British press while this new 142 BACKGROUND PAPERS legislation was being debated, the law produced one of the most dramatic examples of .changed drinking driving behavior resulting from DWI legislation (Ross 1973, 1988). The success of the British Road Safety Act stimulated other nations, such as Canada, the United States, and Australia, to attempt similar programs. The implementation of the system in Canada had a much smaller effect because of limitations on the authority of police officers to require field tests. In Britain, roadside breath tests could be required of anyone in an accident or guilty of a driving infraction. In Canada, the officers could test only when there was cause to believe the individual was impaired. In the United States, the success of the British Road Safety Act increased interest in roadside breath testing of drivers. Implementation of roadside testing was held back because of a challenge to the accuracy of the tube-type testers (Prouty and O'Neilll971) and the question of whether roadside breath tests could be administered without reason to believe that a DWI offense had been committed. The first problem was overcome through the development of miniature, electronic test devices using fuel cell or semi-conductor sensors. By the mid-1970s a small fuel cell test device the "Alco-Sensor"ry became available and was sold to police departments throughout the country. This device permitted roadside breath tests with substantially the same accuracy as could be obtained with the evidential breath test devices in the police station. It also appeared to increase the number of arrests. In Minnesota, in early 1980, the State purchased a large number of roadside breath testers and distributed them to State and local police departments. A time-series analysis performed by Cleary and Rodgers (1986) suggested that this distribution produced a permanent increase in arrests by Minnesota police agencies. The second issue, regarding the authority of the officer to require a preliminary breath test in the absence of probable cause, has not been resolved by the Federal courts. Most police departments use field breath test devices only after the field investigation has been completed and the officer has decided that the driver is impaired and is about to charge him with the DWI offense. The field test device is then used toverifjr the officer's decision and to avoid transporting an individual who later turns out to be below the legal limit to the police station for the evidential test. Rarely, if at all, are these devices used during the second step of the investigation where the officer attempts to determine if the individual has been drinking heavily. Because these devices are not used earlier in the arrest process, many impaired drivers avoid detection because they fail to give thz signs typically observed by police. Field studies (Taubenslag and Taubenslag 1954, Vingilis et al. 1982; Jones and Lund 1985) have demonstrated that police officers miss at least half the impaired drivers with whom they come in contact. This is not surprising, since studies of the ability of physicians to identify drivers with ACs over 0.10 indicate that even they fail to detect half the individuals who would be legally impaired for driving (National Safety Council 1976, page 11). Of all the nations that adapted modifications of this Scandinavian system, the state of New South Wales, Australia has recently made the most rigorous application of what Voas (1982) has labeled the "chemistry-based" enforcement program. In New South Wales and in Tazmania, laws were passed authorizing random breath testing of all drivers using the roads, and the police were provided with funds to establish a policy of vigorous use of sobriety checkpoints. During 1982, the first year of the New South Wales "Staysafe Program," Home1 (1%) reported that nearly 1 million breath tests were made on a driving populationof 3 million, or nearly one in three licensed drivers, and monthly fatalities decreased by an average of 23 percent compared with the previous 6 years. Other data (Ross 1988) suggested that this change has been relatively permanent. As Ross noted, the evidence clearly supports the deterrence theory, since surveys of the driving publicindicate that they are well aware of the law and the police enforcement practices. Moreover, it appears that a fair amount LAW ENFORCEMENT 143 of this information is reaching drivers through their own experience of being tested or through friends who have been tested at checkpoints. The chemistry-based enforcement system implemented in New South Wales, Australia is in sharp contrast to the more behavioral approach used in America. Rather than selecting vehicles from the traffic flow based on aberrant or illegal behavior, the chemistry-based system makes extensive use of roadside checkpoints. A breath test is then conducted on every driver stopped. This makes it unlikely that drinking drivers can drive in a manner that will avoid observation and testing by the police. Once the field test for alcohol indicates that the driver has an illegal alcohol concentration, the individual is charged and taken to the police station for an evidential test. While checkpoints have been used as an occasional feature of the enforcement programs in a number of communities throughout the United States, no jurisdiction has adopted this method as a principal feature of its enforcement activities. More wide- spread use of checkpoints has been constrained by questions regarding the consti- tutionality of the procedure and the manpower required to conduct checkpoints. A series of Federal court decisions (Ifft 1983) have established a "balancing" proce- dure that permits the police to conduct checkpoints under certain highly controlled procedures where the State can demonstrate that this technique is required to protect citizens against the hazards posed by the drunk driver. The procedures required by the court are somewhat limiting. Survey sites must be preselected on the basis of drunk driving incidents and surveyed for safety. A plan must be developed in advance and approved by the highest authority in the police department. Checkpoints must be manned by a number of police officers, with their vehicles, to provide a signifmant "show of force" to reassure drivers that they are not being singled out for investigation. The procedure for selecting vehicles from the traffic flow must prevent individual officer discretion in order to avoid arbitrary or biased selection procedures. Because of these rather elaborate requirements, checkpoints in the United States ' have been relatively expensive operations, Considerable controversy has arisen as to whether they are cost effective. In part, this controversy depends on the objectives of checkpoints. Some police departments hpld that deterrence is accomplished by simply stopping and interviewing a large number of motorists, regardless of the number of arrests made. Other departments stress making DWI arrests in checkpoint operations. Those departments that emphasize driver contacts and the creation of deterrence as the principal role of checkpoints generally employ very brief interviews (lo-l.5 seconds) and only rarely use prearrest breath-testing devices. Such brief interviews make it unlikely that the officer can detect any but the most highly impaired drivers. Other departments conduct somewhat longer interviews (resulting in fewer drivers contacted), but make greater use of prearrest test devices, with a resulting higher arrest rate. Voas, Rhodenizer, and Lynn (1985) demonstrated that a checkpoint can produce more DWI arrests per hour than traditional patrol procedures. The use of prearrest breath-test devices at checkpoints has been limited by the continuing issue as to whether a test can be required without probable cause, or at least "reason-to-believe" that the driver is impaired by alcohol. In an effort to overcome this limitation, passive sensors have been developed (Voas 1983; Jones 1986, Jones and Lund 1985). These handheld units pump mixed environmental and expired air from in front of the driver's face into the sensor and can be made sufticiently sensitive to reliably detect, those individuals who are over the legal limit (Jones and Lund 1985). Legal analysis of these devices (Fields and Henricko 1986) suggested that they are not limited by the provisions of the Fourth Amendment prohibition against warrantless searches and could be used without establishing probable cause that an offense had been committed. This should make it possible for police to use such devices at sobriety checkpoints. When this is done, a drinking driver can be detected in lo-15 seconds (Voas 144 BACKGROUND PAPERS and Layfield 1983). Passive sensors are currently being tested by a number of police departments but the courts have not yet ruled on their constitutionality. - Current Status of DWI Enforcement in the United States Beginning in 1980, a new element entered the DWI enforcement picture. This was the emergence of citizens' activist groups, such as Mothers Against Drunk Driving (MADD) and Remove Intoxicated Drivers (RID). These groups succeeded in calling public attention to the drunk driving problem and in motivating legislators to pass substantial DWI legislation. Most of the legislation dealt with increasing penalties and making them mandatory, or with prohibiting plea bargaining. The general effect of this type of legislation, with respect to enforcement, was to increase the efforts of defendants to avoid conviction, thereby putting increased stress on the quality of the evidence provided by the police officer in court. This increased pressure on the police investiga- tion was counterbalanced somewhat by the adoption of illegal per se laws in 45 States. These laws made it an offense to be at an illegal AC while in control of a vehicle. An illegal per se law reduces the requirement on the police officer to present evidence of impairment, though it does not eliminate it entirely. It is still necessary to show probable cause for administering the evidential breath test in the frst place. In addition, despite the per se law, many courts continue to accept arguments regarding the behavior of the defendant. A second significant element in the new wave of legislation was the passage of "administrative per se" laws that empowered motor vehicle departments to suspend the licenses of drivers not only for refusal to take a chemical test but also for failing a test. Many of these laws permit the police officer to seize the driving permit at the time of arrest and substitute a notice of hearing, which serves as a temporary license. The license is then forwarded to the motor vehicle administrator. The suspension takes place unless the hearing determines that the police officer did not have probable cause to require a chemical test or that the chemical test procedure was faulty. Administrative per se laws provided an additional incentive for the police to make arrests by ensuring that arrests for drunk driving will result in an immediate consequence and that the efforts of the police will not be invalidated by plea bargaining or some other limitation in judicial procedures. At the same time, such laws add somewhat to the paperwork required at the time of arrest. This wave of legislation also produced an increase in the number of States that specifically provided for the use of prearrest breath tests at the roadside by police officers. However, most police forces continue to use these devices as they had before, only at the end of the investigation. Perhaps the most significant effect of this wave of legislation and the public attention given to drunk driving was the fact that it reminded police departments and individual police officers of the extent of public support for rigorous DWI enforcement. This public support also resulted in additional funds for many police departments for DWI enforce- ment and political support to pursue drunk driving arrests more rigorously. Currently, the traditional behavioral system of enforcement (shown in the upper portion of figure 1) remains the primary method of apprehending drinking drivers in the United States. Considerable technology has been applied to thii system since its initia- tion early in the century, particularly in the area of breath testing. However, the system remains basically dependent upon the experience andjudgment of the off&xx in selecting the vehicles to be stopped and identifying drinking drivers, because breath test technol- LAW ENFORCEMENT 145 ogy is not applied until near the end of the investigative process. The chemistry-based system shown in the Iower portion of figure 1 is not used in this country except for partial application in occasional checkpoints conducted in some jurisdictions. Total arrests peaked in 1983 and have decreased slightly since then. The number of arrests seems unlikely to increase unless considerable additional funding becomes available to police departments to augment their traditional behavioral system or to pursue more extensive use of sobriety checkpoints. Evaluation of Enforcement Efforts in the United States While considerable effort has gone into the enforcement of drunk driving laws in the United States, and significant sums have been spent on equipment and overtime pay- ments to special DWI patrols, relatively little rigorous scientific evaluation of these efforts has occurred (Jonah and Wilson 1983). Several factors mitigate against such evaluations. First, most enforcement efforts are implemented as part of a package of DWI legislative programs, making it diflicult to separate the effect of the increased enforcement effort from other changes in the DWI control system. Second, the public appears to accept relatively superficial evaluations and shows little appreciation of the need for rigorous scientific evaluation. Many enforcement programs are evaluated on the basis of a change in the number of DWI arrests. This is a completely inadequate basis for such evaluations, since the estimated arrest rate is 1 for every 1,000 offenses. Doubling such a rate can hardly have much impact in and of itself. Further, changes in the arrest rate are subject to differing interpretations. Increases in arrests are often cited as evidence that the increased enforcement is achieving its goal. Decreases in arrests are sometimes also cited as evidence that the enforcement process is achieving its goal, because (so the reasoning goes) fewer drunk drivers are on the road. Thus, this measure of enforcement effectiveness is completely circular and useless for the purposes of research, except as an intervening variable when an effort is made to determine the actual reduction in alcohol-related crashes. The alcohol-related crash criterion is a difficult one to apply because AC data are principally available only for fatally injured drivers. Other AC data are available only for that small proportion of less severe crashes in which a DWI arrest is made, a clearly biased statistic. To obtain a more objective measure of enforcement impact, crash series, such as single vehicle crashes occurring late at night, are frequently used. Such crashes are more likely to involve a drinking driver than multivehicle crashes occurring during daylight hours. In many communities, relevant crash records are too poorly kept or the numbers of crashes are too few to provide a good basis for evaluating enforcement programs. A better, but much more costly, measure is to use voluntary roadside surveys in which the drivers are asked to voluntarily provide a breath sample for research purposes. These surveys provide a measure of the number of drivers who are impaired during those times when most drinking and driving occurs, an important measure of the impact of an enforcement program. Two studies have shown changes in the average alcohol concentrations of drivers using the roadways as a result of enforcement programs. The first of these, Levy et al. (1978), evaluated the changes in roadside surveys at 19 of the 35 Alcohol Safety Action Projects. They found a statistically significant reduction in the number of drivers with illegal ACs in roadside surveys conducted after the projects were initiated, compared with results obtained before program implementation. These results, however, were undoubtedly influenced by elements of the ASAPs in addition to the increased enforce- ment of DWI laws. A demonstration of impact more specifically traceable to increased enforcement was 146 BACKGROUND PAPERS reported by Voas and Hause (1987) in a study of a special enforcement program in the city of Stockton, CaIifornia. Over a 3-year period, they reported a drop of as much as one-third in the number of drivers above theO.10 legal limit. This drop in,impaired drivers was accompanied by a significant reduction in nighttime crashes, compared to four other similar cities in the central valley of California. The reduced level of alcohol-related crashes was maintained during the 3 years of enforcement activity, but tended to disappear when the special enforcement project came to an end. Several other scientific evaluations of enforcement programs that have found positive results are available in the literature (e.g., Klein 1982; Lacey et al. 1986). Overall, the studies of traditional enforcement programs in the United States have tended to be similar to those covered in Ross' (1984) international review of DWf programs. Short-term reductions in drinking driving crashes were obtained in some cases, particularly where enforcement was accompanied by considerable publicity. However, the changes tended to be transitory, maybe because the police failed to fully utilize the Rowers provided to them by the law (as in the case of the British Road Safety Act) or because, after an initial intensification, enforcement efforts returned to previous levels (as in Stockton). Finally, as Borkenstein (1975) hypothesized, it may be necessary to keep changing enforcement procedures to make them "new" and newsworthy and to attract the attention of the public. Few chemistry-based checkpoint systems in the United States have been scientifically evaluated. Voas, Rhodenizer, and Lynn (1985) reported on a year-long enforcement program in which checkpoints were implemented every weekend within the city of Charlottesville, Virginia. Their evaluation indicated that the police apprehended fox DWl approximately 1 percent of the drivers stopped at the checkpoint. In addition, another 1 percent of the drivers were arrested for driving without a license. Random digit dialing telephone surveys indicated that approximately one-fourth of the nighttime drivers in Charlottesville came into direct contact with a checkpoint and that more than 90 percent of all drivers were aware of the checkpoint program. Comparison of the nighttime and alcohol-related crash rates in Charlottesville with those of a similar community that did not employ checkpoints revealed that this procedure reduced such crashes by approximately 15 percent. A similar reduction was apparent when crash rates for Charlottesville were compared with those for the State of Virginia as a whole. Additional evidence for the impact of checkpoint procedures was obtained by Wil- liams and Lund (19&J), who conducted a random digit dialing survey and compared the attitudes and knowledge of the driving public in those communities that used checkpoints with communities that did not use checkpoints. Where checkpoints were used, drivers reported higher levels of deterrence to drinking and driving than citizens of counties where they were not used. These studies provide some evidence for the effectiveness of the chemistry-based enforcement system. A full evaluation of the chemistry-based enforcement system awaits an adequate application of this technique in the United States. Eight Issues for Future Enforcement Programs This summary of the status of DWl enforcement in the United States suggests that it is having a significant general (but unmeasured) impact on deterrence to drunk driving. However, little additional effectiveness can be expected unless new resources are committed to, or new technology and procedures are employed in, the enforcement effort. Among the issues that are currently being debated and the proposals for new enforcement methodology being considered, the following 10 items should provide the subject matter for recommendations to be made in the Surgeon General's Report. LAW ENFORCEMENT Issue I: Increasing the use of sobriety checkpoints 147 As noted, the Federal courts have provided for the use of sobriety checkpoints under certain constraints (Ifft 1983); however, the legality of thii procedure under the Federal constitution does not necessarily mean that it meets the requirements of each of the 50 State constitutions. As of this date, the Supreme Courts of 18 of the 50 States have made favorable decisions regarding checkpoints, while the Supreme Courts of 9 other States have made unfavorable decisions. These unfavorable decisions, however, frequently resulted from the consideration of checkpoint programs that did not meet Federal guidelines. Where there is full compliance with the Federal guidelines, it is probable that most States will find that checkpoint procedures meet the appropriate constitutional tests. Perhaps more significant than constitutional issues is the acceptability to the public of checkpoint programs. Police administrators tend to be highly sensitive to public j opinion. While police departments are often interested in new and novel procedures, police organizations tend to be basically conservative. Most public surveys show sig- nificant support for use of the checkpoint procedure. Voas, Rhodenizer, and Lynn (1985) found that the public in Charlottesville, where checkpoints were regularly con- ducted, were more in favor of checkpoints than the public in the comparison community which had not experienced a checkpoint program. Most available evidence suggests that police departments will be supported by the public if they implement checkpoint programs. Nevertheless, the concern with public relations remains a major drawback to checkpoint programs in the minds of many police administrators. A third problem in mounting checkpoint operations is the issue of cost effectiveness. The Federal court specifications for checkpoint operations require the assembly of a number of police vehicles and the use of a minimum of four to six officers. (In contrast, a checkpoint can be conducted in New South Wales, Australia by a single officer). This requirement for a relatively large force presents personnel and cost problems for many localities. Some jurisdictions have addressed this problem by combining resources from State, county, and local police. Others have used overtime or diverted officers from other duties. Depending on the procedures used and the policies implemented, a checkpoint may result in relatively few arrests or, alternatively, in more arrests per man hour than would be achieved in an equal amount of traditional enforcement activity. In any case, the impact of a checkpoint should not be assessed on the basis of the number of arrests produced. As demonstrated by Williams and Lund (19&t), the impact of sobriety checkpoints on the general driving population is more important in creating deterrence than the number of arrests made by traditional enforcement procedures. States and communities could be encouraged to promote checkpoint operations. This could be done both by influencing police policy and by providing additional funds for checkpoint equipment and operations. At issue is whether this procedure, which is a basic part of the chemistry-based enforcement system, would be cost effective in increas- ing deterrence to drunk driving. Evidence from Scandinavia and Australia suggests that it may be the most cost-effective procedure. Issue 2: Using portable breath tests earlier in the DW enforcement process Portable breath test devices, about the size of a cigarette package, have been available to the police for over a decade. Twenty-six of the 50 States have passed legislation specifically authorizing their use. However, they are rarely used early in the investigative procedure. This occurs partly because of the general assumption made by police departments that the preliminary breath tester (PBT) cannot be used prior to obtaining 146 BACKGROUND PAPERS reason to believe that the individual has been driving while impaired. Thus, the devices are used only after the officer has completed his investigation. They are used only to verify the officer's decision, with the result that many over-the-limit drivers who did not appear to be intoxicated escape detection. The prearrest breath test is clearly a search in'the constitutional sense because the suspect is required to blow through a mouthpiece and provide an active breath sample. The Federal courts, however, have been willing to accept a compromise or a balancing test when the need of the State to protect its citizens is sufficiently great and the intrusion provided by the search is sufficiently smaIl to make the search reasonable. Some constitutional experts believe that the act of blowing into a mouthpiece is such a smaIl intrusion that it would meet this test. They predict that the Federal courts would find it acceptable to administer a breath test to motorists in accidents or to motorists guilty of driving offenses without requiring specific evidence that they are imp,aired by alcohol. Should the courts find that the PBT does not meet constitutional standards without probable cause to suspect a DWI offense, then the passive sensing technology is available for use early in the arrest process. These devices, while somewhat less accurate than the PBT, would almost certainly pass constitutional tests since most experts agree that the passive sensor does not involve a search within the meaning of the Fourth Amendment (Fields and Henrico 1986). The use of such sensors on all individuals stopped at sobriety checkpoints or all individuals stopped for speeding or other traffic infractions would result in a significant increase in DWI arrests, since current evidence indicates that police engaged in these enforcement activities fail to detect half or more of the intoxicated drivers with whom they come in contact. States and localities could encourage the use of PBTs and passive sensors through their influence on police policies and by providing funding for the purchase of this type of equipment. The important issue, however, is not providing additionaI equipment, but more significantly, encouraging police ofticers to use this technology at the beginning, rather than at the end, of their DWI investigation procedure. Issue 3: Expanding DW7 enforcement through new legislation Borkenstein (1975) noted that a typical community with a population of 1 milIion will have 1,000 patrol officers. This same hypothetical city, would have approximateIy325,OOO hazardous moving violations per year. He proposed that, to increase deterrence to impaired driving, every driver stopped for such a hazardous moving offense should be tested for alcohol. If such a driver were found to have an AC over a minimum level (e.g., O.OS), he would be given a special aggrevated-by-alcohol traffic citation. The offense of speeding might carry a line of $20.00 or $30.00, but the offense of speeding while aggrevated-by-alcohol would carry a higher fine (e.g., $50.00) and would also result in a notation on the driving record that an alcohol-related offense had occurred. Drivers with ACs over the per se limit could also be charged with the DWI offense. The principal issue that arises in this procedure is the Fourth Amendment limitation on conducting searches without probable cause. The proposal to test all individuals guilty of serious driving infractions is similar to the British Road Safety Act of 1%7. Enforce- ment would require court acceptance of a driving infraction as meeting the requirements for permitting a "search" such as the use of a prearrest breath test device. If the courts determined that such a search was not permitted without specific evidence that the individual was impaired, passive sensors could be used to obtain evidence of drinking, followed by the use of a PBT. This procedure would likely result in a significant increase in the number of drivers arrested for DWI and an increase in the number of drivers receiving an alcohol-related offense citation on their driving records. Another type of program directed at increasing the number of drinking drivers apprehended is the "Roadside License Suspension" program used in several Canadian LAW ENFORCEMENT 149 provinces. This legislation has two forms. In the first, as practiced in New Brunswick, Ontario, Manitoba, and Saskatchewan, the driver can be required to take a roadside breath test if the officer has reason to believe the driver is impaired. If the result is over 0.05, the officer can suspend the individual's driving permit for periods varying from 6 to 24 hours. Another approach, used in Alberta, British Columbia, and the Yukon Territories, places the testing decision on the driver. Relevant to this approach, the British Columbia Motor Vehicle Act (Section 214) states that (a) if a police officer believes a driver's ability is affected by alcohol (or other drug), he may request the driver to surrender his license; and (b) the period of suspension is 24 hours unless the driver voluntarily submits to a test that determines an AC not exceeding 50 mg percent. Thus, the suspected offender has the choice of submitting to the test to demonstrate that he is not over the 0.05 limit or of surrendering his driving license to the police officer, finding other means to get home, and returning to the police station the next morning to retrieve his driving license. One potential problem with such lesser offenses is that they can be used inap- propriately to avoid the paperwork and hassle involved in the prosecution of more serious drunk driving offenses. If the offense of a traffic violation aggrevated-by-alcohol were created with lower penalty levels, it might well be used as an opportunity for plea bargaining with individuals apprehended with illegal ACs being allowed to plea down to this lesser offense. Police oflicers might also use the short-term suspensions to avoid the paperwork involved in bringing more serious charges in the first place. The basic issue is whether the enactment of such a law would increase the deterrence of drunk driving. Issue 4: Expanding oficerparticipation in alcohol enforcement As noted earlier, the ASAP program popularized the use of special, dedicated DWI enforcement teams. This system was intended to stimulate the apprehension of impaired drivers by all members of the traffic patrol by offering the opportunity to earn overtime pay on the special patrols. Sometimes this opportunity was extended only to those officers who achieved a high arrest rate on their normal duty hours. To a certain extent, this general concept worked as intended. However, in the long run, it tended to establish a policy of allowing a few highly motivated officers to specialize in DWI, while permitting the large majority of traffic officers to make few, if any, DWI arrests. The fact that a two- or three-man special enforcement team active on Friday and Saturday nights could make as many arrests as the rest of the traffic department was more an indication of the lack of attention to drunk driving by the average officer on patrol than an indication of the level of skill of the special patrol members. It will be difficult to greatly increase arrest rates as long as the pursuit of the drunk driver is seen as a specialist activity for a few officers. Expanding the role of the rest of the traffic department in DWI enforcement activities would not only increase the number of arrests but would also broaden the impact on the driver, since individuals would be investigated for impairment, not only on Friday and Saturday nights, but throughout the week. States and localities should be able to increase the number of arrests by the regular traffic patrols through training and by providing the officers with prearrest breath testers or passive sensors; With a passive sensor backed up by an active preliminary breath tester to use in the field, the traffic officer need not be an expert in conducting sobriety tests or in detecting evidence of impairment from the appearance of the driver. Using these devices, he can identify individuals who are over the limit and bring them in for evidential tests. With police management emphasizing the importance of drunk driving enforcement, 150 BACKGROUND PAPERS all officers who are trained and equipped with passive sensors and PBTs should be able to contribute substantially to the drinking driving enforcement effort within a com- munity. The principal issue here is whether the State and locality should fund police departments to train officers, purchase passive sensors and PBTs, and encourage police administrators to make DWI enforcement a high-priority activity for all offtcers engaged in traffic patrol activities. Issue 5: Lowering AC limits for private vehicles Of the 50 States, 41 currently have per se limits at 0.10, while 2 others have 0.08 limits. The National Safety Council Committee on Alcohol and Drugs has taken the position that all individuals at 0.08 AC are impaired. The American Medical Association has gone even further, recommending that the National limit be 0.05 AC. Considerable contro- versy exists regarding the desirability of lowering the AC limit from O.iO, which is the current standard in the United States. The fast issue is whether, in fact, ACs below this level increase the probability of accident involvement. The second issue is whether lowering the illegal AC level reduces the number of high AC drivers on the road or the number of alcohol-related accidents. The third issue is whether lower AC levels can be enforced, and, if so, what such enforcement costs. With regard to the first issue, laboratory and epidemiological studies have indicated that increases in impairment and crash risk begin at low AC levels (Moskowitz and Robinson 1988). Hurst (1973) found that no matter how experienced with alcohol, any individual is at higher risk for involvement in a crash at any AC level over zero. This provides a basis for arguing that the AC limit should be zero, since any level above that will increase the probability of crash involvement. On the other hand, an attempt to eliminate all driving with any positive AC is beyond the resources and capability of the criminal justice system, as shown by the Nation's experience with prohibition. With respect to the second issue, no adequate scientific studies demonstrate the effectiveness of lower AC levels per se. Data from Scandinavia (where the AC limit is 0.05) indicate that the number of high AC drivers on the roadways is clearly lower than in the United States, Canada, and the Netherlands, where the AC limits are 0.10,0.08, and 0.08, respectively (Snortum 1984). However, many other differences between these nations could contribute to these differences, and it is not possible to determine the relative role of the AC limit compared to the differences in enforcement policy and procedure, the penalties for the offense, and the general cultural attitudes toward alcohol use and drinking and driving. Overall, no reliable research evidence clearly demonstrates that lowering the AC alone produces a reduction in alcohol-related accidents. Rarely in real lie is a single countermeasure feature implemented so that it can be evaluated on its own without the interacting effect of other changes in the law or in enforcement. A change in the AC level would probably also be accompanied by a change in enforcement, since enforcing lower AC limits may well require different enforcement techniques. Relative to the issue of enforcing a lower AC limit, some police forces are currently experiencing considerable success in apprehending drivers at lower ACs with traditional patrols. In North Carolina, approximately 10 to 15 percent of all drivers arrested are below 0.10. A significant number of arrests below the 0.10 level are also made by other police departments, such as in the District of Columbia. A serendipitous impact of lowering the AC limit is to increase the probability of conviction for those at 0.10 and higher, since many prosecutors provide for a buffer zone at the legal limit, whatever it may be. In many cities, for example, the prosecutors operating under the current 0.10 law only move those cases where the measured AC is 0.12 or 0.13 and above, because of the difficulty in obtaining convictions where the AC is near the legal limit. Moving the LAW ENFORCEMENT 151 legal AC limit to 0.08 could lower this buffer zone and increase convictions for those with 0.10 ACs. If the legal limit is reduced to 0.05 or 0.08, greater reliance will probably have to be placed on the chemistry-based system shown in the lower portion of figure 1. The use of checkpoints and passive sensors will likely be increased, since the individuals who are at these levels are less likely to present the signs of impairment on which the police normally rely. Whatever final decision is made regarding lowering the AC limit, it is important to repeal the section of many State laws that provides that an AC below 0.05 is presumptive evidence that the individual isnor intoxicated. More recent studies of the effect of alcohol on performance (Moskowitz and Robinson 1988) clearly demonstrated that some individuals are impaired below that level. Further, a number of individuals with low ACs may also have consumed drugs so that the combination of alcohol and drugs produce an impairing effect. This provision could make an AC below 0.05 an obstacle to prosecution of these cases. Further, a number of States have established zero AC limits in which any measurable AC (generally 0.01 to 0.02) is an offense for teenage drivers who are not permitted, under law, to drink. The presumption that an individual below 0.05 is not impaired is in clear conflict with this type of legislation. Issue 6: Suspending driverk licenses Historically, the suspension of the driving privilege has been the most salient penalty for a conviction of driving-while-impaired. The laws of the States varied in their proce- dures for administering the license suspension penalty. In some cases, this became a province of the court with the court seizing the license and, perhaps, substituting a limited driving permit. Usually, the court would forward an order to the Motor Vehicle Admin- istrator to suspend the license. Because the driving public feared the loss of license and saw that as a significant penalty, it became the basis for considerable plea bargaining, to such an extent that in the early 197Os, license suspension occurred only irregularly, and many defendants got off with only fines. Studies conducted principally in the States of California, Washington, and North Carolina (Popkin et al. 1983; Salzberg and Klingberg 1983; Sadler and Perrine 1984; Peck et al. 1985) demonstrated that while as many as one-half to two-thirds of those who receive suspensions continue to drive, suspended drivers were involved in fewer total accidents and fewer non-alcohol related accidents than individuals who retained their licenses in return for attending education or treatment programs. Partly as a result of this evidence and partly as a result of the public concern and attention to the DWI problem stimulated by citizen activist groups such as MADD and RID, administrative suspension laws, which bypassed the courts, were enacted during the late 1970s and early 1980s. The lead was taken by the State of Minnesota, which enacted so-called administrative per se legislation in 1976. This legislation operated in conjunction with the implied consent law, so that the driver was not only to lose his license if he refused a chemical test, but also if he failed such a test. This administrative withdrawal procedure was adjudicated to be constitutional, provided the offender had the opportunity to have a hearing to determine that there was probable cause for his arrest and that the chemical test had been properly conducted (Reeder 1981). An important feature of a number of these laws was that the police officer was allowed to pick up the license on the spot, upon either refusal or failure of the test, and replace the license with a hearing notice. The license was then returned to the Motor Vehicle Administrator by the police officer. In this way, an initial penalty for drunk driving was administered on the spot, a feature that could enhance the deterrence to DWI by reducing the time between offense and punishment and that has been shown to increase the motivation of police officers in some States. 152 BACKGROUND PAPERS Zador et al. (1988) conducted a study of the impact of per se legislation and the administrative per se laws and found evidence that the administrative per se law contributed to some of the national reduction in alcohol-related accidents that occurred between 1982 and 1986. The issue of the effectiveness of administrative license revoca- tion as a penalty lies beyond the scope of this paper. However, this law is significant to the DWI enforcement effort in that it places an additional requirement on police officers but, also, provides them with a potentially motivating element in that they can be assured that the offender is receiving a significant penalty. Too often iu the past, when the application of sanctions has been dependent upon litigation, plea bargaining, and other delaying tactics, the offender has received minor punishment or none at all. Currently, 23 States have administrative per se laws. The success of this procedure suggests that it would be desirable for the remaining States to enact similar legislation. Issue 7: Enforcing driver's license suspensions Suspension or revocation of the driver's license is considered to be the single most effective DWI sanction for reducing subsequent traffic offenses and accidents. However, this sanction comes under attack from practitioners and citizens groups alike because of compelling evidence that a majority of DWI offenders continue to drive to some degree during the period of their license revocation or suspension (Sadler and Perrine 1984). Despite these indications that many offenders continue to drive while under suspen- sion or revocation, there is evidence that those under suspension, as a group, have significantly lower rates of rearrest for DWI and of crash involvement (Popkin et al. 1983; Peck et al. 1985). However, those who do continue to drive (even if they are driving more safely than otherwise) are flaunting the sanction imposed on them and should be arrested and punished. The problem has been that driving with a suspended or revoked license has been a relatively invisible offense. In other words, it generally does not come to a police officer's attention unless some other violation of the law is detected. Even when it is detected and a citation for driving without a license is issued, the court frequently does not convict because of the inability of the prosecutor to demonstrate that the driver received legal notice of his or her suspension. This occurs because the notices frequently go through the mail, and there is no acceptable evidence that they were received. The State of Virginia, among others, has attempted to use police in surveillance of the residences of suspended drivers in an effort to apprehend those who continue to drive. This procedure has yet to be adequately evaluated, but obviously involves fairly high costs in police manhours. Another traditional approach to dealing with this problem is to increase the penalties for driving while suspended by providing for vehicle impound- ment or jail time for suspended drivers. The effectiveness of these more salient penalties procedures is unknown, but their deterrent effect is most probably highly dependent upon the effectiveness of the enforcement system in apprehending suspended drivers. In urban areas with large numbers of automobiles, there is simply no way in which the police can, without some technological assistance, determine whether the individual operating a vehicle is properly licensed. One aid employed by the State of Minnesota and in certain other jurisdictions, such as New Philadelphia, Ohio, is to confiscate the vehicle tags of the drivers who are convicted of driving without a license or convicted of DWI and replace them with distinctive plates that call attention to the vehicle. This provides a means for identifying those vehicles that may be driven by a suspended driver. Several States are currently placing more emphasis on this approach. A more technological approach to the identification of vehicles driven by suspended drivers is the use of the so-called TAGS system, which has been evaluated by the Insurance Institute for Highway Safety (Miller 1978). In that system, police officers are provided with a keyboard on which to enter vehicle tag numbers at random. These LAW ENFORCEMENT 153 numbers are transmitted to a central data file that checks to determine if the tag belongs to a vehicle that has been stolen, if the driver is wanted for other criminal offenses, or if the vehicle is owned and driven by a suspended driver. If a match occurs, a signal is sent to the patrol vehicle so that it can stop the car and interview the suspected driver. Using this TAGS system in Maryland, Miller found that 9.6 offenders were identified per officer hour compared to only .5 offenders per officer hour using traditional patrol methods. The sobriety checkpoint also offers a method of enforcing the laws-against driving while suspended. In the CharIottesviIle checkpoint program (Voas et al. 1985), 1 percent of the drivers stopped were given citations for driving without a proper license, equal to the number arrested for DWI. (For a full discussion of the technical problems in enforcing license suspension, see Voas 1988). It is clear that if suspension of driver's licenses is to be the principal penalty for drunk driving, and if a large number of drivers are to be arrested each year, it will be important to enforce this restriction effectively. Approximately one-half of the drivers suspended do not reapply for licenses when they are eligible, apparently because of the high cost of automobile insurance to offenders. Therefore, between 500,030 and 1 million drivers come off the rolls of State driving license registers each year, but these dangerous individuals continue to drive. This places them outside the normallicense control system. An efficient enforcement procedure that can deal with this problem needs to be developed. Issue 8: Managhg license penalties automatically One function of sanctions is to incapacitate the offender and prevent a repetition of the offense by making it impossible to commit the same crime. Incarceration is the classic method for ensuring that an offender will not repeat his offense, at least during the prison term. In DWI adjudication, jail terms are generally far too short to have any significant effect through incapacitation. The offender is soon released and able to operate his automobile. Suspension of the driving license is intended to continue the incapacitation for a significant period, usually several months to a year or more. However, this type of incapacitation is only partially effective.because it is difficult to enforce. A recently developed alternative to traditional enforcement methods, which will incapacitate the individual from repeating his offense, is the alcohol safety interlock. The concept for a device that would be mounted on the car and test the operator's performance or AC was first proposed in the Secretary of Transportation's report, AIcohoI and Highway Safety, in 1968 (U.S. DOT, 1968). In 1970, Voas reviewed this concept, describing the oppor- tunities and the problems posed by what he dubbed as "Alcohol Safety Interlock Systems" (Voas 1970) The idea of an in-vehicle system that can determine the impairment of the driver and prevent vehicle operation is such a parsimonious and attractive approach to the solution of the DWI problem that this concept has long enjoyed considerable support among safety specialists and politicians. As a result, the Federal Government undertook a decade of research directed at developing an interlock system (Compton 1988). This research was primarily directed at using performance tests as a method for identifying the impaired driver. This is attractive because performance tests can detect drivers impaired by drugs as well as alcohol. However, the first commercially developed devices have all been based on the measurement of breath alcohol using simple semiconductor sensors. Currently, at least 10 States have passed legislation authorizing the testing of these devices, and individual courts in a number of other States have established demonstra- tion programs. To this date, however, there has not been a sufficient number of these devices in the field to provide an adequate scientific test of their effectiveness. Slightly over 200 units are currently authorized in two experimental counties in California in an 154 BACKGROUND PAPERS evaluation program that is being supervised by the Office of Highway Safety of the State of California. In another year or two, this program should provide scientific evaluation of these devices. While these devices control only the driving in the vehicles in which they are placed and therefore do not control driving by offenders in other vehicles, they offer the potential for taking over much of the supervision problem. Since the systems are predicated upon the offender paying for their cost and the monitoring being done by .commercial companies, these systems relieve the State of considerable expense. They will free the police to use their time in enforcing basic drinking driving and other hazardous driving laws. The principal concern of the highway safety community is that these devices will be implemented widely before they are fully evaluated and will be substituted by the courts for the full suspension penalty, which has been proven to be effective. Conclusion The issues listed above provide an important agenda for consideration in action programs and research studies. The first half of this decade has brought considerable progress in the reduction of alcohol-related accidents as a portion of total fatal accidents. It has provided the Erst evidence in history that it is possible to ameliorate the drinking driving problem through the criminal justice system. The recent leveling off of DWI arrests and alcohol-related fatalities suggest that new initiatives are needed if the progress seen during the first half of thii decade is to be continued. REFERENCES Andenaes, J. The Scandinavian Experience. In: Laurence, M.D.; Snortum, J.R.; and Zimring, F-E., eds. Social C&r01 of the Drinking Dn'ver. Chicago: University of Chicago Press, 1988. pp. 43-63. Beitel, GA.; Sharp, M.C.; and Glauz, W.D. Probability of arrest while driving under the influence of alcohol. JournaZofStudies on Alcohol 36(1):109-l 15, 1975. Borkenstein, R.F. Problems of enforcement, adjudication and sanctioning. In: Israelstam, S., and Lambert, S., eds. Alcoholhugs and Trajjic Safety. Proceedings of the Sixth International Conference on Alcohol, Drugs and Traffic Safety. Toronto, Ontario: Addiction Research Foundation of Ontario, 197.5. Cleat-y, J., and Rodgers, A. Analysis of the Effects of Recent Changes in Minnesota's DWI Laws: Pan III Longitudinal Analysis of Policy Impacts. St. Paul, MN: Research Department, Minnesota House of Representatives, 1986. Compton, R.P. Potential for Application of Ignition Interlock Desices to Prohibit Operation of Motor Vehicles by Intoxicated Individuals. Washington, DC: National Highway Traffic Safety Administration., 1988. Department of Transportation. AZcohol and Highway Safety. Report to the U.S. Congress. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1968. Fields, M., and Henricko, AR. Passive Alcohol Sensors-Constitutional Implications. The Prosecutor 20( 1):45-50, 1986. Harris, D.H.; Hewlett, J.B.; and Ridgeway, R-G., fir& Defection of&kg while InfoaiCUfed. Project Interim Report: Identification of V&al Cues and Development of Detection Methods. NHTSA Report No. DOT HS 805-051. Washington, DC: NHTSA, 1979. Available from NTIS, Springfield, VA 22161 Hause, J.; Voas, R.; and Chavez, E. Conducting voluntary roadside surveys: The Stockton experience. In: Valverius, M-R., ed. Proceedings of the Satellire Conference to the 8th International Conference on Alcohol Drug, and Trafic Safety, 104-13. Stockholm: The Swedish Council for Information on Alcohol and Other Drugs, 1982. LAW ENFORCEMENT 155 Homei, R. Policing the Drinking Driver: Random Breath Testing Md the Process of Deterrence. Sydney, Australia: Federal Office of Road Safety, 1986. Howland, J. Social norms and drunk driving countermeasures. In: Graham, Ed. Preventing Automobile Injwy; New Findings from Evaluation Research, Dover, MAz Auburn House, 1988. pp. 163-180. Hurst, P. Epidemiological aspects of alcohol in driver crashes and citations. Journal of Sufeq Research 5(3(:130-148,1973. Ifft, R.k Curbing the drunk driver under the fourth amendment: The constitutionality of roadblock seizures. Georgetown Law Journal 71:1457X& 1983. Jonah, B.k, and Wilson, RJ. Improving the effectivenessof drinkingdrivingenforcement through increased efficiency. Accident Analysis & Prevention. 15(6):463-482,1983. Jones, I.S. The Development and Evaluation of a Passive Alcohol Sensor. Washington, DC: Insurance Institute for Highway Safety, 1986. Jones, IS., and Lund, AK. Detection of alcohol-impaired drivers using passive alcohol sensor. Journal of Police Science and Administration 145(2): 153-f&),1985. Klein, T. The Effect of Changes in DWI Legislation in the State of Mqband. NHTSA Technical Report. Washington, DC: NHTSA, 1982. Laccy, J.H.; Steward, J.R.; Marchetti, L.M.; Popkin, C.L.; Murphy, P.V.; Lucke, R.E.; and Jones, R.K. Enforcement and Public Infotmution Strategies foT DWI General Deterrence: Arrest DrunkDriving-The Clear-water and Large, Florida Eqerience. NHTSA Report No. DOT HS 807066. Washington, DC: NHTSA, 1986. Laurence, M.D. The legal context in the United States. In: Laurence, M.D.; Snorturn, J.R.; and Zimring, F.E., eds.Social Control of the Drinking Driver. Chicago: University of Chicago Press, 1988. pp. 136-168. Levy, P.; Voas, R.B.; Johnson, P.; and Klein T. Evaluation of the AsAPs. JoumalSafety Research 10(4):162-176, 1978. Miller, G. Summary Report on Project TAGS-An Eqxriment in Mass Screening of License Plates to Zdentiji, Motor Vehicle Law Violators. Washington, DC: Insurance Institute for Highway Safety, 1978. Moskowitz, H., and C. Robinson. Effects of Low Doses of Alcohol on Driving Reiated Sk&: A Review of the Evidence. NHTSA Report, DOT HS 807-280. Washington, DC: NHTSA, 1988. National Highway Traffic Safety Administration. Summury of Nutionul AZcohoZ Safety Action Frojecfs. Washington, DC: Supt. of Dots., US Govt. Print. Off., 1979. National Safety Council. Alcoholand the Impaired Driver. Chicago: National Safety Council, 1976. Oates, J.F., Jr. Factors InfluencingArrestsforAkohol-Related Traffic ~olations. National Highway Traffic Safety Administration, Technical Report, DOT-HS-801~230. Washington, DC: NHTSA, 1974. Peck, R.C.; Sadler, D.D.; and Perrine, M.W., The Comparative Effectiveness of Alcohol Rehabilitation and Licensing Control Actions for Drunk Driving Offenders: A Review of the Literature.Alcohol, Drugs, and Drivtitg;Abstracts and Reviews 1(4);15-40, 1985. Popkin, C.L.; Li, L.R; Iacey, J.H; Stewart, J.R.; and Wailer, P.F.AnInitiuZEvaluation oftheNorth Carolina Alcohol and Drug Education Traffic SchooLs. Volume 1: Technical Report. Chapel Hill, NC: University of North Carolina Highway Safety Research Center, 1983. Prouty, R.W., and O'Neill, B. An Evaluation of Some Quantitative Breath Screening Tests for Alcohol. Technical Report. Washington, DC: Insurance Institute for Highway Safety, 1971. Reeder, R.H. Analytical Study of the Legal and OperationalAspects of the Minnesota Law Entitled "Chemical Test for Intoxication. " NHTSA Report No. DOT-HS-806-170. Washington, DC:NHTSA, 1981. Ross, H.L. Law, science and accidents: The British Road Safety Act of 1%7. The Journalof Legal Studies 2(1):1-78, 1973. Ross, H.L. The Scandinavian myth: The effectiveness of drinking-and-driving legislation in Sweden and Norway. The Journal of Legal Studies 4(2):285-310,1975. Ross, H.L. Deterring the Drinking Driver: Legal Policy and Social Control. Lexington, MA: Lexington Books, 1984. Ross, H.L. Deterrence-based policies in Britain, Canada and Australia. In: Lawrence, M.D.; Snortum, J.R., and Zimring, F.E., eds. Social Control of the Drinking Driver. Chicago: University of Chicago Press, 1988. pp. 64-78. 156 Sadler, D.D., and Perrine, M.W. An EvaIuation of the California Dmnk Driving Countermeasure System: Volwne 2 The Long-Term Trajjic Safep Impact of a PilotAlcoholAblcse Treatment as an Alternarive to License Suspensions. Report No. 90. Sacramento, CA: California Department of Motor Vehicles., 1984. Salzberg, P.M.; and Kiingberg, C.L The effectiveness of deferred prosecution for driving while intoxicated. Journal of Studies on AkohoI 44:299-306,1983. Snortum, J.R. Alcohol impaired driving in Norway and Sweden: Another look at "the Scadinavian Myth." Law and Policy 6(1):5-37,1984. Taubenslag, W.N., and Taubenslag, MJ. Selective Trafic Enforcement Program (STEP): Fort Lauderdale, Pasco Services, Inc. Final Report. Washington, DC: NHTSA, 1975. Tharp, V.; Bums, M.; and Moskowitz, H. Development of Field Test of Psychophysical Test for DWArrest. Technical Report no. DOT-HS805-864. Washington, DC: NHTSA, 1981. Vingilis, E.R.; Adlaf, E.M.; and Chung, L. Comparison of age and sex characteristics of police- suspected impaired drivers and roadside-surveyed impaired drivers.AccidenrAnufyrfi and Prevention 14:4X-30,1982 Voas, R.B. "CarsThat Drunks Can't Drive." Paper presented at the annual meeting of the Human Factors Society, San Francisco, 1970. Voas, R.B. Selective enforcement during prime-time drinking-driving hours: A proposal for increasing deterrence without increasing enforcement costs. Abstracts and Reviews in Alcohol and Driving 4(3):3-21, 1982. Voas, R.B. Laboratory and field tests of a passive alcohol sensing system. Abstracts and Reviews in Alcohol and Driving 4(3):3-Z& 1983. Voas, R.B. Emerging technologies for controlling the drunk driver. In: Laurence, M.D., Snortum, J.R.; Zimring, F.E., eds. Social Control of the Drinking Driver. Chicago: University of Chicago Press, 1988. Voas, R.B., and Hause, J.M. Deterring the drunken driver: the Stockton experience. Accident Analysis and Prevention 19(2):81-90, 1987. Voas, R.B., and Layfield, WA. Creating general deterrence: Can passive sensing help? The Police Chief 50:56-61, 1983. Voas, R.B.; Rhodenizer, AE.; and Lynn, C. Evaluation of Char-lottesvilZe Checkpoint Operations. Technical Report, DOT Contract DTNH-22-83-C-05088. Washington, DC NHTSA, 1985. Williams, A-F., and Lund, A-K Deterrent effect of roadblocks on drinking and driving- Traffic Safeq Evaluation Research Review 3:7-18,1984. Zador, P.R.; Lund, A-K-; Fields, M.; and Weinberg, K.; Alcohol-Impaired Driving Luws and Fatal Crash Involvement. Washington, DC: Insurance Institute for Highway Safety, 1988. 157 Transportation and Alcohol Service Policies Transportation Alternatives for Drinkers Robert Apsler, Ph.D. Department of Psychiatry, Harvard Medical School Drunk driving persists at stubbornly high rates despite continuing efforts to reduce its occurrence (Reed 1981; NHTSA 1985; National Commission Against Drunk Driving 1987). One of the most compelling explanations for this phenomenon is the observation that alcohol and automobiles have become such integral features of our society that drunk driving is virtually inevitable (Gusfield 1981; Mosher 1985; Ross 1987). If substan- tial reductions in drunk driving are to occur, dramatic changes must take place. Either drivers must sharply curtail their drinking, or ways must be found to stop intoxicated individuals from driving. This chapter focuses on a class of prevention strategies that take the latter approach. These strategies share a philosophy of attempting to provide drinkers with safe transpor- tation while requiring as little modification as possible in drinking practices. There is, of course, nothing new about informal efforts to find safe transportation for intoxicated individuals. Surely, efforts to help intoxicated individuals get home safely began long before the invention of the automobile. Today, incidents of hosts, fellow drinkers, sober associates, and police assisting intoxicated individuals in obtaining safe transportation have become part of our drinking lore, even though they have received little study. In contrast to these informal interventions, formal programs designed either to provide safe transport for individuals or to encourage informal actions appear to have originated only in the last few years. Unfortunately, they too have received almost no attention from researchers. These efforts to implement transportation alternatives for intoxicated drivers can be divided into two groups: (1) those in which the individuals participating in the drinking activity supply both vehicles and drivers- the designated driver tactic, and (2) those where neither vehicles nor drivers are typically provided by the individuals taking part in the drinking activity- the safe rides tactic. Designated Drivers With a simple and inexpensive tactic, groups of drinkers can assure themselves that a sober driver will be available when needed. Before drinking commences, they deter- mine the number of drivers necessary to transport the entire group. Then, that number of individuals in the group remain sober and drive all of the others. This tactic, the use of "designated drivers," has been championed by a diverse array of sources and is currently receiving wide dissemination through the mass media. The tactic of drinkers designating individuals to remain sober and do the driving has great appeal. First, the tactic can be used in any settingwhere people drive together after drinking. For example, it can be employed in private homes, bars, sporting events, and restaurants. Second, underage youths can adopt the tactic without requiring the authority or approval of adults. Third, no cost need be associated with exercising the tactic. In fact, designated drivers save money by drinking nonalcoholic beverages. Even 158 BACKGROUND PAPERS when incentives are offered to designated drivers by drinking establishments, the cost of the incentives tends to be inconsequential. Fourth, if light drinkers or abstainers are unavailable, only the designated drivers need change their drinking behavior for the tactic to be successful. Other group members are free to drink in whatever way they choose. Surprisingly, virtually no research has been done on the designated driver tactic. Nevertheless, even the sparse information that does exist reveals both the great potential and some important limitations of the tactic. The potential for widespread use of the designated driver tactic is apparent from the results of a 1987 Gallop Poll (Gallup 1987). Nearly all Americans (91 percent) who participated in social events where alcohol was available wanted the people with whom they associated to employ the designated driver tactic. Furthermore, results from another item in the survey suggested that ample numbers of individuals are willing to serve as designated drivers. Nationally, 78 percent of the individuals who visit settings where alcohol is served indicated that they would be willing, on occasion, to serve as a designated driver. Interestingly, drinkers were more inclined than nondrinkers to serve as designated drivers (84 percent versus 67 percent, respectively). Although the Gallop Poll indicated that most people approve of the designated driver concept, the question remains as to whether their favorable disposition translates into practice. How often and under what circumstances do groups of drinkers designate some of their members to remain sober and do the driving? Informal Designated Drivers One of the advantages of the designated driver tactic is that formal programs are not necessary. Any group of individuals in any drinking environment can designate someone to remain sober and do the driving. Unfortunately, information about these "informal" designated drivers is even more scarce than about those who participate in formal programs. It consists of responses to a few survey items included in studies directed primarily at other aspects of the drinking/driving problem. For example, as part of an unpublished telephone survey that Wayne Harding and I conducted in 1987, people living in a Boston suburb were asked about the designated driver tactic. Respondents were randomly selected from a list of licensed drivers and then screened to produce a sample of 502 individuals who reported having used alcohol and having driven (not necessarily together) in each of the 2 past years. To estimate use of the designated driver tactic, we asked respondents, "During the last 12 months, how many times were you part of a group of drinkers in which someone didn't drink so they could drive others in the group ?" Over half (53 percent) of the entire sample indicated that they had been part of such a group. Eighty-four percent of the individuals who had been in such a group reported that it happened 12 or less times during the past 12 months (nearly half said 3 or fewer times). Another item asked how groups made the decision to use the designated driver tactic. Only 3 percent of the respondents reported that they had been in a group that was "encouraged by a bar or restaurant" to designate someone to remain sober and do the driving. This low figure fits with the finding reported below that formal programs produce few designated drivers. All other respondents reported that their groups made the decision to designate a sober driver on their own. Snortum, Hauge, and Berger (1986) also conducted surveys bearing on use of the designated driver tactic for reducing drunk driving. They estimated that 12 percent of American groups of drinkers always appoint one person to remain sober to drive, while 42 percent of such groups never designate a driver to remain sober. In sharp contrast, 76 percent of Norwegian groups were estimated to always employ the designated driver Tf&4NSPORTATlON AND ALCOHOL SEFMCE POLICIES 159 tactic, while only 4 percent never used it. Unfortunately, the findings were clouded by a 58-percent response rate in their U.S. sample and an unknown response rate in Norway. Survey of Formal Designated Driver Programs In 1985, the National Highway Traffic Safety Administration funded a study specifi- cally aimed at answering questions concerning designated drivers (Apsler et al. 1987). The major focus of the project became the examination of formal designated driver programs (DDPs) - systematic efforts by drinking establishments or organizations to actively promote the designated driver tactic. A "snowball" approach was employed in an effort to locate as many DDPs as possible throughout the United States. As various individuals and organizations were contacted, they identified some DDPs and suggested other individuals who might know of more. Ultimately, these sources produced the ,, names of 431 alleged DDPs. To verify the existence of some programs and to learn more about them, we then began telephoning a geographically diverse sample of the programs that had been praised most highly by our sources. Telephone conversations were held with spokespersons from 37 operational and 3 defunct DDPs. Four of the 37 programs were also visited by the investigators. A second series of phone calls was made to membership organizations, such as fraternal clubs, veterans organizations, and fraternities and sororities. Some of these organizations are notorious for the heavy drinking that takes place, yet virtually none appeared on the list of DDPs identified through the snowball survey. Fii-four mem- bership organizations randomly selected from seven major U.S. cities were contacted, resulting in five DDPs with whom extended conversations were held. Characteristics of DDPs The formal DDPs that were contacted tended to be similar in many ways, despite assorted variations. For example, most DDPs operated whenever the drinking estab- lishment was open, though a few restricted the hours, and two operated onlyon holidays. All respondents claimed that the cost of operation for their DDP was minimal and inconsequential. All DDPs utilized some form of in-house publicity, such as posters, table tents, employee buttons, and promotion by the server or doorman. When publicity occurred in the mass media, it tended to be donated or consisted of news items. Over half of the DDPs stated that they had eligibility requirements for participation, usually in the form of a minimum group size that ranged from two to six people. One program also specified a maximum group size of six people, reasoning that larger groups would not fit in one automobile. Nearly all DDPs required that the designated driver abstain from alcohol, though some permitted the driver to have up to two drinks. All but two DDPs gave incentives to designated drivers, usually in the form of free nonalcoholic drinks. A few DDPs offered free food, and others gave coupons that could be redeemed in the future for free food or drinks. In many establishments, a patron wanting to obtain the incentive for becoming a designated driver had to approach the server and make the request. However, in roughly a third of the DDPs, one of the staff was expected to approach eligible parties with an explicit request that someone be designated as the sober driver. Once selected, the designated driver often received some form of iden- tification, such as a button or hand-stamp. The Number of Designated Drivers It appears that few individuals are participating in the DDPs offered by drinking establishments. Precise figures were unavailable, since few establishments kept reliable records of the numbers of designated drivers, and none recorded the number of eligible 160 BACKGROUND PAPERS groups served. Nevertheless, estimates offered by drinking establishments provide at least a ballpark indication of the numbers involved. The majority of drinking estab- lishments reported serving 20 or fewer designated drivers per week. A'much more meaningful figure is, of course, the percentage of eligible groups in which someone served as a designated driver. Here the figures are even more tenuous, since only a few respondents made a guess at the number of eligible groups served by their establishment. Typically, less than 10 percent of the eligible groups participated in a DDP. Limited Appeal of the Designated Driver Tactic The vast majority of Americans approve of the designated driver tactic and are willing, on occasion, to serve as a designated driver. Yet existing evidence indicates that relatively few groups of drinkers in the United States actually employ the tactic. Part of the explanation may be that there are many circumstances in which serving as a designated driver is unappealing and/or impractical. Who Will Abstain if Everyone Wants to Drink? Individuals' willingness to serve as a designated driver is probably associated with their perception of how necessary drinking is for enjoying an activity. At sporting events, for example, beer drinking may be at least as important for some people (especially young males) as the contest. In such settings where drinking, particularly heavy drinking, is a central part of the experience, remaining sober may be seen by many individuals as too great a sacrifice. In principle, taking turns can make serving as the designated driver more palatable. But if group membership is fluid, equitable sharing of this role becomes difficult. The designated driver tactic is likely to be unappealing in numerous other circum- stances. For instance, when two people go out for drinks, will one of them drink alone? Social activities intended to bring together people who do not know each other well are another example in which drinkers might be reluctant to forego alcohol, since drinking is viewed by many as a social lubricant that facilitates meeting strangers. People who generally feel that alcohol helps them relax and become more outgoing may be unlikely to accept the handicap of not drinking. Problem drinkers are also poor candidates for the role of designated driver. Even people willing to remain sober may resist serving as a designated driver out of reluctance to transport a bunch of drunks. Who wants a car full of potentially belligerent individuals? Who will clean up the mess and get rid of the smell if someone vomits? Logistical Problems Even when individuals are willing to remain sober, they may often be reluctant to serve as a designated driver for logistical reasons. The designated driver tactic is most attractive only when the starting points of group members are geographically close and also when their ultimate destinations after drinking are near each other. It is un- reasonable to expect one individual to drive long distances to pick up and drop off other group members. A compromise is for everyone to drive to a central location from which the designated driver ferries the group to and from the location where drinking occurs. This compromise does reduce the number of miles driven by the intoxicated members of the group. However, if intoxicated group members use the availability of a designated driver as justification for drinking more than they usually would, then their risk of a crash could be even greater than if they had drunk less and driven the entire distance themselves. TRANSPORTATION AND ALCOHOL SERVICE POLICIES The Need for Planning This last example points to another limitation of the designated driver concept: the need for organization and planning in advance of the activity. Once individuals have arrived in their own cars at the drinking location, it is probably too late in most circumstances to employ the designated driver tactic. Although it would still be possible for the drinkers to leave their cars at the drinking location and. ride home with the designated driver, most people are probably reluctant to do so. Clearly, it is better to determine in advance who will attend an activity that will include drinking, how many vehicles will be needed, and who will drive. Timing is another critical aspect of the planning issue. The designated driver tactic requires all members of a group to arrive and depart at the same times. This extensive planning seems practical mainly in struc- tured activities, such as staged events or regular social functions. 161 Other Disadvantages of the Designated Driver Tactic / Resistance From Servers Alcohol servers in the drinking establishments we visited explained that they receive smaller tips from groups with designated drivers. The nonalcoholic beverages that designated drivers drink are usually free, or if purchased, they cost much less than alcoholic beverages. Consequently, alcohol servers, who rely on tips for much of their income, suffer economically when they serve designated drivers, unless drinking estab- lishments make arrangements to compensate them for lost tips. Obviously, without a subsidy, servers may be reluctant to encourage patrons to adopt the designated driver tactic, and we found no establishment that provided such a subsidy. Reliance on an Honor System In many drinking situations, the success of the designated driver tactic depends entirely on the commitment of both the.designated driver and other members of the group. Even when alcohol servers and hosts encourage use of the tactic, they can do little to ensure either that designated drivers begin sober, remain sober, or that they do the driving. Servers typically must rely on buttons or hand-stamps to identify designated drivers, and these devices can easily be hidden by someone determined to obtain alcoholic beverages. Similarly, servers or hosts have no way to guarantee that the designated driver drives all members of the group. We did find rare instances in our study of DDPs when doormen would occasionally follow patrons into the parking lot to make certain that an apparently sober individual got behind the wheel. Nevertheless, even such extreme efforts can easily be circumvented. Determination of "Sober" While most formal DDPs do not serve alcohol to designated drivers, individuals employing the designated driver tactic informally are free to determine the degree of sobriety that the designated driver must maintain. A potential danger is that some groups will be mistakenly complacent as long as the designated driver is less intoxicated than the others, even if not completely sober. Excludes the Solitary Drinker By definition, the designated driver tactic works with a group of individuals- not with a single individual. Consequently, the tactic cannot help provide safe transportation for the solitary drinker. 162 Safe Rides Another transportation-based approach for reducing alcohol-related automobile crashes is to provide both vehicles and drivers for intoxicated individuals who would otherwise drive themselves or ride with an intoxicated driver. This approach is usually referred to as "safe rides." Formal safe ride programs (SRPs) encompass a wide variety of transportation alternatives, such as taxicabs, limousines, tow trucks, buses, and automobiles, while the informal provision of safe rides generally relies on taxicabs and automobiles. Typically, when transportation is needed by an intoxicated person, either that person or someone else, such as a server or host, obtains transportation from outside the drinking environment. In some instances, contact is made directly with a company that provides the transportation, such as a taxicab company, while in others a wm- munications service is contacted, and it, in turn, makes arrangements for transportation. The safe rides tactic is a theoretically perfect solution to the drinking/driving problem. A primary attraction from a drinker's standpoint is that no modification of drinking behavior is necessary. No one need remain sober or even moderate his or her alcohol consumption, and still all drinkers can be transported home without endangering either themselves or others. In addition, the safe rides tactic works with solitary drinkers. Unlike the designated driver tactic, the safe rides tactic does not depend on the existence of groups of individuals. The safe rides tactic appears to receive the same high level of approval that was found for the designated driver tactic. Caudill, Kaufman Kantor, and Ungerleider (1988) interviewed 1,522 patrons as they entered bars and nightclubs in Sacramento and San Jose, California. The survey was conducted to obtain baseline data for a study of SRPs. Nearly all of their respondents (% percent) believed that the availability of SRPs would be useful; 63 percent of the respondents selected the number "10" on a l-10 scale of usefulness. In addition, 79 percent of all respondents reported that they might use such a service if it were available. An even larger percentage of heavy drinkers, 87 percent, indicated that they might take advantage of a safe rides service. On the other hand, they also found that a large number of respondents (38 percent) had not heard about SRPs, and few had actually used one. Seven percent of all respondents and 12 percent of heavy drinkers reported that they had used the services of a SRP sometime in the past. Informal Safe Rides As in the case of the designated driver tactic, formal programs are not necessary for the safe rides tactic to be used. Intoxicated individuals do not need a SRP to use taxicabs or receive rides from sober friends in order to avoid drunk driving. However, this informal use of the safe rides tactic has received almost no attention from researchers. One exception is an ethnographic study of bar settings conducted by Gusfield, Rasmus- sen, and Kotarba (1984). They recount observing incidents in four drinking settings where bartenders would sometimes call a taxicab for intoxicated patrons wanting to avoid driving. The likelihood of bartenders assisting patrons in avoiding drunk driving depended largely on the relationship between patron and bartender. Furthermore, the authors noted considerable variation both among and within drinking establishments in bartenders' efforts to help patrons obtain a safe ride. Hernandez and Rabotis (1987) study of interventions in drunk driving situations also provides information about the informal use of safe rides. They questioned 247 college student volunteers from an introductory sociology class to learn about incidents in which someone had tried to stop them from driving after drinking. Ninety-seven of the students reported experiencing such an incident. Most of the 89 reported interventions in which TRANSPORTATION AND ALCOHOL SERVICE POLICIES 163 someone drove the respondent home occurred at parties (49 percent), and a somewhat smaller portion (33 percent) occurred at friends' homes. Many fewer instances of respondents being driven home took place at either bars (12 percent) or restaurants (6 percent). The results of these two studies substantiate informal use of the safe rides tactic but obviously leave a great many questions unanswered. Survey of Formal Safe Ride Programs In 1986, the National Highway Traffic Safety Administration funded a survey of SRPs (Harding, Apsler, and Goldfein, 1988u, b) that followed the same procedure described above for our earlier study of the designated driver tactic. First, leads on 515 SRPs across the country were obtained using a snowball survey. After information for 325 programs were verified, detailed data were collected on 52 programs (see Harding, Apsler, and Goldfein 1987 for a directory that summarizes key features of the 325 programs). Twelve of the 52 safe ride programs were then visited by the investigators. Characteristics of SRPs We found SRPs existing in communities ranging from small towns to large cities. They were operated by numerous types of organizations, including cab and bus companies, charitable organizations, trade associations, hospitals, government agencies, and non- profit organizations set up specifically for this purpose. Many operated year-round, often providing service every day of the week. Safe ride programs advertised themselves both through the local media and with signs in drinking establishments and other locations. Programs run by transportation companies typically used their own dispatchers and drivers, while other programs used various combinations of paid and volunteer staff. Often, SRPs were staffed largely by volunteers and operated in conjunction with a transportation company that provided the vehicles and drivers. Even though nearly all programs provided their service at no cost to riders, the average annual cost of the year-round programs was under $12,000. Most programs obtained funding from a wide variety of sources, the most common of which were donations in the form of free advertising from the media, member fees paid by drinking establishments and/or corporations, donations from alcohol distributors, and fundrais- ing activities. Most programs accepted requests for rides either from drinkers or someone calling for the drinker, while some took calls only from drinkers and others only from alcohol servers. Many programs screened riders to make sure they fit their requirements, such as whether the rider was intoxicated, drove his/her own vehicle to the drinking site, intended to go directly home, and whether the origin and destination fit within the program's operating range. The Number of Riders As was the case with DDPs, reports of the number of riders transported by SRPs must be interpreted with caution. For instance, some programs could only make estimates, and others could not separate the number of requests received from the number of rides provided. Nevertheless, about half of the programs reported delivering roughly 400 or more rides per year, and about a quarter delivered 1,000 or more rides per year. Unfortunately, almost no ride programs gave estimates for the size of the target popula- tion. Since SRPs tend to cover entire communities or large sections of major cities, the number of eligible intoxicated drivers within their operating boundaries could be quite large. BACKGROUND PAPERS Disadvantages of Ride Programs Cars Are Usually Left Behind Only a small percentage of SRPs (about 15 percent of those we contacted) transport drivers' cars or provide free transportation the next day to help drivers retrieve their vehicles. Obviously, some drivers may forgo a ride home knowing that they must leave their car behind. They may be concerned that returning home without the car will incriminate them; they may fear that their car will be stolen and/or vandalized if left overnight; or they may simply want to avoid the inconvenience of retrieving their car. Determination of Level of Intoxication With rare exceptions, SRPs rely on either the drinker or someone else at the driig site to determine whether the drier is too intoxicated to drive safely. Intoxicated individuals are notoriously incapable of accurately judging their level of intoxication, and research shows that even people who often observe intoxicated drivers, such as alcohol servers, tend to be poor judges of level of intoxication (Langenbucher and Nathan 1983). As a result, it may be the more cautious drivers who tend to seek out rides for themselves, while observers, such as servers, may tend to single out only the most obviously intoxi- cated individuals for a safe ride home. Various devices, such as "Know-Your-Limit" cards and breathtesting machines, could easily be made available to patrons wanting assistance in judging their level of intoxication. Even so, it might still be only the more cautious individuals who would use these devices. Potential for Abuse People who are not intoxicated can easily take advantage of many SRPs, as can intoxicated individuals who do not have cars. In an effort to minimize abuse, some RSPs screen clients. For example, some take requests for rides only from servers as one way of insuring that riders are intoxicated, and some ask to see driver's licenses and keys as at least partial assurance that the rider would otherwise drive. However, no information exists about the prevalence of inappropriate use of SRPs. Staff in some programs also reported efforts to screen out individuals who made frequent requests for rides. Unfor- tunately, those individuals (often referred to by program staff as "alcoholics") may be the ones who present the greatest danger on the roads. A Special Case: Transportation to and From Drinking Locations A rare variation of SRP transports drinkers in both directions-first bringing them to drinking locations and then taking them home from those locations. This "round trip" version of safe rides overcomes several important limitations of the "home-only' programs. For instance, the question of what to do about drinkers' cars disappears, since people do not drive their cars to the drinking locations. In addition, these SRPs make no attempt to determine level of intoxication, so the issue of abuse of the program disappears - anyone, intoxicated or not, can use these programs, In those situations where the "round trip" SRP is practical, safe rides can become a nearly perfect solution to the drunk driving problem. The most common examples are special buses or trains that transport people to and from a scheduled event, such as a sporting event. Another variation occurs in Boston, where the mass transit system is kept TRANSPORTATION AND ALCOHOL SERVICE POLICIES running later than usual on New Year's Eve to help transport the thousands of people attending events held in the downtown area. The problem with these versions of the safe rides tactic is that many drinkers may drive themselves home from the points where buses, trains, or subways deposit them. Another example of round-trip SRPs is the increasing popularity of limousine service on high school prom nights in some com- munities. If door-to-door service is provided, then there is obviously no driving after drinking. The major obstacles to wider use of round-trip SRPs based on mass transit are the same ones that prevent wider use of mass transit in general. Given the widespread ownership of automobiles in most locations and the existence of an extensive highway system, mass transit cannot compete with automobiles bn convenience and cost. Round- trip SRPs are a realistic option when (1) a large number of individuals live in relatively dense areas, and (2) they travel to drinking sites located in a relatively small geographical area. These conditions are typical of some college towns. Dormitories might be clustered in one area and fraternities and sororities in another, while many of the favorite drinking establishments congregate in one section of the nearby town. However, now that the legal ' drinking age has been raised across the country, most undergraduates cannot legally drink in drinking establishments. Consequently, there may no longer be sufficient traffic between campuses and drinking establishments to make round-trip programs practical. The Costs of Expanding Safe Ride Programs If SRPs are to be more effective, they must transport more riders. Yet, it is not clear whether the existing sources of funds and/or volunteers could keep pace with a significant increase in the use of SRPs. The key, of course, is determining how much the ridership would increase if efforts were made to expand the role of SRPs. The answer to that question depends largely on the nature of the population of potential riders that is chosen as a target. For example, SRPs could transport anyone who wants a ride, or they could be restricted to intoxicated individuals. If riders must be intoxicated, then a level of intoxication should be chosen. The commonly used BAC of 0.10 percent is an obvious candidate, though a much lower level, such as the 0.05 percent BAC recommended by the American Medical Association (Council on Scientific Affairs 1986) must be con- sidered. Will potential riders have to prove that they would drive if they are refused use the safe ride program? Will passengers trying to avoid riding with an intoxicated driver be transported, as is usually the case with existing SRPs? Will underage drinkers be transported? The maximum length of rides and permissible destinations of trips are other factors that markedly impact cost estimates. At present, little is known about how the number of eligible individuals would vary with the selection of different target populations. Nor is there information on the percentage of eligible people who might be persuaded to take advantage of SRPs. At least two other types of costs must be considered in planning for broader use of safe rides programs. First, drinking establishments might have to increase their parking facilities to hold the cars that accumulate as their intoxicated drivers receive alternative transportation home. Second, the cost of transportation back to the drinking site at a later time for drivers to retrieve their cars must be included in the overall equation. These costs can be built directly into the SRPs by having them transport cars along with their drivers. For example, a few SRPs transport riders' cars with tow trucks or provide a second, sober driver. On the other side of the cost issue is the question of who would pay for an expanded system of SRPs in the event that existing funding sources could not cover the costs. There are several possibilities. Additional expenses could be borne by those who obtain rides, or .the costs could be spread across a larger population, such as all drinking estab- 165 166 BACKGROUND PAPERS lishments, all drinkers, or even across all taxpayers. Cost sharing, say between riders and drinking establishments, is another option. Finally, an effort to markedly expand SRPs will almost certainly depend heavily on taxicabs. Consequently, it will have to contend with the 50-percent drop in the number of taxicabs and the 4Opercent decline in taxicab operators that occurred over a recent M-year period (Gilbert et al. 1984). On a more positive note, Teal (1985) reported on developments in the taxicab industry that may improve its financial health. Conclusions The designated driver and safe rides tactics comprise a class of transportation alternatives that help reduce the number of intoxicated drivers on our roads. Both tactics enjoy broad support from potential users and can be employed in virtuaily any drinking setting. Safe rides programs make even fewer demands on drinkers than the designated driver tactic and can be successful without advance planning. Hundreds of designated driver and safe rides programs have been established in a broad array of settings. They cost little to operate in their present forms and receive broad support from drinkers, drinking establishments, community organizations, activist groups, and the alcohol beverage industry. Yet, two key questions remain: (1) Just how much impact have the designated driver and safe rides tactics had so far in reducing numbers of intoxicated drivers? (2) What is their potential for making a further reduction? Formal DDPs, according to reports of the programs themselves, produce relatively few designated drivers. While some people employ the designated driver tactic on their own, they may do so infrequently. Based on scanty results, the main effect of publicity about the designated driver tactic may be to encourage drinkers to ride with abstainers or light drinkers when such individuals happen to be available and willing to transport others. The most important disadvantages of the designated driver tactic are probably the need for planning and the existence of many circumstances where the tactic is unappealing to drinkers and/or Iogistically impractical. Nationally, SRPs transport thousands of individuals each year. However, riders' levels of intoxication have not yet been documented, nor has the number who would have either driven or obtained a ride from an intoxicated driver in the absence of SRPs been verified. Furthermore, little is known about the numbers of eligible riders and the feasibility of markedly expanding the scope of SRPs. One likely obstacle to increasing ridership is the requirement in many SRPs that riders leave their cars at the drinking site. The policy implications of existing data are that the designated driver and safe rides tactics should continue to be encouraged and supported with the clear understanding that these strategies are limited in what they can be expected to accomplish. Until additional research shows otherwise, it appears that these strategies are unlikely to fulfill what, at first glance, appears to be their enormous potential for reducing the numbers of intoxicated drivers. Of the two strategies, the designated driver tactic is the more questionable. At present, the only prudent position is to remain extremely skeptical about the impact that it can have. In those circumstances when use of the tactic is both appealing and practical, it can be completely effective in eliminating intoxicated drivers from the roads. Research is necessary to determine how often those circumstances exist and to explore the possibilities for increasing the frequency with which they occur. Safe rides programs, especially when coupled with servers and hosts assuming the responsibility for detecting intoxication and ensuring the use of alternative transportation, can be extremely effec- tive. Here, too, research is necessary to determine just how effective SRPs are in practice and how serious are the obstacles to their wider use. T~SPORTATION AND ALCOHOL SERVICE POLICIES 167 Use of the designated driver and safe rides tactics could be increased by addressing some of the problems discussed above. For example, participation in the designated driver tactic could be bolstered by stressing the need for planning and encouraging drinking establishments to provide incentives for groups of driers who arrive in a single car and also participate in a DDP. More people would probably take advantage of SRPs if their cars were transported. Employers could help by following the lead of those who distribute coupons for a free ride to their employees, thereby making it increasingly difficult for them to justify driving while intoxicated. Motivation Is a Key Factor Ultimately, however, the prospects for expanded use of the two strategies may depend less on their specific characteristics than on the level of motivation among drinkers to avoid drunk driving. Both strategies are primarily procedures that can be employed by drinkers who are already motivated to avoid drunk driving. While publicity about DDPs and SRPs may reinforce concern with drunk driving and may trigger action when presented during drinking ictivities, it probably contributes relatively little to overall motivation. Results from surveys cited earlier are consistent with thii line of reasoning. For example, Snortum, Hauge, and Berger (1986) attribute the much greater use of transpor- tation alternatives in Norway than in the United States to national differences in attitudes toward drinking and driving. They make no mention of differences between the two countries in either publicity about transportation alternatives or in availability of these services. Similarly, the discrepancy between American's widespread approval of both the designated driver and safe rides tactics and their infrequent use of the tactic points to lack of motivation as a likely explanation. More Research Is Essential The tentative conclusions presented here are largely speculative due to the paucity of data. Thus, the one clear message that emerges from the area of transportation alternatives for intoxicated drivers is the need for additional research. The scarcity of research is surprising given the central role that alternative transportation will have to play if drunk driving is to be substantially reduced. Drinking practices appear to be relatively immune to change. Consequently, the success of efforts to motivate people to avoid drunk driving will depend heavily on the availability of attractive and practical alternatives to driving. REFERENCES Apsler, R.; Harding, W.M.; and Goldfein, J. The review and assessment of designated driver programs as an alcohol countermeasure approach. Technical Report DOT HS 807 108. Washington, DC: National Highway Traffic Safety Administration, 1987. Caudill, B.D.; Kaufman Kantor, G.; and Ungerleider, S. "Safe Rides: A Controlled Investigation in Two Major California Cities." Paper presented at American Psychological Association Annual Conference, Atlanta, 1988. Gxmcil on Scientific Affairs. Alcohol and the driver. Journal of the American MedicalAssociation 255(4):522-527, 1986. Gallup, G., Jr. Designated drier program: Who holds the key to safety?Alcohoftim &Addicrion 12:16,1987. Gilbert, G.; Burby, RJ.; and Feibel, C.E. Taxicab operating characteristics in the United StateS. Tranportation 12(2):173-182, 1984. Gusfield, J.R. Kinking-Driving and the Symboiic Order. Chicago: University of Chicago Press, 1981. Gusfield, J.R.; Rasmussen, P.; and Kotarba, J.A. The social control of drinking-driving: An ethnographic study of bar settings. Law & Policy 6(1):4%6,1984. 168 BACKGROUND PAPERS Harding, W.M.; Apsler, R.; and Goldfein, J. A Directory ofRide Service &ograms. Interim Report DOT-HS-807-146. Washington, DC: National Highway Traffic Safety Administration, 1987. Harding, W.M.; Apsler, R.; and Goldfein, J. ~AssessmenrofRideSeniceProgramsar MAlcohol Cowrtmmeanue. Technical Report DOT HS 807 290. Washington, DC: National High- way Traffic Safety Administration, 198&z. Harding, W.M.; Apsler, R.; Goldfein, J.A Users'Guide toRia% ServiceProgra??zs. Technical Report DOT HS 807 291. Washington, DC: National Highway Safety Traffic Administration, 1988b. Hemandez, A.C.R., and Rabow, J. Passive and assertive student interventionsin public and private drunken driving situations. Journal of Studies on Alcoholism 48(3)5X59-271,1987. Langenbucher, J.W., and Nathan, P.E. Psychology, public policy, and the evidence for alcohol intoxicationAmerica hychobgkt 38(10):1070-1077,1983. Mosher, J.F. Alcohol policy and the presidentialcommission ondrunkdriving: The pathsnot taken. Accident Analysir & Prevention 17(3):239-250,1985. National Commission Against Drunk Driving. Progress Repott on Rec~mmendatiom Proposed by the Presidential Commission on Drrurk tiving. DOT HS 807.. 043. Washington, DC: National Highway Traffic Safety Administration, 1987. National Highway Traffic Safety Administration. Alcohol and Highway Safety 1984: A Review oj the Stole of the Knowledge. DOT HS 806 569. Washington, DC: National Highway Traffic Safety Administration, 1985. Reed, D.S. Reducing the costs of drinking and driving. In: Moore, M.H., and Gerstein, D.R., ed. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: National Academy Press, 1981, pp.336-387. Ross, HI,. Brewers view drunk driving: A critique.AccidentAnalysir & Prevention 19(6):475-477, 1987. Snortum, J.R.; Hauge, R.; and Berger, D.E. Deterring alcohol-impaired driving: A comparative analysis of compliance in Norway and the United States. Justice Quarterly 3(2):139-165, 1986. Teal, R.F. Private enterprise in public transportation: The case of the taxi industry. Trunsportution Quarter& 39(2):235-252 1985. Transportation and Alcohol Service Policies Server Intervention and Responsible Beverage Service Programs Robert F. Salk, Ph.D. Prevention Research Center Berkeley, California The recent rise in alcohol-impaired automobile crashes, injuries, and fatalities after a dip in those rates over the past few years reminds us how intractable this problem is. School-based education and mass media programs, intensified law enforcement, and court-ordered treatment programs have no doubt changed the normative climate regarding drinking and driving, and yet something about the phenomenon limits the effectiveness of these attempts to change individual attitudes and behavior. Recognizing that people's ability to alter their habits depends greatly on situational influences, prevention specialists have turned their attention to reducing the risk of alcohol-impaired driving through modification of the drinking environment itself. Alter- ing those contexts can, in conjunction with the educational approaches, reduce risk to a far greater degree than would either strategy alone. Server intervention refers to a broad set of strategies to create safer drinkingenviron- ments that first, reduce the risk of intoxication and,second, reduce the risk that intoxi- cated persons will harm themselves or others. These strategies include specialized training for servers and managers, but could also comprise raising the prices of alcoholic beverages, promoting food, and altering decor to foster safe drinking. Since approxi- mately half of those driving while intoxicated (DWI) come from a place licensed to sell alcoholic beverages (O'Donnell 1985), it seems natural to look at ways to intervene in those places to prevent the problem. In a series of articles, Mosher (1979,1983,1984u) has laid out a conceptual framework for server intervention that addresses environmental reforms at two basic levels: the legal environment and the specific environment of the licensed establishment. The broad goal of server intervention is to work in a coordinated fashion at both levels to achieve consistent and effective prevention. The first and most encompassing level, the legal environment, includes dram shop (civil) liability law, State and local Alcoholic Beverage Control (ABC) codes, and criminal statutes that affect serving practices. Dram shop (liquor liability) laws are those that hold commercial servers (and sometimes social hosts) liable if they serve obviously intoxicated or underage persons who subsequently cause harm to others or themselves. Mosher (19&k) as argued that current liability laws are vague and pay little attention to their potential in preventing alcohol-related deaths and injuries, and has coauthored a model dram-shop law that would correct these deficiencies (Colman et al. 1985). ABC NOTE: preparation of this article was supported by the National Institute on Alcohol Abuse and Alcoholism Research Center grant AA06282 to the Prevention Research Center, Pacific Institute for Research and Evaluation. The material is this caapter is taken from a larger paper appearing in Health Education Quarterly. 170 BACKGROUND PAPERS statutes and regulations also, of course, determine how, when, and where alcoholic beverages may be served, but there again, little heed is paid to prevention. Indeed, many view these provisions as "quaint but outdated remnants of a past era" when the primary concern wasivith controlling vice and other criminal activity (Mosher 1984b). Criminal statutes constitute a third facet of the legal environment, usually drafted to state explicitly what is often contained in ABC codes and dram shop laws (e.g., laws prohibiting sale to minors or obviously intoxicated persons). The level that has received more attention is the environment of the establishment itself. The earliest server intervention programs concentrated primarily on training servers to recognize intoxication and refuse service to any customer who appeared intoxicated. `As they gained experience with such programs, howeve? many trainers felt that server training alone was not sufficient to prevent intoxication. First, interven- tion took place after the onset of intoxication, and second, servers seemed unable to carry out their new responsibilities unless management and management policies were solidly behind the prevention effort. Thus, more comprehensive programs were developed to include review and modification of management policies and operations, in addition to training for employees. Reflecting the evolution, one now hears more about "responsible beverage service" than server intervention per se. The review and revision of management policies are not limited to those prohibiting alcohol sales to minors and obviously intoxicated customers. They also focus on the availability and promotion of nonalcoholic beverages and food, standards for customer behavior, minimum staffing levels, transportation for intoxicated customers, and full management support for servers who limit their customers' consumption. Server training is necessary because most new policies require the server to accom- modate to several major changes. Servers must redefine their role with respect to the customer and learn a new set of skills for monitoring and controlling customers' driig. In addition to concrete knowledge and skills, however, training must help servers understand the program goals, modify their own attitudes about alcohol and its service, and overcome any fear or anxiety they may have about their new duties. Research to Date In the last couple of years, a handful of server intervention studies have been conducted. Although their aims and methods differ, each has tried to estimate the impact of server training or server intervention on either the server's behavior or the customer's consumption of alcohol. Very little (if any) systematic research has yet been conducted regarding specific components of a program, or how the program can be delivered for maximum impact and efficiency. One of the first evaluations of server intervention, the Navy Server Study (Saltz 1987), sought simply to determine whether the concept of server interveition had potential merit as a prevention strategy. Two similar Navy Clubs for enlisted personnel were selected, with one serving as a program site and the other as a comparison. The test site employed approximately 50 people who had direct contact with customers. It took in 1 For a brief history of Server Intervention programs and evaluation, see Saltz, RF. Server Intervention: Conceptual overview and current developments.dkohol, Drugs, andDriving Abstracts andReviews 1(4):1-14, 1985. 2 One may compare the proceedings from the 1984 and 1987 Responsible Beverage Service Forums sponsored by the Responsible Hospitality Institute of Springfield, MA. The proceedings from the latter meeting of program and research specialists revealed a much greater concern for management systems to support server training. T&W3PORTATlON AND ALCOHOL SERVICE POLICIES over half a million dollars from alcohol sales in 1985 (ivhen the data were collected), and would get as many as 800 customers on a busy night. The program itself involved extensive consultation with the club manager, producing several changes in club policies and practices, and an l&hour training course for all staff. The policy changes included promoting nonalcoholic beverages and food, overtly delaying service of an alcoholic beverage if it would put the patron at or above the legal limit for intoxication, and discontinuing the sale of beer in pitchers. Food service, which had previously been segregated from the bar area, was installed in the barroom, and money incentives were provided for servers and cooks to promote food sales. In addition, where servers had been free to serve customers anywhere in the building, the new program assigned them to specific sections of optimal size so that customers' consump- tion could be monitored. The food and beverage menus were expanded and drink prices raised marginally to cover the program costs. The training course, broken into five modules and spread out over as many weeks, covered the reasons for change, alcohol's effects on the body, monitoring customers' consumption to know when they had reached the limit, and techniques to pace service and refuse it when necessary. Group discussions and visual presentations were used throughout the l&hour program, with role-play exercises dominating the last two sessions. Data to measure the program's impact came from interviews with randomly selected customers, structured observations of selected customers' consumption, and archival data of alcohol and food sales provided by the clubs. The project did not measure changes in server behavior, primarily because the researchers could not agree on a method, but also because the prime question was the program's effect on patron consumption. Data were collected for 2 months prior to program implementation and 2 months following. Results have been reported from the interview data. Customers were interviewed on Thursday, Friday, and Saturday nights for 3 to 5 minutes. Questions included arrival time, mode of transportation, consumption of speciftc foods and beverages, frequency of patronage, age, height, and weight. The primary dependent measure was whether the patrons were over their "limit" as defined by a drink-counting system introduced in the training (a limit based on the number of drinks, the duration of time drinking, and the patron's weight category). Thii limit, incidentally, corresponds very closely to the BAC estimate derived from a formula given in Segal and Sisson (1985). A logistic regression analysis that statistically controlled for intervening variables showed that the risk of intoxication, which was as high as 32 percent for males at the test site, was cut in half (to 15 percent) after the program was implemented. For females, the rate dropped from 5 percent to 2 percent. Figure 1 shows the cumulative distributions of BAC levels (pre- and post-program) at the experimental site. Note, for example, that the BAC for the 70th percentile dropped from approximately 0.12 percent to 0.07 percent. While the Navy study accomplished its goal of showing the potential for server intervention, many questions remained about the generalizability of its fmdii, the relative effectiveness of the training and policy changes, and the need for such extensive consultation and training to achieve the results. The servers'behavior was not monitored, the evaluation only assessed the short-term impact, af3d no one knows how the Navy Club setting may differ from commercial establishments. A second study, reported by Russ and Geller (1987; Geller et al. 1987), concerned 171 3 While some wonder whether the Navy Club was a more controlled environment, it was not unusual to see fistfights break out on the premises, along with the usual attempts by underage patrons to obtain an alcoholic beverage. The club was also under constant pressure lo produce the profit necessary to keep various other recreational base operations funded. 172 BACKGROUND PAPERS 60 Cumulative w Percent 50 30 20 10 6 10 11 12 13 14 15 16 17 Total Alcohol Drinks 22 23 24 25 >25 Figure 1. Experimental site interview data, pre versus post-training one of the commercially available server training programs, TIPS (Training for Inter- vention Procedures by Servers of Alcohol) (Chafetz 1984). This program comprises a 6-hour course that includes video vignettes, group discussion, and role-playing, with an emphasis on identifying signs of impairment, pacing service, checking patrons' age, and promoting alternatives to alcoholic beverages. The trainees must then score at least 70-percent correct on a 40-question written test to become certified servers. The authors recruited 17 trainees from two commercial establishments, ending up with about half the employees having been trained. Research assistants, posing as customers, entered the establishments 24 times before and 25 times after the training and attempted to order and consume a drink every 20 minutes over a 2-hour period. If the training were effective, the server should intervene in some way to slow down or terminate the "pseudopatron's" consumption. Russ and Geller counted the type and number of interventions for each drink ordered (up to the maximum of 6 drinks) and compared the type and frequency of intervention for the trained staff (n = 17) and the untrained staff (n = 9) against the pretraining baseline (type and frequency of interven- tions) with all staff (n = 24). They found that while the untrained staff intervened no more frequently than at baseline (about 0.75 interventions), the trained staff intervened more frequently (3.24 interventions). Interventions included the offer of food or water, check- ing ID, delaying service, commenting on the quantity or speed of the customer's alcohol consumption, and making a driving-related comment. A second measure of the program's impact was the pseudopatron's BAC taken after the 2-hour drinking period. Whereas those served by the untrained staff had BACs as TRANSPORTATION AND ALCOHOL SERVICE POLICIES 173 high as baseline (mean of 0.103, with four of nine pseudopatrons above the legal limit), none of those served by the trained staff was over the legal liiit (mean of 0.059 BAC). These results were obtained without the benefit of policy changes at the management level or having all staff at an establishment working together to make the program a success. It would certain be interesting to compare these findings with those obtained when all of an establishment's staff was trained at once. In contrast to the Navy Server Project, the TIPS evaluation addressed training per se rather than training as part of a broader program. Its advantages include a direct measure of the server's behavior and its impact on the pseudopatron's BAC but, as with the Navy study, it has a few weaknesses as well. It, too, measured program impact in the short term, collecting baseline and posttraining data over an 11-week period. Though the trained servers were more likely to intervene, the interventions themselves were fairly mild. For instance, onIy 10 of the 55 interventions involved delay of service, and half of those were during the first three drinks, with no server delaying service of the sixth drink. The modal intervention was the offer of food or water, which accounted for about a third of the interventions. Eleven of the I3 interventions occurring on the sixth drink were either offers of food and water, or comments about the pseudopatron's consumption or driving. At no time was service refused. The mildness of the interventions was reinforced by the few examples given of how they were coded. The pseudopatrons were accompanied by a confederate who activated a small tape recorder when the server and pseudopatron interacted. Two research assistants, blind to the pre- or post-training condition, independently coded the inter- ventions. "Delay of service" could mean that the server offered to refill the confederate's nonalcoholic beverage without offering to get the pseudopatron another alcoholic drink. Since the confederate ordered only one drink during the 2-hour period, it's hard to interpret the significance of the "delay." A "driving-related comment" could be asking who was driving or suggesting that a nondrinking partner drive carefully. These inter- ventions do not seem capable of having a major impact on driving while intoxicated. What, then, of the pseudopatron's BAC? Here again, interpretation is complicated. The pseudopatrons were instructed to order a drink every 20 minutes for 2 hours. If the server intervened, they were to react in's manner similar to their normal drinking behavior. If offered food, for example, the pseudopatrons were told to accept it if they were hungry. This leaves the BAC measure to be a result of the interaction of the server's behavior and the pseudopatron's (unmeasured) inclination to accept the offer or heed the comment, whatever the case may be. As an example, two pseudopatrons could enter the same establishment, one could accept the offer of food while the other wasn't hungry. They would then presumably exit with different BACs despite identical "interventions" by the server. If the pseudopatron's behavior is not consistent, the generalizability of the results is uncertain. While we know the mean level of intervention increased for 17 trainees, we cannot tell from the reported data whether the increase in intervention was widespread among the trainees, or whether, say, 2 or 3 trainees were especially active and accounted for all the interventions while the other 14 or 15 remained unaffected by the training. Finally, individual trainees were not compared pre- and post-training, since the baseline data did not identify servers. Since the pool of trainees was basically self- selected, we don't know the degree to which the serving practices between the trained and untrained servers were different even before the training. The National Highway Traffic Safety Administration (NHTSA) has sponsored two demonstration and evaluation studies of sewer intervention. The first, called TEAM (Techniques of Effective Alcohol Management), represented a collaborative effort of NHTSA and several other organizations, and focused on alcohol service at seven selected arenas associated with the National Basketball Association (NHTSA 1986). The 174 BACKGROUND PAPERS program called for policy review and revision, followed by a Chour training covering the drinking and driving problem, liability law, alcohol's effects, recognizing impairment, policies and practices, and dealing with alcohol and drug related incidents. All arena employees were included in the training, not just those who served beverages. The TEAM evaluation report is hard toassess because the evaluation activities were directed toward program development. The program was constantly changing as dif- ferent data were collected at subsets of the seven participating arenas. The program was evaluated through a combination of a followup review of arena management policies (at five sites), surveys of staff and patron attitudes and reported behavior (at seven and three sites, respectively), and a review of sales data from two sites. The study showed that the program did result in policy revisions at the participating arenas, and that alcohol consumption (especially beer) declined in two sites while food and nonalcoholic beverage sales increased. 4 Through data collected from the staff and patrons, the researchers also concluded that management support was critical to the success of the program, and furthermore, that the support had to be visible for the-staff to carry out their own responsibilities. The TEAM evaluation is best thought of as an informal summary of loosely organized quantitative and qualitative data, much of which was apparently collected after the programs were in place. It provides many suggestions to program designers and trainers, but should not be considered a formal impact evaluation. The authors of the report, in fact, state that a formal evaluation design was inappropriate for their purposes and needlessly constraining. NHTSA also sponsored a study conducted by M&night (1987) that involved the development and delivery of a responsible beverage service program to 32 estab- lishments in Louisiana and Michigan. In this study, a 3-hour training was given to servers and managers, with 3 additional hours for the managers alone; 245 people were trained in all. A specially selected group of 10 establishments in each State was used for comparison, along with 24 establishments that were invited to participate but did not. The emphasis of the training was on prevention, providing the servers with strategies to prevent customers from becoming intoxicated. If service is performed responsibly, it should not be necessary to refuse service to anyone. The server's trainiig used videotapes followed by discussion of the material shown in the tapes, The training covered the concept of server liability, the moral and legal responsibility to prevent intoxicated patrons from driving, and the physiological effects of alcohol. The course then moved to prevention, including checking ID, serving food and nonalcoholic beverages, providing activities, and observing patrons for signs of impairment. The final module for servers covered intervention-what to do when customers became intoxicated-and included such tactics as delaying service, providing alternative transportation, and refusing serv- ice. The servers were expected to know when intervention was needed, but the managers were expected to carry it out. The extra 3 hours for managers covered intervention skills and strategies (with role-play exercises) and a section on responsible alcohol service business practices, where managers were encouraged to formulate policies relevant to their own establishments. The program's effectiveness was measured via pre-post differences in scores on a lo-item knowledge test and a stt of 10 items measuring opinions about the service of alcohol. The knowledge test comprised different, but equivalent, items for the pre and post-tests, while the opinion scale remained the same for both administrations. In a 4 The report does not offer consistent data on these changes, nor are there significant tests. Given the lack of comparison sites, it would be difficult to attribute changes in the test sites to the TEAM program alone. TRANSPORTATION AND ALCOHOL SERVICE POLICIES 175 separate set of items, servers were also asked about the frequency with which they engaged in several different types of activities related to prevention (e.g., offering coffee, inquiring as to who is driving, refusing service). Managers were given a checklist of beverage service policies (e.g., closing hours, availability of snacks) to indicate which they employed at their establishment. The same forms were given to servers and managers 4 months after the training. Research assistants were used in this study, also, to pose as customers, but here they were to feign intoxication when entering the establishment to see if they would be served a drink despite their condition. The pseudopatrons were trained to maximize the consistency of their behavior, which included staggering to their table, missing the chair or stool when sitting, slurred speech, and exhibiting difficulty in handling the money to pay for the drink. After 15 minutes, the pseudopatron would leave the establishment and record details of the encounter, whether any intervention had taken place, and whether any customers were intoxicated or drinking despite appearing underage. All estab- lishments were visited four times by four different assistants before and after the training. The results of the Louisiana and Michigan programs differed somewhat. Knowledge scores increased in both States, with Louisiana trainees (n = 120) improving their scores from a mean of 6.35 to 7.65, and Michigan scores (n = 95) improving from a mean of 6.24 to 8.23. The Michigan score change was (statistically) significantly greater than the other State's. The trainees' opinions became more favorable, too, after the training, with the Michigan trainees starting out with more favorable opinions than the Louisiana trainees. A self-report, serving practices questionnaire was completed by 55 percent of the Louisiana servers and only 29 percent of the Michigan servers. Apparently, many of the servers had quit working at their original establishments, and some had been promoted to managers. Both sets of servers reported a statistically significant increase in respon- sible serving practices. The manager's reports on serving policies showed a significant change in policies in Michigan but not Louisiana. Table 1 summarizes the level of intervention by servers confronted by the "intoxi- cated" pseudopatron. One can see that `in the best of circumstances (the Michigan treatment post period), the pseudopatrons were served 72 percent of the time with no intervention of any kind. On the other hand, outright refusal of service jumped from 3 Table 1. Distribution of server action by experimental group, site, and period (in percents) Server action Louisiana Michigan Treatment Control Treatment Control Pre Post Pre Post Pre Post Pre Post N=62 N=62 N=127 N=126 N=63 N=61 N-141 N=135 Service, no intervention 92 90 93 85 87 72 86 86 Service, status 3 3 4 2 3 7 6 5 Service, alternative 3 3 3 10 3 0 3 4 Service, slow 2 0 0 0 3 2 3 0 Service, transport 0 0 0 0 0 0 0 1 Service, final 0 0 0 0 0 3 1 0 No service 0 3 0 2 3 16 2 4 Source: M&night 1987. 176 BACKGROUND PAPERS Table 2. Mean intervention level by experimental group, site, and period for collapsed intervention score Louisiana Michigan Time Treatment Control Both Treatment Control Both Total Pre .08 .07 .07 .16 .16 .16 .l2 Post .l3 .17 .16 A4 .18 26 21 Diff .05 .lO .09 23 .02. .lO .16 Source: McKnight 1987. to 16 percent in the same group of establishments. While the Louisiana treatment group was more likely to refuse service after the training, so was the control group. To test the outcomes for statistical significance, the author collapsed the intervention levels into "service without intervention, " "service with some form of intervention," and "no service," with scores of 0, 1, and 2, respectively. Table 2 shows the resulting differences across sites and conditions. An analysis of variance (ANOVA) showed that the program produced in a significant increase in interventions in Michigan, but not Louisiana. Unaccountably, the Louisiana control group's increase in intervention was greater than the treatment group, primarily in their suggesting alternatives. M&night concluded that the program can improve knowledge and attitudes and can produce a small increase in interventions, but, depending on situational variables, changes in management policy may be small and limited, and finally, the type of establishment influences the program's chance of success. In particular, the program seems to be most successful in places with a smaller volume of sales, or that serve affluent clientele. The latest reported evaluation of a responsible beverage service program was con- ducted by researchers of the Addiction Research Foundation in Toronto, Canada (Glicksman and Single 1988; Simpson et al. 1986). Here, manager and complementary server training courses were given in four different types of establishments in Thunder Bay, Ontario, with four other sites used for comparison. Managers and owners were trained to implement specific policies of which servers were aware and which would reinforce the desired serving practices. Training for servers included such topics as serving and the law, health, preventing intoxication, and managing intoxicated persons. The training emphasized clear and concise steps for servers to take. The program also set a limit on the number of drinks a customer could have. A 35-item true/false test was used along with three open-ended items to assess changes in trainees knowledge of appropriate serving practices. A t-test showed sig- nificant increases in both portions of the test, with true/false scores improving from 24.1 to 30.2, and the open-ended items from 1.3 to 5.3 (out of a perfect score of 11). The study also adapted and expanded on the pseudopatron approach used in the Geller and M&night studies by devising seven alternative scripts for the pseudopatrons, covering different situations that would require intervention, as follows: Being too "young" to be served Ordering too many drinks at once Ordering too often Displaying drunken behavior and disorderly conduct T~NSPORTATION AND ALCOHOL SERVICE POLICIES Displaying drunken behavior but quiet conduct Preparing to drive home when obviously intoxicated Ordering drinks when intoxicated The research team also constructed a 1Zpoint scale of server responses to these incidents, ranging from a -6 for unsolicited service of more alcohol (when service should be denied) to + 6 for calling the manager over. Intermediate scores were assigned for responses that fell in between-for example, a -1 for ignoring the customer and a +2 for commenting on the pseudopatron's behavior. If the server's actions involved more than one response, the scores were added together. Pre- and post-measures of knowledge about alcohol and good serving practices showed a significant improvement among those who had been trained, and measures of receptivity to the training were also positive. ANOVA was conducted with three pairs of matched bars (the fourth pair had to be eliminated because of untrained staff in the experimental site) using a 2 X 2 repeated measures design using time (pre versus post intervention) and group (experimental versus comparison). The dependent variable was ' the server's response scores (with a constant added to make all scores positive). As with the Navy Server study, results showed both a time effect and an interaction of time and group. The mean score of the comparison sites increased slightly from about 16.3 to 16.9, while the experimental sites' mean score rose from about 15 to over 21 (see figure 2). It seems clear that the trained servers had moved toward more appropriate responses to the problematic scenes acted out by the pseudopatrons. 22 r 21 - 20 - 19 - 18 - 17 - 16 - Experimental Comparison I I I Pretest Posttest Time 177 Figure 2. Server behavior- interaction of time by group Source: Gliksman and Single 1988. 178 BACKGROUND PAPERS As with the other evaluations, only short-term effects were measured, and again, one cannot tell from the report how widespread the interventions were across trainees. Since the server could respond in several ways, it is theoretically possible with an additive score for a few servers to have exceptionally high scores while others remained unchanged, resulting in an overall gain in the mean intervention score. In summarizing the existing research, we should point out that these studies differed in their aims. The TEAM study did not employ a strict evaluation design partly because the researchers felt it was not appropriate for program development, but also because they wished to remain open to any opportunities to get a "feel" of how the program was working. The Navy study was trying to measure the impact of a comprehensive program that included more than server training on customer consumption, and thus, did not focus on the servers' intervention so much as on whether the overall program both reduced customer demand and limited the supply of alcohol. The TIPS evaluation and the M&night study, on the other hand,,were explicitly concerned with whether the server training had increased the likelihood of intervention by the server directly. It is unclear why the TIPS program would have seemed somewhat successful despite having only some of the servers trained and no particular management support, while the M&night program had a limited impact in only one of the two States. Perhaps the difference was due to the different definitions of "intervention." For the TIPS study, mild forms (comments, offering food and water, etc.) were weighted alike, whereas for M&night's analysis, mild forms were scored lower than refusals. On the other hand, the TIPS pseudopatrons did not necessarily show overt signs of intoxication as did the other study's staff. One might guess that refusal of service to obviously intoxicated customers would be one of the easier objectives to achieve in the training. Research Recommendations Obviously, we have only just begun to explore this promising avenue for prevention. The studies summarized above were designed to assess the potential for server interven- tion. There are, of course, a host of specific questions remaining regarding the proper emphasis for such prevention strategies and questions regarding the social and legal environments that may encourage the intensity and growth of responsible beverage service. Among these research questions are the following: o Trainingcuniculum. How much emphasis is needed on "affective" topics versus specific skills in intervention? Which modes of training (e.g., lecture, videotape, group discussion, role play) are best suited for each topic in the curriculum? How long must the training be? Who should be trained? What kind of followup training is required and how often should it be offered? o Establishments. What program modifications are necessary for very large or small businesses? Should the program be tailored for different clientele (e.g., upscale versus casual). o Management policies. Which policies and practices should be considered the minimum necessary to create an environment conducive to the prevention aims of the training'? Which specific practices pose the largest risks for intoxication (e.g., happy hours or other promotions)? What is the impact of patron educa- tion? o Social and legal environment. What role does dram shop liability play in en- couraging effective programs? What is necessary for insurance companies to offer meaningful discounts to businesses that participate in responsible beverage service programs? TRANSPORTATION AND ALCOHOL SEFIVICE POLICIES Summary 179 It should be clear that much remains to be done to refine the design and implemen- tation of server intervention or responsible beverage service programs. While current results are somewhat mixed, there does seem to be an opportunity to reduce the risk of intoxication, or at least the level of intoxication, among customers at licensed establishments. Obviously, research and evaluation of server intervention or responsible beverage service is in its infancy. While we now have reason to believe that server intervention can reduce intoxication and subsequent alcohol-impaired driving, the results are mixed, especially regarding the size of that impact. When results differ, we naturally turn to questions about the nature of the programs being evaluated and how they were imple- mented. Further research can take the materials that were developed in the programs designed to date, compare their features, and begin a systematic exploration of which features should be kept and which discarded, which methods are best suited for deliver- ing those elements, and what situational and environmental influences help or hinder an effective program's implementation. REFERENCES Chafe&, M.E. Training in intervention procedures: A prevention program. Abstracts: Review of Alcohol Md Driving 5: 17-19,1984. Colman, V.; Well, B.; and Mosher, J.M. Preventing alcohol-related injuries: Dram shop liability in a public health perspective. Western StateLaw Review 12:417-517, 1985. Geiler, E.S.; Russ, N.W.; and Delphos, WA, Doesserver intervention make a difference?AZcohol Health and Research World 11(4):6469,1987. Gliksman, L., and Single, E." A Field Evaluation of a Server Intervention Program: Accommodat- ing Reality." Paper presented at the Canadian Evaluation Society Meetings, Montreal, Canada, May, 1988. M&night, A.J. Development and Field Test of a Respkible Alcohol Service Progrant Voiume I: Research Findings. Report No. DOT HS 807 221, Washington, DC: National Highway Traffic Safety Administration, Department of Transportation, 1987. Mosher, J.M. Dram shop liability and the prevention of alcohol-related problems. Jomal of Studies on Alcohol 40:773-798, 1979. Mosher, J.M. Senter intervention: a new approach for preventing drinking driving. Accident Analysis and Prevention 15:483-497,1983. Mosher, J.M. Server Intervention: Present Status and Fuhue Prospects. Berkeley, CA: Prevention Research Center, 19840. Mosher, J.M. Legal Liabilities of Licensed Alcoholic Beverage Establtihments: Recent Develop- ments and Policy Zmplications. Berkeley, CA: Prevention Research Center, 1984b. Mosher, J.M. The impact of legal provisions on barroom behavior: Toward an alcohol-problems prevention policy. Alcohol 1:205-211, 1984.c. National Highway Traffic Safety Administration. TEAM: Techniques ofEffectiveA&ohofManage- ment: Findings from the First Year. Washington, DC: National Highway Traffic Safety Administration, Department of Transportation, 1986. O'Donnell, M. Research on drinking locations of alcohol-impaired drivers: Implication for preven- tion policies. Journal of Public Hen&h Policy 6:510-52.5, 1985. Russ, N.W., and Geller, E.S. Training bar personnel to prevent drunken driving: A field evaluation. American Journal ofPublic Health 77:952-954,1987. Saltz, R.F. The roles of bars and restaurants in preventing alcohol-impaired driving: An evaluation of server intervention. Evaizuztion and the Health Professions lO( 1):5-27, 1987. Segal, R., and S&on, B. Medical complications associated with alcohol use and alcoholism. In: Bratter, T.E., and Forrest, G., eds. Alcohol&m and Substance Abuse. New York: Free Press, 1985. Simpson, R.; Stanghetta, P.; Brunet, S.; Single, E.; Solomon, R.; and Van de duet, W. A Guide to the Responsible Service ofAlcohol. Toronto: Addiction Research Foundation, 1986. 180 Injury Control Injury and Disability Prevention and Alcohol-Related Crashes Julian A. Waller, M.D., M.P.H. Depa&rmnt of Medicine, University of Vermont, Burlington, Vermont This chapter is built around four assumptions. First, contrary to the general impres- sion of the public, and even of many working in the field of highway safety, about 80 percent of the people fatally injured in alcohol-related crashes are individuals who themselves had been drinking and are either drivers, impaired passengers of alcohol- impaired drivers, or impaired pedestrians (Waller 1985). About two-thirds of injuries in alcohol-related highway crashes also involve such individuals. The image of the innocent victim who is run down by a drunken driver who escapes all injury himself has great emotional impact but only limited support from the real world. We have a responsibility to be fully aware that injury to the innocent is not the largest aspect of the problem. Second, the tendency over the years has been to see issues involving alcohol and injury primarily as a safety problem, rather than a public health concern. As work over the past two decades has shown, reinforced by the recent National Academy of Sciences report I+lryirr America, injury is the primary cause of lost productive years of life in the United States, and alcohol abuse is an important contributor to that toll (National Research Council Committee on Trauma Research 1985). A very substantial proportion-probably a majority- of persons who get into trouble with alcohol on the highway are not typical social drinkers, but rather problem drinkers or alcoholics. The safety issue is only part of a much larger public health problem. In looking at the safety aspect, we must constantly be aware of this broader perspec- tive. For example, an important effect of increasing the age for alcohol consumption to 21 was the reduction in the highway crash toll among teenagers. But, although at least some short-term data suggest otherwise (Vingilii and Smart 1981), it may be that the most important long-term result of such legislation will be to reduce a wide range of alcohol-related problems among teenagers and the onslaught of teenage and young adult alcoholics that we began to see when the drinking age was lowered to 18. Third, during the late 196Os, Dr. William Haddon developed a matrix of highway crash analysis in which human, vehicular, and environmental factors interact during the precrash, crash, and postcrash phases to determine the occurrence, initial severity, and ultimate outcome of these events. He subsequently described generic interventions relevant to these phases in his brilliant paper, "On the Escape of Tigers" (Haddon 1970). In talking about injury prevention and amelioration, this chapter utilizes Haddon's basic approach, with an awareness of all three types of factors across all three phases of crash events. Finally, this review is clearly not alone in discussing components of Haddon's nine cells. Therefore, this examination is quite selective, focusing on aspects that will not be considered by others or that will be considered in a different context. INJURY CONTROL 181 Behavioral Issues in the Precrash Phase How do we alter behavior to prevent the occurrence of crashes? All the other panels are concerned with one or another aspect of this area. This chapter focuses particularly on three aspects. One is the characteristics of the populations we are trying to reach. As data expand, it is becoming increasingly clear that many people begin to be impaired at blood alcohol concentrations (BACs) below 0.05 gm/dl. But the data about who gets into trouble are equally clear. Over three-quarters of alcohol-related fatalities, and almost as high a proportion of alcohol-related injury crashes, involve drivers or pedestrians with BACs of 0.10 gm/dl or higher. Usually the BAC is much higher, averaging 0.16 gm/dl (NHTSA 1986). My own work and that of others has focused on two groups of heavy drinkers, namely, teenage and young adult males, and problem drinkers and alcoholics. A new group causing increasing concern is teenage and young adult females who are drinking more and driving more, at least in part as aspects of their new independence (Fell 1987). This discussion, however, focuses on a different issue, namely, the interdependence of lifestyle components. Studies show that those teenagers most likely to drive after drinking are also the ones most likely not to use seatbelts, to drive recklessly, to have less exercise, to prefer more fat in their diet, to have had premarital sexual experience, and to have done so without benefit of contraceptives (Jessor 1987; Maron et al. 1986). We know that heavy drinkers tend also to be heavy smokers, a fact not lost on the tobacco and alcohol beverage manufacturers, which often are divisions of single companies and thus are able to consolidate marketing strategies. A recent study by the Internal Revenue Service shows that individuals most likely to cheat seriously on their taxes are also more likely to be taking chances in other aspects of their lives, including speeding, drunken driving, adultery, and risky investments and sports (Goleman 1988). This is precisely why it is so important to take a broader public health viewpoint. Specific behavioral interventions often show at least partial success. But such interven- tions all tend to bog down on the fact that they have minimal effect on those hard core segments of the population that are most `overrepresented in whatever behavioral aspect they are trying to change. The problem, simply stated, is that we are just beginning to learn which behaviors interact, and how, so as to create that entity we called lifestyle. We know relatively little about how lifestyles are either transmitted or altered between generations or within generations, and whether it is possible to alter selectively one aspect of lifestyle, even if it creates dissonance with other aspects. Frankly, we need such information if we are to do more than simply nibble around the edges of behavioral modification. Alcohol- related behavior is far too intimately and intricately tied to other life beliefs and behaviors to permit us to expect success with simplistic approaches to modifying behavior, espe- cially as it relates only to the activity of driving. The foregoing and subsequent comments may sound like a plea for more research, and on one level they are. But the ultimate goal is to achieve intervention programs that work. In addition to basic research, we need to try out interventions in such a way that they can be adequately evaluated to answer the following questions, among others (Wailer 1980). o Does the program alter behavior, morbidity, or mortality overall? Are any other activities or events going on in the community that may explain any changes observed? For example, the rise of activities by MADD, SADD, RID, and other groups concerned with alcohol and highway safety coincided with the end of the baby boom, and it appears that at least a part of the reduction in alcohol- related crashes can be attributed to the reduction in young males, rather than 182 BACKGROUND PAPERS to specific anti-alcohol activities. This is not a criticism of the activities per se, but rather of the method of determining effect. . . o Is the programmore or less successful with some segments of the population, or under some circumstances, than with others? . What proportion of the total problem do-those segments account for? . What factors seem to explain the population or circumstance differences? Are the factors intrinsic to the population or &cum- stances itself, to the way the intervention was carried out, or to both? o Can the program be modified in a way more likely to reach the hard to reach, for example, by addressing more specifically those unique aspects of knowledge, attitudes, and practices that make them a more difficult target? a Given the availability of other nonbehavioral options for injury control, is the cost-effectiveness of the program sufficient to warrant its continuation, or are other options preferable? The methods for answering these questions must be built into the initial design of the intervention rather than being tacked on as an afterthought. This requires early involve- ment of someone with skills and knowledge in process and outcome components of program evaluation. It has been said that knowledge without action is futile; but action without knowledge is fatal. We need sound use of both. The third concern is about alcohol-impaired pedestrians, who account for 7 percent of all alcohol-related highway fatalities (NHTSA 1986) but a substantially higher propor- tion of those in urban areas. Most attention in the efforts to reduce the alcohol-related highway tolI has been paid to the impaired driver, as indeed it should. But we must not overlook the pedestrian. Efforts to educate people about alcohol-related risks should also mention the risks to this group of road users. An excellent study by Blomberg et al. (1979) in New Orleans showed that, unfor- tunately, it is not possible to identify high-risk locations where impaired pedestrians are more likely to be found. Nonetheless, some efforts aimed at high-risk populations may be of use, Efforts to provide alcohol in controlled settings on campuses, for example, may reduce the likelihood that college students will travel as either drivers or pedestrians after becoming impaired off campus. But apparently no evaluation of such activities has taken place. Clearly, more needs to be done in both the research and program areas. Vehicular Issues in the Precrash Phase There is a tendency to assign ail responsibility for a crash to the driver or the pedestrian if alcohol is present. But, as Patricia Wailer cogently notes (Haight et al. 1976). The fact that the human error involved in accidents is frequently related to information failure (including recognition errors) strongly suggests that the demands of the driving situation are more than the driver can handle. There is considerable need to recognize that the human being varies in his perfor- mance and that on the whole it can be assumed that he probably does about as well as he can be expected to, given the circumstances. Accident inves- tigations should be conducted in which the human element is taken as a given and the vehicle and environment are analyzed to determine the extent to which they need to be modified so that the human can function satisfactorily. Thus, simply because a vehicle is performing up to the manufacturer's standards does not mean that the vehicle performance is satisfactory. Per- haps the manufacturer's standards need to be modified and the vehicle performance enhanced. Simply because the roadway signing meets the criteria set down in a highway design handbook does not mean that the INJURY CONTROL 183 signing is adequate. The criteria often used call for signs that can be readily viewed by drivers with 20/30 vision. Furthermore, the standards must be met only by the signs when they are new. Most highway signs remain iu place for years, and many legally licensed drivers cannot meet a vision criterion of 20/30. Under such circumstances, when the driver fails to read the sign in time to make a decision, is it a driver failure? Most humari factors experts would not agree, yet accident causation studiespersist inperpetuatingthemyth that dnvers are somehow supposed to be able to compensate under any conditions for. the shortcomings (legalizedfaikres?) of the vehicles and driving environment. (p. 48-49) Ample evidence exists that vehicle design or other characteristics contribute to the occurrence of highway crashes. a Perhaps the most obvious example involves the motorcycle, which combines special problems with stability and handling, reduced conspicuousness so that . . other drivers are less able to avoid it and, once in a crash, reduced survivability for the occupants. Recent work from California and New York indicates that special training and licensing requirements for motorcycle drivers are inade- quate to compensate for these problems (Insurance Institute for Highway Safety). In 1986, fully 54 percent of motorcycle fatalities had been drinking (NHTSA 1986), but whether the unique features of the motorcycle exacerbate the effects of the alcohol impairment is not known. In other words, if a driver has a crash at a given BAC it is not known to what extent the vehicle handling characteristics versus the driver's handling capabilities contribute to the crash. o One modification of motorcycle design that has aided crash avoidance has been linking the ignition switch to lights so that motorcycles always operate with their lights on during the daytime, thus increasing their visibility. The addition of reflectors to bicycles also improves the nighttime visibility of these vehicles, thus giving a car driver impaired by alcohol additional time to react. o Issues of vehicle handling characteristics also are relevant to the crashes of certain utility vehicles that have. a propensity to roll over in crashes (Reinfurt et al. 1982). When faced with the rollover data, persons representing the automo- tive industry on a related lawsuit countered that the overrepresentation of rollovers did not reflect the vehicle characteristics per se, but rather the sorts of people who drive these vehicles, for example, the fact that such vehicles are often driven by the young.' As seen in figure 1, the propensity to rollover is associated with age. Both the young driver and the "older" driver are at higher risk of rollover in crashes for two of the three models examined. If the issue were solely driver characteristics, one would expect to see overrepresentation only of young drivers, but not of those who have reached the "senile" years of 35 or older! o But it is also known that the automotive industry- as does any industry- tries to target its buying audience by selective placement and design of advertise- ments to highlight vehicle characteristics that are likely to attract certain buyers. It is unlikely that many 45-year-old business executives drove around during the 1960s and 1970s in VW bugs, or currently use pickup trucks as their transpor- tation of choice. It is equally unlikely that many 19-year-old laborers ?? college students are in the market for a Mercedes or Volvo or Lincoln Continental, and advertisements for these vehicles that appear to be aimed at these audiences are seldom, if ever, seen. o How does all this relate to alcohol? A study from the General Motors research staff shows a strong correlation between "Youth Sport," "High Sport," "Sixties 1 Comments about the Reinfurt et al. paper by Kent Joscelyn and others during post-presentation question and answer period at AAAM meeting, 1982. 184 BACKGROUND PAPERS % 40 - 30 - 64 Jeep CJ-5 10 - +-----I, Ford Bronco A.. . . . . . . . A Chev. Blazer - All Utility Vehicles . I I I I I <20 20-24 25-29 30-34 >34 DRIVER AGE Figure 1. Utility vehicle rollover percentages by driver age for single vehicle crashes. Generation," and "Wild West/Cool Bravado" vehicles and likelihood of involve- ment or death in an alcohol-related crash, and either negative or no association for "Sedate, Self-restrained" or "Domestic Economy" vehicles (Angell and Von Buseck 1985). The authors of this research were not able to divulge which vehicles lit into such categories. The question, of course, is to what extent do handling characteristics of the vehicles involved make driving easier, more difficult, or have no effect for drivers with alcohol in their systems? o Issues related to braking provide an additional vehicular example. Recent improvements of braking systems in more expensive vehicle models through the use of computers are aimed at a longstanding problem of brakes that lock up and permit the vehicle to skid. In considering the anticipation or early iden- tification of a skid situation, Olsen (1978) has noted that suspension systems in many U.S. vehicles are sufficiently "soft" so the driver may realize he is begin- ning to skid when there is barely time, or it is already too late, to make an appropriate response. o Given the fact that alcohol alters perception, judgment, and slows response INJURY CONTROL 185 time, "soft" suspension and standard brake systems can only add to the demands of the driving task for a person who is already impaired. The recent installation of elevated rear brake lights is an important improvement to quickly alert drivers to potential danger and should especially benefit those with modest impairment by alcohol. o Side and rear view vision systems have been faulted in many vehicles, especially trucks, and in 1972 an estimated 900,000 crashes per year were attributed to inadequate vision on trucks, with 500,000 of those believed to result from poor but correctable mirror design (Reiss et al, 1972). Alcohol severely affects vision, including narrowing of visual fields. Inadequacies of vehicle vision systems, therefore, are likely to be even more of a problem for the alcohol-impaired driver than for the average driver. o Trucks present their own set of difficulties. In 1986, almost 5,100 medium or heavy trucks (over 10,000 lbs) were involved in crashes, in which 4,881 persons died. Only 16 percent of the fatalities were truck occupants (NHTSA 1986). ' Crash rates for trucks exceed those of other vehicles on all types of roads, including freeways, and when trucks account for 25 percent of traffic volume on an expressway system, a 3-percent increase in truck traftic produces a 23-per- cent increase in crash frequency per mile (U.S. Bureau of Public Roads 1963). Overwhelmingly, such crashes are more serious because the energy loadings are greater. o Other drivers find it particularly difficult to maneuver in the presence of large trucks, not only because they are behemoths, but because they often are traveling faster or slower than the traffic flow, may be poorly visible at night, create aerodynamic turbulence and, in bad weather, may obscure the vision of other drivers by splashing their windshields. Each of these characteristics adds to the alcohol impairments of the driver. Therefore, the very presence of more trucks on the roads quite likely exacerbates the problem of alcohol-related crashes. Add to this the safety effects of deregulation and of increases in truck size and double trailers that have been approved at the Federal level and it is clear why we have a serious problem. Roadway Issues in the Preinjury Phase The earlier quote by Patricia Waller is relevant to issues of highway design and construction as well as to vehicles. It is not enough to know that individuals crashed because they were impaired by alcohol, The question also is, if they were so impaired, why did they crash where they did and not at an earlier location along their path? The answer often is that only at this location did the driving task become too demanding. Much is already known about the types of roadway characteristics that increase demands on the driver and are associated with more crashes. In addition, some recent work has pinpointed the interreration between some of these factors and alcohol impairment. Here are a few examples. 0 Several studies done for the Insurance Institute for Highway Safety pinpointed roadway characteristics that are overrepresented in single vehicle crashes. One study in Georgia showed that such crashes are most likely to occur at or near curves of greater than 6 degrees, especially if associated with grades of 2 percent or more (Wright and Robertson 1976). These data, however, are not specific to drivers using alcohol, and it is reasonable to assume - but currently not known - that a person with a BAC of 0.10 gm/dl might be at greater risk at somewhat less demanding locations, e.g., curves of 4 degrees or greater with or without an associated grade. o Presumably, the higher the BAC the less demanding the roadway that will prove 186 BACKGROUND PAPERS excessive. However, since the Georgia study examined single vehicle crashes, which commonly involve alcohol, it may also be that unimpaired individuals tend not to crash until the road is much more demanding and, in fact, the combination of a 6-degree curve and 2-percent grade is a good measure of the point at which the demand becomes too great for the typical driver impaired by alcohol. This can only be determined, however, by a BAC-specific analysis of crash environment. o Increasing the speed with which a traffic light changes from green to red is associated with an increase in intersection crashes, whereas slowing the change cycle reduces the crash rate (Zador et al. 1985). A change of 20 percent, from lo-percent slower than that recommended in traffic management manuals to lo-percent faster, is associated with a fourfold difference in crash rate. Again, impaired drivers are uniquely likely to be affected by such changes because they have slower reaction times, especially for situations that involve making choices. a A report from the National Transportation Safety Board (1980) identified the State of Utah as having two to three times as many crashes per day of wet weather as do the surrounding States. It does not seem reasonable to assume that Utah has more careless or alcohol-impaired drivers than these other States. In fact, given its large Mormon population, one would anticipate less alcohol use. The difference appears to be attributable to the Utah roads, which are built with a dense aggregate so that water doesn't drain off readily and vehicles can hydroplane easily. While this situation has been documented for Utah, it is not known how often such differences in roadway construction contribute in extremely subtle ways to the occurrence of crashes, or to those involving alcohol. Police investigations simply take the roadway construction as an immutable fact and, if drivers skid, they assume it must have been something they did wrong, rather than consider- ing the possibility that something may also have been wrong with the way the road was built. o Two cases provide information about the effect of roadway characteristics on drivers at different BACs. One study indicated that drivers at BACs of 0.05 gm/dl, and, to a lesser extent at 0.10 gm/dl, are less likely to wander off the sides of roads or across the center line if the painted stripes are 6 or 8 inches wide instead of the usual four inches (Nedas et al. 1982). Test areas of roadway with wider side and center stripes had significantly fewer crashes than did standard roads. o Other work shows that putting reflectors on signs or on pedestrians permits drivers with moderate BACs to recognize these features from a further distance (Hazlett and Allen 1969). Since the inadequacies of road sign placement, size, readability, and information presented have been well documented, such greater visibility and consequently longer time for the impaired driver to react can only be helpful (Tamburri 1969). Studies of wrong-way driving by alcohol- impaired drivers about 20 years ago identified problems of poor signage and inadequate separation of some on- and off-ramps on freeways and resulted in important and successful corrective actions (Alcohol and Highway Safety 1968). The Missing Link in Alcohol-Related Crash Avoidance Research Excellent information has existed for at least the past two decades about the relative risk of crashing that a driver faces at given BACs (Jones and Joscelyn 1978; Haight et al. 1976). Information is also available about the crash contribution, and sometimes even the relative crash risk, of certain vehicular or roadway characteristics. What is not known is how the three sets of data interact. To date, no study has specifically examined the INJURY CONTROL 187 interaction of alcohol, roadway, and vehicular characteristics in crashes without making the methodologic error of assuming that the roadway and vehicular aspects were functioning optimally if they were working as designed and constructed. The fact that they may have been designed or constructed so as to increase crash risk has been considered irrelevant by prior researchers. Again quoting Patricia Wailer on this issue (Haight et al. 1976), It is a serious error to consider the vehicle and environment as noncon- tributory simply because they meet the currently established standards. Most drivers also meet the currently established standards. After all, driver licens- ing programs exist precisely to see that this is the case. Critics may quickly point out that our licensing standards are not adequate, but cannot the same criticism be made with equal validity of the standards concerning vehicles and roadways? Why are these standards somehow sacrosanct while the standards used to license drivers are so readily the object of criticism? If the driver in a crash holds a valid license, then the argument can be made that he, too, has met existing standards and can be held no more culpable than the inadequate signing that is nonetheless legal. (p. 31) Such unnecessary additions to the demands on the driver or the pedestrian, of course, create a problem not only for the person impaired by alcohol, but also for the elderly, those just learning to drive, persons unfamiliar with the specific road, and all the rest of us at times when traffic is particularly heavy, weather is bad, or when, for whatever reason, we happen to be functioning at less than our personal optimum capacity. If the goal is truly the prevention of crashes, rather than simply the excoriation of those who consume alcohol, we will look to all possible options in crash prevention, especially if they have the wider benefits described above. Human Issues During the Crash Phase A common axiom says, "Drunks don't get hurt when they fall because they are so relaxed." Unfortunately, this assumption not only is inaccurate; it is the exact opposite of what occurs. Both laboratory and epidemiologic evidence has been growing rapidly documenting that, especially at lower levels of kinetic impact, persons who have con- sumed alcohol are more likely to be injured, to have severe injury, to die, and to die before definitive treatment can be brought to bear (Waller 1985). The reason is that alcohol adversely affects both the internal function of individual cells and the functioning of organ systems to make it easier for them to become injured and more difficult for them to respond properly to injuries that do occur. Clearly, this myth of invulnerability for alcohol-impaired persons needs to be dispelled. Vehicular Issues During the Crash Phase The past two decades have seen considerable attention to, and correction of, vehicular components that increase crash severity. Examples include laws requiring seatbelt use and the move toward wider availability of airbags, collapsible steering columns designed to absorb damaging kinetic energy before it reaches the driver's chest, development of high-penetration windshields, dashboard construction that is less hazardous on impact, removal of pedestrian-spearing hood ornaments and tail tins, and attempts to alter the front design of vehicles to reduce pedestrian damage on impact. Some major hiatuses still exist, however. Attempts to reduce the effects of side impacts on vehicle occupants are as yet unsuccessful. Most States have not reinstated laws requiring use of helmets by motorcyclists, despite the documented savings of life, limb, and money that accompanied initial passage of such laws and the loss of such savings where these laws were repealed. 188 BACKGROUND PAPERS Unfortunately, with both helmet and seatbelt laws, among those least likely to obey the law are persons under the influence of alcohol (e.g., Jessor 1987; Maron et al. 1986). Research has shown that excessive nonuse of seatbelts is to be found as often among persons entering bars as among those leaving and, therefore, greater nonuse is not an acute effect of the alcohol per se but rather appears to reflect underlying lifestyle issues (Preusser et al. 1986). Despite this relatively greater tendency toward nonuse, however, persons using alcohol more often wear helmets and seatbelts where the law rquires them than where use is voluntary, indicating that this high-risk group may be harder to reach, but not impossible to reach. Another major hiatus is the faihue to improve truck de&n or construction to reduce crash severity. It has been known for years, for example, that sides and rear ends of trucks can be constructed to prevent underride by automobiles during crashes and, conse- quently, to avoid decapitation of automobile occupants. But the trucking industry, to date, has successfully avoided regulations that would require this small addition to truck weight (Minahan and O'Day 1977). The industry may have gained, but public safety has lost. In similar fashion, energy-absorbing features universally found in automobiles are often not required by NHTSA in pickup trucks or various recreational vehicles, which are probably the vehicles preferred by high-risk young, heavy-drinking males. Roadway Issues During the Crash Phase The past two decades have seen both substantial progress and continuing problems in roadway safety. Energy-absorbing construction of guardrails, bridges, and under- passes, and of light- and sign-posts has become far more common, especially on primary roads built for higher speeds. These changes reflect the decades-old data indicating that such design and construction reduce crash severity. Little of this, however, has been applied to secondary roads, and in many places trees, rock outcroppings, and telephone poles remain - or in some cases continue to be planted or placed- dangerously close to the roadway, even for roads that have posted speed limits of 50 mph or higher. One continuing problem is that even the best energy-absorbing devices that were designed for one generation of vehicles may be inappropriate for subsequent generations of vehicles that might follow during the expected lifetime of these devices. For example, devices built to deal effectively with crashes of the large and mid-size cars that preceded the 1973 oil embargo are no longer adequate for the large population of compact and subcompact automobiles, especially as these share the roadways with more and larger trucks. Again, the Missing Link Alcohol increases injury severity, especially at lower rates of energy exchange. AU the vehicular and roadway research and designs for alteration of energy transfer to date have been based on a presumption of normal tissue response. It is not known to what extent, if at all, assumptions need to be altered in considering the reduction of injury for alcohol-impaired persons or, as well, for the elderly, rather than just for unimpaired young individuals. Issues During the Postcrash Phase Only hvo issues during the postcrash phase are considered, namely, the acute manage- ment of injured persons who have consumed alcohol, and the application of rehabilita- tion concepts. Treatment of underlying alcoholism postinjury is also an extremely important issue, but it is being considered by another panel. Before turning to alcohol-specific issues of acute management, a brief comment should be made about the general status of trauma care systems in the United States. INJURY CONTROL 189 Over the past two and a half decades, substantial improvements have occurred in the trauma care system. Modified hearses and police cars to transport the injured have been replaced by modular, well-equipped ambulances. The SO-hour or beyond emergency medical technician (EMT) training has replaced the &hour standard Red Cross course. Ambulance-to-hospital radio communication has become commonplace. Emergency department nurses and many physicians are being trained in advanced life support, and the whole specialty of emergency medicine has developed. What is still missing is the widespread regionalization of such improvements, espe- cially as they involve cooperation between ambulance services, between ambulances and hospitals and, most assuredly, between hospitals themselves. A tremendous economic power struggle is taking place in many States concerned with designating specific hospitals as trauma centers because they are better equipped and better staffed, while neighboring hospitals perceive them as likely to "steal" lucrative cases. Until we see a system of regionalization of trauma care in every State of the Nation, instead of the current spotty distribution of such systems, we will not be able to say that this Nation's acute care system is beginning to be adequate. We turn now to alcohol-specific issues in acute care. As long ago as 1928, Bogen documented the need to determine whether a person had consumed alcohol in order to provide effective emergency care. It is becoming increasingly apparent that the com- petent physician must determine if the individual is likely, not only to be acutely under the influence of alcohol, but also to be an alcoholic. Acutely alcohol-impaired patients are more likely to have cardiac arrhythmia that may be lie-threatening either at the crash site or shortly after arrival in the emergency department. They may have greater respiratory distress both because of the acute effects of alcohol and because they are likely to be smokers. Alcohol may cause severe hypoten- sion. The presence of alcohol may, on the one hand, cause serious masking of symptoms of intra-abdominal injuries because of clouded sensorium and, on the other hand, result in overdiagnosis of the severity of head injury because alcohol-related altered conscious- ness is confused with trauma-related effects. If patients are alcoholics, their wounds are more likely to become infected, because alcohol depresses immune response. But, at the same time, sudden fever may not be a sign of infection, but rather of impending delirium tremens, which carries a 25-percent mortality rate. Many physicians have been concerned about their legal liability if they do a blood alcohol determination without specific patient approval. It is becoming increasingly obvious, however, for all of the above reasons, that the physician who does not obtain a BAC for clinical management is courting a suit for malpractice if problems occur that might have been foreseeable had the BAC been known. Lastly, attention must be turned to the subject of rehabilitation. The whole field of rehabilitation medicine developed as a result of the progressively higher survival rates of severely injured soldiers that was achieved during World War II and the Korean Conflict. The years since have seen tremendous advances in basic research in this field as well as in the design and development of new techniques, skills, equipment, and prostheses and the formation of rehabilitation teams. Unfortunately, as was pointed out in the National Research Council (1985) report, Zujury in America, much of what is known is not being applied. Patients may not be referred to a physiatrist, or the referral may be sufficiently delayed so that disability is unnecessarily prolonged or may even have become permanent. Complete rehabilitation centers are few and far between, and many physicians do not know how to access them. This is especially true for the services needed for severe head injury, which is a more common outcome in the presence of alcohol. The plea of this chapter, therefore, is not so much for new research, but rather for the wider application of what is already ~IIOWLL 190 BACKGROUND PAPERS Summary Alcohol contributes not only to the occurrence of crashes but also to the initial severity of injuries, problems in treatment, and ultimate outcome. The exact relationship of alcohol use and abuse and other lifestyle issues is just beginning to be examined, and the dearth of knowledge remains an important obstacle to achieving behavioral change for drivers and pedestrians who drink. Characteristics of the vehicle'and the road environment also contribute to crash occurrence and crash severity. Because littIe research, either qualitative or quantitative, has examined the interaction of vehicular and road environment characteristics in crashes at various blood alcohol concentrations, the exact relationship remains conjec- ture, and potential improvements in injury and disability prevention may be missed. Specifically, the following are recommended: . Research needs to be carried out and the results used to intervene in the interaction between alcohol use practices and other aspects of lifestyle. o Persons knowledgable in evaluation concepts and methods must be utilized early in the design of intervention programs, and evaluation should be an integral part of all such programs. 0 Research should be undertaken into the interaction between different BACs, vehicle handling characteristics, and aspects of road design, construction, and maintenance to determine which vehicular and roadway aspects exacerbate the effects of impairment caused by BACs below 0.10 gm/dl, and to set vehicle and road standards that will take this knowledge into consideration. Similar re- search is needed to determine quantitatively the relation between BAC and vehicle and roadway features relevant to injury severity, and to apply such knowledge for injury reduction standards. 0 Motorcycle helmet laws need to be reinstated in those States where they were rescinded.. 0 Federal crash and injury prevention standards currently applicable to automobiles should be extended to pickup trucks and recreational xhicles. Rear and side guards on trucks should be mandated to prevent other vehicles from underriding medium and large trucks in two-vehicle crashes. 0 All necessary steps must be undertaken, including legislation if nxzary, to ensure that regionalization of adequate trauma systems is achiexd throughout the Nation. o Physicians must be educated through hospital quality assurance m and other means about the need to determine BAC and to screen for &&ism as part of the proper management of the trauma patient. Sii -onal methods should be used to get physicians to seek prompt and Lzpriate assistance for the rehabilitative aspects of trauma care. REFERENCES ,Q&ol and Highway Safety. Report to the U.S. Congress. Washington, DC: L'L ant of Transportarion, 1968. Angel& L.S., and Von Buseck, C.R. Anexploratorystudyofvehicle type inalab%-zcrashes. In: Evans, L., and Schwing, R.C., eds. Human Behavior and Traffic Q@- X&u York: Plenum Press, 1985. pp. 285-328. Blomberg, R.D.; Preusser, D.F.; Hale,k; and Ulmer, R.G.A Comparison ofk.ktil in Pedestrians and Pedeshian Casualties. Final Report on Contraa DTT-Z%&XMfj. Norwalk, (=T: Dunlap and Assoc. East, 1979. Bogen, E. Drunkenness.American Journal of Medical Science 176:153-167,2X Fell, J.C. Alcohol involvement rates in fatal crashes: A focus on young drive axe izze tiers. INJURY CONTROL 191 In: Proceedings of 3Ist Annual Conference, American Association for Automotive Medicine. New Orleans, m the Association, Sept. 28-30, 1987. pp. I-30. Goleman, D. The Tax Cheats: Selfish to the Bottom Line. NY 7IIs, April 1 l:Al, D2,1988. Haddon, W., Jr. On the escape of tigers: An ecologic note. Technological Review 72:44-53,197O. Haight, F.k; Joksch, H.C.; O'Day, J.; and Wailer, P.F. Review of Methods for StudyingPre-Crash Factors. Report HSRC-PR38. Chapel Hill, NC: Highway Safety Research Center, Univer- sity of North Carolina, 1976. pp. 31,48-49. Hdett, R-D., and Allen, MJ. The ability to see a pedestrian at night: The effects of clothing, reflectorization and driver intoxication. American Jountal of Optometry and Archives of rhe American Academy of Optometry 45:246-258,1%9. Insurance Institute for Highway Safety. Two Studies Question V&e of Motorcycle Licensing ProgTam. Status Report 23(8):1,2,6. Washington, D.C.: the Institute, August 13, 1988. Jessor, R. Risky driving and adolescent problem behavior: An extension of problem-behavior theory. Alcohol, Drugs, and Driving 3(3-4): l-l 1, July-Dec., 1987. Jones, R.K., and Joscelyn, KB. Alcohol and Highway Safev 1978. A Review of the State of Knowledge. Report DOTHS-803 764. Ann Arbor, MI: Highway Safety Research Institute, University of Michigan, 1978. Maron, D J.; Telch, M J.; Killen, J.D.; et al. Correlates of seatbelt use by adolescents: Implications for health promotion. Preventive Medicine 15614623, 1986. Minahan, DJ.; and O'Day, J. Fatal Car-into-Ttuck/Trailer Underruie CoUi.rions. HSRI Research Review 8(3) Nov./Dee. 19771-16. Ann Arbor, MI: Highway Safety Research Institute, University of Michigan. National Highway Traffic Safety Administration. Fatal Accident Reporting System 1986. Washington, DC: U.S. Department of Transportation, 1986. National Research Council Committee on Trauma Research.Injwy in America. Washington, DC: National Academy Press, 1985. National Transportation Safety Board. Special Study: Fatal Highway Accidents on Wet Pavement - The Magnitude, Location, and Characteristics. Report NTSB-HSS-80-1. Washington, DC: the Board, 1980. Nedas, N.D.; Balcar, G.P.; and Macy, R.P. Road markings as an alcohol countermeasure for highway safety: Field study of standard and wide edgelines. Transportation Research Record 847:43-46,1982. Olsen, R.k The driver as cause or victim in vehicle skidding accidents. Accident Analysti and Prevention, lCk61-68, 1978. Preusser, D.F.; Williams, AF.; and Lund, AK Seatbelt use among New York bar patrons.Jountal of Public Health Policy 7:470479, 1986. Reinfurt, D.W.; Li, LX.; Popkin, C.L.; O'Neill, B.; Burchman, P.F.; and Wells, J.K Rollover and serious driver injury differences among various utilityvehicles, pickup trucks and passenger car group&In: Proceedings of 26thAnnual Conference, American Association forAutomo- five Medicine. Ottawa, Ontario, Canada: the Association, Ott 44297-312, 1982. Reiss, M.L.; Lunenfeld, H.; and Morton, G.W. Field of view Requirements Directly Behind Trucks and Buses. Contract DOT-HS-112-1- 162. Deer Park, NY: AIL, Division of Cutler-Ham- mer, 1972. Tamburri, T.N. Wrong- Way DrivingAccidentsAre Reduced. Highway Research Record 292:24-50. Washington, DC: Highway Research Board, 1969. U.S. Bureau of Public Roads. Traffic Control and Road Elements. Washington, DC: the Bureau, 1%3. Vingilis, E., and Smart, R. Effects of raising the legal drinking age in Ontario. Britain Journal of Addictions 76:415-424, 1981. Wailer, J.A. A systems model for safety program evaluation. Accident Analysis and Prevention 121-5, March 1980. Wailer, J.A. Injury Control. Lexington, m Lexington Books, 1985. pp. 125-128. Wailer, J.k Research needs and opportunities concerning human-environmental interactions in crashes involving alcohol. Joumal of Studies on Alcohol Supplement 10:54-60, 1985. Wright, P.H., and Robertson, L.S. Priorities for roadside hazard modification. TraJic Engineering 46(8):24-30, 1976. Zador, P.; et al. The effect of signal timing on traffic flow and crashes at signalized intersections. Transportation Research Review. lOlfk1-8, 1985. 192 Youth and Other Special Populations Youth Impaired Driving: Causes and Cotintermeasuies Michael Klitzner, Ph.D. Pacific Institute for Research and Evaluation Traumatic injury is responsible for more deaths among American adolescents and young adults age 14-24 than all other causes combined (Paulson 1983). Far and away the leading cause of traumatic injuries is traffic crashes (Robertson 1981; Lewis 1987). Traffic crashes have been cited as the cause of about half of all accidental deaths in adolescents and young adults and have also been cited as the cause of half of all spinal cord injuries (Robertson 1981). Not only adolescent drivers, but also their passengers (who tend to be adolescents) are at significantly increased risk when compared to older age groups (Insurance Institute for Highway Safety 1984). The exact contribution of alcohol use to youth traffic crashes has been debated (e.g., Zylman 1973; Cameron 1982). However, the conclusion appears inescapable that alcohol is a major causal factor (Cameron 1982, Lewis 1987). Young drivers are overrepresented in alcohol-related fatal crashes even when driving exposure is controlled (Vegega 1984). Although teen alcohol-related traffic fatalities steadily decreased from 1982 to 1985, they increased again in 1986 to a level just below that of 1983 (DOT 1987). Unpublished Fatal Accident Reporting System data for 1987 suggest another downturn, but overall death rates still exceed those observed in 1985. The reasons for this downturn are not clear but may reflect increasing public awareness and intolerance and recent changes in legislation. The most compelling evidence for a causal link between alcohol and youthful crash involvement comes from studies of changes in minimum alcohol purchase age (e.g., Fell 1988; Smith et al. 1984; Hingson et al. 1983; Wagenaar 1982u, b, 1983b). Although results have varied from State to State and from study to study, consistent reductions in youthful crash involvement have been observed following increases in the minimum purchase age. Interestingly, most studies that examined the effects of purchase age changes failed to find meaningful differences in youthful alcohol consumption (Moskowitz 1989). This may simply reflect the difficulties associated with measuring changes in consumption as opposed to changes in crash rates. Alternately, it may be that the relationship between minimum purchase age and crash involvement is not mediated simply by consumption, but rather reflects more complex changes in youth drinking patterns and drinking locations. Crash data reflect only a small segment of the youth cl&king/driving problem. Overall, youthful drivers are much more likely than their older counterparts to report driving after drinking (Hingson et al. 1988). Recent survey data gathered from the Nation's high schools (Bachman et al. 1987) revealed that approximately one in four seniors had driven after drinking in the 2 weeks predating the survey, and approximately one in six had driven after having live or more drinks in a row. During the same 2-week period, two in five seniors had ridden with a drinking driver, and one in five believed the driver had consumed live or more drinks. Driving while impaired (DWI) and riding with YOUTH AND OTHER SPECIAL POPULATIONS 193 impaired drivers (RWID) would appear to be a regular occurrence for a significant minority of American youth. Causes and Correlates Any successful attempt to reduce the extent of impaired driving and riding with impaired drivers among youth must be rooted in an understanding of the factors that predispose, reinforce, and enable these behaviors. However, current knowledge of the causes and correlates of youth DWURWID is incomplete. The majority of relevant studies have focused on alcohol consumption and related problems rather than on DWI per se, and only a very limited number of studies have focused on factors related to riding with impaired drivers. In addition, predisposing, reinforcing, and enabling factors have often been studied in isolation, complicating assessments of the relative contribution of different variables or classes of variables to DWI/RWID. Individual Characteristics Perhaps the largest body of correlate research has focused on characteristics of individual youths. These studies noted that personality factors such as aggressiveness, intolerance of authority, nonconformity, escapism, and immaturity may be associated with increased probability of driving after drinking (Lightsey and Sweeney 1985; Boyd and Huffman 1984; DOT 1975; Kraus et al. 1970). DWI has also been associated with poor academic performance, greater participation in social activities, access to cars including car ownership, more discretionary income, and working part time (Klitzner et al. 1987, 1988; Williams et al. 1986). Other individual-level studies have focused on the stresses of transition from adoles- cence to adulthood (Pelt and Schuman 1971) and 0" the relationship between stress and alcohol consumption (e.g., Fomey et al. 1984; Wagenaar 1983u; Koningsberg et al. 1983; Cameron 1982). A recent study of adolescent DWI offenders (Farrow 1987) suggested that offenders are more likely than nonoffinders to use risky driving as a stress manage- ment technique. Studies that assessed young people's awareness of the physiological and psychological effects of alcohol revealed that young people are generally ignorant of the effects of alcohol (Forney et al. 1984; Blane 1983; Hetherington et al. 1979) and are unable to identify the amount of alcohol that impairs driving performance (Pawlowski 1982). Several studies have explored the effects on DWI risk of positive attitudes toward drinking and drinking and driving. Most have focused on attitudes toward alcohol (Krohn et al. 1982; Milgram 1982; Lowrnan 1981; Douglass 1983; DOT 1975; Kraus et al. 1970). These studies suggest that normative acceptance of drinking by youth increases both alcohol consumption and DWI risk. Two recent studies by Klitzner et al. (1987, 1988) found that normative acceptance of DWI was also strongly related to both alcohol consumption and actual DWI/RWID behavior. Social Influences A second broad area of correlate research has focused on social influences, especially those associated with peer groups. Numerous studies have reported increased alcohol consumption among youth who associate with peers who drink and/or approve of drinking (Vejnoska 1982; Stoles and Fine 1981; Krohn et al. 1982; Nusbaumer and Zusman 1981; Biddle et al. 1980). A study by Finley (1983) implied that peer influence may be so pervasive as to negate the effect of countervailing influences such as fear of legal sanctions or parental disapproval. Group centeredness, a probable component of 194 BACKGROUND PAPERS susceptibility to peer influence, has also been found to increase DWI risk (Kraus et al. 1970; DOT 1975), and a recent study by DiBlasio (1988) found that peer modeling plays an important role in decisions to ride with intoxicated drivers. Jessor (1987) recently extended Problem-Behavior Theory (Jessor and Jessor 1977) to youthful risky driving, including DWI. Problem-Behavior Theory has a 25year history as a major theoretical orientation for understanding youth substance abuse and related problems. Jessor's recent work demonstrated that youth who are more influenced by friends than parents, and whose friends model risky driving behaviors, are more likely to report risky driving. Studies of social influence have also focused on the effects of mass media, especially alcohol advertising, on youth alcohol consumption and DWI. Most advertising research has examined general populations and has failed to find consistent effects (Frankena et al. 1985). Studies that looked specifically at youth (Atkin et al. 1983, 1984, Strickland 1983) reported possible effects of advertising exposure on both alcoholconsumption and DWI, but methodological weaknesses in these studies limit the strength of conclusions that may be drawn from them (Moskowitz 1989). Characteristics of Youth Drinking and Youth Driving A third broad approach to understanding the youth DWI problem has been to explore the special characteristics of youth drinking and youth driving. For young drivers, risk of crash involvement begins to increase at very low blood alcohol concentrations (BACs), and studies suggest that any measurable BAC can result in a significantly increased risk for younger drivers (Simpson et al. 1982; Farris et al. 1976; Perrine et al. 1971). Thus, the gap between risky and illegal BACs for youth in most States is large, and "safe" consumption guidelines publicized for adults may be dangerously misleading for youth. The more rapid impairment of the younger drinker is reflected in the fact that crash- involved adolescents are likely to have lower BACs than their older counterparts (Cameron 1982), and in the'higher risk of fatal crashes for young drivers when compared to adults with comparable BACs (Bergeron and Joly 1986). The simultaneous acquisition of driving skills and drinking experience may further increase the likelihood of crashes (O'Day 1970; Lewis 1987), and youth who DWI may tend to be riskier drivers in general (Bergeron and Joly 1986). Nataanen and Summala (1976) also noted the importance of considering the "extra motives" (beyond simple transportation) that driving may fulfil for youth. These include tension reduction, meeting the need for competition, showing off, and deliberate risk-taking. Summala (1987) found that these extra motives may be more important than lack of driving skill in contributing to poor youth driving performance. Research conducted in preparation for a National Institute on Alcohol Abuse and Alcoholism (NIAAA) youth and alcohol media campaign pointed to a number of structural and contextual factors that may serve to associate youth drinking with youth driving (URSmacificon 1980). These data suggest that, for many youth, the automobile represents the only place where privacy may be relatively certain. Drinking and other negatively sanctioned behaviors are most likely go undetected when undertaken in cars. Consistent with this assumption, data from the 1986 yearly survey of high school seniors (Bachman et al. 1987) revealed that over half of all seniors who drank had done so in cars, and approximately 28 percent reported doing so "some of the times" or "most of the times." Similarly, a national survey conducted by Grey Advertising (DOT 1975) revealed that among students who drank, 38 percent reported drinking "while driving around," and 20 percent reported drinking at drive-in movies within the previous 3 months. Driving constitutes a social occasion for youth, and the ride to and from a social event constitutes a prelude to and continuation of that event (Farrow 1987). Thus, drinking in cars may be a simple extension of other teen drinking. YOUTH AND OTHER SPECIAL POPULATIONS 195 The NIAAA planning data indicated that youth are more likely than adults to drink all that they possess at any given time, thus eliinating problems of storage or hiding of contraband alcohol. Moreover, data reported by Vegega and Klitzner (in press), Farrow (1987), and Bachman et al. (1987) showed that the great majority of teen drinking occured outside the home. Thus, the structure of teen drinking may lead to the consump- tion of large quantities of alcohol in settings that subsequently require some sort of transportation home. Multiple Correlate Studies In an effort to assess the relative contribution to DWI/RWID of a variety of risk factors discussed in the literature, Klitzner and colleagues (1988; Vegega and Klitzner in press) surveyed and/or interviewed a convenience sample of approximately 1,550 youth in grades 7 through college in six U.S. cities. In one study (Klitzner et al. 1988), 1,323 youth completed anonymous questionnaires that assessed lifestyle variables (friends' drinking ' practices, participation in parties and dates, access to cars), alcohol use variables, DWI/RWID risk factors, and self-reported DWI/RWID behavior. Of nine risk factors studied, only one -perceived deviance of DWI- was strongly related to DWI and RWID. Two other factors-use of alternative modes of transportation and decisionmak- ing skills-were related to DWI and RWID, but only insofar as they predicted drinking practices. The remaining six risk factors- awareness of alternative modes of transpor- tation, self-concept, communications skills, alcohol knowledge, knowledge of local DWI laws, and susceptibility to peer influence-predicted neither drinking practices nor DWI/RWID. Despite the failure of these risks factors to predict drinking or DWI/RWID directly, all nine risk factors were interrelated. Thus, the factors that did not directly predict drinking, DWI, or RWID may still contribute to overall risk. Drinking practices were themselves strong predictors of both DWI and RWID, a point discussed later. In a second study (Vegega and Klitzner in press), indepth interviews were conducted with 120 youth who reported DWI and 121 youth who reported having ridden with an intoxicated driver. This study focused on the contribution of situational factors to youth DWI/RWID. Among the factors assessed were social context variables; social pressures to drink, drive, and/or ride; perception of immediate risk; destination variables; and availability of alternative transportation. In general, the results showed that DWI/RWID is largely a function of the role alcohol plays in the youth culture. Many respondents suggested that DWI and RWID are "inevitable" because drinking is an "inevitable" component of adolescent lifestyles. Despite the current popularity of "peer pressure" as an explanation of youth drinking and related problems, only about 15 percent of Vegega and Klitzner's DWIs reported any pressure to drink, and only 13 percent reported any pressure to drive after drinking. Among RWIDs, less than 7 percent reported that peer pressure contributed to their decision to ride with the impaired driver. To the extent that DWI and RWID were situationally determined, they were controlled largely by a perceived need to get home or to get a passenger home. This finding is consistent with data reported by Farrow (1987), who found that home was the most common destination for youth engaging in a variety of risky driving behaviors, including DWI/RWID. Vegega and Klitzner described a special case of RWTD that occurred when the impaired driver was a parent or other adult relative. In this case, which represented slightly more than a quarter of the reported RWID incidents, the youths' apparent inability to affect parental DWI or to utilize alternative modes of transportation effec- tively precluded any protective action (other than fastening a safety belt) on the part of the affected youngster. Here, parents and other adult relatives appeared to make a significant contribution to the DWI-related risk experienced by youth. 196 BACKGROUND PAPERS Summary In general, the factors that predispose, reinforce, and enable youth DWI and RWID appear to be similar to those risk factors associated with other adolescent health risk behavior (Jessor 1987). Social and normative influences, risk-taking orientation, and individual differences in attitudes toward and beliefs about drinking and drinking and driving all appear to contribute to increased or decreased risk. Of particular import in considering DWI and RWID specifically, however, is the powerful role played by alcohol consumption per se in increasing risk of both DWI and of RWID (for which consumption is not a prerequisite). Indeed, Klitzner et al's (1988) data revealed that DWI/RWID risks increased directly and potentlyas a function ofboth quantity and frequency of alcohol consumption. Moreover, among Vegega and Klitzner's (in press) sample of DWIs and RWlDs, alcohol use was perceived to be an inextricable part of the youth culture, and DWI/RWID were viewed as "inevitable" results of the strong association between youth socializing and youth drinking. Thus, it seems unlikely that meaningful reductions in youth DWURWID can be realized without significant attention to changes in youth drinking practices. Countermeasures for Youth The past two decades have witnessed a rapid expansion in the number and types of programs and strategies employed to prevent youthful DWI and RWID. A review of 133 youth DWI prevention models (Vegega and Klitzner 1988) revealed enormous diversity of focus, underlying assumptions, and activities. Youth DWI/RWID countermeasures include school curricula, clubs, alternative transportation, alternative (alcohol-free) parties, teen retreats, andsyouth-focused legislation and regulation. Current DWI/RWID prevention strategies can be grouped into three major categories - those mainly concerned with the prevention of drinking, those mainly concerned with separating drinking and driving, and those concerned with preventing mortality and morbidity when and if DWb'RWID occur. The differences among these approaches can be illustrated by considering the natural history of DWI and RWID. Figure 1 presents, in highly simplified flow diagrams, the processes that lead to DWl/RWID and related mortality and morbidity. Figure 1. Processes that lead to DWVRWID NATURAL PROBLEM HISTORY OF YOUTH DWI Youth Who-l- > Youth Who-2- > Youth Who -3 - > Mortality & Don't Drink Do Drink Drink & Drive Morbidity NATURAL PROBLEM HISTORY OF YOUTH RWID Youth Who-l- > Youth Who-2- > Youth Who-3- > Mortality & Don't Associate Do Associate Ride With Morbidity With Drinkers With Drinkers Drinkers These flow diagrams indicate three points at which DWI/RWID strategies and programming can be directed. Point 1 represents strategies that have as their primary objective the prevention of youth drinking and the establishment of nondrinkinglifestyles among youth. Such programs attempt to alter the factors that either predispose, rein- force, or enable drinking among individual youth (e.g., school curricula, "say no" organizations, intervention programs for users) and attempt to reduce youth access to YOUTH AND OTHER SPECIAL POPULATIONS 197 alcohol (e.g., alcohol-free alternative parties, minimum purchase age increases, server training, limited outlets, education of retail clerks). Strategies at point 1 would not, of course, address the problem of youth who RWID when parents or other adults are the drivers. 1 Point 2 represents strategies that attempt to disassociate drinking and driving. Here, although youth alcohol use may still be of concern, the major objective is to address risk factors that lead drinking youth to drive, or that lead youth who associate with drinkers to be passengers. Examples of strategies at point 2 include SafeRides, designated driver, alternative transportation, direct intervention (e.g., taking keys), parent/student transportation "contracts," general and specific deterrence, and a variety of licensing strategies. Point 3 represents strategies that attempt to limit morbidity and mortality among drinking drivers, their passengers, and those with whom they crash. Examples of these strategies include passive restraints, other vehicle-;elated technologies, highway design elements such as breakaway sign posts, and so on. Point 1 Strategies Many point 1 strategies (those that attempt to reduce youth drinking) have been extensively studied. In particular, school-based strategies of various types have been the object of intense research scrutiny for at least two decades. Programs have been developed and evaluated that focus on arousal of fear of negative consequences, provision of information, development of "life skills" (e.g., positive self-regard, com- munication skills, assertiveness, decisionmaking, coping), clarification of values, and, most recently, "resistance" training. The literature on school-based alcohol and other drug prevention programs has been repeatedly reviewed (e.g., Moskowitz 1989; Klitzner 1987; Goodstadt 1985; Wittman 1982). In general, these reviews concur that evidence in support of school-based alcohol and drug prevention programs is sparse. Although increases in knowledge and changes in attitudes are often reported, effects on behavior have been weak, inconsistent, transient, and sometimes in the wrong direction. The failure to demonstrate educational program effects has been attributed to failures in program models, to failures in program implementation, and (more optimistically) to weak or inappropriate research designs (Klitzner and Bell 1987, Moskowitz 1983). Whatever the causes, no scientific mandate currently exists for adopting any particular school-based approach to alcohol use reduction and prevention. Emerging strategies focusing on family education (e.g., DeMarsh and Kumpfer 1985), management of early antisocial behavior (e.g., Hawkins and Lishner in press), changes in school and classroom structure (Gottfredson 1987), and school drug and alcohol policies (Moskowitz 1987) have shown promise and hold out the hope of more effective responses to youth alcohol-related problems in the future. Until such time as these strategies are thoroughly researched, however, their appeal remains largely theoretical. Several strategies to control youth access to alcohol have been studied with varying results. The uniform purchase age of 21 (e.g., Fell 1988; Hingson et al. 1983; Wagenaar 1982, b, 1983) and increased taxation (Saffer and Grossman 1985; Coate and Gross- man 1985) have been shown to have an impact on the sequelae of consumption including 1 Some DWI/RWID program developers have labeled point 1 programs as "prevention programs" in order to distinguish them from point 2 programs, which have been labeled as "intervention programs."This distinction seems somewhat artificial, since both types of strategies seek to prevent the occurrence of DWIJRWID. 2 These strategies are the topic of a separate background chapter and will not be discussed further. 198 BACKGROUND PAPERS youthful crashes, although effects on consumption per se, have been difficult to docu- ment. Failure to document effects on consumption may be due to methodological difficulties in measuring such effects, or it may result from the inadequacy of a simple, direct model of the effects of youth access to alcohol on consumption and related problems. The effect of numbers of alcohol outlets per capita on consumption has also received `some scrutiny, although no studies have focused directly on youth. The results of these studies have been mixed, with one study demonstrating lower consumption in States with fewer outlets (Ornstein and Hannens 1985) and hvo studies failing to find such effects (Hoadley et al. 1984, Schweitzer et al. 1983). A fourth study revealed a correlation between numbers of outlets and alcohol-related problems including felony and mis- demeanor DWI arrests in 213 California cities (Watts and Rabow 1983). This study did not include direct measures of consumption. I Student assistance programs to intervene with alcohol- and drug-using youth (e.g., Chambers and Morehouse 1983; Morehouse 1982) have been studied from a process perspective, but rigorous assessments of student drinking outcomes are not available. Other popular approaches (e.g., prevention "clubs," alcohol-free recreation, concerned parent groups) have received minimal research attention. Point 2 Strategies Of the available strategies aimed at point 2 (separating drinking from driving), perhaps the most extensively discussed is deterrence. Unfortunately for the current discussion, existing research does not generally address specific effects on youth. Ross (1984,1985) and Moskowitz (1989) provided extensive reviews of various deterrence- based strategies, including increased penalties, per se laws, enforcement crackdowns, and administrative license revocation. In general, these reviews showed that enforcement crackdowns, especially when accompanied by extensive media coverage, can have short-term (months to a few years) effects. On the other hand, a study of increased enforcement in France that focused specifically on drivers under 25 (Jayet 1986) failed to find a deterrent effect. Recently, concern over the risk of crashes associated with even very low BACs in youth has motivated some States to adopt a lower legal BAC limit for youth than for adults. In most of these States, license revocation is either an automatic or discretionary penalty for violations. Drummond et al. studied a zero BAC limit for first-year drivers in Australia. Preliminary data disclosed that this law reduced nighttime, weekend driving- a peak time for youth crash involvement (e.g., Farrow 1987; Robertson 1981). However, actual crash data concerning the Australian zero BAC law were not available at the time the research report was prepared. Hingson et al. (1986, in press) studied the effects of a 0.02 BAC limit and admini- strative license revocation for 1 year on youth in Maine. Initial results (Hingson et al. 1986) revealed that self-reported DWI and self-reported nonfatal crash involvement among drivers 19 and under declined significantly when compared to Massachusetts teens and Maine adults. Declines were most dramatic for teens who were aware of the law. In addition, actual injury and fatal crashes among Maine teens increased at a much lower rate than for drivers 20 years old and over. Followup results (Hingson et al. in press) have generally mirrored the 1986 findings, although differences between Maine and Massachusetts teens have declined to a nonsignificant level. This lack of difference appears due to a "catching-up" on the part of Massachusetts teens, perhaps owing to the high level of antidrug and anti-DWI activity in that State. Hingson et al. also noted that enforcement of the 0.02 BAC law has become sporadic, and police appear to arrest juveniles with less regularity than adult offenders. Several States have experimented with license revocation as a sanction against youth YOUTH AND OTHER SPECIAL POPULATIONS 199 possession of alcohol and other drugs. In recent testimony before the National Commis- sion Against Drunk Driving (NHTSA, October 198'7), Judge C. Foley of Milwaukee, Wisconsin, credited such a law with significant reductions in youth DWI between 1982 and 1986. However, the existence of a zero BAC law in Wisconsin, increased public awareness of the youth DWI problem, and the lack of comparison data render inter- pretation of these reductions difficult. At this time, the effects of license revocation as a sanction for youthful alcohol and other drug possession are unproven. A recently popular strategy for separating drinking from driving is to issue youth restricted licenses that limit the hours during which they may operate a vehicle. Impetus for such a strategy derives from the previously cited observation that youth DWI as well as youth fatal crashes are most likely in the evenings, especially weekend evenings. As reported by Williams (1987), at least 18 States have some sort of curfew restrictions. Williams cited a study of restrictions in four States by Preusser et al. (1984) that reported dramatic reductions in crashes during the restricted hours. He also cited additional data from New York, Louisiana, and Maryland that supported the efficacy of restricted driving hours for youth. Despite one study of the Maryland law that did not find effects on crash rates (M&night et al. 1983), Williams concluded that curfew restrictions can substantially reduce youth crash involvement. Other licensing approaches to reducing youth crashes include making drivers' licenses more difficult to obtain and presenting the license in juvenile court to both the youths and their parents. Preliminary data from California (Hagge and Marsh 1986) suggested positive results from making licenses more difficult to obtain, although, as noted by Williams (1987), the California program had so many facets that it was impossible to determine which elements contributed to the positive results and which did not. Separation of youth drinking from youth driving has also been attempted through educational strategies. There is little evidence that such programs reduce crash rates (Williams 1987; Moskowitz 1989). One well-conducted Canadian evaluation of a drink- ing/driving education program (Albert and Simpson 1985) demonstrated decreased intentions to DWI, but these decreases were realized at the cost of an increase in reported drinking frequency. Some critics of drivers' education (Robertson 1980; Robertson and Zador 1978) have suggested that such programs may actually increase crash rates by increasing the liccnsure of 16- and 17-year-olds. However, as discussed by Moskowitz (1989), these claims are based on short-term results and may not justify possible long-term negative effects of discontinuing drivers' education. Cognizant of the general failure of drinking/driving education programs, the National Highway Traffic Safety Administration (NHTSA) sponsored the development of a Peer Intervention Program (McPherson et al. 1983) aimed at enabling and motivating youngsters to intervene in the drinking and driving behavior of their peers. The program provided 8 hours of role-playing as well as 1 hour of alcohol and traffic safety information. A true experiment with random assignment compared the Peer Intervention Program to a traditional drinking and driving education program (McKnight and McPherson 1986). Students in the Peer Intervention Program reported statistically significant gains in intervention behavior at followup intervals of 1 to 4 months. The actual magnitude of these effects appeared small, although the description of the behavioral measure provided by M&night and McPherson is too sketchy to determine the meaning of the differences reported. Students Against Driving Drunk (SADD) (Anastas 1983) represents an attempt to change school and community norms with regard to youth DWVRWID. Klitzner et al. (1987) conducted an evaluation of SADD in two cities in the Western United States. This quasi-experimental study failed to find effects of SADD on any drinking or drinking/driving variables. However, weak program implementation in the SADD BACKGROUND PAPERS schools, high subject attrition from the research study, and other design confounds limit the strength of these conclusions. \ Alternative transportation (e.g., SafeRides, designated driver) as a means of separat- ing drinking from driving has not been well evaluated (Klitzner et al. 1988). Klitzner et al. (1987) provided preliminary data on parent/student contracting. These data showed that signing contracts increases the likelihood that youth will call parents for a ride. However, no differences in DWI or RWID as a result of signing the contracts were observed. This somewhat puzzling result suggests that although signers are calling home, safer transportation does not result. Critics of alternative transportation strategies have objected to these approaches on the grounds that they implicitly sanction youth drinking. Klitzner et al. (1987) failed to find evidence that signing parent/student contracts had effects on youth drinking or related problems. On the other hand, Klitzner et al. (1988) found that heavier driers also reported using more transportation alternatives. The meaning'of this latter result is unclear. It may, indeed, confirm the fears of critics of alternative transportation strategies, or it may simply reflect the fact that heavier drinkers have more reasons for using and opportunities to use transportation alternatives. Multicomponent Strategies One common indictment of many attempts to prevent alcohol- and drug-related problems among youth is too narrow a programmatic focus (Klitzner 1987; Klitzner and Bell 1987; Goodstadt 1986, Huba et al. 1980). That is, communities have tended to focus on one kind of response (e.g., a school curriculum, a SADD club, a police crackdown) to the exclusion of other types of responses. Recently, some communities have attempted to overcome the narrowness and frag- mentation of past responses to youth DWI by instituting communitywide, systemic responses that attempt to institute a coordinated and comprehensive package of mutual- ly reinforcing countermeasures. Thus, a community might institute a strong anti-alcohol use school policy, work to restrict alcohol sales to minors through increasing alcohol beverage control enforcement, rigorously enforce DWI laws, institute roadblocks, ag- gressively prosecute and heavily sanction youthful DWI offenders, and develop com- munity resources for the treatment of addicted teens. Ten communities that are attempting to implement communitywide responses are described by Pacific Institute (in press). The communitywide model has considerable theoretical and conceptual appeal, and many of the strategies communities appear to be using have been shown to be effective in their own right (e.g., increased enforcement, reductions in alcohol availability to youth). To date, however, rigorous evaluations of multicomponent, communitywide anti-DWI programs have been extremely limited. Perhaps the most relevant research is the Lackland Air Force Base Experiment (Barmark and Payne l%l), which effectively reduced DWI among airmen through a variety of normative, informational, and enforcement strategies. However, strategies shown effective in the highly insular and controlled environment of a m&try hstdation will not necessarily be effective in the less well-controlled environments of most American communities. The communitywide model has shown promise in other health areas, notably the reduction of risk factors associated with cardiovascular disease (Farquhar et al. 1977; Puska et al. 1985). However, the effectiveness and feasibility of systemwide responses to the youth DWI problem awaits further research. YOUTH AND OTHER SPECIAL POPULATIONS 201 Summary Current research into youth drinking and driving countermeasures suggests that effective strategies are available for reducing youth access to alcohol and for separating youth drinking from youth driving. However, despite the continuing popularity of strategies focused on developing or changing individual knowledge, attitudes, and skills, aimost all the effective countermeasures reviewed in this chapter have been regulatory or legislative in nature. Recent evidence indicates that the most effective countermeasures will likely be those that focus on minimum purchase age, alcohoi pricing, limiting alcohol outlets, lower legal BACs for youth, curfew restrictions on youth driving and other licensing restrictions, and enforcement. This is not to imply that the search for effective programs focused on individual youth should be discontinued. However, many individually focused prevention approaches popular in the 1970s and 1980s appear to have outlived their usefulness. Issues and Recommendations Proven technologies exist for reducing the death and disability suffered by youth as a result of drinking and driving. As discussed, these include restrictions on youth access to alcohol and restrictions on youth driving. The problem is not so much one of finding effective countermeasures as it is overcoming societal inertia to implement them. Thus far, the uniform alcohol purchase age is the only proven countermeasure to be adopted nationwide. In some States, even the threatened loss of Federal highway funds did not guarantee speedy legislative action. Williams (1987) posed the question of whether society is ready to take the steps necessary to improve the current situation with regard to youth drinking and driving. He responded: "To the extent that legislative restrictions are necessary to rectify the situation, [this] question can at present be answered in the negative." A major item on the Nation's public health agenda should be to educate parents, legislators, and other concerned citizens about the regulatory measures that can be taken to realize additional meaningful reductions in youth DWI. Of course, regulatory responses will only be effective to the extent that they are enforced (Ross 1984). In general, the quality of enforcement of novel DWI laws decreases over time, an effect observed in Maine's experience with 0.02 legislation enacted in 1982 (Hingson et al. in press). Public support must be developed for the vigorous and continued enforcement of new laws as well as for their enactment. It is also clear that regulation alone will never be a complete answer to the youth DWI/RWID problem. Youth wiIl always have access to alcohol and cars. Indeed, licensing restrictions will never affect all teenagers, since a significant minority of teen drivers are unlicensed (Klitzner et al. 1988; Williams et al. 1985). Moreover, the high crash rates of teens continues into the early twenties-an age group to whom purchase age restrictions do not apply. Thus, in addition to regulatory responses, continued efforts should be made to develop prevention programs that affect the drinking and drinking/ driving choices of individual young people. Prevention program development and research need to break away from the unsuc- cessful models of the past. New approaches are needed that are firmly grounded in an understanding of the factors that predispose, reinforce, and enable youth alcohol use and DWURWID. Given the current state of knowledge, such an understanding will require a program of additional research into the etiology of youth drinking and DWURWID. This is not to imply that the testing of new program models should be delayed until a comprehensive and widely accepted set of etiological models is available. 202 BACKGROUND PAPERS Rather, program research and etiologic research should be seen as complementary endeavors, with data from one area of inquiry informing theory development and research activities in the other. Significant programmatic attention must be paid to youth alcohol consumption per se. It is possible, in theory, to separate youth drinking from youth driving. However, these behaviors are currently so inextricably intertwined that successful DWI/RWID preven- tion programs may ultimately be those with a heavy emphasis on reducing alcohol consumption. The youth DWURWID problem is not limited to impairment due to alcohol. Data from a 1983 survey of 18- to 24-year-olds (Elliot 1987) revealed that one in five respon- dents had driven while high on marijuana, and nearly one in ten had driven while high on other drugs. Moreover, the prevalence rate for DWI was twice as high for multiple drug users as it was for youth who only used alcohol. More research is needed on the contribution of marijuana and other drugs to crash-related mortality and morbidity, and future discussions of the youth DWI/RWID problem should specifically address these substances. An effort should be made to involve physicians in the national effort to combat youth drinking/driving. Questions about alcohol use patterns, driving, and seatbelt use should be part of all doctor visits for teens and should be specifically explored when a presenting complaint suggests aIcoho1 involvement (AAP 1987; Klitzner and Schonberg 1988). Moreover, parents of preteens should be counseled concerning the effects of their own drinking and drinking/driving attitudes and behavior on the behavior of their children. Parents should also be encouraged by physicians to disallow alcohol consumption by their adolescent children, including restrictions on attending parties where alcohol is served (AAP 1987). Finally, continued efforts should be made to change social norms regarding youth alcohol use and DWURWID. It has been argued that changes in social norms and values as a result of two decades of antismoking activities and programs have contributed significantly to the efficacy of smoking cessation and prevention programs (Polich et al. 1984, Moskowitz 1983; Leventhal and Cleary 1980). Similar changes in drinking and DWURWID norms toward greater intolerance can facilitate the adoption of effective regulatory measures (Moskowitz 1989) and can also have a direct impact on youth behavior (Klitzner et al. 1988). The communitywide approach discussed earlier is one appealing strategy for effecting normative change because it attempts to involve all segments of the community in combating the youth drinking/driving problem. REFERENCES Albert, W., and Simpson, R. Evaluating an educational program for the prevention of impaired driving among grade 11 students. Journal of Drug Education 15(1):57-71,198s. American Academy of Pediatrics Committee on Adolescence. Alcohol use and abuse: A pediatric concern. Pediatrics 79(3):450-453, 1987. Anastas, R. SADD Chapter Handbook and CumUum Guide. Marlboro, MA: Students Against Driving Drunk, 1983. Atkin, C.; Neuendorf, K, and McDermott, S. The role of alcohol advertising in excessive and hazardous drinking. Joumal of DrugEducation 13:313-326,1983. Atkin, C.; Hocking, J.; and Block, M. Teenage drinking: Does advertising make a difference? Journal of Communications 34: 157-167, 1984. Bachman, J.; Johnston, L.; and O'Malley, P. Monitoringthe Future: Questionnaire Responvesfiom the Nation's High School Seniors. Ann Arbor, MI: Institute for Social Research, 1987. Barmark, J., and Payne, D. The Lackland accident countermeasures experiment. In: Proceedings of the 40th Annual Meeting of the Highway Research Board, Volume 45, Washington, DC, 1961. pp. 221-232. Bergeron, J., and Joly, P. Young drivers' attitudes and drunk driving habits as a function of their YOUTH AND OTHER SPECIAL POPUIATIONS 203 accident involvement and violation record. In: Benjamin, T., ed. Young tivers Impaired by Alcohol and Other Drugs. London: Royal Society of Medicine Services, 1986. Biddle, B.; Biddle, B.; Bank, J.; and Martin, M. Social determinants of adolescent drinking. Journal of Studies on Alcohol 41(3):215-241,198O. Blanc, H. Problem drinking in delinquent and nondelinquent adolescent males. llte American Journal of Lhqs and Alcohol 9(2):221-232,1983. Boyd, N.R., and Huffian, W. The relationship between adult maturity and drinking-driving involvement among young adults. JournaZ of Safety Resemch 15( 1): 16, 1984. Cameron, T. Drinking and driving among American youth: Beliefs and behaviors. Drug and Alcohol Dependence 10: 1 l-33,1982. Chambers, J., and Morehouse, E. A cooperative model for preventing drug and alcohol abuse. National Association of Secondary School Principah'BulZetin January:8187,1983. Coate, D., and Grossman, M. "Effects of Alcohol Beverage Prices and Legal Drinking Ages on Youth Alcohol Use: Results from the Second National Health and Nutrition Examination Survey." Unpublished manuscript. New York: National Bureau of Economic Research, 1985. DeMarsh, J., and Kumpfer, K Family oriented interventions for the prevention of chemical dependency in children and adolescents. Journal of Children in Contemporasy Society 18(1 and 2): 117-182, 1985. Department of Transportation. Communications Strategies on Alcohol and Highway Safety, Volume ZZ, by Grey Advertising. Report No. DOT HS-803-714. Washington, DC: the Department, 1975. Department of Transportation. Fatal Accident Reporting System 1986 Summary Report. Washington, DC: The Department, 1987. Department of Transportation. Norms and Attitudes Related to AlcohoI Usage and Driving: A Review of the Rekvant Literature, Volume III and ZV, by Pawlowski, W. DTNH 22-81- 07385. Washington, DC: the Department, 1982. DiBlasio, F. Predriving riders and drinking drivers. Journal of Studies on Alcohol 49(1):11-15, 1988. Douglass, R. Youth, alcohol, and traffic accidents: Current status. Recent Developments in AlcohoZisnq Vohune I. New York: Plenum Press, 1983. Drummond, AE., Cave, T.C.; and Healy, DJ. The risk of accident involvement by time of week-an assessment of the effects of zero BAC legislation and the potential of driving curfews. In: Benjamin, T., ed. Young Drivers Impaired by Alcohol and Other Drugs. London: Royal Society of Medicine Services, 1987. Elliot, D. Self-reported driving while under the influence of alcohol/drugs and the risk of al- cohol/drug-related accidents. Alcohol, Drugs, and Driving 3(3-4):314%, 1987. Farquhar, J.; Maccoby, N.; Wood, P.; Alexander, J.; Breitrose, H.; Brown, P.; Haskell, W.; McAlister, A; Meyer, A; Nash, J.; and Stern, M. Community education for cardiovascular health. Lancet June:1192-1195, 1977. Farris, R.; Malone, T.; and Lilliefors, H. A Comparison of Alcohol-Znvolvement in Exposed and Injured drivers: Phases Z a& II. (Report No. DOT-HS-4-06954). Washington, DC: National Highway Traffic Safety Administration, 1976. Farrow, J. Young driver risk taking: A description of dangerous driving situations among 16- to 19-year-old drivers. International Journal of the Addictions 22(12): 1255-1267, 1987. Fell, J. "Effectivenessof Raising the Drinking Age to21." Paper presented to the 14th International Forum on Traffic Records Systems, San Diego, CA, July, 1988. Finley, B. Drinking and driving behaviors in grades l-10 in one rural high school. Current ZsslCRF in Alcohol and Drug Nursing:\ Research, Education, and Clinical Practice. Washington, DC, May, 1983. Forney, M.; Forney, P.; Davis, H.; Van Hoose, J.; Gafferty, T.; and Allen, H. "Factors Affecting Knowledge, Attitude, and Behavior Toward the Use of Alcohol Among Middle School Students." Paper presented to the American Educational Research Association Annual Conference, April, 1984. Frankena, M.; Cohen, M.; Daniel, T.; Ehrlich, L.; Greenspun, N.; and Kelman, D. Alcohol advertising, consumption and abuse. In: Federal Trade Commission. Recommendations of the Stafl of the Federal Trade Commission: Omnibus Petition for Regulation of Unfair and Deceptive Alcohol Beverage Marketing Practices (Docket Number 209-46). Washington, DC: Federal Trade Commission, 1985. 204 BACKGROUND PAPERS Goodstadt, M. Shaping drinking practices through education. In: von Wartburg, J.; Magnenat, P.; Muller, R.; and Wyss, S., eds. C'nts in Alcohol Research and the Revenfion ofArcoho1 hblems. Beme, Switzerland: Hans Huber Publishers, 1985. Goodstadt, M. School-based drug education in North America: What iswrong? What can be done? Journal of School Health 56(7):278-281,1986. Gottfredson, D. "Changing School Structures to Benefit High-Risk Youths" Paper presented at the National Center on Effective Secondary School's Symposium on Structural Change in Secondary Education, Madison, WI, May, 1987. Hagge, R., and Marsh, W. The Traffic Safev Impact of Proti&ui kenring. Sacramento, CA: California Department of Motor Vehicles, 1986. Hawkins, J.D., and Lishner, D.M. Schooling and delinquency. In: Johnson, E.H., ed. Handbook on Crime ami Winquency Prevention. Westport, CT.: Greenwood Press, in press. Hetherington, R.; Dickinson, J.; Cipywnyk, K; and Hay, D. Attitudesand knowledge about alcohol among Saskatchewan adolescents. Canadian Journal of Public Heahh 70(4):247-259, 1979. Hingson, R.; Heeren, T.; and Morelock, S. Preliminary effects of Maine's 1982 0.02 law to reduce teenage driving after drinking. In: Benjamin, T., ed. Young Drivers Impaired by AIcohoi and Other tigs. London: Royal Society of Medicine Services, 1986. Hingson, R.; Heeren, T.; and Morelock, S. Effects of Maine's 1983 -02 law to reduce teenage driving after drinking. Alcohol, L%ugs, and Driving, in press. Hingson, R.; Howland, J.; Morelock, S.; and Heeren, T. Legal interventions to reduce drunken driving and related fatalities among youthful drivers.AZcohol Dmgs, Md Dn'ving 4(2):87- 98,1988. Hingson, R.; Scotch, N.; Mangione, T.; Meyers, A.; Glantz, L.; Heeren, T.; Liu, N.; Muscatel, M.; and Pierce, G. Impact of legislation raising the legal drinking age in Massachusetts from 18 to 20. American Journal of Public Health 73: 163-170, 1983. Hoadley, J.; Fuchs, B.; and Holder, H.The effects of alcohol beverage restrictionson consumption: AZ-year longitudinal anatysis.Aimetican JournalofDrugandAlcohoIAbzue 10~375401, 1984. Huba, G.; Wingard, J.; and Bentler, P. Applications of a theory of drug use to prevention programs. Joumai of Drug Education lo(l):2538,198O. Insurance Institute for Highway Safety. Srnrus Report 19(3): 1-3, 1984. Jayet, M. Penal policy for preventing young drivers from drinking: Scope, limits, and principles. In: Benjamin, T.; ed. Young Drivers Impaired by Alcohol and Other Drugs. London: Royal Society of Medicine Services, 1986. Jessor, R. Risky driving and adolescent problem behavior: An extension of problem-behavior theory. Alcohol, Drugs, and Driving 3(3-4): l-12,1987. Jessor, R., and Jessor, S. Problem Behavior and Psychosocial Development: A Longitudinal Study of Your/t. New York: Academic Press, 1977. Klitzner, M. An assessment of the research on school-based prevention programs. In: Report to Congress on the Nature and Effectiveness of Federd State, and Local Drug Preven- tion/Education Z+ogra.mr. Washington, DC U.S. Department of Education, 1987. Klitzner, M., and Bell, C. "Youth Drug and Alcohol Abuse Prevention: Why We Can't Answer the Question, `What Works?"' Paper presented to the 115th Annual Meeting of the American Public Health Association, New Orleans, LA, October, 1987. Klitzner, M., and Schonberg, K Concerns regarding the indirect assessment for drug abuse among adolescents. Journal of DeverOpmetttal and Behavioral Pediatrics 9( 1):30-31, 1988. Klitzner, M.; Gruenewald, P.; Rossiter, C.; Bamberger, E.; and Roth, T. Students Against Driving Drunk: A National Study -Final Report. Vienna, Vk Pacific Institute for Research and Evaluation, 1987. Klitzner, M.; Vegega, M.; and Gruenewald, P. An empirical examination of the assumptions underlying youth drinking/driving prevention programs. EvaIuarionandProgram Pkznning 11:219-235,1988. Koningsberg, D.; Weinhouse, B.; and Wechsler, J. Teenagers and alcohol: Holiday hazard, year-round tragedy. L.&es Home Journal lOO(12): 1983. Kraus, A; Steele, R.; Ghent, W.; and Thompson, M. Predriving identification of young drivers with high risk of accidents. Journal of Safety Research 255.56, 1970. Krohn, M.; Akers, R.; Radosevich, M.; and Lanze-Kaduce, L. Norm qualities and adolescent driving and drug behavior. Journal of Drug Issues Fal1:349-359,1982 YOUTH AND OTHER SPECIAL POPULATIONS 205 Leventhal, H., and Cleary, P. The smoking problem: A review of the research and theory in behavioral risk modification. Psychological Bulletin 88:370-405,1980. Lewis, C. "Preventing Traffic Casualties Among Youth: What Is Our Knowledge Base?" Paper presented to the International Symposium on Youth at Risk for Traffic Accidents, Santa Monica, CA, May 1987. Lightsey, M., and Sweeney, M. Life problems experienced from drinking: Factors associated with level of problem drinking among youthful DWI offenders. Journal of Alcohol and Drug Educarion 30(3):6582,1985. Lowman, C. Facts for planning, no. 3: U.S. teenage alcohol use in unsupetvised social settings. Alcohol Heahh and Resemch World 6(2): 1981. McKnight, A; Hyle, P.; and Aibrecht, L Youth License Conbol Hemonstralion Project. NTIS No. DOT-H!+806 616. Washington, DC: National Highway Traffic Safety Administration, 1983. M&night, J., and McPherson, K Evaluation of peer intervention training for high school alcohol safety education.AccidentAn and Prevention 18(4):339-347,1986. McPherson, K.; M&night, A; and Weidman, J. SupplementalDriver Safety Program Development: Final Report, Volume I, Developmental Research and Evaluation. Washington, DC: National Highway Transportation Safety Administration, 1983. Milgram, G. Societal attitudes toward youthful drinking. Journal of Drug E&cation 12(4), 1982. Morehouse, E. The student assistance program: An alcohol and drug abuse prevention model. In: Amowitz, E., ed. fievenrion Strategies in Memal Heahh. New York: PRODIST, 1982 Moskcnvitz, J. Preventing adolescent substance abuse through drug education. In: Glynn, T.; Leukefeld, C.; and Ludford, J., eds. Preventing Adolescent Drug Abuse: Intervention S@&e@es. National Institute on Drug Abuse Research Monograph No. 47, DHHS Pub. No. (ADM)83-1280. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1983. Moskowitz, J. The primary prevention of alcohol problems: A critical review of the research literature. Journal of Studies on Alcohol 50( 1):54-88, 1989. Moskowitz, J. School Dnrg and Alcohol Policy: A Prehminary Model Relating Policy Implementa- rion ro School Problem. Berkeley, CA Pacific Institute for Research and Evaiua- tion/`Prevention Research Center, 1987. Nataanen, R., and Summala, H. Road User Behavior and Traffic Accidents. Amsterdam and New York: North Holland/American Eisevier, 1976. Nusbaumer, M., and Zusman, M. Autos, alcohol, and adolescents: Forgotten concerns and overlooked linkages. Journal of Drug Education 1 l(2): 167-178,198l. O'Day, J. Drinking involvement and age ofyoung drivers in fatal accidents. Hit Laboratory Reports October:13-14, 1970. Omstein, S., and Hannens, D. Alcohol control laws and the consumption of distilled spirits and beer. Journal of Consumer Research 12:200-213,198s. Pacific Institute for Research and Evaluation. The Introduction of Effective Systemwide Strategies to Combat Youth Drug and Alcohol Abuse: Assessment Report. Vienna, VA Pacific Institute for Research and Evaluation, in press. Paulson, J. Accidental injuries. In: Behrman, R., and Vaughan, V., eds. Nelson Handbook of Pediatrics. Philadelphia, PA Saunders Company, 1983. Peiz, D., and Schuman, S. Are young drivers really more dangerous after controlling for exposure and experience? Journal of Safety Research 368-79, 1971. Perrine, M.; Waller, J.; and Harris, LAkoholandHighway Safety: Behavioraland MedicaiAspecls. Burlington, VT: University of Vermont, 1971. Polich. J.; Ellickson, P.; Reuter, P.; and Kahan, J. Strategies for ControllingAdolescent Drug Use. Santa Monica, CA Rand Corporation, 1984. Preusser, D.; Williams, A; and Zador, P. The effects of curfew laws on motor vehicle crashes. Law end Policy 6115128.1984. Puska, P.; Nissinen, A.; Tuomilehto, J.; Salonen, J.; Koskela, K; McAiister, A; Kottke, T.; Maccoby, N.; and Farquhar, J. The community-based strategy to prevent coronary heart disease: Conclusions from the ten years of the North Karelia Project. Annual Review of Public Health 6: 147-193,1985. Robertson, L. Patterns of teenage driver involvement in fatal motor vehicle crashes: Implications for policy choice. Journal of Heahh Politics, Policy und Law 6(2):303-314, 1981. Robertson, L, and Zador, P. Driver education and fatal crash involvement of teenage drivers. American Journal of Public Health 73:959-%5,1978. 206 BACKGROUND PAPERS Robertson, L.S. Crash involvement of teenaged driverswhen driver education is eliminated from high school. American Journal of Public Health 70:599603,1980. Ross, H. Deterring drunken driving: An analysis of current efforts. Journal of Studies on Alcohol Supplement 10:122-128, 1985. Ross, H. Deleting the Drinking Dkfm: Legal Policy and Social Conb-ol. Lexington, MA: Lexington Books, 1984. Saffer, H., and Grossman, M. Effecfi of Beer Prices.and Legal DrinkingAges on Youth Motor Vehicle Fa&itzks. Unpublished manuscript. New York: National Bureau of Economic Research, 1985. Schweitzer, S.; Intriligator, M.; and Salehi, H. Alcoholism, an econometric model of its causes, its effects, and its control. In: Grant, M.; Plant, M.; and Williams, A., eds. Economics and Alcohol Consumption and Control. London: Croom Helm, 1983. Stoles, P., and Fine, E. Substance abuse patterns among youthful drinking drivers. Alcoholism Clinical and Eqxrimental Research 5(l), 1981. Simpson, H.; Mayhew, D.; and Warren, R. Epidemiology of road accidents involving young adults: Alcohol, drugs, and other factors. Drug and Alcohol Lkpendence 10:3563, 1982. Strickland, D. Advertising exposure, alcohol consumption, and misuse of alcohol. In: Grant, M.; Plant, M.; and Williams, k, eds. Economics and Alcohol Consumption and Controls. London: Croom Helm, 1983. Summala, H. Young driver accidents: Risk taking or failure of skills. Alcohol Drugs, and Driving 3(3-4):79-91, 1987. URSA/Pacificon. Campaign Strategy Repoti for the IVUAA I%blic information Campaign. Rock- ville, MD: National Institute on Alcohol Abuse and Alcoholism, 1980. Vegega, M, "Deterring Drinking-Driving Among Youth: Some Research Needs." Paper presented at the annual convention of the American Psychological Association, Toronto, Canada, 1984. Vegega, M., and Klitzner, M. Drinking and driving among youth: Astudy of situational risk factors. Health Education Quarterly, in press. Vegega, M., and IUitzner, M. What have we learned about youth anti-drinking-driving programs? Evaluation and Program Planning 11:203-217,1988. Vejnoska, I. Putting the brakes on teenage drunk driving. Police Chief 49(12), 1982 Wagenaar, k Alcohol, Young Lkivers, and Traffic Accidents: Effects of Minimum Age Laws. Lexington,, MA Lexington Books, 198%. Wagenaar, A. "Youth Drinking and Driving: One Effective Countermeasure." Paper presented at the Drinking and Driving Symposium, American Association for Advancement of Science Meetings, Detroit, MI, May 31, 19836. Wagenaar, A Preventing highway crashes by raising the legal minimum age for drinking: An empirical confirmation. Journal of Safety Research 13:57-71.19826. Wagenaar, A. Aggregate beer and wine consumption: Effects of changes in the minimum legal drinking age and a mandatory beverage container deposit law in Michigan. Journal of Alcohol Studies 43:469-4%,1982a. Watts, R., and Rabow, J. Alcohol availability and alcohol-related problems in 213 California cities. Alcoholism: CLinical and Experimental Research 7( 1):47-58,1983. Williams, A; Lund, A; and Preusser, D. Teenage driver licensing in relation to state laws.Accident Analysis and fievention 17(2):135-145,1985. Williams, A.; Lund, A.; and Preusser, D. Drinking and driving among high school students. International Journal of the Addictions 21(6):643-655,1986. Williams, A. Effective and ineffective policies for reducing injuries associated with youthful drivers. Alcohol Drugs, and Driving 3(3-4):109-l 17,1987. W&man, F. Current status of research demonstration programs in the primary prevention of alcohol problems. In: Prevention, Intervention, and Treatment: Concerns and Models. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1982. Zylman, R. Youth, alcohol, and collision involvement. Jozunal ofsafety Research 5(2), 1973. 207 Youth and Other Special Populations Motor Vehicle Crashes and Alcohol Among American Indians and Alaska Natives Philip A. May, Ph.D. University of New Mexico The American Indian population, made up of well over 300 tribes, now numbers more than 1.5 million or over 0.6 percent of the U.S. population. Throughout the recent past, it has been a rapidly growing population that has consistently had a birth rate twice that of the rest of the U.S. population (IHS 1988; May 1988; Broudy and May 1983). Because of this high fertility, the median age of the Indian population was 22.9 in 1980, compared to 30.3 for the overaIl U.S. population. The median age of the population living on reservation was 19 years in 1980 (U.S. Congress 1986). The young age composition has definite implications for the study of both motor vehicle injuries and alcohol. Some other general considerations of the U.S. Indian population are also vital here. While Indians and Alaska Natives live in every State of the Union, most live in 32 "reservation States" (see exhibit l), the vast majority (over 1 million) in Western States (IHS 1988). Many Indians live in rural areas (46 percent) or other urbanized settings such as small towns (32 percent), while 22percent live in central cities. In 1980,37 percent lived on reservations or in Indian communities. Many who live in cities or off reservations are characterized as highly mobile (U.S: Congress 1986). Average economic indicators for Indians as a group remain poor. The 1980 Census indicated unemployment rates twice the national average, and the median family income for Indians was about half the average for the entire U.S. population (U.S. Congress 1986; U.S. Bureau of Census 1984q b). Finally, educational attainment of many Indians remains lower than the national averages (Brod and McQuiston 1983). Overview of Accident Statistics When reviewing accident data for Indians, one must first consider accidents as they affect the overall Indian population and then move to specific considerations of gender, age, and local and regional data, Only recently have accidents fallen to second place among causes of mortality. Accidents were the leading cause of death throughout the 195Os, 196Os, and 1970s for Indians and Alaska Natives in the 32 reservation States (IHS 1988). In 1983-85? accidents caused 3,218 deaths or 16 percent of all mortality (see table 1). Of these deaths, 1,753 or 54.5 percent were caused by motor vehicles. When measured by crude rates, accidents are the fourth leading cause of death among other U.S. residents. Motor vehicle accidents take a higher toll among Indian males than females, and this is even more true of other accidents. Table 2 presents age-adjusted death rates that allow comparisons between Indians and other groups in the United States. When age-adjusted rates are considered, accident 208 BACKGROUND PAPERS Exhibit 1 Indian Health Service (IHS) reservation States numbered L?4 from 1969 to 1974. Prior to that, IHS calculated rates on SO-percent samples of U.S. records. Since 1974, eight more States have heen added. The current 32 stateslisted by administrative area and the year added for the most recent additions are: Nashville Area: Alabama (added in 1984),.Connecticut (added in 1983), Florida, Louisiana, Maine (added in 1979), Mississippi, New York (added in 1979), North Carolina, Pennsylvania (added in 1979), Rhode Island (added in 1983). Bemidji Area: Michigan, Minnesota, Wisconsin Oklahoma Area: Oklahoma, Texas (added in 1983), Kansas Aberdeen Area: Iowa, Nebraska, North Dakota, South Dakota. Billings Area: Montana, Wyoming Albuquerque Area: Colorado, New Mexico Navajo Area: Main Navajo reservation (parts of Arizona,NewMexico and Utah) Phoenix Area: Arizona (most), Nevada, Utah Tucson Area: Southern Ariina (Papago reservation) Portland Area: Idaho, Oregon, Washington Alaska Area: Alaska mortality becomes the third leading cause of death for aU Indians and Alaska Natives. The rate of 42.2 per 100,000 for motor vehicle accidents is 2.3 times the rate for alI races combined. Other accidents, which in the distant and relatively recent past have played a major role in mortality among Indians (Kunitz 1983) and particularly Alaska Natives (Boyd et al. 1968), are 2.2 times higher than U.S. averages. The trend in accident mortality rates for Indians and Alaska Natives has been down over the years that accurate, Indian-specific rates have been kept, but the exact mag- nitude of this drop is difficult to determine . Many procedural changes have affected the calculation of Indian mortality rates, confusing the interpretation of trends. Some of these changes include: alterations in census enumeration methodologies for American Indians and Alaska Natives; changes in methodologies for estimating the Indian popula- tion (which was necessary to correct suspected undercounts;); debate over whether tribal enrollment or resident population figures were the most appropriate for use as denom- inators; and, most importantly, the addition of Indians from States such as Pennsylvania, Rhode Island, and Connecticut to Indian Health Service "reservation States" (see U.S. Congress 1986, pp. 59-82; IHS 1988; Passe11 1976; PasseII and Berman 1986, May 1988). Taken at face value the Indian Health Service (IHS) rates indicate that accident mortality has fallen dramatically. The age-adjusted rate of motor vehicle accident mortality declined from a high of 106.2 in 1956 to 42.6 in 1985 (IHS 19@3), a drop of 59.8 percent. However, a more realistic trend figure might be obtained by using m-year figures from 1956 through 1976, when the major reshaping of the II-IS denominator began. From 1956 to 1976, the Indian rate of motor vehicle accident death dropped from 106.2 to 74.5 or 29.8 percent. This is quite comparable to the overall U.S. drop of 14.7 percent from 1956 to 1976 or 34.0 percent from 1956 to 1985. Therefore, it is safe to conclude that the American Indian rates have dropped, probably in a magnitude similar to or slightly greater than the drop for the overall U.S. population. 1 One positive reason for this drop could certainly be the upgrading of emergency medical services begun in the middle 1970s on many reservations. 1 Indian vital statistics in many areas tend to parallel U.S. statistics, although some rates usually remain higher. For example, most tribes experienced a baby boom similar to the total U.S. population, only the birth rates were substantially higher and the peak was several years later (see Broudyand May 1983). Also, motor vehicle accident rates were reduced in 1974 on the Navajo reservation when the oil embargo forced the lowering of the national speed limit (Katz and May 1979). YOUTH AND OTHER SPECIAL POPULATIONS 209 Table I. Leading causes of death for American Indian and Alaska Natives by frequency, gender, and rates per 100,000, 1983-85 Males Females Mortality Mortality Causes of death Number rate Causes of death Number rate All causes 11,894 Diseases of the heart 2,727 Accidents 2,385 Motor vehicle 1,263 Other accidents 1,122 Malignant neoplasms 1,424 Chronic liver disease and cirrhosis 520 Suicide 485 Cerebrovascular diseases 456 Homicide 453 Pneumonia and influenza 387 Diabetes mellitus 284 Chronic obstructive pulmonary diseases and allied conditions 254 Certain conditions originating in the perinatal period 216 All other causes 2,303 568.4 130.3 114.0 60.4 53.6 68.0 24.8 23.2 21.8 21.6 18.5 13.6 12.1 10.3 All causes Diseases of the heart Malignant neoplasms Accidents Motor vehicle Other accidents Cerebrovascular diseases Chronic liver disease and cirrhosis Diabetes mellitus Pneumonia and influenza Certain conditions originating in the perinatal period Nephritis, nephrotic, and nephrosis Homicide All other causes 8,216 382.2 1,958 91.1 1,301 60.5 833 38.7 490 22.8 343 16.0 513 23.9 404 18.8 392 18.2 295 33.7 171 8.0 151 144 2,054 7.0 6.7 Source: Indian Health Service, Chart Book Series, 1988. The greater magnitude of the rate of inotor vehicle accident death warrants further discussion here. The overall aggregated data for 1983-85 indicate that Indians and Alaska Natives die twice (2.3 times) as frequently from motor vehicle crashes. But since these are overall figures that include both males and females, eastern and western Indians, Alaska Natives and Indians, rural and urban Indians, and a great diversity of tribes with differing social and cultural traditions, it is imperative to examine data that are more specific to local, tribal, and gender variations. Table 3 gives age- and gender-specific data for U.S. Indians and Alaska Natives. Indian males' rates of motor vehicle accident mortality are substantially higher than Indian females, in each age category. Further, Indian males age 15 to 85 years and older die three times more frequently than Indian females in motor vehicle crashes. Both Indian males and Indian females have higher rates of death at virtually every age category than all races combined, the exception being Indian women 75 and over. This higher rate of accidental death for Indian males, when expressed in years of potential life lost or life expectancy, is dramatic. Among the Navajo in 1972-74, motor vehicle deaths reduced the male life expectancy at birth by5.2 years, and other accidents reduced male longevity by another 3.1 years (Carr and Lee 1978). For Navajo females, the respective figures were 2.7 years and less than 1 year (Carr and Lee 1978). The comparable figures for the U.S. population at this time were reductions of less than 1 year for motor vehicle and other accidents for both males and females (Tsai et al. 1978). An examination of geographical and cultural variations in unintentional injury death rates (table 4) reveals three items that need to be emphasized. First, the highest rates of 210 BACKGROUND PAPERS Table 2. Age-adjusted mortality rates (per 100,000 population) for American Indians and Alaska Natives in reservation States, 1983-1985, and U.S. populations, 1985 Indians United States Ratio of and Alaska AlI Indians to Natives All races White other all US races All causes ~ 542.7 Major cardiovascular disease 174.6 Diseases of heart 141.6 Cerebrovascular diseases 26.8 Atherosclerosis 2.5 Hypertension 1.0 Malignant neoplasms 84.9 Accidents 77.7 Motor vehicle 42.6 AII other 35.1 Chronic liver disease and cirrhosis 29.2 Diabetes mellitus 22.9 Pneumonia and influenza 17.9 Homicide 14.3 Suicide 14.1 Chronic obstructive pulmonary diseases and allied conditions 11.5 Tuberculosis, all forms 1.6 546.1 523.1 697.8 1.0 224.0 216.9 273.4 0.8 180.5 176.1 210.4 0.8 32.3 30.1 49.4 0.8 4.0 4.0 3.9 0.6 1.8 1.4 4.7 0.6 133.6 130.7 155.7 0.6 34.7 34.1 39.7 2.2 18.8 19.1 17.4 2.3 16.0 15.0 22.3 2.2 9.6 8.9 14.4 3.0 9.6 8.6 17.7 2.4 13.4 12.8 16.8 1.3 8.3 5.4 24.4 1.7 11.5 12.3 6.7 1.2 18.7 19.2 13.8 0.6 0.5 0.3 1.8 3.2 Source: Indian Health S&vice. Chart Book Series, 1988. motor vehicle accident deaths occur in the West (Rocky Mountain, Plains, and Upper Midwest). Indians in the Billings, Aberdeen, Tucson, and Navajo areas have the highest rates, which are 6.7, 5.4, 4.9, and 4.6 times, respectively, that of the general U.S. population. Alaska has the lowest rate. Second, Alaska Natives, while they have lower rates of motor vehicle mortality, have much higher rates of other accident mortality such as from water transport, firearm incidents, and air transport. Third, females, particularly those living in Alaska and the West, are at increased risk for motor vehicle and other injury-related death. In summary, even though the overall age-adjusted death rates for accidents among Indians seem to have improved dramatically over the years and are now only 2.2 to 2.3 times the national averages, many tribes, cultures, and subcultures, particularly in the West, still have &maIjdy I&her accident mortality rates. The reasons for this are sociocultural (May 1982) as well as geographical and envirqnmental. Local Mortality Data From the West Isolated State data confirm, in another way, the extremely high accident mortality rates for western Indians. From 1982 through 1986 in Montana, the Indian crude rate of death from motor vehicle accidents was 79.7 per lOO,fKKl, while the non-Indian rate was YOUTH AND OTHER SPECIAL POPULATIONS 211 Table 3. Age-and gender-specific motor vehicle accident death rates (per 100,000 population) for American Indians and Alaska Natives in reservation States, 1983-85, and U.S. all races, 1984 MALE FEMALE American Indian U.S. Ratio American Indian -U.S. Ratio and all Indian to and au Indian to Alaska Native races U.S. all Alaska Native races U.S. all (1983-1985) (1984) races (1983-1985) (1984) races Under 1 year 19.5 43 4.5 12.0 4.6 2.6 1 to 4 years 23.8 8.0 3.0 18.8 5.7 3.3 5 to 14 years 9.9 8.3 1.2 7.9 4.9 1.6 15 to 24 years 87.7 54.9 1.6 31.9 18.3 1.7 25 to 34 years 99.1 37.0 2.7 30.7 10.7 2.9 / 3.5 to 44 years 78.4 25.2 3.1 29.4 9.2 3.2 45 to 54years 71.0 22.4 3.2 25.7 8.8 2.9 55 to 64 years 59.0 21.9 2.7 22.1 10.1 2.2 65 to 74 years 57.0 24.9 2.3 15.2 12.7 1.2 75 to 84 years 89.4 42.0 2.1 14.4 20.0 0.7 Over 84 years 71.9 53.3 1.3 9.5 13.5 0.7 Source: Indian Health Service, Chart Book Series, 1988. 26.5 (calculated from data in the State of Montana, 1983-88). Even though these wey not age-adjusted, one can conclude that the Indian rate was approximately 2.5 to 3 times the non-Indian rate in the same State. In New Mexico, similar trends are found. In both 1976-78 and 1984-86, the age-adjusted motor vehicle accident death rates for Indians were considerably higher than for the other three major ethnic groups (see table 5). Indian rates in both periods were 2.5 to 5.2 times higher than the other groups, even though the rates for all groups decreased with time. The rate of decrease was 35.5 percent for Indians, which was a more substaritial reduction than among Hispanics and blacks, but not as desirable as the decline among non-Hispanic whites. Among Indian and non-Indian populations in the same State, Indians have higher rates of motor vehicle accident death. Even in the same county, Indians and non-Indians have differential rates of motor vehicle accident death. For example, in Montana in 1974, a colleague of mine and I calculated crude accidental death rates for both Indians and non-Indians on a county-by- county basis. This was facilitated by the fact that reservation boundaries cross several counties, and each county therefore has a substantial number of both Indian and non-Indian residents living in a similar natural ecology. Of the 17 counties characterized by this situation, Indians had higher rates of accidental death (motor vehicle and other) in 11. The overall mortality rate for Indians in these counties was 253 per 100,000 compared to 70 per 100,000 for non-Indians (May and Morigeau 1976). Morbidity Statistics Accidents among Indians put a great burden on the health care system as measured by hospital discharge data. In fiscal year 1987, injuries and poisonings was the leading diagnostic category for Indian and Alaska Native males at IHS and contract hospitals, 2 Actually3.6 times, but since the rates are not age-adjusted, too specific acomparison is not appropriate. Table 4. Age-adjusted death rates (per 100,000 population) by type of Injury, sex, and area for Indians and Alaska Natives, 1980-82 Type of injury EASTERN WESTERN FAR WEST Nashville Oklahoma Aberdeen Bemidji Billings Albuquerque Navajo Phoenix Tucson Portland Alaska Male Motor vehicle traffic Ratio to U.S. all races Other unintentional injury Ratio to U.S. all races Female Motor vehicle traffic Ratio to U.S. all races Other unintentional injury Ratio to U.S. all races Total (both sexes) Motor vehicle traffic Ratio to U.S. all races Other unintentional injury Ratio to U.S. all races 74.4 64.5 160.4 109.9 204.7 104.3 156.8 120.7 162.8 99.5 42.5 2.3 2.0 4.9 3.4 6.3 3.2 4.8 3.7 5.0 3.0 1.3 84.5 36.7 103.0 79.7 149.8 85.6 114.3 106.5 59.7 76.6 277.1 3.1 1.3 3.7 2.9 5.4 3.1 4.2 3.9 2.2 2.8 10.1 22.2 27.8 77.2 2.0 2.5 6.8 4.2 6.7 30.8 0.5 0.7 3.4 46.6 4.1 28.3 3.1 87.2 48.0 51.2 54.1 55.0 49.0 86.7 7.7 4.3 4.5 4.8 4.9 4.3 9.5 35.2 17.6 20.1 24.8 19.2 27.1 11.0 3.9 1.9 2.2 2.7 2.1 3.0 1.0 48.2 45.8 116.8 77.6 145.0 74.8 101.3 85.9 106.2 74.0 27.0 2.2 2.1 5.4 3.6 6.7 3.4 4.6 3.9 4.9 3.4 1.2 42.7 21.1 65.5 53.1 91.1 49.6 64.4 63.9 37.5 51.6 183.5 2.4 1.2 2.3 3.0 5.1 2.8 3.6 3.6 2.1 2.9 10.2 Source: U.S. Congress, Indian Health Care, 1986. YOUTH AND OTHER SPECIAL POPULATIONS Table 5. Age-adjusted and crude mortality motor vehicle accident rates (per 100,000 population) for New Mexico ethnic groups 197878 and 1984-86 213 Crude rate Age-adjusted rates Percent Ratio change Total I&Other Male Female 1976-86 1976 through 1978 Indian Hispanic White non-Hispanic Black 1984 through 1986 Indian Hispanic White non-Hispanic Black 125.5 350.4 - 44.6 51.2 2.9 39.1 42.6 3.5 23.5 29.2 5.2 85.9 97.0 - 1483 36.3 38.8 2.5 58.6 24.9 25.2 3.8 35.6 23.7 27.7 3.5 43.7 243.5 67.9 - 85.7 22.8 - 61.4 26.2 - 39.7 19.4 - 52.0 -35.5 19.8 -24.2 14.9 a.8 10.3 - 5.1 Source: New Mtico Selected He&h Statistics, 1978 and 1986 accounting for 18.7 percent of ail inpatient stays. Among females, the same category was the fourth most frequent inpatient diagnosis, accounting for 6.8 percent of all stays. Within these statistics, motor vehicle-related injuries are the largest category in most areas of the country, rivaled and/or surpassed only in certain age groups by falls (IHS 1988; unpublished IHS program statistics.). Available Literature and Explanatory Themes All the above material raises some interesting questions about why Indians and Alaska Natives are plagued by high rates of accidental death and injury. Realizing that the word "accident" is both unpopular and also inaccurate because vehicle crashes have objective, recognizable, and predictable precursors, it is now expedient to examine the research literature. Few major studies are available on motor vehicle accidents among Indians or Alaska Natives. However, some studies have examin ed all types of accidents combined and also violent death. Most of these articles include motor vehicle accidents as one of a number of variables. Schmitt et al. (1966) studied accidental deaths among Indians in British Columbia (Canada) . In this study, motor vehicle deaths accounted for 24 percent of all Indian accident deaths as opposed to 35 percent for non-Indians. Drownings and burns were the most common causes of death. Boyd et al. (1968) studied accident mortality in Alaska in 1958-62 and found drowning (water transport) as the leading cause of death (62.7 per 100,000) and motor vehicle accidents as the third leading cause, behind fires. Brown et al. (1970) also studied all types of accidents, but used clinic records to examine both morbidity and mortality among the Navajo. Motor vehicle accident deaths accounted for 48 percent of Navajo accident mortality in the late 1960s and 19.3 percent of the injuries. Factors related to motor vehicle crashes were unlicensed drivers, alcohol use, lack of driver education, lax enforcement of laws, and poor roads. Omran and Laughlin (1972) also studied injury morbidity and mortality among the 214 Navajo, but focused only on clinic records from an earlier period (1950s and early 1960s) from a remote, central portion of the reservation. All the studies cited above are rather descriptive and in many ways serve to place the motor vehicle accident experience in a historical context (e.g., in relation to horse and wagon-induced injuries) and in the broader context of all injuries (cuts, falls, etc.). A more recent study of all types of injury was undertaken by Leon Robertson for the Indian Health Service (1985). Using national IHS chic data and existing literature, Robertson reviewed the nature of injury morbidity and mortality as they related to magnitude in all areas served by IHS. Also, there was an attempt to identify individual risk factors that relate to injury. Such high-risk diagnoses are venereal diseases, diabetes, etc. While this was an interesting attempt to identify a general population at risk, the epidemiological and public health implications of this study were limited by a lack of focus on local and specific conditions and socio-cultural variables. i Four major articles/works have been published about the Papago of Southern Arizona. Each work examined the total range of injuries from clinic records in various modern and traditional communities. Hackenberg and Gallagher (1972) found that injury rates were twice as high in the more modern villages and explained this as a function of social change correlating with wage labor, Protestant religion, and higher education. Change produced stress that, coupled with alcohol use, produced more injury. Stull(l972,1973) also studied Papago clinic records and found that the highest injury rates were typical of modern individuals in progressive communities, but that rates of accident and injury were lower for traditional individuals in progressive communities and lowest for traditional people in traditional communities. In a later article, Stull (1977) modified this argument to say that injuries were more prevalent in modern settings "simply because the environment and associated lifestyle place individuals at a greater risk for injury." Alcohol was also mentioned as a contributing factor to injury. One other article that explored the theme of modernization and social stress was written by James Wills (1969), w h o worked among the Oglala Sioux of Pine Ridge, South Dakota. Wills states that a comprehensive, baseline self-report survey showed that accident and injury were higher among fullblood Indians. Further, one half of those individuals involved in vehicular accidents had been arrested at least once for drinking and were also prone to other social and legal problems. The bulk of this study, however, was concerned with brief psychological autopsies of three male Sioux drivers killed in motor vehicle crashes. These drivers were characterized as having problems of adjust- ment, low self-esteem, failure to achieve, and feelings of a lack of control over their own lives. Another study among the Navajo focused solely on motor vehicle accidents. Using police data from three States and from tribal police, this project produced three documents - one monograph (Katz and May 1979) and two article/pamphlets (May and Katz 1979u, b) -that described and analyzed the multivariate nature and epidemiologic features of both Indian and non-Indian fatal (n = 500) and nonfatal crashes (n = 2,347). Data were collected on both the Navajo reservation and roads leading to it in 1973-75 Some of the key findings regarding fatal crashes in this study include: o Mortality rates per 100 million miles traveled (which control for driving ex- posure) were higher for Navajos (12.6 per 100 million) than for the State of Arizona (5.0), New Mexico (6.1), or the U.S. population (3.8). o Of all the fatal Indian crashes, 51 percent were single vehicle, 29 percent were multiple vehicle, and 20 percent were pedestrian. o In fatal single vehicle crashes, Navajos were significantly more likely than non-Indians to be driving a pickup truck, have an invalid license, be younger, and have been drinking. YOUTH AND OTHER SPECIAL POPULATIONS 215 o When comparing Navajo fatal single versus multiple vehicle crashes, single vehicle drivers were significantly more likely to have been drinking (41 percent versus 28 percent), less likely to have a valid license, were younger, and single vehicle fatalities were more likely to happen at any time of day and in the daylight. o Regarding multiple vehicle crashes, the only signifitit difference between Navajo and non-Indian experience was that Navajos were more likely to be driving a pickup. No differences were found between fatal Indian .and non-Indian crashes regarding number of persons in the vehicle, condition of vehicle, or environmental conditions (weather, light, hour of day, weekend, or season). In nonfatal crashes, weather, road, and environment were more important. Therefore, this study pointed heavily, but not exclusively, to the individual characteristics (e.g., youth, alcohol, license) of the Indian driver. Local belief at the time of the study maintained that the roads, environmental conditions, vehicles in poor condition, and driving greater distances were the major determinants of serious crashes. In the study, however, the rates per hundred million miles traveled and other comparisons did not support these beliefs, but rather showed that most fatal crashes were related to driver characteristics. In addition to the specific findings of the Katz and May (1979) Navajo studies, general conclusions were reached as to why Indians have a higher rate of motor vehicle death than others in the United States or surrounding areas. First of all rurality is a key factor. As Waller (1967) and Wailer et al. (1964) have shown, accidents in rural areas produce four times the amount of death than urban accidents of similar severity. Since the Navajo reservation is possibly the most remote area in the U.S. with a very sparse population (150,000 to an area the size of West Virginia), high rates of death are to be expected. The U.S. in general is about 70 percent urban and had a rate of death of 4.20 per 100 million miles traveled in 1973. One would, then, expect the total rate for the Navajo reservation (approximately lo-20 percent urban) to be approximately 12 per 100 million overall and 16 or higher on many roads. Thus lack of access to medical care due to remoteness and time and distance to medical care influence the fatalities greatly. . . . A second major factor in the high rates is the age of the population. Young persons in the U.S. are frequently involved in and killed in accidents. To a similar extent Navajo under 30 years of age are also highly involved in accidents, particularly, single car accidents. The difference, which has the effect of raising Navajo rates, is the sheer number of people who are under 30 years old. While only 50 percent of the U.S. population is under 30, 70 percent of the Navajo population is under 30. Thus, unless age specific rates are calculated, crude rates and rates per 100 million miles will be higher for the Navajo population. The preponderance of young Navajo results in a high risk population. A third factor in the high rates is alcohol and the drinking patterns. Younger persons are more likely to be experimenters with alcohol in the U.S. in general and this is also true with the Navajo. Most Navajo who drink are under 45 years of age (Levy and Kunitz 1974) and the modal drinking style of those who do is sporadic, binge drinking (Ferguson 1968). This style of drinking, coupled with prohibition, is a factor in high accident rates. Under prohibition, with most cultural and ethnic groups, drinking styles emphasize hurried drinking and intoxication (see May 1976). The Navajo are no excep- tion and drinking in border areas such as Gallup, Farmington, Flagstaff, etc., 216 BACKGROUND PAPERS takes this form. After drinking, however, a long drive or walk home of 10 to over 100 miles becomes a more difficult task, especially on two-lane roads. Thus, the higher rates of motor vehicle and pedestrian accidents and alcohol relatedness among the Navajo are influenced by alcohol and particularly the laws of prohibition on the reservation. A final variable in explaining high accident rates is culture. As many authors have pointed out, cultural variables are important in many areas of deviance or "social pathology" (Levy and Kunitz 1974), and for accidents this is true (Hackenberg and Gallagher 1972; Selzer and Viokur 1974). Nevertheless the nature of Navajo acculturation and assimilation should not be ignored. Acculturation and assimilation of cultural items are slow processes and are never uniform. Presently in Navajo society, acculturation has occurred in that the motor vehicle (particularly the pickup truck) has been adopted and is used with great vigor. Assimilation, on the other hand, the internalizing of values from another culture, is a much slower process and has not occurred among many Navajo with regard to motor vehicle use. The Navajo use pickups in a fashion which is strongly based in traditional values. One very prominent value surrounding accidents is fatalism. . . . Related to this is a lack of adoption of safe or defensive driving, many people disregard licensing procedures and laws, and seat belts are rarely used. This is not to deny that overall U.S. and European populations could be accused of these behaviors, but the degree is different. Due to differences in culture and world view, fewer Navajo have adopted safe practices for motor vehicles than some other groups in the U.S. population. (Katz and May 1979) Alcohol Involvement Because of drunk-Indian stereotypes and other folk beliefs about the West, a strong relationship between alcohol and motor vehicle accidents among Indians has been assumed by many (see May and Smith 1988; May 1982). The fact is, however, that there is little concrete evidence in print to document the true extent of alcohol involvement in motor vehicle accidents involving Indians and Alaska Natives. There is, however, enough evidence to generalize here. First of all, a little known fact is that Indian drinking behavior is highly variable from one tribe to the next. In some tribes, more youth and adults drink than the general population but in others, fewer individuals drink. For example, among the Navajo in 1970 only 31 percent of adults drank at all, and many individuals in the over-40 age categories were abstainers (Levy and Kunitz 1974). By 1985, some increase had been registered in Navajo drinking prevalence, which was then believed to be as high as 52 percent, but the prevalence was still lower than the 67 percent reported in the overall U.S. population (May and Smith 1988). The point here is that Indian drinking patterns arehighlyvariable from one tribe to the next, and drinking styles are also unique and may not always conform to popular notions put forth by mainstream stereotypes. Particular drinking styles, however, are important to consider when studying and intervening with the unintentional injury complex. Going back to earlier accident studies, Schmitt et al. (1966) in British Columbia used hospital insurance data to find that 28 percent of the 300 accidental (of all types) native deaths in their study were alcohol-related. Further, they found that among those over 20 years of age, 39 percent of the male accident deaths and 49 percent of the female deaths were alcohol-related. Alcohol intoxication, they concluded, was an important factor in native driving and transport accidents. YOUTH AND OTHER SPECIAL POPULATIONS 217 Other studies dealing with all types of accidents have found and/or speculated upon a native alcohol and injury relationship. Boyd et al. (1968) found that 12 percent of all motor vehicle accidents and drownings in Alaska were alcohol-related. Wii (1969) found a significant correlation between Sioux accident victims and alcohol arrest. Brown et al. (1970) listed alcohol as the most influential factor in Navajo motor vehicle accidents, but provided no data. Omran and Laughlin (1972) reported that of 10 motor vehicle accident deaths studied among the Navajo, 5 were alcohol-related. Among the Papago, both Hackenberg and Gallagher (1972) and Stull(l973,1977) related alcohol consumption to modernization and acculturation stress, and these in turn combined to increase injury rates. Katz and May (1979) found that police records baked alcohol to 41 percent of the Indian drivers in single vehicle crashes, 46 percent of multiple vehicle crashes, and driver or pedestrians in 44 percent 4 of pedestrian fatalities. With all of these studies, however, the role of alcohol was, no doubt, underestimated, for police reports in the 197Os, hospital and clinic records, and death records seldom ., reported the true magnitude of the problem (Zylman and Bacon 1968). Further, police data might be biased in that Indians might be tested more frequently than others as Westermeyer and Brantner (1972) found, or recording of alcohol involvement on Indians may be more common or acceptable in terms of societal norms. For example, in the Navajo studies (Katz and May 1979), the Indian alcohol involvement was 2.8 times higher in single vehicle crashes, 1.7 times higher in multiplevehicle crashes, and 1.3 times higher in pedestrian fatalities than non-Indian events. Maybe this was the actual magnitude, but underreporting and selective reporting of all alcohol involvement must be considered as a possibility with all these studies. Some investigators used more accurate or valid data sources. In an intensive autopsy study of sudden death in British Columbia, 43 of the subjects were Native (Cutler and Morrison 1971). Of the accidents, suicides, and homicides of these Indians, 54.8 percent involved BACs at or above to the legal intoxication limit (0.08) in Canada. The authors concluded that the "high rate of Indian sudden death is primarily the result of occasional, but intense intoxication in high risk situations" (Cutler and Morrison 1971). In another study of sudden death, `the Kenora (Ontario) Planning Council found, through inquiry of knowledgeable parties, that alcohol was a factor in more than 70 percent of the nearly 200 sudden Indian deaths, including even 50 percent of the victims of homicide. This report concluded that the 70 percent may be an underestimate. In Manitoba, Trott et al. (1981) found sudden death to be more common among Native Indians, with 52.5 to 54.3 percent of the accidental deaths alcohol- and/or drug-related. These data, however, included many non-Indians as well as Indians. Jarvis and Boldt (1982) studied the deaths of Alberta Province Indians in a prospective fashion, using interviewers in 35 native communities. Alcohol was involved in 70.6 percent of all motor vehicle accident deaths as reported by informants immediately following a death. The four Canadian studies have increased our knowledge considerably, but the question remains, what about Indians in the United States? In New Mexico, the medical investigator system and the State laboratories are centralized in both authority and in operation. Blood and breath testing are strictly overseen by these centralized bureaus, so the quality of information is good. Since more than 8 percent of the State is populated by American Indians, and several counties are heavily Indian, one might closely examine the alcohol and accident situation among Indians using routine data. Realizing the real risk of committing the ecological correla- tion fallacy in examining county data without linking it to individuals (Robinson 1950), 3 Includes data on all drivers of both vehicles, so that if one or the other or both were cited as intoxicated, it was counted as an alcohol-related crash. 4 The pedestrian involvement included 22 percent of the drivers (excluding the 255percent hit-and-run who maywell have been intoxicated) and 22 percent of the pedestrians. 218 Table 6. Motor vehicle accidents and alcohol in three selected counties in New Mexico, 1982-84 Percent Indian Alcohol involvement (percent) County (major tribe) Fatal accidents Severe accidents* A 66 (Navajo) 78.0 33.1 B 33 (Navajo) 78.0 28.8 C 27 (Pueblo) 73.7 24.2 Entire State 8 63.7 21.2 *Severe accidents are those that involve injury or death. Source: New Mexico Traffic Safety Bureau 1984 and New Medco vital Statistics, 1984. table 6 shows that serious accidents are slightly more alcohol-related in those counties with a substantial Indian population. County A, which is 66 percent Indian, is probabl! the best comparison. In the 3 years covered, 1982 through 1984,78 percent of all fatal accidents and 33 percent of all serious accidents were alcohol-involved as measured by blood or breath alcohol. One final statistic from New Mexico is also available. Several studies, particularly in Canada, referred to one particular Indian pattern or style of drinking-binge drinking- that is characterized by rapid consumption and leads to high blood alcohol levels. Autopsy data examining alcohol levels in blood and vitreous fluid samples from 253 accidental deaths in New Mexico indicated that Indian crash victims had high blood- alcohol levels. The 68 Indian crash victims autopsied in 1986 who were positive for alcohol had a mean blood alcohol level of 0.201, compared to 0.183 for Hispanic positives and 0.126 for comparable non-Hispanic whites `(Guerin, in process). Solutions and Partial Solutions Given all the data that indicate that alcohol and driving are tragically linked in American Indian populations, with more severe consequences than among other groups in the United States, a comprehensive, multivariate prevention strategy must be planned and launched. Because the available information is not as complete as public health professionals might desire, more research to further specify the relationship between alcohol, injury, and Indians is needed in the future. However, current knowledge is clearly adequate for a solid beginning in prevention and intervention. Below, general trends and/or ideas in prevention are highlighted, using as a guide the usual scheme of primary, secondary, and tertiary levels of prevention. Primary preven- tion attacks root causes of a problem in its basic environmental, social, cultural, and political structure. Secondary prevention is directed at eliminating the onset of a problem and/or taking action in the early stages of manifestation. Tertiary prevention seeks to provide curative and remedial action to deal with and minimize the illness, injury, and other negative outcomes of a major health problem. 5 The legal intoxication level in New Mexico is 0.10. YOUTH AND OTHER SPECIAL POPULATIONS Primary Prevention 219 The so&I, economic, and educational status of many American Indians can undoubt- edlybe improved to the overall benefit of many tribes and individuals alike (Dozier 1966). As several authors have pointed out, many Indian groups are undergoing epidemiologic transition (Omran 1971; Kunitz 1983; Broudy and May 1983), and this movement from traditional to modern is related to the high accident rates now prevalent, particularly on Western reservations (Hackenberg and Gallagher 1972; Stulll972,1977,1973; Omran and Laughlin 1972). As Omran and Laughlin (1972) pointed out, people in transition "exhibit many insecurities and inabilities to cope with their changing way of life," and this results in the accident and violence complex among particular segments of Indian societies. Therefore, assisting Indian groups and individuals in their efforts at economic, social and educational enhancement will ultimately benefit many areas, including alcohol abuse and accident and violence rates. In the Native American Adolescent Injury Prevention Program (unpublished data) : run in the rural areas of New Mexico, rural Indian youth were found to have the lowest scores on social psychological scales, such as locus of control and self-esteem, that are predictive of accident susceptibility. However, these scores are highly influenced by the educational level of the parents of the Indian children (see also Liban and Smart 1982). Indian children from families of high educational attainment not only score better on social psychological scales, but they aIso report less substance abuse and risk-taking behavior common to accidents. Therefore, broad social enhancement, such as education, which may strengthen the family, also enhances the individual in ways that should reduce the accident rate over an extended period (Mail 1985; Dozier 1966). Other primary prevention is being, and can be further, pursued. The HIS, Office of Environmental Health, is now working at national, State, and local levels to improve roads, encourage seatbelt use, and promote injury control and state-of-the-art highway safety all over the Indian lands of the Nation. It appears likely that some of this road improFment has already begun to reduce the accident fatality rates in some reservation areas. A final area of primary prevention concerns tribal and community law and policy. Tribes have great lega power to control alcohol policy, taxation, and enforcement on their reservations and communities. More than 60 percent have retained prohibition but are reluctant to enforce strong penalties to support it (May 1975,1976,1977; May and Smith 1988). Tribes need to strictly examine and change alcohol and social policies so that they influence new norms and guidelines of behavior regarding alcohol and sub- stance use, particularly for the minority who abuse alcohol. Whether the policy adopted is a more strict, comprehensive, and effective approach to prohibition, or whether it is legalization that is comprehensive and specific in its prescriptive qualities, a new, more explicit set of policies is needed to guide behavior (see May and Smith 1988; May 1986). Secondary Prevention Alcohol and heahh education with youth must continue and be further expanded. Recent years have seen a great increase in this area, but more has yet to be done. Youth need to be educated, not only about alcohol abuse, but also taught in a social learning mode that emphasizes self-esteem enhancement, values clarification, coping, and prob- lem-solving skills (Bach and Bornstein 1981; Winfree and Griffiths 1983; May 1986). 6 For example, the motorvehicle accident death rate for the Crow and Northern Cheyenne Indians appears to have declined in 19f31-8.5 (IHS unpublished statistics) from what it was in 1959-75 (see May 1976). Some of this reduction may be due to the construction of Interstate Highway 25 from Crow Agency to Billings, Montana, for the old road was a notoriously narrow, hilly, winding, and dangerous stretch of more than 60 miles. 220 BACKGROUND PAPERS Health education with adult Indians should take a different approach. Most adult Indians are well aware of the health implications of alcohol abuse but could benefit greatly from education and exposure to public health principles and the possibilities of public policy for alcohol-abuse and accident prevention (May and Smith 1988). A continuing and open forum on alcohol policy needs to be pursued, and widespread public education will feed this. Policies, laws, and ~paigns that are specific to accident and alcohol abuse hold great promise for a tribe. Positive action in this area is now beginning with safety campaigns, but routiniig the dialog and concern regarding these issues is necessary. Recently, for example, the Navajo Tribe took a major step in this direction by passing a mandatory seatbelt law for theii reservation, the largest in the United States (Landon 1988). The enforcement of such laws and the perception of sure and swift penalties must be emphasized as well (Ross 1982). Safety programs such as infant car-seat purchase and rental have been promoted by the IHS and a number of tribes. Generally, these programs seem to be well received, for most Indians are particularly responsive to child health and weifare issues. The programs, however, need to be expanded to all babies of all tribes. Targeting youth in safety programs may produce a whole new generation of Indian seatbelt users. Currently, surveys of the New Mexico Native American Adolescent Injury Prevention Program (unpublished data) indicate that young Indian males have the lowest seatbelt use of any group in New Mexico. But these data also indicate that increasing seatbelt use is the most common and significant change in behavior of the injury prevention program. Tertiary Prevention Until the middle 197Os, few reservations had any emergency medical resources such as ambulances and trained and employed emergency medical technicians. Now, most reservations and many rural Indian communities have some level of emergency medical services, and these small programs may well have had an influence on the drop in Indian motor vehicle crash fatality rates registered since 1976. However, these services greatly need to be continually upgraded, improved, and expanded so that more accident victims can be saved and/or their injuries minimized, and the high disability rate of Indians can be lowered. Further, emergency room capabilities need to be enhanced, so that accident victims receive state-of-the-art trauma care whenever possible. New Mexico has only one level-one trauma center and some western States have none. The most desirable tertiary prevention would be improved interdiction of drunk drivers before they cause crashes and expanded alcohol treatment capabilities. Tribal, Bureau of Indian Affairs, local sheriff, and police forces in Indian country are notoriously understaffed and in need of further assistance of all kinds. Increasing their training and, particularly, their manpower and resources should be a priority. Further, if the laws and policies regarding alcohol and driving under the influence (primary and secondary prevention) were improved so that the police efforts were facilitated, then enforcement would be greatly aided. For example, on many "dry" reservations, bootlegging is so common and institutionalized that the tribal police can tell you who sells and how much it costs, and can even estimate their income. But to eliminate the practice would require considerably more manpower, time, and new, stronger, currently exist on reservations and in most western States. yd more explicit laws than Alcohol treatment programs in Indian country are characterized by too few resources and capabilities of all kinds (manpower, money, level of training for the counselors, and 7 On the Naajo reservation, for example, the maximum penally for bootlegging is 6 months in jail and a SXlOfine.Policereport, however, that tochargeapersonwith bootleggingtakesanumberofofficermanpower days to build a case, search, collect evidence, and bring it to court. Seldom, however, is any major jail time spent by the accused, and the maximum fine for conviction ($500) is little more than an overhead expense in the cost of doing a profitable business. YOUTH AND OTHER SPECIAL POPULATIONS 221 treatment regimens). A total upgrading and new commitment is needed in this arena, so that all first-time driving-under-the-influence offenders can be screened and placed in an appropriate treatment modality (May 1986). Conclusion In summary, then, the problem of alcohol abuse and accidents exists in many Indian communities. It takes a tremendously large toll in terms of injury and lives lost. Since it has multiple causes and related factors, its solution must also be multivariate. Any program designed to reduce alcohol-related motor vehicle accidents among Indians must be comprehensive and must address both general conditions and specific preven- tion and intervention tasks. REFERENCES Bach, P.J., and Bornstein, P.H. A social learning rationaleand suggestions for behavioral treatment with American Indian alcohol abusers.Addictive Behaviors 6:75&U, 1981. Boyd, D.L.; Maynard, J.E.; and Holmes, L.M. Accident mortality in Alaska.Archives ofEnviron- mental Health 17: 10 1 - 106, 1968. Brod, R.L., and McQuiston, J.M. American Indian adult education and literacy: The first national survey. Journal ofAmerican Indian Education l:l-16, 1983. Broudy, D.W., and May, PA. Demographic and epidemiologic transition among the Navajo Indians. Social Biology 30:1-16, 1983. Brown, R.C.; Gurunanjappa, B.S.; Hawk;RJ.; and Bitsuie, D. The epidemiology of accidents among the Navajo Indians. Public Health Reports 85881-888, 1970. Carr, B., and Lee, ES. Navajo tribal mortality: A life table analysis of the leading causes of death. Social Biology 25:279-287, 1978. Cutler, R., and Morrison, N. Sudden Death: A Study of Charactetitics of Victims and Events Leading to Sudded Death in British Cohanbia Wth Primary Emphasis on ApparentAlcohol Involvement and Indian Sudden Deaths. Vancouver, BC: Alcoholism Foundation of British Columbia, 1971. Dozier, E.P. Problem drinking among American Indians: The role of so&cultural deprivation. Quarterly Journal of Studies on Alcohol 27:72-87, 1966. Ferguson, EN. Navajo drinking: Some tentative hypotheses. Human Organization 27159-167, 1968. Guerin, P.E. "Alcohol Related Traffic Fatalities in New Mexico." MA. thesis, Department of Sociology, University of New Mexico, in process, Hackenberg, R.A., and Gallagher, M.M. The costs of cultural change: Accident injury and modernization among the Papago Indians. Human Organization 31(2):211-226, 1972. Indian Health Service. Chart Book Series, 1988 edition. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1988. Jarvis, G.K, and Boldt, M. Death styles among Canada's Indians. Social Science and Medicine 16( 14): 1345-52, 1982. Katz, P.S., and May, PA. Motor VehicleAccidents on the Navajo Reservation: Z973-1975. Window Rock, AZ: The Navajo Health Authority, 1979. Kenora Social Planning Council. WWe People Sleep: Sudden Deaths in the Kenora Area. Kenora, Ontario: Grand Treaty Council, No. 3, 1973. Kunitz, SJ. Disease Change and the Role of Medicine: The Navajo E@?rience. Berkeley, CA: University of California Press, 1983. Landon, S. Navajos pass reservation seat belt law. Albuquerque Journal July 31, 1988. Levy, J.E., and Kunitz, SJ. Indian Drinking. New York: Wiley Interscience, 1974. Liban, C.B., and Smart, R.G. Drinking and drug use among Ontario Indian students. Drug and Alcohol Dependence 9:161-171,1982. Mail, P.D. Closing the circle: A prevention model for Indian communities with alcohol problems. ZHS Primary Care Provider 10~2-5, 1985. 222 BACKGROUND PAPERS May, PA Arrests, alcohol and alcohol legalization among an American Indian tribe. PZuins Anu'tropoZogist 20(68):129-134, 1975. May, PA. "Alcohol Legalization and Native Americans: A Sociological Inquiry". Ph.D. disserta- tion, University of Montana, 1976. May, PA. Alcohol beverage control: A survey of tribal alcohol statutes. American Indian Luw Review 5:217-228,1977. May, P.A Substance abuse and American Indians: Prevalence and susceptibility. Infern~bnaZ Journal of the Addictions 17: 11851209, 1982. May, PA Alcohol and drug misuse prevention programs for American Indians: Needs and opportunities. Journal of Studies on Alcohol 47(3):187-195,1986. May, PA. The health status of Indian children: Problems and prevention in early life. In: Manson, S., and Dinges, N., eds. Health and Behavior: A Research Agenda for American Indians andAlaska Natives. Denver, CO: University of Colorado Health Sciences University Press, 1988. May, P.A, and Morigeau, G. Unpublished statistics, University of Montana, 1976. May, P.A, and Katz, P.S. Motor Vehicle Accidents on the Navajo Reservation, 1973-1975: Health Planning Swnmary. Window Rock, Arizona: Navajo Health Authority, 1979a. May, P.A, and Katz, P.S. Motor Vehicle Accidents on the Navajo Reservation, 1973-1975: Public Information Pamphlet. Window Rock, Arizona: Navajo Health Authority, 1979b. May, P.A., and Smith, M.B. Some Navajo Indian opinions about alcohol abuse and prohibition: A survey and recommendations for policy. Journal of Stidies on Alcohol 49:324-334, 1988. Native American Adolescent Injury Prevention Program. Unpublished program data. Health and Environment Department, Public Health Division. A SPRANS-funded project. Santa Fe, NM. New Mexico Traffic Safety Bureau. Diving While Znroticured in New Mexico. Santa Fe: Depart- ment of Transportation, 1984. Omran, AR. The epidemiologic transition. Milbank Memorial Fund Quarterly 49509-538, 1971. Omran A.R., and Laughlin, B. An epidemiologic study of accidents among the Navajo Indians. Journal of Esyptian MedicalAssociation 55: l-22, 1972. Passcll, J.S., Provisional evaluation of the 1970 Census count of American Indians. Demography 13(3):397-409,1976. Passell, J.S., and Berman, PA. Quality of 1980 census data for American Indians. Social BioZogv 33(3-4):163-182, 1986. Robertson, L.S. "Epidemiological Assessment of the Contributing Factors of Injury Mortality and Morbidity Among Native Americans." Manuscript. New Haven, CT: Yale University, 1985. Robinson, W.S. Ecological correlations and the behavior of individuals. American Sociological Review 15(3):351-357, 1950. Ross, H.L. Deterring the Drinking Driver. Lexington, MA Heath, 1982. Schmitt, N.; Hole, L.W.; and Barclay, W.S. Accidental deaths among British Columbia Indians. Canadian Medical Association Journal 94228234, 1966. Seizer, M., and Vinokur, A. Life events, subjective stress, and traffic accidents.American Journal of Psychiahy 13 1903906, 1974. State of Montana. Montana VifuZSfutistics, 1982edition. Helena, MT: State Department of Health and Environmental Sciences, 1983. State of Montana.Monrana I'imlSrarisrics, 1983 edition. Helena, MT: State Department of Health and Environmental Sciences, 1984. State of Montana.Montana VTtaZStatistics, 1984edition. Helena, MT: State Department ofHealth and Environmental Sciences, 1985. State of Montana.Montana ViralSWstics, 1985 edition. Helena, MT: State Department of Health and Environmental Sciences, 1986. State of Montana.Monrana VitalStatistics, 1986edition. Helena, MT: State Department of Health and Environmental Sciences, 1988. State of New Mexico. NewMexico Selected Health Statistics, 1978Annual Report. Santa Fe: Health and Environment, 1980. State of New Mexico. New M&co Selected Health Statistics, 1986AnnuaZReport. Santa Fe: Health and Environment, 1988. Stull, D.D. Victims of modernization: Accident rates and Papago Indian adjustment. Human Organization 31(2):227-240, 1972. YOUTH AND OTHER SPECIAL POPULATIONS 223 Stull, D.D. "Modernization and Symptoms of Stress: Attitudes, Accidents and Alcohol Use Among Urban Papago Indians." Ph.D. dissertation, University of Colorado, 1973. Stull, D.D. New data on accident victim rates among Papago Indians: The urban case. Human Organization 36(4):395-398, 1977. Tsai, S.P.; Lee, E.S.; and Hardy;RJ. The effect of a reduction in leading causes of death: Potential gains in life expectancy. American Journal of Public Health 68(10):9&i-971, 1978. Trott, L.; Barnes, G.; and Dumoff, R. Ethnicity and other demographic characteristics as predic- tors of sudden drug-related deaths. Journal of Studies on Alcohol 42:564-78, 1981. U.S. Bureau of Census. American Indian Areas and Alaska Native Kllages, 1980. Supplementary Report (PCXO-51-13). Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 19%. U.S. Bureau of Census. A Statistical Profile of the American Zndian Population: 1980 Census. Census Fact Sheet. Washington DC: Supt. of Dots., U.S. Govt. Print. Off., 19846. U.S. Congress, Office of Technology Assessment. Indian Health Care. OTA-H-290. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1986. Wager, J.A Control of accidents in rural areas. Journal of the American Medical Association 201(3):94-98, 1967. Waller, J.; Curran, R.; and Noyes, F. Traffic deaths: A preliminary study of urban and rural ' fatalities in California. California Medicine 101:272-276, 1964. Westermeyer, J., and Brantner, J. Violent death and alcohol use among the Chippewa in Min- nesota. Minnesota Medicine 55:749-752, 1972. Wills, J.E. Psychological problems of the Sioux Indians resulting in the accident phenomena. Pine Ridge Research Bulletin 84963, 1969. Winfree, L. T., and Grifiiths, C.T. Youth at risk: Marijuana use among Native American and Caucasian youths. International Journal of the Addictions 18:53-70, 1983. Zylman, R., and Bacon, S.D. Police records and accidents involving alcohol. Qrlartedy Journal of Studies onAlcohol Supplement 4: 178-211, 1968. Youth and Other Special Populations Drunk Driving Among Blacks and Hispanics Jan M. Howard, Ph.D., Elsie D. Taylor, H. Laurence Ross, Ph.D., and Mary L. Ganikos, Ph.D. Prevention Research Branch, National InshMe on Alcohol Abuse a&d Alcoholism Blacks comprise about 12 percent of the U.S. population, and Hispanics about 8 percent (Bureau of Census 1987). Research suggests that problem drinking and asso- ciated mortality rates are higher in these two minority groups than in the general public (Herd 1985; NIAAA 198&z). Yet, few studies have investigated the issue of drunk driving among blacks and Hispanics. This chapter summarizes relevant characteristics of these two groups, their drinking practices, and available information on their drunk driving behavior. Demographic Characteristics Such factors as age, socioeconomic status, and rural/urban distribution may have significant effects on overall consumption patterns and drinking and driving behavior. Thus, the alcohol problems noted among blacks and Hispanics may be partially a function of their particular demographic profiles. On average, the black and Hispanic populations tend to be younger than whites and underrepresented among persons 65 years of age and older (Bureau of Census 1987). For both blacks and Hispanics, socioeconomic status (as measured by income and education) is significantly lower than that for whites (Bureau of Census 1987). Among civilians age 16 and older, 54 percent of blacks are employed, versus 59 percent of Hispanics and 62 percent ofwhites (Bureau of Census 1987). Compared to whites, blacks and Hispanics are much less likely to live in rural areas (Bureau of Census 1984) where driving is a necessity. The Hispanic population in the United States is a very heterogeneous group, with diverse national and sociocultural backgrounds (NIAAA 1987). According to the 1980 census, approximately 60 percent of the U.S. Hispanic population is of Mexican origin; 15 percent are considered Puerto Rican, 6 percent Cuban, and the remaining 20 percent are linked to other countries, including those in Central and South America (NIAAA 1982a; NHTSA 1987; DHHS 19866). The black population of the United States is ah.0 heterogeneous, reflecting different cultural, regional, and socioeconomic perspectives. Consumption Patterns Blacks The data on drinking practices among blacks are somewhat inconsistent and difficult to interpret. Blacks are at high risk for alcohol-related medical problems, especially liver YOUTH AND OTHER SPECIAL POPULATIONS 225 cirrhosis and esophageal cancer (NIAAA 1987), and their cirrhosis mortality rate is nearly twice that for whites (Herd 1985). Yet, blacks of both sexes report higher abstention rates than whites; and among drinkers, black men are less likely than white men to drink heavily (NIAAA 1987). Among female drinkers, the pattern is reversed; black women are more likely than white women to drink heavily (NIAAA 1987). The relationship between age and drinking practices appears to differ by race. Alcohol consumption among white males age 18-29 is high, with a decline after age 30 (NIAAA 1987). Consumption among blacks, however, is relatively low in the 18-29 age group, rises dramatically among those in their 3Os, and declines after age 39 (NIAAA 1987; NIAAA 1982b). According to a recent report prepared for the National Highway Traffic Safety Administration (NHTSA), black males in younger age groups "are at substantially less risk for high rates of heavy drinking than younger whites" (NHTSA 1987). Moreover, studies of adolescents consistently show lower rates of problem drinking and related arrests among blacks than whites (DHHS 19%). On the other hand, data suggest that blacks enter alcohol treatment programs at younger ages than whites, the peak age for blacks being35-44 compared to 45-54 for whites (NIAAA 1987). The relationship between income level and drinking practices also appears to vary by race. For white men, increased income has been associated with increased heavy drinking (NIAAA 1987; Herd 1985). For black men, the reverse is true; increased income has been associated with decreased heavy consumption (NIAAA 1987; Herd 1985).' With respect to women, however, the two races are similar: increased income is as- sociated with more frequent (as opposed to heavy) drinking (NIAAA 1987). Drinking practices among blacks may differ by geographic region. In the Northeast, for example, the proportion of blacks in alcohol treatment is reported to be two to three times higher than their proportion of the regional population (NIAAA 1987; DHHS 1986u). But in the interior South, the number of blacks in treatment is generally proportional to their representation in the population (NIAAA 1987; DHHS 198&z). Hispanics Compared to the general U.S. population, Hispanics in the aggregate have relatively high rates of heavy drinking and alcohol problems (DHHS 19866). Yet, differences can be observed within the Hispanic population. As an example, Hispanic men in this country have relatively high rates of alcohol use and abuse (NIAAA 1987) as well as cirrhosis mortality (NIAAA 1982a), while Hispanic women show high rates of abstention (NIAAA 1987). As is true for whites in general, consumption for both sexes seems to increase with increased income and educational levels (NIAAA 1987; Wilson and Williams 1983). With respect to age-related drinking problems, Hispanics are more similar to blacks than to whites. Whereas drinking problems among whites decline abruptly from their 20s to 3Os, for Hispanics (and blacks) problems increase from their 20s to 30s and then decline gradually in their 40s (NHTSA 1987). First-generation American-born Hispanic men tend to drink more heavily than Hispanic-American men born abroad (NIAAA 1987). Mexican Americans appear to consume more alcohol and report more drinking problems than the other Hispanic groups (Caetano 1988). The rate of "frequent high maximum" drinking is relatively large for Mexican-American women as well as men (Caetano 1988). Both sexes, however, also show high rates of abstention (NIAAA 1987; Caetano 1988). Compared to the other Hispanic females, Puerto Rican women have the lowest abstention rate, but data suggest that they are largely moderate drinkers withvery few heavy drinkers among them (NIAAA 1987). 1 Yet data from the 1983 National Health Interview Surveyshowsimilar income effects for black and white males (Wilson and Williams 1983). 226 BACKGROUND PAPERS Among Hispanic-American men born abroad, Mexican Americans have a low rate of abstention and a rate of heavy drinking six times that of any other national subgroup (NIAAA 1987). Conversely, Mexican-American women born abroad have a relatively high rate of abstention and virtually no heavy drinking (NIAAA 1987). Drunk Driving Behavior One of the barriers to obtaining information on drunk driving among racial and ethnic minorities is the failure of the records system to collect appropriate data. Depending on the purpose for which drunk-driving information is gathered, ascertaining race and ethnicity may be deemed unimportant. Political and legal considerations may further limit the availability of relevant data. Drunk driving is a criminal and civil offense that can have severe personal consequences for the driver involved, Moreover, the stig- matization of individuals accused of drunk driving can also taint the groups to which they belong. Under these circumstances, government authorities may be reluctant to collect pertinent information on specific ethnic groups, or they may decline to release data that has been collected. The use of death certificates to determine the proportion of traffic fatalities at- tributable to drunk driving grossly underestimates the contribution of alcohol to such deaths (Dufour et al. 1985). The physician who completes the death certificate may not be the one who provided medical care to the patient and would not necessarily be aware of alcohol involvement (Dufour et al. 1985). Another problem is that some physicians do not recognize the importance of death certificate data for advancing medical and scientific knowledge and fail to record complete information (Dufour et al. 1985). And even when they have the full picture, they may wish to protect the deceased person's family from the stigma and financial liability associated with an alcohol-induced accident. A more appropriate national data system for investigating alcohol involvement in traffic fatalities is the Fatal Accident Reporting System (FARS), which contains detailed information on the driver, vehicle, and environmental characteristics associated with each traffic death (Dufour et al. 1984). *Unfortunately, FARS does not include data on the race of persons involved in these fatal accidents. Death certificates indicate the race of the deceased, but not FARS. Thus, to generate statistics on racial differences in alcohol-related traffic deaths in the United States, FARS has been linked with the Multiple Cause of Death (MCD) data system, which is based on death certificates (Dufour et al. 1984). This linkage makes it possible to investigate the effects of race on alcohol-related traffic fatalities for the country as a whole. With respect to Hispanics, however, missing data are still a problem, because the majority of States do not currently include the category "Hispanic" on their death certificates. Beginning in 1989, 10 States will pilot test a new death certificate that does contain a Hispanic category, and it is hoped that by 1990, all 50 States will collect such information (personal communication M. Dufour, NIAA4, January 1989). Despite these various limitations, important relationships have been observed be- tween race or ethnicity and drunk driving. Blacks and Hispanics both appear to be at high risk for alcohol-related driving problems. 2 A FARS accident is one that involves a motorvehicle moving on a roadway customarily open to the public and resulting in the death of a person (occupant or nonoccupant of thevehicle) within 30 days of the accident (Dufour et al. 1984). YOUTH AND OTHER SPECIAL POPULATIONS 227 Blacks The report prepared for NHTSA concluded that blacks are at greater risk than whites for traffic accidents due to drinking (NHTSA 1987). Germane to thii conclusion is the Grand Rapids, Michigan study which showed that nonwhite drivers were involved in proportionally more collisions than whites (Zyhnan 1972; Hyman 1968~). Among those experiencing collisions, a greater proportion of nonwhites than whites had been drinking, and the BAC levels for nonwhites were considerably higher than those of whites (Zylman 1972; Hyman 196&r; Cosper and Mozersky 1968). According to Hyman's analysis, the higher accident vulnerability of the nonwhite males extended across every BAC and educational category. Even in the control group (who were not involved in accidents), nonwhites were overrepresented in the high BAC categories. Linkages between FARS and the Multiple Cause Mortality records were effected for 2,700 Oklahoma residents who died in motor vehicle accidents in the late 1970s (Dufour et al. 1984). This facilitated the investigation of relationships between race and alcohol involvement in these deaths. Results indicated that 46 percent of the black deaths and 41 percent of the white deaths were alcohol-related. Alcohol involvement was deter- mined by BAC testing for 91 percent of the blacks and 78 percent of the whites. In the remaining cases, it was determined by the judgment of the investigating officer. A much stronger relationship between race and alcohol involvement in fatalities was reported by Waller and his colleagues (1969). They found that 76 percent of black drivers killed in traffic crashes in California had been drinking, compared to only 56 percent of whites. Moreover, 53 percent of the blacks had BACs above a level of 0.15 percent compared to only 34 percent of the whites. As recently as 1973, the drunk-driving arrest rate for black adults in the United States was nearly twice as high as that for whites, but since then the two rates have converged (Herd 1985). Thus, in 1981 the DUI arrest rate among persons 18 years of age and older was 951 per 100,000 population for blacks and 917 for whites (Caetano 1984). And for persons under 18 years of age, the blacks actually showed a much lower DUI arrest rate than the whites-6.5 arrests per 100,000 population for blacks compared to 47.7 for whites (Caetano 1984). Further data on arrests come from studies in specific localities. In a 1968 report, black males in Columbus, Ohio were twice as likely to be arrested for driving while intoxicated as were other men in the 20-64 age range (Hyman 1968b). But over the 1972-75 period, Rabow and Watts (1982) found no significant correlation between the percentage of black households in 51 California counties and arrest rates for misdemeanor or felony drunk driving. A recent analysis of recidivism patterns in Mississippi (Wells-Parker et al. 1988; Anderson personal communication 1988) suggests that blacks in that particular State are at higher risk of being rearrested for drunk driving than whites. The study found that 61 percent of black offenders under age 25 and 50 percent over age 25 were rearrested for DUI during the 6- to 9-year tracking period beginning in 1976. Recidivism rates for whites in the same age groups were 46 percent and 41 percent, respectively. When the investigators controlled for other offender characteristics such as age, level of education, and severity of drinking problem, the blacks were approximately 1.4 times more likely to recidivate than the whites. For reasons that are unclear, limited survey data indicate that blacks are less likely to report driving while drunk than are whites (MAAA 1987; Herd 1985). Among men, the reported rate for whites is more than two and a half times higher than that for blacks. And among women, the white rate is more than five times higher than the black rate 228 BACKGROUND PAPERS (Herd 1985). 3 The possible influence of differential car ownership on these relation- ships has not been discussed. Hispanics Caetano's analysis of FBI statistics for 1981 indicated that among persons 18 years of age and older, the arrest rate for driving under the influence was more than twice as high for Hispanics as non-Hispanics (Caetano 1984). Specifically, the rate per 100,000 population was 1,712.2 for Hispanics and 742.6 for.non-Hispanics. Yet, the DUI arrest rate among persons younger than 18 years of age was only slightly higher for Hispanics than non-Hispanics (39.8 per 100,000 population compared to342 (Caetano 1984)). An examination of lotal arrests for Hispanics and non-Hispanics in the 18-plus age range showed that 19 percent of the Hispanic arrests were for driving under the influence, compared to 17 percent of the non-Hispanic arrests (Caetano 1984). I Studies of California populations support Caetano's finding that Hispanic arrests for drunk driving are relatively high. Hyman (1968b) reported an overrepresentation of Spanish surnames among persons arrested for drunk driving in Santa Clara County. After controlling for urbanization and income, Rabow and Watts found a significant correlation (.42) between misdemeanor arrests for drunk driving and the percentage of Spanish Americans in 51 California counties (Rabow and Watts 1982). However, there was no significant correlation for felony drunk-driving arrests. Consistent with these findings are three other California studies summarized by Caetano (DHHS 1986Ir). All showed an overrepresentation of Hispanics among arrestees for drunk driving (also see NIAAA 1982u). In addition, several studies indicated that Hispanics are disproportionately involved in alcohol-induced crashes. During the 197Os, May and Baker (1974) found an over- representation of Hispanic drivers in alcohol-related traffic accidents in New Mexico, and Alcocer reported that the rate of traffic accidents resulting in injuries or fatalities was higher for Hispanic than non-Hispanic neighborhoods of Los Angeles (DHHS 1986b). Moreover, a higher proportion of the Hispanics than non-Hispanics who had been arrested- for DWI in Hyman's Santa Clara study had been involved in accidents (Hyman 1968b; Hyman et al. 1972). The BAC levels also tended to be higher for the Hispanic than non-Hispanic arrestees (Hyman 1968b). Interpretation Drunk driving can be measured directly through roadside surveys, including breath- alcohol tests of drivers. It is indirectly measured by crashes ("accidents"), especially fatal crashes, which are strongly correlated with drunk driving, and by arrests, although the latter is a weaker index because of its sensitivity to the level of police activity. In general, the available evidence suggests that black and Hispanic drivers are more likely than members of other groups to be impaired by alcohol. Roadside survey data obtained in the Grand Rapids study (Zylman 1972; Hyman 1968~~; Cosper and Mozersky 1968) revealed that blacks were more likely than whites to drive with high BACs and to have crashes. The study stands alone, as more recent roadside surveys have failed to gather racial and ethnic data. Unfortunately, the Grand Rapids study cannot be considered recent, but its conclusion of black overinvolvement in drunk driving is supported by independent studies of involvement in fatal crashes 3 Data from Cosper and Mozershys study of Grand Rapids and "Greentown" respondents are consistent with these patterns (Cosper and Mozersky 1968). YOUTH AND OTHER SPECIAL POPULATIONS 229 (Dufour et al. 1984, Waller et al. 1%9). These latter studies show that when blacks have fatal crashes, they are more likely than whites to have alcohol in their blood. However, these investigations do not in themselves demonstrate a higher fatal crash rate for blacks on a per capita basis. The studies using arrest (Herd 1985; Caetano 1984, Hyman 1%8b; Rabow and Watts 1982) and rearrest (Wells-Parker et al. 1988) data have produced inconsistent findings in the matter of black overinvolvement, perhaps related to the weakness of the criterion variable (arrests). Since 1965, both whites and blacks have shown dramatic increases in DUI arrest rates among persons 18 years of age and older, but whites have shown the greater change over time, thereby closing the gap between the races (see figure 1). Although the reasons for these patterns are unclear, several other trends suggest that increased law enforcement (DHHS 1986a) is the explanation rather than increased drunk driving. The mileage- based highway mortality rate has been declining over time (National Safety Council 1986) as has the proportion of fatal crashes in which alcohol is implicated (NHTSA 1988), and roadside surveys are finding proportionately fewer alcohol-impaired drivers than in earlier years (Wolfe 1986). 900 - 600 - too - 600 - so0 - A 6kk 0 wtr 200 - ,# I%5 IOX) 1973 1980 YEAR OF ARREST Figure 1. U.S. arrest rates for driving under the influence: Persons 18 years and over, by race, 1965-1982 Source: This figure was reproduced from Herd (DHHS 19841, page 101). The data sources for arrest rateswere the Uniform Crime Reports and Current Population Reports. The 1981 arrest rates portrayed here are lower than those reported by Caetano (1984) because the two authors used somewhat different estimates of the relevant 1981 populations. However, in both cases the arrest rates for blacks and whites were quite similar. 230 BACKGROUND PAPERS Concerning the drinking and driving of Hispanics, no roadside survey-based studies are available. However, several studies reported overinvolvement of Hispanics in alcohol-related crashes (DHHS 1986b; Hyman 196%; Hyman et al. 1972), and large excesses of drunk-driving arrests are reported in the literature (Caetano 1984; Hyman 196%; Rabow and Watts 1982). The consistency of these findings leads to the conclusion that Hispanics are unusually likely to engage in drinking and driving. One possible interpretation of the findings concerning arrests of Hispanics, and also blacks, is that police are disproportionately likely to arrest minority-group members (DHHS 1986b). If so, the reported overinvolvement of minorities in drunk drivingwould be misleading (see Morales cited in DHHS 19863). However, the only study on the issue (Hyman et al. 1972) reported that a higher proportion of Hispanic than non-Hispanic drunk drivers are arrested as a consequence of accidents rather than moving violations, a fact inconsistent with the idea that the excess of arrests stems from police bias. Studies showing excess alcohol-related crashes for minorities may have to contend with different proportions of blood-alcohol testing (Dufour et al. 1984), which again could produce misleading conclusions. However, the finding that fewer of the majority group are tested suggests that their relative rate of impaired driving may be overestimated (assuming that testing is first done on those most reasonably suspected of drinking, and that the rate is calculated on the basis of the population tested). Thus, the estimate of excessive involvement of the minority is likely to be conservative rather than overstated. If the conclusion of minority overrepresentation in drunk driving is valid and not merely a measurement artifact, it requires further explanation. Two straightforward interpretations are that minority status is related to more heavy drinking and to more frequent driving. However, the latter possibility (excessive driving exposure) could potentially explain only population-based indexes, like ordinary arrest rates. A third possible interpretation is that minority-group members, though neither dispropor- tionately drinking nor driving, are more likely to combine these behaviors. The first explanation, heavier drinking, is clearly supported for Hispanics, though not so clearly for blacks. Why Hispanics, in particular, should be disproportionately heavy drinkers may be explicable on cultural grounds. The "machismo" (manliness) attitude of Hispanic men (Ames and Mora 1988) may contribute to their higher rates of alcohol-related arrests and accidents by leading to an increase in risk-taking behavior. Relevant also is the "cruising" practiced by Hispanic men (and occasionally women) in the Southwest. This cultural practice involves driving slowly while drinking alcoholic beverages and flirting with the opposite sex. It is facilitated by a social network that encourages and sanctions the simultaneous use of automobiles and alcohol. It also makes these men more visible to police and more vulnerable to arrest (J. Cuellar, Prevention Research Center, Berkeley, CA, personal communication October 1988). The second explanation - that minorities drive more than other people- has no empirical support. Rather, minority-group family automobile ownership and individual driving appear to be lower, and mileage less, at least among blacks (Cosper and Mozersky 1968; Hyman 196%). The relatively high urbanization of both blacks and Hispanics may reduce their need and opportunity to drive. Thus, the excess of minority arrests for drunk driving would seem to be conservatively stated. Zylman's (1972) suggestion that black overrepresentation in accidents may be due to residing (and hence driving) in congested areas must be regarded as merely speculative. A related speculation, also not yet empirically grounded, is that minority-groupmembers have less access to transportation alternatives when drinking; these could include formal alternatives such as mass transit and taxis, or informal ones such as the practice of accepting designated drivers in drinking groups. Indirect support for the third explanation, mixing drinking and driving, comes from the Grand Rapids study finding that, compared with others, blacks offer a larger estimate of the safe number of drinks that can be consumed before driving (Cosper and Mozersky YOUTH AND OTHER SPECIAL POPULATIONS 231 1968). Moreover, the suggestion has been made that minority-group members may disproportionately drink in locations away from home, such as bars and parking lots, presenting the inducement to drive while impaired. However, a recent national survey found only equivocal support for this idea (Caetano and Herd 1987). In any event, the need to use automobiles for nearly all social purposes is so fundamental in American life that such differences would probably be only marginal in their effect. Both blacks and Hispanics are disproportionately youthful groups, which leads to the suggestion that their overinvolvement in drunk driving may merely reflect their demographic characteristics. However, among young people in both groups, the dii- ferences from the majority in arrest rates are either reduced or reversed (Caetano 1984), a finding that contradicts the demographic explanation. A commonly offered explanation of minority involvement in drunk driving hinges on the low socioeconomic status of blacks and Hispanics in America. However, controls for occupation, education, and income, although possibly successful in reducing the linkage between minority status and indexes of drunk driving, do not seem to eliminate the relationship (Zylman 1972; Hyman 1968~; Rabow and Watts 1982). In brief, the excess involvement of minorities in drunk driving seems to be real and not a statistical artifact. It is not merely the reflection of age or class differences related to race and ethnicity. Rather, it is best interpreted as a consequence of minority-group members' excess involvement in drinking, especially in heavy-drinking episodes. This may be marginally compounded by a relatively greater conjunction of drinking and driving behaviors in the groups discussed. The development of appropriate policy countermeasures would thus seem to be served by a more thorough understanding of the definitions of drinking and of driving in the black and Hispanic subcultures of the United States. Research Priorities The data on black drinking patterns suggest inconsistencies between self-reported consumption practices and medical problems induced by alcohol. Research is needed to clarify the reasons for the apparent inconsistencies-whether, for example, self- reports of blacks are distorted in the direction of social desirability, or whether a longer duration of drinking among blacks (DHHS 198ti) increases the risk of medical problems. Since the full extent of the drunk driving problem is also not clear in either of these minority groups, more precise and complete measures are needed of alcohol use and abuse among blacks and Hispanics, with particular reference to drunk driving. In addition, it is important to determine the extent to which the higher DUI arrest or rearrest rates among these two groups reflect real differences in drunk driving or differential law enforcement by police. The relationship of other variables to drinking patterns and drunk driving among blacks and Hispanics should also be more extensively explored. Such variables include age, income level, geographic and urban/rural distribu- tion, and car ownership. Perhaps most important, researchers should develop and test prevention strategies tailored to specific subcultures in which alcohol contributes to social bonding, social status, social integration, and coping with misfortune. The appropriate choice of inter- vention strategies will be enhanced by a better understanding of the cause of drunk driving problems among blacks and Hispanics. However, the evidence at hand is sufficient to begin the process of prevention research. 232 REFERENCES Ames, G. and Mora, J. "Alcohol Problem Prevention in Mexican-American Populations." Un- published paper. Berkeley, CA: Prevention Research Center, 1988. Bureau of Census. I980 Census of Population, General Social and Economic Characteristics, United States Smnmary, PC-80-l-Cl. Washington, DC Supt. of Dots., U.S. Govt. Print. Off., 1984. pp. I-12, l-13. Bureau of Census. StatisticalAbstract of the United States: 1988,108th Edition. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1987. p. 37. Caetano, R. A note on arrest statistics for alcohol-related offenses. DrinEdng Md Dnrg 13acrices Surveyor 19(Apr.):12-17, 1984. Caetano, R. Alcohol use among Hispanic groups in the United States.j4rnetica?t Jownal of Drug and Alcohol Abuse 14:293-308,1988. Caetano, R. and Herd, D. "Drinking in Different Social Contexts Among White, Black and Hispanic Men." Presented at the Sixth Annual Meeting of &American College of Epidemiology, New Orleans, Oct. 1-2, 1987. Ciqer, R. and Mozersky, K Social correlates of drinking and driing. Quarterly Journal of Studies on Alcohol, Supplement 4:58-l 17, 1968. Department of Health and Human Services. A review of drinking patterns and alcohol problems among U.S. blacks, by Herd, D. In: Report of the Secretary's Task Force on Black ana' Minority Health, Vol. VII: ChemicalDependency and Diabetes. Washington, DC Supt. of Dots., U.S. Govt. Print. Off., 198&r. pp. 7-140. Department of Health and Human Services. Patterns and problems of drinking among U.S. Hispanics, by Caetano, R. In: Report of the Secretaty's Task Force on Black and Minority Health, Vol. VII: Chemical Dependency and Diabetes. Washington, DC Supt. of Dots., U.S. Govt. Print. Off., 19m. pp. 143-186. Dufour, M.; Malin, H.; Bertolucci, D.; and Charles, C. "Death Certification Practices in Aicohol- Related Traffic Fatalities." Presented at the 112 Annual Meeting of the American Public Health Association, Anaheim, CA, Nov. 1984. Dufour, M.C.; Bertolucci; D.; and Weed, J. "Multiple Cause Mortality Data: General Description, Methodological Issues, and Preliminary Findings." Presented at the National Institute on Aicohd Abuse and Alcoholism Conference on the Epidemiology of Alcohol Use and Abuse Among U.S. Ethnic Minority Groups, Bethesda, MD, Sept. ll-14,1985. Herd, D. "The Epidemiology of Drinking Patterns and Alcohol-Related Problems Among U.S. Blacks." Presented at the National Institute on Alcohol Abuse and Alcoholism Conference on the Epidemiology of Alcohol Use and Abuse Among U.S. Ethnic Minority Groups, Bethesda, MD, Sept. 1 l-14,1985. Hyman, M.M. Accident vulnerability and blood alcohol concentrations of drivers by demographic characteristics. Quarterly Journal of Studies on Alcohol, Supplement 4:34-57,196&1. Hyrnan, M.M. The social characteristics of persons arrested for drivingwhile intoxicated. Quarter& Journal of Studies on Alcohol, Supplement 4:138-177,1968b. Hyman, M.M.; Helrich, AR.; and Benson, G. Ascertaining police bias in arrests for drunken driving. Quarterly Journal of Shdies on Alcohol 33: 148-159,1972. May, G.W., and Baker, W.E.Human and environmental factors in alcohol-related traffic accidents. In: Israelstam, S., and Lambert, S., eds. Alcohol, Drugs, and Trafic Safety: Proceedings of the Sirth International Conference on Alcohol, Drugs and Traffic Safety. Toronto: Addic- tion Research Foundation of Ontario, 1974. pp. 129-145. Smith, J. and Shaw, J. "Interim Report Identifying Population Subgroups at High Risk for Impaired Driving." Unpublished report by Triton, Inc., submitted to Zelphia Gouldson, U.S. Department of Transportation, Washington, DC, May, 1987. National Highway Traffic Safety Administration. FatalAccident Reporting System 1986: A Review of Information on Fatal Trafic Accidents in the United States in 1986. Washington, DC: 1988. p. v. National Institute on Alcohol Abuse and Alcoholism. Alcohol and Health, Siah Special Report to the U.S. Congress. DHHS Pub. No. (ADM)87-1519. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1987. National Institute on Alcohol Abuse and Alcoholism. Alcohol use and abuse among the Hispanic American population, by Alwcer, AM. In: Alcohol and Health Monograph No. 4: Special YOUTH AND OTHER SPECIAL POPULATIONS 233 Populafion Issues. DHHS Pub. No. (ADM)82-1193. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 1982. pp. 361-382. National Institute on Alcohol Abuse and Alcoholism. Alcoholism: Studies regarding black Americans 1977-1980, by King, L.M. In: Alcohol and Health Monograph No. 4: Special Popuiution Zssrtes. DHHS Publication No. (ADM)82-1193. Washington, DC: Supt. of Dots., U.S. Govt. Print. Off., 19826. pp. 385407. National Safety CouncilAccidenr Facts. Chicago: the Council, 1986. p. 59. Rabow, J. and Watts, R.K Alcohol availability, alcoholic beverage sales and alcohol-related problems. Journal of Shrdies on Alcohol 43:767-801,1982. Wailer, J.A; King, E.M.; Nielson, G.; and Turkel, H.W. Alcohol and other factors in California highway fatalities. Journal of Forensic Sciences 14:429#, 1%9. Wells-Parker, E.; McMillen, D.L.; Anderson, BJ.; and Ian&urn, J.W. Interactions among DUI offender characteristics and traditional intervention modalities: A long-term recidivism follow-up. Submitted for publication, 1988. Wilson, R.W. and Williams, G.D. "Alcohol Use and Abuse Among U.S. Minority Groups: Results from the 1983 National Health Interview Survey." Unpublished paper. Wolfe, AC. Changes in the incidence of drunk driving in the United States 1973-1986. In: Noordzij, P.C., and Reszbach, R. eds. Alcohol Drugs, and Trafic Safefy-l-86. Amsterdam: EI- sevier, Sept. 9-12,1986. pp. 305-308. Zylman, R. Race and social status discrimination and police action in alcohol-affected collisions. Journal of Safety Resewch 4:75-W, 1972. 234 Treatment Rehabilitation Countermeasures for Drinking Drivers Kathryn Stewart .\ Pacijic Institute for Research and Evaluation Hennq Virginia Vernon S. Ellingstad Department of Psychology, Universiy of South Dakota Vermillion, South Dakota The primary goal of rehabilitative programs for drunk driving offenders is to reduce the probability of subsequent drinking and driving. Punishment, such as licensing penalties, lines, and incarceration, is also designed to prevent subsequent drinking and driving either by making the consequences of arrest so unpleasant and costly as to discourage the offense or by eliminating the offenders' capacity to drive (by putting them in jail or by invalidating their drivers' licenses). Rehabilitation is based on one of two assumptions: That offenders drink and drive because they lack knowledge about the effects of alcohol, the potential consequences of drinking and driving, and strategies for avoiding drinking and driving; or that drinking and driving results from an abusive, addictive, or otherwise uncontrolled pattern of alcohol consumption. Therefore, offenders must receive education to help them ration- ally choose not to drink and drive or they must receive treatment so they can eliminate abusive drinking and thus stop drinking and driving. The goals of rehabilitation are certainly important-almost one-third of convicted drinking drivers have a previous offense (Sweedler and Smith 1984). Preventing some part of this recidivism is desirable. From the perspective of individual offenders, gaining information and skills and receiving treatment to allow them to avoid drinking and driving can save them from additional expense, humiliation, inconvenience, and potential tragedy. Rehabilitation can also have positive effects on other areas of the offenders' lives if abusive drinking is reduced. It is important to emphasize that even programs that are extremely effective in reducing recidivism cannot be expected to have major effects on traffic safety. Reed (1981) estimated that even if all persons arrested for drunk driving were prevented from ever combining drinking and driving again, fatal crashes would decrease by only 3 percent. Other efforts aimed at prevention or general deterrence such as well-publicized enforcement crackdowns have the potential to save many more lives. Characteristics of Rehabilitation Countermeasures There is wide variation in what are referred to as rehabilitative programs. Programs can vary in length, format, content, and structure. Programs may be quite brief (8 to 10 TREATMENT 235 hours) or more lengthy (50 or 100 hours). They maybe presented in a concentrated form over a few days or stretched over several weeks or months or even years. The format may be didactic with offenders sitting through a series of lectures, or the program may include more active participation by the offender. Some programs include group or individual counseling. Some programs involve spouses or other people close to the offender. The content of programs also varies greatly. Some programs' focus on information about the effects of alcohol, the law, potential consequences of drinking and driving, and strategies for avoiding drinking and driving (including both strategies for decreasing drinking and strategies for avoiding driving while intoxicated, such as appointing desig- nated drivers). Other programs focus more heavily on helping offenders to identify abusive or addictive drinking patterns and providing (or persuading participants to seek) alcoholism treatment. Programs also exist that emphasize development of the right hemisphere of the brain as a way of reducing problem drinking or that teach assertiveness skills in hopes that these skills will help participants avoid drinking and driving. Rehabilitative programs also vary considerably in the ways they are used. For ex- ample, in some States, offenders go through an assessment process to determine the nature and severity of their alcohol problems and are assigned to one of a number of rehabilitative programs based on the outcome of the assessment. In other States, offenders are assigned to programs based on other criteria such as blood alcohol concentration at the time of arrest or the number of previous alcohol-related offenses. The manner in which compliance with rehabilitative orders is enforced also varies, as well as other penalties that are applied along with rehabilitation. States vary in the way these programs are administered. Some programs are delivered by State agencies while others are carried out under contract with a wide variety of private agencies (everything from alcoholism treatment facilities to driving schools). The amount of control that the State exerts in determining program content, format, and standards also varies. States monitor program quality and adherence to standards in varying degrees as well. Issues in Rehabilitation Program Evaluation Assessment of the effectiveness and value of rehabilitation programs is at best a complicated endeavor, and a number of important issues must be considered in evaluat- ing such programs and in considering the results of evaluation studies. The criterion issue, or the explicit definition of program success, is a pervasive problem for any complex applied program, but is particularly difficult for rehabilitation programs oper- ated within a larger traffic safety context. In this context an alcohol treatment program would be successful if rehabilitation reduced the probability of subsequent involvement in alcohol-related crashes, or at least, if the frequency of the behavior assumed to lead to such crash involvement (i.e., drunk driving) was diminished. Measures such as alcohol-related crash involvement or driving while intoxicated (DWI) or driving under the influence (DUI) arrests and convictions are frequently chosen as criteria of success. But these measures pose at least two important meth- odological problems. First, and perhaps most importantly, being arrested for DUI or even being involved in an alcohol-related crash should probably be considered as only incidental to the drinking problems toward which many treatment programs are directed; these measures are certainly not comprehensive indicators of the intended effects of treatment. Second, despite the fact that alcohol-related crashes and drunk driving arrests occur frequently enough to justify countermeasures, the probability of these recidivist events is so low that statistical comparisons between treatment and no-treatment groups are usually not sensitive to treatment effects (the comparisons 236 BACKGROUND PAPERS usually have low statistical power). That is, the sample size must be very large or the group differences in recidivism very substantial for these differences to be empirically identified. To address these methodological problems, some evaluations of rehabilitation programs have used other measures of success such as self-reports of drinking behavior, indices of personal adjustment, and other indicators tied more closely to the expectations of the treatment programs. Such measures are not without their methodological shortcomings, including their frequent reliance on unsubstantiated self-reports. Another methodological problem, which has constrained assessments of treatment effectiveness at least as much as measurement shortcomings, concerns the adequacy of the experimental or quasi-experimental designs for contrasting treatment against no- treatment effects. In an ideal case, treatment evaluations would be conducted under carefully controlled experimental conditions, with individuals randomly assigned to treatment and no-treatment conditions, and the posttreatment performance of treat- ment and control groups compared. These conditions have not been uniformly available in rehabilitation evaluation studies, and many of the results reported in the literature represent less than rigorous experiments. Evaluation Results Keeping in mind the variability in the nature of rehabilitative programs and the difficulties in accomplishing a full and fair assessment of their effectiveness, we may proceed to a discussion of the results of evaluations of rehabilitation programs con- ducted during two distinctly different periods. The Alcohol Safety Action Projects (ASAP) of the 1970s introduced rehabilitation modalities as part of an integrated set of alcohol/traffic safety countermeasures. The 1980s brought a number of locally tailored programs, including a program based on skills-building tested in California, the Week- end Intervention Program originated at Wright State University in Ohio, and programs to provide court-mandated alcoholism treatment. The ASAP Era In June of 1970, the National Highway Safety Bureau (later to become the National Highway Traffic Safety Administration- NHTSA) of the new U.S. Department of Transportation introduced nine traffic safety countermeasure demonstration projects, which came to be known as Alcohol Safety Action Projects or ASAPs. Twenty additional ASAPs were funded iu 1971, and a final six projects were initiated during 1972. Each ASAP was designed to operate as a local drinking/driving control system (Joscelyn and Jones, 1971) which coordinated the efforts of traditional traffic safety and driver control agencies such as traffic courts, police departments, motor vehicle departments, and community health resources. Some of the ASAPs operated in single metropolitan areas, others covered large city/county regions, and still others operated as statewide projects. The NHTSA intent in funding these local projects was to provide for a demonstration (or rather, 35 replications of a demonstration) of the feasibility of an integrated systems approach to the alcohol traffic safety problem. The goal of each project was to reduce alcohol-related motor vehicle crashes by reducing the number of persons who drive while intoxicated or impaired. Rehabilitation modalities shared this project goal with law enforcement agencies, judicial systems, and public information and education com- ponents of the ASAPs. The conceptual model that prescribed the general role of rehabilitation in the ASAPs is shown in figure 1. ASAP rehabilitation countermeasures were conceived of as a bridge between the TREATMENT COURTS o CONVICTION . SENTENCING ASAP "TRANSITION" and/or COUNSELING CHEMOTHERAPY I 1 I r&sir- COMMUNITY HEALTH RESOURCES o DETOXIFICATION . IN-PATIENT o INOIV. THERAPY Fhe Jail PREiENTENCE ALCOHOL GROUP __+ SAFETY REFERRAL Revocation INVESTIGATION SCHOOL THERAPY, o ?? Suspensh . Probation I 4 4 - o MENTAL HEALTH I I I CLINICS I 0 ETC.. PROBATION Figure 1. ASAP rehabilitation: A court-referral system traffic court systems that adjudicated drunk driving offenses and various community heahh and mental health resources that provided alcohol treatment. As demonstration projects, the ASAPs were expected to provide rigorous assessments of all counter- measures employed by the projects, including rehabilitation. Each project included an evaluation function to accomplish this purpose. ASAP Rehabilitation Countermeasures ASAP rehabilitation systems were, in each of the 35 projects, designed to supplement the driver control functions of the police, courts, and licensing agencies. A fundamental assumption of ASAP rehabilitation countermeasure programs was that a significant proportion of individuals arrested and convicted of drunk driving offenses were "prob- lem drinkers" whose control over their drinking behavior (and thus drinking/driving behavior) was Iimited. This assumption created a systems requirement to perform at least a minimal diagnosis to discriminate "problem" from"nonproblem" drinkers among the ASAP? drunk driver clientele, and presentence investigations represented a primary liaison between the traffic courts and each project's rehabilitation countermeasures program. A substantial number of rehabilitation programs were conducted by the projects, or received referrals from the ASAPs. Thirty-two of the thirty-five projects used an "alcohol safety school" as a rehabilitation modality. Most of these schools were conducted by the ASAPs themselves. Some projects used the alcohol safety school as a re-education/ rehabilitation modality for nonproblem drinkers, some as a treatment alternative for problem drinkers, and still others as a rehabilitation countermeasure for both problem and nonproblem drinkers. The schools were short-term (2-6 sessions), educationally oriented programs designed to handle a substantial number of drunk driver referrals. The school was frequently the sole rehabilitation assignment for nonproblem drinkers while for problem drinkers, schools were often used in conjunction with other treatment alternatives. 237 238 BACKGROUND PAPERS In 10 ASAPs, special group therapy programs were developed and conducted by the projects themselves. Generally, these programs used weekly or biweekly sessions of an hour or two in length that extended over a period of a month to 6 weeks. The primary source of alcohol rehabilitation services across the ASAP sites was, however, the existing rehabilitation system of the community, and most ASAP treatments were provided by community treatment agencies. Outpatient treatment services provided by these agen- cies included both group therapy and individual counseling. Some projects established cooperative arrangements with local Alcoholics Anonymous chapters and utilized AA as a referral resource. Limited use was made, across projects, of inpatient treatment referrals. A few projects used chemotherapy (primarily disulfiiam) as a treatment modality, usually in combination with some kind of group or individual therapy. Analyses of treatment effectiveness were conducted at the individual project level and also at the overall program level (Ellingstad and Springer 1976). The general approach taken to the evaluation of rehabilitation effectiveness at both levelsinvolved comparing the performance of individuals who had been exposed to ASAP-sponsored or -coor- dinated treatment with the performance of individuals who were not referred to rehabilitation. Unfortunately, with but two exceptions (Nassau County, New York and Phoenix, Arizona), the demonstration projects were not structured to provide robust experimental tests of rehabilitation with random assignment of clients to treatment and no-treatment conditions. The "no-treatment controls" at most ASAP sites were of individuals excluded from treatment because there was no room in the treatment programs when they entered the system, they refused to participate, or for one reason or another they were judged to be unsuitable for entry into treatment. Project-level analyses of treatment program effectiveness were reported to NHTSA annually in Analysis of Akohol Rehabilitation Efforts, a mandated analytic study prepared by the local project evaluation component. A number of summaries of these studies have been reported (see, for example, Ellingstad, 1976; Spiegel and Struckman- Johnson 1978). Evaluations included overall assessments of rehabilitation system effec- tiveness (all treatment modalities combined) as well as assessments of individual rehabilitation countermeasures. Criteria on which these analyses were based included crash recidivism, arrest recidivism, and in some cases, other measures obtained by testing or interviewing program participants. The most common criterion of program success was alcohol-related arrest recidivism. While isolated reports of treatment impact on traffic safety criteria (crash or arrest recidivism) came from some of the 35 projects, an inverse relationship was also apparent between the methodological adequacy of the analytic study and its likelihood of reporting significant results. Despite the generally pessimistic results of these analyses when critically evaluated, some indications of success were present. Process-oriented studies of alcohol safety schools almost universally demonstrated them to be capable of altering levels of knowledge and attitude, even though the effects of these treatment programs on recidivism was equivocal. Most analyses of the more intensive treatment programs showed no clear evidence of treatment effectiveness. A notable exception concerned the Disulliram Clinic operated by the Los Angeles ASAP. This program demonstrated a statistically significant reduction in recidivism associated with disulfiram treatment in a well-controlled and statistically sound analysis. Program-level analyses of ASAP rehabilitation countermeasure effectiveness were also performed by pooling data (mostly arrest recidivism data) submitted by the in- dividual projects (Ellingstad and Springer 1976). Comparisons of survival rates (propor- tions of clients avoiding rearrest) over a 3-year followup period for nonproblem drinkers (as determined in presentence investigations) showed the pooled across-project trcat- ment group to have outperformed the pooled no-treatment group. This program-level result was at least suggestive that treatment may have had some of its intended effect on nonproblem drinkers. A similar comparison for problem drinkers did not show sig- TREATMENT 239 nificant differences in survival rates between individuals who had been referred to ASAP treatments and those who had not received treatment. It seems fair to conclude that, on balance, the results produced by the ASAPs in identifying effective alcohol rehabilitation countermeasures were disappointing. Sig- nificant methodological problems constrained both project- and program-level analyses of rehabilitation system effectiveness and prevented clear tests of treatment effect. The absence of adequate experimental controls seemed to be the principalissue. The Short-Term Rehabilitation Study It became apparent after the first few years of ASAP operations that the methodologi- cal problems alluded to previously were likely to seriously handicap assessments of rehabilitation effectiveness within this program. Because of this concern and because of project-level interest in a relatively new alcohol treatment program called Power Motiva- tion Training (PMT), a series of important changes in the implementation and evaluation i of ASAP rehabilitation countermeasures were introduced beginning in 1973 (Ellingstad 1976b). PMT, developed by McBer and Company, alcohol treatment researchers, was based on a distinct set of theoretical principles and consisted of a well-defined and carefully described set of therapeutic procedures. Moreover, PMT was a short-term modality that did not depend on highly trained professional therapists and could be readily implemented within the ASAP rehabilitation systems (Cutter, et al. 1975). The PMT program was formally begun in eight sites in early 1975. McBer and Company, under contract with NHTSA, trained therapists at the participating sites (Boyatzis 1976). In addition, an evaluation function was created to develop a system to collect, monitor, and process data from the PMT sites and to develop instruments to provide measures of relevant indices of treatment effectiveness. The name was changed to the Short-Term Rehabilitation (STR) Study to reflect the fact that several treatment alternatives in addition to PMT were to be included in the experimental designs at some of the sites, and that an additional three ASAPs that did not use PMT but did employ random assignment procedures and no-treatment control groups were to be added to the study. Each site in the STR study used its presentence investigation procedures to identify a pool of mid-range problem drinkers considered to be the most appropriate clients for PMT and related treatment programs (both social or nonproblem drinkers and alco- holics were excluded). From this pool, clients were randomly assigned to either treat- ment or control conditions. A comprehensive data collection procedure involving extensive interviews, questionnaires, and record checks was conducted at the time of assignment, as well as at 6-, 12-, and l&month followup contacts. A total of 3,663 clients were randomly assigned to treatment and no-treatment conditions at the 11 sites, with 2,462 clients exposed to various short-term rehabilitation modalities and 1,201 clients assigned to no-treatment or "minimum exposure" control groups (some sites required a minimal treatment such as the distribution of literature about alcohol and driving instead of a true no-treatment control condition - this affected only four sites), The extensive data collection employed within the STR study provided for a large battery of outcome criteria including: traffic safety outcome measures such as crash and arrest recidivism; direct indices of drinking behavior such as duration of abstinence, average level of alcohol consumption, and incidence of abusive drinking; lie status measures such as current drinking problems, physical health problems, and employ- ment/economic stability; and measures of personality characteristics (Ellingstad and Struckman-Johnson 1978). Detailed analyses were conducted for each set of dependent variables within the experimental designs of each of the 11 STR sites individually (Struckman-Johnson and Ellingstad 197&z). No compelling evidence of treatment effectiveness was found in any 240 of these analyses and, in fact, statistically significant negative effects were observed in two or three instances. Program-level analyses were also performed on data pooled from the 11 sites (Struck- man-Johnson and Ellingstad 197%). A large number of statistical comparisons revealed some evidence of treatment effectiveness for alcohol safety schools (employed as a treatment alternative by four of the STR sites), and some evidence suggested a negative treatment effect for PMT as a single modality treatment assignment. The CDUI Project Despite the fact that the STR study had involved thousands of drunk driver clients and had been able to achieve the methodological requirements (random assignment and control groups) of a true experiment, the fact that the study encompassed 11 very different jurisdictions presented organizational difficulties that may have prevented as powerful a test of rehabilitation countermeasures in the traffic safety context as might be desired. In late 1976, a massive, single site experimental project called the Com- prehensive Driving Under the Influence of Alcohol Offender Treatment Demonstration (CDUI) Project was initiated in Sacramento, California. The CDUI Project operated from September 1977 through January 1981 in Sacramento County, receiving its referrals from the Sacramento County Municipal Court. The project employed two separate experimental designs, one for first-offense drunk drivers, the other for drivers convicted of multiple DUI offenses. The lirst- offender design provided random assignment of 4,639 individuals convicted of DUI to one of three treatment alternatives: (1) an in-class education program consisting of four classroom sessions of 2 l/2 hours each over a 4-week period, using a standard alcohol education program patterned after others in use around the United States; (2) a home study program consisting of an organized set of reading materials designed as a self-study package, which was presented to the clients in a l-hour orientation session; and (3) a control group who received no treatment. All clients were placed on 2-year informal probation and received a reduced line as an incentive to participate. In addition to the treatment assignments, one-half of each treatment group was randomly assigned to receive quarterly monitoring letters to remind them of their probation status and to encourage them to drive soberly. Half the clients were also randomly assigned to receive followup interviews designed primarily to collect life activities data for treatment out- come analyses. Both the in-class and home study education programs were shown to produce significant reductions in DUI recidivism relative to the no-treatment control group. Neither program, however, had significant impact on crash involvement or on the variety of life status measures collected at followup intervals 10 and 20 months subsequent to treatment entry (Reis 1982). The principal CDUI multiple offender design involved a postconviction presentence (PCPS) procedure under which a guilty plea to DUI was accepted prior to referral, but final disposition and sentencing was postponed 13 months to permit participation in the assigned treatment condition. Those clients who successfully completed the assigned treatment then had the charge reduced to reckless driving, thereby avoiding the man- datory licensing action that would have resulted from the DUI conviction. The 1,103 clients available to the PCPS multiple offender design were randomly assigned to the following conditions: (1) Control (341): No educational or rehabilitative treatment, no educational counseling, no chemotherapy, and no biweekly contacts; (2) Biweekly contacts only (326): Twenty-six 15-minute individual interviews with a probation officer every other week for 1 year; (3) Skills workshop (110): A group educational counseling approach developed for the CDUI project consisting of 34 2-hour group counseling sessions with the first 16 sessions meeting weekly and the final 18 sessions every other TREATMENT 241 week for the remainder of the year of treatment; (4) Skills workshop and chemotherapy (109): Three supervised administrations of disulliram per week for the first 6 months of the assignment were combined with the skills workshop group therapy program; (5) Educational eclectic therapy (109): Counselors conducting eclectic groups had complete freedom to organize group therapy sessions according to their preferred style. The first four 2 l/2-hour sessions were alcohol education classes identical to the first-offender classes. They were followed by 28 weekly 2-hour group therapy sessions; and (6) Educational eclectic therapy and chemotherapy (108): Three supervised admin- istrations of disulliram per week for the first 6 months of the assignment were combined with the educational eclectic therapy program. Followup of multiple offender clients over a 20-month period showed both counseling programs to produce significant reductions in DUI recidivism in comparison to the no-treatment control group. Adding chemotherapy to counseling programs did not improve the recidivism performance of these rehabilitation programs. Chemotherapy was shown to be effective in reducing levels of alcohol consumption for up to 14 months beyond the termination of disultiram treatment in clients who completed a counseling : program. None of the multiple offender treatments affected crash involvement (Reis 1982). In contrast to the earlier ASAP and STR experiences, the CDUI results provided considerably more encouragement with respect to the efficacy of alcohol rehabilitation programs operated within the context of a traffic safety system. Post-ASAP Rehabilitation Programs Despite the mixed and disappointing results of evaluations of the various rehabilita- tive programs carried out in the ASAP era, the concept of rehabilitation still generated interest. Additional program models have been tested in recent years. Skills Building While some ASAP programs were able to show some reduction in recidivism and subsequent crashes, the magnitude of the reductions was disappointing. However, many of the program models evaluated seemed rather weak and not well grounded in theoreti- cal or empirical knowledge about alcohol abuse or behavior change (Kunkel 1983). Moreover, little attention seemed to have been paid to the quality of implementation. Programs as they occur in actual practice often bear little resemblance to programs as they appear on paper (French and Kaufman 1981). It seemed possible that the dii- appointing outcomes might be due in part to weak program models or poor implementation. To give rehabilitation countermeasures the best chance of showing effectiveness, the State of California sponsored an evaluation effort that included an extensive program model development effort and careful attention to quality of implementation (Stewart et al. 1987). The model program that was developed resembled traditional first-offender programs implemented in the ASAPs in many respects. It included information on the effects of alcohol, on drinking and driving laws, on symptoms of alcohol addiction, etc. It also had several distinctive features, including a focus on the development of skills to enable the offenders to separate drinking from driving. This aspect of the program was based on Bandura's Self-Efficacy model (Bandura 1977) in which participants develop strategies for dealing with a series of risky situations of increasing difficulty. Over the course of the program, offenders were helped to develop, rehearse, and practice realistic strategies to avoid drinking and driving. In addition, rather than 242 BACKGROUND PAPERS attempting to deal with serious problems such as alcoholism within the constraints of the program, a strong emphasis was placed on assessing the offenders' problems and referring them to other helping resources in the community. The program was structured to include a great deal of enforced participation so that the offenders were compelled to be actively engaged in the program. Staff at the program sites received extensive training and ongoing technical assistance and monitoring to ensure that the quality of implementation would be high. The program had two segments. The first segment was a6-week (15 hour) educational program and the second was a 7-week (11 hour) counseling program. Offenders were randomly assigned to the education-only or the education-plus-counseling segments to determine whether programs of differing lengths and intensities would have different effects. The two versions of the model program were compared to two existing California programs and to a control group (which participated in community service projects but received no formal program content). Participants were randomly assigned to these program conditions. The driver records of the participants in the four groups were followed over 5 to 11 months to determine recidivism rates. In addition, a sample of participants was interviewed before program entry and again 6 months later to include more sensitive indicators of program success by broadening outcome measures to include drinking behavior, symptoms of alcohol dependency and undetected drinking and driving. Close associates of a sample of the respondents were also interviewed to validate self-reports of drinking and drinking/driving. Even given this carefully designed, well-implemented program, no differences could be detected between the self-reported drinking behavior and drinking/driving behavior of first offenders randomly assigned to the four program conditions, including the control condition. Though some decreases in drinking and in frequency of drinking and driving were reported, these decreases were reported equally by respondents in all program groups. Thus, no evidence was found of the superiority of any program type over any other, including the control group, which received no program at all. In fact, the observed changes could be due to the natural reaction to any intervention or could be the result of a statistical artifact (regression to the mean). The followup time for recidivism was quite short, but no significant differences in recidivism were detected. Thus, the modest effects demonstrated by other evaluations of rehabilitative programs were not improved upon by altering program content and format. Within the range of standard programs, no program type appears to have any outcome advantage over any other, and the potential traffic safety effects of any program are very small indeed. The Weekend Intervention Program The Weekend Intervention Program (WIP) (Siegal and Moore 1985) was based on the assumption that it is unrealistic to expect a long-term pattern of problem drinking to be altered after a short period of treatment or education. The intervention approach does not &y to treat the problem drinker. Instead, it is designed to identify whether a problem exists, to assess its extent and severity, and to refer offenders in need of treatment to appropriate facilities. The offenders follow through on the referral on their own, sometimes with the encouragement or order of the court. As the title implies, the program takes place over the weekend. Educational and counseling activities takes place in a medical school, and clients are housed in a nearby motel in the evenings under police supervision. The goals of the program are to carry out an assessment or diagnosis, to break down denial in those participants who have a serious problem, and to prepare offenders to accept treatment if needed. TREATMENT 243 An evaluation of WIP indicated that the program was effective in lowering the recidivism rate as compared to nonequivalent comparison groups who received a suspended sentence or who were sentenced to jail. The effect was strongest for repeat offenders. During a l- to 2-year followup of repeat offenders, 21.8 percent of WIP participants recidivated compared to 26.8 percent of jailed offenders and 30.4 percent of offenders with suspended sentences. For first-time offenders, the recidivism rate for WIP participants was 9.2 percent while the rate for all other first offenders was 12.7 percent (Siegall987). The program was acceptable to the community in Ohio. It was easily understood by the public and was acceptable to law enforcement and judicial personnel. It was less expensive than traditional incarceration and, apparently, had a more positive effect on subsequent drinking and driving (Siegal and Moore 1985). Clearly, however, the effects on recidivism were modest. The program may serve a function in the community by providing an acceptable form of punishment that may also be advantageous to the offender, but its effects on traffic safety were negligible. Compulsory `keatment Models Some more intensive program models have been tried to deal with habitual offenders or those with severe alcohol problems. One intervention is court mandated alcoholism treatment for offenders who are addicted. Questions have been raised about the appro- priateness or efficacy of compulsory treatment. Some research indicates that the out- come of treatment for patients receiving treatment as part of a suspended sentence for drunk driving compare favorably with improvements in alcoholics treated voluntarily (Ben-Arie et al. 1983). A 7- to g-year followup was carried out with 50 offenders who had been diagnosed alcoholic (most of whom were multiple offenders) and who received compulsory treatment. At the time of followup, 14 of the offenders had been convicted of further driving offenses. Thus, the treatment cannot be considered highly successful in terms of reducing recidivism. Forty percent of the offenders were either sober or generally sober, indicating that the treatment may have been useful in overcoming alcoholism (Ben-Arie, et al. 1986). A second model that has been implemented in several locales is the combination of incarceration and treatment. Special facilities are set up to incarcerate drunk driving offenders, usually repeat offenders. During their incarceration, inmates participate in a highly structured education and treatment program, usually including detoxification (if necessary), educational sessions, and group and individual counseling. A program of this type carried out in Massachusetts reported a recidivism rate of 6 percent compared to a statewide rate of 25 percent and a 19-percent rate for low security institutions similar to the program's (LeClair 1987). In the Netherlands, an educational program for incarcerated drunk drivers used volunteers from various areas of the drunk driving system to teach sessions on the nature and impact of alcohol abuse and provide information on community alcoholism treat- ment services. Positive effects were reported on knowledge, attitudes, and driving behavior when it was evaluated (Bovens 1987). These studies provide some preliminary support for compulsory treatment programs, either as mandated by the courts or as a component of incarceration. Here again, however, while some individual offenders may be helped to overcome addictive drinking or to avoid later drinking and driving, the impact on traffic safety is quite smalL Conclusions Because of the nature of the alcohol-related crash problem, rehabilitative approaches can only have a very small effect on traffic safety, even if maximally effective. A wide 244 BACKGROUND PAPERS variety of rehabilitative programs based on a variety of theoretical models, and delivered in a variety of settings have never been able to'achieve more than modest effects on recidivism. Although many evaluations of these programs suffered from methodological weaknesses, the conclusion seems inescapable that to achieve improvements in traffic safety, other strategies must be employed. Rehabilitative programs may serve other purposes, such as providing an additional appropriate and acceptable form of punishment to offenders, enforcing a general societal message that drlnkiig and driving is unacceptable behavior, and providing a mechanism for intervention into the driig problems of individuals. It is important to keep in mind, however, that these possible benefits must be weighed against the costs of the programs. Rehabilitative programs are not free.In many States, a substantial industry (often a for-profit industry) supportiig hundreds or even thousands of employees has grown up to provide these mandatory programs. Usually, the direct cost of the programs is borne primarily by fees paid by offenders. These fees can be viewed as just another part of the fines and other monetary penalties offenders are required to pay. From this perspective, the effectiveness of the programs may not be an issue. However, the extent to which these fees are purchasing services valuable to the individual or to society may be called into question. While possible benefits to individuals have been discussed, these beneficial effects have not been thoroughly evaluated. A number of evaluations report attitude changes in offenders (Foon 1988). However, evaluations that measured drinking levels or improve- ments in other life areas, found very few effects (Stewart et al. 1987, Reis 1982). The possibility that some programs may actually have harmful effects cannot be ignored. In some States, rehabilitative programs are not closely monitored, and the appropriateness of program content and the qualifications of staff may not be carefully evaluated. Programs must deal with content that is highly charged emotionally. There is risk of harmful effects if these areas are not handled with skill and caution. It is important to compare the effectiveness of rehabilitative programs to other sanctions - specifically license penalties. A 1984 study (Sadler and Perrine) compared the impact of alcohol treatment programs to that of license suspensions on subsequent crash rates and drunk driving recidivism. The study found that license suspensions have a significant positive impact on traffic safety, more so than did the treatment programs (although treatment programs had a greater impact on alcohol-related crashes and arrests). Hagen et al. (1980) found that license suspensions and revocations produced significant reductions in subsequent convictions and crash rates for multiple offenders. In a review of a number of evaluations of the effectiveness of license actions, Peck et al. (1985) drew this conclusion: "... there is no question that license suspensions have a significant effect in reducing the accident and drunk driving frequency of convicted DUI offenders." Recommendations Given the weak traffic safety benefits of rehabilitation countermeasures, it is very important that the continuation of rehabilitative programs not be allowed to deflect attention or resources away from drinking/driving countermeasures that have more powerful effects. Strategies that primarily attempt to bring about specific deterrence will necessarily be limited in their ability to improve traffic safety. Even within that limited realm, licensing penalties have been shown to be more effective in reducing recidivism than rehabilitative programs, whatever their form. In many States, participation in a rehabilitative program is offered as a substitute for license suspension or revocation. TREATMENT Such substitution is clearly counterproductive from the standpoint of traffic safety. If the positive effects of rehabilitative programs are sufficient to justify their continued exist- ence, these programs must be used in addition lo rather than instead of license penalties. A larger question can be raised about all drinking/driving strategies that focus on individual behavior with little consideration of the environment that shapes that behavior (Wallack 1984). It is easier for society to blame the problem of, drinking and driving on a defmed group of individuals rather than on money-making produc'ts, industries, and systems that support drinking and driving and amplify its destructive potential (Vingilis 1987). In our zeal to deal with the population of identified drinking drivers, we should not lose sight of social forces such as the political and economic climate, cultural patterns, and values and norms that all combine to determine how alcohol is used and what consequences that use will have for our health and Gfety. 245 REFERENCES Bandura, A. Self-efficacy: Toward a unifying theory of behavior change. Psychological Review S&191-215, 1977. Ben-Arie, 0.; George, G.C.W.; and Hirschowitz, J. Compulsory treatment of 50alcoholic drunken drivers. South Afican Medical Journal 63~241-243, 1983. Ben-Arie, 0.; Swartz; and George, G.C.W. The compulsory treatment ofalcoholic drunken drivers referred by the courts: A 7 to 9 years outcome study. The Journal of Law and Psychiatry 1986. Bovens, R. Alcohol program: An educational program for drunken drivers in prison. In: Brand- Koolen, M.J.M., ed. Studies on the Dutch Prison System. Berkeley, CA: Kugler Publica- tions, 1987. p.151-157. Boyatzis, R.E. "Implementation of Power Motivation Training as a Rehabilitation Counter- measure for DWIs. Technical report No. DOT-HSSOl- 834, McBer and Company, Boston, February 1976. Cutter, H.S.; McClelland, D.C.; Boyatzis, R.E.; and Blancy, D.D. "The Effectiveness of Power Motivation Training for Rehabilitating Alcoholics." Technical report, McBer and Com- pany, Boston, 1975. Ellingstad, V.S. "1975 Interim Analyses of ASAP Rehabilitation Efforts, Interim Report." Human Factors Laboratory, University of South Dakota, March, 1976a. Ellingstad, V.S. "Program Level Evaluation of ASAP Diagnosis, Referral and Rehabilitation Efforts. Vol. IV: Development of the Short Term Rehabilitation (STR) Study." Technical report DOT-HS-191-3-759- F4, Human Factors Laboratory, University of South Dakota, June, 1976b. Ellingstad, V.S., and Springer, T.J. "Program Level Evaluation of Diagnosis, Referral and Rehabilitation Efforts. Vol. III: Evaluation of Rehabilitation Effectiveness." Technical report DOT-HS-191-3- 759-F3, Human Factors Laboratory, University of South Dakota, June, 1976. Ellingstad, V.S., and Struckman-Johnson, D.L. "The Short Term Rehabilitation Study. Vol.11: Development and Description of Measurement Battery." Report HFL-78-12, Human Factors Laboratory, University of South Dakota, November, 1978. French, J.F., and Kaufman, N. Handbook for Prevention Evahtation. Rockville, MD: National Institute on Drug Abuse, 1981. Foon, AE. The effectiveness of drinking-driving treatment programs: A critical review. Inrerna- rional Journal of the Addictions 23(Z): 15 l-174, 1988. Hagen, R.E.; McConnell, EJ.; and Williams, R.L. Abstract of Suspension and Revocation Eflects on rhe DUI Offender. Sacramento, CA California Department of Motor Vehicles, 1980. Joscelyn, J.D., and Jones, R.K. A Systems Analysis of rhe Trafic Law System: Summary Volume. NHTSA Report No. DOT-HS-800-640, Institute for Research in Public Safety, Indiana State University, October, 1971. Kunkel, E. Driver improvement courses for drinking drivers reconsidered. AccidentAnalysis and Prevention 1.5(6):429, 1983. LeClair, D.P. Use of Prison Conjinemenr for the Treatment ofMuhiple Dnrnk Driver wenders: An Evalumion of the Longwood Treatment Center, Bcutive Summq. Massachusetts 246 BACKGROUND PAPERS Department of Corrections, Division of Research. Washington, DC: National Institute of Justice, U.S. Department of Justice, 1987. Peck, R.C.; Sadler, D.D.; and Perrine, M.W. The comparative effectiveness of alcohol rehabilita- tion and licensing control actions for drunk driving offenders: A review of the literature. Akoho& Drugs and Lkiving l(4): 1539,198s Reed, D.S. Reducing the costs of drinking and driving. In: Moore, M.H., and Get-stein, D.R., eds. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: National Academy Press, 1981. Reis, R. E., Jr. Ike Trafic Safety Eflectiveness of Education Programs for First mense Drunk Drivers. Final report, Comprehensive Driving Under the Influence of Alcohol Offender Treatment Demonstration Program, County of. Sacramento Health Department. Sacramento, CA California Department of Motor Vehicles, 1982. Sadler, D.D., and Perrine, M.L.W., The Long-Tenn Traffic Safety Impact of a PilotAlcoholAbuse Treatment as an Alfemative to License Szqension. Sacramento, CA: California Depart- ment of Motor Vehicles, 1984. Siegal, HA Intervention: A Successful Technique for Repeat offenders: Reyrt of an Evaluation, Alcohol LIrugs and Trafic Safety. Eighth International Conference on Alcohol, Drugs and Traffic Safety. Amsterdam: Elsevier, 1987. Siegal, HA., and Moore, D. "Rehabilitating the Drinking or Drug Impaired Driver: Its Evolution From Education to Intervention." Paper presented at the 34th International Congress on Alcoholism and Drug Dependence, Calgary, Alberta, 1985. Spiegel, D.K., and Struckman-Johnson, D.L. "1977 Interim Assessments of Drinker Diagnosis, Referral and Rehabilitation." Report HFL-78- 5, Human Factors Laboratory, University of South Dakota, July, 1978. Stewart, K; Epstein, L.; Gruenewald, P.; Laurence, S.; and Roth, T. "The California First DUI Offender Evaluation Project." Final Report, Pacific Institute for Research and Evaluation, Walnut Creek, California, 1987. Struckman-Johnson, D.L., and Ellingstad, V.S. "The Short Term Rehabilitation Study. Vol.111: Site Specific Analyses of Effectiveness." Report HFL-78-11, Human Factors Laboratory, University of South Dakota, November 197&z. Struckman-Johnson, D.L., and Ellingstad, V.S. "The Short Term Rehabilitation Study. Vol. IV Program Level Analyses of Effectiveness." Report HFL-78-9, Human Factors Laboratory, University of South Dakota, November 19786. Sweedler, B.M., and Smith, L. "The Repeat Offender Drunk Driver: Where Has the System Failed."Paper presented at the International Workshopon Punishment and/orTreatment for Driving Under the Influence of Alcohol and Other Drugs, Stockholm, Sweden, 1984. Vingilis, E. The six myths of drinking-driving prevention.hlealth Education Research 2(2):145-149, 1987. Wallack, L. Practical issues, ethical concerns and future directions in the prevention of alcohol- related problems. Journal of Primary Prevention 4: 199-224, 1984. Citizen Advocacy 247 Independent Citizen Advocacy: The Past and The, Prospects John D. McCarthy, Ph.D. and Debra S. Harvey, MA ' Depariment of Sociology and Life Cycle Institute The Catholic University ofAmerica During the last decade, a widespread movement of citizen advocate groups has emerged whose members, many of whom are victims of drunk drivers, work to reduce the level and consequences of drunk driving. Their efforts are widely seen by a variety of observers as having had some success. For instance, Senator John Danforth in recent congressional hearings said of MADD, This organization has made the public realize that drunk driving is not a victimless crime. This change in public attitude has made it possible for those of us in Congress and in State legislatures to pass stronger drunk driving laws. (1988) Franklin Zimring (1988), a consistently skeptical social observer, discussing these local advocacy groups says, ",. the mobilization by the groups] of public opinion has been partially responsible for the increased prominence of drunk driving as a public policy issue" (p. 374), and goes on to say, "My guess is that citizen action groups are a more important explanation [than others] of the passage of legislation in the 1980s" (p. 380). Finally, Mark Wolfson (1988) concluded in his systematic evaluation of the effects of local advocacy that the efforts of these groups positively affected State legislative initiatives and "may have [had] some influence on fatalities" (p. 9). Any organized effort to encourage the continued growth and vitality of this inde- pendent citizens' movement must depend upon an adequate description and under- standing of its emergence, the community support and attention it receives, its typical structural forms, the personal lives of its activists, and the nature and extent of its organized activities. It must depend, too, upon an understanding of the typical difficulties that such movements encounter in maintaining continued high levels of citizen advocacy. The Development of the Local Movement * The citizens' movement against drunk driving consists of a number of different national and local organizations. At the national level are two umbrella groups, Mothers 1 The original research described in this chapter was supported, in part, by a grant from the National Science Foundation (SEs-8419767) and continuing support from the Life Cycle Institute at The Catholic University of America. We thank Mark Wolfson for his continuing contributions to this research. 2 Most of the evidence upon which this description is based was gathered during the 1985-86 period and refers to the experience within the United States. Any changes the movement has undergone since then are, therefore, not reflected in this account. 248 BACKGROUND PAPERS Against Drunk Driving (MADD), headquartered in Hurst, Texas, and Remove Intoxi- cated Drivers-USA (RID), headquartered in Schenectady, New York. Both groups have a large number of local chapters spread across many States. In addition to MADD and RID, there are a number of regional and local citizens' groups that are not affiiated with any national umbrella group, which we call "out.liers." After an exhaustive attempt to generate a census of local groups, we estimate that 458 local groups, including MADD chapters, RID chapters, and outliers, existed in 1985. The pattern of foundings of these groups and their present distribution across the United States is described below. Remove Intoxicated Drive; (RID). The citizens' movement against drunk driving began in 1978. During that year, three local groups formed in New York State. These groups were later to become affiiated with a national umbrella group, Remove Intoxi- cated Drivers (RID), started by Doris Aiken in Schenectady, New York in 1979. In 1979, four more groups started in New York that were to affiliate with RID. In 1980, two more RID chapters formed, both in New York State. In 1981,14 RID chapters formed, including four in New York. That was the first year in which RID chapters formed outside of New York State-in Oklahoma, Mas- sachusetts, Tennessee, New Jersey, Wisconsin, Colorado, Connecticut, Pennsylvania, and Texas. The growth of RID peaked in 1982, when 18 new groups formed (see figure 1). As of 1985, RID had 70 active chapters in 23 States, although the majority were in New York (22 chapters), Illinois (6 chapters), and Tennessee (5 chapters). Mothers Against Drunk Driving (MADD). Candy Lightner and others formed a group called Mothers Against Drunk Drivers (later to become Mothers Against Drunk Driv- ing) in Sacramento, California, in 1980 after Candy's daughter was killed by a drunk driver. A second MADD chapter formed in California in 1980. In contrast to RID, MADD was quick to diversify geographically: the nine MADD chapters formed in 1981 were in California (2 chapters), Florida (2 chapters), Ohio (2 chapters), Pennsylvania, Texas, and Kentucky (1 chapter each). The growth of MADD accelerated at a breath- taking pace over the next few years (see figure 1). As of 1985, an estimated 377 MADD chapters existed, with at least one chapter in every State except Idaho and Montana. Chapters of MADD were most heavily concentrated in California (29 chapters), Plorida (25 chapters), and Texas (26 chapters). Orttliers. The pattern of founding of groups that are not affiliated with any national organization has been somewhat harder to estimate. Since they are not affiliated nation- ally, no comprehensive listing of these groups is maintained. Information could be collected directly from only 11 groups, although this is almost certainly an underestimate of their actual number. One outlier group, Concerned Citizens and Victims of Drunk Drivers, of Reno, Nevada, was one of the earliest groups, having formed in 1979. As of 1985, outliers included a number of regional coalitions, such as the Alliance Against Intoxicated Motorists (AAIM), which is concentrated in Illinois, and Rid Arizona of Intoxicated Drivers (RAID), which, as the name suggests, is limited to Arizona. In all, outliers were found in Arizona, California, Illinois, Nevada, New Hampshire, North Carolina, Oregon, Utah, and Virginia. Thus, the citizens' movement against drunk driving began with the formation of a few local groups, two of which developed first into primarily regional movements in New York State and California. However, by 1985 the movement had become truly national in scope, with an estimated 458 local groups and at least one in every State but Montana. 3 3 We have not included certain kinds of groups that devote extensive efforts to the issue of drunk driving in our analysis of advocacy groups. Such groups include locals of Students Against Driving Drunk (SADD) and bt Alcohol Consciousness Concerning the Health of University Students (BACCHUS), which are, respectively, high school and college student groups. CITIZEN ADVOCACY Figure 1: New Foundings and Cummulative Number of Groups by Year 249 0 0 0 0 0 0 0 / / / / / / / / I / / / / / mouPNPES MADD -- RD -m-m__ YEAR 1978 1979 19BO 1981 1982 1983 1984 I.985 Total wcx 2 9 88 104 93 81 377 l-u!2 3 4 2 14 1B l.2 10 7 70 Outlier 1 1 4 3 1 1 11 lbtal 3 5 4 24 110 119 104 89 458 Patterns of Local Group Founding The aggregate temporal pattern of local anti-drunk driving advocacy group founding through 1985 can be seen in figure 1. Only a few groups formed between 1978 and 1980, while a burst of foundings occurred in 1981. The peak year for new foundings was 1983, with the number of new groups founded dropping off through 1984 and 1985. While the rate of new group formation declined during these years, the total number of local groups continued to expand. The pattern of emergence of these local groups was uneven between 1978 and 1985. Many communities lacked a local advocacy group dedicated primarily to the issue of drinking and driving in 1985, while other communities saw a group form rather late in the period. Our analyses (McCarthy et al. 1988; McCarthy and Wolfson 1988) of this process indicated that neither a high rate of alcohol-related motor vehicle fatalities nor the prior existence of an Alcohol Safety Action Project, each of which might be expected to do so, predicted the formation of an advocacy group by 1985 in a local community. Community size was important because larger communities were more likely to see groups formed Bnd formed early than were smaller communities. We concluded that this founding pattern is similar to the diffusion of other kinds of innovations (see Hamblin et al. 1973). To the extent that common understandings concerning drinking and driving as a soluble problem and specific models for citizen advocacy are available, the more citizens in any community, the more likely a local group will emerge. Given that most groups are formed by a single highly motivated individual, family, or small cluster of friends, group foundings are quite unpredictable events. 250 BACKGROUND PAPERS Some observers have interpreted the decline in the rate of founding new local groups to mean that this movement has begun to lose its vitality and general community support. We interpret the pattern, however, to reflect the natural limit on new group formation. First, the majority of large American communities had a local group in 1985. Second, MADD has limited new group formation, to a single group in each county, with a few exceptions. Finally, many local groups see the commnnity they serve as broader than just the city or county where they reside. Our anafysis showed that, in 1985,55 percent of the American population lived in a county which had a local advocacy group; 67 percent lived in a county in which such a group recruited members; and 95 percent lived in a media market that included such a group (McCarthy et al. 1987). These figures strongly support our assessment that local groups had come close to saturating local communities across the United States by 1985. Patterns of Local Group Dissolution Many emergent local advocacy groups do not get off the ground, and others leave almost no trace after very short corporate lives. Our analyses were based upon groups which existed during 1985. As a result, we have no information on those groups that emerged earlier but failed to survive, although we estimate a 5- to 25-percent failure rate. Frank Weed (19%) found a failure rate of about 20 percent for local MADD chapters over the 25 months ending in June 1987. This rate is consistent with our upper limit estimate. Weed's analysis showed, similar to studies of other types of local groups (Freeman et al. 1983), that younger groups fail at much higher rates than older groups. He also found that groups with more independent local leadership and with wider and deeper local community support were more likely to survive than groups without these features. The population of local advocacy groups addressing drunk driving in any period, then, is the result of previous patterns of group founding and group dissolution. As a conse- quence, any effort to maintain a large number of active groups in this movement requires an understanding of the processes of both organizational formation and dissolution. Community Support and Public Attention Strong, consistent national governmental support is demonstrated in several ways, including the efforts of the Presidential Commission on Drunk Driving, wide congres- sional support for legislation such as encouraging States to raise the drinking age and to employ "administrative revocation," and Federal executive agency initiatives on drunken driving demonstrate consistent and strong nationa governmental support. Local group leaders report wide and deep community-level support for their efforts. The national and local media attention to the issue of drunk driving and the activities of citizen advocates has been extensive as well. National and Local Support Several notable aspects of national and local support exist. One is the flowofresonrces from Federal Agencies to local citizens' groups. Such resources can be expected to increase the local capacity to mobilize around the issue. These resources take several forms. 1. Literature designed to help organize a local citizens' group against drunk driving has been produced and widely disseminated by the National Highway Traffic Safety Administration (NHTSA) (1982, 1983). One packet of materials con- sisted of well-organized and simply presented steps to be followed and goals to CITIZEN ADVOCACY 251 be pursued in beginning a local citizens' group. Analysis of information col- lected from local groups suggested that nearly 80 percent of them received some of the literature they used in their activities in 1985 from NHTSA (McCarthy et al. 1987). 2. Financial support has been provided to local groups attempting to organize around the drunk driving issue. During the early days of the movement, NHTSA provided funds to support the development of a few local citizens' organizations (Mann 1983). Since then, NHTSA has continued to provide some level of indirect support. This is evident, for example, in NHTSA's 1984 budget which requested the appropriation of funds for Citizen Support: generating community support for comprehensive programs, thus providing a political base for increased counter- measure activity. As shown recently in New York, Maryland, West Virginia, California, and elsewhere, an organized and informed body of individuals can bring about major change in State laws and alcohol programs. (NHTSA 1984) This description of the activities of NHTSA makes clear its practice of en- couraging the growth of citizens' groups focusing on the drunk driving issue. 3. Extensive efforts were made by a number of Federal Agencies, especially the National Institute on Alcoholism and Alcohol Abuse (NIAAA) (Vejnoska 1982), to create networks of citizens concerned about drunk driving and to encourage them to organize. In addition, NHTSA, in concert with the National Safety Council and others, has continued to help organize the annual "Lifesavers" conferences that bring together local activists; Federal, State, and local officials; industry representatives; and researchers. 4. Early organizers of the anti-drunk driving movement were employed by NHTSA to generate local activity and additional technical support (Golden 1983). 5. The ongoing regional workshops for local advocacy groups sponsored by NHTSA continue one important form of national support for the movement. The national public opinion evidence also shows widespread support for the general goals of these advocates. In 1977,84 percent of the American population agreed with the statement. "There should be stricter laws on drinking and driving;" and in 1982,92 percent of the population agreed with that statement (Public Opinion 1983). In June of 1984, almost 80 percent of the American population favored a national law raising the legal drinking age to 21 years, and even larger majorities of people over the age of 30 supported such legislation. In 1984,30 percent of the public favored laws fining drivers and front seat passengers $50 for not wearing seat belts (Gallup Poll 1984). Even broader support was shown for such laws among younger citizens. A national telephone survey of adults carried out for MADD by a research and consulting firm (Epilon 1985) provided additional evidence of the wide public support for this movement. Large majorities of the respondents believed that MADD: - should be involved in victim assistance programs (81 percent), - should be involved in promotingprevattive legislation (77 percent), - should be involved in promotingpwtirive legislation (83 percent), - should be involved in reviewing court decisions (67 percent), - should be involved in educating youth (92 percent), and - should be involved in educating the public at large (90 percent). Finally, leaders of the local group reported widespread contact and support from local community organizations. Groups working on drinking and driving issues came into contact with a wide variety of individuals and groups, but primarily with the local police, 252 BACKGROUND PAPERS the central offices of MADD and RID, high schools, judges, andstate police. Numerous contacts were also reported with churches, legislators, civic groups, and NHTSA. More than 85 percent of the leaders reported that, overall, their communities supported their activities and goals. Over half the groups reported that the State and local police, NHTSA, the national central offices, and other MADD groups were very supportive. Churches and high schools were seen as very supportive for about half of the groups. When asked to choose the community organization that had been the most suppotive of their work, leaders overwhelmingly chose the local police and the State police. After police, the most supportive groups were the central of&es of MADD or RID. Both offices have worked closely with their local groups, providing support and advice. Local prosecutors were also seen as the most supportive by a large number of groups. Only a few local groups have opposed the work of the anti-drunk driving advocates, and even these groups (e.g., bar and restaurant owners, alcoholic beverage distributors and retailers, and the trial lawyers bar) were mentioned as being antagonistic by only a small minority of local leaders. Moreover, opposition to the movement tends to be issue specific. For example, representatives of the beer industry actively opposed the efforts of these groups in some States to pass 21-year-old drinking age legislation (Wolfson 1988). Nevertheless, this opposition does not extend to the overall goal of the move- ment - the reduction of drunk driving- nor to many of its specific objectives that do not threaten the economic interests of the industry. In fact, segments of the beer industry have at times provided resources, in the form of literature and financial support, to some of the local groups, and have recently supported national legislation aimed in part at facilitating administrative revocation of drivers' licenses (Rumbaugh 1988). National and Local Media Attention Access to the mass media, in particular to print and broadcast news, is an import resource for any advocacy movement. The mass media represent a potential mechanism for "communicating with movement followers, reaching out to potential recruits, neutralizing would-be opponents, and confusing or otherwise immobilizing committed opponents" (Molotoch 1979, p.71). Knowing this, advocates invest extensive effort in attempting to gain positive coverage of their issue and their advocacy. We have been systematically monitoring national and local print media coverage of the drinking and driving issue. We obtained counts of the number of national newspaper stories devoted to the issue from the National Newspaper Index, 1979-1987, which indexes stories in Tire Christian Science Monitor, l7ze Los Angeles Times, T&e New York lTmes, TIte WaN Street Journal, and The Washington Post. We obtained counts of periodical stories on drunk driving from Magazine Indexes, 1979-1987. This data base consists of an index of 370 popular periodicals. Our local newspaper counts of drunk driving stories were obtained from Newsbank, a service that indexes more than 500 daily newspapers, including at least one in every State of the United States. Figure 2 shows the trends in coverage of the drunk driving issue derived from these three sources between 1979 and 1987. The peak year of coverage for each type of print medium was 1983. Coverage of the issue declined quite consistently after 1983. It is unknown whether this trend was also reflected in local and national broadcast media. Nevertheless, the decline in coverage can be seen as handicapping efforts by advocates to generate continuing community awareness of the issue of drunk driving. Whether the trend reflects declining vitality among advocate groups or is responsible for increasing alcohol-related fatalities, as some (Stevens 1987, Dukakis 1988) have suggested, cannot be easily determined. We asked local leaders to give us their evaluations of the extent of media coverage of the drunk driving issue in their communities. We also looked at the local newspaper coverage in 96 communities. Both perceptions of the level of local coverage and actual CITIZEN ADVOCACY Figure 2: Print Media Coverage of Drunk Driving 200 lm --- /----- c_------------____ ------- - -e-- 02 1979 1980 1981 lQ62 1903 1964 l905 1906 1907 253 wa.L- -- NATIoN*L- s----- YEAR 1579 1980 1981 1982 1983 1984 1985 1986 1987 laxal Nevspa;le= 63 94 223 395 423 411 331 224 163 Natimal -papers 17 3 17 El. 169 162 76 68 58 Fericdids 3 1 13 35 50 42 36 26 29 print coverage were strongly related to how active the local group was in attempting to get such coverage. We also found that the more general community support the leaders saw, the more support they saw from the media. Activities related to getting media attention were fundraisers, speeches, and an active membership (as opposed to groups primarily dependent upon leaders). Likewise, when the news was perceived favorably, the leaders reported high levels of recruiting from media sources. Local Groups, Activists, and Activities Information gathered from local advocacy groups across the country allowed us to describe their typical dimensions and the characteristics of their leaders. While a few of the groups were very large and resource rich, the typical group was small, with an average of 35 members and a mailing list of 100 names. About six people beyond the leaders did volunteer work for the typical group during an average month. Seventy percent of the groups had annual revenues of $2,500 or less in 1985, with the median revenue being $1,229. Most of these local groups relied primarily on leaders, volunteers, and donations of money and other resources (e.g., telephones, postage, and supplies) to carry on their 254 BACKGROUND PAPERS work. Though they were widely supported by their communities (Ungerleider et al. 1986), they depended primarily upon their members for labor and financial support (Weed 1989). G iven that we have counted more than 450 groups, this adds up to about 2,250 leaders, 2,700 regular volunteers, 15,750group members and45,OOO people on local mailing lists in 1985. The profile of the typical activist in this movement was quite similar to activists in other advocacy movements (Verba and Nie 1972; Eitzen 1970). Because of the size of the groups, the officers typically did much of the work. Not surprisingly, the typical chapter officer was a woman. She either did not work outside of the home or worked part-time. Often she was married with school-age children at home. Though all officers tended to be highly involved (Weed 1987), the chapter president tended to be the most active (McCarthy, et al. 1987). She was, typically, about 43 years old and had had some college education. Weed (1987), in his survey of the MADD chapters, wrote "Presidents were less apt to be in the labor force that other officers, and when they were employed they tended to hold slightly higher status jobs" (p. 265). Victims were heavily represented in all leadership positions in local groups, but the presidents were the most likely to report being victims (McCarthy et al. 1987). About one-fourth of the typical local group members were victims. The majority of presidents were also at least partly responsible for starting the organization. This was consistent with other studies (Weed 1987; Ungerleider et al. 1986) on the leaders of MADD chapters. Further, many leaders were already involved in other volunteer groups. This led Weed to conclude that "MADD tends to be run by activists who have been victimized rather than victims who have become activists" (1988b, p. 19). The most important program emphases at the local level were public awareness, youth education, and attitude change on the issue of drunk driving (McCarthy et al. 1987). Victim support and changing laws were emphasized less, as was recruiting new members. When asked to rate the program area in which they were most successful, the leaders overwhelmingly chose public awareness and changing public attitudes. About a quarter of the groups indicated that they had been very active in attempting to change laws during 1985. The groups used activities such as candlelight vigils, public booths, project gradua- tion,4safe rides, and poster contests extensively. Leaders spoke widely and gained substantial access to the broadcast media. The Structure of the National Movement National advocacy movements exhibit wide variations in form. Some, for instance, have very large and strong national advocacy operations but very little organized, local, grassroots strength, while others have the obverse. In some movements, local groups are under the very close control and supervision of a national organization, while in others, these ties are very weak. Figure 3 represents the national structure of the citizens' movement against drunk driving in 1985. Both RID and MADD national offices provided support services of many kinds to their local groups, although MADD had substantially more resources for doing so. Both groups had organized intermediate levels of coordination at the State level. The formal tie between MADD national and local groups was, at least in theory, tighter than that between RID groups because of the Internal Revenue Service status of MADD. Each local group was a subgroup of a single national 5Ol(c)3 organization. This status meant that the national organization bear some level of financial responsibility for local groups. Frank Weed reported that "The opera- 4 Project graduation is a variety of activities that may include offering high school students free rides to and from graduation parties or arranged overnight lodging to sponsoring alternative alcohol-free parties. CITIZEN ADVOCACY 255 Gcoqraphicsl Level of Ornanitat ior2 National Figure 3: The Structure of the National Citizens' Advocacy Movement Against Drunken Driving state Local Cl-OUPS Emup Structure tions of the local chapters are virtually autonomous and the central office of MADD has some difficulty getting reports from all its chapters in a clear and timely manner" (1987, p.7). Local RID chapters, on the other hand, were not linked to RID national in this way, and each group made its own decision about its I.R.S. status. In fact, in 1985, the majority of local RID groups, as well as the outliers we contacted, had already received or had applied for formal nonprofit status. Regardless of the formal nature of the tie between locals and nationals, however, local leaders reported extensive contacts and high levels of support from their national group as well as from State coordinators where they existed. Likewise, lateral contacts and support were common at the grassroots level; that is, local leaders reported high levels of contact and support from other local groups working on the drinking and driving issue in their areas regardless of the national affiliation of the other group. The majority of local leaders reported some contact with local SADD groups, and about a third reported extensive contact (McCarthy et al. 1987). The major strength of the anti-drunk driving advocacy movement lies in its extensive local groups. Weed's analysis of the failure of local MADD chapters between 1985 and 1987 demonstrated that local groups are more likely to survive when they are more independent of the central office on operating policies but cooperative in carrying out common activities (198&).This suggests that the typical fierce independence of the local groups throughout this movement is also one of its major strengths. Lessons from Independent Citizen Advocacy Movements Observations across many citizen advocacy movements make obvious the difficulties in predicting their emergence and their ebbs and flows. Many thousands of fatalities in 256 BACKGROUND PAPERS alcohol-related automobile crashes occurred, for instance, before the first citizen a.d- vocacy group emerged to address this problem. It took the development of a widespread belief that the number of these incidents could be reduced by collective efforts before local groups began to form (Gusfield 1975, 1981). The continuing formation of new groups, the continuity of older groups, and the perseverance of the individual activist leaders and volunteers who staff them depend-upon a variety of factors beyond the objective extent of the drinking and driving problem itself. These factors are central to an understanding of how advocacy movements grow and decline. A number of them are particuldy important for assessing the future prospects for growth, stability, or decline of the citizens' advocacy movement against drunk driving. ~crivisr lives. Few activists understand the potentially massive Ievel of commitment of time and energy they are making when they begin their advocacy careers. We know (McCarthy and Zald 1977; McAdam 1987) that one's ability to devote such large amounts of time to advocacy are, for most people, constrained by other obligations such as jobs and parenting. Movements that depend on leaders who have little availabIe time to devote to activism are handicapped. The drunk driving movement benefits from its heavy reliance upon women leaders who do not work outside the home. Our research showed that local leaders who were employed full-time devoted substantially less time to the activities of their local group than those who were not. Having children in the home did not seem to hinder activism in this movement, however, since many of the local groups were physically based in the home of the president and integrated family members into the activities of the group (Harvey and Wolfson 1987). nle temporal and spatial dimensions of victimhood. Most advocacy movements form around a commonly timed victim experience. The citizens' movement that formed in response to the Three Mile Island nuclear "accident," for instance, depended upon many citizens responding in unison to a common event (Walsh and Warland 1983). Such movements experience common cycles of increasing and then declining enthusiasm among victim activists. The anti-drunk driving movement, on the other hand, given the discrete and disconnected nature of the victim experience, can expect to see a constant replenishment of the pool of new victims who might become activists. This is so because alcohol-related fatalities continue at high rates in most communities. To the extent that new pools of victims and other concerned citizens can be integrated into the activities of ongoing local advocacy groups, this movement should transcend the pattern of decline that results from the typical common timing of victimization. Bureaucratization and goal displacement. A typical pattern of transformation charac- terizcs the history of many citizen advocacy groups. As such organizations acquire more and more resources, leaders, who have a financial and personal stake in their operations, begin to lose track of their original goals (Perrow 1979). This pattern is especially likely to result from wide success. At the local level, the citizens' movement against drunk driving seems likely to avoid these consequences since most groups depend almost exclusively upon volunteer labor and are resource poor. Our evidence in 1985, however, showed that the vast majority (72 percent) of the presidents of the local groups were original group founders (McCarthy et al. 1987). Since most of the groups were thenvery recently founded, this may not represent a problem of too long-entrenched leadership. Cenfralization and decentralization. A lively debate continues among observers of citizens movements about the relative advantages and disadvantages of decentralized and centralized organizational structure (Gamson 1975; Piven and &ward 1977). Centralization allows more national coordination and the concentration oflocal energies toward common goals. However, it has the disadvantages of bureaucratization described above. But decentralization-that is, a movement composed of quite autonomous local groups -has certain advantages, too (Gerlach and Hine 1970). Most important is the likelihood of more creative innovation in goals and tactics. If a centralized organization undertakes an innovative campaign that is poorly conceived, an entire movement can suffer the consequences. This is part of the reason why large, national advocacy groups CITIZEN ADVOCACY 257 become cautious. However, if an autonomous local organization does SO, the costs of its failure are minimal. If such innovation is successful in one locale, it can be tried in others. This way, new successful forms and goals of advocacy spread rapidly across local groups. Another advantage of decentralization is that an entire movement is less responsible, in the eyes of the public and potential supportive groups, for "rogue" locals, that is, focal groups that deviate widely from the common goals and tactics pf the larger movement. The movement against drunk driving is highly decentralized in comparison with many other advocacy movements. This feature of its structure seems to have been responsible, in part, for the constant development of new programs and approaches in local com- munities. To the extent that solutions to the problem of drinking and driving can be achieved at the local and State levels, its decentralized form can be seen as an advantage. To the extent that solutions to its problems lie at the nationaI Ievel, this feature of its national structure may represent a disadvantage. Support by tzomictims. Many advocacy groups are made up, primarily, of individuals who are not the direct victims of the problem for which the groups seek solutions. At the grassroots level, the drunk driving movement is, as we have seen, heavily peopled by victims. To the extent that their commitment to continued efforts for solutions to the problems of drinking and driving are greater than nonvictims, this movement will benefit from such heavy levels of victim involvement. Likewise, the range and level of support by elite groups is crucial to the success of advocacy movements. This movement has benefited greatly from the strong support of such outside groups as we have shown, and, probably, would have been far less vital without that extensive support. NHTSA has continued to be supportive in many ways. The annual "Lifesavers" conference, supported by a number of outside groups, has been important in networking local leaders with one another as well as with researchers, police representatives, and industry supporters, especially insurance companies, which have a special stake in the issue of drinking and driving. To the extent that the past levels of such support continue, this factor suggests that the movement will not decline in vitality. Media cycles alld adwcacy morpemertt cycles. Many factors contribute to under- standing the extent of media coverage of any issue at a particular time. These include, importantly, major events that focus attention on the issue, concerted campaigns to focus attention on the issue by advocates, and processes internal to the production of the media outcomes themselves. Consequently, media attention to any social issue does not neces- sarily reflect its objective importance or the strength and efforts of the advocates concerned about the issue (Graber 1984). The rapid decline in print media attention to drinking and driving during the last several years, therefore, does not necessarily reflect a decline in public concern about the issue or a declining effort by local groups to continue to generate public awareness around the issue. Cycles of media attention lie the one we see here are typical and have drawn the attention of many observers. Anthony Downs (1972) argued that social problems will suddenly become prominent, grasp the public attention for a short time, and then gradually lose public attention. This cycle, he explained, is embedded in the nature of both problems and media. Problems, once understood, are difficult to solve. And, since the public consumes news partially as entertainment, a problem must be exciting, and continue to be exciting, to maintain the public's interest. Our research, which linked the extent of local leaders' activities in 1985 to the extent of local print media coverage in their communities in 1985, demonstrated that the more effort local leaders invested in attempting to gain such coverage of the issue, the more coverage they actually got. Yet, to the extent that the declining media attention to the issue of drunk driving is the result of general processes of media coverage rather than the level of effort and skill of advocates to bring the issue to public attention, it will be more difficult to reverse the cycle we have observed. 258 Consensus and conflict. Most advocacy movements meet substantial public and organized opposition. The anti-drunk driving movement is unusual in that it has achieved wide and deep support for its goals. But advocacy movements can rapidly gain or lose public support depending upon how they frame the issues and propose their solutions. For instance, Lo (1984) explained how the previously unpopular advocates for property tax reform in California altered their goals'and, as a result, became the successful "tax revolt" movement. The anti-drunk driving movement could easily frame its goaIs in ways that wouId substantially narrow the breadth of its community support. If it is seen as a "prohibitionist" movement, declining support could be expected. The Epsilon Survey (1985) presented evidence showing that citizens who perceived the movement as "anti- alcohol" were far less likely to support its instrumental goals than those who viewed it otherwise. Most organized elements of the movement have insisted on defining the issue of drunk driving to preclude public perceptions of "anti-alcoholism." To the extent that this framing of the problem continues to dominate the goals and rhetoric of the movement, it will probably continue to garner the wide and deep community support it has seen in the past. Encouraging Advocacy Movements Individuals and organized groups can facilitate or, alternatively, attempt to inhibit, the efforts of advocacy groups (Marx 1979; Wolfson 1989). Members of the general public, governmental actors, and representatives of all types of private groups at the national, State, and local level have many avenues for encouraging or discouraging the efforts of advocacy movements. The effectiveness of such efforts depend, in an important part, upon the characteristics of the movement in question, such as its organizational form, its level and location of community support, its leaders, and its typical tactical approaches. The three main types of facilitating efforts aim at (1) increasing the level of general community support, (2) indirectly improving the opportunities for organized advocacy efforts, and (3) making direct support available to organized advocacy groups. Each type of facilitation has been common in the movement against drunk driving. Examples of specific forms for each type follow. 0 Creating general community support for - Advocacy and advocacy groups - Concern for the general issue - Specific goals of advocacy (e.g., "administrative revocation") - Specific advocacy actions (e.g., a Surgeon General's Conference) o Indirectly improving advocacy effectiveness - Make general models for advocacy available - Facilitate communication among advocates - Set broad legislative and regulatory agendas - Provide new opportunities for advocacy (e.g., task forces) o Direct support of advocacy - Provide useful materials (e.g., literature) - Provide expertise - Provide training (e.g., use of volunteers) - Provide direct support (e.g., money or space) CITIZEN ADVOCACY 259 These examples of specific forms of facilitation of advocacy do not exhaust the possibilities. Conclusions about the most effective relative mix of the three types of facilitation and their most effective specific forms for the citizens' movement against drunk driving demand serious deliberation. REFERENCES Danforth, J.C. "Hearing on S. 2.549, The Drunk Driving Prevention Act of 1988." Statement presented at the Senate Committee on Commerce, Science and Transportation, Con- sumer Subcommittee, August 2,1988. Downs, A. Up and down with ecology - The `Issue-Attention Cycle'. Public Interest 2838-50, 1972. Dukakis, M.S. No holds barred (Letter to the Editor). Boston Magaike lO:February, 1988. Eitzen, D.S. A study of voluntary association members among middle