Alcohol and Non-Vehicular Trauma

People who abuse alcohol are both more likely to be involved in physically traumatic events and more likely to be seriously hurt in them than nondrinkers are. In fact, according to the National Hospital Ambulatory Medical Care Survey (Stussman 1997), more than half of all emergency-room visits associated with alcohol or drug use involve traumatic injuries. The report also notes that individuals who die from injuries are more likely to have used alcohol more frequently and heavily than are those who die from disease.

Other studies indicate that alcohol is a factor in 21 to 47 percent of drownings, 35 to 63 percent of deaths from falls, and 12 to 61 percent of fire-related deaths (Hingson and Howland 1993). Burn victims with blood alcohol concentrations (BAC's) above 0.06 percent were more than twice as likely to die as burn victims with BAC's of 0.06 or less (Haum et al. 1995).

Alcohol abuse is particularly prevalent among patients with head trauma. Research findings vary, but generally indicate that more than 50 percent of those who sustain head injuries were drinking alcohol before or at the time of injury (Kraus 1993). To put the consequences in hard dollar terms, in 1994 the National Head Injury Foundation estimated the lifetime medical expenses for a person with a severe head injury at $4.6 million.

Alcohol, Crime, and Violence

The use of alcohol is indisputably associated with homicide, sexual and other assault, domestic violence, and child abuse (Roizen 1997). A review of several studies of drinking patterns and homicides shows that 7 to 85 percent of murderers had been drinking when they committed their crimes; most of the studies reported that 60 percent of homicide offenders were drinking at the time of the offense (Murdoch, Pihl, and Ross 1990). A study of police reports in northern Ontario revealed that among the several hundred incidents reported by community residents, 20 percent involved marital abuse, in which 44 percent of the assailants and 14 percent of the victims had been drinking (Pernanen 1991).

According to a 1990 study by the National Institute on Alcohol Abuse and Alcoholism (NIAAA), alcohol also has factored in approximately a third of all suicides. In addition, positive blood alcohol levels are frequently found among victims of drug overdoses (Commonwealth of Massachusetts 1995).

Alcohol and Youth

Nationwide studies report that more than half of America's 8th graders and 80 percent of 12th graders have tried alcohol, even though buying it is illegal for anyone under 21. Even worse, heavy drinking is reported by 15 percent of 8th graders, 28 percent of 12th graders, and 34 percent of college students (Johnston, O'Malley, and Bachman 1995). While these percentages appear to be on the decline, they remain disturbingly high. Comparing data from 1990 with 1984 surveys, Midanik and Clark (1995) reported a significant increase among 18- to 29-year-olds who have experienced two or more negative social consequences due to drinking. Among all age groups, these young adults also comprised the highest percentage of individuals classified as alcohol abusers and alcohol dependents (Catalano and Room 1994; Grant et al. 1994).

Among young people in particular, alcohol use is known to be associated with a decreased likelihood of condom use and other measures to prevent pregnancy and disease transmission from sexual intercourse (Strunin and Hingson 1992). Heavy consumption of alcohol along with other drug use is also associated with poor school performance, high school dropout, and problems with work productivity later in life (Cook and Moore 1992). Young people entering the work force immediately after high school tend to maintain the drinking patterns established in their last school years; approximately 30 percent report that they are heavy drinkers (Harford 1993).

A survey at 140 colleges found that 44 percent of students reported binge drinking, defined as five or more drinks per sitting. Frequent binge drinkers were 10 times more likely than non-binge drinkers to report trouble with campus police, damages to property, injuries, and unplanned or unprotected sex (Wechsler et al. 1994). A nationwide survey of students in grades 9 through 12 found that the prevalence of drinking and driving increased substantially with alcohol use and especially binge drinking (Escobedo, Chorba, and Waxweilen 1995).

Alcohol Consumption and Related Problems

It is clear that an individual's level of consumption is related to the nature and extent of any problems that result from alcohol use, as demonstrated by risk function analysis, which looks at relationships between volumes of alcohol consumed and the social consequences (NIAAA 1997). A report by Room, Bondy, and Ferris (1995) states that in six life areas, the probability of harm rises steadily with the respondent's alcohol consumption. Considering similar findings by other researchers, Edwards et al. (1994) observe that the lessons "for the individual and society are outstandingly clear — less is better, more drinking carries more risk for a wide range of adverse happenings, and heavy drinking is distinctly dangerous behavior" (p. 68).

The health and social problems associated with alcohol do not apply only to individuals who are clinically dependent on alcoholabout 7 percent of the population (Moore and Gerstein 1981). In fact, many problems occur among individuals who are not dependent on alcohol, while many others use alcohol without experiencing negative health consequences (NIAAA 1993). But problems also occur as a result of community tolerance of the social and economic norms for alcohol availability and use. Therefore, understanding the risks can play a significant role in shaping a community's response to alcohol-related problems.

Researchers in the field are looking for effective methods to curtail these problems while at the same time allowing for the responsible and appropriate distribution of alcohol. This Reference Guide analyzes research and practice findings on efforts to manage alcohol availability and synthesizes the available knowledge on the effectiveness of these approaches in solving specific problems.

Alcohol Availability and General Public Consumption

Studies of natural experiments clarify the relationships among alcohol availability, consumption, and related problems by analyzing the effects of market-driven or other naturally occurring changes in alcohol availability. Such studies have examined strikes by liquor store workers, prohibitions on alcohol sales, and shifts from monopoly to private alcohol sales and vice versa. Despite considerable variations in the ways these natural experiments took place, in general they indicate that a reduction in alcohol availability appears to reduce the rates of consumption and drinking-related problems such as public drunkenness, assaults, and drunk driving, at least immediately following the change in availability (Edwards et al. 1994).

Alcohol Availability and Crime and Safety Issues

Alcohol consumption is a factor in more than half of all homicides and serious assaults (Pernanen 1991). Among the respondents to Pernanen's study who had visited a public site where drinking took place at least twice in the previous 30 days, alcohol-related violence was more likely to be reported at on-sale outlets than in homes, work sites, schools, or the streets. Pernanen points out, however, that the relationship between alcohol and violence is complex and influenced by a variety of cultural, historical, personal, and other situational factors that in combination with the arousing effects of alcohol can turn into aggressive behaviors that lead to sexual and other types of assault.

Studies suggest a strong association between violence and the retail availability of alcohol (Cook and Moore 1993a, b; Fagan 1993a, b; Lester 1993; Parker and Rebhun 1995). As a result, public health officials and legislators as well as researchers are furthering their efforts to understand this relationship. For example, the higher rates of violence at bars have inspired studies of management practices and policies for serving alcohol, leading researchers to look at placing greater responsibility for any resulting problems on the managers of on-sale alcohol outlets (Holder and Wagenaar 1994).

Alcohol Availability and Traffic Crashes

Various links between alcohol availability and traffic crashes have been examined (Gruenewald et al. 1996; Holder 1993a; O'Malley and Wagenaar 1991; Scribner, Mackinnon, and Dwyer 1994; Van Oers and Garretsen 1993), including the relationships between alcohol-related crashes and minimum legal drinking age; density of licensed alcohol outlets; and place of last drink and the resulting incidence of driving while intoxicated (DWI) arrests.

A number of studies have found an inverse relationship between raising the minimum legal drinking age and the number of traffic crashes resulting in personal injury or death (O'Malley and Wagenaar 1991; Wagenaar and Wolfson 1994). The rate of crashes was shown to decline when the minimum age for legally purchasing alcohol was raised from 18 to 20 or 21 years.

As might be expected, researchers have identified a direct relationship between the density of alcohol outlets and the number of traffic crashes in a given area (Scribner, Mackinnon, and Dwyer 1995; Van Oers and Garretsen 1993). Research also has shown that the highest percentage of individuals charged with DWI consumed their last drink in a bar or restaurant; a study by Fell (1988), discussed in McKnight (1991), for example, reported that 30 percent of drivers arrested for DWI in Maryland were en route from bars or restaurants. Furthermore, data collected in roadside surveys indicated that a major share of the drinking by intoxicated drivers occurred in bars and restaurants (Interministerial Committee on Drinking-Driving 1980; McKnight 1991; Palmer 1986; Wolfe 1975). Overall, the proportion of DWI arrests after drinking at bars and restaurants to arrests after drinking at residences is greater than might be expected, even taking into consideration the relative proportions of alcohol sold through on- and off-sale outlets.

Alcohol Availability and Youth Access to Alcohol

In July 1984, Federal legislation resulted in all 50 States and the District of Columbia setting a minimum legal drinking age of 21 years. Despite retailers' awareness of these laws, research reveals that persons under 21 are frequently and even routinely able to purchase alcohol (Forster et al. 1994; Preusser and Williams 1992; Wagenaar and Wolfson 1994). In fact, of a U.S. population of more than 18 million 16- to 20-year-olds, most report drinking alcohol (University of Michigan Institute for Social Research 1998).

State and Local Responsibilities for Alcohol Availability-Related Problems

The 21st Amendment to the U.S. Constitution gave control over alcohol availability to the individual States, which have chosen to share much of the responsibility with local jurisdictions. States generally look to their cities and counties to establish the conditions under which an outlet may sell alcohol (location, hours of operation, type and size of establishment, and business practices) within the State's broader constraints, such as taxes, retail licenses, and prohibitions on sales to minors. States also leave decisions about alcohol sales at public events and in public places owned by the city, county, or other public agencies to the localities. Along the same lines, the States do little to influence the advertising and promotion of alcoholic beverages at the local level and tend to rely on local police for assistance in enforcing existing regulations. As a result, communities have considerable authority to manage local alcohol availability in the interest of public health and safety; these powers are addressed in detail in chapter 2.

Local decisions about alcohol availability are based in part on local perceptions of the potential problems and benefits from specific policies. These perceptions are influenced by factors such as recognition of the relationship between alcohol availability and health and safety problems and the community's values and traditions regarding alcohol. Regional traditions and practices often determine the places, times, and occasions at which drinking occurs. Socioeconomic factors, such as the availability of alcohol and the purchasing power of local consumers, also play a significant role: to a considerable extent a local economy's dependence on alcohol sales and the interests of alcoholic beverage retailers and distributors shape communities' decisions about alcohol availability (Holder and Giesbrecht 1989).

Because community residents are often the first to experience the effects of alcohol-related problems, they are in an ideal position to start solving those problems at the neighborhood level by calling on local governments, schools, neighborhood and religious groups, and voluntary and business organizations. Community-based organizations play a major role in identifying problem areas, linking existing resources, developing recommendations for solutions, and creating effective policies to reduce alcohol-related problems. One example of successful community action occurred in South Central Los Angeles, where problems caused by alcohol availability received considerable media attention in the aftermath of the April 1992 civil disturbances over the acquittal of four white policemen charged with beating black motorist Rodney King (see appendix F). A local community coalition for the prevention of alcohol and other drug problems had been meeting, to little effect, with city officials to address the high density of liquor stores and their proximity to residences in two city council districts. The riots put a third of the off-sale alcohol outlets in these districts out of business, setting the stage for community action against drug-dealing, excessive drinking, and other outlet-related problems — actions supported by public health activists, lawyers, bureaucrats, and elected officials throughout California. The Los Angeles Police Department and researchers from the University of Southern California report that the rate of crimes associated with proximity to liquor stores has declined as a result. (Chapter 4 examines specific problem-solving processes local agencies and community groups have used to tackle alcohol-related problems.)

High-Risk Alcohol Outlets

Many communities across the United States have not only public health and safety but broader "quality-of-life" concerns related to alcohol availability. Some studies have found that the complaints about alcohol outlets most often reported to city planners had to do with noise, traffic, litter, or loitering (Wittman and Hilton 1987). While these may seem relatively minor aesthetic concerns, they can indicate serious underlying problems; poor property maintenance, for example, appears to be closely linked to health and safety problems (Wilson and Kelling 1982), and loitering with drug dealing and prostitution.

Violations of health and safety standards provide a basis for State ABC license sanctions and local agency code enforcement. After all, freedom from unwanted interruptions in one's home and place of business are fundamental legal rights.² [² Land-use law, for example, is dedicated to protecting the occupants' "health, safety, welfare, and morals" (City of Euclid v. Ambler Realty Company 1926).] A basic tenet of law is the right to "quiet enjoyment" of one's own property. When these rights of residents, shopkeepers, and others who live and work in an area are threatened, local zoning laws as well as police enforcement can be brought to bear. (This issue is discussed in detail in chapter 2.)

Communities can prevent or reduce such problems by identifying alcohol outlets at high risk for generating them and then taking action to eliminate or at least reduce the risks. Designating an outlet as high-risk is a matter of both community perceptions and a legal determination by State and local agencies. An outlet is determined to be at high risk by law if it endangers the public health, safety, and well-being of the community through one or more of the activities highlighted in the box at right.

Research continues to show a correlation between a high density of alcohol outlets and a high rate of alcohol-related problems in a community (Edwards et al. 1994). Studies have identified three indicators of local alcohol outlet density that merit the attention of prevention planners (Wittman 1986):

• Infiltration — The introduction of alcohol outlets to commercial establishments from which they had been excluded, such as beer and wine sales in fast-food chains. Infiltration is associated with general increases in alcohol consumption levels and other documented outlet-related problems (Bruun et al. 1975; Edwards et al. 1994; Gruenewald, Ponicki, and Holder 1993).
• Overconcentration — The clustering of alcohol outlets in a confined geographic area. Overconcentration is characterized by the presence of several alcohol outlets within a city block or other area small enough for people to walk easily from one alcohol outlet to another. Overconcentration is associated with crime and public safety problems, economic displacement, and deteriorating quality of life. For example, one city with a population of 90,000 has 44 bars and restaurants in a four-block entertainment district that accounts for some 25 percent of the city's calls for police service (Vasquez, Wittman, and Harding 1998).
• Proliferation — The establishment of retail alcohol outlets in a given area at a faster rate than that of other commercial outlets. The sudden proliferation of alcohol outlets can be especially troublesome in rapidly expanding areas such as suburban housing developments, dislocating other retail businesses. In one community, for example, over some 6 years an eight-block commercial district that became "trendy" experienced a fourfold increase in bars and restaurants, driving out other businesses and boosting criminal activity in the area.

Mixed Messages About Alcohol Consumption

The conflicting messages American pop culture sends about alcohol use complicate community attitudes toward alcohol availability. National marketers use myriad media to encourage the consumption of alcohol: many newspapers, magazines, and radio and television programs put forth themes and characters that promote not only tolerance but a downright positive image of alcohol use. What's more, alcohol manufacturers and distributors broadcast and publish state-of-the-art advertisements aimed to impress youth and push not only brand loyalty but alcohol consumption in general. Although Federal licensing guidelines require radio and television stations to air public service announcements, very few of these messages convey the hazards of drinking or promote the acceptability of temperance.

Of course, it is for obvious economic reasons that the alcoholic beverage industry presents its products in the best possible light and rarely mentions the negative consequences of drinking (Wallack et al. 1993). Those prevention messages the industry does send, such as "Know when to say when," are difficult for the public to use as a practical guideline. Meanwhile, local media interest in news with novelty may lead to inaccurate or incomplete coverage of public health and safety issues concerning alcohol. For example, when research indicated that modest levels of alcohol consumption (up to two drinks per day) may be associated with a lower incidence of heart disease while higher rates of consumption over long periods may be very harmful (Klatsky 1994), media reports appeared on the "health benefits" of alcohol without mentioning the appropriate caveats and clarifications — omissions that could be misleading and even damaging to the public's health.

Society's mixed messages about alcohol are highlighted in many communities' controversies over the availability of alcoholic beverages at public events in public places, such as county fairs and Fourth of July celebrations. Supporters of availability view alcoholic beverages as a fund-raising tool that should be readily available for people to enjoy; prevention workers and some community groups, on the other hand, object to the identification of community and cultural events with alcohol products, while public officials have reason to be concerned about behavior problems, extra clean-up costs, and property damage resulting from alcohol availability at such events.

Overall, mixed messages about alcohol can create an environment hostile to local groups and agencies engaged in prevention activities. Prevention planners are therefore faced with disentangling the conflicting messages about the risks of alcohol consumption and its potential benefits in limited quantities. Effective strategies for responding to such challenges are reviewed in chapter 2.

ABC Licensing and Enforcement Activities

State ABC's began to be established in 1933 in response to the alcohol-related problems that sprang up around Prohibition and the economic needs created by the Great Depression of the early 1930s. These ABC agencies worked quickly to establish controls aimed at eliminating crime and corruption in the alcoholic beverage industry, promptly removing the last vestiges of the lately flourishing illegal bootlegging industry and instituting regular collections of alcohol license fees and taxes the Nation's teetering economy so urgently needed.

The ABC's established licensing and enforcement operations to make sure that retail alcohol outlets complied with new State standards. The States approached alcohol licensing in one of two ways: license States issued private parties licenses to sell alcohol by retail, while control States allowed only their own ABC's to act as alcohol retailers (see Box). Both systems require reviews to ensure that proposed retail outlets conform to published State standards governing alcohol serving practices, permissible locations, hours of sale, and other relevant matters. ABC field officers conduct these licensing investigations and monitor the operations of existing outlets. The ABC field officers operate under the same circumstances and exercise the same powers as sworn local police officers, who are also authorized to enforce State ABC laws.

Today's State ABC's govern the manufacture, production, wholesale distribution, and retail sales of alcohol under uniform standards established in each State and integrated into a single control system designed to promote fair business practices throughout the alcoholic beverage industry. In addition, State ABC's maintain production standards, ensure proper alcohol tax collection, monitor the industry's business practices, and work to reduce alcohol-related criminal activity.

ABC Interactions With Other State Agencies and Local Jurisdictions

By law, States have both the power and the responsibility to regulate all aspects of alcohol availability, and they can preempt local efforts to regulate alcohol outlets. In practice, however, States share much of their power and responsibility with cities and counties; after all, decisions about the distribution and operation of alcohol outlets can best be made within the communities they affect, which are in a far better position than any State agency to manage the nuances of their relationships with alcohol outlets. Similarly, ABC's benefit from the cooperation of local police and planning departments in handling State licensing and enforcement activities.

More than 60 years after the repeal of Prohibition, the primary concerns of State ABC's remain regulation of commerce, collection of revenue, and control of crime and corruption associated with the alcoholic beverage industry (Morgan 1980). Public health and community welfare issues linked to alcohol availability are included in the State charters of many ABC's, but historically have received little attention, as State ABC's typically lack the authorization, staff, or funding to pursue such matters.

Developments over the past 20 years have, however, put increasing pressure on State ABC's to pay attention to public health and community well-being. First, as described in chapter 1, research findings concerning the nexus between communities' alcohol availability and alcohol-related problems suggest that the grouping and distribution of retail alcohol outlets deserve more attention. There is a clear need, for example, to improve assessments of the environmental factors associated with geographically clustered outlets and with those in high-risk locations, such as near college campuses and in areas where drugs are known to be sold.

Second, the advent of Federal block grants to the States in the late 1970s gave State and local alcohol and drug program agencies the added responsibility of defining their own prevention initiatives. Federal prevention grants were awarded directly to communities to stimulate local government agencies and private organizations to draw upon their own resources to solve alcohol-related problems. These developments, coupled with the growing body of research on the relationships between alcohol outlets and alcohol problems, boosted interest in local control (Wittman and Shane 1988). Communities made more and more requests for service to State ABC agencies, and local prevention programs began demanding stronger State and Federal policies to support community-level prevention efforts.

Third, many State ABC's are facing ever-greater workloads as a result of a continuing decline in their budgets — and thus their staffs — relative to the number of retail alcohol outlets and to growing State populations. In California, for example, the ratio of ABC field investigators to retail alcohol outlets soared from one officer per 43 retail outlets in 1955 to one officer per some 340 outlets in 1997, a total of 200 ABC officers for 68,000 retail outlets statewide (George Reece, Deputy Director, California Alcoholic Beverage Control Agency, personal communication to Friedner Wittman, September 1988). It is hardly surprising that many State ABC's are eager to work closely with local jurisdictions and to share the responsibility for licensing and enforcement.

Local Zoning Laws and Land-Use Ordinances

Over the past decade, local zoning laws and other land-use ordinances have gained popularity as a means of controlling the risks posed by alcohol outlets. Such ordinances protect the community by establishing land-use zones — districts or areas that can be used only in certain ways. These zones require operating conditions and facility designs that limit high-risk behaviors on or near the site in question; an established local review process weighs the merits of every zoning application. By setting such limits on permitted land uses and in some cases denying even those permits, zoning ordinances guard communities against a variety of hazards. For example, zoning ordinances separate industrial plants that use dangerous manufacturing processes from residential and other vulnerable areas such as hospitals and schools (Babcock 1966; Bassett et al. 1935; Christoffel and Teret 1993).

Modern zoning ordinances came into widespread use throughout the United States following a 1926 Supreme Court ruling that cities, counties, and other locally governed jurisdictions have the power to protect the public's "health, safety, welfare, and morals" (City of Euclid v. Ambler Realty Co. 1926). Over the past 70 years, local communities have effectively used zoning laws to protect the public health by assisting ABC's and other State agencies in carrying out their duties. It is therefore not surprising that State governments generally recognize the value of local zoning laws and protect their use. Most State ABC statutes provide ample latitude for the application of local zoning ordinances to State-licensed or -operated retail alcohol outlets. Typically, applicants for State ABC licenses are asked to obtain local zoning clearance before submitting their applications to an ABC.

Communities have a significant opportunity to manage the potentially troublesome aspects of retail alcohol availability by limiting alcohol outlets through the local zoning code (Wittman 1994), as follows:

• Density restrictions can limit the number of alcohol outlets by population, geographic area, or as a percentage of all retail alcohol outlets in a given commercial area. (Limits on the number of outlets according to population are more often found in State ABC statutes than in local ordinances.)
• Restrictions on hours of operation can limit the times when alcoholic beverages can be sold.
• Spacing restrictions can specify a minimum distance between alcohol outlets and schools, churches, residences, parks, playgrounds, or other community sites.
• Standards of operation and facility design can ensure that outlet premises are securely maintained and that staff are trained not to sell alcohol to underage youths, inebriates, or those likely to drive. Zoning ordinances may also require certain architectural design features to prevent theft or loitering (Clarke 1992).
• Types of alcohol outlets can be limited to certain zones or disallowed altogether. For example, bars that offer live entertainment generally are not permitted in residential areas, and some communities do not permit drive-through outlets anywhere.

These limits can be stated as "conditions" on the zoning permit issued to any individual establishment. Such limits may be imposed either as standard conditions (a list of the limits typically required for each application) or as discretionary conditions (restrictions applied on a case-by-case basis and tailored to the specific features of the proposal and the surrounding land area). Most communities use standard conditions supplemented by discretionary conditions as circumstances warrant (Wittman 1994).

Standard conditions designed to curtail alcohol-related problems by restricting activities in and around licensed on- and off-sale premises include the following:

• Authorizing police officers to enter a licensed alcohol outlet at any time to determine whether any regulations have been violated.
• Ensuring the presence of an on-duty employee over the age of 21 at all times when alcoholic beverages are sold.
• Hiring security guards to protect patrons and prevent damage to the premises.
• Preventing underage youth from entering the premises during certain hours or without an accompanying adult.
• Requiring Responsible Beverage Service (RBS) training for staff of retail alcohol establishments and short-term permit events.
• Requiring soundproofing and fencing to avoid disturbing neighbors.
• Restricting nude entertainment.
• Restricting the hours and days when alcohol is sold.

"Active" and "Passive" Use of Local Zoning Ordinances

Communities may make active or passive use of local zoning ordinances. Active use requires that every application be reviewed to determine whether the proposed land use complies with the ordinance. This review weighs the risks and benefits to public health, safety, and community welfare associated with the proposal. This type of zoning is administered through conditional-use permits, which impose the conditions described above on the land-use permit application.

Procedures for case-by-case CUP review of local planning and zoning ordinances are generally the same throughout the United States. Applications for zoning permits are initiated by the property owner or a designee. Staff officials in local agencies then conduct reviews to determine whether the proposed land use meets zoning code requirements and whether problems might occur if the permit is granted. Once the official findings are completed, recommendations are made to the local zoning authority (usually a zoning review board). This local authority conducts a public review of the findings and decides whether to grant the use permit as requested, to grant it with some modifications, or to deny the request. This decision may be appealed to the local legislature and then to the courts (Babcock 1966; Bassett et al. 1935; Wittman 1982, 1994).

Under a passive review system, local zoning authorities are restricted to determining whether an application conforms to the published requirements of the local zoning code. If the proposed land use complies, a use permit is issued automatically, or "as of right."

Active zoning is preferable in preventive land use planning for retail alcohol outlets. Passive zoning does not offer the opportunity for careful review of the merits of each case; as a result, problems on an application may be overlooked. Moreover, once granted, a passive use permit cannot easily be modified or revoked unless appropriate conditions were imposed at the outset. Should problems arise later, the enforcement process is likely to take longer and will probably not be as effective as under an active use permit (Wittman and Shane 1988).

Some U.S. cities and counties have been making active use of planning and zoning ordinances to prevent problems with retail alcohol outlets for 20 years (Wittman 1982, 1994). This is particularly true in California, where planning research on local zoning of alcohol outlets got under way in 1979. In 1987, at least 30 percent of the State's approximately 470 municipalities were using CUP ordinances for both on- and off-sale alcohol outlets (Wittman and Hilton 1987); a 1993 estimate by the League of California Cities suggests that about half the State's municipalities were then making active use of CUPs. Community prevention coalitions, local officials, and local police departments have found that CUP conditions for alcohol outlets provide a vital tool for controlling and preventing alcohol outlet-related problems (Wittman, Helmke, and Biderman 1992; Wittman and Shane 1988). Appendix F outlines an example of a CUP ordinance applied to prevention problems.

Zoning Enforcement and Nuisance Abatement

Active zoning through CUP ordinances enhances enforcement efforts by allowing prompt intervention when alcohol outlets are violating their State ABC licenses or local land use permits. If an outlet violates both local zoning conditions and State licensing requirements, local authorities have greater leverage to bring the operator into compliance. If informal discussions with operators prove ineffective, their violation of CUP conditions offers grounds for revoking the local use permit. When it is clear that standards will be enforced, voluntary compliance with community standards and sensitivity to neighborhood concerns can be expected to increase.

Sometimes, however, a more forceful approach may be required. If initial attempts to obtain voluntary compliance fail, the State ABC and the local jurisdiction may take stronger action via inspections, citations, fines, and possibly revocation of the State ABC license or local use permit. Most alcohol outlet owners and operators agree to address problems if economic sanctions and revocation of their licenses and permits are at issue. Owners occasionally will take a "wait and see" attitude to test the community's willingness to enforce the zoning conditions, but most eventually comply or sell the establishment rather than face strong enforcement actions; only rarely are public agencies forced to proceed with license or permit revocations or fines.

Several communities are using CUPs to strengthen and enforce nuisance abatement laws. The nuisances referred to typically include littering, loitering, harassment of passers-by, public urination, open-air drug-dealing, and vice crimes such as prostitution and gambling. Some California communities now include special nuisance abatement provisions in CUP ordinances for alcohol outlets (Mosher and Works 1994; Wittman 1994). These ordinances facilitate prompt action by city officials once a formal determination has been made that an alcohol outlet is creating or abetting public nuisances. The active zoning ordinance's CUP conditions make it possible for zoning authorities to work directly with the outlet operator to eliminate nuisances by bringing the establishment into compliance with CUP conditions as an alternative to going to court. This approach often resolves the issues without resort to protracted and expensive legal proceedings. Under passive zoning, by contrast, the community would be required to file a court case in the first place to address the alleged nuisances (Kiley 1998).

Regulation of Alcohol at Public Places and Events

Although the 21st Amendment to the U.S. Constitution gives States the power to regulate the sale of alcoholic beverages in public places and at public events, States often exert this control only through the minimal standard requirements they apply to one-day or other short-term licenses. What's more, States generally leave the regulation of drinking in public places to local jurisdictions. These State policies leave considerable room for local public agencies to interpret and implement specific policies and ordinances of their own. Such latitude gives communities the opportunity to prevent problems related to alcohol sales during casual uses of public spaces and at special events such as Fourth of July fireworks displays, Saturday-night dances, high-school graduations, and ethnic group celebrations.

Substantial legal liability may rest with the local jurisdiction that owns or controls facilities used by the general public, whether or not alcohol is sold there. Therefore, unless reasonable measures are taken to prevent problems the use of alcohol on public property may cause, the local jurisdiction may be held liable in the event of an alcohol-related injury (Solomon et al. 1996). Public properties subject to such liability claims include the following:

• Common areas, such as plazas, squares, and courtyards.
• Facilities designed for public gatherings, such as fairgrounds, auditoriums, and stadiums.
• Parks, beaches, playgrounds, and natural open spaces.
• Streets, sidewalks, alleys, and transportation facilities.

Community policies and restrictions on drinking in public continue to spark local debate. Should the community prohibit all drinking in public? If public drinking is permitted, what constraints should be imposed to prevent problems? Traditionally, local governments have banned drinking in public to control nuisance inebriates. More recently, restrictions on drinking at public beaches and in automobiles, parks, and other recreational settings have been implemented as part of broader public safety policies that put community policing concepts to work against crime and violence at the neighborhood or district level (Edwards et al. 1994; Goldstein 1990).

Very few studies have been done on local ordinances that apply to drinking in public. Thus debate continues over whether to allow drinking in open public spaces such as parks, sports stadiums, concert halls, and arts centers. Some localities are finding that under certain circumstances controlled drinking in such places is easier to enforce and results in fewer problems than complete bans on drinking. However, local authorities may well prefer to ban all drinking at hard-to-control events such as long parades, large street fairs, and mass open-air events targeted at young people (Wittman 1997a).

For controlled drinking to be allowed, most communities require a special permit stipulating preventive measures such as the following:

• Advance public notification that preventive measures will be in effect.
• Controlled access through a supervised entrance into a fenced enclosure where alcohol is sold and consumed.
• Facility designs and operational procedures that deny underage individuals access to alcoholic beverages and that minimize risks to event attendees, for example by providing alcohol-free "family areas" at sports stadiums and allowing alcohol sales only in stadium concourses, not in the seats.
• Limits on advertising and promotion of alcohol availability in conjunction with the event.
• Prohibiting the public from bringing alcohol onto the premises and cutting off sales of alcoholic beverages at least an hour before the end of an event to avoid loosing inebriated drivers onto the streets.
• Special arrangements for security and enforcement of alcohol restrictions.

Responsible Beverage Service Standards

Many communities' initial experiences with alcohol control policies have led local officials to adopt responsible beverage service standards for those who sell or serve alcoholic beverages at public events. RBS — a general term covering a range of preventive policies and practices for the sale and service of alcohol in retail establishments — is a powerful tool for reducing alcohol-related problems at the local level. Examples of RBS policies and programs are detailed in chapter 3.

One aspect of RBS programs is training servers in responsible beverage service practices. Another involves working with alcohol outlet owners and managers to establish prevention policies and practices that support the servers and minimize excessive or other problematic drinking. The principal goals of RBS training programs are to prevent patrons from becoming intoxicated, to ensure that intoxicated patrons are not served further or exposed to harm, and to prevent the sale of alcohol to minors. Some cities and counties include RBS training as a standard requirement for local planning and zoning ordinances, for short-term event permits, and for special events hosted or sponsored by community agencies (Wittman and Harding 1993).

RBS programs are becoming increasingly available at the local level. Some State ABC's and a number of private organizations now offer training in and consultation on RBS practices and policies. The California State ABC, for example, provides Licensee Education on Alcohol and Drugs training for all ABC licensees, while the State of Oregon requires RBS training for all commercial alcoholic beverage servers. Furthermore, California State-level organizations are being created to certify training and site-management services based on RBS standards (California Department of Alcoholic Beverage Control 1994a; Institute for the Study of Social Change 1995; Wittman et al. 1991). The not-for-profit California Coordinating Council on Responsible Beverage Service certifies or recertifies about 15 RBS training programs in the State each year (California Coordinating Council on Responsible Beverage Service 1995). The implementation and effectiveness of some of these RBS programs — particularly those produced by federally funded research projects — have been formally evaluated; see chapter 3 for details.

Local Public Agencies as Models for Alcohol Prevention Policies

Community organizations may look to their local public agencies not only as sources of information but as models for alcohol availability policies at social events. Similarly, official personnel policies, such as limits on employee drinking at corporate events, can offer private organizations a model for planning and instituting their own workplace drinking policies. Many cities and counties have preventive policies concerning drinking at department-sponsored gatherings such as holiday parties, receptions, and other official functions. These policies contribute to the quality of the events and reduce the potential for unpleasant incidents that may lead to lawsuits or other problems.

Public agencies also can take the lead in securing sponsorship for prevention campaigns and programs that include local organizations and community groups. Anti-drinking-and-driving holiday initiatives, sober-graduation campaigns, and summer recreation programs all require working partnerships among public agencies, local businesses, and voluntary and other community organizations. Support and endorsements from public officials help such programs attain greater visibility, wider participation, and positive results.

Local authorities can also take steps to make sure their jurisdictions' prevention policies on alcohol availability are consistent across all public agencies, widely disseminated to the public, and correctly and fully applied. Consistent policies engender a supportive climate that can resonate throughout the broader community. Such consistency makes for more effective action on alcohol availability problems, which requires coordination among the State ABC; the district attorney's office; and the local police, land-use planning, and community development departments, among others. It takes both cooperation and coordination across a community to facilitate the granting of permits, the distribution and operation of new and transferred alcohol outlets, and the enforcement of local regulations.

Negotiating Agreements With Alcohol Outlet Operators

Problems involving alcohol outlets often can be resolved by direct negotiations between the outlet operator and the affected parties. State ABC agencies, local authorities, neighbors, or competitors can meet with the outlet operator outside the formal hearing process to establish conditions acceptable to all. When good will is present on all sides and there are no insurmountable complications, this approach can succeed. For example, a Los Angeles police officer whose family owned a convenience store in a city in northern San Diego County took immediate action upon hearing from the store's neighbors that it was a major source of calls for police service; the officer promptly replaced the store manager, who had allowed loitering and drinking on and near the premises, and conditions around the store improved dramatically (F. D. Wittman, personal communication, 1994).

Voluntary negotiations between an outlet operator and the establishment's neighbors can be conducted successfully without the presence of State or local officials, although it often helps to include them from the outset. In any case, if initial negotiations fail, the next step is to call in State ABC and local police and planning officials. They can convey to the operator the importance of responding cooperatively to problems as well as the possible consequences of not doing so under the State ABC license and local use permit.

A distinction should be made, however, between voluntary agreements negotiated for a single retail alcohol outlet and those reached with several outlets operating under a single owner or strong trade association, or with several independently owned and operated outlets. Negotiating voluntary agreements with several outlets is far harder than reaching an accord with just one, as independent owners have no way of guaranteeing one another's voluntary compliance or of ensuring that troublesome operators will participate in an agreement.

Efforts to reduce public health problems through voluntary agreements with alcohol outlet operators generally meet with limited success when several outlets in an area have a financial interest in conducting business practices that are detrimental to the public health. For example, convenience stores in a high-risk area may refuse to stop selling inexpensive, high-proof alcohol to public inebriates. Even when agreements are reached in such cases, they are often short-lived; as soon as one outlet stops honoring the voluntary agreement, the rest will likely abandon it as well (Institute for the Study of Social Change 1995).

Preventing Access to Alcohol by Minors

Mounting evidence shows that there is a clear need for communities to increase their efforts to enact and enforce laws prohibiting alcohol sales to minors and to prevent minors from even trying to buy alcohol. Since 1984, every State has had a minimum drinking age of 21, and a number of studies indicate that the increased age limit has reduced the number of alcohol-related traffic crashes among the affected age groups (Wagenaar 1986). The increase in the minimum drinking age was also shown to reduce alcohol consumption among high school seniors — lower rates of drinking that appeared to persist as these youth entered their 20's (O'Malley and Wagenaar 1991).

Research also shows, however, that minors can easily obtain alcohol from friends or family members as well as through shoplifting and direct purchases from retailers (Wagenaar et al. 1992). Nevertheless, outlet operators are rarely punished for selling alcohol to minors, and State ABC offices lack the staff for adequate enforcement (Wagenaar and Wolfson 1994). Although research on minors' access to alcohol is in the early stages, studies of minors' access to tobacco products have resulted in numerous community-based efforts to prevent the sale of tobacco to minors. Such efforts have included merchant education and support for merchant compliance with the laws (Biglan et al. 1994; Keay et al. 1993; Skretny et al. 1990) — methods that have potential for preventing sales of alcohol to minors as well.

One strategy State ABC's and local police are using to prevent youth access to alcohol and tobacco is the "sting" or decoy-buy operation. In accordance with a procedure approved by the State ABC and administered by local law enforcement agencies, decoy-buys send minors to outlets to attempt to buy alcoholic beverages. Clerks who sell to minors are fined and the violation is recorded on the outlet owner's license. Repeat offenses can result in increasing sanctions, leading to revocation of the ABC license or local use permit. Periodic decoy-buy operations (three or more per year) have resulted in decreased sales of alcoholic beverages to decoy buyers (California Department of Alcoholic Beverage Control 1996).

Prevention Through Architectural Design and Property Management

Certain architectural designs and property management procedures lend themselves to crime prevention through environmental design, which can prevent specific problems associated with retail alcohol outlets, such as crime, loitering, shoplifting, public intoxication, vehicular traffic, noise, bright lights, and other neighborhood disturbances. One study showed that several design features and operating practices in convenience stores affected their likelihood of being robbed. Such features and practices include secure cash-handling methods using in-store safes, open architectural designs that permit easy surveillance by police and store personnel, and following a policy of having at least two clerks in the store at night. In Florida, legislation requiring the adoption of such practices is credited with a 10-percent decrease in convenience store robberies over a period when violent crime elsewhere increased by 10 percent (Hunter and Jeffrey 1992). Crime prevention through environmental design techniques can be implemented by owners or operators of alcohol outlets either independently or in consultation with local officials, such as from the planning and police departments.

State ABC's Sharing Authority With Local Jurisdictions

Several natural experiments that arose spontaneously in the course of policy development have strengthened the bonds between State ABC's and local alcohol control systems by increasing the authority or capacity of local jurisdictions to assist State ABC licensing and enforcement. Although these experiments were neither planned nor thoroughly documented, their results offer important lessons for future initiatives:

• A California State Assembly member convened a panel to draft a model zoning ordinance to strengthen local control over alcohol outlets (Model Zoning Ordinance 1987). Many California cities and counties have used this model ordinance when drafting local zoning regulations. The use of planning and zoning ordinances has strengthened the connections between communities and the ABC regarding licensing and enforcement.
• Although local jurisdictions generally have the authority to enforce State ABC laws, many local law enforcement agencies lack training in administering the laws. To rectify this situation, the California Department of Alcoholic Beverage Control now awards $100,000 development grants to local police departments to improve their alcohol licensing and enforcement activities (California Department of Alcoholic Beverage Control 1994b).
• Because decoy-buy operations have been demonstrated to have the short-term effect of suppressing sales of alcohol to minors, State ABC's and local police departments can work jointly to conduct several such operations each year. The California State ABC Department has developed a manual for statewide use that codifies decoy-buy procedures for local police and sheriff's departments.
• The 1994 Nebraska Liquor Control Act gave local governing bodies several specific powers and duties, including the authority to cancel or revoke alcohol licenses for cause or for violations of the act. (Any cancellation or revocation action is subject to appeal to the State Liquor Control Commission.) The act also empowered local governments to make nonbinding recommendations to the Commission, which has the authority to approve or deny applications to serve or sell alcoholic beverages.
• Some State ABC's, such as the California Alcoholic Beverage Control Department and the Pennsylvania Liquor Control Board, have instituted programs to educate local public agencies and school-age youth about alcohol availability-related problems and ABC operations.

Support From Other State and Federal Agencies

Some State alcohol and drug control agencies are encouraging municipalities and other local jurisdictions to become more active in addressing community problems related to alcohol availability (Casswell et al. 1989; Greenfield and Zimmerman 1993). That said, there is wide variation in the extent to which different State agencies support local efforts to improve controls on alcohol availability. Some, like the California Department of Alcohol and Drug Programs, provide technical assistance, information dissemination, statewide conferences, and consultations on county initiatives in environmental prevention (Wittman 1992, 1997b; Wittman and Biderman 1993).

Federal demonstration grants and community trial projects have generated significant support for the development of local controls. From 1990 to 1995, the National Institute on Alcohol Abuse and Alcoholism funded projects for community prevention in several cities in California, Minnesota, and South Carolina, which were systematically evaluated (Addiction June 1997). These projects document problems related to retail alcohol availability and report on community-level strategies for addressing those problems. The U.S. Department of Health and Human Services' Center for Substance Abuse Prevention (CSAP) includes "environmental" and "community-based" approaches to alcohol and drug problem prevention among the six types of local prevention activities that it funds (Center for Substance Abuse Prevention, Federal Register March 31, 1993).

Private foundations have also played a key role in supporting community-level prevention initiatives. The Robert Wood Johnson Foundation, for instance, provides extensive support through "Fighting Back" Partnership programs (administered by Join Together, Boston University School of Public Health) in several cities. Other programs use environmental approaches to prevention and focus on the use of zoning ordinances, RBS programs, and other strategies for working directly with retail alcohol outlets. Foundations also support grassroots efforts to protest alcohol availability in high-risk inner-city and Native American communities (Aguirre-Molina 1991; Mosher and Works 1994).

Knowledge Dissemination

State, Federal, and foundation sources all support knowledge dissemination activities such as this guide. CSAP also supports various information dissemination activities through the National Clearinghouse for Alcohol and Drug Information, the National Center for the Advancement of Prevention, and the regional Centers for the Application of Prevention Technology. These organizations offer conferences, seminars, technical assistance, and publications that include research summaries and information on best practices. The Robert Wood Johnson Foundation's Join Together program also distributes a range of pertinent materials on community-level approaches to prevention.

Limited Use of Taxation and Fees

The debate over taxation continues in every State from one decade to the next: Should alcoholic beverages be taxed at higher rates? Should taxation authority be extended to the local level? Such debates may intensify as research continues to confirm that higher prices for alcoholic beverages are associated with decreases in public health and safety problems (Edwards et al. 1994).

To date, local taxes on alcohol have rarely been imposed because States are generally reluctant to grant such authority to local governments. However, local agencies are typically allowed to charge permit-filing and business-license fees to cover the jurisdiction's full administrative and enforcement costs related to alcohol outlets. Some communities have begun charging nonrefundable permit-filing fees that range from a few hundred to a few thousand dollars and reflect the full cost of processing and enforcing applications for zoning permits (Wittman, Helmke, and Biderman 1992).

The ongoing debate at Federal and State levels over raising taxes on alcoholic beverages affects local prevention efforts in several ways. Higher excise taxes and State fees on alcohol could be used to increase staffing levels and otherwise strengthen the capacity of ABC and other State agencies to assist local communities. Higher taxes and fees would also reduce alcohol consumption and related problems (Edwards et al. 1994). Many cities and counties would probably pass legislation to increase local taxes on alcoholic beverages if permitted to do so by the State.

Constraints on Staffing Resources

The ratios of State ABC staff to the number of retail alcohol outlets, which have declined steadily for decades, are now below the levels originally set for some ABC agencies. Some States also continue to pursue the privatization of alcohol outlets, doing relatively little to restrain their proliferation (see chapter 3). ABC's in many States face such backlogs in satisfying day-to-day licensing demands that they must confine their enforcement efforts to only the most troublesome cases.

Because the resources of State-based ABC's are limited, comparatively little staff time is spent on prevention initiatives and routine surveillance (Gruenewald, Madden, and Janes 1992). The current regulatory process tends to be case-oriented and reactive, triggered by requests from retailers seeking licenses and driven by complaints about existing outlets filed by local officials and neighbors (Wittman 1982).

Operating an ABC on a case-by-case basis is time-consuming and expensive. ABC's must operate according to State law when processing license applications. Case reviews to deny a new license or to sanction an existing establishment require a considerable investment of staff time and expertise. Licenses cannot be denied unless a background investigation shows that an applicant is unfit or that neighbors or public agencies can demonstrate a well-documented threat to public health and safety. ABC's will not deny or sanction licenses simply because of community disapproval of drinking; to influence licensing decisions, protesting local agencies or community groups must demonstrate that the applicant is likely to create specific problems (e.g., that children at play might be endangered by inebriated drivers leaving a neighborhood bar). It is not uncommon for licensing and enforcement actions to take months or even years to complete.

Lack of Community Data on Problems Related to Alcohol Outlets

State agencies do not routinely report data about community alcohol problems in relation to outlet types or geographic areas. Similarly, local police generally do not file regular reports on alcohol-law violations by outlet type or police beat. The State ABC and most local police departments do, however, provide detailed information on police complaints and other problems associated with specific alcohol outlets on an as-needed basis in individual cases. But this information is often provided after the fact, in connection with enforcement activities following numerous complaints about the outlet. Because the information is not collected consistently and in a timely fashion, it is of little value in monitoring alcohol outlet control policies or in planning long-range prevention strategies.

Likewise, the State agencies responsible for assessing and collecting taxes generally do not report alcohol sales separately by retail outlets. This makes it difficult for researchers to include consumption variables in studies of the relationships between alcohol outlets and alcohol- related problems. States could readily report information about alcohol sales volumes and dollar amounts if they chose to do so.

Researchers and information specialists have assembled much information illuminating the underlying nexus of relationships among types of alcohol outlets, geographic locations, and patterns of alcohol-related problems. Until recently, special expertise at high cost was required to manage these data, which typically are collected for a limited time and for the limited purposes of a specific research project. Findings based on this information are useful to local communities and State ABC's for general prevention planning, as described in chapter 3, but this information is not immediately useful for managing alcohol outlets at the community level. In fact, State ABC's and local agencies lack the capacity to collect and analyze such data on their own.

Recent advances in the development of geographic information systems (GIS) and computer software for their use have made it possible to describe alcohol-related problems in relation to retail alcohol outlets at the community level. State ABC's and local police departments can now routinely observe public health and safety problems related to alcohol outlets in the environmental contexts of time, location, and type of setting, including all types of on- and off-sale outlets according to State licensing nomenclature and local land use categories (Harding and Wittman 1995). Following pilot tests in several California cities, a local GIS-format Alcohol/Drug Sensitive Information Planning System has been developed to provide such information based on continuous reporting of alcohol and drug involvement in all calls to the police for service (Harding and Wittman 1995). This system provides local information, by time, location, and type of setting, about emerging "hot spots" and other problem trends associated with alcohol outlets in. It provides the information in a format useful to local planners for reducing alcohol-related problems by pinpointing the settings and circumstances where problems are most likely to occur.

Conceptual Basis for Approach 1

Establishing and enforcing minimum drinking age laws that prohibit alcohol sales to, and the purchase of alcohol by, underage youth tend to reduce alcohol use and related problems. Although all States have already raised the legal drinking age to 21, evidence demonstrating the effectiveness of a high minimum drinking age is pertinent because it not only underscores the importance of enforcement efforts but offers valuable points in the debate over retaining such laws.

Overview of the Research and Practice Evidence on Approach 1

The analysis of the effectiveness of efforts to prevent the availability of alcohol to underage youth is based on 11 research studies and 3 prevention practice cases.

Several studies examined the effects of different State minimum drinking age laws on drinking behaviors and alcohol-related problems. Differences among the States' minimum legal drinking ages were eliminated with the passage of the uniform national minimum age limit in 1984.

• O'Malley and Wagenaar (1991) compared drinking behaviors and the number of alcohol-related traffic crashes in States with high minimum legal drinking ages with those in States with low minimum legal drinking ages. They also examined changes in drinking behavior and the number of traffic crashes in States that raised their minimum legal drinking age to 21 years with those in States that already had set the minimum at 21.
• Similarly, Wagenaar (1986) examined the effects of raising the minimum legal drinking age in Michigan on the number of injury-producing traffic crashes among drivers between the ages of 18 and 20.
• Wagenaar and Maybee (1986) examined the effects on traffic-crash rates of a 1981 Texas State law raising the minimum legal drinking age from 18 to 19 years.
• Mooney and Gramling (1993) compared a group of college students where the minimum legal drinking age was 21 with another group where the minimum was 18. Patterns of drinking behavior were compared in terms of frequency, quantity, and location of consumption.
• In 1987, the U.S. General Accounting Office (GAO) applied evaluation synthesis methodology to then-available research studies that analyzed the effects of raising the minimum drinking age laws on traffic crashes and alcohol consumption among those affected by the laws.

Two studies examined the effects of establishing different limits on blood alcohol concentration (BAC) rates for adults and youth.

• Hingson, Heeren, and Winter (1994) compared the rate of alcohol-related traffic crashes in 12 States where the BAC limit was lower for young drivers than for adults with those in 12 comparison States where the BAC limit was the same for youths and adults.
• The National Highway Traffic Safety Administration (1992) examined the effectiveness of a Maryland State law that prohibits driving by those younger than 21 who have a BAC of 0.02 percent or more and of a public information campaign designed to promote the law's effects.

Two additional studies examined alcohol purchase attempts by underage youth.

• Preusser, Williams, and Weinstein (1994) conducted underage alcohol purchase attempts on a subset of all alcohol outlets in Denver, Colorado.
• Preusser et al. (1995) compared the behaviors of underage youth attempting to purchase alcohol in two States that have substantially different laws regarding the purchase, possession, and consumption of alcohol and the use of false identification by underage youth.

One research study and three practice cases examined prevention programs.

• Hingson et al. (1996) evaluated the effectiveness of the Saving Lives Program, a community initiative that organized multiple city departments and private citizens to reduce alcohol-impaired driving, alcohol-related driving risks, and traffic deaths and injuries.
• The Pennsylvania Liquor Control Board conducted activities designed to control the density of alcohol outlets, support the prevention of alcohol abuse, disseminate information on the responsible use of alcohol and responsible alcohol sales and service, and establish partnerships with other State agencies and organizations.
• The Under 21 Enforcement Project of the Town of Yorktown Police Department was designed to enforce regulations prohibiting the sale of alcohol to persons under 21 years of age, to heighten public awareness of alcohol abuse by youth, to aggressively enforce DWI regulations for underage youth, and to increase the accuracy of police reports concerning traffic crashes.
• Cops in Shops was an enforcement and media campaign designed to reduce alcohol purchases by underage youth in Las Cruces, New Mexico. The campaign involved police officers posing as employees of licensed establishments.

More detailed abstracts of the above studies appear at the end of this chapter (pp. 60-67).

Lessons Learned From the Evidence Reviewed for Approach 1

The following are the lessons learned from the research and practice evidence reviewed in this section:

• Laws restricting alcohol sales to underage youth are often ignored by both merchants and underage youth (Mooney and Gramling 1993; Preusser, Williams, and Weinstein 1994; Preusser et al. 1995). Most Americans who drink alcohol begin in their early adolescent years.
• Enforcement of minimum drinking age laws diminishes but does not eliminate underage drinking (Hingson, Heeren, and Winter 1994; Mooney and Gramling 1993; National Highway Traffic Safety Administration 1992; O'Malley and Wagenaar 1991; Preusser, Williams, and Weinstein 1994; U.S. General Accounting Office 1987; Wagenaar 1986; Wagenaar and Maybee 1986; Cops in Shops; Under 21 Enforcement Project).
• Enforcement efforts with the greatest public health impact are those that aim to reduce the harm resulting from drinking as well as to reduce drinking itself, such as efforts to counteract the link between drinking and driving (Hingson, Heeren, and Winter 1994; Hingson et al. 1996; National Highway Traffic Safety Administration 1992; O'Malley and Wagenaar 1991; Wagenaar 1986; Wagenaar and Maybee 1986; Pennsylvania Liquor Control Board; Under 21 Enforcement Project).

Suggestions for Future Research on Approach 1

The expert panel made the following suggestions for future research on preventing availability to underage youth:

• Studies of tobacco purchases by adolescents show that when one avenue of access is closed, others are devised. Similar research is needed on sources of alcohol other than direct purchase from retail outlets and how such they may be blocked. Such sources include family members, other adults, "shoulder tapping" of strangers at retail outlets, and shoplifting (Preusser et al. 1995; Cops in Shops).
• Research needs to be conducted on the role of alcohol in teenage subcultures and social life, particularly as a rite of passage in various ethnic, cultural, and regional groups of underage youth. These aspects of social life include team bonding, sexuality, and partying.
• There is a need for research on the nature, extent, mechanisms, and effects of the advertising and marketing of alcohol to underage youth.
• Additional research is called for on ways to enforce minimum drinking age laws more effectively and to prevent underage youth from buying alcohol. These research efforts should examine clerks, outlets, and adolescent characteristics associated with successful and failed purchase attempts (Wagenaar et al. 1992).
• Some universities maintain licensed alcohol sales and service on campus, while others have reduced alcohol availability through community coalitions that include university administrators, students, and local businesses. Other university communities have keg registration laws. Research is needed to examine the enforcement and effectiveness of these laws and policies on underage youth access to alcohol, especially in college areas (Hingson et al. 1996; Mooney and Gramling 1993; Preusser et al. 1995).
• Research needs to be conducted on potential unintended and untoward effects of the minimum legal drinking age. For example, although the law has contributed to important outcomes such as a decrease in traffic casualties, it may foster a pattern of mildly illicit behavior, such as learning to drink in illegal circumstances. Thus, research should examine whether long-term problems, especially those related to the acceptability of illicit alcohol use by underage youth, are associated with a minimum legal drinking age of 21 years (Hingson, Heeren, and Winter 1994; National Highway Traffic Safety Administration 1992; O'Malley and Wagenaar 1991; Wagenaar 1986).
• Some areas, such as the Province of Ontario, Canada, have adopted a graduated driver's licensing approach, which typically requires a BAC of 0 percent for novices in their first one or two years of driving. Research should be conducted to examine whether the graduated licensing approach prevents traffic casualties related to youth drinking, as does the minimum drinking age law (Hingson, Heeren, and Winter 1994; National Highway Traffic Safety Administration 1992).
• Students Against Drunk Driving (SADD) and other organizations promote activities such as assigning designated drivers and having "no-questions-asked" parental transportation policies as alternatives to drinking and driving. These activities may reduce incidents of youth drinking and driving, but research is needed to determine how widespread these strategies are and the extent to which they accomplish their objectives (Hingson et al. 1996).

Conceptual Basis for Approach 2

Because alcohol is a discretionary purchase, especially for social drinkers, demand for it is affected by price fluctuations. An increase in the price as well as in the taxes levied on alcohol should therefore effectively depress its consumption. This should be especially true among underage youth, who generally have relatively low disposable funds.

Overview of the Research and Practice Evidence on Approach 2

The analysis of the effects of raising prices and taxes on alcohol sales and consumption and alcohol-related problems is based on 10 studies.

Four of these studies were econometric analyses.

• Grossman et al. (1987) and Coate and Grossman (1985) developed estimates of the responsiveness of alcohol use by youth aged 16 to 21 years to variations in the price of alcohol.
• Levy and Sheflin (1983) examined the importance of cross-price effects and provided estimates of price elasticity based on demand.
• Similarly, Wette et al. (1993) examined the relationship between price and the consumption of beer, wine, and spirits to assess changes in alcohol consumption in New Zealand as a function of increases in the price of alcoholic beverages.
• Ornstein and Hanssens (1985) estimated the impact of a variety of control measures on the demand for distilled spirits and beer.

Two studies examined the effects of alcohol taxes on alcohol-related problems.

• Cook and Moore (1993) analyzed the effect of beer excise taxes on rates of homicide, rape, assault, and robbery in the 48 contiguous United States.
• Saffer and Grossman (1987) examined the relationships of beer taxes and the minimum legal drinking age to traffic fatalities among youth aged 15 to 24 in the 48 contiguous United States from 1975 through 1981.

The common goal of the seven studies outlined above was to examine the effects of State alcohol taxes and retail alcohol costs on alcohol consumption and its impact on behavior as well as physical and social health.

Three studies examined different aspects of "happy-hour" discount drink policies.

• Babor et al. (1978) examined the effect on drinking behavior of a controlled setting for a happy-hour discount drink policy.
• Kohn, Smart, and Adlaf (1985) surveyed the prevalence, content, and timing of happy hours in metropolitan Toronto, Canada, during the summer of 1983.
• Smart and Adlaf (1986) analyzed the impact of a ban on happy-hour discount drink promotions on drinking behavior, overall alcohol sales, and impaired-driving charges in Ontario, Canada.

More detailed abstracts of these studies appear at the end of this chapter (pp. 67-71). The research evidence reviewed indicates the following:

Lessons Learned From the Evidence Reviewed for Approach 2

The following are the lessons learned from the research and practice evidence reviewed in this section:

• Increases in beer taxes are associated with decreases in beer consumption (Cook and Moore 1993; Saffer and Grossman 1987), violent crime (Cook and Moore 1993), and deaths from traffic crashes (Saffer and Grossman 1987).
• Increases in the price of beer are associated with decreases in beer consumption (Coate and Grossman 1985; Levy and Sheflin 1983; Wette et al. 1993).
• Increases in the minimum legal drinking age are associated with decreases in beer consumption (Coate and Grossman 1985; Grossman, Coate, and Arluck 1987; Ornstein and Hanssens 1985) and decreases in death rates from traffic crashes (Saffer and Grossman 1987).

Suggestions for Future Research on Approach 2

• Additional research is needed on the short- and long-term effects of alcohol taxation, with particular attention to population subgroups such as heavy alcohol consumers (Coate and Grossman 1985; Cook and Moore 1993; Saffer and Grossman 1987).
• Some research suggests that alcohol tax increases promote reductions in drinking in approximately equivalent proportions among heavy and light drinkers (Coate and Grossman 1985). Although State tax increases may result in a moderate decrease in alcohol consumption, it seems plausible that other prevention activities may be required to sustain such decreases. Research is required to examine which other prevention activities can boost as well as sustain those decreases over time.

Conceptual Basis for Approach 3

Research shows that the behavior of alcohol servers and the policies of drinking establishments can have an impact on the behavior of their patrons. For example, servers can engage in activities that encourage or increase heavy drinking, allow heavy drinking to be ignored, promote intoxication, and foster problems associated with intoxication. Server interventions, such as server and management training, are a broad set of strategies designed to reduce the risk that patrons will become intoxicated and that intoxicated persons will harm themselves or others. Interventions can include the review and modification of management policies and operations. The combined efforts of server intervention and management policies is referred to as responsible beverage service. RBS should create safer drinking environments and reduce the likelihood of intoxication and related problems.

Overview of the Research and Practice Evidence on Approach 3

The analysis of the effectiveness of RBS is based on eight research studies.

• Gliksman et al. (1993) evaluated the knowledge, attitudes, and behaviors of alcohol providers at four types of bars (roadhouses, hotel lounges, neighborhood taverns, and skid-row bars) immediately before and after participation in a server training program.
• Similarly, McKnight (1991) evaluated the effect of server intervention education and the effect of various situational variables on program effectiveness.
• McKnight and Streff (1994) examined the effects of enforcing the law prohibiting alcohol service to intoxicated patrons on service to patrons, DWI rates, and enforcement costs and benefits.
• Saltz (1987) compared patrons' consumption of alcoholic beverages in a U.S. Navy enlisted club that had a server intervention program with that in a similar Navy enlisted club that did not have such a program.
• Holder and Wagenaar (1994) evaluated the effects of the State of Oregon's mandatory Oregon Server Training Program on the number of single-vehicle nighttime crashes in which alcohol was involved that resulted in injuries or fatalities.
• Similarly, Mosher et al. (1989) examined the efficacy of an intensive community-based RBS training program intended to reduce intoxication rates among patrons of participating establishments. They also conducted formative evaluations to determine community and industry support for an RBS training program.
• Russ and Geller (1987) evaluated the impact of an alcohol server intervention program on the type and frequency of server interventions, the number of drinks consumed by pseudopatrons, and the BAC's of exiting pseudopatrons who drank three alcoholic beverages per hour for 2 consecutive hours.
• Wagenaar and Holder (1991a) examined the effects of a sudden change in a Texas law that resulted in exposure to legal liability for alcohol servers in relation to the frequency of single-vehicle, injury- producing traffic crashes.

More detailed abstracts for these studies appear at the end of this chapter (pp. 71-74).

Lessons Learned From the Evidence Reviewed for Approach 3

The following are lessons learned from the research and practice evidence reviewed in this section.

• Server training programs differ in type, intensity, length, and focus (McKnight 1991; Mosher et al. 1989; Russ and Geller 1987; Saltz 1987). There is no evidence that certain server training program characteristics are associated with greater or lesser effectiveness. Evidence does, however, suggest that server training programs are more likely to exist in situations where there is support, organization, and interest among stakeholders (Mosher et al. 1989).
• Server training and responsible beverage service programs can be established through the efforts of States, counties, and other jurisdictions (McKnight 1991; Mosher et al. 1989; Russ and Geller 1987; Saltz 1987).
• Responsible beverage service is most likely to be successful when servers and managers know that the law will be enforced or realize that they assume significant liability when they serve intoxicated or underage individuals (Holder et al. 1993; McKnight and Streff 1994; Wagenaar and Holder 1991a).

Suggestions for Future Research on Approach 3

The expert panel made the following suggestions for future research on responsible beverage service:

• Because server training programs with varying characteristics seem to be yielding similar results, there is a need to evaluate the specific components that constitute the essentials of an effective server training program, as well as the settings and conditions that enhance the effectiveness of these programs (Holder et al. 1993; McKnight 1991; Mosher et al. 1989; Saltz 1987).
• Research is needed to determine whether server training programs result in long-term changes in patterns of patron intoxication and related problems such as traffic crashes and DWI arrests (Holder and Wagenaar 1994; McKnight and Streff 1994; Mosher et al. 1989; Russ and Geller 1987; Saltz 1987; Wagenaar and Holder 1991a).
• So far, most research has focused on training servers in drinking establishments. There is a need for research on effective interventions and approaches for retail clerks and others engaged in off-sale alcohol distribution (Holder and Wagenaar 1994; McKnight 1991; McKnight and Streff 1994; Mosher et al. 1989; Russ and Geller 1987; Saltz 1987).

Conceptual Basis for Approach 4

Research suggests that there is an association between alcohol availability, such as outlet density, and alcohol-related social, civic, and health problems. Decreasing alcohol availability should therefore reduce alcohol-related problems such as homicide, suicide, and other types of violence; cirrhosis and other health problems; and traffic crashes.

In addition to availability through permanent outlets, a particular area of concern is alcohol availability at special events and locations. Alcohol is often brought to or purchased and consumed at special occasions such as professional sporting events. Similarly, alcohol is often purchased and consumed in municipally owned facilities such as city parks. Laws established to regulate the sale, purchase, and consumption of alcoholic beverages should affect access, consumption, and alcohol-related behaviors during and after special events and in public places.

Overview of the Research and Practice Evidence on Approach 4

The research studies reviewed on changing the conditions of availability did not evaluate researcher-directed interventions. Rather, they examined outcomes of natural experiments, particularly changes in legislation regarding alcohol sales restrictions and regional differences in laws.

The analysis of the effects of changing the general conditions of alcohol availability is based on nine studies and one prevention practice case. Several studies examined the effects of privatizing Iowa's State retail monopoly on alcohol sales.

• The effects of a natural experiment involving change in a California State law to remove restrictions on sales of beer and wine was studied on two university campuses by Fillmore and Wittman (1982). They assessed its impact on the quantity of alcoholic beverages sold near the campuses and on student drinking.
• Fitzgerald and Mulford (1992, 1993) evaluated whether a sudden and dramatic increase in the availability of wine and distilled spirits in Iowa resulted in an increase in alcohol consumption and alcohol-related problems among individuals aged 18 years and older.
• Using data from 38 States over 12 years, Gruenewald and Ponicki (1993, 1995) evaluated the impact of changes in alcohol sales and the physical availability of alcohol on fatalities from single-vehicle nighttime crashes.
• Holder and Wagenaar (1990) evaluated the effect on consumption of the privatization of Iowa's State retail monopoly on sales of distilled spirits for off-premise consumption.
• MacKinnon, Scribner, and Taft (1995) analyzed data on alcohol availability and problems in unincorporated areas and in 84 cities in Los Angeles County to assess the impact of alcohol availability on alcohol-related civic and health outcomes.
• In a replication of this study, Gruenewald et al. (1995) analyzed data on alcohol availability and alcohol-related traffic crashes across 102 areas of four communities in California.
• Mulford and Fitzgerald (1988) evaluated the effect of increasing the number of off-sale wine outlets in Iowa on the rates of alcohol purchases, consumption, heavy drinking, and problem drinking.
• Mulford, Ledolter, and Fitzgerald (1992) examined changes in monthly sales of wholesale wine and spirits in relation to the privatization of State-run wholesale wine and retail wine and spirits sales.
• Similarly, Wagenaar and Holder (1991b) evaluated whether the privatization of retail wine sales in Iowa and West Virginia affected wine consumption, beer and distilled spirits sales, and alcohol consumption in other States.
• Laxer et al. (1994) studied the impact of privatization of retail and wholesale distribution of alcohol in Alberta, Canada. They assessed the outcomes of this natural experiment on the availability and price of alcohol and on crimes against outlets.

The analysis of the effects of changing the specific conditions of alcohol availability at special events and locations is based on one study and four practice cases.

• West et al. (1993) studied drinking, intoxication, and verbally and physically abusive behavior at Ontario's Maple Leaf Gardens during Toronto Maple Leaf hockey games to examine the effects of legislation permitting alcohol sales at sports arenas.
• Similarly, Fisher and Single (1983) observed drinking and drug consumption, alcohol- and drug-related incidents, and crowd composition at Toronto Blue Jays baseball games following an Ontario provincial law permitting the sale of beer at sporting events.
• Bjor, Knutsson, and Kuhlhorn (1990) examined interventions related to the celebration of Midsummer's Eve, a popular event in Sweden that has traditionally been associated with alcohol consumption and sexuality. The interventions included an alcohol-rationing rule, prohibitions against congregating at campgrounds or in parking lots near the center of the city, and prohibitions against possessing alcohol or knives in public places.
• Gliksman et al. (1990) reported the implementation and effects on residents of a social control policy related to the use of alcohol on city-owned properties and facilities.
• Gliksman et al. (1995) surveyed recreation directors, facility managers, and other appropriate individuals in all Ontario cities, towns, and townships to evaluate the development of alcohol-related policies and their impact on problems and facility rentals.

More detailed abstracts of the above studies appear at the end of this chapter (pp. 74-82).

Lessons Learned From the Evidence Reviewed for Approach 4

The following are the lessons learned from the research and practice evidence reviewed for this section.

• Although the body of research is not very strong, there is sufficient evidence to suggest that increasing alcohol availability, such as through the number and density of outlets and hours of sale, can increase alcohol consumption levels (Fillmore and Wittman 1982; Holder and Wagenaar 1990; Mulford, Ledolter, and Fitzgerald 1992; Wagenaar and Holder 1991b, 1995) and rates of alcohol-related problems (Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995). Many of the studies specifically examining outlet density and its impact on alcohol-related problems were also ecological or correlational studies, which do not provide proof of causality (Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995).
• A wide range of restrictions can be placed on special events, including restrictions on operating hours, noise levels, general locations of events, locations of alcohol sales or places of consumption (e.g., beer gardens), alcohol advertising and corporate sponsorship, age of servers, quantity of sales, size of containers, and the condition of the patrons (Bjor, Knutsson, and Kuhlhorn 1990; Gliksman et al. 1990, 1995).
• There is conflicting evidence on whether restricting the sale of alcohol at professional sporting events reduces alcohol-related problems (Fisher and Single 1983; Gliksman et al. 1995; West et al. 1993). The conflicting findings may result largely from the fact that this evidence came from studies based on opportunistic initiatives, leaving much room for improvement in study design and execution as well as measures of outcomes.

Suggestions for Future Research on Approach 4

The expert panel made the following suggestions for future research on changing the conditions of availability:

• There is a need to better understand how different aspects of availability control affect specific patterns of drinking and alcohol-related problems (Fillmore and Wittman 1982; Gruenewald and Ponicki 1993, 1995; Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995).
• Although many cities and States have established requirements regarding the spacing between alcohol outlets and between outlets and churches and schools, the extensive search undertaken for this guide did not turn up any research or practice case studies that examined this phenomenon. The effectiveness of such strategies therefore needs to be evaluated.
• Although a substantial amount of research has been done on the effects of increasing alcohol availability, there is virtually none on the effectiveness of restricting outlet density to prevent alcohol-related problems (Fillmore and Wittman 1982; Gruenewald and Ponicki 1993, 1995; Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995).
• Crime statistics should be examined from neighborhoods that eliminated previous problem outlets and compared with statistics from neighborhoods that still have problem outlets (Fillmore and Wittman 1982; Gruenewald and Ponicki 1993, 1995; Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995).
• Studies of geographic density should be further developed to include economic factors related to beverage prices. To date, studies have focused either on the geographic aspects of outlet density or on the effects of price on alcohol purchases, consumption, or related problems (Fillmore and Wittman 1982; Gruenewald and Ponicki 1993, 1995; Gruenewald et al. 1995; MacKinnon, Scribner, and Taft 1995). Further studies are needed to determine the simultaneous as well as the separate effects of both these factors.
• More studies are needed to fill in gaps and resolve inconsistencies in the research on changing the conditions of alcohol availability (Ferris et al. 1994; Mulford and Fitzgerald 1988; Mulford, Ledolter, and Fitzgerald 1992; Wagenaar and Holder 1991b).
• Research is needed on the scope and types of problems experienced at events where alcohol is sold.
• A better understanding is needed of alcohol use and related problems at different types of special events, such as block parties, wedding receptions in municipal halls, sports events, and city celebrations after winning major sports championships such as the Super Bowl or the World Series.
• Research is needed on the effects of short-term use permits for special events.
• Further research is required to provide qualitative and quantitative analyses of tailgate parking-lot parties at sporting and popular music events, with a focus on associated alcohol-related problems.

Conceptual Basis for Approach 5

Research on the effects of changing the hours and days of alcohol sales is based on evaluations of natural experiments. These experiments start with government decisions to extend or decrease the hours or days of alcohol sales at specific sites such as bars or wine and spirits shops. Although these decisions are often unrelated to specific prevention goals, they provide an opportunity to examine the effects of changing sale hours and days on overall alcohol sales and consumption patterns and on related problems.

Overview of the Research and Practice Evidence on Approach 5

The analysis of the effects of changing the hours and days of alcohol sales is based on six research studies. These studies primarily evaluated outcomes that resulted from natural experiments, particularly changes in legislation regarding alcohol sales restrictions.

Four studies examined the effects of changing the hours of sales in taverns in Australia.

• Smith (1987) evaluated the effect on traffic crashes of a New South Wales, Australia, state law that allowed the sale of alcohol at taverns on Sunday from noon to 10:00 p.m.
• Smith (1988a) examined the effects on traffic crashes of a legislative change that introduced flexible tavern operating hours in the Australian state of Tasmania. The study was designed to evaluate whether later closing times (and thus later drinking hours) might decrease the number of crashes from 10:00 p.m. to midnight and increase the number from midnight to 6:00 a.m.
• In the Australian state of Victoria, a February 1966 regulation changed tavern closing times on Mondays through Saturdays from 6:00 p.m. to 10:00 p.m. Smith (1988b) analyzed data on all injury-producing traffic crashes in Victoria before and after the regulation allowing bar patrons to consume alcohol in taverns later than before.
• In the Australian city of Brisbane, a regulation allowed the sale of alcohol on Sundays from 11:00 a.m. to 1:00 p.m. and from 4:00 p.m. to 6:00 p.m. Smith (1988c) evaluated the effect of these Sunday alcohol sales on traffic crashes, including resulting casualties and reported property damage.

Two studies in Scandinavia examined the effects of closing retail alcohol stores.

• Nordlund (1985) examined the effects of forcing wine and spirits shops in six Norwegian towns to close on Saturdays by comparing the outcomes with those in six control towns that allowed Saturday sales. The study measured reports to police of and arrests for drunkenness, reports of incidents of violence, admissions to alcohol detoxification programs, and alcohol sales.
• In Sweden, all sales of export beer, wine, and spirits take place in special state-owned retail liquor stores. Olsson and Wikstrom (1982) examined whether Saturday closings of these stores would have an impact on total alcohol consumption, drunkenness, public order, and certain types of crime.

More detailed abstracts of the above case studies appear at the end of this chapter (pp. 82-84).

Lessons Learned From the Evidence Reviewed for Approach 5

The following are the lessons learned from the research and practice evidence reviewed for this section.

• Increasing alcohol availability by extending hours of sale tends to increase rates of alcohol consumption and related problems (Nordlund 1985; Olsson and Wikstrom 1982; Smith 1987, 1988a, b, c).
• Most of the research in this area reflects recent experiences with extending rather than reducing hours or days of alcohol sales (Smith 1987, 1988a, b, c). Moving in the other direction (reducing hours or days of sale) is politically difficult in an era when consumers demand convenience (Nordlund 1985; Olsson and Wikstrom 1982). Even though the Norwegian experiment clearly showed positive results from Saturday closings, the political will was lacking to continue or extend such closings (Nordlund 1985). Proposals to extend the hours or days of sale should be evaluated in light of the near-impossibility of reversing any such changes.

Suggestions for Future Research on Approach 5

The expert panel made the following suggestions for future research on changing the hours and days of alcohol sales:

• More studies are required on the impacts of late-night closing times, which put drivers and pedestrians on the street under the potential effects not only of an evening of drinking but also of fatigue. The interplay of drinking and fatigue on related casualties needs to be systematically studied (Smith 1987, 1988a, b).
• Research is needed regarding regulation changes that increase alcohol availability in a community, such as by allowing previously prohibited Sunday sales (Smith 1987, 1988a, b, c).
• In order to devise effective prevention strategies, research is needed on the characteristics of alcohol users, their preferred hours and days of alcohol consumption, and the rates and types of alcohol-related problems they generate.
• Where on-premise drinking is a major social pattern, violence and trouble from drinkers outside alcohol outlets have sometimes been attributed to the fact that large numbers of drinkers are on the street at the same fixed closing time. Experiments have been conducted in the Netherlands and Scotland on staggered closing times for different drinking establishments; such research may be valuable in the United States as well.

Conceptual Basis for Approach 6

Many alcohol-related problems are local in nature; many prevention activities, therefore, are best conducted at the local level. Although some community coalitions work to identify and evaluate local problems and then select those most appropriate to address, many such organizations emerge in direct response to a specific local problem. These and other community-based activities can be directed toward reducing the availability of alcohol.

Overview of the Research and Practice Evidence for Approach 6

The analysis of the effectiveness of community-based alcohol prevention activities is based on three research studies and six practice cases.

Three research studies evaluated the effectiveness of community-based prevention programs.

• Giesbrecht, Pranovi, and Wood (1990) evaluated the effectiveness of a community-based project to reduce alcohol consumption among heavy drinkers. The study analyzed whether a change in the proportion of heavy drinkers has a measurable impact on the overall distribution of alcohol consumption. The research also examined interactions between heavy and moderate drinkers and assessed the impact of drinking management interventions on heavier drinkers.
• Hingson et al. (1996) evaluated the effectiveness of the Saving Lives Program, a community effort that organized multiple city departments and private citizens to reduce alcohol-impaired driving, alcohol-related driving risks, and traffic deaths and injuries. The activities included media campaigns, speeding and drunk driving awareness days, "speedwatch" telephone hotlines, police training, SADD chapters, alcohol-free proms, beer keg registration, increased surveillance of liquor outlets, preschool education programs, and training for staff at hospitals and prenatal clinics.
• Pentz et al. (1989) evaluated the effects of the Midwestern Prevention Project as administered to the initial 1984 cohort of sixth- and seventh-grade adolescents in Kansas City communities from September 1984 through January 1986. The program components delivered during this period consisted of a 10-session youth education program on skills training for resisting drug abuse, 10 homework sessions involving active interviews and role-playing with parents and family members, and mass-media coverage.

The practice evidence reviewed on community-based prevention activities included six practice case studies.

• The Association for Responsible Alcohol Control was established by a group of Latino and other activists who were concerned about the relatively high density and rapid growth of alcohol outlets in Latino neighborhoods and the disproportionately high arrest rates for alcohol-related offenses among Latinos. The group focused on passing a law that would require new businesses seeking permission for off-site alcohol sales to undergo a public hearing, giving residents a voice in decisions regarding alcohol availability in their neighborhoods.
• The Central Precinct Neighborhood Alliance comprised numerous neighborhood associations that worked with the Portland Police Bureau to establish a community-policing, problem-solving action plan and partnership agreement. The alliance and its efforts were designed to reduce alcohol abuse among problem street drinkers and thereby reduce alcohol-related local criminal activity.
• The Community Coalition for Substance Abuse Prevention and Treatment involved grassroots community activism organized to influence the alcohol licensing process in Los Angeles. The goal was to decrease the density of alcohol outlets and diminish the violence associated with high outlet density.
• The Escondido Community Alcohol Planning Project set up a city-based planning committee that received technical assistance to document alcohol availability in retail, public, and social environments; to identify strategies for addressing problems in these environments; to select intervention policies and programs; and to implement such interventions.
• The Minors in Night Clubs public education campaign in New Jersey was organized in response to an action by the Wildwood City Council that would permit youth aged 18 to 20 years access to drinking establishments during normal operating hours but that would not permit them to drink alcohol there.
• The New Zealand Community Action Project (Casswell and Gilmore 1989; Stewart and Casswell 1993) was designed to examine the relative effect on alcohol-related community activities and public attitudes of a mass-media campaign, with or without the input of alcohol-focused community organizers. The major objective of the project was to increase public support for alcohol policies such as restrictions on availability and advertising.

More detailed abstracts of the above studies appear at the end of this chapter (pp. 84-89).

Lessons Learned From the Evidence Reviewed for Approach 6

The following are the lessons learned from the research and practice evidence reviewed for this section:

• Community-based activities to control alcohol availability can lead to the development of other alcohol-related prevention activities.
• Community-based activities to control alcohol availability can be used to enhance the effectiveness of prevention programs aimed at reducing drinking by individuals.
• Community coalitions formed in response to specific problems can result in the establishment of permanent entities for the maintenance of changes in alcohol availability policies.
• Community organizing initiatives can be combined with mass media campaigns to increase coverage of and debate on alcohol availability issues and proposed changes in local laws, regulations, or policies.

Suggestions for Future Research on Approach 6

The expert panel made the following suggestions for future research on community-based prevention activities:

• Research is needed to examine the extent to which community coalitions can change local norms and standards and the processes for doing so.
• Further research is needed to study ways for community groups and public agencies to develop mutually supportive relationships to achieve and implement public policy goals.
• Research also is needed to better understand the factors that account for ongoing success by community-based organizations and the ways to avoid "burnout."

Recommendations for Prevention Approach 1: Preventing Availability to Underage Youth

The expert panel's recommendations for preventing availability to underage youth focus on laws relating to alcohol purchases and sales and the enforcement of these laws. Their five recommendations in this area are as follows:

1. Enhance Enforcement. There is strong evidence that laws increasing the minimum legal drinking age result in decreases in traffic casualties, and moderate evidence that such laws result in decreases in alcohol consumption and alcohol-related problems. These laws, however, are often poorly or inconsistently enforced. The expert panel recommends that prevention efforts focus on enhancing enforcement. Such efforts can include creating linkages among several entities, including State ABC's, local police, college administrations, and other groups that promote community values and support enforcement of minimum drinking age laws.
2. Be Consistent. It is recommended that local community laws, such as on land use, and local community policies, such as on implementation and enforcement, be consistent with local prevention messages.
3. Unite Efforts. Experience suggests that having a high minimum drinking age is more effective when complemented by adjunct efforts such as implementing land use laws consistent with drinking laws.
4. Anticipate Crises. Because the enforcement of minimum drinking age laws is often driven by tragedies and crises, there is a practical advantage in anticipating crises, such as those likely to occur during special events, high school and college activities, and holidays. Many college fraternities and sororities, for example, are involved in activities centered around drinking alcohol, promoting heavy drinking as normative behavior and frequently leading to serious alcohol-related problems. When fraternities and sororities are reluctant to respond to efforts aimed at reducing alcohol-related problems, pressure can be applied to their national organizations and insurance companies, which can in turn apply significant pressure and leverage to seek ways to solve the problems. In this way, the self-interest of a fraternity or sorority becomes an important aspect of prevention efforts.
5. Educate Underage Youth. Research has shown that some young people often mistakenly believe that drinking beer impairs driving ability less than drinking spirits, and they often underestimate the amount of alcohol required to make driving unsafe (Lang, Kass, and Barnes 1983; Williams, Lund, and Preusser 1986). These beliefs are in sharp contrast with research results showing that one or two drinks will impair driving-related skills (Moskowitz, Burns, and Williams 1984); that four or five beers often produce BAC's in excess of the legal limit of 0.10 percent (O'Neill, Williams, and Dukowski 1983); and that compared with older drivers, teenagers with low and moderate BAC's are much more likely to be in traffic crashes (Mayhew et al. 1981). Special efforts should therefore be made to provide underage youth with accurate information on the effects of drinking alcoholic beverages and driving.

Recommendations for Prevention Approach 2: Raising Alcohol Taxes and Prices

The expert panel's recommendations regarding alcohol taxes and prices address issues such as taxation methods and how to prevent the erosion of tax benefits. Their four recommendations in this area are as follows:

1. Levy Local Taxes. Where State law allows, local taxes such as "nickel-a-drink" levies on on-sale purchases may be an effective way of financing local initiatives to address alcohol problems while reducing rates. Local license fees also may be used as a kind of "user fee" device to mitigate alcohol problems. Because price increases affect the sale of alcoholic beverages, community prevention efforts can include strategies that increase alcohol prices. Although excise taxes have many benefits, resistance to tax increases by voters and beverage industries suggests that a range of approaches to raising prices should be considered.
2. Equalize Taxes. Most of the adverse consequences of drinking result from the amount of alcohol in each drink, whether the alcohol is concentrated or diluted. To maximize public health benefits, taxes on different beverages can be set so that the retail prices of the most inexpensive form of each beverage are approximately equal. This may require a higher tax level per unit of alcohol on beverages such as spirits, which are more concentrated and thus cheaper to manufacture and distribute than beer and wine.
3. default State Taxes. The benefits of increases in State alcohol taxes, such as reductions in alcohol-related health and social problems, are likely to diminish as inflation erodes the real value of the tax increase. To avoid this erosion, the tax can be defaulted so that the nominal tax rate rises in step with prices. The expert panel suggests that defaulting alcohol taxes to the consumer price default should make the public health gains of higher taxes permanent.
4. Seek Allies. The benefits of State alcohol taxes erode when the strategy behind them is out of step with neighboring jurisdictions. Where this is the case, countervailing effects such as cross-border shopping, cross-border drinking and driving, theft, and black-market sales can occur. When neighboring jurisdictions adopt equivalent regulations, however, such problems greatly diminish. The expert panel recommends that individuals engaged in planning prevention strategies including alcohol taxes initiate discussions and possible collaboration with their counterparts in neighboring regions.

Recommendations for Prevention Approach 3: Responsible Beverage Service

The expert panel recommendations regarding responsible beverage service focus on enforcement, liability, licensing, and general activities. Their 10 recommendations in this area are as follows:

1. Enforce the Laws. All States and most jurisdictions already have laws prohibiting alcohol sales to intoxicated and underage individuals. The panel's principal recommendation concerning responsible beverage service is that jurisdictions strictly and uniformly enforce the laws regarding the sale of alcohol to such individuals.
2. Target Trouble Spots. Although the panel recommends uniform enforcement of drinking laws, it also recommends targeting high-risk drinking establishments for opportunities for enforcement. Lax enforcement of State and local liquor laws may compel some businesses to engage in illegal practices, such as serving intoxicated patrons or underage youth. This may be especially true among drinking establishments in areas with a high density of alcohol outlets and a relatively small consumer base. Establishments that serve intoxicated or underage individuals may attract high-risk patrons, thereby increasing the risk for alcohol-related problems in the community. There are several ways to identify high-risk outlets; in particular, greater attention can be paid to adherence to the laws at establishments linked to substantial incidents of drunk driving arrests. Such high-risk establishments can be singled out by identifying the locations where arrested drunk drivers purchased their last drinks. This information can be collected by the police during arrests, by counselors during hearings, or by trainers during DWI education sessions.
3. Keep the Burden on the Owner. Since the 1970's, most States have increasingly recognized some form of liability on the part of drinking establishments, whether by legislation, court decisions, or both. In the past few years, however, under the influence of alcohol-retailer lobbies a countertrend has emerged to reverse the court-inspired progress toward increased liability (Wagenaar and Holder 1991a). What's more, in a few States the civil liability for intoxication-related problems following alcohol consumption in a drinking establishment has shifted from establishment owners to their servers. Typically, this shift in liability is accompanied by a mandated server training program, which means that in essence owner liability has been traded for server training. Experience suggests that the strongest incentive to stop service to intoxicated or underage individuals appears to be the threat of business loss due to revocation of an establishment's alcohol license. The expert panel recommends that States and jurisdictions undertake efforts to keep the burden of legal responsibility on the owners of drinking establishments and alcohol licenses rather than their employees, such as servers. Jurisdictions might, in fact, consider increasing such liability burdens, not decreasing them. Interested readers can refer to the model alcoholic beverage retail licensee liability act of 1985 ("Model Alcoholic," 1985), and to the work of Coleman, Krell, and Mosher (1985) and Coleman and Kleinman (1986).
4. Provide Incentives. Experience suggests that drinking establishments generally will not participate in RBS activities unless they have significant incentives to do so, such as the following:
   
   
   • Keeping a license to sell alcohol in order to remain open and avoiding license revocation are the greatest incentives for drinking establishments to participate in RBS training. Because alcohol licenses include specific restrictions, such as prohibiting alcohol sales to intoxicated or underage individuals, enforcement of the relevant laws governing alcohol sales seems to be the best prevention intervention method in this area.
   • Drinking establishments have vested interests in safety and the prevention of injuries for which they are legally liable. As a result, they are naturally concerned with the liabilities associated with serving intoxicated patrons and serving patrons to the point of intoxication. The expert panel recommends that prevention activities involving outlet owners and managers play on these inherent concerns.
   • Although drinking establishments are in the business of selling alcoholic beverages, this does not preclude them from wanting to prevent alcohol abuse, addiction, and related problems. Indeed, individuals who serve and sell alcohol can have strong interests in the overall health of their customers that encourage active partnerships with prevention practitioners.
   
   
   
5. Close Loopholes. In some cases, the owner of a problem establishment may transfer ownership of the establishment to a friend or relative while continuing to operate it. The panel recommends that when a State or jurisdiction revokes a license, the license should not be allowed to transfer to a new establishment owner. To cut down on such tactics, it is recommended that States and jurisdictions permanently revoke the license to a specific location and force any potential new owner to go through a new formal license-application process. Restricting the transfer of licenses does not, however, guarantee that a previously irresponsible owner will not become involved with a new business. Therefore, it is critical to screen all potential owners and managers of drinking establishments carefully.
6. Avoid Grandfather Exceptions. In some locations, purchasing an existing business and its alcohol license requires a different legal process than purchasing a new license. In some situations, restrictions that would apply to new licenses do not apply to existing ones. The expert panel recommends that jurisdictions refuse to transfer licenses to new owners unless all restrictions that apply to new licenses are transferred as well.
7. Intervene Early. States and other jurisdictions are often reluctant to revoke or suspend alcohol licenses, waiting until numerous complaints and severe problems are reported before acting. The panel recommends that States and jurisdictions intervene in problem establishments early rather than waiting until problems associated with a drinking establishment get even worse.
8. Provide Server Training. It is recommended that RBS training be provided to individuals who serve alcohol at special occasions such as sports and community events, street fairs, and business and private parties. Prevention practitioners can develop productive relationships with hospitality-related organizations, such as professional and trade associations, in developing server training programs. Experience suggests that there is a significant rate of turnover among alcohol servers in the hospitality business. As a result, the expert panel recommends that continuous server training rather than one-time sessions be offered to entrants into the business, and that periodic refresher sessions be held for all servers.
9. Focus on All Components of Responsible Beverage Service Training. Several national organizations and State agencies are currently delineating which RBS components are essential, optimal, and effective. The panel recommends that server training be understood as one aspect of responsible hospitality, which encompasses the following:
   
   
   • Creating positive incentives by recognizing and rewarding businesses and events that practice responsible beverage service.
   • Developing community norms on the principles and practices of responsible hospitality.
   • Developing guidelines for employers and social hosts.
   • Enforcing regulations governing the sales and serving of alcoholic beverages consistently.
   • Instituting professional development programs for management and service staff.
   • Providing information and training for organizers and volunteers at community events.
   
   
   
10. Offer Alternatives to Alcohol. Providing food and nonalcoholic beverages appears to be a useful way to cut down on alcohol consumption. Depending on the drinking establishment, however, selling these products can range from profitable to revenue-neutral or even unprofitable; as a result, some establishments see little incentive to provide such services. Nevertheless, RBS programs can encourage managers and servers to offer nonalcoholic products by promoting them as ways to encourage designated nondrinking drivers and to reduce intoxication and related problems and liabilities, as well as a potential source of revenue. Experience suggests that server interventions in this regard work only when they are part of management policies that support responsible beverage service; these interventions are most effective when supported by community-based programs for reducing alcohol-related problems. It also appears that RBS is effective only where there is strict enforcement of laws prohibiting sales to intoxicated or underage individuals.

Recommendations for Prevention Approach 4: Changing the Conditions of Availability

The expert panel's recommendations regarding changing the conditions of alcohol availability address general issues such as sales practices and the geographic spacing of outlets as well as the regulation of alcohol at special events and locations. The panel's nine recommendations in this area are as follows:

1. Collect Data. When communities first perceive problems related to alcohol outlet density, it is important that they set an individual skilled in data collection and analysis to assembling facts relevant to the problems. Important information to obtain includes police activities, citizen complaints, and State license violations.
2. Learn the Ropes. Research suggests that communities that initiate and enforce restrictions on the spacing of alcohol outlets generally experience a reduction in police problems with off- and on-site sales, at least for several years. The expert panel thus recommends that community-based prevention groups become involved in local control issues such as outlet density and spacing. It is important for these groups to learn about licensing laws and processes and how they can become involved in developing them. Few local laws are changed without such activism, often at the community level. When considering prevention efforts related to alcohol outlet density or spacing, it is recommended that local communities become familiar with their State laws and issues related to State preemption; communities can, for example, explore the use of small claims courts as a remedy for dealing with nuisance outlets. Along the same lines, the panel recommends that State efforts stay sensitive to the needs and requirements of local communities; it is, after all, optimal for States and local communities to work together when either initiates a prevention effort related to outlet density or spacing. In areas with high densities of alcohol outlets, the clustered outlets are likely to be similar in type and to engage in vigorous competition, which can result in conditions conducive to availability-related problems as follows:
   
   
   • Competition in high-density areas can lead to increases in high-risk sales practices such as discounting drinks, holding events to attract large crowds, providing marginal entertainment, and lessening supervision of drinkers. Other competitive marketing strategies include providing free food, live music, "happy hours," multiple televisions for viewing sporting events, "ladies' nights," bikini contests, and female mud wrestling. These sorts of stunts are, of course, far less prevalent in high-density areas such as restaurant rows and high-end commercial specialty areas.
   • In fiercely competitive markets where on-sale outlets are competing for the same patrons, successful marketing strategies are likely to be replicated — that is, if one outlet becomes successful by using a specific marketing technique as described above, nearby outlets will try the same or a similar strategy. Merchants can, however, be encouraged to seek alternative strategies for marketing to different types of patrons.
   
   
   
3. Consider Compatibility. Neighborhood compatibility is an important aspect of prevention efforts aimed at controlling outlet density. For instance, a specific outlet may be incompatible with other surrounding businesses or local residents, such as by providing late-night entertainment or late hours of sale. The expert panel recommends that every alcohol-license applicant be interviewed by local residents and businesses and charged with demonstrating to the city or licensing body that the immediate neighborhood does not object to the business moving in.
4. Intervene Early. The panel recommends that discussions regarding alcohol control activities such as the provision of nonalcoholic beverages and other measures be introduced early in the planning stages of community-sponsored festivals, street fairs, and other special events.
5. Train Servers. It is recommended that all individuals responsible for dispensing alcoholic beverages at special events receive server training to learn the relevant laws and policies, including policies for resolving alcohol-related problems. Guidelines for problem resolution should thus be developed in advance of special events and should be understood by all staff working at such events.
6. Disseminate Rules. Rules regarding alcohol consumption at a special event can be printed and disseminated on signs, brochures, and event tickets. Publishing such rules may diminish the likelihood of trouble related to the enforcement of alcohol regulations.
7. Use Physical Aids. Experience has shown that age-based and other restrictions on alcohol sales at special events can be more easily and strictly enforced through the use of hand stamps or identification bands issued at locations other than the alcohol purchase site. Experience also suggests that limiting beer sales (e.g., not selling to people in their seats at sporting events and cutting off sales after the seventh inning or third quarter) and providing nondrinking "family areas" decrease alcohol-related problems and attract attendance by families averse to drunken rowdiness.
8. Educate Promoters. Alcohol-related problems at sporting events, rock concerts, and other large special events fall into two categories. The first includes unpleasant behavior, rowdiness, fights, and personal and property damage at the site before and during the event; the second includes DWI arrests, drunk driving casualties, street fights, and other incidents that occur after the event, away from the event site, and on patrons' way home. Although many event promoters have interests in addressing both types of alcohol-related problems, some promoters do not adequately address the latter. Prevention practitioners can play an important role by providing information to promoters regarding all types of alcohol-related problems associated with special events.
9. Balance Interests. Permitting and controlling alcohol use at sporting and similar events involves balancing several needs and concerns. Alcohol sales are a substantial source of profit for stadium and arena owners, sports teams, and catering companies. If alcohol-related problems become too numerous and severe, however, attendance may suffer and liability issues may arise. Because community interests can influence those of private businesses, prevention efforts may be more effective if they embrace the need for balance.

Recommendations for Prevention Approach 5: Changing Hours and Days of Sale

The expert panel's one recommendation regarding changing hours and days of sale concerns familiarity with existing laws is as follows:

• Know the Laws. It is important for communities to be familiar with State and local laws regarding local control over alcohol outlets' hours and days of operation in order to increase where possible the power localities can exercise in this area.

Recommendations for Prevention Approach 6: Community-Based Prevention Approach

The expert panel's recommendations regarding community-based prevention activities address issues such as forming coalitions, recruiting participation by the retail and wholesale alcohol beverage industry, and sustaining policies. The panel's six recommendations in this area are as follows:

1. Use Multiple, Integrated Strategies. Research and experience suggest that multicomponent community-based strategies are more effective than those with only one; the individual elements of a multicomponent strategy can strengthen, complement, and support each other in a cumulative effect whose overall impact is greater than the sum of the impacts of the individual components. In addition, if one aspect of a multicomponent strategy is eliminated, the combination of the remaining components may continue to exert a significant preventive effect. For example, a multicomponent RBS strategy could involve server intervention training, the development of management policies and procedures in drinking establishments, and enforcement efforts. The panel therefore recommends that community-based coalitions and practitioners identify and utilize complementary multiple prevention strategies.
2. Support Policies. Experience suggests that it is important for community coalitions to see prevention activities as having two phases: implementation and sustainment. Community coalitions that focus only on the implementation of a policy or regulation will often see their gains diminish if attention is not paid to sustaining the policy changes and related prevention activities. One way to promote the sustainability of such activities is to institutionalize prevention efforts. For example, ad hoc grassroots groups can evolve into permanent committees within city or county governments. Where this occurs, however, ongoing community participation is important to protect against a loss of vigor in committee activities.
3. Cultivate Membership. It is recommended that prevention practitioners engage in ongoing efforts to recruit new members, maintain existing ones, and respond to new requests. The panel recommends that community coalitions stay involved in prevention and keep paying attention to the issues. Coalition leaders can take steps to keep members involved, motivated, and educated; government agencies can provide technical assistance to community coalitions to help them stay involved, flexible, tolerant of problems such as staff changes, and willing to examine new goals when previous ones have been attained.
4. Use Practitioners. Professional prevention practitioners can serve as a primary information source for community-based coalitions. Over time, specific issues will change, partnerships will evolve, and information needs will vary; prevention practitioners can offer important information, guidance, and resources for community coalitions despite such changes.
5. Organize Mindfully. The expert panel recommends that grassroots community coalitions have control and authority over the direction of local prevention efforts equal to those of service providers, city officials, and other professional participants. In general, the success of community coalitions depends on the participation and collaboration of public agencies, local community groups, and organizers whose mission is to achieve specific goals. Although organizers can be volunteers or paid professionals and may come from public agencies, it is generally not appropriate for organizers to represent any particular agency. Experience demonstrates that some agencies and groups are good partners in certain coalitions but not in others, depending on the issues at hand. For example, entities such as police departments make good coalition partners for dealing with public safety issues, but as public employees police need to remain impartial on certain politicized issues, which would prohibit them from joining coalitions devoted to those issues.
6. Identify, Secure, and Organize Data. In preparation for community coalition interventions, it is recommended that data be collected on the distribution of retail alcohol availability and the problems associated with it. The goal should be to gather baseline data showing the relationship between outlet density and resulting problems. In many cases, public agencies have such data in public records that can be easily retrieved.

General Recommendations for Alcohol Availability

In addition to its specific recommendations for each of the six prevention approaches considered in this chapter, the expert panel provided six general recommendations regarding alcohol availability, as follows:

1. Experiment. Many of the variables that make an intervention effective in one situation do not work in others. For this reason, communities would do well to keep an open mind about experimental prevention efforts appropriate to their particular situation and region. If such interventions do not seem to work, expert advice can be sought and adjustments made. Throughout any such experiments, it is important for communities to document the processes followed, including any adjustments to them.
2. Head Off Problems. Sometimes the motivation for instituting specific prevention efforts is to address a particular crisis; at other times it may be to identify and resolve problems early in their manifestation. Whenever possible, it is important for community practitioners to focus on identifying and intervening in problems before they become crises. Local groups can seek collaboration with other components of the prevention system, such as internal liquor control committees, to obtain evidence of burgeoning problems with an eye toward early intervention. At the same time, when possible it should be determined whether prevention efforts will be short- or long-term so that appropriate approaches and tools are used.
3. Foster Trust. It is important in prevention efforts to take steps designed to promote a trusting relationship between law enforcement and local businesses and merchants. This can be accomplished through formal or informal committees or boards set up to reduce specific alcohol-related problems or to study general problems related to alcohol availability.
4. Promote Cooperation. Community members and retailers may be unaware that police powers supersede the authority of retail alcohol merchants, and that prevention efforts might be enhanced by cooperation among merchants, police, and other regulatory authorities. When considering policy changes at the local level, it is advisable for prevention practitioners to work with State agencies to establish joint regulations to ensure coordination with State and local policies. A useful approach for enhancing cooperation and collaboration among State and local efforts involves establishing a single point of contact, such as a coalition, committee, or working group within a city government that focuses on alcohol availability issues through the city manager's office.
5. Empower Community Members. Empowerment activities can be used to develop new power centers in underrepresented parts of a community. Under such strategies, community members are encouraged to pool their collective resources to define and solve shared problems and to develop action plans for solving or preventing identified problems with alcohol availability. By sharing the process of problem assessment, a common vision can emerge and agreements can be reached on how to proceed.
6. Use Free Legal Aid. Nuisance abatement can be pursued through the civil section at the county attorney level. Whereas many city attorneys provide contract services at an hourly rate, State and county attorneys may provide pro bono assistance to community groups seeking to do something about nuisances related to alcohol availability.

Recommendations for ABC-Related Activities

The expert panel formulated the following two general recommendations for ABC-related activities:

1. Communicate With ABC's. The orientations and missions of ABC's differ among the States with regard to staffing, police powers, legal mandates, and policy direction from legislative and executive offices. Many States, such as California, have a public health, welfare, and safety mandate in addition to a legal mandate to prevent the development of marketing conditions that disrupt the orderly distribution and sale of alcoholic beverages. Both mandates are important to the effective and efficient regulation of the industry; they also share a common purpose, which is to regulate the industry in a manner consistent with the welfare of the public as a whole. Therefore, some ABC's place greater emphasis on enforcing marketing regulations than others because of State legislative policies. The expert panel recommends that communities establish consistent and open communication with ABC's, supporting and encouraging the agency to recognize public health, welfare, and safety mandates while supporting legislation for alcohol licensing that promotes and emphasizes public health goals. When prevention groups have difficulties getting appropriate responses from an ABC, State and local officials can be called upon to urge the board to respond. When working with ABC's, the best approach is cooperation, collaboration, and a focus on common grounds.
2. Alert ABC's. It is recommended that State ABC's be alerted to the initiation of new prevention initiatives so they can provide support for them when appropriate.

Recommendations for Outlet-Related Activities

The expert panel formulated three recommendations for alcohol outlet-related activities, as follows:

1. Exercise Police Powers. Communities have significant police powers to support public health issues, a fact that is often overlooked. Some individuals in the alcohol industry believe that public health authorities cannot exert significant control over their activities; in fact, there is considerable control at the local level under the authority of public health and safety agencies.
2. Pressure Problem Outlets. Although many communities that issue conditional-use permits have been successful in slowing the opening of new alcohol outlets, many remain frustrated with the problems associated with existing outlets. Groups may need to establish problem prevention strategies for existing outlets that differ from those for new outlets. For example, groups may collaborate with the local police and may request assistance with priority problem outlets. Letters can be sent to these outlets requesting reasons why the prevention group should not announce a public hearing concerning the outlet. Merchants will often cooperate under these circumstances. When they do not, hearings can be held to discuss problems and potential solutions regarding the outlet. If the problems are not resolved within a year, another hearing can be held during which a recommendation can be made to close any problem establishments.
3. Explore Alternative Fixes. Economic incentives can be effective when working with problem outlets. For example, owners of problem drinking establishments can be offered extra incentives to reorient their businesses, such as by turning a bar into a coin-operated laundromat. Community economic development departments can be helpful during these negotiations. Overall, alcohol outlets are owned by individuals with substantial financial and commercial stakes, keeping which can be a powerful incentive to cooperate with local values. The expert panel recommends that prevention efforts consider the needs of these businesspeople, including their need to avoid liability problems.

Recommendations for Zoning-Related Activities

The expert panel made the following two recommendations for zoning-related activities:

1. Document and Test Successful Practices. Local officials and community groups in thousands of cities and counties often use planning and zoning ordinances to prevent, reduce, and respond to problems related to retail alcohol outlets. A substantial disparity exists between the wealth of undocumented practice experience and the level of scientific attention applied to understanding and improving the preventive uses of that experience. The value of this work to prevention efforts may be clear to the many officials, citizens, and outlet operators who are directly involved, but their practical experience has yet to be fully documented or rigorously tested.
2. Enhance Zoning-Related Activities. Prevention practice and some research on the uses of local planning and zoning ordinances suggest that enhancing the application of local planning and zoning to alcohol outlets can be effective in preventing alcohol-related problems. Various hypotheses can be tested to evaluate the effectiveness of zoning-related activities and thus to help communities make informed decisions about strengthening their prevention efforts by expanding enforcement activities directed at alcohol outlets; making greater and more efficient use of existing ordinances; and developing stronger local ordinances to reduce the density of alcohol outlets, as follows:
   
   
   • If problems related to alcohol outlets can be reduced or eliminated through a variety of enforcement actions, and if those enforcement actions are applied, then the problems of individual outlets should be reduced or eliminated. Despite insufficient documentation and analysis practice, experience suggests that this hypothesis holds true.
   • If local planning and ordinances related to alcohol outlets and availability reduce alcohol-related problems, and if communities use these interventions, then alcohol problems in the community should decrease after community-level initiatives are applied through local planning and zoning ordinances.
   • If higher densities of alcohol outlets in a community are associated with higher problem rates, and if the introduction of local CUP zoning ordinances will reduce the density of alcohol outlets, then alcohol problems in the community should decrease when such ordinances are applied.

Research Evidence Reviewed for Approach 1

O'Malley and Wagenaar (1991) conducted two studies designed to (1) compare drinking behaviors and the number of alcohol-related motor vehicle collisions in States with a high minimum legal drinking age with those in States with a low minimum legal drinking age, and (2) examine changes in drinking behavior and the number of motor vehicle collisions in States that raised the minimum legal drinking age to 21 years with those in States that maintained the minimum legal drinking age at 21 years. The first study was a cross-sectional survey comparing two conditions across time; the second was a time-series analysis examining data collected annually over a period during which a policy change occurred. The findings included the following:

• Before the Federal Government required the States to raise the minimum legal drinking age, high school seniors in States with a minimum legal drinking age of 21 consumed less alcohol than their counterparts in States where the minimum legal drinking age was lower than 21. In the former States, lower rates of drinking among teenagers appeared to persist as these youth entered their 20s.
• Raising the minimum legal drinking age from 18 to 21 years reduced alcohol consumption among high school seniors.
• Alcohol-involved highway collisions declined among those aged 18 through 20 years after the minimum legal drinking age was raised.
• High school seniors spent less time in bars and taverns in States that raised the minimum legal drinking age to 21 years.

In December 1978, Michigan raised its minimum legal drinking age from 18 to 21 years. Wagenaar (1986) examined the effects of raising the minimum legal drinking age on the number of injury-producing traffic crashes among drivers between the ages of 18 and 20. This time-series study compared the rates of motor vehicle collisions before and after the change among drivers police officers said "had been drinking," the rates of single-vehicle nighttime traffic crashes before and after the change, and the rates of injury-producing crashes among drivers 21 years and older who "had been drinking" with those for drivers aged 18 through 20 years. The study found the following after the minimum legal drinking age was increased to 21:

• The number of crashes involving drinking drivers between the ages of 18 and 20 decreased by 6 percent, while crashes involving drinking drivers aged 21 and older rose by 13 percent.
• The overall reduction in crashes attributable to raising the drinking age among drinking drivers was 19 percent.
• The number of single-vehicle nighttime crashes among drivers aged 21 and older increased by 16 percent.

Wagenaar and Maybee (1986) examined the effects on motor vehicle collision rates of a 1981 Texas State law raising the minimum legal drinking age from 18 to 19 years. Their study, which covered 1978 through 1984, employed a quasi-experimental, nonequivalent, interrupted, multiple time-series design. The rates of motor vehicle collisions per 100,000 licensed drivers were examined for three levels of severity: (1) serious injury; (2) minor injury; and (3) property damage only for four age groups: (1) 16 to 17 years; (2) 18 years; (3) 19 to 20 years; and (4) 21 years and older. Among the study's findings were the following:

• The group of licensed drivers aged 18 years experienced significant reductions in rates of single-vehicle nighttime collisions at all levels of collision severity (serious injury collisions, 10.8 percent; minor-injury collisions, 14.3 percent; property damage-only collisions, 12.8 percent). Drivers aged 16 through 17 years experienced similar reductions.
• The relative rate of decline in single-vehicle nighttime collisions between drivers aged 18 years and those 21 and older was 9.5 percent for serious-injury collisions, 7.6 percent for minor-injury collisions, and 7.8 percent for property damage-only collisions. There was no significant change in single-vehicle nighttime collisions among drivers aged 19 years and older.
• Reductions in rates of nonsingle-vehicle nighttime collisions for drivers aged 18 years were consistently smaller than the reductions in the rates of single-vehicle nighttime collisions.

Many States have established lower legal BAC limits for drivers younger than age 21 than for older drivers. Hingson, Heeren, and Winter (1994) compared the rate of alcohol-related automobile crashes in 12 States where the BAC limit was lower for young drivers than for adults with those in 12 comparison States where the BAC limit was the same for youths and adults. Matches were made for States that had BAC limits of 0.00 percent, 0.02 percent, and 0.04 to 0.06 percent. Data on fatal crashes were based on the U.S. Department of Transportation's Fatal Accident Reporting System. Data were collected over a 12-month period, beginning with the month in which the reduced-BAC law was enacted. Log linear analyses were used to test the significance of differential shifts in the proportion of single-vehicle fatal nighttime crashes. The study's findings included the following:

• Among drivers aged 15 through 20 years, single-vehicle fatal nighttime crashes were three times more likely than other fatal crashes to be alcohol-related.
• The proportion of single-vehicle fatal nighttime crashes declined by 16 percent among young drivers in States that had lowered BAC limits for underage youth, whereas it rose 1 percent among drivers of the same age in comparison States where the BAC limits were not changed.
• After the BAC laws were enacted, the proportion of single-vehicle fatal nighttime crashes declined by 5 percent in States that had lower levels for young drivers and by 6 percent in the group of neighboring comparison States.
• The proportion of single-vehicle fatal nighttime crashes declined by 22 percent among drivers in States with a BAC limit of 0.00 percent but by only 2 percent among drivers of the same age in the comparison States.
• Among those targeted by 0.02 percent BAC limits, the proportion of single-vehicle fatal nighttime crashes declined by 17 percent but rose by 4 percent in the comparison States.

A 1988 Maryland State law prohibits driving by those younger than 21 who have a BAC of 0.02 percent or more (in contrast to a 0.10 percent BAC limit for drivers older than 21). This law effectively prohibits underage youth from operating motor vehicles after consuming even a small amount of alcohol. Their driver's licenses are imprinted with the phrase "Under 21 Alcohol Restricted." Underage youth with a 0.02 percent BAC can be charged with violating the restriction and may be punished with a fine up to $500 and suspension of their driver's licenses for up to a year. Underage youth in Maryland can be charged with driving under the influence (DUI) if their BAC is 0.07 percent or more and with DWI if it is 0.10 percent or more.

The National Highway Traffic Safety Administration (1992) examined the effectiveness of the Maryland BAC law and of the State's public information campaign to promote its effects. The experimental campaign was launched in six counties; two others served as comparison counties. The campaign included broadcasting television and radio public service announcements, brochures, and posters; informing police agencies of the importance of the law and promoting its enforcement; and encouraging local groups to help teach the public about the BAC law. Statewide crash statistics from 1985 to 1990 were obtained from the Maryland State police; measurements of the public information campaign's exposure were derived from surveys conducted during driver's license applications and renewals at colleges and high schools. Among the findings were the following:

• The Maryland BAC law was associated with a significant statewide reduction (about 11 percent below the baseline mean) in the number of accident-involved drivers who were under age 21 and judged to have been drinking.
• In the six experimental counties, the percentage of all underage drivers involved in crashes was 8.9 percent at baseline, 7.4 percent during the early law period, 6.3 percent after licenses were imprinted, and 5 percent after the public information campaign. This 44 percent drop from the pre-law period through the information campaign compares with a 30 percent drop in the comparison counties.
• In the experimental counties, significant baseline-to-posttest increases appeared in public recall of information about alcohol license restrictions being provided through print media (23 percent), television (25 percent), and radio (26 percent).

Hingson et al. (1996) evaluated the effectiveness of the Saving Lives Program, a community effort to organize multiple city departments and private citizens to reduce alcohol-impaired driving, alcohol-related driving risks, and traffic deaths and injuries. In each of the six program communities a full-time coordinator from the mayor's or city manager's office organized a task force of concerned private citizens, organizations, and officials representing various city departments. The communities developed initiatives such as media campaigns, speeding and drunk driving awareness days, "speedwatch" telephone hotlines, police training, SADD chapters, alcohol-free proms, beer keg registration, increased surveillance of liquor outlets, preschool education programs, and training for staff at hospitals and prenatal clinics. The intervention cities were compared with the rest of Massachusetts and with five cities that prepared high-quality proposals that were not funded. This quasi-experimental study evaluated the impact of the intervention on traffic crashes and injuries, safety belt use, vehicle travel speeds, and driving after drinking.

Monitoring of crashes was based on data from the Department of Transportation's Fatal Accident Reporting System. These data were collected 5 years before and 5 years after the start of the program. Additional data came from the Massachusetts Registry of Motor Vehicles and were collected 4 years before and 5 years after the start of the program. Five annual direct observation surveys of speeding and safety belt use were conducted 1 year before and 4 years after the start of the program; these surveys involved using radar from unmarked cars for detecting speeding vehicles at randomly selected roadways at intervention and comparison cities, as well as direct observation at signalized road intersections and freeway off-ramps using a stratified random sampling procedure. To assess program awareness, beliefs about police enforcement, and the frequency of driving after drinking, four independent cross-sectional, random-digit-dial telephone surveys of 15,188 young people aged 16 to 19 years and adults aged 20 years and above were conducted in intervention cities and the rest of the State between 1988 and 1993.

Compared with the rest of the State and with the 5 years before the intervention, the following findings were noted in the Massachusetts intervention cities during the 5 years of the program:

• Alcohol-related fatal crashes declined by 42 percent.
• Fatal crashes declined by 25 percent.
• Fatal crashes involving drivers aged 15 to 25 years declined by 39 percent.
• The number of fatally injured drivers with positive BAC's declined by 47 percent.

Compared with the rest of the State and with the 4 years before the intervention, the following findings were noted in the Massachusetts intervention cities during the 5 years of the program:

• The number of pedestrian injuries per 100 crashes declined by 10 percent.
• The number of visible injuries declined by 5 percent.

In contrast to five nonintervention comparison cities, the Massachusetts intervention cities experienced a 33 percent decline in fatal crashes and a 42 percent decline in alcohol-related fatal crashes. The following outcomes were also reported:

• Safety belt use increased from 22 percent in 1989 to 29 percent in subsequent years, an increase 17 percent proportionately greater than in the rest of the State.
• The proportion of vehicles observed at speeds more than 10 miles per hour over the speed limit declined from 19 percent in 1989 to 9 percent in subsequent years. This decline was 43 percent greater than in comparison areas.
• The proportion of youths aged 16 to 19 years who reported driving after drinking in the month before being interviewed declined from 19 percent in 1988 to 9 percent in 1993, a decline 43 percent greater than in the rest of the State.
• From 1987 to 1992, combined police traffic citations declined by 12 percent, speeding citations declined by 14 percent, and citations for DWI declined by 13 percent.

Preusser, Williams, and Weinstein (1994) conducted underage alcohol-purchase attempts on a subset of all alcohol outlets in Denver, Colorado. In this staggered quasi-experimental design with random selection, underage male police cadets attempted to purchase six-packs of beer. Three sets of 100 stores were randomly selected from alcohol beverage license lists. Three days after the first set was visited to establish a baseline, a press conference announced the results and warning letters were sent to violators. One month later, all visited stores were revisited (the first enforcement wave), and clerks and licensees were cited at violating stores. Three months later, a second set of letters warned licensees that enforcement would continue. Shortly thereafter, a second sample of 100 licensees was randomly selected and visited (the second enforcement wave); violators were cited as before. Six months later, a third sample of 100 licensees randomly selected from updated license lists were visited (the third enforcement wave). Violators were cited. Due to closings and time constraints, the actual number of stores visited was 88 at baseline and in the first enforcement wave, 84 in the second enforcement wave, and 85 in the third enforcement wave. The study's findings included the following:

• At baseline, 59 percent of the stores sold beer to underage youth. When sales were totaled across the three enforcement waves, it was found that 28 percent of the stores sold beer to underage youth.
• During the first enforcement wave, 32 percent of the stores sold beer to underage youth. Of the stores that did, half had done so during the baseline visit.
• During the second enforcement wave, 26 percent of the stores sold beer to underage youth. Of the stores not visited during the baseline or in the first wave of enforcement, 29 percent sold beer to underage youth. Among the remaining 15 stores, which by chance had been on the first two licensee lists, only 2 sold beer to the underage youth.
• During the third enforcement wave, 26 percent of the stores sold beer to underage youth. Of these 22 sales, 12 occurred at stores that had not been visited before, while 10 occurred at stores that had been visited before.
• The 11 stores that had sold beer to underage youth during the baseline and in the first enforcement wave were revisited during the third enforcement wave. Three of them sold beer to underage youth, seven did not, and one was closed.

Using a self-reporting, cross-sectional survey design, Mooney and Gramling (1993) examined two samples of college students: one from North Carolina, where the minimum legal drinking age was 21, and the other from Louisiana, where the minimum legal drinking age was 18. Patterns of drinking behavior were compared in terms of frequency, quantity, and location of consumption (defined as either controlled locations, where proof of age is required or where there is a higher likelihood of sanctions, or uncontrolled locations such as dormitories and fraternity houses, where sanctions are unlikely). Among the survey's findings were the following:

• In North Carolina, drinking was deterred in controlled locations but not in uncontrolled locations.
• In controlled locations, the State/age interactions for frequency and monthly consumption were significant.
• In uncontrolled locations, none of the State interactions was significant, suggesting that such laws have little effect in situations where the threat of formal social control is absent or minimized.

A study by Preusser et. al (1995) looked at alcohol policies in New York and Pennsylvania, which have substantially different State laws regarding the purchase, possession, and consumption of alcohol and the use of false identification by underage individuals. New York has less stringent laws than Pennsylvania, which has strong laws and State-controlled liquor stores. Because of these differences, and because the prevalence of the use of false identification by underage youth could be related to these variations, the study compared the behaviors of underage youth attempting to purchase alcohol in the two States. The subjects were aged 20 years and younger and included high school juniors and seniors as well as college students. This cross-sectional study was based on surveys that focused on the subjects' (1) frequency of drinking, (2) sources of alcohol, and (3) use of false identification. Completed surveys were obtained from 2,167 underage youth in New York and 2,223 underage youth in Pennsylvania. Among the findings were the following:

• Although more high school students in New York than in Pennsylvania reported that they had ever consumed alcohol (other than in the presence of their immediate family) and more students in New York than in Pennsylvania reported drinking two or three times per month, the differences in the rates were not significant.
• High school students in New York (43 percent) were more likely than those in Pennsylvania (30 percent) to have attempted to purchase alcohol at an outlet.
• College students in New York (75 percent) were more likely than their Pennsylvania counterparts (59 percent) to have attempted to purchase alcohol at an outlet.
• One-third of all students in both States (about 21 percent of the high school students and 48 percent of the college students) had used false identification in attempts to purchase alcohol. High school students in New York (28 percent) were more likely than those in Pennsylvania (14 percent) to report having used false identification.
• Among underage youth who did not directly purchase alcohol, the most likely sources of alcohol were friends under 21; other sources were friends over 21 and parents. However, New York underage youth were more likely than their counterparts in Pennsylvania to obtain alcohol from parents.

The U.S. General Accounting Office (1987) applied evaluation synthesis methodology to then-available research studies that analyzed the effects of raising minimum drinking age laws on traffic crashes and alcohol consumption among those affected by the laws, generally individuals aged 18 to 20 years. The primary measurements used in the crash studies were police and coroners' reports and surrogate indicators of alcohol use such as single-vehicle nighttime crashes. The crash data were grouped in categories, such as by driver fatalities and injuries. The measures used in the consumption studies were self-reporting surveys, interviews, and alcohol sales data. Findings were reported for the following categories:

• Consumption After Minimum Age Increases. Four studies evaluated the relationships between increasing drinking age laws and the levels or frequency of consumption by the groups affected by minimum age laws (Coate and Grossman 1985; Lillis 1984; Perkins and Berkowitz 1985; Williams and Lillis 1985). These studies suggest that raising the minimum drinking age by law promotes a significant decrease in the frequency and amount of drinking among the age groups affected by such laws.
• Driver Fatality Crashes (Multi-State). Four studies evaluated the effects of increasing the drinking age on crashes involving driver fatalities across multiple States (Arnold 1985; DuMouchel, Williams, and Zador 1985; Hoskin, Yalung-Mathews, and Carraro 1986; Williams et al. 1983). Each of these studies showed statistically significant reductions in crashes resulting in driver fatalities, ranging from 5 to 28 percent, for the age groups affected by the law during the study periods. In addition, most of the individual States in each study's subject pool showed significant reductions in such crashes.
• Driver Fatality Crashes (Selected States). Five studies evaluated the effects of increasing the drinking age on crashes resulting in driver fatalities in individual States (Emery 1983; Florida Department of Community Affairs 1983; Hingson et al. 1983; Lillis 1984; Schroeder and Meyer 1983). All five studies showed reductions in the number of crashes attributable to higher drinking ages, ranging from a 1-percent reduction in "driver-fatal" crashes in Massachusetts to a 35-percent reduction in "driver had been drinking-fatal" crashes in New York. The results were statistically significant in four of the studies.
• Driver Fatality or Injury Crashes. Four studies in four States evaluated the effects of increasing the drinking age on crashes in which the driver was either killed or injured (Florida Department of Community Affairs 1983; Lillis 1984; Wagenaar 1981, 1987). Each study found reductions among those affected by the law after the minimum drinking age was increased. Results were significant for Florida, Michigan, and New York, ranging from about 10 percent in New York to a 28-percent reduction in Michigan.
• Driver Injury Crashes. One study examined the effects of increasing the drinking age on crashes resulting in driver injury (Florida Department of Community Affairs 1983). This study observed a statistically significant net reduction of approximately 2 percent during the study period among age groups affected by the law.
• Driver-Involved Crashes. Four studies evaluated the effects of increasing the drinking age on driver-involved crashes in three States (Klein 1981; Maxwell 1981; Schroeder and Meyer 1983; Wagenaar 1981). All four studies found reductions in driver-involved crashes among the age groups affected by the laws. The reductions in Illinois, Maine, and Michigan were significant, ranging from 9 percent in Illinois to 22 percent in Michigan.
• Total Crash Fatalities. One study evaluated the effects of increasing the minimum legal drinking age on total crash fatalities (Saffer and Grossman 1985). Analysis of national data during a period after many States increased their minimum legal drinking ages revealed a 7-percent average reduction in fatalities in States with higher drinking ages.

Practice Evidence Reviewed for Approach 1

Through the Pennsylvania Liquor Control Board, the State of Pennsylvania is engaged in activities designed to control the density of alcohol outlets, support the prevention of alcohol abuse, disseminate information on the responsible use of alcohol and responsible alcohol sales and service, and establish partnerships with other State agencies and organizations. The State's results include the following:

• In 1992, 66 youth DWI arrests were made in Pennsylvania, in contrast to the national average of 134 such arrests.
• In 1992, a smaller proportion of youths were killed in alcohol-related vehicle collisions in Pennsylvania than in most other States.
• Underage youth are served alcohol less frequently in Pennsylvania than in open States, where alcohol is sold in drug, grocery, and other retail stores.

The Under 21 Enforcement Project was designed by the Town of Yorktown Police Department to enforce regulations prohibiting the sale of alcohol to persons under 21 years of age, to heighten public awareness of alcohol abuse by youth, to aggressively enforce DWI regulations for underage youth, and to increase the accuracy of police reports concerning motor vehicle collisions. The project targeted premises suspected of selling alcohol to underage youth, issued press releases on violations and enforcement efforts, patrolled youth gathering places, screened drivers, and provided information regarding ABC laws and reporting techniques. The Under 21 Enforcement Project proceeded in three steps. First, letters were mailed to merchants whose clerks requested identification during purchase attempts by underage patrons. Second, patrol supervisors received special orders requesting aggressive investigations of complaints about underage drinking at private residences. Third, inservice police training sessions were held to advise patrol officers of the importance of accurately completing vehicle collision reports. The following outcomes were noted during the project's 1-year grant period:

• The proportion of alcohol-related violations by underage youth that were not associated with driving declined from 62.7 to 50 percent.
• The proportion of all drivers between 16 and 20 years of age who were involved in vehicle collisions declined from 18.3 to 17 percent.
• The proportion of DWI arrests among underage youth increased from 8.1 to 8.4 percent.

Cops in Shops was an enforcement and media campaign designed to diminish alcohol purchases by underage youth in Las Cruces, New Mexico. In the enforcement component, plainclothes police officers posed as employees at seven licensed establishments but did not handle transactions. Customers who appeared to be under 25 years of age and who attempted to purchase alcohol had their identification referred to the police officer in the store. Those patrons who were underage were refused alcohol sales. As each underage customer left the store, another officer stationed outside issued the youth a citation. The intervention was conducted at seven stores in rotation, generally between 5:00 p.m. and 9:00 p.m. A total of 50 hours of enforcement occurred during the course of a month. The media campaign included the use of signs at retail alcohol stores announcing that the police would verify the age of alcohol purchasers and would cite underage youth who attempted to purchase alcohol. Public service announcements were developed for use by local television stations and the local cable company, and advertisements were produced for the city newspaper and the student newspaper at New Mexico State University. Two weeks before the Cops in Shops program was implemented, 75 public service announcements were aired on cable television channels including Cable News Network, Lifetime, MTV, Nickelodeon, and USA Network.

Among the 95 total purchases and purchase attempts made before the program's implementation and the 70 total purchases and purchase attempts made during its implementation, the following was reported:

• Four fake identifications were used in purchase attempts by underage youth before implementation, while none were used during implementation.
• Three purchase attempts were aborted when patrons were asked for identification before implementation, while six were aborted after implementation.
• Transactions between youths and adults in the parking lots outside stores during the enforcement program led police to issue 10 citations to adults for purchasing and distributing alcohol to underage youth and 4 citations to underage purchasers.

Research Evidence Reviewed for Approach 2

A study by Cook and Moore (1993) analyzed the effect of beer excise taxes on rates of homicide, rape, assault, and robbery. They used annual data from 1979 through 1988 on these violent crimes in the 48 contiguous United States to generate three sets of closely related estimates: (1) on the effects of drinking on violent crime; (2) on the effect of the beer tax on drinking; and (3) on the effect of the beer tax on violent crime. The study's findings included the following:

• A 10-percent increase in per capita consumption was associated with the following increases in violent crimes: homicide, 0.87 percent; assault, 5.85 percent; rape, 6.47 percent; robbery, 9.13 percent.
• A 100-percent increase in alcohol taxes was associated with the following reductions: assault, 0.26 percent; homicide, 0.32 percent; alcohol consumption, 0.48 percent; robbery, 0.87 percent; rape, 1.32 percent.

Saffer and Grossman (1987) examined the relationships between beer taxes and the minimum legal drinking age on motor vehicle fatalities among youth aged 15 to 24 in the 48 contiguous United States from 1975 through 1981. One goal of this study was to investigate the responsiveness of motor vehicle death rates of youth aged 15 to 24 to variations in the cost of beer as reflected by differences in State excise tax rates on beer. Another goal was to examine the effect of an increase in the legal drinking age on youth motor vehicle deaths. Using a time-series cross-sectional design with historical cohorts, study data were used to predict the effects on fatalities of a uniform drinking age of 21 and of policies that fixed the Federal beer tax in real terms to its 1951 level or taxed the alcohol in beer at the same rate as that in spirits. Among the findings were the following:

• Higher beer taxes were associated with lower death rates from motor vehicle collisions for youth aged 15 to 17, 18 to 20, and 21 to 24 years.
• Higher legal drinking ages were associated with lower death rates from motor vehicle collisions only for youth aged 18 to 20.
• It was projected that the enactment of a uniform drinking age of 21 in all States would have reduced the number of youth aged 18 through 20 killed in motor vehicle collisions by 8 percent from 1975 through 1981.
• It was projected that policies fixing the Federal beer tax in real terms since 1951 would have reduced the number of lives lost in fatal collisions by 15 percent, whereas a policy that taxed the alcohol in beer at the same rate as that in spirits would have reduced the number of lives lost by 21 percent. A combination of the two tax policies would have caused a 54 percent decline in the number of youth killed.

Using a cross-sectional design and a historical cohort, Grossman, Coate, & Arluck (1987) conducted an econometric analysis to develop estimates of the responsiveness of alcohol use by youth aged 16 to 21 years to variations in the price of alcohol. Estimates of the effects of price were also compared with estimates of the effect on alcohol use by youth of raising the minimum legal drinking age. The analysis was based on the first National Health and Nutrition Examination Survey, which was conducted from 1971 through 1974. The study's findings included the following:

• A 10-cent increase in the price of beer would reduce the number of youth aged 16 through 21 years who drink by about 11 percent, the number who drink two or three times weekly by 8 percent, and the number who consume as many as three to five cans of beer on a typical drinking day by 15 percent.
• A 30-cent increase in the price of spirits would reduce the number of youth who drink spirits by 23 percent and the number who consume at least three to five drinks on a typical drinking day by 27 percent.
• Increases in both the minimum legal drinking age and the prices of beer and distilled spirits were associated with decreases in beer consumption.
• Increases in the price of beer were associated with decreases in wine consumption.
• Increases in the price of distilled spirits were significantly associated with decreases in their consumption.
• Overall, the study suggested that the prices of beer and distilled spirits and the minimum legal age for their consumption are likely to affect alcohol use by youth.

Similarly, using a time-series design with a historical cohort, Coate and Grossman (1985) conducted an econometric analysis to develop estimates of the responsiveness of alcohol use by youth aged 16 to 21 years to variations in the price of alcohol. The study was also designed to compare estimates of the effects of price with estimates of the effects of an increase in the legal drinking age on alcohol use by youth. The analysis was based on the second National Health and Nutrition Examination Survey, which was conducted from 1976 through 1980. The study's findings included the following:

• A Federal policy that simultaneously taxes the alcohol in beer and spirits at the same rates offsets the erosion in the real beer tax since 1951; fixing the beer tax in real terms would reduce the number of youth who drink beer four to seven times weekly (approximately 11 percent of all youth) between 32 and 35 percent. Such a Federal policy would also reduce the number of youth who drink beer one to three times weekly (approximately 28 percent of all youth) by 24 to 28 percent.
• Increases in alcohol prices and the minimum legal drinking age were not significantly associated with decreases in consumption of spirits for any other outcome category.
• The inverse effects of alcohol prices and minimum legal drinking age were not limited to reductions among youth who consume beer less than once per week. The numbers of youth who consume beer infrequently (one to three times per week) or frequently (four to seven times per week) fall more in absolute or percentage terms than does the number of infrequent drinkers (less than once weekly) when price or the drinking age rises.
• Overall, the study suggests that the frequency of the consumption of beer, the most popular alcoholic beverage among youth, is inversely related to the real price of beer and to the minimum legal age for its purchase and consumption.

In an econometric analysis, Levy and Sheflin (1983) examined the importance of cross-price effects and provided estimates of the elasticity of demand according to price. Using annual U.S. data from 1940 through 1980, they estimated a standard-demand equation with per capita alcohol consumption broken down by real per capita income and the real prices of different types of alcoholic beverages. Two variables were used to measure consumption: the volume of absolute alcohol consumed and expenditures on alcoholic beverages. The period of analysis precisely coincided with a period during which consumption rose steadily; it has been falling since 1981. The study's findings included the following:

• From 1940 through 1980, per capita alcohol consumption increased by an average of 1.2 percent a year according to a volume measure and by 0.95 percent per year according to an expenditure measure.
• In the same period, real per capita disposable income increased at an annual rate of 1.9 percent, while the relative price measure for alcohol beverages indicated a 1.4 percent decline for the volume measure.
• Provided real per capita disposable income continues to increase, the income elasticity of 0.5 suggests that each 2-percent increase in per capita income will lead to a 1-percent increase in per capita alcohol consumption. Price policies can be a very effective means of reducing alcohol consumption. A price increase of 2.5 percent, for example, would reduce per capita alcohol consumption by 1 percent. A 2.5-percent increase in price would be necessary to offset a 2-percent increase in income; increases of the same magnitude in price and income would approximately offset each other.

Using an econometric analysis, Ornstein and Hanssens (1985) estimated the impact of a variety of control measures on the demand for distilled spirits and beer. The analysis considered variables such as economic and sociodemographic levels, retail availability, price controls, advertising restrictions, and whether a State had a monopoly on alcohol sales. The sample for distilled spirits was based on 255 observations made in the 50 States and the District of Columbia from 1974 through 1978. For beer, price data were missing for Hawaii and Alaska and were available for the other 48 States only for 1976 through 1978, yielding an initial sample of 147 observations. The authors used a pooled time-series cross-sectional sample. Among their econometric findings and predictions were the following:

• Control laws are either unrelated to distilled spirits consumption, as in the cases of minimum legal drinking age and Sunday sales, or are related but with very low elasticities, as in the cases of resale price maintenance and print and billboard price advertising. This suggests that control laws that affect price have the greatest impact on consumption.
• Control laws with the strongest relationship to beer are those regarding minimum legal drinking age and Sunday sales.
• The main determinants of interstate differences in per capita consumption of distilled spirits are price, income, and interstate travel — not differences in alcohol-control laws.
• Price and income are far less elastic for beer than for distilled spirits, implying that control laws that affect price will have a relatively lesser effect on the consumption of beer than on that of spirits.
• Price intervention through tax increases, resale price maintenance, or bans on price advertising will have a much larger effect on the consumption of spirits than on that of beer.

Using a time-series design with a historical cohort, Wette et al. (1993) conducted an econometric analysis to determine the relationship between price and consumption of beer, wine, and spirits from 1983 through 1991. In particular, the study was designed to assess changes in alcohol consumption in New Zealand as a function of increases in the price of alcoholic beverages. Among the findings were the following:

• For beer, wine, and alcoholic beverages (but not spirits) and for total alcohol, an increase in price led to a decrease in alcohol consumption.
• Income was not a significant determinant of consumption, suggesting that recent decreases in disposable income in New Zealand have not reduced alcohol consumption.
• The price of one alcoholic beverage (e.g., wine) had little impact on the consumption of another alcoholic beverage (e.g., spirits), with one exception: an increase in the price of beer appeared to lead to a decrease in the consumption of both beer and wine.

Happy-hour discount drink promotions were banned in Ontario, Canada, in December 1982. Taking advantage of this natural experiment, Smart and Adlaf (1986) conducted measures before and after the ban and analyzed its impact on drinking behavior, impaired-driving charges, and overall alcohol sales in Toronto. An observational component of the study examined alcohol consumption by patrons in five drinking establishments before and after the ban. A baseline evaluation was conducted 2 days before the banning of happy hours; a post-ban evaluation was conducted 4 weeks later. The baseline and post-ban evaluations focused on variables such as the number of patrons at the bar on arrival, the number of tables with one patron, the number of tables with two or more patrons, and the number of patrons arriving during the observation period. Observers selected two tables at which each patron's consumption would be observed. Data were collected on the type and number of alcoholic beverages each patron consumed at the observed tables. Aggregate data on alcohol sales were collected for the study period (October 1984 through February 1985) and for a comparison period (October 1983 through February 1984). Data on the daily number of impaired- driving charges were collected for the study period (November 1984 through January 1985) and for a comparison period (November 1983 through January 1984). The analysis produced the following findings:

• A decline in impaired-driving charges could not be causally attributed to the independent effect of the ban on happy hours.
• An analysis of the aggregate data on alcohol sales indicated no significant trends over a similar period.
• No significant differences in alcohol consumption before and after the ban were noted at either the individual or the aggregate level.

Babor et al. (1978) examined the effect on drinking behavior of a controlled setting for the happy hour discount drink policy. In this quasi-experimental study, conducted in the research ward of a hospital, 34 male subjects were given free access to alcohol during a 20-day period. Half of the subjects could purchase alcohol for a single price, while a matched group was offered a 3-hour period in the afternoon during which the price was reduced by 50 percent (the happy hour condition). Subjects were identified as having a history of either casual or heavy drinking. A secondary purpose of the study was to determine whether alcohol consumption during a period of reduced cost would have a kindling or "priming" effect on subsequent drinking and whether casual and heavy users would be affected differently by this variable. Evaluations conducted during and after drinking occasions focused on the time of day, type of beverage, and consecutive number of drinks. BAC's were estimated through multiple daily Breathalyzer tests. The study's findings included the following:

• Casual drinkers in the happy hour condition consumed twice as much alcohol as their counterparts in the nonhappy hour condition.
• Casual drinkers in the nonhappy hour condition had very few incidents of intoxication. In contrast, casual drinkers in the happy hour condition showed a greater frequency of intoxication during the afternoon.
• Heavy drinkers in the happy hour condition drank twice as much alcohol as their counterparts in the nonhappy hour condition.
• Heavy drinkers in both conditions experienced greater incidences of intoxication during the later hours of the day, and heavy users in the happy hour condition experienced intoxication more often than did heavy users in the nonhappy hour condition.
• Reinstatement of the standard alcohol purchase price effectively suppressed happy hourrelated increases in consumption.

Research Evidence Reviewed for Approach 3

Employing a quasi-experimental pre- and posttest design, Gliksman et al. (1993) evaluated the knowledge, attitudes, and behaviors of alcohol providers at four types of bars (roadhouses, hotel lounges, neighborhood taverns, and skid-row bars) immediately before and after participation in the Addiction Research Foundation's Server Intervention Program. Pre- and posttests included scored behavioral observations of servers' responses to actors who posed as patrons and engaged in six typical inappropriate behaviors. The study found the following about server knowledge, attitudes, and behavior:

• After training, servers' knowledge increased regarding alcohol and its effects, their own legal obligations, and appropriate and inappropriate practices for alcohol service.
• After training, servers exhibited more appropriate responses to the problem behaviors of customers. However, their responses were less seriously inappropriate rather than as appropriate as the responses presented during the training.
• Before training, servers were known to encourage, promote, and orchestrate heavy drinking via activities centered on the rapid consumption of alcoholic beverages, such as pouring drinks down patrons' throats. After training, servers did not participate in such activities.
• Servers scored high on measures of positive attitudes toward the training, perception of management support in implementing server intervention, and sufficient knowledge about alcohol, its effects, and the signs that indicate a need for server intervention.

McKnight (1991) evaluated the effect of server intervention education and the effect of various situational variables on RBS program effectiveness. A 6-hour server education program was administered to 876 servers and 203 managers at 100 licensed drinking establishments in eight States. Each establishment was visited three to four times by staff observers who exhibited visible signs of intoxication. These observers also visited 138 comparison establishments in the same communities. In this quasi-experimental study with matched comparison groups, baseline and postprogram evaluations focused on content knowledge, attitudes regarding responsible beverage service, and frequency of adherence to responsible service practices. Among the findings were the following:

• After server training and policy changes, server intervention was observed in only 20 percent of the 1,590 visits, and only 7 percent of the visits resulted in service termination.
• In contrast to the comparison sites, the experimental sites reported positive treatment trends.
• Overall improvements were noted in knowledge of, attitude toward, and policies involving responsible alcohol service at the experimental sites, but not at the comparison sites.
• When server intervention was examined State by State, positive changes were reported in only five of the eight States. The differences were related to prior intervention levels, type of establishment, and business volume.

A study by McKnight and Streff (1994) examined the effects of enforcing the law prohibiting alcohol service to intoxicated patrons on service to patrons, DWI rates, and enforcement costs and benefits. The study used a quasi-experimental design with a nonequivalent comparison group and pre- and postintervention measurements. Among its findings were the following:

• In the county receiving the intervention, refusal of service to observers posing as patrons ("pseudopatrons") and simulating intoxication rose from 17.5 percent to 54.3 percent, eventually declining to 41 percent. In a comparison county, service refusal rose from 11.5 percent to 32.7 percent.
• In the county receiving the intervention, the number of DWI arrests involving patrons exiting bars and restaurants declined from 31.7 percent to 23.3 percent. No changes in DWI arrests among patrons exiting bars and restaurants occurred in the comparison county.
• Service refusals were related to volume of business and the number of intoxicated patrons in an establishment at the time of observation. DWI arrest numbers were related to the nature of the establishment's clientele, policies, and practices.
• The estimated benefits from an alcohol service enforcement program capable of a one-fourth reduction in DWIs from bars and restaurants range from $90 (direct monetary savings) to $260 (total savings) for each dollar invested in enforcement.

Saltz (1987) compared customers' consumption of alcoholic beverages in a U.S. Navy enlisted club that had a server intervention program with that in a similar Navy enlisted club that did not have such a program. The server intervention program included an 18-hour training course for management and staff and a revision of established alcohol service policies. The revised policies promoted the consumption of nonalcoholic beverages and food, the overt delaying of alcohol service to nearly intoxicated patrons, and the discontinuation of selling beer in pitchers. In this study, which used a quasi-experimental design with a nonequivalent control group, baseline and posttest interviews were conducted with 603 randomly selected customers 2 months before and after implementation of the program. Measures focused on customer characteristics and rates of alcohol consumption, specifically whether the patrons were intoxicated according to a system that examined the number of drinks consumed, the duration of time spent consuming them, and the patron's weight. The study's findings included the following:

• At the intervention site, the rate of intoxication for males was 32 percent at baseline and 15 percent posttest.
• For females, the rate of intoxication dropped from 5 to 2 percent at the intervention site.
• While the likelihood of customer intoxication was reduced at the intervention site, the absolute consumption and the rate of consumption of alcoholic beverages were unaffected by the server intervention program.

Holder and Wagenaar (1994) evaluated a natural experiment that started in 1986 when the State of Oregon introduced legislation mandating a training program for all alcohol servers as well as for all owners and managers of establishments that serve alcohol. Using an interrupted, quasi-experimental, time-series design, the authors analyzed the effect of the training program on the number of single-vehicle nighttime collisions in Oregon that resulted in injuries or fatalities and in which alcohol was involved. The study found that the mandatory training program significantly reduced the incidence of single-vehicle nighttime collisions in the State. This effect increased over the first 3 years of the program as the proportion of trained servers increased.

Mosher et al. (1989) conducted a study with two components: outcome-based and community-based. The first component was a quasi-experimental, outcome-based study designed to examine the efficacy of an intensive, community-based RBS training program aimed at reducing intoxication rates among patrons of participating establishments. In the second component, formative and process evaluations were devised to establish community and industry support for a responsible beverage service training program. The findings of the outcome-based study included the following:

• A risk-assessment survey of servers identified problematic practices in both of the sites studied (Santa Cruz County and Monterey County, California).
• A significant increase in knowledge was noted after the training among trainees in one of the study sites (Santa Cruz County) but not in the other (Monterey County).
• Patrons leaving the experimental sites were equally likely to be intoxicated as those leaving the control sites.
• Patrons who were "censored" by servers for their drinking experienced significant reductions in the risk of becoming intoxicated in Santa Cruz County but not in Monterey County. Reduced probability of intoxication was associated with higher patron weights and female gender. Increased probability of intoxication was associated with spending more time in the drinking establishment.

The findings of the community-based study included the following:

• Preliminary pre- and posttests of managers and servers initially showed no gains in RBS knowledge. After the measurement instrument was revised to correct ambiguities and unclear wording, however, significant gains were noted.
• Project activities had greater success in Santa Cruz County, where there was strong preexisting interest in alcohol issues, than in Monterey County, where interest in such issues was low and the hospitality industry was poorly organized.
• Task force activities, media coverage, and project responses generated community and industry support for establishing a responsible beverage service program in Santa Cruz County, whereas similar efforts in Monterey County failed.

Russ and Geller (1987) evaluated the impact of an alcohol server intervention program called Training for Intervention Procedures by Servers of Alcohol. Using a quasi-experimental design, the authors evaluated pre- and posttest measurements of an intervention group and a comparison group. The study examined the type and frequency of server interventions, the number of drinks consumed by pseudopatrons, and the BAC's of exiting pseudopatrons who drank three alcoholic beverages per hour for 2 consecutive hours. Among the study's findings were the following:

• Pseudopatrons who had been served by trained personnel had substantially lower BAC's than did those served by untrained personnel, with average levels of 0.059 and 0.103 percent, respectively. None of the pseudopatrons served by trained personnel exceeded the legal limit for intoxication (a BAC of 0.10 percent); by contrast, nearly 45 percent of the pseudopatrons served by untrained personnel had BAC's at or above the legal limit for intoxication.
• Servers who had undergone training intervened more frequently and in a different way than did untrained servers. Upon serving a customer's first drink, trained servers were more likely than untrained ones to ask for identification and to offer food or water. After a customer's fourth drink, trained servers were more likely than untrained servers to offer food or water, to delay service, or to make a driving-related comment; untrained servers initiated no intervention at this point.
• The interventions trained servers made most frequently upon serving a customer's first and second drinks were checking identification and offering food or water. Interventions made at the fourth, fifth, and sixth drinks consisted of offering food or water, delaying service, and making driving-related comments.

Before 1983, the State of Texas had no statutory provisions allowing or prohibiting liability for negligent service of alcoholic beverages. There were thus no case-law precedents clearly establishing that retail establishments licensed to sell alcohol were liable for damages resulting from irresponsible serving practices. In 1983 and 1984, however, two cases involving service to intoxicated individuals drastically changed the liability situation in Texas. These two cases (El Chico Corp. v. Poole and Joleemo v. Evans) progressed through the State courts, resulting in a landmark 1987 Texas Supreme Court decision allowing common-law actions against licensed alcohol outlets to proceed. Wagenaar and Holder (1991a) examined the effects of this sudden change in exposure to legal liability for alcohol servers on the major ultimate outcome of interest: the frequency of nighttime, single-vehicle, injury-producing traffic crashes. Using a multiple time-series, quasi-experimental design, data on injuries occurring between 1978 and 1988 were analyzed. The analysis controlled for broader crash trends reflected in data from other States as well as for the effects of other major policy changes in Texas, such as those raising the legal drinking age, strengthening drunk driving laws, and requiring the use of safety belts. The study's findings included the following:

• Final estimates of the parameters in the time-series model revealed significant reductions in the frequency of nighttime, single-vehicle, injury-producing traffic crashes after the January 1983 and November 1984 filings of major court cases involving server liability.
• The rate of traffic crashes decreased by 6.5 percent immediately after the 1983 case was filed.
• The rate of traffic crashes decreased by 5.3 percent after the 1984 case was filed.

Research Evidence Reviewed for Approach 4

Holder and Wagenaar (1990) conducted interrupted time-series analyses of apparent consumption of spirits in Iowa over a 20-year period (1968 to 1989) to evaluate the effect on consumption of the privatization of Iowa's State retail monopoly on distilled spirits sales for off-premise consumption. Their study's findings included the following:

• During the month in which distilled spirits sales were privatized, sales of spirits rose by 9.5 percent, a statistically significant increase.
• Even though there was a corresponding decline of 12.1 percent in wine sales and no change in beer sales, privatization of distilled spirits retail sales yielded a net increase in total alcohol consumption in Iowa. The observed decline in wine sales followed a 70.3 percent increase after Iowa's privatization of wine sales in 1985.
• No changes were found in spirits sales in States bordering Iowa during the period when spirits sales were privatized.
• Significant stocking effects were found for the month before spirits sales privatization (21.8 percent) and for the first month in which private sales were allowed (15.2 percent).
• The privatization of distilled spirits sales was associated with an apparent net annual increase in consumption of 24,000 liters of pure ethanol in Iowa.

Using an interrupted, time-series design, Wagenaar and Holder (1991b) examined natural experiments brought about when Iowa and West Virginia eliminated their State monopolies and privatized retail wine sales. The study was designed to assess whether the privatization of retail wine sales in Iowa and West Virginia increased the apparent consumption of wine, affected sales of beer and distilled spirits, or affected alcohol consumption in other States. Among the findings were the following:

• Compared with 5-year baseline averages, net annual increases in ethanol use per month after the privatization of retail wine sales in Iowa and West Virginia were 28,602 and 25,234 liters, respectively.
• Sales of beer decreased by 3.1 percent in Iowa and increased by 12.0 percent in West Virginia during the first 18 months after privatization.
• Sales of distilled spirits significantly decreased, by 5.4 percent in Iowa and 13.8 percent in West Virginia, during the first 18 months after privatization.
• Sales of wine increased by 93 and 48.2 percent in Iowa and West Virginia, respectively, during the 18 months after privatization.
• No identifiable decrease in wine sales was reported in other States concomitant with the significant increases in Iowa and West Virginia.

Fitzgerald and Mulford (1992, 1993) conducted three cross-sectional, self-reporting surveys of individuals aged 18 years and older at three points during a 4-year period (1985, 1986, and 1989). In July 1985, the State of Iowa abandoned its monopoly on wholesale and retail sales of bottled wine, and in March 1987 gave up its monopoly on retail but not wholesale sales of bottled spirits. The study was designed to determine whether a sudden and dramatic increase in the availability of wine and distilled spirits in Iowa resulted in an increase in alcohol consumption and alcohol-related problems among individuals aged 18 years and older. Among the findings were the following:

• No dramatic changes occurred in the rank order of drinking frequencies in different places (bars, restaurants, homes, others' homes, sports events, and outdoor recreation sites) from one year to another.
• No statistically significant changes occurred in any of the problem indicators of heavy or problem drinking.
• No statistically significant changes occurred in problems due to drinking associated with privatization of either wine or spirits sales.
• No statistically significant changes occurred in the frequency of drinking alone or with different companions (spouse, other relative, coworkers, or close friends).
• The number of off-sale outlets for bottled wine and spirits substantially increased after privatization.

Mulford and Fitzgerald (1988) conducted a study designed to evaluate the effect of increasing the number of off-sale wine outlets in Iowa on rates of consumption and purchase of alcohol and on heavy drinking and problem drinking. The study included a baseline self-reporting survey of adults aged 18 years and older from February through April 1985 (4 months before the wine outlet increases) and a follow-up survey in April 1986, about 9 months after the privatization of wine sales. The study also examined monthly State alcohol sales statistics for table wines, beer, and distilled spirits from 1983 through August 1987, and for wine and distilled spirits coolers from July 1985 through August 1987. The study's findings included the following:

• Expanded alcohol availability through retail outlets did not result in overall lasting sales increases, greater alcohol consumption, or increased numbers of heavy and problem drinkers.
• Monthly wine sales surged immediately after the increase in outlets, and the prevalence of self-reported wine purchasers was higher when measured 9 months later. Per capita consumption and purchases were virtually unchanged, however, and monthly wine sales had dropped back to preprivatization levels.
• Self-reported consumption and sales of beer and distilled spirits were unaffected by the increase in wine outlets.
• Self-reported total alcohol consumption was unchanged, as were rates of heavy drinking and problem drinking.

Mulford, Ledolter, and Fitzgerald (1992) used a multiple time-series design to analyze changes in monthly sales of wholesale wine and spirits (an approximation of apparent consumption) in relation to the privatization of State-run retail sales of wine and spirits and wholesale wine sales. In addition, the researchers conducted a supplementary quasi-experimental analysis in which Iowa sales were compared with national sales. This study was designed in part to resolve, through a time-series analysis, the contradictory findings of previous research suggesting that privatization did not increase alcohol consumption (Mulford and Fitzgerald 1988) or that it permanently increased wine consumption by 93 percent (Wagenaar and Holder 1991b). The study's findings included the following:

• Distilled spirits sales increased in February and March 1987 due to stocking effects. Sales increased by 26 percent in September 1985 due to a retailer-announced Federal tax increase in October. The following month, sales decreased by about the same amount. Spirits sales decreased at a slightly faster rate in Iowa than they did nationally from 1980 to 1989.
• Overall, privatization temporarily increased sales of wine in 1985 and of spirits in 1987, but had no lasting impact.
• The effect of privatization on wine sales was temporary. Sales diminished over the next 2 years and gradually returned to preprivatization levels.

In an ecological study, MacKinnon, Scribner, and Taft (1995) analyzed data on alcohol availability and related problems in unincorporated areas and 84 cities in Los Angeles County. The sources of data were as follows: ABC reports from 1970 to 1991; California State Department of Justice police crime reports from 1970 to 1990; census data for 1970, 1980, and 1990; Los Angeles County Health Department data on causes of death from 1973 through 1987; data on alcohol-related vehicle collisions from 1970 to 1990; and legislative measures of community involvement in alcohol availability issues in 1990. The study was designed to assess the impact of alcohol availability (i.e., alcohol outlet density) on alcohol-related civil outcomes (e.g., arrests for drunken driving and public drunkenness) and alcohol-related health outcomes (e.g., deaths from liver cirrhosis as well as alcohol-related traffic-crash injuries and fatalities). Among the study's findings were the following:

• Off-sale alcohol availability was substantially related to rates of arrests for public drunkenness and disturbing the peace.
• On-sale alcohol availability was substantially related to rates of death from liver cirrhosis and to arrests for drunken driving, public drunkenness, and disturbing the peace. A 1-percent increase in on-sale alcohol availability was associated with a 0.35- to 0.51-percent increase in deaths from liver cirrhosis and a 0.51- to 0.66-percent increase in arrests for drunk driving.
• Retail alcohol outlet density was found to be substantially and positively related to several alcohol related problems.
• The density of on-sale outlets was more strongly related to alcohol problems than was the density of off-sale outlets.

In an ecological study, Gruenewald et al. (1995) analyzed data on alcohol availability and alcohol-related traffic crashes across 102 areas of four communities in California. This study replicated the alcohol-related crash component of the study by MacKinnon, Scribner, and Taft (1995), but used improved statistical analyses. The sources of data were ABC records on premise types and locations; California State Highway Patrol data on locations of single-vehicle nighttime crashes; census data on population densities; self-reported data on population demographics, consumption patterns, and drinking and driving patterns among the subject populations (aggregated into the 102 target areas); and extensive measures of the traffic flow and road network characteristics of the selected areas. The study was designed to assess geographical relationships between the physical availability of alcohol (e.g., at bars and restaurants) and alcohol-related traffic crashes within and across geographic areas of the communities. The study provided a statistical means of correcting for biases that arose in the use of these ecological data. The study's findings included the following:

• Across the studied communities, a 1-percent increase in restaurant densities was associated with a 0.06-percent increase in alcohol-related crashes. Within specific community areas, a 1-percent increase in restaurant densities was associated with a range of changes in crash rates from -0.59 percent to 0.35 percent.
• Retail outlet density (measured along the roadway network) was substantially and positively related to rates of alcohol-related traffic crashes.
• The density of restaurants was most strongly related to alcohol-related traffic crashes. This was attributed to the use of alcohol in these relatively unique drinking-and-driving environments.
• The effect of outlet density on traffic crashes was shown to radiate across community areas away from the areas of greatest density toward areas of lesser density.
• The effects of outlet density on traffic crashes was shown to depend on the location of the change in density within the community areas.

Research by Gruenewald and Ponicki (1993, 1995) involved a series of analyses of cross-sectional time-series data from 38 States over 12 years (1975 to 1986). The purpose of the studies was to evaluate the impact of changes in alcohol sales and the physical availability of alcohol on fatalities from single-vehicle nighttime collisions. Among the findings were the following:

• Collision rates were strongly determined by per capita alcohol use, particularly by the use of beer and spirits.
• Fatality rates from single-vehicle nighttime collisions were not directly related to the physical availability of alcohol, independent of a number of related economic and demographic variables.
• The effect of beer sales on collision rates exceeded those of either spirits or wine sales. This effect was sustained when the estimates were adjusted for market share, where the impact of beer proved 2.75 times that of spirits and 3.33 times that of wine.

Taking advantage of a natural experiment, Fillmore and Wittman (1982) conducted a study to observe the responses of two university communities to a change in a State law removing off-sale restrictions on the sale of wine and beer near campuses. To assess whether an increase in the availability of alcohol would affect student drinking practices, questionnaires were administered to students enrolled at the University of California at Berkeley and the University of California at Davis shortly after the law took effect in 1979 and again 2 years later. During this time, the number of retail outlets located near the campuses substantially increased. The study's findings included the following:

• Between 1979 and 1981, the percentage of respondents who had not consumed alcohol in the preceding week increased from 24 to 31 percent among students residing outside of Berkeley and from 26 to 37 percent among those in geographical areas with no more than three alcohol outlets. Among students living in areas with higher outlet densities, the percentage who had not consumed alcohol in the preceding week decreased by 3 to 1 percent, regardless of whether the number of outlets in the study area had increased between 1979 and 1981.
• Three months after the repeal of off-sale restrictions, the city of Berkeley experienced a 15 percent increase in the number of gallons of alcoholic beverages sold at retail outlets. This increase was concentrated within a 1-mile area surrounding the campus. Almost 4,000 gallons of the increase and three of the five new off-sale outlets were located within a half-mile of the campus.
• Three months after the repeal, the city of Davis experienced a 140 percent increase in the number of gallons of alcoholic beverages sold at retail outlets and an increase in the number of off-sale outlets from 15 to 21.
• The mean number of alcohol purchases decreased marginally at both sites between the two measurement points.
• The percentage of respondents in Davis who had consumed 14 or more beers in the preceding week increased from 14 percent in 1979 to 23 percent in 1981.
• There were no major changes in the frequency or the volume of student drinking from the baseline to the followup 2 years later.

To assess the effects of legislation permitting alcohol sales at a sports arena, West et al. (1993) conducted a quasi-experimental pre- and posttest observational study to examine drinking, intoxication, and verbally and physically abusive behavior at the Maple Leaf Gardens in Ontario, Canada, during Toronto Maple Leaf hockey games. Ten observers attended 15 games to make pretest (1992) and posttest (1993) observations; a questionnaire was also mailed to a sample of season ticket holders. In addition, a survey of arena security staff was conducted to determine their experiences with and attitudes toward alcohol consumption before and after the legislation was enacted. The study's findings included the following:

• Men comprised 75 percent of the patrons and committed 95 percent of the reported alcohol-related incidents. Individuals between the ages of 20 and 30 years were responsible for most of the inappropriate behaviors.
• The most common form of disruptive behavior observed during the 1992 and 1993 games was intoxication followed by verbal abuse.
• The pattern of disruptive incidents changed. In 1992, when alcohol sales were illegal and many patrons did their drinking before entering the arena, there were more incidents of intoxication before the hockey games. The proportion of intoxicated patrons during the third periods of the games increased from 9 percent of all intoxicated patrons in 1992 to 28 percent in 1993.
• There was no increase in the number of incidents of inappropriate behavior after the introduction of alcohol sales.

Practice Evidence Reviewed for Approach 4

In 1993, the Alberta Liquor Control Board in Alberta, Canada, decided to privatize retail and wholesale sales and the distribution of alcohol. In an examination of this natural experiment, Laxer et al. (1994) studied the impact of the privatization on consumers, alcohol-related social problems, and government finances. A comparison was made of the previously existing public monopoly system of alcohol control and regulation, the current stage of private independent retailing, and the likely future system of alcohol distribution through private groceries and other chains. The Alberta study's findings included the following:

• Criminal offenses against liquor stores, including robbery and breaking and entering, increased after privatization.
• Hours of operation were extended at many alcohol outlets after privatization.
• In July 1994, retail alcohol prices were 5.9 percent higher than before privatization.

In 1982, an Ontario provincial law permitted the sale of beer at sporting events. To evaluate the effect of the new law, Fisher and Single (1983) conducted a pre- and postlegislation study in which alcohol-related behavior was observed at Toronto's Exhibition Stadium during Blue Jays baseball games. At 7 games before and 11 games after the legislation was enacted, observations were recorded regarding drinking and drug consumption, alcohol- and drug-related incidents, and crowd composition. The study's findings included the following:

• After enactment of the legislation, the crowd comprised more adolescent and young adult males and fewer women.
• After enactment of the legislation, there were more incidents of rowdiness, pushing and shoving, harassment, fights, vomiting, and disruptive fans being asked to leave.
• During the 11 games played after the legislation was enacted, the mean rate of consumption ranged from 0.1 to 2.8 beers per person and averaged 0.9 beers per person. Twelve percent of all the observed drinkers at the games consumed five or more beers.
• The games at which there was the greatest amount of intoxication and rowdiness were the doubleheaders.

The Province of Ontario offers municipalities the option to sell alcohol. Increasingly, Ontario municipalities are taking this option at municipally owned or managed facilities such as arenas, stadiums, community centers, performance halls, senior centers, parks, beaches, and sports fields, at all of which alcohol may be served during specially licensed social and recreational events. Because such events are usually staffed by inexperienced volunteers who occasionally serve participants to the point of intoxication, the potential exists for alcohol-related fights, injuries, vandalism, and impaired driving — any of which can lead to license suspensions, criminal charges, police involvement, and civil litigation.

A policy model for responding to such problems was therefore developed and disseminated to Ontario's municipalities. The model included recommendations such as making lists of the facilities at which alcohol can and cannot be served; imposing restrictions on serving patrons to the point of intoxication; providing transportation for intoxicated patrons; prohibiting alcohol service to underage patrons; making food and low-alcohol drinks available; and developing a list of required management practices. These recommended practices included providing adequate door and floor supervision, limiting the number of drinks per patron, serving drinks in paper cups, and requiring server training. In a cross-sectional study, Gliksman et al. (1995) surveyed recreation directors, facility managers, and other appropriate individuals from all of Ontario's municipalities to evaluate the development of local alcohol policies and their impact on alcohol-related problems and rentals of public facilities. The study's findings included the following:

• Among the 477 municipalities with eligible facilities, 59 percent did not have a policy in place and 8 percent had an informal policy that was either implicit or mentioned in city bylaws.
• Among the remaining municipalities, 22 percent had a formal written policy in place and 12 percent had a formal policy in development. The average time the policies had been in place was 25 months, while the average time to complete a policy was 10 months.
• Among the 162 municipalities with a formal policy in place or in development, 40 percent were in townships, 35 were in towns, 7 percent were in villages, and 19 percent were in cities. These percentages are consistent with the proportions of the types of communities.
• Among the 162 municipalities with formal policies in place or in development, only 7 percent did not receive assistance from outside sources and developed their policies independently. The rest received assistance from sources such as the Addiction Research Foundation Community Program Offices (73 percent); provincial or municipal police representatives (50 percent); Public Health Unit staff (41 percent); Liquor License Board of Ontario inspectors (22 percent); and other nonprofit and private sources.
• Among the 107 municipalities that had adopted formal policies, 49 percent reported that facility rentals had remained the same or had increased, 22 percent reported a decrease, and 7 percent were unsure.
• Of these 107 formal policies, 47 percent had been in place for a year or less. Twenty-six percent had been in place for 13 months to 2 years, 16 percent for 25 months to 5 years, and 11 percent for more than 5 years.
• Of the 107 municipalities that had adopted formal policies, 44 perceived a reduction in alcohol-related problems and 7 did not. Of the remainder, 34 said it was too soon to know or that they had not yet implemented their policies, 14 were uncertain, and 8 said they had never experienced problems.
• The major problem areas in which the 44 respondents indicated a noticeable reduction were fighting, vandalism, and underage drinking. Significant decreases were also seen in the numbers of police interventions and public complaints.

During the mid-1970s, the city of Thunder Bay, Ontario, experienced problems related to the inappropriate and excessive use of alcohol at city-owned parks and other recreational facilities. Some citizens demanded the removal of alcohol from such facilities, whereas others wanted more alcohol-related functions for social enjoyment and for fundraising activities. The city developed a social control policy for the use of alcohol on city-owned property and in facilities operated by its Parks and Recreation Department. The process of implementing this policy included a promotional component designed to increase awareness of and subsequent voluntary compliance with the policy. Gliksman et al. (1990) used before- and after-policy questionnaires to examine the implementation of the policy and to evaluate the impact of the promotional campaign on Thunder Bay residents. The responses from random samples of Thunder Bay residents were compared with those of comparison samples drawn from a nonintervention community. The study's findings included the following:

• After the policy was implemented, the attitudes of subjects in the intervention group became less liberal regarding legal control over the sale and use of alcohol, while those of the control group did not change over time.
• After the policy, subjects in both the intervention and control groups became more positive than they had been before about renting facilities where alcohol would not be available.
• After the policy, subjects in both the intervention and control groups reported greater intentions than they had before to attend functions at facilities where alcohol would not be served.
• Attitudes among the intervention group regarding underage drinking became less positive over time, whereas those of the control group changed very little.
• Subjects in the intervention group became less positive regarding the use of alcohol at recreational facilities.
• Subjects in the intervention group became more knowledgeable about the laws concerning alcohol use in general and the Thunder Bay policy in particular.
• Subjects in the intervention group demonstrated an increase in their intentions to comply with alcohol-control laws, whereas those in the comparison group demonstrated little change in such intentions over time.

In a nonexperimental study, Bjor, Knutsson, and Kuhlhorn (1990) examined interventions related to the celebration of Midsummer's Eve in Borgholm, a small Swedish town on the island of Oland in the Baltic Sea. Midsummer's Eve is a popular event in Sweden that has traditionally been associated with alcohol consumption and open sexuality. The partying often begins the day before Midsummer's Eve and lasts for 3 days. During these celebrations, large crowds gather and threaten public order with disorderly and drunken behavior. At Borgholm in 1987, a baseline was established through participant observation. That celebration of Midsummer's Eve began with drinking parties at camping sites outside of town and progressed to public intoxication within the town. Ultimately, thousands converged upon the town, creating a crush of people and a tensely charged atmosphere. Fearing a riot, police marched into the crowd and arrested individuals for intoxication and disorderly conduct. An intervention was introduced for the next year's celebration.

First, an alcohol-rationing rule was established under which each visitor was limited to one bottle of distilled spirits and 24 cans of beer. Second, participants were prohibited from congregating at the campgrounds or in the parking lots near the center of the city. Finally, the celebration was declared a public event, under which conditions individuals are forbidden to possess alcohol or knives, and alcohol consumption was forbidden in public places. These new rules were publicized via the Swedish media. After the 1988 Midsummer's Eve celebration, interviews were conducted with 34 key people, including both those who had an interest in promoting the celebrations (generally local businessmen) and those who did not. Among the findings regarding the 1988 intervention were the following:

• Fewer people went to Borgholm for Midsummer's Eve than in the previous year.
• The troubles experienced the previous year did not materialize.
• Celebrants who attempted to enter campgrounds were sent away. Many appeared to leave the event entirely, some going to a different location at which public intoxication and disorderly conduct were observed.
• The number of people arrested for intoxication dropped from 131 in 1987 to 120 in 1988.
• The number of people arrested for disorderly conduct dropped from 42 in 1987 to 15 in 1988.
• In interviews with 34 key individuals, nearly all reported that the negative consequences associated with drunkenness, accidents, public disorder, and vandalism at the celebration were "better" in 1988 than in 1987.

Research Evidence Reviewed for Approach 5

In New South Wales, Australia, a December 1979 state regulation allowed the sale of alcohol at taverns on Sundays from noon to 10:00 p.m. Using a quasi-experimental, multiple time-series design with a comparison area, Smith (1987) evaluated the effect on vehicle collisions of Sunday sales of alcohol at taverns. Because Sunday sales were already permitted at sports and returned-servicemen's clubs, which have large memberships in Australia, the objective of the study was to determine the effect of further increases in Sunday sales of alcohol. A comparison of Sunday vehicle collisions in New South Wales in the 3 years before and the 2 years after the regulation went into effect indicated the following outcomes:

• A 22-percent increase in fatal collisions between noon and 11:50 p.m.
• A 28-percent increase in collisions between 6:00 p.m. and 11:50 p.m. resulting in injury but not in hospital admission.
• A 21-percent increase in collisions between 6:00 p.m. and 11:59 p.m. resulting in injury but not in hospital admission.
• A 12-percent increase in tow-away collisions between 6:00 p.m. and 11:59 p.m.
• A 16-percent increase in collisions resulting in death or injury between 6:00 p.m. and 7:59 p.m.
• A 38-percent increase in collisions resulting in death or injury between 8:00 p.m. and 9:59 p.m. and a 22-percent increase in such collisions between 10:00 p.m. and 11:59 p.m.

Using a quasi-experimental, multiple time-series design, Smith (1988a) examined the effects on vehicle collisions of a legislative change that introduced flexible tavern operating hours in the Australian state of Tasmania. Prior to the legislation, tavern operating hours in Tasmania were 10:00 a.m. to 10:00 p.m. Mondays through Saturdays and noon to 8:00 p.m. on Sundays. After the introduction of flexible outlet operating hours, taverns could remain open for as few or as many hours as they chose, with certain exceptions: they could not open from 5:00 a.m. to noon on Sundays or from 8:00 p.m. to midnight on Sundays or Good Friday. Thus, the study was designed to evaluate whether later closing times (and hence later drinking hours) might decrease the number of vehicle collisions from 10:00 p.m. to midnight and increase the number of collisions from midnight to 6:00 a.m. The study found the following to be the case after the introduction of flexible tavern hours:

• No significant changes occurred in the number of vehicle collisions resulting in casualties between 6:00 p.m. and 10:00 p.m.
• The rate of vehicle collisions between 10:00 p.m. and 6:00 a.m. increased by nearly 11 percent.
• The rate of vehicle collisions resulting in casualties increased from 49 to 56 percent.

In the Australian state of Victoria, a February 1966 regulation changed the times at which taverns closed on Mondays through Saturdays from 6:00 p.m. to 10:00 p.m., allowing bar patrons to consume alcohol in taverns later than before. To evaluate the effect of this regulation on vehicle collisions resulting in casualties, Smith (1988b) analyzed data on all such collisions that occurred in Victoria before and after the regulation went into effect. In addition, in December 1966 it became compulsory for the police to use a Breathalyzer test to determine the BAC's of drivers suspected of drinking. Thus, this quasi-experimental, multiple time-series study was also designed to evaluate whether compulsory tests for suspected drinking drivers had an effect on the number of collisions. Among the study's findings for the year after the introduction of the 10:00 p.m. tavern-closing regulation were the following:

• A considerable and immediate increase occurred in the number of collisions resulting in casualties between 10:01 p.m. and 2:00 a.m.
• Annual collisions resulting in casualties between 6:01 p.m. and 2:00 a.m. were 11.5 percent higher during the 8-hour experimental period than during the 16-hour control period.
• During the second year, the increase in collisions resulting in casualties was apparently offset by the compulsory use of the Breathalyzer. The number of casualty-causing collisions between 10:01 p.m. and 2:00 a.m. decreased by 9.2 percent.

In the Australian city of Brisbane, an April 1970 regulation allowed the sale of alcohol on Sundays from 11:00 a.m. to 1:00 p.m. and from 4:00 p.m. to 6:00 p.m. Using a quasi-experimental, multiple time-series design, Smith (1988c) evaluated the effect of Sunday alcohol sales on vehicle collisions, including reported casualties and property damage. The study's findings included the following:

• In the 2 hours after the period from 4:00 p.m. to 6:00 p.m., the number of collisions resulting in property damage increased by 85 percent and the number resulting in casualties by 129 percent.
• On Sundays between noon and 1:50 p.m. (after the period from 11:00 a.m. to 1:00 p.m.), the number of collisions resulting in property damage increased by 52 percent.
• Overall, the introduction of Sunday alcohol sales had a close temporal relationship with the increase of collisions resulting in casualties or property damage in the Brisbane area, but not in the rest of the state of Queensland.

Nordlund (1985) examined the effects of closing wine and spirits shops in Norway on Saturdays. In an experimental study, wine and spirits shops were forced to close in six Norwegian towns receiving the intervention, whereas shops were allowed to remain open in six matched control towns. Measures included police records on arrests for drunkenness, reports of drunkenness, and reports of violent incidents; admissions to alcohol detoxification programs; and alcohol sales. The study's findings for towns receiving the intervention included the following:

• Sales of pure alcohol decreased by 3 percent and those of beer by 6 percent compared with the year before the intervention.
• The number of arrests for drunkenness decreased on Saturday nights and in the early hours of Sunday mornings.
• The number of reports to the police regarding drunkenness decreased markedly on Saturdays and early Sundays.
• The number of reports to the police regarding violence increased considerably on Fridays and the early hours of Saturday mornings.
• The number of admissions to alcohol detoxification facilities in Oslo decreased considerably more for weekends than for weekdays.

In Sweden, all sales of export beers, wines, and spirits take place in special State-owned retail liquor stores. In May 1981, the Swedish Riksdag decided to close these stores on Saturdays between June and September of that year. In a quasi-experimental study, Olsson and Wikstrom (1982) examined whether these Saturday closings would have an impact on total alcohol consumption, drunkenness, public order, and certain types of crime. Data were collected primarily from official statistics, generally covering the study period (June through September 1981) and the same period during the prior year. Among the findings on outcomes compared with those in the same period during the prior year were the following:

• The mean number of domestic disturbances during the study period was substantially lower on Saturdays (25 percent lower) and Sundays (26 percent).
• The mean number of indoor assaults decreased by 18 percent on Saturdays and by 12 percent on Sundays, but increased by 4.3 percent on Tuesdays, Wednesdays, and Thursdays.
• The mean number of outdoor assaults diminished substantially on Saturdays (by 25 percent) and Sundays (by 11 percent).
• The mean number of police interventions against intoxicated individuals during the study period decreased by 39 percent on Saturdays and Sundays.
• The mean number of police interventions against public disturbances during the study period decreased substantially on Saturdays (by 17 percent) and Sundays (by 21 percent).
• The number of disturbances on the Stockholm Underground diminished substantially on Saturdays (by 19 percent) and Sundays (by 11 percent).
• It was estimated that the number of detentions of intoxicated persons decreased by about 5,000, the number of domestic disturbances by 900, the number of assaults by 600, and the number of "other disturbances" by 700.

Research Evidence Reviewed for Approach 6

As cited in the research evidence reviewed on approach 1, Hingson et al. (1996) evaluated the effectiveness of the Saving Lives Program, a community effort to organize multiple city departments and private citizens to reduce alcohol-impaired driving, alcohol-related driving risks, and traffic deaths and injuries. In each of the six program communities a full-time coordinator from the mayor's or city manager's office organized a task force of concerned private citizens, organizations, and officials representing various city departments. The communities developed initiatives such as media campaigns, speeding and drunk driving awareness days, "speedwatch" telephone hotlines, police training, SADD chapters, alcohol-free proms, beer keg registration, increased surveillance of liquor outlets, preschool education programs, and training for staff at hospitals and prenatal clinics. The intervention cities were compared with the rest of Massachusetts and with five cities that prepared high-quality proposals that were not funded. This quasi-experimental study evaluated the impact of the intervention on traffic crashes and injuries, safety belt use, vehicle travel speeds, and driving after drinking.

Monitoring of crashes was based on data from the Department of Transportation's Fatal Accident Reporting System. These data were collected 5 years before and 5 years after the start of the program. Additional data came from the Massachusetts Registry of Motor Vehicles and were collected 4 years before and 5 years after the start of the program. Five annual direct observation surveys of speeding and safety belt use were conducted 1 year before and 4 years after the start of the program; these surveys involved using radar from unmarked cars for detecting speeding vehicles at randomly selected roadways at intervention and comparison cities, as well as direct observation at signalized road intersections and freeway off-ramps using a stratified random sampling procedure. To assess program awareness, beliefs about police enforcement, and the frequency of driving after drinking, four independent cross-sectional, random-digit-dial telephone surveys of 15,188 young people aged 16 to 19 years and adults aged 20 years and above were conducted in intervention cities and the rest of the State between 1988 and 1993.

Compared with the rest of the State and with the 5 years before the intervention, the following findings were noted in the Massachusetts intervention cities during the 5 years of the program:

• Fatal crashes declined by 25 percent and alcohol-related fatal crashes by 42 percent.
• The proportion of vehicles observed speeding was reduced by 43 percent and the use of seat belts increased 17 percent.
• The proportion of teenagers who drove after drinking declined by 40 percent.

The Midwestern Prevention Project was a 6-year longitudinal study of community-based efforts to prevent alcohol, cigarette, and marijuana use among adolescents. The project included mass media programming, a school-based educational program for youth, parent education and organizing, community organizing, and health policy programming components, all of which were introduced sequentially into communities over the 6-year period. Pentz et al. (1989) evaluated the effects of the project components administered to the initial 1984 cohort of sixth- and seventh-grade adolescents in Kansas City communities from September 1984 through January 1986. The overall research design for the Midwestern Prevention Project included a quasi-experimental study in Kansas City and a randomized experimental study in Indianapolis.

The project components administered from September 1984 through January 1986 consisted of a 10-session youth education program on skills for resisting drug abuse, 10 homework sessions involving active interviews and role-playing with parents and family members, and mass-media coverage. The Midwestern Prevention Project's effectiveness was evaluated through annual assessments of adolescent drug use in schools assigned either an immediate-intervention or a delayed-intervention control condition. Measurements focused on demographic characteristics, drug use, and psychosocial variables related to drug use. Immediately before administration of the questionnaire, a breath test was used to measure carbon monoxide concentrations in subjects' expired air to increase the accuracy of subsequent self-reports. The study's findings included the following:

• Analyses of 42 schools indicated that the prevalence rates of alcohol (and marijuana and cigarette) use were lower at the 1-year followup in the immediate-intervention condition schools than in the delayed-intervention condition schools.
• These lower prevalence rates were demonstrated with and without controlling for race, grade, socioeconomic status, and urbanism (11 percent versus 16 percent for alcohol use).
• The net increase in the prevalence of drug use at immediate-intervention schools was half that in delayed-intervention schools.

Giesbrecht, Pranovi, and Wood (1990) evaluated the effectiveness of a community-based project to reduce alcohol consumption among heavy drinkers. They analyzed whether a change in the proportion of heavy drinkers has a measurable impact on the overall distribution of alcohol consumption. They also examined the interaction between heavy and moderate drinkers and assessed the impact of a drinking-management intervention on heavier drinkers that offered them an educational and counseling program. Behavioral change was also pursued through more general educational measures, influence on local alcohol policies, server intervention programs, and training workshops. In addition, project staff joined several community health and social service committees.

The community-based, multisite study was conducted with an intervention community and two control communities, the last being the "pure" control community. Community 1 (the intervention community) and community 2 (the regular control community) were surveyed twice; the second survey was done 24 months after the baseline. During the interval between surveys, an 18-month, multifaceted alcohol reduction project was implemented in community 1. Surveys conducted in the midwinters of 1984 and 1986 assessed individual alcohol consumption through various indicators of typical and heavy drinking and attitudes toward alcohol use, prevention, and treatment. The surveys evaluated the volumes and frequency of alcohol use; the context and nature of drinking occasions; the pressures to drink or to stop doing so; the health and social consequences related to alcohol use; attitudes and norms about drinking and alcohol-related behavior; views on alcohol addiction and treatment; and demographics. Among the study's findings were the following:

• Members of the subject group reduced their consumption of alcohol during the course of their participation in the study. The percentage of the client group consuming more than 14 drinks per week declined between the week before the program began and the last week for which data were available for each client. The median dropped from 20 to 9 drinks per week, and the mean from 15 to 7 drinks per week during this interval.
• Overall, the percentage of subjects who consumed more than 14 drinks per week declined from 54 percent before the program began to 43 percent after the first counseling session and to 21 percent in the last week for which data were available.
• The number of opportunities to promote the intervention through public service announcements, speaking engagements, training sessions, committee work, and consultations on alcohol issues in the community increased over time.
• A local server intervention committee was established to set guidelines for serving alcoholic beverages in licensed premises and at special occasions.

Practice Evidence Reviewed for Approach 6

The aim of the New Zealand Community Action Project (Casswell and Gilmore 1989; Stewart and Casswell 1993) was to examine the relative effect on alcohol-related community activities and public attitudes of a mass-media campaign, with or without the participation of alcohol-focused community organizers. The project's main objective was to increase the level of public support for alcohol-control policies such as restrictions on availability and advertising. An intermediate objective was to increase the amount of alcohol-related material in the local print media and on local radio programs.

The project involved print media advertising campaigns and community organizing strategies. The latter involved community activists who worked closely with local media and alcohol councils, which included representatives from community institutions such as social welfare, police, traffic enforcement, and alcohol treatment agencies. The demonstration project, which took place between October 1982 and February 1985, used a quasi-experimental design in which six provincial cities were roughly matched to form two groups of three cities. Each group included a comparison city in which no planned alcohol-prevention activities took place, a target city that was exposed to a mass-media campaign largely directed toward reducing high-volume drinking by young males through television and radio commercials, and an intensive-intervention city exposed to the same mass-media campaign along with a community organizing effort. The study's findings included the following:

• In the four cities in which mass media only or intensive intervention were employed, local newspaper coverage of alcohol-related issues such as moderation and social policies increased.
• More cautious attitudes regarding the harmfulness of alcohol increased in the intervention but not in the reference cities.
• Support increased for alcohol policies regarding age limits and limits on alcohol sales in supermarkets in the intervention but not in the reference cities. Support declined for alcohol policies regarding advertising, price, and general restrictions on alcohol sales in the reference but not in the intervention cities.
• Surveys of both key-participant and man-on-the-street interviews suggested that the program had a positive effect in the media-only and intensive-intervention cities compared with the reference cities.

In response to an increase in the numbers of transients and chronic street drinkers along with associated crimes and other alcohol-related problems in Portland, Oregon, merchants, citizens, and city bureaus entered into a partnership agreement to address those and other problems affecting the community. The resulting Central Precinct Neighborhood Alliance included numerous neighborhood associations that combined their efforts with those of the Portland Police Bureau to establish a community-policing action plan. The alliance and its efforts were designed to reduce alcohol abuse among problem street drinkers and thereby reduce alcohol-related criminal activity. Their efforts included the following:

• Merchants, community organizations, and the police entered into a Community Policing Partnership Agreement.
• Merchants, police officers, and community organizations worked together to identify the alcoholic beverages associated with street drinking.
• Portland police vigorously enforced city ordinances and State statutes related to liquor law violations.
• The partnership agreement included an understanding that sales of domestic beers and malt beverages in containers larger than 16 ounces would be phased out and then discontinued.

The Minors in Night Clubs public education campaign was organized by the Cape May, New Jersey, County Council on Alcoholism and Drug Abuse in response to an action by the Wildwood City Council, acting as the city's ABC authority. Under the action, which was approved for a trial period only, underage youth aged 18 to 20 years would be permitted access to drinking establishments during normal operating hours but would not be permitted to drink alcohol there. The City Council was expected to pass a law making the ordinance permanent after the trial period ended. The campaign's efforts included the following:

• Alcohol-free activities were established for underage youth, as were an adventure course and peer leadership groups.
• The local Hotel/Motel Association collaborated with the Prosecutor's Working Group on Substance Abuse to resolve problems associated with underage youth drinking alcohol on hotel or motel premises.
• The local Municipal Alliance sponsored a campaign against underage drinking ("We 4 for 21").
• With broad support from local organizations and agencies, the Cape May County Council on Alcoholism and Drug Abuse mounted a vigorous media campaign against the law. In response to public outcry, the City Council decided to terminate the trial period and not to make the ordinance permanent.

The Community Coalition for Substance Abuse Prevention and Treatment involved grassroots community activism organized to influence the alcohol licensing and regulation process in Los Angeles. The goal was to reduce the density of alcohol outlets and to diminish the violence associated with high outlet densities. This activism intensified after the 1992 race riots, during which 400 alcohol outlets were damaged or destroyed, prompting the city to inaugurate a fast-track process to reduce the bureaucratic red tape associated with rebuilding businesses. The coalition's aims included giving a voice to citizen opposition to the licensing of new alcohol outlets; ensuring that those outlets would not be rebuilt without community input; urging the Los Angeles City Council to exempt alcohol outlets from ordinances designed to facilitate rapid rebuilding after the riots; and developing a plan to support alcohol outlet owners who agreed to convert their property to alternative land uses. The coalition's efforts produced the following results:

• After 15 days of pressure, the city enacted an ordinance with amendments that excluded alcohol outlets from the fast-track rebuilding process.
• Citizens were mobilized to force the city to withdraw proposed amendments that would have exempted grocery stores and other businesses occupying 1,400 or more square feet from the outlet-overconcentration ordinance review process.
• Community pressure prompted the city to establish a Mayor's Task Force on Problem Liquor Stores, through which citizens had the opportunity to air their concerns to the mayor, the planning commissioner, and the zoning administrator.
• Only 56 rebuilding permits were granted as of 1994 for the 400 alcohol outlets that were burned or damaged during the 1992 riots.

The Escondido Community Alcohol Planning Project (ECAPP) was a prevention program designed by the Community Prevention Planning Demonstration Project of the Institute for the Study of Social Change at the University of California at Berkeley. ECAPP consisted of a city-based planning committee that received technical assistance regarding the documentation of alcohol availability in retail, public, and social environments; the identification of strategies to address problems in these environments; the selection of intervention policies and programs; and the implementation of interventions.

The project found the following regarding implementation, integration, and institutionalization of alcohol prevention policies:

• The City Council of Escondido approved the city's participation in ECAPP and assigned a council member to the planning group.
• A representative of the Community Services Department was invited to join the local group responsible for reviewing alcohol license applications.
• The council endorsed ECAPP's mission, adopted prevention policies, directed staff to develop implementation strategies, and incorporated elements of ECAPP's initiatives into its annual program.
• By the second year, alcohol prevention was incorporated into the city's organizational policies and procedures. Examples included limiting alcohol outlets through passage of an interim ordinance, prohibiting banners that included alcohol or tobacco sponsorship or advertising, and limiting conditional-use permits for outlets near schools and residential areas.
• A request to serve beer at the annual Escondido street fair was denied.

ECAPP also reported the following proofs of community support:

• The city developed and implemented prevention-oriented special-permit guidelines governing the use of alcohol at city parks and recreation areas, and extended the ban on alcohol use to city lakes and recreation areas.
• The city developed a draft alcohol outlet permit ordinance.
• The city developed a database of problem and prevention indicators for prevention planning, monitoring, and evaluation.
• The city developed and approved a policy for the California Center for the Arts in Escondido regarding substance abuse.
• The city incorporated alcohol policies into its criteria for designating apartment complexes as "Drug-Free Zones."
• A resolution supporting beer-keg registration was passed and forwarded to the State legislature.

The Association for Responsible Alcohol Control (ARAC) was established by a group of Latino and other activists in San Jose, California. The activists had heard a report by a representative of the Prevention Division of the Santa Clara County Bureau of Alcohol Services describing the relatively high density and rapid growth of alcohol outlets in Latino neighborhoods as well as the disproportionately high arrest rates for alcohol-related offenses among Latinos. ARAC members focused on changing alcohol policies at the city level, beginning with a concerted effort to pass a law that would require new businesses seeking permission for off-site alcohol sales to apply for a CUP, a process that requires a public hearing and gives residents a voice in decisions regarding alcohol availability in their neighborhoods. ARAC's activities produced results including the following:

• The alcohol industry mounted little opposition to the proposed CUP law.
• The San Jose City Council voted unanimously in favor of the CUP law.
• ARAC created media interest by issuing news releases concerning the relationship between alcohol accessibility and alcohol-related problems.
• ARAC convinced local merchants to discontinue selling inexpensive fortified wines.
• The county health department created a Responsible Hospitality Council to examine alcohol use in public and private settings and to develop guidelines for alcohol use on county property.