What Colleges Are Doing About Student Binge Drinking

Public health is increasingly making use of a variety of behavior change strategies to support a comprehensive, systems-wide approach for prevention. Examples of multidimensional and systems-change approaches to prevention are found in programs targeting high-risk behaviors, such as heavy episodic or "binge" drinking, smoking, risky sexual behavior, and violence among youth and young adults.^1,2 Systems-change programs share an assumption that behavior is shaped by many factors, including those associated with individuals and their environments.³ Consequently, administrators and prevention specialists use strategies that span educational and environmental efforts. In preventing alcohol abuse, systems-change approaches are coupling education about alcohol with efforts to alter relevant policies, reduce supply, and change media and marketing practices that shape or reflect norms (Weitzman ER, Wechsler H. Reducing high-risk drinking and related harms using a social ecological approach: Research overview for the "Matter of Degree" program evaluation. Unpublished manuscript).^4-6

Today, efforts to prevent binge drinking in colleges are advanced through a variety of programs that range from unique single-school projects to cross-college initiatives and some surveys of programming.⁷ Newer initiatives include the counter-advertising campaigns of the National Association of State Universities and Land Grant Colleges, private philanthropic efforts (eg, social ecology interventions supported by the Robert Wood Johnson Foundation's A Matter of Degree Program⁸ ), federal initiatives (eg, the US Department of Education Fund for Improvement of Secondary Education programs), and prepared curricula from various sources.

Determining the efficacy of the various efforts to reduce binge drinking and related harms calls for rigorous program evaluation. Although a number of descriptions of prevention programs exist, detailed research findings are rare. Until quality evaluation data are available, only a comprehensive examination of existing prevention programs will allow administrators to assess strengths, weaknesses, and gaps in their approaches to campus alcohol problems. In this article, we offer a national picture of current efforts to prevent and curtail binge drinking at more than 700 colleges and universities and categorize these efforts into distinct areas in a systems-change approach. The purposes of our study were to

• facilitate a "taking stock" of current efforts, options, and program trends;
• provide a comprehensive overview of current prevention efforts nationally, with a summary of institutional capabilities for prevention, using various indicators of program evaluation, technical support, and campus/community cooperation;
• assess whether implementation of prevention programs differed by the degree of perceived drinking problems on each campus and by other college characteristics.

METHOD

Sample of Colleges

We surveyed presidents (or individuals they designated) at 4-year, American liberal arts universities and colleges. Initially, we selected 1105 schools from the American Council of Education's list of accredited institutions of higher education. Later, we dropped 141 institutions from the list because they were considered inappropriate for our study (ie, institutions that did not offer baccalaureate liberal arts degrees, such as seminaries, theological schools, professional institutes, and schools outside of the United States or US territories). Seven hundred thirty-four schools responded to our questionnaire, a response rate of 76.1%. Respondents were presidents (10%), designated student affairs (57.3%) and health promotion staff, including persons from wellness, counseling, alcohol and other drug education programs (21.9%), and other administrative officers (10.9%).

The Questionnaire

We developed an eight-page questionnaire based on measures developed in the Harvard School of Public Health (HSPH) College Alcohol Study⁹ and the HSPH's A Matter of Degree program evaluation (Weitzman ER, Wechsler H, Reducing high-risk drinking and related harms using a social ecological approach; Research review for the "Matter of Degree" program evaluation. Personal communication , 1999). The questionnaire consisted of 31 questions, some with both forced-choice and open-ended parts. We asked respondents to rate the perceived severity of alcohol-abuse problems among students on their campuses, to specify approaches their colleges used to address binge drinking, and to describe prevention resources. Thirteen questions were open-ended, designed to elicit descriptive information about the types of alcohol education and prevention programs offered to students, as well as the range of campus alcohol-control policies and their enforcement.

Mailing and Response Rate

We mailed questionnaires to college presidents in February 1999, with an accompanying cover letter. The presidents' options were to have the survey filled out on-line for electronic submission to a secure HSPH server or to return the completed survey by regular mail or fax. Over the next 2 months, we sent nonrespondents a second and a third questionnaire and an accompanying cover letter notifying them that the response deadline had been extended. Responses were voluntary and respondents were not required to provide their names. By the end of March, 41.7% of the questionnaires had been returned; an additional 33.5% arrived by the end of May; and the remaining 24.8% arrived by the end of August. More than 60% of the respondents submitted their responses on-line, and 39.4% used regular mail or fax to return the survey.

Data Analysis

We used the SAS for Unix systems¹⁰ to carry out all statistical analyses and used chi-square analyses to compare the respondents ' reported perceptions of alcohol problems on their campuses with their reports about alcohol policies, sanctioning, comprehensiveness of alcohol education programs, and approaches to address high-risk drinking. Qualitative data describing characteristics of prevention programming were postcoded to reflect major typological categories of prevention programming, using a framework developed for another project. Descriptive data were then included as quantifiable measures within the frequency analyses.

RESULTS

Sample Characteristics

The 734 schools in the sample represented a cross-section of US institutions of higher education located in 50 states and 2 US territories (Table 1). Twenty-eight percent were in the North Central region , 26.6% in the Northeast, 35.6% in the South, and 9.3% in the West; 40% were public institutions and 60% were private. We divided school enrollment into three categories: large schools with more than 10 000 students (16.9%), medium-sized schools with 5001 to 10 000 students (13.9%), and small schools with fewer than 5000 students (69.2%). Approximately three fifths of the schools (61.4%) were in an urban or suburban area, and two fifths (38.6%) were in a small town or rural setting. Fifty-six percent had a religious affiliation, 2.2% enrolled only women, and 7.2% were commuter institutions with no residence halls. Sixty percent had fraternities or sororities or both, and 94% participated in intercollegiate sports.

Prevention Initiatives Nationally,
by College Characteristics

Wide variations existed in types of college prevention programming and preventive interventions associated with college characteristics (Table 2 a b). Almost all schools (97%) provided general alcohol education programs, although proportionately fewer schools with small enrollments and institutions with a religious affiliation did so. Despite high levels of these general alcohol education programs, fewer schools provided more targeted education (eg, to Greek-affiliated students or athletes, even though these groups are known to represent above-average alcohol consumption and alcohol-related problem levels on many campuses^11,12).

Although approximately 60% of the schools surveyed had fraternities/sororities, only 67% of those institutions targeted Greeks. The same situation occurred in 94% of the schools that had athletic programs, with only 59% providing programs that targeted athletes. Programs for Greek-affiliated or athlete groups were more likely to be found at large (p < .001), public (p < .05, and p < .001, respectively), and secular (p < .001) schools. Less common were education programs comprising outreach to students' families (about one third did so) or high school students (fewer than one fifth). These efforts, again, were more likely to be found at large, public, secular schools than at the smaller schools or institutions with a religious affiliation.

Restrictions on the supply of alcohol were prevalent at most of the schools we surveyed. About 98% of the administrators at residential campuses reported that their schools or communities prohibited keg delivery to dormitories. Administrations at highly competitive schools (in terms of admission) were less likely to prohibit keg deliveries, as were schools in suburban or urban settings (p < .01 and p < . 0 5, respectively). Similarly, a majority of the schools reported restrictions on alcohol sales at intercollegiate sports events. Large, public, secular, and suburban/urban schools were all significantly less likely to impose controls on sales (p < .001 for all comparisons) than were smaller, religious, and rural schools.

Keg delivery to fraternity or sorority houses was prohibited at 87% of the surveyed schools, most commonly at schools in the Northeast (p < .01) and less competitive schools (p < .05). We found similar patterns for prohibitions of alcohol at home intercollegiate sports events and tailgate parties at three quarters or more of the respondents' institutions, with significant differences by college type for these efforts. Compared with small schools, the colleges with large enrollments were about one quarter less likely to ban alcohol at intercollegiate sports events and about half as likely to do so at home tailgate events, resulting in a clear gradient by school size (p < .001 for both comparisons). Thus, the smaller the school, the more likely that restrictions on alcohol at sporting events were to be in effect. Secular schools were also substantially less likely to use these controls than institutions affiliated with a religious denomination (p < .001 for both comparisons). Perhaps the highest use of supply-side restrictions was evident among the small group (n = 16) of women's colleges, whose administrators reported complete (100%) use of restrictions for three out of five survey items in this area.

Restrictions on alcohol advertising varied widely by institutional characteristics. Although approximately 90% of the schools reported they restricted alcohol advertisements at home sporting events, only 1 in 2 (51%) reported prohibiting advertisements for off-campus bars or clubs in campus newspapers or on bulletin boards. Public schools were considerably less likely than their private counterparts to limit advertising (p < .001 in both cases), whereas schools in the Northeast and those with a religious affiliation were more likely to do so (p < .001 in both cases).

Among the 681 residential campuses (93% of the total sample) we surveyed, approximately two thirds offered alcohol-free dormitories and living spaces. These arrangements were more likely to exist at public (p < .05), large (p < .001), northeastern (p < .01), highly competitive (p < .001), and secular (p < .001) campuses. Women's colleges were markedly less likely to provide these arrangements, although the small size of the sample limited the significance of the analysis.

The last category of initiatives represents institutional investments in prevention. Rather than prevention efforts per se, these initiatives reflect infrastructure to monitor and respond to students' problem drinking. Seventy-seven percent of the college administrators we surveyed reported that their schools had a designated individual who was in charge of issues related to alcohol and other drug abuse. Institutions that were small (p < .001), private (p < .001), in the West (p < .01), and had a religious affiliation (p < .001) were least likely to have such an alcohol and other drug specialist.

Approximately three fifths (60%) of the respondents reported that they had a task force, working group, or coalition charged to work in this area. Small (p < .001), private (p < .001), southern (p < .01), noncompetitive institutions (in terms of admissions) (p < .001), and commuter schools (p < .001) were least likely to have a task force.

Somewhat more than half (55%) of the surveyed respondents regularly measured the extent of the campus drinking problem through some type of survey mechanism: large (p < .001), public (p < .001) schools, and those in the Northeast (p < .05) were most likely to report monitoring alcohol use, as were more competitive (p < .001), residential (p < .001), and secular (p < .001) schools. Of those schools that regularly measured the extent of student high-risk drinking and associated harms, only half (46.7%) reported these findings back to students. Twenty percent incorporated those findings into "social norming" campaigns designed to influence student perceptions about drinking norms on campus. A minority (40%) of respondents reported that they had a designated specialist or office to evaluate preventive interventions on campus. Small (p < .001), private (p < .01), and religious institutions (p < .001) were least likely to have an in-house evaluator.

Forty percent of the respondents reported their institutions had cooperative agreements with community agencies to share data, resources, or strategies. These arrangements were most common at large (p < .001), public (p < .01), and residential campuses (p < .05). Last, only about one quarter (24%) of the respondents reported meeting regularly with neighbors or community groups to address issues related to student drinking. Large (p < .001), public (p < .01), northeastern, more competitive (p < .001), residential, (p < .05), and secular (p < .001) colleges were more likely to report involvement in community exchanges.

Prevention Initiatives, by Degree
of Perceived Problem

The degree to which college administrators perceived student alcohol abuse as a problem was positively associated with the comprehensiveness of institutional education and prevention programs (Table 3). The respondents who reported that student alcohol abuse was a problem or a major problem on their campuses were considerably more likely than other respondents to report that their institution provided general education programs about alcohol (p < .001), targeted programs for high-risk groups (p < .05 - p < .001), and initiated outreach to families and high school students (p < .05 - p < .001). In addition, those respondents who reported that alcohol abuse was a major problem for students at their institution were 14% less likely to have an institutional policy prohibiting alcohol at home tailgate parties (p < .001) than those who reported less severity of alcohol-abuse problems.

A strong positive association existed between perceived severity of student alcohol abuse and the institutional investment in prevention. Respondents who perceived student alcohol abuse to be a major problem were from 60% to 250% more likely to support implementation of various institutional actions than were their peers who reported low perceived severity of alcohol-related problems (p < .001 in all cases).

COMMENT

College administrators across the country reported that their schools engaged in a wide variety of efforts designed to prevent binge drinking. General educational interventions-long the hallmark of alcohol and other drug prevention ^7,13,14-were most common and were present in virtually all surveyed schools. Proportionately fewer schools engaged in social-norms campaigns designed to change student perception of normative drinking on campus, had educational efforts that targeted high-risk groups, or had programs that extended beyond the campus to reach families or high school students. Not surprisingly, schools that implemented the more targeted educational initiatives were those whose administrators perceived that student alcohol abuse was a major problem.

A majority of the respondents reported that their schools restricted student access to alcohol by prohibiting keg delivery to dormitories or fraternity/sorority houses and imposing sales controls or bans on tailgate parties. Large, public, highly competitive schools were less likely to report these supply-side efforts, a finding that was echoed in the restrictions on advertising. Why these institutional differences existed is unclear. However, administrators at these types of institutions were significantly more likely to report that student alcohol abuse was a major (as opposed to a minor or moderate) problem. This finding might suggest that problems were greater where supply-side controls were absent. Administrative perceptions of a more severe alcohol-abuse problem may also reflect unmeasured variables, such as greater outlet density in a community or local norms more accepting of high-volume consumption (suggested by the absence of advertising controls), all factors that may impede supply-side interventions.

Unfortunately, these hypotheses cannot be tested in a cross-sectional study such as this. Nevertheless, high rates of drinking problems reported in other national survey research,¹⁵ including those of the College Alcohol Study reported in this issue of the Journal of American College Health (pp 199-210) , may suggest that at least some of the supply-side controls reported by college administrators were not necessarily implemented or enforced.

Institutional investments in efforts to reduce binge drinking through various prevention initiatives were mixed. A majority of the college administrators reported that their schools had a designated official in charge of prevention initiatives. However, we observed a clear gradient in the prevalence of institutional investment and the response required. Institutional investments that required greater specificity in function (eg, in-house program evaluation), more personnel (eg, a task force or coalition), and more community involvement (eg, cooperative agreements and community meetings) were less common, with cooperative agreements relatively rare.

The tendency for a college administration to provide more inclusive, activist investments was directly related to how severe a problem student drinking was perceived to be. Perhaps this was a sign of the significance of a high level buy-in for building capacity in prevention in an institution.

Although evidence is emerging that the national picture of binge-drinking prevention efforts includes many elements of a more comprehensive, systems approach, several gaps are evident. Targeted education and outreach lag behind more general educational approaches. Policy and advertising restrictions are reportedly prevalent but belied by what is known about student drinking behaviors. Institutional capability to respond is mixed, with indications that community partnership is low. Of course, the least affluent schools-and perhaps many of the smaller institutions in the sample-may not be able to assign personnel and resources to deal with their problems internally or through community cooperation.

This national picture of binge-drinking-prevention programs in colleges should be considered in light of several of the study's limitations. The questions we used to obtain information about prevention efforts were derived from a systems-approach perspective and stressed environmental and policy issues. We may have omitted some program elements that adherents of social marketing or norming approaches would stress. These approaches use marketing tools to create demand for lifestyles that do not focus on drinking and disseminate information on low-risk drinking styles to encourage modeling of healthy peer norms.

Because of limitations inherent in collecting survey data and postcoding large quantities of qualitative reports, the listing of program elements and approaches presented here may be more succinct than reality might suggest. Prevention programming is varied, even within a program category; and crucial dimensions of policies and programming, including enforcement and implementation, cannot be captured in a survey of this kind. Depending on their position in the college administration, respondents may not be fully aware of the degree to which certain programs are implemented and policies enforced. Moreover, this was not an anonymous survey, in that colleges were identified and respondents were college presidents or other administrators chosen to respond for their institution, which may have helped shape the responses.

Because of the relative absence of scientific program evaluation results, we cannot guarantee that adopting program elements we included will result in a major reduction in college binge drinking and associated harms. Most important, the data presented are cross-sectional and we cannot infer causality in any of the associations we observed. Nevertheless, the very high rate of responses to the survey and the scope of information provided afford an important vantage point from which to assess prevention trends and needs in 2000 and beyond.

NOTE

For further information, please address correspondence to Henry Wechsler, PhD, Department of Health and Social Behavior, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115 (e-mail:hwechsle@hsph.harvard.edu).

REFERENCES