1 STADIUM ALCOHOL POLICY CHARACTERISTICS: AN EXAMINATION OF ALCOHOL POLICY IMPLEMENTATION, DIFFERENCES, AND EFFECTIVENESS By BRIAN EDWARD MENAKER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Brian E. Menaker
3 To all people who have guided me
4 ACKNOWLEDGMENTS I wou ld like to thank all of those who contributed both indirectly and directly to this dissertation. I thank my parents, Faith Ritchie and Richard Menaker for supporting me. I am grateful to my committee members and all those who have contributed to this work. I am thankful to Dr. Charles Wi lliams for advising me, chairing my committee, and instilling and reinforcing the values of hard work and determination, Dr. Beth Chaney for all of her help in statistical analyses and evaluation, Dr. Michael Sagas for his input and support, Dr. Matthew Walker for his encouragement and contributions and Dr. Kaplanidou for her help I would like to thank Dr. Ronald Akers and Dr. Ruth Steiner whose classes contributed to the framework of this paper. This project would have been impossible without the help of records departments of the law enforcement agencies who searched, sorted, and located crime reports. Also, I would like to acknowledge all of those who participated in interviews for their honest and thoughtful responses. Additionally, the research proj ect was conducted with financial support from the Bill Simms Endowment Graduate Student Research Grant through the Eric Friedheim Tourism Institute. Finally, I would like to thank my peers for keeping me sane and grounded and allowing me to do the same for them.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF ABBREVIATIONS ................................ ................................ ............................. 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Motivation of the Study ................................ ................................ ........................... 12 Overview of Study ................................ ................................ ................................ ... 14 Theoretical Overview ................................ ................................ ........................ 14 Statement of Purpose ................................ ................................ ....................... 15 Significance of Study ................................ ................................ ........................ 16 Research Questions ................................ ................................ ......................... 18 Limitations ................................ ................................ ................................ ........ 19 Definition of Terms ................................ ................................ ........................... 19 Basic Assumptions ................................ ................................ ........................... 20 2 REVIEW OF LITERATURE ................................ ................................ .................... 22 Multidisci plinary Nature of Stadium Alcohol Policy Literature ................................ 22 Concepts Associated with the Nature of this Study ................................ .......... 22 Deficiencies and Previous Focus in Stadium Alcohol Policy Literature ............ 23 Alcohol and Crime ................................ ................................ ................................ .. 23 Relationship between Alcohol and Crime ................................ ......................... 23 Alcohol Availability and Crime ................................ ................................ .......... 24 Relationships between Sport, Alcohol, and Aggression ................................ .......... 25 Sport, Leisure, and Alcohol ................................ ................................ .............. 25 Alcohol Consumption Considerations ................................ ............................... 26 Sport, Alcohol Consumption, and Aggression ................................ .................. 28 University and College Alcohol Policy ................................ .............................. 29 Stadiums and Tailgate Policy ................................ ................................ ........... 31 Stadium Alcohol Policy ................................ ................................ ..................... 32 College Football and Alcohol Consumption ................................ ...................... 34 College Football Games and Alcohol Related Crime and Injuries .......................... 37 College Football Games and Community Crime ................................ .............. 37 Alcohol Policy Effect on College Football Stadium Crime and Injury ................ 38 Theoretical Considerations ................................ ................................ ..................... 39 Social Cognitive Theory ................................ ................................ .................... 40 Influence of Alcohol on Aggression and Criminal Deviance ............................. 42
6 Temporal Regulation and Intertemporal Substitution ................................ ....... 43 Closing Remarks Regarding Previous Literature ................................ .................... 45 Summary of Research Questions ................................ ................................ ........... 46 3 METHODOLOGY ................................ ................................ ................................ ... 48 Design ................................ ................................ ................................ ..................... 49 Data Collection ................................ ................................ ................................ ....... 50 Collection of Qualitative Data ................................ ................................ ........... 50 Co llection of Quantitative Data ................................ ................................ ......... 53 Variables ................................ ................................ ................................ .......... 54 Indepe ndent v ariables ................................ ................................ ................ 55 Dependent v ariables ................................ ................................ .................. 56 Data Treatme nt ................................ ................................ ................................ ....... 57 Interview Analysis ................................ ................................ ............................. 57 Statistical Analyses ................................ ................................ .......................... 57 4 RESULTS ................................ ................................ ................................ ............... 60 Part 1: College Football Game Administrative Alcohol ................................ ........... 60 Consumption Containment and Enforcement ................................ ......................... 60 Policy Interpretation and Enforcement ................................ ................................ .... 61 Policy Origins and Development ................................ ................................ ...... 61 Stadium Location Differences ................................ ................................ ........... 62 Enforcement Perceptions ................................ ................................ ................. 63 On campus Compared with Off campus Stadiums ................................ ................. 67 Alcohol Sales versus Alcohol Prohibition in General Seating ................................ 68 Financial Benefits ................................ ................................ ............................. 70 Having it Both Ways: Access for Few, Prohibition for Most .............................. 71 Control ................................ ................................ ................................ .................... 73 Control through Prohibition in General Seating ................................ ................ 73 Control Where Sales are Permitted in General Seating ................................ ... 74 NCAA Status as a Control ................................ ................................ ................ 74 Education and Alcohol Policy ................................ ................................ .................. 75 Tradition and Culture ................................ ................................ .............................. 77 Alcohol related Issues Experienced on Game Day ................................ ................. 79 Overall Summary of Policies ................................ ................................ ................... 81 Part 2: Relationship among Game Day Characteristics and Game Day Behavioral Outcome Va riables ................................ ................................ ............ 83 Canonical Correlation Results ................................ ................................ .......... 83 Multiple Regression Results ................................ ................................ ............. 86 Crim e a s dependent v ariable ................................ ................................ ..... 86 Ejections as dependent v ariable ................................ ................................ 86 5 DISCUSSION AND ANALYSIS ................................ ................................ .............. 92 Linking Alcohol Policy Consideration with Social Cognitive Theory ........................ 92
7 Environmental Determinants ................................ ................................ ............ 93 Facilitation and Self regulation ................................ ................................ ... 96 Observational Learning ................................ ................................ .............. 96 Outcome Ex pectations ................................ ................................ ..................... 97 Self efficacy and Collective Efficacy ................................ ................................ 99 Avoiding Moral Disengagement ................................ ................................ ...... 100 Intertemporal Substitution ................................ ................................ ............... 101 Relating Policies to Outcome ................................ ................................ ................ 102 Canonical Correlation Findings ................................ ................................ ...... 102 Multiple Regression Findings ................................ ................................ ......... 104 Crime ................................ ................................ ................................ ....... 104 Ejections ................................ ................................ ................................ .. 105 The Connection between Qualitative and Quantitative Analyses .......................... 106 Conclusions ................................ ................................ ................................ .......... 106 Summary of Findings ................................ ................................ ..................... 106 Implications ................................ ................................ ................................ .... 107 Limitations ................................ ................................ ................................ ...... 109 Future Study ................................ ................................ ................................ ... 110 APPENDIX A INFORMED CONSENT FORM ................................ ................................ ............. 111 B SAMPLE RECRUITMENT LETTER ................................ ................................ ...... 113 LIST OF REFERENCES ................................ ................................ ............................. 115 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 121
8 LIST OF TABLES Table page 3 1 Description of participants in interviews. ................................ ............................. 59 4 1 Tests of canonical dimensions ................................ ................................ ............ 87 4 2 Calculation of Redundancy Indices for the First Canonical Function .................. 87 4 3 Calculation of Redundancy Indices for the Second Ca nonical Function ............. 88 4 4 Standardized Variance of the Dependent Variables Explained by ..................... 88 4 5 Standardized Variance of the Independent Variables Explained by ................... 88 4 6 Standardized Canonical Coefficients (Weights) ................................ .................. 89 4 7 Canonical Loadings ................................ ................................ ............................ 89 4 8 Canonical Cross Loadings ................................ ................................ ................. 90 4 9 Regression results for crime as dependent variable ................................ ........... 90 4 10 Regression results for ejections as dependent variable ................................ ..... 91
9 LIST OF ABBREVIATIONS BAC Blood Alcohol Content DUI Driving under the influence FBS Football Bowl Subdivision NCAA National Collegiate Athletic Association SCT Social Cognitive Theory SLT Social Learning Theory TPB Theory of Planned Behavior
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STADIUM ALCOHOL POLICY CHARACTERISTICS: AN EXAMINATION OF ALCOHOL POLICY IMPLEMENTATION, DIFFERENCES, AND EFFECTIVENESS By Brian Edward Menaker August 2011 Chair: Charles S. Williams Major: Health and Human Performance College football stadiums are affected by game day alcohol consumption which contributes to misconduct and law enforcement activity in these venues. Alcohol related misconduct has become a concern for many university officials, athletic departments, law en forcement officers, and spectators who experience problems that threaten the safety and enjoyment of all of those involved in college football games. This mixed methods study endeavored to explore the rationales for alcohol policy at these stadiums along w on campus or off campus start time of game, temperature, quality of opponent, and other game day characteristics affect the law enforcement figures within stadiums located in one Southeastern state in the U. S. The participants of the qualitative part of the study were facility administrators of the universities who played in the seven stadiums in the state. Reported crime and ejections were compiled from games played in those venues over three years along with game day information compiled from box scores such as team ranking, start time, collect and analyze the interview data. Canonical correlation analysis and multiple
11 regression ana lysis were implemented as the quantitative methods in considering the effect between game day variables and reported misconduct. There are fou r main categories of findings. First, t here were discernable differences between factors that affect the difference in polic i es between stadiums identified by stadium administrators Alcohol is served in all of the sampled venues, but only three sold to th e general population, while the other limited consumption to those club ticketholders or suite ticketholders. Social Cognitive Theory serves as a tacit influence on the policy development process. A relationship existed between policy factors and the relat ive numbers of repo rted crime and ejections with time of game and attendance being the most significant variables in all models analyzed. I t can be concluded that more reported alcohol related law enforcement incidents occur within off campus stadiums tha n on campus stadiums while the later the sta rt time of game, the more incidents that occur inside a college football stadium.
12 CHAPTER 1 INTRODUCTION Motivation of the Study Alcohol consumption in sport venues has become a concern for stadium managers, law enforcement, community officials, spectators, and community members (Madensen & Eck, 2008). This is especially true for managers of venues that host college football games during autumn Saturdays in the United States (Wieberg, 2005; Opdyke & Kesmodel, 2009). Venues that host large scale sporting events are subject to crowd control issues as a result of alcohol consumption before, during, and after events. Alcohol policies are implemented to diminish the impact of alcohol abuse by spectators and to prote ct all spectators, including those who do not consume alcohol. Administrators operate under policies and procedures depending on their experiences, crowd demographics, and expected fan behavior. It has been documented that, alcohol related traffic injuries and deaths have occurred stemming from sporting events held in venues that sell alcohol (Southall & Sharp, 2006 ). College football games are popular and important social occurrences for many fans in university towns across the United States. Fans would a gree that NCAA Division I Football Bowl Subdivision football is at the height of popularity, drawing over 48 million spectators to games during the 2009 season (Johnson, 2010). Many college football teams draw over 90,000 spectators, bringing in large amou nts of revenue to universities, and adding an influx of people to small municipalities. High risk alcohol consumption is often associated with these games and some would argue this is what brings about the festive atmosphere. However, some would argue that the high attendance and alcohol consumption combines to produce many related problems.
13 These issues include arrests and citations for public intoxication, disorderly conduct, assault, batte ry, personal injury, and in some cases death (Rees & Schnepel, 200 9). As a measure to control this potential crime, misconduct, and injury, regulations and policies are put into place to limit the impact of alcohol related issues on game day (Toomey & Wagenaar, 1999). Approximately three dozen of the major college footba ll stadiums allow alcohol sales, while most limit sales or consumption to high priced luxury suites and seats (Opdyke & Kesmodel, 2009). The locations of stadiums vary by being located on campus and off campus. In one southeastern state, the on campus stad iums do not sell alcohol or allow alcohol within their general seating, while alcohol is sold at the off campus stadiums Large crowds and the festive college football experience is a major concern for university administrators, stadium managers, and law enforcement officials. Tailgating and alcohol consumption are a part of the college football spectator experience, yet policies are different depending on location and ownership of stadium, primarily whether t he stadium is owned by the university or not. T he majority of stadiums do not allow alcohol in their general seating areas as per published alcohol policies on websites (Menaker & Connaughton, 2010). However, most colleges struggle with underage drinking among their students and have a negative outlook of alcohol sales in their sport facilities. The alcohol policies governing each stadium are different due to local alcohol laws, location, and responsible party for control of the venue. On a general level the built environment, that is, planned infrastru cture, spatial relations, and building characteristics can have an impact on alcohol sales, alcohol consumption, and related
14 crime and injury (Bernstein et al 2007). This phenomenon is no different with stadium planning and policy development. From a cro wd control standpoint, on campus stadiums most often are controlled by a combination of university athletic departments, university law enforcement, and contracted security, and do not allow alcohol in their general seating area. Off campus stadiums are ge nerally controlled by private entities and allow alcohol sales and consumption in the venue (Opdyke & Kesmodel, 2009). Therefore, there are differences between on campus and off campus stadiums in alcohol consumption patterns. However, it is unclear whethe r the levels of alcohol consumption occurring during the games, or the policies surrounding alcohol consumption are directly responsible for the crime and injury that occurs with college football stadiums. Regardless of the reasons for implementing differi ng policies, stadium managers are concerned with developing effective risk management policies and implementing these policies (Fried & Ammon, 2009) The policies, in turn, may influence the alcohol consumption decisions that patrons make when attending ga mes. Changing the drinking behavior of individuals can limit the amount of injury and crime that occurs at these events. So, the focus of this study is to analyze managerial rationale for stadium alcohol policy and the outcomes of the interaction between p olicy and alcohol consumption behavior, that is, reported ejection and crime figures within stadiums. Overview of Study Theoretical Overview Theories of aggression and alcohol consumption may be applied to explain the rationale for policy content as well as the relationship between alcohol policy and fan behavior. Social Learning/Social Cognitive Theory (SLT/SCT) which suggests that
15 behavior i s learned through interacting with their environment, has been used to explain alcohol consumption behavior and aggressive behavior (Akers, 2009; Bandura, 1973). The theory of intertemporal substitution/temporal regulation which explains how individuals s ubstitute for a n inability to perform a specific behavior during a regulated period, has been applied to how changes in alcohol policy affect alcohol consumption behavior at college football games (Boyes & Faith, 1993). There has been little research on ap plying this economic model to alcohol consumption behavior beyond this study. SL T has been used in few studies reg arding alcohol policy, specific ally dealing with modeling behavior (Abar, Turrisi, & Abar, 20 10 ). This is one theoretical explanation for the implementation differing alcohol policy and the relationship to aggressive behavior. In earlier works on aggression, Bandura (1973) suggested that regardless of outcome, witnessing a sporting event can evoke a cts of aggression which can lead to injury causing behavior and crime. The addition of the concept of self efficacy is a possible use of SCT in explaining policy implementation. Temporal regulation and intertemporal substitution has been applied to changes in outcomes of alcohol consumption in blood alcohol content ( BAC ) and driving under the influence ( DUI ) (Boyes & Faith, 1993). The dissertation uses an exploratory approach to determine the reasons for alcohol policy development and the effect alcohol pol icy and other game day factors has on game day alcohol related crime and ejections. Statement of Purpose The purpose of this study consists of an exploration of the environmental factors that affect alcohol policy and measure s alcohol related behavioral ou tcomes that might be affected these policies. First, t h is consists of inquiry into the composition of policy and administrator rationales regarding development and implementation of various
16 college football stadium alcohol policies including their own. Se cond, a consideration of the relationship between alcohol policy and stadium alcohol related law enforcement problems as measured by reported law enforcement activity (e.g., crime and ejections) is considered . The overarching purpose is to explore what af fects policy implementation and observed repercussions of alcohol policy at college football games. Significance of Study There are a number of rationales for conducting these complementary analyses. College football stadiums experience many spectator beha vioral problems, crimes, and crowd control issues related to alcohol consumption as is experienced in any place where individuals gather for sporting events (Madensen & Eck, 2008). C ollege football stadiums have varying posted alcohol policies (Menaker & C onnaughton, 2010), but often deal with alcohol related problems which burden university officials and law enforcement, and can create major risk management issues ( Fried & Metchi ck, 2005; Southall & Sharp, 2006 ; Rees & Schnepel, 1993 ). The enacted policies are not informed by theory but are often reactive. However inductive reasoning can be used to apply theory to explanations of why the components of college football stadiums differ from venue to venue. There has been no effective attempt to determine rea soning for policy and the outcomes of differing policy among different venues. So, this study will help administrators determine what game day factors, includi ng location of their venue, con tribute to the number of incidents occurring at football games. Li mited studies have compared the effect of policy on spectator alcohol consumption and associated misconduct at sporting events Previous research has analyzed the relationship between college football games and alcohol consumption, how alcohol consumption at college football games impacts perception of policy, and
17 the relationship between college football games and crime. Some studies have used TPB to explain alcohol consumption behavior during these events (Glassman et. al 2008; Glassman et al., 2010). Lit tle attention has been paid to the influence of environmental factors toward differing policies on engaging in problematic alcohol related behavior. The significance of this study is that college football stadiums have differing alcohol polici es, but often deal with alcohol related problems which burden university officials and law enforcement, and can create major risk management issues. Few studies have compared the effect of policy on those intentions. There has been research that has analyzed the relati onship between college football games and alcohol consumption, how alcohol consumption at college football games impacts perception of policy, and the relationship between college football games and crime. However, an effort to understand administrative re asoning and rationale for alcohol policies along with a comparison of the outcome of college football game day drinking behavior as a result of differing policies, among a number of stadiums has not been carried out. Additionally, the accepted best practic e in on campus stadiums is the prohibition of alcohol in general bowl seating while off campus stadiums allow alcohol consumption (Opdyke & Kesmodel, 2009). Studies exploring the impact of alcohol po licy on crime, injury, and other health outcomes within individual stadiums have been previously conducted (Bormann & Stone, 2001; Boyes & Faith, 1993; Spaite et al., 1990). However, comparisons of alcohol policy on reported crime and ejection within stadiums have not been conducted. The se significances influence t he following research questions
18 Research Questions The research questi ons in this study are influenced by SCT and the theory of intertemporal substitution. behavior while (Bandura, 1986). So based on SCT as a guiding theory of the study, how does the policy? With this in mi nd the following questions for research consider how the interaction between individuals and their environment affects development and implementation of college football stadium alcohol policy along with some behavioral outcomes that are influenced by poli cy. The research questions for analysis are as follows. Is there a relationship between game day environment characteristics and alcohol policy? Is there a relationship between college football stadium location (on campus o r off campus) and alcohol policy ? What factors affect alcohol policy enforcement? Does size of stadium as signified by stadium capacity affect alcohol policy considerations? Do game day characteristics a ffect alcohol related deviance outcomes as measured by r eported law enforcement recor ds? These previous questions relate how the environment interacts The following questions are influenced by th e theory of intertemporal substitu tion which suggests individuals will substitute for their inability to p erform a behavior during a regulated period of time, during unregulated periods (Boyes & Faith, 1993). Does alcohol policy correlate with number of ejections or reported crime within a stadium? Related to this, does the type of policy (alcohol sales permit ted v ersus alcohol sales prohibited) affect crime and ejection counts?
19 Limitations There are a number of inherent limitations to this study. The first component of the dissertation relies on interview data from one type of administrator which may not paint a complete picture of alcohol policy development and implementation. However, it h ighlights the thought process of a key policy decision maker. The quantitative component does not account for impact of college football game day outside of stadiums since reported crime data does not give the complete picture of alcohol related misbehavio r within stadiums. Additionally, the inability to gauge the direct impact of football games on crime to a consistent distance from venues, among all venues being compared. Alcohol policies are not enforced to the same degree in each municipality subject to study. Definition of Terms Alcohol related crime. Alcohol related crime refers to criminal acts such as public drunkenness, unlawful possession of alcohol, possession of alcohol under legal age, vandalism, disorderly conduct, and other aggressive crimes r elated to the consumption of alcohol. Alcohol related problems. These are hardships experienced by law enforcement officers and stadium administrators caused by alcohol consumption and spectator misconduct of stadium patrons. Reported crime. Criminal acts that have been reported, cited, and included in law enforcement ledgers are considered reported crime Stadium alcohol policy. These are restrictions or allowance of alcohol consumption within a stadium. This also pertains to the enforcement of alcohol con sumption violations. In addition it refers to the presence or lack of sales of alcohol.
20 Incident. This is a specific type of crime that results in law enforcement repercussions, including citation, ejection, arrest, or other recorded measure of law enforc ement action on game day. Open container laws. Local laws that stipulate individuals cannot carry open alcohol in public are known as open container laws. Senior athletic administrators. These individuals are administrators who work for university athleti c department and include athletic directors, athletic facility directors, or any upper level athletic administrator. This can include the senior facility administrator, the highest level of operations or facility manager of a college football stadium. Tail gating. This is a phenomenon that occurs outside sporting events in the United States It is a type of party that usually occurs at the rear of a vehicle in a parking lot near a venue where spectators share food and beverages in conjunction with a football game. This activity occurs before, during, and/or after the event and some revelers choose to participate during the course of the game as opposed to actually attending games Law enforcement. The process through which statutes and public safety policie s are enforced by uniformed police officers. At a college football game multiple law enforcement agencies work to ensure crowd control and public safety. These can and the alcohol control agency officers. Basic Assumptions This study operates under a number of basic assumptions. The assumption is made that the responses by participants in the interview and survey based stud ies are
21 accurate and truthful. Administra consistent with actual game day policy and operations. It is assumed that crime is enforced consistently by police officers with the same consistency on game day and the law is enforced consistently from game to game. Perception and rationales can be reliably measured and demonstrated via interview Also, part icipants are representative of typical college football game attendees. Crime and ejection statistics from law enforcement were accurately and consistently compiled over the time period observed.
22 CHAPTER 2 REVIEW OF LITERATURE Multidisciplinary Nature of Stadium Alcohol Policy Literature Concepts Associated with the Nature of this Stud y The topic of alcohol consumption surrounding college football games, alcohol policy development and enactment in college football stadiums literature, and other a ssociated literature is multidisciplinary in nature and, therefore, can be influenced by many concepts, theories, and disciplines. This review will cover an array of concepts and topics. The primary launching point, the relationship between alcohol and agg ression has been well documented, along with the impact of alcohol fueled aggression on crime (Bandura, 1973, Akers, 2009). Alcohol availability and distribution planning of alcoholic sales outlets has received attention for its impact on varying levels of crime. Sport has been related to alcohol consumption, and concurrently related to increases in crime rate, either due to watching it on television, or attending in person (Bandura, 1973) Treatment of alcohol policy in general, institutional alcohol polic y, and sporting event policy is integral to understanding this overall project. Alcohol consumption has become commonplace at sporting events, especially before college football games and during professional events. Thus alcohol policy in sporting events s ets the tone for the alcohol policies in college venues. College football games played on campus are subject to college or university alcohol polic y, which often influences policies of alcohol consumption during sporting events Along those lines are the c oncepts of institutional and/or environmental alcohol management and stadium alcohol risk management
23 strategies. Additionally social theories that can be applied to aggression, alcohol consumption, and alcohol policy but are not explicitly tied to those t opics are discussed. Deficiencies and Previous Focus in Stadium Alcohol Policy Literature Little scholarly literature exists on stadium policy, but there is a recent growth in its study. Much of the literature is applied from economics, criminology, epide miology, and health promotion as there is little sport management discussion of alcohol policy in college football stadiums. Alcohol consumption at college football games remains a major environmental health and risk management issue. Accompanying issues a t college football stadiums are not isolated to game day. The following discussion will establish how the main scholarly treatment of alcohol policy falls outside of sport management literature, and the necessity for exploring it with a sport management po licy lens. The relationship between alcohol consumption and crime is well established in criminology, economic, public health, sociological, and limited sport management literature, among others. Studies discussing risk management approaches to alcohol pol icy in sport venues remain among the few sport management studies in existence relating to this topic Therefore, it is necessary to venture outside of management literature to develop our theoretical framework for empirical study of alcohol policy conside rations at college football stadiums. Alcohol and Crime Relationship between Alcohol and Crime Empirical studies have evidence that excessive public alcohol consumption contributes to violence and increases the fear of crime (Makkai 2001). A United States Department of Justice study showed a heavy association between the use of alcohol on crime convictions (Greenfield, 1998). Approximately 36 % of the 5.3 convicted offenders
24 under the jurisdiction of the Bureau of Justice corrections agencies in 1996 had been consuming alcohol at the time of arrest. Close to 75% of those who committed public order crimes had been using alcohol. Alcohol was also related to assault, illustrated by the statistic that 4 out of 10 violent crimes involved a lcohol use by the offender. The study also found that only 19 % reported consuming alcohol alone, while the other 40 % used illegal drugs in combination with alcohol (Greenfield, 1998). This data can provide reasoning for the decision to remove alcohol from college football stadiums due to the large number of people in the venue comparable to the size of a small city the tendency for alcohol consumption, and the link between drunkenness and crime Alcohol Availability and Crime A relationship has been shown between how the built environment impacts alcohol consumption and related crime as illustrated by the association between alcohol outlet density and crime in Milwaukee (DiIulio, 1995). Research has found a positive relationship between the density of alcoh ol sellers and crime in the inner city by using census tract data. Findings propose that having more alcohol outlets in densely populated areas contributes to crime. A similar relationship exists between alcohol outlet density and crime in Detroit (Gyimah Brempong, 2001). This analysis showed higher crime rates, which included the total, violent, property, and homicide crime rates, being related to alcohol availability. There appears to be a relationship between more reported crimes in areas with higher con centrations of liquor stores and other alcohol sellers (Gyimah Brempong, 2001). An additional study found an acceleration of crime rate when licenses granted reaches 10 in a specific census tract, showing a density point for alcohol outlets. One study foun d there was a negative effect of alcohol selling, which could be limited through even dispersal of alcohol availability (Gyimah Brempong
25 & Racine, 2006). Therefore, alcohol policy restructuring should impact the quantity of alcohol consumed to counteract t he previously shown positive correlation between the availability of alcohol and crime. Another study also consi de red the relationship between alcohol outlet density and crime in Austin, Texas and San Antonio, Texas (Zhu, Gorman, & Horel, 2004). However, sociocultural variables were also considered. The authors analyzed neighborhood data describing social structure data, alcohol density, and violent crime. This study revealed that socio structure covariates explained approximately 59% of crime variance in Austin, and about 39% in San Antonio (Zhu, Gorman, & Horel, 2004). Adding alcohol to the model improved its statistical power and alcohol outlet was a significant predictor of crime with 71% of variance explained by the two variables in Austin and 59% in S an Antonio. While social demographics of an area contribute to suggesting that limiting access and availability to alcohol should have a positive impact on crime redu ction efforts (Zhu, Gorman, & Horel, 2004). Relationships between Sport, Alcohol, and Aggression Sport, Leisure, and Alcohol Many research inquiries have considered whether being a sport fans affects levels of aggression. This stems from observations that there is a connection between alcohol and leisure (Carruthers, 1993). Spectator sport is leisure and the expectancy that alcohol is a part of leisure behavior makes sense. Carruthers asserts that individuals believe that alcohol enhances their leisure expe riences. Many believe that the amount of alcohol is proportional to the level of enjoyment of the activity. Alcohol served three functions in surveyed participants. It heightened the engagement of the leisure
26 experience, increased social impulsiveness and personal comfort, and released the drinker from responsibility (Carruthers, 1993). Spectators are generally unreceptive to alcohol restrictions at sporting events even though they support general societal level restrictions away from the venue (McAllister, 1995). Individuals supported alcohol sponsorship restrictions except at sporting events. Additionally, they found a connection between sport viewing and alcohol consumption. One manifestation of this connection is the following consideration of alcohol consumption patterns at two major league baseball stadiums (Wolfe, Martinez, & Scott, 1998). Alcoholic beverage drinking behavior was measured at games played in two major league baseball stadiums. Since men are more likely to purchase alcohol than women a s evidenced in demographic information obtained in one of the stadiums, male spectators were given breathalyzer tests at the entrance gate and during the fifth inning as to determine alcohol consumption. Participants between the age of 20 and 35 were most likely to have consumed alcohol. The researchers observed that 41 % of the sample drank alcohol, but only 8.3 % were legally intoxicated at the .08 % blood alcohol level. Close to 5 % of individuals tested during the fifth inning planned to drive home from the game, which illustrates the link between alcohol consumption behavior and potentially risky behavior after the game (Wolfe, Martinez, & Scott, 1998). Alcohol Consumption Considerations Some scholars have made attempts to empirically link alcohol consumption to personal spectator sport involvement, or attachment or support for a sport or certain comparison of the degree of fandom of 180 college students ranging from not supporting sports at all, to being very much invested in following sports. The frequency
27 of their consumption of alcohol failed to show a relationship between support of athletics and degree of alcohol consumption. While no correlation existed in this cas e, it seemed worthwhile to continue to examine the potential link. Nelson and Wechsler (2003) first established a link between fan identification and alcohol consumption, in their analysis of alcohol consumption behaviors and attitudes. Their research show ed that sport fans had a tendency to get into conflicts, and suggested a relationship between level of fan (fan or non fan) and alcohol abuse. Collegiate fans may be more prone to participate in heavy drinking than other kinds of spectators. The study suggested a relationship between identifying oneself as a sports fan and alcohol abuse, and the conclusion was that sports fans drink more than non fans and acknowledg ed being involved in more conflicts than non fans. Wakefield and Wann (2006) explored the difference between those who they identify as dysfunctional fans and those identified as non dysfunctional. Dysfunctional behaviors can be categorized as complaining and confronting other spectators at sporting events, or other types of aggressive behavior in those situations (Wakefield & Wann, 2006). Observations predicted that dysfunctional fans of high identification would find alcohol more of a necessity of the spo rt consumption environment than highly involved low dysfunctional fans. The study explained the link between alcohol consumption and aggressive behavior is dysfunctional fans may drink to decrease inhibitions and increase efficacy in confrontational or dys functional behavior. Since highly identified fans are likely to attend sporting events in person a relationship exists between fan identification level and favorability toward alcohol consumption. The
28 authors concluded that college football games are likel y to attract individuals who are inclined to consume alcohol (Wakefield & Wann, 2006). significant difference between alcohol consumption amounts between fans and non fans. They bel identification equates with attendance of sporting events, but not all spectators are fans, a lso conducted an analysis according to fan and team identification level and found no significant difference between alcohol consumption amounts and experiencing negative alcohol related incidents. The authors acknowledged that more research should be cond ucted to further analyze attitudes toward alcohol consumption and behavior. Sport, Alcohol Consumption, and Aggression Kaplowitz and Campo (2004) analyzed alcohol consumption and attitudes toward alcohol policy in the context of a sports related riot. Foll owing a Michigan State University loss in the Final Four of the NCAA Basketball Tournament, the campus experienced fan violence, which led to rioting. The study sought to link the amount of alcohol consumption to student attitudes towards both campus alcoh ol policies and the posit ive effect on enjoying the riot drank more objected heavily to restrictive alcohol policies. In general, those who consume more alcohol object to policies restricting alcohol consumption. Related to these concepts, Dimmock and Grove (2005) investigated whether fan identification (i.e., the psychological bond that an individual has to a team or sport ) and associated aggression are related with constructs of the TPB, e.g., attitudes, social norms, and
29 perceived control. In testing Australian fans of different teams in differing sports, attitudes toward aggression or their beliefs about social norms wit h respect to aggression did not vary between highly identified fans and lowly identified fans. Conversely, highly identified fans felt less control over their actions at games than fans with low levels of identification. One noteworthy explanation for this relationship is that fans who sense a threat or enhancement to their identity, emote in ways that lead to a lack of behavioral control. Celebratory riots would be attractive for highly identified fans desiring identity enhancement as opposed to lowly iden tified fans (Dimmock and Grove, 2005). Camp Randall Stadium in 1993, suggested that the violence and disruption was fueled by alcohol abuse by spectators (Fried & Metchi ck, 2005). A riot ensued at the end of a game between Wisconsin and Michigan, which caused a fence to collapse within the stadium injuring 73 spectators. The authors offered suggestions for preventing similar incidents in the future. Included within these were: increasing penalties for alcohol related misconduct, encouraging Universities to work with local bars to ban inexpensive alcohol promotions, enforcing of open container laws, offering students who abide by the rules, special incentives or preferentia l seating, and revocation of season tickets for poor behavior. University and College Alcohol Policy Previous discussion has established the issues surrounding alcohol consumption related to leisure sporting events. Crime, aggressive behavior, injury, rio ting, and death can stem from alcohol abuse activities. Therefore, an alcohol control policy is
30 necessary. The following discussions highlight alcohol policy considerations on university campuses and in sporting venues. Universities and colleges face crime vandalism, injury, and behavioral problems due to alcohol abuse on their campuses. Therefore, administrators formulate policies to deter, decrease, and discourage alcohol abuse and the related environmental health hazard that accompany problem drinking. However, these policies should endeavor to change attitudes and norms regarding drinking, not simply serve as a deterrent (Toomey & Wagenaar, 1999). Environmental policies and strategies should be directed at college students and the community as a whole ( Toomey, Lenk, & Wagenaar, 2007). These strategies include reducing alcohol related problems and consumption among underage individuals, reducing alcohol consumption by all college students, and lowering the emphasis on the role of alcohol in campus life. E nvironmental alcohol consumption control measures could also be applied to the college football game day environment to limit or decrease potential alcohol related problems. Gaining support for such strategies might aid the environmental management of alco hol abuse. Environmental control of alcohol on campuses is imperative due to reports that on campus crime rates rise when alcohol availability increases. Increases in beer prices and related taxes could limit the amount of binge drinking and underage drink ing by college aged females (Chaloupka & Wechsler, 1996) College age males were unresponsive to price increase. Accordingly, controlling availability and increasing price decrease s crime rates, depending on the situation. However, other factors such as living in a sorority or fraternity, the ease of availability of alcohol, and living on campus might be better predictors of alcohol consumption by college students (Chaloupka &
31 Wechsl er, 1996). Results of this study parallel previous and subsequent findings that alcohol availability is related to crime. DeJong, Towvim, & Schneider (2007) give empirical support for the posited policy suggestions by analyzing student support for alcohol enforcement policies and strategies among 32 four year institutions of higher education in Pennsylvania. A majority of the college students sampled supported five of 12 alcohol problem reduction policy proposals which included: stricter disciplinary sancti ons for alcohol related violence and repeat offenders of campus alcohol policy, stronger repercussions for those who purchase alcohol with false identification, keg prohibition on campus, and restricting alcohol advertising. The following policy proposals did not receive majority s upport but were supported by 30 % or more: undercover visits to bars to increase legal drinking age compliance, banning alcohol company sponsored events on campus, elimination of bar and liquor store specials, alcohol tax increases to support programs that keep minors from drinking. This study suggested that administrators should poll students to determine what policies their students are likely to support (DeJong, Towvim, & Schneider, 2007). Stadiums and Tailgate Policy Stadium al cohol and tailgating policies vary by venue. Miller and Gillentine (2006) inquired into the risk management efforts that major Division I college football stadiums made with regard to tailgating policies. Research determined responsibility for who actually Gillentine, 2006, p. 203). They concluded that stadiums should employ a risk management plan in order to keep patrons from unnecessary harm, and protect thems elves from negl igence lawsuits. Variations exist between alcohol policies at
32 different professional venues (Fried & Ammon, 2009). After surveying professional stadium ma nagers, the authors found that a lcohol is sold at most professional facilities. Most enforce alcohol p olicies that include confiscation of outside alcohol and two beers per transaction. Tailgating was found at 50% of responding facilities. However, the responsibility for parking lots where tailgating occurred varied depending on whether parking lots were i n control of stadium management and where the lots were located. While alcohol policies vary between venues, attitudes directed toward alcohol policies at one venue can also vary among patrons. Glassman, Werch, Jobli, and Bian (2007) conducted a compariso n of game day policies that included allowing open containers in designated tailgate areas and increasing the amount of law enforcement, and limiting tailgating hours. Over half of spectators surveyed did not drink on game day. Those fans supported restric tive alcohol policies. However, the degree of alcohol consumed on game day negatively affected the support for alcohol restriction. Students opposed more underage drinking enforcement, while non students supported it, but neither group strongly opposed thi s proposed measure. Stadium Alcohol Policy Lenk, Toomey, and Erickson (2009) assessed the variety and intensity of professional sport venue alcohol enforcement and related problems. In a telephone survey of state alcohol beverage control and local law enfo rcement agencies in host cities and states in the U.S. that have professional sports stadiums during 2005 06 seasons (n=98), they found that compliance checks were the most common type of alcohol enforcement in slightly over 50% of agencies. Reports of fig hting were the most common complaints received by almost three quarters of agencies. Complaints about intoxicated individuals were received by 65% of state agencies. Venues that hosted
33 basketball were less likely to have complaints than other venues that d id not host professional basketball, while those hosting professional football were more likely to have complaints as opposed to venues with an absence of football played within. The study concluded that high rates of alcohol related problems but low level s of enforcement occurred at sports stadiums. Additionally, states with lower heavy episodic drinking rates were more likely to have more types of enforcement than those with higher rates. The authors suggest ed that states with lower rates of heavy episodi c drinking may have less alcohol related issue and a lower frequency of alcohol related problems, as a result of more systems of alcohol enforcement implemented Therefore, the culture (or lack) of drinking, along with type of enforcement of alcohol relate d laws may impact the levels of alcohol consumption and alcohol relat ed problems. (Lenk, Toomey, & Erickson, 2009) Lenk et al. (2010) built upon the previous stud ies by examining alcohol control policies employed by professional stadiums, by surveying 66 o f the 100 venues that host hockey, bask etball, baseball, or football games The majority of stadiums reported allowing no more than two alcoholic beverages for purchase by patrons, and checking identification of individuals who appeared under the age of 30 However, about half of the venues prohibit servers under 21 to serve alcohol to guests. Additionally, only one third have sections where alcohol sales and consumption are prohibited. Thus, alcohol control policies in stadiums are not standardized, and it is important to determine what policies are the most successful in preventing alcohol related problems at sports venues.
34 Erickson et al. (2011) conducted a preliminary study to test the feasibility of collecting the blood alcohol content (BAC) levels of a ttendees of sporting events and also observe the levels of intoxication of individuals leaving those events. They tested 362 patrons leaving 13 baseball games and 3 football games, considering the factors related to higher measurements of BAC. They conclud ed that 8% of those in their sample were legally intoxicated, above the .08% BAC level. Additionally, many young adults who participate in pregame tailgate festivities were 14 times more likely to have BAC levels over .08%. The authors suggest that the pra ctices of serving individuals with the stadiums along with tailgating contribute to the elevated BAC levels of patrons, which could lead to more alcohol related problems after games (Erickson, et al., 2011). College Football and Alcohol Consumption Previou s literature that relates TPB to alcohol consumption has helped inform studies about alcohol consumption at college football games. Conner et al. (1999) concluded that TPB was a suitable model in showing that alcohol consumption intentions, attitudes towar d behavior, subjective norms, and perceived behavioral control explained a fair portion of the variance in intentions to consume alcohol. Perceived behavioral control correlated highly with actual drinking behavior in some cases, but not all. The authors f ound that regularity of drinking, or not drinking alcohol, was associated with intention to consume but lower perceived control over alcohol consumption. Intentions to consume alcohol and actual alcohol consumption were partially explained by a perceived l ack of control (Conner et al, 1999). So the less perceived control the more likely one was to possess intentions to consume alcohol. A handful of research has focused on college football game day alcohol consumption behavior. TPB might be a sufficient pre dictor of game day alcohol
35 consumption (Glassman, et al. 2008). The study explored the fit of TPB in predicting alcohol consumption on game day among college students. A sample of 740 college and the number of alcoholic beverages consumed on game day using path analysis. Behavioral intentions, attitude, and subjective norms were all statistically significant; however, perceived behavioral control was not significant. Since perceived behavioral control was shown to be of questionable applicability as compared with the precursor to TPB, the Theory of Reasoned Action, the latter might prove to be a better predictor for game day alcohol consumption. More effective measures of perceived behavioral c ontrol are necessary to conclude that the TPB can predict game day college student alcohol consumption (Glassman et al., 2008). A related study assessed how well TPB can predict college based on consumption rates. Three groups were selected from three rivalry games at University of Florida home football games to complete an anonymous survey on the Monday after those chosen games (Glassman et al. 2010). Path analyses were conducted to tes t which of the constructs of TPB predicted behavioral intentions and alcohol consumption patterns among social, high risk, and extreme ritualistic drinking, defined as individuals who drank 1 4, 5 9, and over 10 drinks on game day for males, and 1 3, 4 8, and over 9 consume alcohol across all three groups, but perceived behavioral control was inconsistent in predicting intentions and alcohol consumption (Glassman et al., 2010). Select constructs may be used in predicting alcohol consumption behavior, but the
36 Oster Aaland and Neighbors (2007) evaluated the impact of alcohol consumption po licy change during football game tailgating activities. They analyzed the prevalence of alcohol consumption during tailgating at college football games and how expectations of drinking compared to actual drinking during tailgate parties, and found that man y fans considered alcohol consumption a critical part of being a sports spectator. However, attitudes about attending college football games were not changed by completely restricting alcohol consumption at tailgates, nor did it change reported alcohol con sumption. Often spectator expectations and attitudes shape the drinking behavior and related legal consequences on game day. So a resulting conclusion remained that overestimation of peer drinking can lead to an environment of excessive alcohol abuse. One study shows empirically when individuals drink before college football games, they generally consume large quantities of alcohol. Heavy episodic drinking, or 5+ drinks for males or 4+ drinks for females in a drinking session, is prevalent at tailgates bef ore college football games (Merlo, et al. 2011). In a study conducted at two breathe alcohol content half of the 466 participants engaged in heavy drinking prior to games (Merlo, et al., 2011). Only 54 individuals from the sample did not consume alcohol at all at tailgates This shows how widespread alcohol consumption is surrounding college football game festivities. Therefore, public health and responsible alcohol consump tion interventions would be helpful in controlling heavy game day drinking and promote a safe environment. Role modeling behavior as described by Social Learning Theory can also impact the quantity of alcohol consumed by individuals at pre game football t ailgate events
37 (Abar, Turrisi, & Abar, 2010). Results suggest that drunken behavior by parents at drunken behavior by parents at tailgates was a better predictor of drink ing and adverse consequences of college student drinking than the influence of heavy episodic drinking heavy drinking, in this case tailgating at sporting events, can serv e a modeling behavior that students later emulate (Abar, Turrisi, & Abar, 2010). Col lege Football Games and Alcohol Related Crime and Injuries College Football Games and Community Crime Two studies have established a relationship between alcohol related crime and college football game days, suggesting that the amount of alcohol related crime increases in comparison to non game days (Merlo, Hong, & Cottler, 2009; Rees & Schnepel, 2009). An alysis suggested that the relationship between college football games and alcohol related arrests found that more crimes related to alcohol consumption occurred on game days, as opposed to normal days or holidays (Merlo, Hong, & Cottler, 2009). There were nearly seven times more arrests on game days than on normal days or holidays that were included in the sample. They compared college football game days, holidays, and control days, and found more crimes on game days as compared to holidays and non special days. Arrests occurred closer to the stadium on game day than on other days. Rees and Schnepel (2009) suggested that host cities of college football teams experience increased crime on game days. They found a positive association between home game days and vandalism, assaults, disorderly conduct arrests, and alcohol related arrests. The largest estimated effect of college football games on crime occurred when an upset occurred. The spike in crime might be
38 explained by the dramatic increase of people into th ese cities, and different degrees of law enforcement by police, as opposed to alcohol consumption (Rees & Schnepel, 2009). Alcohol Policy Effect on College Football Stadium Crime and Injury crime when alcohol consumption was prohibited after many years of allowing its sale (Bormann & Stone, 2001). Banning alcohol within a stadium has been shown to decrease crime, arrests, and injuries at college football games (Bormann & Stone, 2001). Their Folsom Field yielded a decrease in delinquent behavior and crime at the games observing a 50% decrease in ejections and 45% in arrests, and an 89% drop in student conduct violations re ferrals to judicial affairs office (Bormann & Stone, 2001). Thus, bans of alcohol may be considered the best policy by on campus stadiums, due to empirical evidence. However, another study has found alcohol policy change might not eradicate alcohol related problems, particularly injury, just change the type of outcome. Spaite, et al. (1990) observed no signif icant overall change in alcohol related injury rate after the alcohol ban but their stadium of interest experienced a difference in the pattern or type of injury. This study considered the effect of banning alcohol on reports of illness and injuries to spectators at a major college football stadium between 1983 and 1986. Alcohol sales were never permitted, but alcohol was allowed to be carried into the v enue until 1985. The results suggested no significant overall change in incident rate after the alcohol ban ended, but there was a difference in the pattern or type of injury. (Spaite, et al. 1990)
39 A case study further supports the aforementioned findings that alcohol policy can have an impact on game day alcohol consumption behavior and related injury (Johannessen, Glider, Collins, Hueston, & DeJong, 2001). The University of Arizona after administrators observed a noticeable pattern of increased crime and violence during homecoming weekends concluded that implementing manageable, enforceable, and consistent alcohol control policies surrounding football festivities would decrease the associated incidents experienced before ( Johannessen, et al., 2001). The outcomes of stricter alcohol restrictions at college football games and associated activities were less crime, fewer neighborhood calls to law enforcement, and a reduction of other alcohol related problems in the year the new policies were enacted. Yet, in cidents rose to levels observed before the new policies in following years. Therefore, it is difficult to confirm the efficacy of the policies because of the possible impact of other factors such as varying law enforcement deployment over the years. The au thors suggested that more oversight and restrictions surrounding homecoming activities where alcohol was served at the University of Arizona yielded less crime and fewer incidents. Theoretical Considerations There are two theories gleaned from the previous alcohol policy and stadium policy literature that direct the empirical study to be described in the following ch apters. The frameworks include Social L earning T heory/Social Cognitive Theory (Bandura, 1986) and the economic theory of intertemporal substitu tion. SCT and social learning theories can explain the connection between sport and aggression, alcohol consumption, and related crime. Intertemporal substitution can inform whether individuals are likely to consume more alcohol during unregulated periods to make up
40 for alcohol prohibitions, and in turn contribute to higher crime rates in stadiums that do not consume alcohol as opposed to stadiums that do sell alcohol. Social Cognitive Theory SCT serves as a guide to explain how individuals interact with t heir environment and vice versa to influence behavior. This explanation can be extended to help observe how alcohol policies are informed and implemented by extending explanations of how people interact with their environment in situations where alcohol is consumed and controlled. SCT is based on SLT principles established by Miller and Dollard (1941) and efficacy and perceived environmental impediments and facilitators in the regulation of human motivation, action, and well the ability to modify and regulate their behavior based on their learning his tory, perceptions of the environment, and physical and intellectual capacities. Learning and perceptions and cognition, especially alcohol related behavior (Bandura, 1969). Utilizi ng these ideas can lead to successful alcohol policies at college football games. The concepts of SCT can be grouped into five categories: Psychological determinants of behavior, which include outcome expectations, self efficacy, collective efficacy; obser vational learning; environmental determinants which include reciprocal determinism, facilitation, and incentive motivation; self regulation; and moral disengagement (McAlister, Perry, & Percel, 2008). Outcome expectations are the probability and impact of the consequences of their
41 perform or do not perform a certain behavior. Self efficacy consists of personal beliefs o behave in a way that brings desired outcomes. There are four ways through which self efficacy can be developed: Mastery experience enables the person to succeed (Bandura, 2004; Bandura, 1997). Collective efficacy extends the self belief of self efficacy to groups. Observational learning occurs when individuals learn to perform new behaviors by exposure to those behaviors, especially by viewing peer models. Incentive motivation stems from providing rewards and/or punishments to change behavior. Reciprocal determinism illustrates how the environment influences individuals and groups, but in addition also considers addresses how people interact with their environment while regulating their own behavior (McAlister, Perry, & Parcel, 2008). Facilitation provides new structures, resources, or environmental changes that enable ease of new behavior performance (Bandura, 1998). Self regulation refers to an control through self goal setting, feedback (i nformation about behavior performance and how it could differ), self reward, self instruction (self talk before and during a behavioral performance), and finding people to provide social support. Moral disengagement occurs when people think about harmful b ehaviors and breach moral standards for self regulation (McAlister, Perry, & Parcel, 2008). This theory can be applied to situations involving aggressive behavior: Alcohol consumption, college football viewing, and explanations of criminal behavior. Accord ing to Social Cognitive Theory (SCT) regardless of outcome, witnessing a sporting event can evoke acts of aggression which can lead to injury causing behavior and crime (Bandura, 1973). At a general level, the theory asserts that people can work together in
42 social systems to change an environment to benefit an entire group (Bandura, 1997). SCT can be used to explain behavior surrounding college football games while helping to inform alcohol control policies to implement before, during, and after those even ts. Influence of Alcohol on Aggression and Criminal Deviance An attempt to apply SLT a precursor and essentially complementary to SCT without self efficacy, as analyzed in criminology, to college football alcohol policies is worth a brief overview in the review of literature. While these theories are generally applied to treatment programs, there are some ways to explain general stadium alcohol policies through social learning. The theory suggests that the learning processes that lead to alcohol abuse als o encourage criminal and deviant behavior. Substance abuse In studies focusing on the variable of peer association, a relationship exists between it and delinquency, a lcohol use, violent crime, property crime, and other types of deviant behavior (Akers & Jensen, 2007). Along those lines, Akers (1979) shows that the probability of abstinence from alcohol decreases, while frequency increases, when there is more exposure t o models who were users as opposed to abstainers. Positive definitions have more of an effect on amount of use than negative ones. Data from Akers' Boys Town study focusing on alcohol and marijuana use in adolescents can help shed light on decisions to use alcohol by Midwestern adolescents, abstention from alcohol can be explained through differential reinforcement, that is, re rewards and punishments
43 favorable or unfavorable (p. 171). Individuals learn attit udes about use through it is accepted within the family and close circles of friends (p. 171). They also learn the behavior of how to use substances to enhance it ef fects from social influences. If the individuals define the use as beneficial and acceptable, rather than conforming to negative attitudes, they are more likely to consume alcohol. Additionally, there is more social acceptability to extreme consumption of alcohol, due to the prevalence of its availability. definitions and other factors, had no effect. Additionally effect on varia tions in individual abstention from alcohol consumption and marijuana use (p. 187). Thus, social learning can be a persuasive explanation between the relationship between alcohol use and crime (Akers, 2009). Temporal Regulation and Intertemporal Substituti on Temporal regulation and intertemporal substitution serves as a potential explanation of alcohol consumption at college football games, especially keeping varying alcohol sales allowances or prohibitions. The following study detailing the effects of alco hol policy amendments described the effect of changing game day policy on blood alcohol content and law enforcement reports of driving while intoxicated. (Boyes & Faith, 1993)
44 One economic theory suggests that the control of alcohol use during a specific t ime period may shift the negative consequences of alcohol related delinquency to another uncontrolled period (Boyes and Faith, 1993). Intertemporal substitution refers to a displacement of consumption created by regulation. This shift in alcohol consumptio n could be applied to the policies that prohibit alcohol within general stadium seating areas. The behavior is allowed during some periods, but the consumable that causes the problem behavior (e.g., alcohol) is taken away for a regulated period of time. Bo yes and Faith (1993) apply the economic theory of intertemporal substitution and temporal regulation to alcohol consumption and policy. Public drinking imposes costs on others. hift at least a portion of that consumption into non consumption (p. 596). This shift may be desirable or inconsequential for man y goods. For instance, not allowing smoking in a certain place may not lessen the number of smokers, but limit the amount of tobacco smoke and the social cost to others. The substitution may not be socially harmless in the case of alcohol since intoxicatio n increases during non ced a shift in intoxication to the unregulated periods. More car accidents after games, as well as more arrests from driving under the influence, could result from a higher level of intoxication, since intertemporal substitution of alcohol consumption can lead to drinking and driving.
45 One might suppose that regulating alcohol during a football game should help g for increased caused by intertemporal substitution is not inconsequential. There are three arguments for the unintended effects shifting the externalities. First, the effe cts of increasing consumption of alcohol in the non regulated period before the game carries over into the regulated period due to the fact that alcohol concentration does not reach maximum until one hour after the final drink is consumed. Second, the rate and volume of even if there is not a one for one substitution of consumption from the restricted period to the adjacent unregulated periods, average intoxication taken o ver the adjacent and increase of the external effect of drunk driving to after t he game (p. 596). Acknowledging that intertemporal substitution may not have the desired effect may give credence to those who assert a preference for the availability of alcohol sales at college football games. C losing Remarks Regarding Previous Literatur e Based on the previous literature, two studies will be conducted to analyze and explain the development, and crime and health outcomes of college football alcohol policy. Based on previous considerations, SCT is the most relevant explanation for interven tions to counteract aggression and alcohol abuses. Therefore the first study
46 focuses on revealing themes related to alcohol development and implementation which could possibly be connected to theory. No further study has built upon the Boyes and Faith (19 93) effort in exploring the effect of alcohol policy on spectator misconduct, nor has there been a comparison of the effect of alcohol sales on stadium crime and injury between different venues. Temporal regulation and intertemporal substitution serves as a means of looking at the unintended consequences of stadium alcohol policies created by regulation. The underlying rationales for a policy and its expectations for desired effects or consequence (namely reduction of alcohol related problems at the game an d on campus) may have inadvertent consequences of shifting crime to another time and place. This means that a policy with the intent to help limit the harmful byproducts of alcohol consumption and analogous behaviors associated with heavy drinking simply s hifts the problems away from the stadiums plus creating new problems in and around it. Therefore, the second study will test this effect by comparing stadiums with differing policies. In other words, does the likely shift of alcohol availability in stadium s without alcohol sales yield higher crime and ejection rates than stadiums that allow its sales. Summary of Research Questions The aforementioned literature review influences the research questions for analysis and are as follows. What relationship exists between game day environment characteristics and alcohol policy? Is there a relationship between college football stadium location (on campus o r off campus) and alcohol policy? What factors affect alcohol policy enforcement? Studies that detail alcohol policy at sporting events guide those three questions (Abar, Turrisi, & Abar, 2010 ; Erickson et al., 2011; Lenk et al. 2010; Lenk et al. 2009). Does size of stadium as signified by stadium capacity affect
47 alcohol policy considerations? Do game day characteristics affect alcohol related deviance outcomes (crime and ejections) as measured by reported law enforcement records? Since these questions are based on environmental influences on behavior SCT and SLT help guide these questions (Bandura, 1986; B andura 1973; Akers, 2009). The following questions are influenced by intertemporal substitution (Boyes & Faith, 1993). Does alcohol policy correlate with the count of ejections or reported crime within a stadium? Additionally, does the type of alcohol sale s policy affect crime and ejection counts? (Spaite et al., 1990; Boyes & Faith, 1993; Bormann & Stone, 2001; Johannessen et al., 2001 ; Rees & Schne pel, 2009 ) T he next section will detail the methods for analyzing the research questions.
48 CHAPTER 3 METHODOLOGY The purpose of this dissertation is to identify and explore major factors that influence the alcohol policy development and implementation by athletic adminsitrators for college football stadiums and determine what characteristics of game day affect alcohol related crime and ejections The research is carried out using a twofold analysis. The first step is to determine what impacts the reasoning for administrative alcohol policy. The second examines what environmental factors affect in stadium ejections and alcohol related crime. The benefit of mixed methods is that multiple approaches provide a comprehensive analysis that will help us determine the role that alcohol policy and other variables play in spectator game day behavior. The qualitative component seeks to determine the rationale for stadium alcohol policy Previous literature had identified SCT/SLT as a beneficial explanation for the reasons for alcohol consumption at sporting events ( Abar, Turrisi, & Abar, 20 10 ; Rees & Schnepel, 2009; Bandura, 1973) and t he theory of intertemporal substitution has been used to posit the differences in alcohol related problems between stadiums that sell alcohol and those that prohibit sales (Boyes Faith, 1993) but have stopped short of determining alco hol policy development and its impact at these events This inquiry seeks to resolve the following question s for research : What relationship is present between game day environment characteristics and alcohol policy? Is there a relationship between locatio n of a college football stadium (on campus o r off campus) and alcohol policy? What fac tors affect enforcement of alcohol policies ? Does the size of the stadium affect alcohol policy development by the decision makers ?
49 The quantitative component compares t he relationship between alcohol policies at different stadiums to reported crime within stadiums in one Southeastern state over a three year period. The purpose of this component is to determine whether the environmental characteristics of college football game days affect stadium policies affect the amount of spectator e jections and crimes. Q uestions for research based on the quantitative inquiry include the following. Does the type of policy (alcohol sales permitted v. alcohol sales prohibited) affect reported stadium crime and ejection totals ? Is there a correlation between alcohol policy and crime and ejections?. Additionally, is there a relationship between game day characteristic variables (i.e., temperature, start time, conference opponent) and rep orted spectator cr ime and ejections ? Design This study follow ed a mixed methods analysis using a phenomenological approach to explore administrative viewpoints of college foo tball game alcohol policy, a canonical correlation analysis, and multi ple regress ion analysis to model the relationship between college football game day variables on spectator deviant behavior as measured by alcohol related crime. The qualitative interview based approach help ed to inform the framework and variable selection for the ca nonical correlation analysis and multiple regression models. These analyses are intended to show the relationship between and among location of stadium, alcohol policy, and other dependent variables. The first component explore d the phenomenon of alcohol p olicy control at col lege football stadiums within venues from a Southeastern state. phenomenological approach was used. Using this framework, c riteria were established for selecting participants Subsequently, i nterviews conducted using a guided semi structured schedule I ndividual, composite meanings were constructed to create a
50 synthesis of meaning and essences of the experiences A summarization of the study and findings were sub sequently produced (Moustakas, 1994) The quantitative com ponent consisted of testing the relationship between college football game day characteristics and alcohol related crime and ejections A canonical correlation was chosen to establish a relationship between the set of independent variables (Hair et al., 2009 ) (i.e., the game day characteristics ) with the set of dependent variables (i.e., crime and ejections) ejections and crime while regression model s were used to predict the relationship of independent variables on ejection and crime se parately as dependent variables. The subsequent sections detail the data collection for each component of the study followed by data treatment for each respective part of the study. Data C ollection Collection of Qualitative Data The collection technique t hat was used was semi structured face to face interviews (Berg, 2002) with senior athletic administrators with direct experience with college football stadium alcohol policy development, implementation, and enforcement. The purposive sampling procedure con sist ed of choosing athletic administrators in seven athletic departments. The participants in the study include senior level stadium administrators in charge of game day operations of college football stadiums. These individuals were systematically chosen based on their title and involvement in the athletic departments of the sites selected for study. The sites are all in the same state, at universities that sponsor Division I Football Bowl Subdivision( FBS ) teams. For a detailed table of stadium characteris tics along with a description of the individuals representing each stadium, see Table 3 1.
51 Eight individuals from seven athletic department s were interviewed. Four include on campus stadiums that do not sell, serve, or permit consumption of alcohol in the ir general seating areas. One of these venues has a capacity of 20,000 and a program with less than ten years of football competition, referred to as Stadium A in the study. Two individuals from this location participated in interviews An administrator fr om each of the following locations participated in the study. The second has an administrator involved in game day facility operations at a stadium with a capacity of approximately 90,000, hosting a successful program that has a long established football t radition, labeled as Stadium B in the study. Stadium C has a capacity of approximately 48,000, and Stadium D holds 83,000. Three of the universities host their games off campus in stadiums that sell alcohol. Stadium E has a capacity of 20,000, Stadium F ha s a capacity of 70,000, and Stadium G has a capacity just over 75,000. Individual administrators who had personal involvement in implementing, carrying out, and evaluating game day operations policies and procedures, including the stadium alcohol policy w ere interviewed. These individuals were found by scanning the athletic department directory websites for the individual in charge of football event operations. Individuals were contacted by electronic mail and/or telephone to determine whether they indeed oversaw game day operation. Once an individual was selected, an appointment for a face to face interview was arranged. The researcher travelled to the campuses of each administrator and met them for an interview held in their offices. Data was collected via face to face semi structured interviews which last ed between 45 and 90 minutes.
52 Institutional Review Board (See Appendix A) approval was granted by the University of Florida and participants were given an informed consent form to review t he risks, benefits, goals and purpose of the study, and outline of the overall process of the interviews. All interviewees provided informed consent to participate in the project. The interviews follow ed a semi st ructured schedule which involved implementi ng predetermined questions but the interviewer probes into the answers of the prepared standardized queries (Berg, 2002). The interview questions focused on the participants knowledge of alcohol policy at their stadium, rationales for their policy, and oth er aspects of alcohol policy. The interview schedule included the following questions : What is your stadium alcohol policy? What are the different aspects of the alcohol policy that are enforced? Do you allow alcohol sales within your stadium? What are some reasons for allowing/not allowing alcohol sales within your stadium? Is stadium management responsible for implementing the alcohol policy? If not, who implements the policy? Is alcohol allowed in some portions of the stadium? If yes, where is it allowed? Who are the responsible parties for carrying out the alcohol policies on game day? Who is responsible for enforcing alcohol policy outside the stadium? Are law enforcement officials judicious about citing people for certain alcohol violations?
53 What is your personal attitude toward alcohol consumption at college football games? What is the amount of alcohol related problems the stadium experiences on game days? How are stadium alcohol violations cited and treated? How do you perceive the relationship between alcohol consumption and fan behavior? Other relevant questions that followed from the conduction of the interview were asked if deemed necessary by the interviewer Since the interviews were conducted in a semi structured manner, not a ll questions in the schedule were asked to all interviewee. Questions were used as a guide to encourage the interviewees to talk about alcohol policy issues on game day in general and within and around their stadium. Interviews were audiorecord ed and trans cribed verbatim. Final transcripts of each interview were sent to participants for them to confirm whether their comments were accurate and representative of their session. Collection of Quantitative Data The objective of the second part of the study was t o establish a relationship between game day variables and reported alcohol related crime and ejections at college football stadiums. Location was determined by locating stadiums on a map and confirming with interview participants as to whether the stadium was considered on or off campus. Reporting of ejections was determined through the interviews with administrators used in the qualitative componen t of the study along with the disclosure from law enforcement agencies in charge of those stadiums on game day that they did
54 not report escorting individuals out of the stadium if they were not charged with a crime. The other game day variables of start ti me, temperature, home team ranked, visiting team ranked, attendance, conference opponent, and in state rival opponent were determined from box scores available from team websites. Variables Each variable is defined based on the previously reviewed literat ure, a review of alcohol policies, and the semi structured interviews. The independent variables are measures of characteristics of college football game day at the stadium. These include alcohol policy related and other environmental variables related to the game day aspects and surrounding atmosphere of a college football stadium. Alcohol policy related variables include location of stadium (dichotomous), measured based on being located on campus or off campus, and whether ejections are reported The foll owing variables are infl uenced by characteristics of game day : time of game as measured by the hour of kickoff the temperature at kickoff (Boyes & Faith, 1993) whether the home team is ranked, whether the visiting team is r anked (Rees & Schnepel, 2009), a ttendance, in state rivalry, and conference game s The dependent variables are reported crime and ejections from the stadium. Types of individual reported crimes cited by law enforcement officers within the stadium were defined and deemed as noteworthy by the semi structured interviews with administrators and analysis of violations listed on crime ledgers/reports from each law enforcement group. These include public intoxication, possession of alcohol under the legal age, disorderly conduct, and assault am ong other infractions which result in arrest by law enforcement officers (Rees & Schnepel, 2009). Ejections involve removal of an individual from the venue for violating a stadium policy, but are not considered arrests
55 and are therefore not defined as repo rted crime for the purpose of this study The following is a list of how the variables were identified in the data analysis and how each was measured. Independent v ariables Reported Ejections. Law enforcement officers in some locations record the ejections of patrons, while others will escort perpetrators out of the stadium without recording the offense E jections are reported (1) a nd ejections are not reported (0 ). Location. Location refers to where the stadium is located, either on campus (dummy coded as 1) or off campus (0) Off campus stadiums permit the selling of alcohol, while on campus stadiums prohibit sales. Home team ranking This variable describes whether the home team playing the game is ranked in top 25 of a national poll The variable is meas ured by if the home team participating in the game measured was ranked in top 25, yes (1) or no (0), based on the ranking listed in the box score. Visiting team ranking. Whether the visiting team is ranked is also an independent variable. The variable is m easured by if the visiting team is measured was ranked in top 25, yes (1) or no (0), based on the ranking listed in the box score. Game time temperature. The temperature at game start in degrees Fahrenheit based on temperature listed on box score. Attendan ce This is the attendance of each game listed in the box score. Rivalry game. A game between in state rivals. The variable is measured as yes (1) or no (0). Conference game. A game played between two members of the same conference. The variable is measure d as yes (1) or no (0).
56 Start time. The start time of the game is signified by the hour of start of game (0 24). Year. This is the year in which the game was played. The years of the study are 2008, 200 9, and 2010. Dependent v ariables Crime. This is the co unt of alcohol related arrests reported within the stadiums of study. The total number of alcohol related arrests where individuals were charged with a crime carried out by law enforcement officers within the stadium per game was counted. Ejections. Refers to the count of ejections reported at college football stadiums. Those who are ejected are forcibly removed from the stadium by law enforcement officers but are not arrested or charged with a crime. Accessing crime and ejection data varied by individual s tadium and law enforcement agency. Crime a law enforcement incident that results in an arrest, was counted from incident reports archived by the police departments with jurisdiction over the stadium at five of the stadiums. Those five university athletic departments did not compile those records or permit the researcher to access them. As a result, the police departments in charge of game day law enforcement gave records of arrests and all other alcohol related law enforcement act ivity to the researcher and he conducted a count of the relevant crimes. Records archivists pulled crime report documents from the day of each game of interest. Two stadiums kept records of all law enforcement activity as reported by the police departments These records were obtained from the stadium administrators. O ne was via spreadsheet with a detailed breakdown of all incidents recorded per game while t he other had a count of total incidents, including alcohol related crimes and ejections in tabular fo rmat. Ejections counts were given for four
57 stadiums, while three stadiums did not provide that information due to the policy decision of removing individuals from the venue without producing a written report After the records were analyzed, a count of tot al crime and ejections within the gates of each individual stadium for each game date was obtained. Data Treatment Interview Analysis Data was analyzed inductively using criteria establi shed Moustakas (1994) and Berg (2002). The procedure used for analyzing data was phenomenolog y ( Moustakas, 1994 ). All data was coded by hand due to the small sample size (n = 8) E ach interview was read one time to expose higher order themes. Common themes that repeatedly emerged were identified a mong the intervi e ws Quotations that were representative and correspondent to those themes were chosen and highlighted by the researcher. A copy of each interview was printed out to facilitate determination of relationships between corresponding quotations in which themes were identified. Every statement relevant to the topic was considered to have equal value and subsequently placed into a theme (Moustakas, 1994). The relationship among quotat ions within the themes was noted and categorized accordingly to provide meanings and the fundamental nature of the phenomenon. The results are presented in Chapter 4. Statistical Analyses Two statistical analyses were chosen to aide in the analysis of the relationship between college football game day characteristics a nd reported crime and ejections canonical correlation analysis and multiple regression A c anonical correlation analysis was conducted to determine the relationship among variables within two set of variables, along with the correlation between the indepe ndent and dep endent variables
58 Tests of canonical dimension to determine the significance of the canonical dimensions as influenced by the dependent variables, calculation of redundancy indices, canonical weights, canonical loadings, and canonical cross lo adings were all obtained to determine the strength of the relationship between game day characteristic variables and reported law enforcement activity, ejections and crime, as a part of the same model. A fter considering the relationship between sets of variables, the impact of the independent variables on each dependent variable separately crime and ejections was determined using multiple regression. The first model regressed crime on the 10 independent variables. There we r e 126 observations of games within seven stadiums. The independent variables include ejections, location of the stadium (on campus or off campus), time of game, temperature, attendance, conference opponent, in state rival, ranked home team, ranked opponent, and year. The second model regressed ejections on 9 independent variables for the 80 games in four stadiums that reported ejection. The independent variable of reporting of ejections was excluded as a variabl e in this model. All of the statistical analyses were conducted using R version 2.11.1
59 Table 3 1 Description of participants in interviews. Venue Participant Position Stadium Location Interview Location Approx. Capacity Stadium A Ass istant AD/ Events Director On campus In office 20,000 Stadium A Facility Manager On campus In office 20,000 Stadium B Assoc iate AD/ Operations Director On campus In office 90,000 Stadium C Associate AD/ Operations Director On campus In office 48,000 Stadium D Facility Manager On campus In office 83,000 Stadium E Ass istant AD/ Operations Director Off campus In office 20,000 Stadium F Events Director Off campus In office 70,000 Stadium G Events Director Off campus In office 75,000
60 CHAPTER 4 RESULTS The results of this study are presented in two parts The first part presents a report of the interview responses of the participants categorized by theme. Discussion, explanation, and analysis of this content are covered in Chapt er 5. The second part of this chapter presents the statistical analyses which include canonical correlation analysis and multiple regression As with the qualitative component, d iscussion, explanation, and analysis of proposed relationships between the qua litative and quantitative results will be presented in Chapter 5. Part 1: College Football Game Administrative Alcohol Consumption Containment and Enforcement Interviews were conducted on the phenomenon of alcohol consumption policy enforcement and implem entation at college football stadiums to determine the emerging themes and resonating perceptions of facility and operations administrators at the seven schools that play football in NCAA Division I FBS in a Southeastern state Analysis of the interviews y ielded a number o f interconnected themes related to the process of policy enforcement, efficacy of policy, rationales for stadium policy, difference between stadium location and perceptions of issues related to college football game day alcohol consumption by patrons and spectators. These themes include policy rationale and enforcement, the difference between on campus and off campus facilities, sales of alcohol versus no sales in the general area, control, education of fans, tradition and culture, and alco hol related issues experienced on game day. C ommon themes emerge despite differing perceptions and insights disclosed by the participants.
61 Policy Interpretation and Enforcement Overall, the stadiums reported in this study are split between allowing a) alcohol consumption in the entire stadium with the opportunity to purchase alcohol available to all fans, and b) allowing alcohol consumption only within the luxury seating areas. These latter areas include club seating an d suites, which are set back from the general seating area. All of the stadiums that sell alcohol to the general public allow beer sales, while Stadium E allowed liquor sales to the general public additionally. Among all of the respondents alcohol policy and crowd control enforcement are t he responsibility of local law enforcement agencies. Policy Origins and Development For each administrator interviewed, the historical circumstances behind why the stadium alcohol policies were developed were considered relevant to policy implementation an d en forcement Three of the stadiums studied s old alcohol and are located away from their university campuses One stadium administrator stated: alcohol policy is they definitely sell beer and actually liquor. We were fortunate enough to operate under sub vendors on the stadium grounds and actually liquor as well ( Stadium F). In stadiums that sell alcohol, the administrators had differing levels of awareness to the purchasing limits. Two venues allowed two beverages per person, while one individual was unaware of the limits that existed in his off campus stadium. One stadium they cut off alcohol sales at the beginning of the fourth quarter, so they would sell through
62 [O]ur policy was just adapted from what they already do [at the stadium]. their concessions people, which is an outside agency which comes in and, they are overseen by [stadium] concessions. Their policy is, obviously they are checking ID regardless of age, so able to buy two alcoholic beverages per time. Our policy is that, we have a soft shutoff at the end of half time, which means at the end of half time we have either security and/or police go to the back of the line and walk down d shutoff where at the end of half time we start, anyone in line able to buy alcohol up to that person. That soft shutoff lasts within that fifteen that line is done serving alcohol at that point. (Stadium G) According to this participant t he shut off time is implemented in order to give individuals time for alcohol to metabolize through their systems. This was their policy in their previous stadium and they continued it when they moved into their new venue. Stadium Location Differences The interviewees made distinctions between alcohol access policies i n different parts of their stadiums. Four stadiums located on university campuses, do not sell alcohol within their general areas but allow for sales or consumption within their luxury suites. When characterizing their alcohol policies a number of the administrators state that there is no alcohol in the stadium, but this only refers to the general bowl area. This sentiment is clear in the following comments from an administrator with an on campus stadium : Our alcohol policy here is there i s no alcohol i n the stadium. There [are] two places you can have alcohol, which is the stadium club, which is for all our club seat holders. They get wrist bands, they go into their own room, where they purchase their drinks, and drink their drinks in the stadium club, [and] seat holders, and also in our suites. We have alcohol in all suites. Once the alcohol stays in the suite hallway. (Stadium A)
63 Other stadiums that do not sell alcohol have similar policies. Stadium B established a bowl areas. The only pl ace that alcohol is served is in private suites. And that alcohol and the beverages are delivered at the beginning of the year or during the week. It is not delivered on game day as alcohol cannot enter the stadium on those days due to the stadium. No alcohol is sold in general patron areas of the stadium. The only area that alcohol lounge but is not permitted in the seating area The reason for this is the lack of a clear de alcohol into the seat s which is a typical policy. It has to stay indoo rs in that premium private area (Stadium C). Stadium D allows alcohol sales within its luxury skyboxes T he box holders decide whether they want to purchase alcohol and do so through a private entity that operates a club that is open year round for events and meals. Additionally, individu als may purchase alcohol in the club during games, provided they purchas e a membership. Enforcement Perceptions The ability to contain problems through careful implementation of their policies was carried out through stadium staff public law enforcement, and private security officers ministrator as follows: [S] omebody goes into Gate 13 and has got a bottle of whiskey on him and d and they are inside the gate, a nd the officer stops
64 [have] a prohibited item, they are cited for possession of alcohol then go to the booking room. That booking room is on the east side of the stadium. that, arrested for alcohol possession on campus inside the stadium, and if standpoint. They could have their season tickets taken away, get else front of a judge, remains to be seen as to how the severity of that charge of other law issue that you wou ld have outside the stadium. So, we just happen to have the opportunity here, locally in our stadium on game days because we are processing a bunch of different people. We have 90,000 people all at once; lice and we get a list of everyone who has been arrested or ejected and we cross check it against our season ticket holder list. And if anybody that was ejection letter or arrest letter, it depends. If you get have a conversation one on one with that person. If it happens a third time Other administrators ackn owledged that the revocation of season tickets for the behavior of any individual sitting in the seat of the ticket holder was also done specifically in Stadium C and G. Stadium G policy enforcement is comprehensive outside the stadium in parking lots, as well as within the stadium. Ejecting and arresting individuals who violate policy or laws is supposed to be part of the containment process. We have a decent number of ejections and even sometimes arrests, depending the severity of our, we note whether th ey are alcohol related or not. We actually have a fair amount of ejections that happen prior to the game being started. Not allowing fans into the stadium because they appear overly intoxicated. So, since we are on private property, we are able to eject pe ople right away and not even allow them into the stadium where you know they could cause even more issues once they started drinking. (Stadium G) Mobilizing p ersonnel in multiple areas is another strategy to enhance enforcement.
65 re going through and if they see fans not abiding they are able to hand [the policy pamphlet] directly to them and give in the parking lot. We got police in the parking lots. We hav e ABT in the I think the stadium does a good job in training their personnel as well as working with police to make g the best we can. I think that the concessions do a very d circuit cameras all over. Really able to get pretty much anywhere to think they do a pretty good job cameras t (Stadium G) Several of the stadiums enhance enforcement through a text mess aging system. Stadium security cannot be everywhere at once, and the cameras will not pick up misconduct unless they know where to focus. have a text message service, so fans are able text in any issues and then right away as soon as those texts come in we put a camera on it and we got it until someone, either law enforcement or security can get to I think it also allows us to check up to see make sure that security or police are following through the policies as well a s the concessions [I] n the concourses we have cameras who are able to get in onto the concession lines to make sure the shut off i giving out to many drinks of anything like that. (Stadium G) T hese stadiums provide a service where individuals can text their location and problem s they are experiencing. really good for us, because the tough thing about holding an have 45,000 staff. So we have to find a way to get feedback. So, this text messaging tool has helped us do that. Peopl e anonymously alert us about anything they want, and they do. (Stadium C) The game day operations command center gets the information, can zoom the cameras into the problem area and also dispatch security and/or law enforcement into the area.
66 [W] e found t game. And, those are issues that we never would have heard of. You know, would have to go to an usher or security person that person then if they clean up, or a medical issue, or security issue. Now it just happens instantaneous. (St adium B) It is unclear if this has led to a decrease in incidents, but it gives them another tool in containing alcohol related spectator misconduct. We can put the camera on them or we can [focus our] binoculars on them, whichever is easier, identify th at location quickly, and then dispatch someone there within a minute even it be non sworn security, or security. yone can text for the most part. And you can do it anonymously. If someone is sitting in front of you smoking, and we have a no smoking policy, you can say the going to be someone th ere to take care of it. Same thing if there is ask them. And (Stadium B) The system is intended to protect fa ns from possible repercussions or apprehension associated with reporting unruly fan behavio r. Administrators have found that if somebody does report a possible infraction to law enforcement and the perpetrator finds out it might lead to a confrontation between those parties potential altercations, and possible arrests Are we getting better, or improving. However, I think as the years go by and more people will be used to it, and more people will utilize it. So we may end up getting more
67 O n campus Compared with O ff campus Stadiums The administrators saw differences between on campus and off campus a lcohol policies, whether they were the actual variations in policies or perceptions of alcohol consumption behavior and alcohol policy efficacy. Stadium A administrators determined that their lack of alcohol sales in the general seating stemmed from campus policy that prohibited alcohol on campus. This follows from the idea of keeping students from consuming alcohol in connection with a sporting event Other venue administrators had campus is a little bit different. You bring it right (Stadium C). Additionally, the Stadium C administrator saw more responsibility being on [because] Thus, the location of stadium had much to do with the sales policy. I think the fact that we are an on campus has a lot to do with whether we have the ability to sell alcohol in our venue I think a lot of times stadiums that are off opportunity to do that. Maybe it could be a college team that plays at a professional stadium. (Stadium B) iolence in with not allowing alcohol sales ( as opposed to allowing limi ted consumption in luxury areas) at football games in his venue. The sentiment that off campus venues may sell alcohol and on campus ones refrain from sales was shared by the administrator at Stadium D when he responded
68 These were the on campus venues perceptions. Off campus venue administrators had the following comments on the differences between venues on campus and those not within the boundaries of universities. I guess the general rule is if you have an on general beer or alcoho l sales outside of club level or special are a. But since campus they alleviated the pressure there. We saw the value of the revenue we get from it and in our situation revenue drive s the train in most cases. (Stadium E) The participant from Stadi um F suggested that moving the crowd they currently had at an off campus venue would not change much with regard to game day problems. He know if it would result in more a rrests or ejections but you would have a slightly different impact on the alcohol sales policy decision. Alcohol Sales versus Alcohol Prohib iti on in General Seating The re was considerable tension between the participants perceptions of the financial benefits of alcohol sales as opposed to the risks that such sales created of increased misconduct. The ra tionale for not selling alcohol was generally associated with the potential for problems to occur within the stadium. The interviewees recognized that access to alcohol in or around a stadium, one way or another is going to cause misconduct and other related problems. trator implied this. because the student mentality is going to be, you know a binge mentality
69 their buzz on, and try to drink more so it will last them through the game, get it inside. It seems to be a little different than what a lot people have, are in the tailgating areas, by correlation there. I think it actually might be quite the opposite. It may be t you know. (Stadium E) Yet, t his administrator also thought that the idea that on campus stadiums do not sell alcohol hurt the smaller schools financially. They took in a percentag e of alcohol sales at their off campus venue which totaled around 8,000 to 9,000 dollars per game. a lot of money to be mad e in it, and I think in a situation where you have an ecision to make when a commented on how the atmosphere of the game is affected by the presenc e of sales in the stadium. e. I think just about every stadium we go to, at least some of the smaller [venues] our conference schools. Everybody that we go to, it seems like the students are right behind us, right behind the bench. You know, there are some places that are better th correlation? It could be. But, any other place i t could be alcohol. (Stadium E) Another conference rival sells alcohol within their on campus venue and is considered to be a dangerous environment by this administrator. ght on top of you. We had an incident a couple of years ago where one of our coaches was on the verge of going in to the stands because the kids were spitting on the players and saying things. They had those ces and hitting them with literally over the top of you ... it was pretty rough,
70 but, we travelled with a couple of campus police so we were keen to it. We H owever, this administrator believes that allowing alcohol sales with careful enforcement and control would lead to a low level of game day issues. Financial Benefits Another administrator believed that allowing alcohol sales makes his school a good partner with the off consumption is something the fans want to partake in during games. He also believed that the amount of a lcohol related issues on game day would not vary between venues that sold alcohol and those that did not. the people who are prone to use alcohol or consume it to the level th at it becomes a situation I think are going to do it whether a stadium sells it or get a beer inside once they get in then they might have the extra beer outside. (Stadium F ) So, he did not perceive there to be much of a difference in alcohol abuse on game day between venues that sold alcohol and those that did not. Additionally, he b elieved that a stadium that prohibited alcohol sales would not necessarily have fewer issues. On the contrary, it meant that misconduct would be greater outside the stadium. part to go in with th at assumption because I think that people would that impression it would probably be the wrong impression. (Stadium F) This above statement consistent with the belief of Stadium E alcohol sales are unlikely to lead to more alcohol related issues within those stadiums.
71 focused on financial advantages to having alcohol sales with the stadium which outweighed any behavioral issues. and take there. You have some increased fan behavior issues, and yo u just [have] to weigh the pros a nd the cons verses the revenue created from with the stadium and a partnership with the concessionaires, and some contractual things as far as alcohol sales there. Ultimately they a re being hurt too if we stop vending alcohol, they get hurt too, so they have a vested interest as well, in our concession numbers. Alcohol sales create revenue, and that is the tradeoff that the off campus stadium make s betw een raising money for the depar tment or more behavioral control in the stadium. Stadium F does not make any revenue from alcohol sales. However, Stadium G does profit from alcohol sales but struggle s to reconcile making money with containing problems. ultimately operating budget is affected somewhat in part by our concession from those sales, Therefore, a ll of the administrators in stadiums that sell alcohol saw some benefits to allowing sales within their venues whether it was from a revenue s standpoint or not Having it Both Ways: Access for Few, Prohibition for Most The venues that sell alcohol seem to think that prohibiting alcohol sales in their general seating makes their stadiums prone to fewer issues, even though there is an acknowledgement that they are likely turning down increased revenue for their athletic [S]omething that is unique to this environment that does not sell alcohol, as opposed to a stadium that does, is that we find during the game that our
72 issue and concerns regarding health and security, medical and security issues will peak at the beginning of the game then go down as the game goes on. Whereas if you visit with folks that sell a lcohol in their stadiums, a drinking during the game. So we find our spike early in the game and it stadiums, or where the beginning and continue as the game goes on as it relates to people I think that not allowing alcohol sales njoyment of the part of the disruption for a lot of fans of the event itself. I think on the flip side, for alcohol sales, a positive aspect is I think there is a demand for that as it related to the fans now days, and I also think from a monetary standpoint you can generate increasing revenue from alcohol sales. Pretty substantial revenue you can generate from alcohol sales, beer and alcohol, no doubt. (Stadium B) Other stadiums admini strators agree with the demand for alcohol at college football games. The thinking is alcohol sales might change how one approach es law enforcement and develop and implement policy. The restrictions would change. usly an age parameter. When are you serving? When are you cutting off? What type of training? Who are the employees? Are they university employees? Are they [concessionaire employees]? Is there a level of insulation between the university and the alcohol s ales? What type of funding benefit could come as result of how many budget problems could you solve through alcohol sales in your stands. (Stadium G) These stadiums come up with t he compromise of allowing alcohol sales in luxury seating, which is the case in all of the on campus stadiums. The stadium administrators understand that they are forgoing a large portion of revenue by not allowing general seating alcohol sales. Overall, t he distinctions between atmosphere and enforcement potential lead to the concept of control.
73 Contro l The need to c ontrol alcohol consumption because of the issues and byproducts of alcohol induced mis behavior emerged as an important theme. Locations with fewer incidents attributed this outcome to management having achieved a high level of control. The administrator at Stadium A felt that their level of control was a reason for their low number of incidents. people enjoy themselves, p roblem with it. (Stadium A) H aving a sense control over alcohol consumption provided a justification for allowing sales in any area of a stadium. Control t hrough Prohibition in General Seating Keeping alcohol within the luxury seating and away from the ge neral seating, clearly present in on alcohol in limited means was a part of this, but so was higher level administrative policy making. [and] limit the alcohol to just the premium areas, the private areas and not to a public ow why demand from those patrons to have the ability to have alcohol. Now I guess guess there was a decisi on made that we would allow for it but we would do it just in a controlled environment where you would have that just in these premium areas inside the suites, inside the private club lounge. So again, the president might be the only one to answer why that was done. But again, you have more control over that than the entire seating area of the stadium, if that makes sense. (Stadium C)
74 Not allowing alcohol in the general seating area was a product of control for other stadiums as well. a level of containment of where the alcohol is going. Having those 89 suites, the large booster box, and the university club, it becomes a fairly significant popular. You could be looking upwards of 2000 to 2500 patrons, so a pretty good chunk of your est capacity within [our] s tadium. So the idea is to keep those controls within those confines, therefore we can monitor patrons that have had too much to drink whereas if the patrons are taking the drinks out and handing them to frie nds and family or in some cases creating their own little cottage industry patrons. So that was the principle reason for why those spots were designated for alcoh ol not to leave from. (Stadium D ) Therefore, the ability to control dictated the limitations on where alcohol was sold. Control Where Sales are Permitted in General Seatin g Even at a stadium that sells alcohol in its general area, the rationale for doing so was that cont rol could be maintained. The interviewees perc e i ved a small number of people exhibiting problematic behavior at the games. that type of person as a patron at your game and there are also other people around that person helping to maintain that control. Other fans and friends of that person. I thin part of the reason that our [incident] numbers are so low. (Stadium F) There is a sense that mutual fan responsibility leads to control, not only the law enforcement and policy initiatives put into place within that venue. NCAA Status as a Control The status of being an NCAA institution also affects their alcohol policy along with being on campus and in turn impacts the type of con trol they have over policy making
75 alcohol sales at its championship events and also encourages institut ions to not sell alcohol at their home events. Education and Alcohol Policy All interviewees mentioned the role o f education as an important theme in relation to stadium alcohol policy, and it emerged in two ways: First, in terms of the need to educate fans about behavior w hic h stems from the responsibility of universities as educational entities and second, in terms of t he educational environment in which college football operates Stad ium policy makers claim they have a role to play in educat ing fans about alcohol consumption and its problematic byproducts. All of the venues have fan guides that have their policy information available for fans, and have game day announcements over publi Other venues feature game day messages on their video boards about excessive alcohol use a nd game day safety. Educational initiatives were considered virtually educate people on how to act, and sometimes keep people aware of their surroundings, to know who to look out for Stadium G also found education of patron s to be an extension of control. procedures and hope that they are able to act in an adult manne to they can do it in a manner that they are still able to enjoy the game, and not we have sec urity, and police and we have all means we at our disposal. (Stadium G)
76 While seemingly a forced exercise, this administrator thought education was essential to alcohol policy control within and around their home stadium even though potential difficulties in educating patrons were acknowledged. y car to make and e njoyable atmosphere and doing so responsibly. (Stadium G) The importance of educating was echoed in Stadium C. One of the reasons was fans before having to resort to arres t or ejection. somehow, the reason that person cooperates is because they know, we knew that rule bu Stadium A also put an emphasis on working with local law enforcement with educational communication initiatives We work with university police on, like I said, getting the word out and these are the policies, procedures you follow on campus on games days as whatever those rules are. (Stadium B) The police are also present outside the stadium with bicycle teams asses sing the
77 (Stadium B). Also an alcohol component exists on the public service announcements played before and during games. Additionally, the educational environment that surrounds college football as an event affiliated with places of higher learning also became apparent. The idea of education resonated more with the administrators affiliated with on campus facilities. is more ownership on campus. I think the university feels they have more responsibility campus sanctioned university activity, a controlling alcohol consumption in a differe nt way on campus, specifically not allowing sales within their stadiums, stems from this educational atmosphere. campus is because binge drinking is really a prevalent problem on ma ny college campuses and [our institution] is included in that, is the idea of do towards. (Stadium D) T hus the educational environment creates pressure toward keep ing alcohol sales out of university owned football stadiums or sharply limiting such sales Tradition and C ulture Tradition and culture surrounding college football game day were cited as a reas on for (in some instances) the prevalence or (in other instances) lack of alcohol related incidents on game day. There are certain accepted behaviors, practices, or ongoing traditions that might increase alcohol consumption. Additionally the culture surro unding a team or certain opponents can have an influence on game day behavior by fans. One administrator attributed alcohol consumption issues to excitement and
78 anticipation of the beginning of the season as well as more popular opponents, often rivals. Yo u certainly have a factor that plays in that you add in for the early season Maybe a little bit of an additional factor that goes in there. When those factors were combined with the game being against the [major in state certainly say quite a different bit of behavior from our fans than if those two This administrator perceived that the time of game may have some bearing on the amount of incidents or amount of drinking that occurs. People have more time to rised when I walk by tailgates at seven Numerous participants agreed that t he start time of the game, temperature and the opponent can impact w hen people arrive for games and start consuming alcohol. For a 7:00 p.m. game, folks may show up in the middle of the afternoon about three or four, and you still have three or four hours before kickoff. For :00 am is still plenty early in the that point, anyway. (Stadium D) Numerous administrators acknowledged that the lack of a tradition can affect the alcohol over ten years, and a fourth had only had football in an on campus venue for four years, with not much of a traditi on of school support. For two of the programs, the administrators cited not having a tradition as a reason for having more problems in the early stages of their home games. The other cited not having a tradition of tailgating or
79 football success as a reaso n for relatively few game day incidents at their venue. was a reason for the atmosphere on game day even though the games are off campus. The culture of [our] football fans for a lack of a better word has been tame. itude that our department has always had, and our program, our football program, and the university itself, you know kind of being a friendly place to be and trying to treat people the way we be treated when we go other places. primarily been. (Stadium F) The students have generally been responsible for fewer issues than general non student, non alumni fans, by his indication. The presence of a tradition can impact game day atmosphere and related fans have a reputation for poor behavior not fan friendly, but they Many of the incidents are perceived by the administrators to be caused by a few fans who take their roles as supporters of their teams too far. Alcohol related Issues Experienced on Game D ay The administrators admitted that alcohol was likely a culprit in game day issues in their stadiums. However, they differed on the level of seriousness of these issues and where they came from. The smaller stadiums generally had problems surrounding tailgating events, but the stadiu ms rarely had problems. Stadium s A and E did not see
80 alcohol consumption as a major problem with in their stadium, and felt that they experienced limited issues within their venue. Stadium E had noteworthy problems when they moved to their new stadium due to an enforcement crackdown by the Bureau of Alcoholic Beverages and Tobacco (ABT) They experienced over 100 arrests for underage alcohol consump tion at their games, troublesome due to the fact that the nistrator perceived his stadium had few alcohol related incidents. The other venue administrators acknowledged an array of alcohol related probl ems that stressed stadium operation or public safety staff, and/or led to law enforcement activity by game day security officials. Alcohol consumption by fans led to many of the disruptions experienced on game day. I would probably say, if we are looking at ejections and arrests, probably our aid stations, and at times during hot games, it can be overwh elming for out fire rescue folks. The majority, 50% 60% of those folks that are going dehydration or a fall, or vomiting, or something to that nature. So, I would say over 50% of the i ssues we deal with both security and medically, have to deal with alcohol use. (Stadium B). While issues arose due to alcohol consumption, he considered the 100 arrests or ejections occurring on game day to be minor in a stadium with 90,000 spectators pres related problems to be was a factor. A 1:00 pm game in November compared to a 7 :00 pm game in September which would be worse. this. Again most of our issues are the early season games that are at night. dictates it.
81 games have been in Septe the bigger alcohol issues. They usually occur, honestly they occur in the lots or before they get into the stadium. That is again, some of it may be heat relat ed because there [are individuals] who function we ll after a certain point and the police usually catch them before. The goal is if there are any been our biggest thing. (Stadium C) Other stadiums have found issues with rivals, no matter what time of year. Often those games are at night so that may impact the amount of incidents regardless of opponent. T he presence of empty bottles of alcohol in the stadium following games was an issues that still occurred even though police performed che ck of all individuals entering the venue So patrons would still smuggle alcohol into the stadiums. Additionally one of the stadiums allowed reentry. As a result of this policy individuals could consume alcohol outside the stadium as opposed to purchasing it. So, alcohol consumption by individuals in the general seating areas by underage individuals at stadiums with sales, and by those who could not buy alcohol in the other venues, was still widespread Overall Summary of Policies In sum, the interviews yie lded an assortment of policies that were e nacted by the many of stadiums and the reasoning for implementing them. The most high profile policy is the decision of whether to permit alcohol sales or not. Controlling alcohol abuse of patrons which may cause f an disruptions or crime and ejections in the stadium is rationale for prohibition of alcohol sales, while conversely; controlling alcohol abuse outside the stadium is cited as a reason for selling alcohol to patrons. The stadiums with general seating area prohibition allow alcohol sales or consumption in luxury suites or
82 club seats because it is a source of revenue, those fans are less likely to buy seats without the added perk of alcohol consumption, security can control the drinking in a smaller area, and these fans are considered less likely to be arrested or ejected for alcohol related infractions than students or othe r general seating fans. When alcohol is sold a number of policies are put in place. Checking identification of potential buyers is enforced to prevent underage patrons from buying alcohol, and in some stadiums individuals who want to purchase alcohol must receive a wristband to show they are over the legal age. Additionally, concessionaires put limits of number of beverages per sale (e.g., 2) along with a time during the game when sales are discontinued (e.g., the third quarter). Policies implemented outsid e the stadium are aimed at creating a safer environment inside the venue. The media, such as newspapers and websites contain messages to fans about the stadium policies. Not allowing any food or drink into the stadium serves as an attempt to limit or eradi cate any alcohol smuggled into the stadium. Educational programs before the games include bicycle police reminding individuals of alcohol consumption laws outside the stadium as well as distributing game day policy pamphlet to stadium patrons. This informs spectators about what behavior is permissible and what is not tolerated by the stadium owners. Within the stadium the text message system has been utilized to help fans patrol their own sections and summon police officers anonymously without fear of reper cussions. Security cameras enable a limit security and law enforcement staff to view the whole stadium without dispatching officers to every corner of the venue. Public service announcement, aimed spectator safety, before and during the game remind fans ab out what behavior is tolerate along
83 with reminders to not drink and drive after the game. Ejecting patrons and recording this action serves as the ultimate deterrent for alcohol related misconduct. Other stadiums do not reported ejecting patrons as a means of controlling potentially unruly fans. Season ticket accountability programs that revoke tickets if a seat user is arrested or ejected a certain number of times serves as an incentive for keeping tickets out of the hand of potential binge drinkers who mi ght cause game day problems. Also, promote good behavior at games and make it a point of pride to support a safe game day environment for the university community. Thus, these policies detailed by the administrators emerged from the quotations gleaned from the interviews categorized in the previously mentioned corresponding themes. Part 2: Relationship among Game Day Characteristics and Game Day Behavioral Outcome Variable s Canonical Correlation Results Canonical correlation analysis was performed using the R version 2.11.1 packages of CCA fda zoo fields and catspec to determine the relationship between two sets of variables, the game day characteristics (independent var iables) and crime and ejections (dependent variables) Tests of dimensionality for the canonical correlation analysis, as show n in Table 4 1 indicate that the two canonical dimensions are statistically significant at the .05 level. Dimension 1 had a canon ical correlation of 0.73 between the sets of variables, while for dimension 2 the canonic al correlation was lower at 0.46 The canonical correlations were of sufficient size to be practically significant. A redundancy analysis was performed on each canonic al function (dimension) to reiterate practical significance.
84 Tables 4 2 and 4 3 show the calculation of redundancy indices for each of the canonical dimensions, also known as functions. Tables 4 4 and 4 5 show the redundancy analysis of dependent and inde pendent variables for both canonical functions. The redundancy index for the independent variables in the first function is .3654, explaining 36.54 % of the dependent variable variance The dependent variables explain ed 13.50% of the variance in the indepen dent variables. For the second function, the independent variables explain ed 22.36% of the variance in the dependent variables while the dependent variables only explain ed 2.88% of the independent variable variance. Together 58.9% of the variance in the de pendent variables wa s explained by the independent variables. Practically, the independent variables accounting for close to 59% of variance is adequate to declare significance in the redundancy index. Table 4 6 presents the standardized canonical coefficients for the two dimensions across both sets of variables. These results are in terms of standard deviation units for one standard deviation change in the canonical dimension. There is no rule for reporting the size of a standardized canonical coefficient, but .30 is often considered a good cutoff (StatSoft, 2011) For the game day characteristic variables, the first canonical dimension wa s most strongly influenc ed by reporting ejections ( .502 ), conference game ( .449 ), attendance ( .366), start t ime ( .338), and r ivalry game ( .334 ); and the second dimension wa s most strongly influenced by attendance ( 1.756 ), reporting ejections (1.332), location ( .608), home team ranking (.498), and start time ( .460). For in cident variables, the first dimension was comprised of eje ctions as the dominating variable ( 0.990) For th e second dimension crime ( 0.994) was the dominating variable
85 Tables 4 7 and 4 8 show the structure for the two canonical functions canonical load ings and canonical cross loadings An interpretation of the canonical loadings finds a loading of.9935 showing that ejections were a representative response to the variables in the first canonical function. The first independent variate has a range of load ings from .0527 to .5020 with all relationships negative except for one, with a very small value. Reporting ejections ( .8144), attendance ( .6728), and home team ranking ( .6723) were above the accepted canonical loading cutoff of .40 for canonical loadi ngs (Wong & Lan, 2001) For the second canonical function crime was highly negative ( .9989) in the second dependent variate. Away team ranking ( .4915) and the time of game ( .5132) had the highest loading on the independent variate. All other loadings we re considered too small for interpretation. An analysis of the cross loadings was also performed In the first canonical function the dependent variable of ejections had the highest cross loading of .725. This means that 52.6 % of the variance in ejections is explained in the Dimension 1. The independent variables with the highest correlations with the dependent variables were reporting ejections ( .5946) which accounted for 35 .4 % of dependent variable variance, attendance ( .5002) explaining 25.0 % of variance home ranking (.4913) accounted for 24.1 % of variance explained. In the second canonical function, the only cross loading above .with a relationship over .3 0 was crime ( .4561), accounting for 21% of variance in crime explained in the second fu nction. A discussion of the canonical correlation analysis and implications on the relationship between the independent variables, game day characteristics, and the dependent variables, misconduct behavior outcomes are discussed in Chapter 5.
86 Multiple Regression Results Crime as dependent v ariable Multiple regression analyses were conducted to examine the relationship between reported crime within the stadiums and potential predictors. Table 4 10 summarizes analysis results. The multiple regression mode l with all 10 predictors produced R = .22, F(10, 114) = 3.21, p < .001. T he reporting of ejections ( b = 14.08, = .433, r s = .001) had a significant weight indicat ing that less reported crime occurred in stadiums that reported ejections, but a very sm all structure coefficient L ocation ( b = 7.646, = .298, r s = .060) has a significant weight indicating showing that games played off campus experience an incr ease in crime but a small structure coefficient. Attendance ( b = .00039, = .842, r s = .340) h ad the most statistical significance of any other variable on reported crimes and a large structure coefficient. Also, the start time of the game ( b = 1.058, = .248, r s = .536 contrib uted to an increase of reported crimes the later the start time and had the largest structure coefficient The variables of year, temperature, rivalry game, conference opponent, home team ranking and away team ranking did not contribute significantly to the multiple regression model Ejections as dependent v ariable Mu ltiple regression analyses were conducted to examine the relationship between reported ejections within the stadiums and potential predictors. The multiple regression model with 9 predictors produced R = .52 52 F( 9 69 ) = 8.481 p < 0 001. As illustrated in Table 4 9 location of stadium ( b = 23.98, = .493, r s = .477) was significant suggesting that off campus stadiums experienced more ejections than those on campus. Attendance ( b = .0005 = 55 3 r s = . 428 ) was significant indicating that higher game time temperatures are associated with more ejections. Rivalry ( b = 1 3 75
87 264 r s = .3 65 ) had a significant impact on ejections with games featuring in state opponents experiencing more ejections Conference games ( b = 1 6 8 6 = .410 r s = .306. ) contributed to a significant increase in reported ejections. Also, the start time of the game contributed to an increase of ejections the later the start time. The variables of year in which the game was played home team ranking, away team ranking, and temperature did not contribute to the multiple regression model. Table 4 1. Tests of canonical dimensions Dimension Canonical Multiple Correlation Wilks Lambda F df1 df2 p value 1 .73 .367 7.41 20 228 <.001 2 .46 .786 3.47 9 115 <.001 Table 4 2. Calculation of Redundancy Indices for the First Canonical Function Variable Canonical Loading Canonical Loading Squared Avg. Loading Squared Canonical R 2 Redundancy Index Independent variables Report .8144 .6631 Year .0896 .0080 Location .1855 .0344 Home Rank .6728 .4527 Away Rank .2898 .0840 Attendance .6850 .4692 Temperature .0243 .0006 Rivalry .3262 .1064 Conference .0703 .0049 Time .1603 .0257 Independent variate 1.8490 .1849 .730 .1350 Dependent variables Crime .1483 .0220 Ejections .9935 .9870 Dependent variate 1.009 0 .5005 .730 .3653
88 Table 4 3. Calculation of Redundancy Indices for the Second Canonical Function Variable Canonical Loading Canonical Loading Squared Avg. Loading Squared Canonical R 2 Redundancy Index Independent variables Report .1936 .0375 Year .0099 .0001 Location .1054 .0111 Home Rank .1614 .0260 Away Rank .4195 .1760 Attendance .1875 .0352 Temperature .0676 .0046 Rivalry .1629 .0265 Conference .2066 .0427 Time .5132 .2634 Independent Variate .6230 .0623 .4622 .0288 Dependent variables Crime .9989 .0220 Ejections .1139 .9870 Dependent Variate 1.0090 .5054 .4622 .2336 Table 4 4 Standardized Variance of the D ependent Variables Explained by Their Own Canonical Variate (Shared Variance) The Opposite Canonical Variate (Redundancy) Canonical Function Percentage Cumulative Percentage Canonical R 2 Percentage Redundancy Index 1 .5005 .5005 .730 .3653 .3653 2 .5054 1.0059 .462 .2236 .5889 Table 4 5. Standardized Variance of the Independent Variables Explained by Their Own Canonical Variate (Shared Variance) The Opposite Canonical Variate (Redundancy) Canonical Function Percentage Cumulative Percentage Canonical R 2 Percentage Cumulative 1 .1936 .1926 .730 .1350 .1350 2 .0099 .2035 .462 .0288 .1638
89 Table 4 6. Standardized Canonical Coefficients (Weights) Dimension 1 Dimension 2 Independent Variables Report .5020 1.3202 Year .0818 .0241 Location .1765 .6106 Home Rank .1868 .5115 Away Rank .0527 .2856 Attendance .3658 1.7558 Temperature .2168 .0501 Rivalry .3382 .1158 Conference .4493 .1828 Time .3338 .4641 Dependent Variables Crime .1139 .9941 Ejections .9895 .1484 Table 4 7. Canonical Loadings Dimension 1 Dimension 2 Correlations between the independent variables and their canonical variates Report .8144 .1936 Year .0896 .0099 Location .1855 .1054 Home Rank .6728 .1614 Away Rank .2898 .4195 Attendance .6850 .1875 Temperature .0243 .0676 Rivalry .3262 .1629 Conference .0703 .2066 Time .1603 .5132 Correlations between the dependent variables and their canonical variates Crime .1483 .9989 Ejections .9935 .1139
90 Table 4 8. Canonical Cross Loadings Dimension 1 Dimension 2 Correlations between the independent variables and their dependent canonical variates Report .5946 .0895 Year .0654 .0045 Location .1354 .0487 Home Rank .4913 .0746 Away Rank .2116 .1939 Attendance .5002 .0867 Temperature .0177 .0313 Rivalry .2381 .0075 Conference .0514 .0955 Time .1170 .2372 Correlations between the dependent variables and their independent canonical variates Crime .1083 .4561 Ejections .7254 0526 Table 4 9 Regression results for crime as dependent variable Variable b Std. Error r s t value p value Report 1 4 08 4.539 .549 .001 3.102 .0024 Year .03 1.279 002 .03 0 .026 .9794 Location 7.6 5 3.149 .298 .06 0 2.428 .0167 Home Rank 5.54 3.042 .213 .002 1.821 .0712 Away Rank 4.3 8 3.184 .125 .475 1.375 .1720 Attendance .00 .000 842 .340 4.144 <.0001 Temperature .0 7 .13 2 .046 .060 .500 .6181 Rivalry .57 3.230 .016 .091 .177 .8597 Conference .90 2.602 03 5 .185 .347 .7294 Time 1.0 6 3.704 2 48 .536 2.856 0051 Residual standard error: 11.59 on 115 degrees of freedom Multiple R squared: 0.2207, Adjusted R squared: 0.1529 F stat istic: 3.256 on 10 and 115 DF, p value < 0.00 1 n= 126
91 Table 4 10 Regression results for ejections as dependent variable Variable b Std. Error r s t value p value Year 1.47 2.112 .0 59 .046 .698 488 Location 23.98 4.933 .493 .477 4.860 <.001 Home Rank 7.4 2 4.812 .018 .498 .154 878 Away Rank 3.29 4.818 061 .218 .683 497 Attendance .00 .000 .553 .428 4.341 <.001 Temperature 31 .201 .142 .001 1.546 127 Rivalry 13.75 4.910 .264 .365 2.801 001 Conference 16.86 4.081 .410 .306 4.131 .003 Time 1 .40 .583 .208 .243 2.404 .019 Residual standard error: 15.05 on 69 degrees of freedom Multiple R squared: 0.5252, Adjusted R squared: 0.4633 F statistic: 8.481 on 9 and 69 DF, p value < .0001 n =80
92 CHAPTER 5 DISCUSSION AND ANALYSIS The objective of this mixed methods study was to explore the relationship s between and among alcohol policy development, implementation, and game day behavioral outcomes. Qualitative findings of this study ranged from determining rationales for alcohol policies on game day, differing perceptions for why alcohol should be sold in college football stadiums, and reasoning for why certain policies are in place, and the relationship between alcohol policy and spectator behavior. The concepts used in SCT provided a basis for organizing the qualitative inquiry. Quantitative data showed a relati onship between the game day characteristi c variables and alcohol related deviance outcomes. The following presents a discussion and analysis of the findings. Linking Alcohol Policy Consideration with Social Cognitive Theory The interviews with eight administrators at NCAA Division I FBS athletic departments provided insight into the thought processes of policy development but also the social behavioral issues that administrators need to plan for when planning game day crowd control. Control issues are generally geared toward providing a safe environment for patrons. However, alcohol consumption is often at the forefront of game day public safety policies. Interview results showed a focus on eight themes: Policy rationales policy enforcement, the locat ion of stadiums, sale of alcohol dependent on stadium location, control, education, tradition and culture, and alcohol related issues. These themes can be interpreted and best explained as environmental strategies that are comprehensible to the theoretical concepts of SCT The concepts consist of reciprocal determinism, incentive motivation, facilitation, self regulation outcome expectations, self efficacy, collective efficacy, observational learning, and moral
93 disengagement ( Bandura, 1977, 1986, 1997, 199 9; McAlister et al., 2008) In applying this theory that explains the experiential, observational capacities of individuals, function and adaptation of groups can explain the behavioral processes (Bandura, 1986; Bandura 1997) we are able to evaluate the n ecessary components that go into developing, implementing, and improving alcohol policies in college football stadiums The following will show how the adaptive capacity of a group, in this case the college football stadium inhabitants which include admini strators, stadium employees, and spectators, can influence how the extent to which alcohol consumption problems can be remedied, controlled, and prevented. Environmental D eterminants Policy makers look at the physical space and rules for behavior that su rround a person that might be affected by a policy and they try to alter an environment that encourages preferred behavior and discourages disapproved behavior. The policy rationales and enforcement protocol are influenced by environmental determinants of behavior: reciprocal determinism (Bandura, 2002) which leads to the behavioral modification concepts of incentive motivation (Bandura, 1969, 1986) and facilitation (Bandura, 1988) Reciprocal determinism can be seen in the planning of alcohol policy it its elf. This can be done through policies that change behaviors or at least enable individuals to make choices that influence drinking decisions. Uniformly among the stadiums of interest, patrons were prohibited from bringing alcohol into the stadium. With th e exception of one stadium, active searches or pat downs were conducted by security staff. The alcohol sales policies differ between two groups of venues. The on campus venues do not allow alcohol consumption or sales in general seating, while those locate d off
94 (2005) findings that only 15% of institutions sampled allowed alcohol sales at on campus sporting events. However, all the stadiums subject to examination in the interviews allowed alcoh ol consumption in at least one section of the stadium. Three stadiums allowed sales in club seating, while another allowed consumption in the club and suites areas, but did not have sales during the game. The allowance of alcohol in higher end seating is c consumption poli cies for luxury seating (Opdyke & Kesmodel, 2009). There was an agreement among participants that schools did not sell alcohol in the general seating on campus because of th e amount of student patrons, many who are under the legal drinking age, and the responsibility that the university has as an academic institution. Those selling alcohol had limitations on number of b everages per transaction, along with cutting off sales at a prescribed time. Based on the sales policies patrons who are of age may make choices as to their alcohol consumption. In areas where alcohol consumption is prohibited, some spectators choose to drink either before the game in tailgating parties, or smug gle alcohol into the stadium. This is especially the case for underage students shown in numerous previous literature (Glassman et al., 2007; Glassman et al., 2008; Glassman et al., 2010 ). Many stadium administrators noted that individuals also consume hea vily before the game, even when alcohol sales were permitted inside the gates. So, whether alcohol is completely limited, or available with control, individuals still have an influence on the game day environment with regard to drinking. Control of the env ironment is a major facet of alcohol policy which was agreed upon by administrators Limitations on alcohol intake and the choices that individuals
95 make in their drinking exemplify reciprocal determinism. I ncentive motivation is one ecological component can be applied to these policy implementations. A number of policy initiatives were outlined and detailed by the administ rators which follow the use of incentives, whether consciously or not. The use of punishment in dealing with excessive alcohol consump tion and related issues is widely known and dealt through arrest but can also be related to the levels of reported crime (DiIulio, 1995 ; Makkai, 2001 ) Ejecting or arresting patrons who do not abide by policy or laws were used by all seven stadiums explore d is the most time tested incentive. The question of whether arresting or ejecting unruly fans is actually enough of an incentive or abated excessive alcohol consumption remains to be debated For the most part the majority of individuals at games are not being dealt with by security or law enforcement even if they do drink heavily. Some locations found arresting or ejecting people before the game began due to over intoxication to prevent more issues from occurring before the game. Yet, there was a sense t hat more control of alcohol consumption led to less problems. All tickets have stipulations on the back that the privilege of viewing a game may be revoked if policies are violated. However, some locations have taken this one step further. Three of the sta diums implemented a program that includes written reprimands, face to face meeting, or at worst revocation of season ticket privileges of the holder if any individual sitting in that seat is arrested or ejected, regardless of if the season ticket subscribe r was present or not. This puts an onus on the season ticket holder to not only ensure that he/she and the people in the immediate party behave and do not encounter
96 law enforcement repercussions. Also, it makes people reconsider putting tickets on the seco ndary market Facilitation and Self regulation Bandura (2002) suggests that the environment must support new behaviors or behavioral change. If individuals are going to be influenced to cease abusing alcohol, the environment must facilitate behavior modif ication. Policy enforcement, education, and control themes can be considered informed by the concepts of facilitation and self regulation. Limiting intake of alcohol can also be a form of self regulation. Keeping alcohol away from students in the on campu s environment can be a form of facilitating members of the university community to not consume alcohol. For those who consu me alcoholic beverages in stadiums that provide it, having sales limits such as the two beverages per visit and cutoff time like the end of halftime at Stadium G exemplify facilitation. With regard to spectators being able to ensure a safe area in their seating along with contributing to the enforcement process, the text messaging programs provide a means to point out potential problems This can also be considered a form of self regulation. In addition, administrators in Stadium C and Stadium G pointed believed that people looked out for each other as a result of a culture instilled at their institutions, either encouraging individuals to act responsibly as a form of self regulation. Observational Learning Observational learning which is analogous to learning to perform a behavior through peer modeling (Bandura, 1986, 2002; Akers, 1979, 2009), was also a key guiding principle that emer ged from all participants in discussing policy rationale, enforcement, education and tradition and culture. The administrators noted that
97 targeting the learning processes at play with patrons attending college football games were as important as law enforc ement crack downs. One example of public services announcements to encourage safe drinking practices featured the coaches, or other well known members of the university community played before and during games Utilizing newspaper advertisement, articles, or distributing policy materials to fans were also a form of modeling to inform visitors about policies and procedures was favored by some of the participants. Policy guidelines were often handed to pregame tailgaters when they were viewed disobeying the laws/guidelines. The direct efficacy of these interventions is unknown. Prior research has shown that interventions using modeling behavior often have less effect on individuals in high frequency drinking conditions. (Ca rey, Scott Sheldon, Carey, & DeM arti ni, 2007). The culture and tradition of tailgating and alcohol consumption also lead to observational learning through peer modeling, whether it was to curtail problem drinking or actually encourage it varied by location. Younger students see older students abusing alcohol at tailgating parties and assume that is the way to participate in college game day. For one stadium, individuals seem to display more restraint that models less aggressive behavior, which may be a reason for a perceived lack of problems. Stadium perception. O utcome E xpectations Outcome expectations can follow from the modeled behavior or beliefs that individuals possess about specific behaviors T his means a determination of the consequences of certain behaviors the willingness of individuals to be influenced by
98 performing that behavior (M cAlister et al., 2008) Many of the major preconceived notions of the administrators were, if a venue has a large group of individuals in a small area, surrounded by a competitive atmosphere and alcohol consumption, problems will occur and they have to be ready to handle them. There is an acceptance that a rrests and ejections are going to happen in this environment People choose to consume alcohol, but the university and off campus venues also permitted its consumption before and after games. The decision of whether to serve alcohol or not and how it affected crow behavior was split among administrators. The locations of stadiums bred differing outcome expectations for the administrators. Off campus stadium administrators thought it was likely for more pro b lems to occur in stadiums that prohibited alcohol sales This was due to a in tailgating alcohol consumption or attending football games, an example of high risk celebratory drinking which is shown in numerous studies (Neighbors et al., 2006; Oster Aaland & Neighbors, 2007; Glassman et. al 2007). The Stadium E administrator cited the perception that there was an idea that not selling alcohol would decrease issues in the tailgating areas. The issue that arises is individuals try to consume more alcohol before a game in order to increase their level of intoxication Prior study has observed confirmed higher intoxication of individuals due to drinking more before games (Boyes & Faith, 1993). Other administrators cited that they believed that not allowing alcohol sales within stadiums would yield a similar amount of incidents, and those who believed not allowing sales meant fewer problems were mistaken. The other off campus stadium administrator perceived more problems to occur in venues that a llowed alcohol for purchase.
99 On campus administrators perceived fewer issues in venues that did not allow alcohol sales than their off campus peers. This was due to their ability to control alcohol consumption in a small, generally enclosed area. This was administrator. He expected fewer issues to occur within the stadium later in the game due to the lack of presence of alcohol in contrast to his observation that professional stadiums experience problems throughout the game. Howe ver things like time of game of hot temperatures were likely to produce more incidents in the on campus stadiums, and the administrators acknowledged they could not prevent this escalation in problems with no control over weather and scheduling. Self e ffic acy and Collective E fficacy Self misconduct and other alcohol related problems arise from themes of policy interpretation and enforcement, control, and education. The ability to influence individuals to not over consume alcohol becomes an important part of policy implementation. Collective efficacy is clear in the overall consensus that the stadiums have control over alcohol consumption and problems through providing it to the ge neral seating area with limitations and cut offs, not allowing the majority of spectators to consume alcohol, or containing it to one area of the stadium. Another illustration of this concept is cooperation between security and law enforcement to decrease alcohol related problems which occurs in all of the stadiums. One stadium administrator enabled patrons and game day personnel to feel like they could work together to create a safe environment on game day (Stadium C) Also, these stadium administrators also play a role as educational administrators So it is possible that they
100 consider it to be part of their job to educat e students on the benefits of controlling their intake of alcohol. Avoiding M oral Disengagement Violating standards for avoidance of violence or cruelty towards others is moral disengagement (Bandura, 1999). The tension between law enforcement officers and citizens can often lead to p otential problems, especially in heavily attended sporting events As shown in economic studies, alcohol availability is related to crime totals (DiIulio, 1995; Greenfield, 1998; Makkai, 2001) The rationalization for removed unruly fans from a stadium can be explained through moral disenga g ement. A number of administrators acknowledge that individuals must be removed from the venue when violating law or policy. Vilifying those who violate policy or creating a culture where individuals are arrested or eject ed might create a worse environment for fans. An arrest for violating a policy is likely to be inherently dehumanizing. But, the moral justification for arresting and holding fans in a cell during the game is as some of the participating stadiums do, thos e individuals broke the law and therefore should be punished. On the other hand avoiding moral disengagement seemed to be an implicit tone of the role of education and proactive law enforcement in game day policy. The main goal of policy was to enable to create a safe and enjoyable atmosphere for patrons, understanding that the majority of guests do not misbehave. The stated goal of not creating a police state while also giving individuals the opportunity to leave t he venue without any legal repercussions was one way that a venue avoided moral disengagement. Additionally some administrators might also day regardless of whether a lcohol is sold or not.
101 Intertemporal Substitution Participants observed that w hen there is a prohibitionary sales policy it encourages excessive alcohol consumption before and after games. SCT does not adequately provide concepts that completely explain the participants reactions to these observations. This part of the pheno menon is explained by the theory of inter te mporal substitution which suggests individuals will substitute for the inability to perform a behavior during a restricted period, by engag ing in the restricted behavior before or after regulation (Boyes & Faith, 1993). This is illustrated in the present study through the idea that p roviding a strict enforcement regime against alcohol consumption within the stadium risks encouraging patrons t o compensate for their inability to drink during the game by engaging in excessive consumption before or after the game. This causes administrators to treat the idea of absolute prohibition of alcohol consumption within a stadium as ineffective because it does not accomplish the objective of halting misbehavior. Yet, if there is a collective perception among the on campus stadium administrators interviewed that alcohol consumption causes problems beyond what people can control in the general seating if there are alcohol sales. They believe they can control misbehavior by only allowing alcohol for a few people in an enclosed area. However, based on the phenomenological approach, the small stadiums experie nce few problems, regardless of sales policy, and both on campus and off campus stadium experience alcohol related problems on game day. Therefore, what evidence exists that not allowing alcohol to the general public actually limits law enforcement activit y within the stadium? Predictive statistical models using canonical correlation and multiple regression analyses we re utilized to determine this.
102 Relating Policies to Outcome Canonical Correlation Findings The statistical portion of this study has shown th at there is a relationship between the stadium alcohol policies and other characteristics of college football game day and spectator conduct and law enforcement outcomes. Canonical correlation analysis found a strong relationship between the game day varia bles (independent variables) and alcohol related deviance outcome variables (dependent variables) Both dimensions tested together were statistically significant at the p <.001 level, as was the second dimension by itself, also at <.001 level. This means t hat both dimensions alone are statistically significant. Redundancy tests showed practical significance of the model, with over 58 % of the variance in the dependent variables explained by independent variates. The standardized canonical coefficients showed the following effects on the canonical dimensions, holding all other variables in the model constant. Canonical coefficient s (weights) are considered suitable for prediction (Hair et al., 2009 ). Wh ile there is no set rule for a suitable or significant ca nonical weight, a canonical coefficient of .30 or higher is considered noteworthy (StatSoft, 2011) For the dependent variables, ejections was the dominant variable in the first function accounting for almost 98 % of variance explained in the dependent variables, while crime was d ominant for the second function For t he variable of reporting ejections, a one standard deviation increase leads to a 0.50 standard deviation decrease in the score on the f irst canonical variate for the independent variable set when the other variables in the model are held constant. The other variables that had notable impacts on the first canonical variate were conference games yielding a .45 decrease, attendance with a .37 standard
103 dev iation decrease, rivalry games yielding a .33 decrease, and time yielding a .33 decrease. One standard deviation increase in ejections leads to a .99 decrease in the first canonical variate. The second canonical variate experienced the following effects h olding the other predictors constant in relationship to an increase of standard deviation of the variable : Attendance evoked a 1.76 decrease in the second dimension. Reporting ejections led to a 1.32 increase in the dimension 2 for the independent variable set. Location of the stadium accounted for a .61 decrease in the second variate set. Home ranking yielded a .52 increase in the second set. Additionally, start time of game led to a .46 decrease in the second variate set. While canonical loadings were rep orted, interpretation of them in this case is not as appropriate as weights, since the loadings are more useful in interpreting underlying constructs. Cross loadings are preferable over weights and loadings as they offer a direct interpretation (Hair et al 2009 ). Therefore, canonical cross loadings for the first function showed that ejections (.73) were related to the independent canonical variates, while reporting ejections ( .59) attendance ( .50) and home rankings ( .49) were related to their dependen t canonical variates The second dimension only had crime as a noteworthy value ( dependent variable. B ased on the canonical correlation analysis reporting ejections yielded the greatest decre ase (inverse relationship) in the reported crime variate (dimension 2) followed by home team ranking A ttendance followed by location of stadium, then start time of game had a direct relationship on and influenced an increase in crime. For the
104 first dimens ion ( t he variate related to ejections) reporting ejections had the greatest impact, followed by temperature, attendance, conference games, and rivalry games. Multiple Regression Findings Two multiple regression models were analyzed to determine the relatio nship between game day characteristic variables (e.g., temperature, time of game, etc.) and alcohol related deviance variables (e.g., crime and ejection s ) Each model had a different outcome variable : reported crime and reported ejections. The models showed that some of the independent variables had more of an impact or effect on the dependent variables. Crime The first of the multiple regression models considered the relationship between game day variables and report ed crime. Reporting ejections, location, attendance, and time of game were significant predictors of crime. Attendance was the most significant predictor of crime, with the beta weight (.84) and the third largest structure coefficient (.34). For every 1,00 0 spectators, a stadium was likely to have .39 more reported crimes. Venues reporting ejections are likely to have approximately 14 fewer crimes than those not reporting ejections based on this model. As corroborated by interviews with Stadium C, schools t hat did not report ejections were likely to arrest individuals if they did not leave the premises when, or were school that experienced low attendances and therefore experienced fewer crimes period. Off campus stadiums experienced approximately 7.6 more cr imes per game than on campus stadiums, however the structure coefficient was quite low ( .06). Also, of every hour later that a game started, 1.05 more crimes were observed in the total report, with a beta weight of .25 and the
105 highest structure coefficien t (.54). So, attendance and start time of game are considered the two strongest factors for increased crime on game day. The year in which the game was played, away team ranking, temperature at start time, and whether the game was a rivalry or conference g ame were not significant predictors. Home ranking was significant at the p = .10 level ( p = .07) with home teams being ranked yielding 5.5 less crimes than when they were unranked, but this was not significant at the designated .05 and therefore did not co ntribute to the model. Ejections The second regression analysis showed the relationship between the policy and game day characteristic variables on reported ejections within the stadiums. The multiple regression model from the included ejection as depende nt variable was a stronger fit than the model with crime as dependent variable ( R 2 = .52 v. R 2 = 22 ). These results possessed similarities and differences to the game day characteristic variables on crime Location, attendance, and start time of game were all significant, just like in the model with ejection as dependent variable in addition to rivalry and conference opponents. Off campus stadiums accounted for 24 more ejections than on campus stadiums, with beta weight of .49 and structure coefficient of .48. Games experienced .50 more ejections per 1,000 additional fans with a beta weight of .55 and structure coefficient of .43. In state rivalry games accounted for 13.75 more ejections than games without an in state opponent with a lower beta weight of 26 a nd structure coefficient of .37 Later start times accounted for 1. 4 ejections per hour later of start time but a lower beta weight (.21) and structure coefficient (.24 ). The canonical correlation results correspond well with the regression models since standardized canonical weights can be interpreted like beta weights (StatSoft, 2011)
106 Based on both types of analyses attendance was the strongest predictor of crime along with start time of game. Attendance and conference games were the strongest p redictors of ejections, as were conference games and rivalry games T ime of game being significant but have the smallest weights and structure coefficients of the five noteworthy predictors. However, due to being significant across both canonical dimension s and the two regression models, it is concluded that greater attendance and later start times of games have the biggest influence on alcohol related ejections and crime at college football games. T he Connection b etween Qualitative and Quantitative Analy ses A number of findings in the multiple regression analysis corroborate information gleaned from administrator interviews. Fewer reported alcohol related crimes and ejections were observed in on campus stadiums which is consistent with perceptions of on c ampus administ rators. Location affect ed alcohol related crime and ejections. Start time a ffected crime and ejections which also corresponds with administrator perceptions that they experience more alcohol related issues during night games Many administra tors discussed the issues they had with rivals, and in state and conference rivalry games significantly affected ejections. Attendan ce a ffected crime and ejections which is also a reasonable conclusion since the stadiums with higher attendances reported mo re problems in interviews. Conclusions Summary of Findings The author has been unable to locate literature prior to the present study, to determine the relationship between alcohol policy and game day variables on the one hand, and alcohol related incidents on the other hand within multiple stadiums. There
107 are fou r main categories of findings: First, t here are difference s in policies between stadiums. All of the stadiums allow alcohol consumption within their gates, but the areas in which they permit it varies. Three stadiums allow game day sales and consumption within all seating areas, while four do not allow alcohol sales or consumption in general seating but do allow suite and club seat patrons to buy or consume alcohol. Second, t here are discernable factors that affect the difference in polices. These factors include the presence of a stadium on campus, ther efore located in the educational environment, and the conflicting argument that allowing sales either contributes to or diminishes control of potential alcohol related misconduct. Through SCT it was possible to show how theory tacitly influences policy dev elopment process. Third, i t was possible to establish a relationship between policy factors and the relative numbers of reported alcohol related misconduct. Finally, t here was a relationship between factors other than policy on reported alcohol related mis conduct. The off campus stadiums sampled that sell alcohol to their general seating areas experience more alcohol related crimes and ejections problems B ased on empirical evidence shown in th is study, on campus stadiums experience less alcohol related pro blem as a whole, but the starting times of the games and attendance also have a strong impact on crime and ejection figures as well. Implications This study has a number of implications for stadium managers, law enforcement, and other practitioners who im plement and employ alcohol policies at college football stadiums. A relationship was established between greater attendance and crime and ejections. Since large crowds are desirable for college football games, stadium officials should implement other envir onmental strategies. Stadiums with high attendances
108 should consider introducing or continuing to impl ement more educational policies, such as utilizing local newspapers, university websites, and other forms of communication to inform patrons about their al cohol policies. Many stadiums uti lize the text messaging system to help lessen the strain on law enforcement and enable fans to report minor behavioral issues around them before they may become greater problems that affect more spectators. This may be a ta ctic that stadiums which are not currently employing want to look into implementing. The results also suggested that l ater games produce d more crime and ejections Often game time is determined by television companies rather than by the stadium or universi ty officials Also games featuring in state opponents and conference games yield ed higher ejection figures. F or later games and/or contests of increased interest to fans (rivalry games) i n addition to other environmental policies, stadium administrators should consider implementing tailgating restrictions that do not allow patrons to drink on stadium property before a certain time. This could eliminate the amount of alcohol abuse that occurs on stadium property before a game. Interviews revealed that some stadiums did not report ejections. In one case those who did not leave the premises when asked by law enforcement were arrested and charged with trespass. In other cases ejections were the bulk of law enforcement activity in lieu of charging individuals w ith crime. Thus, making an attempt to standardize law enforcement practices among stadiums might benefit law enforcement officers and fans. Fans who travel between stadiums would know what to expect and what behavior is considered appropriate. Spectators f rom a venue with more lenient ejection policies may experience more trouble and possible arrest or ejections when
109 travelling to other venues due to the expectations they established at their home stadium. Report ing ejections rather than allowing individual s to leave without documenting the incident would help law enforcement and stadium officials to obtain a better idea of the amount of misconduct occurring at games. In addition, implementing an acco untability program that punishes season ticket holders who violate stadium policy would also serve as a deterrent to individuals who might be prone to excess drinking and related misconduct. Conclusions of this study suggest that o ff campus games experience more ejections. Less of a relationship exists between of f campus games and crime but it was still statistically significant An explanation for the difference in total crimes and ejections in off campus stadiums is alcohol sales in general seating yields higher amounts of crime which is consistent with research on the impact of alcohol outlets on crime (DiIulio, 1995; Gyimah Brempong, 2001, Gyimah Brempong & Racine, 2006) This suggests prohibiting alcohol sales within the stadiums off campus could yield lower instances of crime and ejections Limitations This s tudy has a number of limitations due to the nature of the data collected. The qualitative portion relies on interviews and potential guardedness of interview participants due to the perceived sensitive nature of the content is a possible concern. One of th e interviews recorded occurred with two administrators participating at the same time which could have possibly biased the responses. For the data collection of crime reports, measurement of crime and ejections differs by agency. Since different agencies w ork each stadium, the type of enforcement and whether ejections and crimes are cited consistently from game to game may vary. Using crime and ejections is the
110 only means to measure gam e day alcohol related problems. Some departments did not report ejection s but rather escorted individuals out of the stadium or arrested them for non compliance. Deeming a law enforcement infraction an alcohol related crime was a judgment call on the part of law enforcement officers as well. For the game day characteristics it is virtually impossible to isolate every variable related to alcohol excessive drinking that occurs before a game. However, the models do not account for pregame drink ing including drinking that might occur in stadium owned parking lots In addition, a ccess to data was also a limitation. Some stadiums kept records that were obtained from police departments and made it accessible for review, while others were unwilling to share their reports. Finally, the data is g eneralizable to those stadiums in the sample but not necessarily to all NCAA DI FBS stadiums. Future Study This study lays a foundation for more exploration into the influence of alcohol policy, alcohol consumption, and other game day variables on spectator behavior. Expand ing this approach to include more stadiums is recommended If feasible, a comparison of a ll NCAA Division I FBS stadiums should be conducted In addition, c onsidering an analysis of college basketball arenas, since many sell alcohol, but a different game day culture exists which might yield varying results Stadiums with enclosed lots with def ined areas so crime related to tailgating can be determined. In addition, using the s ame va riables and characteristics to replicate this study in professional games that allow alcohol sales might also prove worthwhile
111 APPENDIX A INFORMED CONSENT FO RM Informed Consent Protocol Title: Stadium Alcohol Policy Characteristics: An Examination of Alcohol Policy Implementation, Differences, and Effectiveness Please read this consent document carefully before you decide to participate in this study. Purpose of the research study: To determine the reasons for differing alcohol policies at on campus college football stadiums and off campus football stadiums. Moreover, how do policies address alcohol related injury, reported crime, arrests, traffic accidents, strains on law enforcement, and impact the quality of life for nearby residents on game days. What you will be asked to do in the study: You will be asked to take part in an interview with the researcher about your role, attitude, and involvement in alcohol policy implementation at the college football stadium you supervise during home games. Also you will be asked information about policy enforcement and reported crime at your venue on game day. Time required: hour 1 hour Risks and Benefits: There are no risks in taking part in this interview, nor are there any anticipated direct benefits, however, you may learn about the types of policies b eing implemented at other stadiums Additionally, participants will be offered a copy of the study when completed. Compensation: You will not be compensated for participating in this research. Confidentiality: Your identity will be kept confid ential to the extent provided by law. Your information will not be shared with anyone else. Your interview will be classified by the size and the status of the stadium (on campus or off campus) that you supervise. Your name will not be used in any report. Voluntary participation: Your participation in this study is completely voluntary. There is no penalty for not participating. Right to withdraw from the study: You have the right to withdraw from the study at anytime without consequence. Whom to conta ct if you have questions about the study:
112 Brian Menaker, M.A. Ph D Student, Department of Tourism, Recreation, and Sport Management. 300 FLG Dr. Charles S. Williams, Ed.D., Department of Tourism, Recreation and Sport Management. Whom to contact about you r rights as a research participant in the study: IRB02 Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; phone 392 0433.
113 APPENDIX B SAMPLE RECRUITMENT L ETTER Dear Participant, I am a Docto ral student in the Department of Tourism, Recreation, and Sport Management at the University of Florida with a research interest in facility management. I am currently working on my dissertation which is looking at the role of athletic departments in the i mplementation of alcohol policy at college football stadiums on home game days. I obtained your contact information from the athletic department directory on your department's website and believe you could possibly be of assistance to me with your experien ce as Associate Director of Athletics / Facilities & Event Management. An integral part of my study includes discussions with higher level administration involved in facility management as to their role and attitudes toward alcohol policy at college footba ll stadiums. As a part of this, I would like to request your participation in a thirty minute face to face interview at a time and day of your convenience. I value your input and would like to include your comments as a part of this study. Your name will b e kept confidential in the study, when the results are written up. Please let me know if you would be willing to participate or if you are unable to participate, if anyone else familiar with football game day alcohol policy and implementation might be willing to be interviewed. Feel free to contact me if you have any questions regarding your participation in the study. If you agree to participate I will forward a copy of the informed consent document with additional information about the study to review. Please respond to this message to acknowledge receipt, regardless of your decision to participate. Thank you for your time, I look forward to meeting with you, and I hope you are able to be a part of my research.
114 Respectfully, Brian Menaker PhD. Candidate Dept. of Tourism, Recreation, and Sport Management University of Florida Gainesville, FL 32611
115 LIST OF REFERENCES Abar, C., Turrisi, R., & Abar, B. (2010). Brief report: Tailgating a s a unique context for Parental modeling on college student alcohol use, Journal of Adolescence xx, 1 4. Ajzen I. (1985). From intentions to actions: A theory of plann ed behavior. In J. Kuhl & J. Beckmann (Eds.), Action control: From cognition to behavior Berlin: Springer. 11 39. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179 211. Ajzen, I. (2006) Constructing a TpB Questionnaire: Conceptual and methodological considerations. Retrieved April 13. 2010 from http://www.people.umass.edu/aizen/pdf/tpb.measurement.pdf Akers, R.L. (1991). Self contro l as a general theory of crime. Journal of Quantitative Criminology, 7, 201 11. Akers, R.L. (2009). Social learning and social structure: a general theory of crime and deviance. Transaction Books. Akers, R. L. & Jensen, G. F. (2007). Empirical status of social learning theory of crime a nd deviance: the past, present, and future. Chapter in Social Learning Th eory and Crime: A Progress Report. Akers, R.L. and Sellers, C.S. (2009) Criminological theo ries: Introduction, evaluation, a nd application 5 th ed. New York: Oxford University Press. Bandura, A. (1969). Principles of behavior modification New York: Ho l t, Rhinehart, & Wilson. Bandura, A. (1973). Aggression: A social learning analysis Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1986). Social foundation of thought and action : A Social Cognitive Theory Englewood Cliffs, N.J.: Prentice Hall, 1986. Bandura, A. (1997). Self efficacy: The exercise of control New York: W .H. Freeman a nd Company. Bandura, A. (1998). Health promotion from the perspective of social cognitive theory. Psychology and Health 1998, 13 623 649. Bandura, A. (1999). Moral disengagement in th e perpetration of inhumanities. Personality and Psychology Review, 3 (3). 193 209.
116 Bandura, A ( 2002). A Social Cognitive Theory of Mass Communications. In J. Bryant and D. Zillman (Eds.), M edia e ffects : Advances in Theory and Research (2 nd Ed.) (pp. 121 153). Hillsdale, N. J.: Erlbaum. Bandura, A. (2004). Health promotion through social cognitive means. Health Education and Behavior 31 143 164. Berg, B. L. (2004). Qualitative research methods for social sciences. (5 th ed.). Boston, MA: Allyn and Bacon. Bernstein, K. T., Galea, S., Ahern, J., Tracy, M. & Vlahov, D. (2007) The built environment and alcohol consumption in urban neighborhoods. Drug and Alcohol Dependence, 91 244 252. Bormann, C. A., & Stone M. H. (2001). The effects of e liminating alcohol in a college stadium: The Folsom Field beer ban. Journal of American College Health, 50, 81 87. Boyes, W. J., & Faith, R. L. (1993). Temporal regulation and intertemporal substitution: The effect of banning alcohol at college football games. Public Choice, 77 585 609. Carey, K. B., Scott Sheldon, L. A. J, Carey, M. P., & DeMa rtini (2007). Individual level interventions to reduce college student drinking: A meta analytic r eview Addict ive Behav iors, 32 (11): 2469 2494. Carruthers, C. P. (1993). Leisure and alcohol expectancies. Journal of Leisure Research 25 229 244. Chaloupka, F. J., & Wechsler, H. (1996). Binge drinking in college: the impact of price, availability an d alcohol control policies, Contemporary Economic Policy XIV 112 124. Conner, M., Warren, R., Close, S., Sparks, P. (1999) Alcohol Consumption and the Theory of Planned Behavior: An Examination of the Cognitive Mediation of Past Behavior Journal of Applied Social Psychology, 29 (8), 1676 1704. DeJong, W., DeRicco, B. Schneider, S. K. (2010). Pregaming: An exploratory study of strategic drinking by college students in Pennsylvania. Journal of American College Health, 58 4, 307 316. DiIulio Jr., J. J (1995). Broken bottles: Liquor, disorder, and crime in Wisconsin. Wisconsin Policy Research Institute Report 5 (4), 1 27.
117 DeJong, W., Towvim, L. G. & Schneider, S. K. (2007). Support for alcohol control policies and enforcement strategies among U.S. col lege students at 4 Year institutions. Journal of American College Health, 56 (3), 231 236. Dimmock, J. A. & Grove, J. R. (2005). Relationship of fan identification to determinants of aggression. Journal of Applied Sport Psychology 17(1), 37 47. End, C. M., Davis, M. C., Kretschmar, J. M., Campbell, J., Mueller, D. G., Worthman, usage and alcohol related experiences. Sociological Spectrum, 269, 649 658. Erickson, D. J ., Toomey, T. L., Lenk, K. M., Kilian, G. R., Fabian, L. E. A (2011). Can we assess blood alcohol levels of attendees leavin g professional sporting events? Alcoholism:Clinical & Experimental Research, 35 (4), 689 694. Fried, G., & Ammon, R. (2009). Alcohol management: Boon or boondoggle? Journal of Crowd Safety and Security Management, 1 (2), 63 80. Fried, G., & Metchick, R. (2005). Camp Randall Memorial Stadium case study: October 30, 1993. Journal of Legal Aspects of Sport, 15 (1). 139 176. Glassman, T., Braun, R. E., Dodd, V. Miller, J. M., & Miller, E. M. (2010). Using the theory of planned behavior to explain the drinking motivations of social, high risk, and extreme drinkers on game day. Journal of Community Health 35 (1), 172 181. Glassman, T., Dodd V., Sheu, J.J., Rienzo, B.A., & Wagenaar, A.C. (2008). Using the theory of planned behavior to predict alcohol consumption among college students on game day. Journal of Global Drug Policy and Practice, 2 (3), Paper 4. Retrieved from http: // www.globaldrugpolicy.org/2/3/4.php Glassman, T., Werch C.H., Jobli, E., & Bian H, (2007) Alcohol related fan behavior on college football game day. Journal of American College of Health, 56 (3), 256 261. Greenfield, L. A. (1998). Alcohol and crime: An analysis of national data on the prevalence of alcohol involvement in crime. Annapolis Junction, MD: US Department of Justice. Gyimah Brempong, K. (2001). Alcohol availability and crime: Evidence from census tract data. Southern Economic Journal 68 (1), 2 21. Gyimah Brempong, K., & Racine, J. (2006). Alcohol availability and crime: a robust approach. Applied Economics 38 1293 1307.
118 Hair, J. F., Black, W. C., Babin, B. J., An derson, R. E. (2009). Multivariate data analysis ( 7 th edition) Upper Saddle River, NJ: Prentice Hall. Johannessen, K, Glider P., Collins, C., Hueston H, & DeJong W., (2001). Preventing alcohol environmental management case study. American Journal of Drug and Alcohol Abuse 27 (3), 587 597. Johnson, G. K. (2010). NCAA football attendance plateaus after record run. The NCAA News Retr ieved April 25, 2010 from http://www.ncaa.org/wps/por tal/ncaahome?WCM_GLOBAL_CONTEXT=/ncaa/nc aa/ncaa+news/ncaa+news+online/2010/association wide/ncaa+football+attendance+plateaus+after+record+run_02_22_10_ncaa_ne ws Kaplowitz, S.A. & Campo S. (2004). Drinking, alcohol policy, and attitudes toward a campus riot. Journal of College Student Development, 45 (5), 591 613 Lenk, K. M., Toomey, T. L., Erickson, D. J. (2009 ). Alcohol related problems and enforcement at profesional sports stadiums. Drugs: education, prevention, and policy, 16 (5), 451 462. Lenk K. M., Toomey, T. L., Erickson, D. J., Kilian, G. R., Nelson, T. F., & Fabian, L. E. (2010). Alcohol control policies and practices at professional sports stadiums. Public Health Report, 125 (5), 665 73. Madensen, T. D. & Eck, J. E. (2008) Spectator viol ence in stadiums. Problem oriented guides for police problem specific guides series 54 Department of Justice. McAlister, A. L., Perry, C. L., Parcel, G. S (2008). How individuals, environments, and health behaviors interact. In K. Glanz, B. K. Ritter, & K. Viswanath (Eds.). Health Behavior and Health Education: Theory, research, and practice. San Francisco: Jossey Bass. McAllister, I. Public attitudes to the regulation of alcohol. (1995). Drug and Alcohol Review 14 179 186. Menaker, B. E. & Connaugh ton, D. P. (2010). Stadium alcohol policies: A comparison of policies available on college athletic department web sites. International Journal of Sport Communication, 3 (2), 151 162. Merlo, L. J., Hong, J., & Cottler, L. (2009). The association between al cohol related arrests and college football game days. Drug and Alcohol Dependence, 106 (1),
119 Miller, J. & Gillentine, A. (2006). An analysis of risk management policies for tailgating activities at selected NCAA division I football games. Journal of Legal Aspects of Sport, 16, 197 215. Miller, N. E., & Dollard, J. (1941). Social learning and imitation. New Haven, Conn.: Yale University Press. Moustakas, C. E. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage. Neighbors, C., Oster Aaland, L., Bergstrom, R. L., & Lewis, M. A. Event specific and context specific normative misperceptions and high risk drinking: 21st birthday celebrations and football tailgating. Journal of Studies on Alcohol 67 (2), 282 289. Nelson T.F. & Wechsler H (2003). School spirits: Alcohol and collegiate sports fans. Addictive Behaviors, 28 (1), 1 11. 69 71. Opdyke, J. D. and Kesmodel, D. (2009, 12 September). Beer sales make a comeback at college stadiums: struggling cities and schools hope to cash in wh ile avoiding underage drinking and violence; No suds after halftime. The Wall Street Journal A5. Oster Aaland, L. K. & Neighbors, C. (2007). The impact of a tailgating policy on drinking behavior and perceptions. Journal of American College Health. 56 (3), 281 285. Rees, D. I., & Schnepel, K. T. (2009). College football and crime. Journal of Sport Economics, 10 (1), 68 87. Rotter, J. B. (1954). Social Learning and Clinical Psychology. Englewood Cliffs, N. J.: Prentice Hall. Smith, J. R., Te rry, D. J., Manstead, A. S. R., Louis, W. R., Kotterman, D. Wolfs, J. (2008). The Attitude Behavior Relationship in Consumer Conduct: The Role of Norms, Past Behavior, and Self Identity. The Journal of Social Psychology, 148 (3), 311 348. Southall, R. M. & Sharp, L. A. (2006). The National Football League and its "Culture of Intoxication:" A negligent marketing analysis of Verni v. Lanzaro. Journal of Legal Aspects of Sport 16 (1), 121 148. Spaite, D. W., Meislin, H., Valenzuela, T., Criss, E., Smith, R. & Nelson, A. (1990). Banning alcohol in a major college stadium: Impact on the incidence and patterns of injury and illness. Journal of American College Health 39 125 128.
120 StatSoft, Inc. (2011). Electronic Statistics Textbook. Tulsa, OK: StatSoft. Ret rieved from http://www.statsoft.c om/textbook/ Stevens, J. P. (2002). Applied multivariate statistics for the social science. Mahwah, New Jersey: Lawrence Erlbaum Associates. Toomey T. L., Lenk, K. M., & Wagenaar, A. C. (2007). Environmental Policies to Reduce College Drinking: An Update of Research Findings. Journ al of Studies of Alcohol and Drug, 68 (2), 208 219. Toomey, T. L., & Wagenaar, A. C. (1999). Policy options for prevention: The case of alcohol. Journal of Public Health Policy, 20 (2), 192 213. Wakefield K.L. and Wann D.L. (2006). An examination of dysfunctional sport fans: method of classification and relationships with problem behaviors. Journal of Leisure Research 38(2), 168 186. Wann, D. L (1998). A preliminary investigation of the relationships between alcohol use and sport fandom. Social Behavior and Personality 26, 287 290. Wieberg, S. (2005, 21 November). Colleges are reaching their limit on alcohol. USA Today online edition Retrieved August 5, 2010 from http://www.usatoday.com/news/nation/2005 11 16 colleges alcohol_x.htm Wolfe, J., Martinez, R., & Scott, W. A. (1998). Baseball and beer: An analysis of alcohol Consumption patterns among male spectators at major league sporting events. Wong, S. & Lau, E. (2001). Understanding the behavior of Hong Kong Chi nese tourists on group tour p ackages Journal of Travel Research, 40 (1), 57 67. Zhu, L Gorman, D. M., & Horel, S. (2004). Alcohol density and violence: A geospatial analysis. Alcohol & Alcoholism, 39 (4), 369 375.
121 BIOGRAPHICAL SKETCH Brian Menaker received his Ph.D in Health and Human Performance with a concentration in sport managem ent from the University of Florida in August of 2011. He studied in the Department of Tourism, Recreation, and Sport Management in the College of Health in Human Performance at the University of Florida from 2008 through 2011 He is originally f rom Pelham, NY. He received a BA in h istor y from Grinnell College in 2004 and a n MA in sport s tudies in 2007 from the University of Iowa.