<%BANNER%>

Drug and Alcohol Offenders in a College Town

Permanent Link: http://ufdc.ufl.edu/UFE0023815/00001

Material Information

Title: Drug and Alcohol Offenders in a College Town Exploring Multiple-Agency Official Data to Assess the Impact of Official Processing and Sanctioning on Future Delinquency and Academic Careers
Physical Description: 1 online resource (153 p.)
Language: english
Creator: Khey, David
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: alcohol, campus, crime, drug, university
Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: 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 DRUG AND ALCOHOL OFFENDERS IN A COLLEGE TOWN: EXPLORING MULTIPLE-AGENCY OFFICIAL DATA TO ASSESS THE IMPACT OF OFFICIAL PROCESSING AND SANCTIONING ON FUTURE DELINQUENCY AND ACADEMIC CAREERS By David Nicolaus Khey May 2009 Chair: Lonn Lanza-Kaduce Major: Criminology The purpose of this dissertation is to learn more about the role of official processing for drug- and alcohol- related offenses in impacting academic careers and future offending of university students. Of particular interest is the effect of sanctions on these outcomes. The study is conducted using two subpopulations entering the University of Florida in the summer or fall of 2004 as first-year students. The first subpopulation consists of those who were arrested in Alachua County, Florida between 2004 and December 2007 for drug and/or alcohol offenses (N = 292). The second subpopulation consists of those who were officially referred to the university for drug or alcohol violations during that time frame (N = 351). Various comparisons are made. The pattern of arrests for those in the first subpopulation is generally compared with the arrests of similarly aged persons in the county. Each of the two subpopulations of students is compared with students who have not been referred or arrested. The two subpopulations are also compared with each other to see if those referred differ from those arrested and to examine how university or criminal justice processing/sanctioning affect academic performance (e.g., grades, change in academic performance, completion, dropout) and subsequent trouble with the university and criminal justice authorities (e.g., additional referral or arrest). Several risk factors are examined in order unveil the importance of each in predicting negative changes in academic achievement and participation in future offending. Overall, this study revealed that in conjunction with lower levels of sanctioning, students processed solely by the university system for drug and alcohol offenses express an increase in odds of completing their undergraduate degrees on-time. Yet the same process regardless of the level of sanctioning increased levels of recidivism. This study also revealed that criminal justice processing and sanctioning may offer advantages over the university sanctioning system in improving academic performance after a student s first officially recognized offense. The results of this analysis suggest that university administrators should take a closer look at how alcohol and drug offenders are processed by their internal judicial affairs system and by the criminal justice system.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by David Khey.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Lanza-Kaduce, Lonn M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0023815:00001

Permanent Link: http://ufdc.ufl.edu/UFE0023815/00001

Material Information

Title: Drug and Alcohol Offenders in a College Town Exploring Multiple-Agency Official Data to Assess the Impact of Official Processing and Sanctioning on Future Delinquency and Academic Careers
Physical Description: 1 online resource (153 p.)
Language: english
Creator: Khey, David
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: alcohol, campus, crime, drug, university
Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: 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 DRUG AND ALCOHOL OFFENDERS IN A COLLEGE TOWN: EXPLORING MULTIPLE-AGENCY OFFICIAL DATA TO ASSESS THE IMPACT OF OFFICIAL PROCESSING AND SANCTIONING ON FUTURE DELINQUENCY AND ACADEMIC CAREERS By David Nicolaus Khey May 2009 Chair: Lonn Lanza-Kaduce Major: Criminology The purpose of this dissertation is to learn more about the role of official processing for drug- and alcohol- related offenses in impacting academic careers and future offending of university students. Of particular interest is the effect of sanctions on these outcomes. The study is conducted using two subpopulations entering the University of Florida in the summer or fall of 2004 as first-year students. The first subpopulation consists of those who were arrested in Alachua County, Florida between 2004 and December 2007 for drug and/or alcohol offenses (N = 292). The second subpopulation consists of those who were officially referred to the university for drug or alcohol violations during that time frame (N = 351). Various comparisons are made. The pattern of arrests for those in the first subpopulation is generally compared with the arrests of similarly aged persons in the county. Each of the two subpopulations of students is compared with students who have not been referred or arrested. The two subpopulations are also compared with each other to see if those referred differ from those arrested and to examine how university or criminal justice processing/sanctioning affect academic performance (e.g., grades, change in academic performance, completion, dropout) and subsequent trouble with the university and criminal justice authorities (e.g., additional referral or arrest). Several risk factors are examined in order unveil the importance of each in predicting negative changes in academic achievement and participation in future offending. Overall, this study revealed that in conjunction with lower levels of sanctioning, students processed solely by the university system for drug and alcohol offenses express an increase in odds of completing their undergraduate degrees on-time. Yet the same process regardless of the level of sanctioning increased levels of recidivism. This study also revealed that criminal justice processing and sanctioning may offer advantages over the university sanctioning system in improving academic performance after a student s first officially recognized offense. The results of this analysis suggest that university administrators should take a closer look at how alcohol and drug offenders are processed by their internal judicial affairs system and by the criminal justice system.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by David Khey.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Lanza-Kaduce, Lonn M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0023815:00001


This item has the following downloads:


Full Text

PAGE 1

DRUG AND ALCOHOL OFFE NDERS IN A COLLEGE TO WN: EXPLORING MULTIPLEAGENCY OFFICIAL DATA TO ASSESS THE IMPACT OF OFFICIAL PROCESSING AND SANCTIONING ON FUTURE DELI NQUENCY AND ACADEMIC CAREERS By DAVID NICOLAUS KHEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 David Nicolaus Khey 2

PAGE 3

To all the folks that kept me strong in my y ears of study, this dissertation is dedicated to you. 3

PAGE 4

ACKNOWLEDGMENTS First and foremost, I would like to thank my dissertation chair and faculty advisor, Lonn Lanza-Kaduce. I owe him a great debt of grat itude for his leadership and scholastic counsel throughout my graduate years. I would like to thank Alisha Tabag and Daintry Cleary from Student Legal Services for their patience through a three-year process of discovering the data found in this dissertation and to Kate Fox fo r her aid in compiling and sorting the mass of information we collected. Lastly, a special thanks to my dissertation committee Joseph Spillane, Chris Gibson, Chuck Frazier, and Bruce Gol dberger for their support in my final stages of my graduate career. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................7LIST OF FIGURES .........................................................................................................................9ABSTRACT ...................................................................................................................... .............10CHAPTER 1 THE IMPACT OF ARREST AND S ANCTIONS ON COLLEGE STUDENTS .................12Introduction .................................................................................................................. ...........12Drug and Alcohol Use by the Youth of America: A Historical Brief ....................................15Present Study ..........................................................................................................................17The Impact of Arrest and Court Sanctions on High School Students .............................18Dissertation Roadmap .....................................................................................................192 REVIEW OF THE LITERATURE AND THEORETICAL PERSPECTIVES .....................26Drug and Alcohol Use and our College Students ...................................................................26Substance Use over the Life-Course .......................................................................................31Reintegrative Shaming and Labeling ......................................................................................35Shaming in a Culture of Ambivalence ....................................................................................39One Journalists Perspective at the Study Institution .............................................................44Summary ....................................................................................................................... ..........463 PRELIMINARY ANALYSIS: STUDENTS A NON-STUDENTS IN A COLLEGE TOWN .......................................................................................................................... ..........49Introduction .................................................................................................................. ...........49Study Population .....................................................................................................................49Descriptive Analysis: Comparing Student versus Non-Student Populations .........................50Presentation and Discussion of Findings .........................................................................53Possible Explanations of Discrepant Findings ................................................................56A Missing Link: Jointly Controlli ng for Age, Gender, and Race ...................................59Summary ....................................................................................................................... ..........614 ASSESSING SIMILARITIES AND DIFFERENCES AMONG OFFENDING STUDENTS ARRESTED BY LOCAL AUTHORITIES OR REFERRED TO JUDICIAL AFFAIRS .............................................................................................................6 8Introduction .................................................................................................................. ...........68The Data ..................................................................................................................................70 5

PAGE 6

A Profile of Freshman, 2004 ..................................................................................................72Summary ....................................................................................................................... ..........895 MULTIVARIATE ANALYSIS OF A RRESTEE AND REFERRAL GROUPS: A FOCUS ON SANCTIONS .....................................................................................................91Research Questions and Strategies .........................................................................................91The Impact of Different Sanctions ..................................................................................91Dependent Variables .......................................................................................................92Independent and Control Variables .................................................................................93Introduction to Multivariate Analysis .....................................................................................94Predicting On-Time Completion ............................................................................................95Predicting Dropout or Dismissal ..........................................................................................100Predicting Recidivism ......................................................................................................... ..101Predicting Overall Mean GPA Differences and Change in Mean GPA Differences Before and After First Intervention ...................................................................................108Summary ....................................................................................................................... ........1116 DISCUSSION OF FINDINGS, POLICY IMPLICATIONS, AND CONCLUSIONS ........114Introduction .................................................................................................................. .........114Disciplinary Systems Imbedded in a Culture of Ambivalence .............................................118Labeling, Deterrence, and M odeling Success and Failure ...................................................115Policy Implications ...............................................................................................................122Factors Tied to Drugand Alcohol Related Offending and Re-offending ..........................124Limitations ................................................................................................................... .........125Conclusions ...........................................................................................................................129 APPENDIX A DESCRIBING A CULTURE OF SUBSTANCE USE ........................................................130B PERCEPTION OF STUDENTS ...........................................................................................134C MULTIVARIATE TABLES IN CLUDING STANDAR D ERRORS .................................136LIST OF REFERENCES .............................................................................................................143BIOGRAPHICAL SKETCH .......................................................................................................152 6

PAGE 7

LIST OF TABLES Table page 1-1 Selected drug law and reform since 1970 ..........................................................................213-1 List of Offenses Included in Alcohol, Drug, Violent, and Non-Violent Categories .........633-2 Arrests in Alachua County, Florida by Agency and Type .................................................653-3 Arrests and Arrest Rates per 1,000 In dividuals Aged 18-24 for All Reporting Agencies in Alachua County, Florid a in 2003-2008 by Gender and Race. .......................663-4 Arrests and Arrest Rates per 1,000 In dividuals Aged 18-24 for All Reporting Agencies in Alachua County, Florida in 2003-2007 by Gender and Race Jointly. ...........674-1 Variables and Sources of Data ...........................................................................................704-2 Overall 2004 Freshman Profile in Rela tion to Those of the Arrestee and Referral Subpopulations ................................................................................................................ ...764-3 Changes in the Difference in Individual Student GPA Relative to their Resident College Before and After First Arrest or Referral .............................................................875-1 Logistic Regression Estimates of the Effect of Official Processing on On-Time Graduation..........................................................................................................................965-2 Logistic Regression Estimates of the E ffect of Official In tervention on On-Time Graduation (N = 570) .........................................................................................................9 95-3 Logistic Regression Estimates of the Effect of Official In tervention on Dropout or Dismissal ..................................................................................................................... .....1015-4 Ordinary Least Squares Estimates of the E ffect of Official Pro cessing on Recidivism ..1035-5 Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism .................................................................................................................... ...1045-6 Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism for Arrestees While F actoring in Juvenile Arrest .........................................1075-7 Ordinary Least Squares Estimates of the Effect of Official Processing and Intervention on Overall GPA Difference (Models 1 & 2) and Change in GPA Difference Before and After Firs t Intervention (Models 3 & 4) ......................................109C-1 Logistic Regression Estimates of the Effect of Official Processing on On-Time Graduation........................................................................................................................136 7

PAGE 8

C-2 Logistic Regression Estimates of the E ffect of Official In tervention on On-Time Graduation (N = 570) .......................................................................................................137C-3 Logistic Regression Estimates of the E ffect of Official In tervention on Dropout or Dismissal ..................................................................................................................... .....138C-4 Ordinary Least Squares Estimates of the E ffect of Official Pro cessing on Recidivism ..139C-5 Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism .................................................................................................................... ...140C-6 Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism for Arrestees While F actoring in Juvenile Arrest .........................................141C-7 Ordinary Least Squares Estimates of the Effect of Official Processing and Intervention on Overall GPA Difference (Models 1 & 2) and Change in GPA Difference Before and After Fi rst Intervention (Models 3 & 4) .....................................142 8

PAGE 9

LIST OF FIGURES Figure page 2-1 Monitoring the Future: Percentage of Co llege Students and Twelfth-Graders that Used Illicit Drugs, Alcohol, and Been Drunk Within the Past 30 Days ............................272-2 Monitoring the Future: Percentage of Colle ge Students versus Others One to Four Years Beyond High School that Drank Al cohol in the Past 30 Days ................................282-3 Monitoring the Future: Percentage of Colle ge Students versus Others One to Four Years Beyond High School that Drank Five or More Drinks in a Row within the Last Two Weeks ..................................................................................................................... ...292-4 Bernburg and Krohns Hypothe sized Effects of Official Intervention in Adolescence on Crime in Early Adulthood .............................................................................................362-5 University of Colorado at Boulder 420 at Farrand Field Police Surveillance ................413-1 Alcohol-Related Offenses in Alachua County, 2003-2007 ...............................................633-2 Alcohol-Related Offenses in Alachua County, Controlling for Football Home Games, 2003-2007 .............................................................................................................634-1 Visual Representation of Study Population as Embedded in their Wider Cohort. ............73 4-2 Comparison of Average GPA Differences. (Comparison sample is on the left; Study sample on the right) .......................................................................................................... .794-3 Average GPA Differences of the Comparison Sample with Dropouts and Dismissals Removed. ...................................................................................................................... .....804-4 Depiction of Timing of First Offense ................................................................................81 B-1 Summary of Students A ttitudes and Perceptions of Drug and Alcohol Use at the Study University. .............................................................................................................135 9

PAGE 10

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DRUG AND ALCOHOL OFFE NDERS IN A COLLEGE TO WN: EXPLORING MULTIPLEAGENCY OFFICIAL DATA TO ASSESS THE IMPACT OF OFFICIAL PROCESSING AND SANCTIONING ON FUTURE DELI NQUENCY AND ACADEMIC CAREERS By David Nicolaus Khey May 2009 Chair: Lonn Lanza-Kaduce Major: Criminology, Law, and Society The purpose of this dissertation is to learn mo re about the role of official processing for drugand alcoholrelated offenses in imp acting academic careers and future offending of university students. Of particular interest is th e effect of sanctions on these outcomes. The study is conducted using two subpopulations entering the Univ ersity of Florida in the summer or fall of 2004 as first-year students. The first subpopulation consists of those who were arrested in Alachua County, Florida between 2004 and Decembe r 2007 for drug and/or alcohol offenses (N = 292). The second subpopulation consists of those who were officially re ferred to the university for drug or alcohol violations dur ing that time frame (N = 351). Various comparisons are made. The pattern of arrests for those in the first subpop ulation is generally compared with the arrests of similarly aged persons in the county. Each of the two subpopulations of students is compared with students who have not been referred or arre sted. The two subpopulations are also compared with each other to see if those referred differ from those arrested and to examine how university or criminal justice processing/sanctioning affect academic performance (e.g., grades, change in academic performance, completion, dropout) and subsequent trouble with the university and criminal justice authorities (e.g., additional referral or arrest). Several ri sk factors are examined 10

PAGE 11

11 in order unveil the importance of each in predicting ne gative changes in academic achievement and participation in future offending. Overall, th is study revealed that in conjunction with lower levels of sanctioning, students processed solely by the university system for drug and alcohol offenses express an increase in odds of completing their undergra duate degrees on-time. Yet the same process regardless of the le vel of sanctioning increased leve ls of recidivism. This study also revealed that criminal justice processi ng and sanctioning may offer advantages over the university sanctioning system in improving acad emic performance after a students first officially recognized offense. The results of this analysis suggest that university administrators should take a closer look at how alcohol and drug offenders are processed by their internal judicial affairs system and by the criminal justice system.

PAGE 12

CHAPTER 1 THE IMPACT OF ARREST AND S ANCTIONS ON COLLEGE STUDENTS Introduction Several criminological investigations have elaborated on the impact of arrest and of sanctions on outcomes in later life (Esben sen, Thornberry, & Huizi nga, 1991; Huizinga & Esbensen, 1992; Huizinga, Esbensen, & We iher, 1996; Klein, 1986; Sherman, Gottfredson, MacKenzie, Eck, Reuter, & Bushway, 1997). One im portant effort derived its data from the Denver Youth Survey led by David Huizinga at the University of Colorado. Beginning in 1986, the Denver Youth Survey researchers sampled households in high-risk areas in and around the Denver area as a part of a larger study spons ored by the Office of Juvenile Justice and Delinquency Prevention. Overall, 806 boys and 721 girls responded to the researchers requests, many continued to report annually from 1988 to 1992 and again from 1995 to 1999. In several offshoots from this major project, Huizinga and his collea gues have consistently determined that arrest has little impact on future delinquency (Esbensen, Thornberry, & Huizinga, 1991; Huizinga & Esbensen, 1992; Huizinga, Esbensen, & Weiher, 1996). Additionally, these researchers sugge st that when arrest does have an impact, it seems to act as an aggravating factor in that it may increase subsequent offending and delinquency. This is in accord with research projects external to the De nver Youth Survey as well. In a classic study by Malcolm Klein on labeling theory, this amplifying effect was determined to be linked to the official delinquent label being imposed on youths, an increase in the amount of formal processing by the criminal justice system, and a subsequent self-fulfilling prophecy being played out to conform to the label of delinquent (Klein, 1986). Perhaps more compelling were the findings of Lawrence Sherman and his colleagues in a report given to the United States Congress in 1997 (Sherman, Gottfredson, MacKenzie, Eck, 12

PAGE 13

Reuter, & Bushway, 1997). Harnessing the collec tive intelligence of so cial scientists and examining the extant research available at the time, Sherman and his colleagues delivered the same message: overall, many of our reactions to crime overwhelmingl y have little to no effect on subsequent offending or may actua lly increase offending. Here are a few examples: increased arrests for domestic violence was found to increase the incidence of similar offenses in areas with relatively high unemployment rates and low marriage rates (see Marciniak, 1994); despite the dedicated work of the Chicago Youth Developmen t Project, arrest rates of the juveniles in the experimental program increased with time (s ee Gold & Mattick, 1974); many youths that have been arrested tend to offend more after arrest than delinquent youths not arrested (Farrington, 1977; Gold & Williams, 1970). In the section of the report Sherman authored, he remarks: the effects of police on crime are complex, a nd often surprising (Sherman et al., 1997: 839). This runs counterintuitiv e to the tenets of specific deterrence and common thinking: punishment for an individual criminal act(s) will deter them from future offending. This intuition and cornerstone of many criminal ju stice systems across the globe is not lacking convincing empirical support, either. For exam ple, Smith and Gartin (1989) find evidence supporting this perspective rather th an an arrest event serving as an aggravating factor for future offending in a sample of males born in 1949 followed through age twenty five. By and large, there does not seem to be a clear answer as to when and in what context arrest and punishment can have a therapeutic or toxic impact on offenders. The multidimensionality and complexity of this phenomenon is self-apparent when review ing the literature on these two broad, yet conflicting, areas of rese arch in criminology. To date, research has not yet be en extrapolated to college students. More specifically, little is known about the effects of arrest and punishment on future offending or on students academic 13

PAGE 14

performance. While a wealth of criminological res earch guides this dissertation and its hypotheses regarding the reactions to arrest and punishment and its effects on future offending (or lack thereof), this project ex tends this previous research into a population that is yet to be studied in this light. Does stude nt offending shift if handled through the criminal justice system? Likewise, are there different rates of offending when students are disciplined by means outside of the criminal justice system such as a univers itys judicial affairs sy stem? Since alcohol and drug offenses are consistently the most promin ent on college campuses across the nation (U.S Department of Education, 2008), this dissertation concentrates on th ese violations in effort to begin to resolve some of these questions. Narrowing the scope to include drug and alcohol offenses also serves to directly address trends that are encroaching on our college students welfare in the United States. College campuses are a notable microcosm of a culture that vastly accepts alcohol consumption, and to a lesser extent, drug use, and stude nts have reported vast uninte nded consequences as a result including having a hangover, poor performan ce on exams and projects, damaging property, involvement in arguments or physical confrontations, becoming nauseous and/or vomiting, missing class, and contact with law enforcement or other authorities which may lead to arrest (Core Institute, 2008; Dowdall, 2006). The aggr egate problems of this population have been worsening in recent years and the emerging pa tterns of alcohol and illicit substances consumption have alarmed administrators, rese archers, and the public (Wechsler & Wuethrich, 2002). In terms of alcohol consumption, a nation ally representative st udy of college students suggests the percentage of this population that binge drink as defined by having five or more alcoholic drinks in one occasion for men or four or more alcoholic drinks in one occasion for women at least once in the previous two weeks has significantly increased from 1993 to 2001 14

PAGE 15

(Weschler, Kuo, Nelson, & Lee, 2002). According to different indicators of illicit drug use among this population (e.g., the National Survey of Drug Use and Health, Monitoring the Future, the Core Institute, et cetera elaborated in Chapter 3), modest increases in drug use have also been seen over the same time frame. Paradoxi cally, these increases al so run parallel with increases of potential punitive responses that may result from arrest of a drug and/or alcohol offense for college students. When considering the potential physical, mental, and social harms that substance use (immoderate use particularly) may have on deve loping teenagers or ine xperienced young adults (see Kinney, 2006), the involvement of law enfor cement and the courts can be viewed from a parens patriae / loco parentis perspective. Several things remain unclear. Do punitive responses make positive impacts on youths lives within the context of American youth culture? Are punitive actions in the best interest of the arrestee? If the goal is to act within the best interest of the student, the overall outcome should be positive and reintegrative rather than disruptive to a students academic career. According to crimin ological literature on the life course, significant life events can have an influential impact on ones life trajectory (Laub & Sampson, 2003; Sampson & Laub, 2005), and events such as an arrest and criminal justice processing are likely to have long-lasting impacts on one s life. This dissertation inve stigates these impacts for drugand alcohol-related offenses on student acad emic performance and future offending. Drug and Alcohol Use by the Youth of America: A Historical Brief As far back as public health officials and gove rnmental agencies have been collecting data on substance use and abuse, the scale of use by the youth population has ebbed and flowed. The current cycle seems to span back to the 1970s an d continues through current times. For the most part, the relative amount of curre nt use by the youthful population is drastically less than the latter half of the 1970s this point in time dema rcates the highest contemporary rates of use and 15

PAGE 16

abuse in America (Johnston, O'Malley, Bachman, & Schulenberg, 2007). As use continued to decline into the 1980s through the early 1990s, ma jor shifts in the c ountrys drug policy took effect doling out more punitive and aggressive re sponses to trafficking, use, and possession of various substances, new approaches to drug educa tion (just say no), and new controls for use in the workplace. This so-called war on drugs was catalyzed by the Reagan administration with continuity in the George H.W. Bush and C linton administrations. Accordingly, the 1980-2000 timeframe has seen much notable landmark legisl ation in the realm of drug law at local, state, and federal levels (see Table 1-1 at the end of this chapter). These policies have vastly impacted the youth of America and college students we re not immune from the increasing punitive legislation as the drug war years progressed. For example, an amendment to the Higher Education Act (passed during the Clinton admini stration in 1998, went in to effect in 2000) restricts student loans, grants, and work st udy positions from college students with drug possession or sale convictions unle ss certain criteria are met. Th is policy was further refined in 2006 to only include drug convictions during the periods in which students are receiving Title IV (federal) financial aid. While the position of the George W. Bush administration aligns with the previous three regimes, the concentration on te rrorism as a platform has over shadowed and usurped the war on drugs effort.1 Remarkably since the early quarter of the 1990s, levels of use and abuse have been generally shifting back to hi gher prevalence across a spectrum of psychoactive substances (Johnston, O'Malley, Bachman, & Schulenberg, 2007) These increases occur in spite of increased sanctions and anti-dr ug efforts, economic booms and federal budget surpluses, and the largest amount of incarcerated drug offenders in contemporary history. If the country has 1 Note that the softening of the drug conviction restricti ons of the Higher Education Act occurred during the years of the George W. Bush administration. 16

PAGE 17

prospered, has identified chronic drug offenders and took them off the streets, and has budgeted a multi-faceted war on drugs program at several in stitutional levels, then why is substance use and abuse flourishing in current ti mes? While this is a tangential question to the current work, it leads us to ask whether our reactio ns to drug and alcohol offenses ar e effective or not, or are they perhaps making matters worse. Present Study This dissertation derives its da ta from a collaborative effort between the Department of Criminology, Law, and Society (currently the Di vision of Criminology & Law, a part of the Department of Sociology and Criminology & Law) a nd Student Legal Services at the University of Florida, dubbed the UF Student Crime Project The resulting datase t, the University of Florida Comprehensive Student Crime Study, 2003 2007 (Khey, Fox, & Lanza-Kaduce, 2008), combines official arrest data fr om local law enforcement agencies,2 student data derived from the Office of the Registrar, information about soci al greek affiliation from the Office of Sorority and Fraternity Affairs, and adj udication data from the Dean of Students Office and the State of Florida District Eight Circuit Court. This current project seeks to extend the existing literature in three ways. Volumes of selfreport data on the differences of the college st udent population re lative to the same age group not attending an institution of highe r education on the leve ls of substance use and abuse exist (see Kinney, 2006), but it is rare to find official data on college students The first task of the present study is to formulate a thorough description and assessment of the patterns of student alcoholand drug-related arrest versus that of the wider community (as a referenc e group) over the age of 2 At the time of this writing, the following agencies released data to the UF Student Crime Project: Gainesville Police Department, University of Florida Police Department, Alachua County Sheriffs Office, and the Florida Division of Alcoholic Beverages and Tobacco. Pending requests exist for Florida Highway Patrol (FHP), the Drug Enforcement Administration (DEA), and Santa Fe Community College Police (SFCCP). In verbal communication with the latter agencies officials, very few University of Florida students are included in their annual arrest data. 17

PAGE 18

18. This will yield information about the diffe rences between these gr oups on a basic level by utilizing official data and will serve to fill this gap in current literature. Second, many researchers have argued that the environment of hi gher education contains certain risk factors for deviance and criminality, such as belonging to social greek orga nizations (McCabe, Schulenberg, Johnston, O'Malley, Bachman, & Kl oska, 2005) and involvement w ith athletics (Kinney, 2006). Accordingly, the present study will isolate the risk factors of arrest associated with these crimes for university students with the added ability to ob serve participation in other official criminal activity across students academic career. Las tly, and most importantly for the present study, this research will estimate the effect of arrest and of punishment by the criminal justice system and/or by university sanctions on student academic performance and re-arrest to extend the work of the Denver Youth Survey. This will assist policy-makers in better understanding the impact of official sanctions on future o ffending and educational attainment. The Impact of Arrest and Court Sanctions on High School Students Two notable studies offer guidance to this disse rtation by observing the e ffect of arrest and sanctions on a different subset of students high school students. First, Bernburg and Krohn (2003) examine labeling theory as a developmenta l theory of structural disadvantage by noting that arrest and sanctions have a tendency to block access to conventional individuals and opportunities. These disadvantages, in turn, accumula te over time to result in a larger and more difficult barrier to conventional life. Be rnberg and Krohn concentrate on conventional opportunities, in the form of educational attainme nt and employment, as mediating events which can predict further participation in crime in early adulthood. After analyses, they do in fact find that arrest and court sanctions increase the probability of future criminal activity as partly mediated by lower educational attainment and employment rates among the disadvantaged youth in their sample. Second, Sweeten (2006) refi nes the work of Bernburg and Krohn by broadening 18

PAGE 19

to a nationally representative sample and incl uding additional control variables to address selection bias issues. In his analysis, Sweeten replicates th e work of Bernburg and Krohn and concludes that court appearance increases the chances of high school dropout, regardless of the actual levels of delinquency re ported by each respondent. In bot h circumstances, arrest and official sanctions have a detrimental impact on 1) future offendi ng, and 2) educational attainment. The present study will return back to these two fo undational works to see if these effects remain consistent for college students. Dissertation Roadmap Chapter 2 will proceed to review extant research in substance use and abuse will be proffered to obtain a snapshot of the current issues facing college administrators, parents, and the students themselves and to explore the current expl anations for differences that exist between the college population and those not at institutions of higher education. Efforts will be concentrated on examining substance use and abuse over an impor tant phase of the life-course the transition and socialization into the college years to gain a better understanding on how the etiologies of criminal activity and substance use and abuse ru n roughly parallel. With this literature as a foundation, Chapter 2 continues by e xploring the theoretical relevanc e of two perspectives to the study of the impact of sanctions. The labe ling perspective, incl uding the revitalized specifications its renaissance in r ecent years, will be juxtaposed with the deterrence perspective. This chapter will conclude with an expanded discussion of the impacts of criminal justice sanctions on offenders, with a narrowed focus on Braithwaites process of reintegrative shaming. Chapter 3 presents an analysis of arrested stude nts in context of a wider population of arrestees in a college town. This analysis guides the fo llowing two chapters that seek to analyze the student population of arrestees and contrast th is group with students w ho offend but were never arrested or punished by the criminal justice syst em. This comparison will assist in determining 19

PAGE 20

20 whether arrest and sanctions for drug and alcohol offenses make a difference in students lives. Policy implications of the findings are offered in Chapter 5 to assist decision-makers in their efforts to ameliorate problems among university-student populations This chapter also gives overall summary of the findings, suggests future research endeavors, an d concludes the current work.

PAGE 21

Table 1-1: Selected drug law and reform since 1970 Year Title Summary 1965 Drug Abuse Control Amendments Define and control depressant and stimulant drugs. This included barbiturates, amphetamines, and the ability to add others after investigation. LSD was one of these extraneous drug s. Also restricted manufacture and made possession of these drugs illegal, and defined penalties for law-breakers. 1970 Comprehensive Drug Abuse Prevention and Control Act Landmark legislation that has governed drug control in contemporary times. Specifically, Title II of this Act, the Cont rolled Substances Act, spells out the current control regime used by the United States federal government and serves as the model for state legislation. Sets forward the scheduling of all controlled substances from Schedule I meaning illicit status except under tightly controlled medical studies to Schedule V making a drug available by prescription from a doctor with few restrictions. Organized Crime Control Act (RICO and CCE) Two powerful prosecutorial tools that define the coordinated acts of individuals who may be considered drug traffickers or organized crime syndicate members. This allows their co llective actions to be scrutinized in court hearings and increases the penalties for participation in certain crimes. 1972 Drug Abuse Office and Treatment Act Established the Special Action Offi ce for Drug Abuse Prevention (SAODAP) within the executive branch and the Na tional Institute of Drug Abuse (NIDA) as a subsidiary of the Na tional Institute of Mental Health (NIMH). SAODAP was to coordinate federal efforts rega rding prevention, educ ation, treatment, rehabilitation, training, a nd research of drug abuse. NIDA was charged with (re-)creating a national community-based treatment system. Research funding was also allocated for such development. This Act also legitimized maintenance for treatment of narcotic dependence under strict conditions and required any hospital that received fe deral funding not to refuse treating patients conditions direc tly or indirectly due to drug abuse or dependence. 21

PAGE 22

Table 1-1: Continued. Alcohol, Drug Abuse, and Mental Health Administration The Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA) was charged to reside over the developm ent of research in a wide area of mental health and substance use areas. Drug Enforcement Administration By executive order of Pres ident Nixon, the Bureau of Narcotics and Dangerous Drugs (BNDD), the Office for Drug Abuse Law Enforcement, and the Office of National Narcotics Intelligence were disbanded and the Drug Enforcement Administration was formed to merge th ese agencies under th e Department of Justice. Essentially, the latter Offices were merged into the BNDD to form the new agency. 1974/1978 Drug Abuse Treatment and Control Amendments Expands the 1972 Act (funding in particular). 1978 Alcohol and Drug Abuse Education Amendments Expands the Department of Educations role in drug education efforts. 1980 Drug Abuse Prevention, Treatment, and Rehabilitation Amendments Additional expansion of drug education and treatment programs (and funding). 1984 Uniform Drinking Age Act Encouraged states to se t the drinking age to 21, wh ich was now required to receive federal funding for highways and roads. Drug Offenders Act Expanded programmi ng and treatment for drug offenders. Comprehensive Crime Control Act A major overhaul of bail and asset forfeitu re procedures at the federal level. Most importantly, this Act abolished pa role at the federa l level and created stiff, often mandatory, sentences for drug offenders which on average lengthened the sentence for an indi vidual convicted of a drug crime. 22

PAGE 23

Table 1-1: Continued. 1986 Analogue Act Brings uncontrolled synthetic substan ces that pharmacologically mirror the effects of a particular controlled substa nce under regulation. This is to prevent illicit chemists from creating psychoactiv e substances that can be sold as a drug of abuse but technically are not i llegal because of chemical structural changes. Anti-Drug Abuse Act (1986) Addressed the foreign drug tr afficking threat by creati ng avenues to encourage partner governments to erad icate illicit crops and s ubstance production. This Act also sought to encourage stronger federal coordination of prevention and education programs, increased block gr ant funding for NIDA research with particular focus on Acquired Immuno -Deficiency Syndrome (AIDS), and imposed mandatory minimum sentences (a t the federal level) for heroin and cocaine. Created a powerful law enforcement tool in the form of asset forfeiture and generally granted law enfo rcement, prosecutors, and the courts increased power to process drug offenders. 1988 Anti-Drug Abuse Act (1988) Extended the Anti-Drug A buse Act of 1986 by encouraging private employers to maintain a drug-free workplace. Workplace drug testing to accept and maintain employment would become a popular choice for many national chains across the nation. Also establ ishes the Office of National Drug Control Policy (ONDCP and the so-called Amer ican drug czar) und er the executive branch. Crime Control Act In regards to drug regulation and enforcement, this act created drug-free school zones, aided rural drug enforcement effo rts, increased efforts at controlling money laundering, refined asset forfeitu re, defined chemical diversion and trafficking, and generally increased drug enforcement funding. This Act also began efforts in increasing sanctions for methamphetamine. 23

PAGE 24

Table 1-1: Continued. Crime Control Act In regards to drug regulation and enforcement, this act created drug-free school zones, aided rural drug enforcement effo rts, increased efforts at controlling money laundering, refined asset forfeitu re, defined chemical diversion and trafficking, and generally increased drug enforcement funding. This Act also began efforts in increasing sanctions for methamphetamine. 1992 ADAMHA Reorganization Organizes the Alcohol, Drug Abuse, and Mental Hea lth Administration under the Department of Health and Human Services (HHS) and integrates the National Institute of Alcohol Abuse (N IAA), the National Institute of Drug Abuse (NIDA), and the National Institute of Mental Health (NIMH) under the National Institutes of Health (NIH). The revised agency name becomes the Substance Abuse and Mental Health Services Administration (SAMHSA). Increased funding efforts directed toward s drug abuse research and treatment, except for needle exchange programs. 1998 Higher Education Act (Amendment) Disqualifies college students from Title IV (federal) fina ncial aid which includes Stafford, Perkins, and PLUS loans, Pell grants, and work study programs if one is convicted of drug possession or sale. Th is restriction may be lifted if the student completes an acceptable drug rehabilitation program. An acceptable program is one that includes at least two unannounced drug tests and is either qualified to receive funds from federal, state, or local governments or from a federalor state-licensed insuranc e company or is administered or accepted by a federal, stat e, or local agency or court, or a federalor state-licensed hospital, clinic, or doctor. This Act was later amended to include convictions during periods of which Title IV funding was being received by the student. 24

PAGE 25

25 Table 1-1: Continued. 2005 Combat Methamphetamine Epidemic Act* Restricts the sale of precursor subs tances previously found over-the-counter into a class onto its own. While not scheduled substances, ephedrine and pseudoephedrine were required to be be hind a merchants c ounter and out of direct reach of customers. Merchants would have to monitor sales to help prevent these substances from being converted to illicit methamphetamine products. Part of the Patriot Act

PAGE 26

CHAPTER 2 REVIEW OF THE LITERATURE AND THEORETICAL PERSPECTIVES Drug and Alcohol Use and our College Students In a comprehensive history of alcohol in America, historians Mark Lender and James Martin (1982) noted that Amer icans were ambivalent about al cohol and how to handle alcohol problems. In the microcosm of the college cam pus, conflicts can readily be seen among the administration, students, and alumni over the best ways to handle alcohol-related problems. As Lender and Martin remark: These controversies reflect, at least to some ex tent, the lack of cons istent public attitudes toward drinking. American ambivalence toward the subject is undeniable. Perhaps 30 percent of the nations citizens do not drink at al l, yet others tolerate considerable latitude in drinking behavior. Getting loaded at a college fraterni ty party, for example, more often than not is simply shrugged off as an instance of adolescent conduct. Hard-drinking, even to the point of drunkenness, is still accepted as the sign of being a real man in some social circles. In other cases, drinkers who see nothing wrong with their own imbibing have doubts about it in othersand although the nation spends huge sums annually in consumer purchases of beverage alcohol, and generally accepts the idea that alcoholism is a disease, Americans still place a severe soci al stigma on alcoholism. Thus, there is no clear sense of what the public really thinks or wants in regard to alcoholism prevention and control (1982: 191). Although this was written in the early 1980s, similar attitudes resound in current times on college campuses for both alcohol and soft drugs, part icularly marijuana. Of the large literature that exists to help assess the reality and myth s of substance use and abuse on college campus, three sources are especially help ful in monitoring the pulse of use and problem use for college students: (1) Monitoring the Future (National Institute of Drug Abuse), (2) the Core Survey (Core Institute Southern Illinois University at Carbondale), and (3 ) College Alcohol Study (Robert Wood Johnson Foundation). The results ar e undeniable. In a report given to college presidents in 2004, the Core Inst itute found that respondents in th eir sample consisting of over 141,121 students from about 300 institutions of highe r education reported that they perceived that drinking was a central part of the student social life, for both male and female students 26

PAGE 27

(Presley, Cheng, & Pimentel, 2004). Additionally Monitoring the Future (Johnston, O'Malley, Bachman, & Schulenberg, 2007) reports that rece nt graduating seniors from high school have more liberal attitudes towards drug and alcohol use relative to c ohorts of the past. Thus, these two very comprehensive and large-scale studies suggest the ambivalence noted by Lender and Martin historically is alive and well on college campuses across the States. Figure 2-1: Monitoring the Future: Percentage of College Students and Twelfth-Graders that Used Illicit Drugs, Alcohol, and Be en Drunk Within the Past 30 Days The most recent estimates show that college students have, for a long time, been more likely to drink and get drunk than are high school seniors. As a reference before further evaluation, Figure 2-1 shows the results of the latest Monito ring the Future study for the prevalence of use of various substances with in the past 30 days from 1975 2006. Generally, 27

PAGE 28

these data also show that college students share similar patterns of use and abuse of psychoactive substances when compared with individuals of the same age that do not attend an institution of higher education with a few exceptions. In a more complete comparison, high school seniors are contrasted with both college students and others who are up to four years past high school. College students are much more likely than similarly-aged non-students and high school seniors to drink overall and also to engage in binge drinking1 (see Figures 2-2 and 2-3). Figure 2-2: Monitoring the Future: Percentage of College Students versus Others One to Four Years Beyond High School that Drank Alcohol in the Past 30 Days 1 Binge drinking for this figure is defi ned as having five or more drinks on one occasion within the last two weeks, regardless if male or female. 28

PAGE 29

Figure 2-3: Monitoring the Future: Percentage of College Students versus Others One to Four Years Beyond High School that Drank Five or More Drinks in a Row within the Last Two Weeks Remarkable stability exists in the overall amount of college student binge drinking since the inception of these data sources. The only ap preciable trend in Figure 2-3 seems to suggest a slight decrease in colle ge student binge drinking from the 1980s to the 1990s, but persisting rates from that time forward. Yet recently popular medi a outlets have reported an alarming trend in binge drinking and prescription drug use in this population (see Critcher 2008; Weschler, Lee, Kuo, and Lee, 2000). These reports are not without merit one ju st has to examine the nuances underneath the overall rates of binge drinking to realize what warrants such alarm to see competing trends. The rates of frequent binge drinking increased, but the rates of abstention have increased as well (Weschler, Kuo, Nels on, and Lee, 2002). The Harvard Alcohol Study 29

PAGE 30

finds that about 19.7% of students surveyed frequently binge drank2 in 1993 compared with 22.8% in 2001 (odds ratio of 1.21); 16.4% abstai ned from drinking in 1993 compared with 19.3% in 2001 (odds ratio of 1.22). This suggests movement of some of those who used alcohol moderately to two polar extrem es frequent binge drinking a nd abstinence (Wechsler et al., 2002). Overall, Wechsler et al. (2002) report an increase in the harm among collegiate drinkers in spite of reform to address binge drinking on ca mpus something that is consistent with the increase in frequent binge drinking (see also Kinney, 2006). Such harms may include missing class, blackouts, altercations, and actions students subsequently regret particularly sexual activity that may put ones health at risk or lead to unplanned pregnancy. One of the problems associated with alcohol, especia lly heavy use of alcohol, is arrest. The official arrest data reported by institutions of higher education since 1992 show a slight annual increase in liquor-law viol ations/arrests reported by these institutions (U.S. Department of Education, 2008). In 2003, 31,234 on-campus criminal violations of alcohol laws were reported a 1.3% increase from the previous year.3 The increase seen in 2004 is roughly 4.8% (34,394 violations). Some media and government reports reference st eep long-term increases since 1992/1993, the years with the lowest contemporary prevalence of drug and alcohol use/misuse among student populations (s ee Johnston et al., 2007). The total number of on-campus drug-law viol ations reported by institutions of higher education in 2002, 2003, and 2004 show similar in creases as alcohol-l aw violations: 12,393, 2 Binge drinking in this study is defined by five or more alcoholic drinks on one occasion within the past two weeks for men or having four or more alcoholic drinks on one occasion within the past two weeks for women. 3 These numbers include on campus violations for all repor ting institutions, including less than 2-year institutions. The total number of violations for all reporting instituti ons for on-campus and off-campus violations was 48,807 in 2002, 47,904 in 2003, and 50,642 in 2004. Even with the slight decrease in 2003, the overall trend shows an increase when one examines the data from 1992 to the most recent available (C hronicle of Higher Education, 2007; US Department of Education, 2007). 30

PAGE 31

12,854, and 13,235 respectively (US Department of Education, 2008).4 Results from the National Household Survey on Drug Use and H ealth (SAMHSA, 2008) as well as Monitoring the Future (Johnston et al., 2008) i ndicate that use and abuse trends have remained stable within college student populations. It is important to note that the official statistics on drug and alcohol violations seem to greatly underestimate the amount of alcohol and drug problems in higher education. A quick comparison of these rates ve rsus that of the aforementioned self-report sources of data suggest problems are more widespr ead than what is requir ed to be reported. One reason for the discrepancy is that students tend to move off-campus to consume alcohol and the institutions of higher education remain blind to the amount of offenses committed away from campus. Many factors explain this behavior, incl uding alcohol availability, local economics, and university policies mandating a dry campus or areas. As students move off-campus to drink and use drugs, it is also important to remember that both their offending and victimization patterns may be altered. Substance Use over the Life-Course While college students are not immune from substance addic tion or dependence, students generally fare better than others due to protective factors that eith er ameliorate the levels of use or prevent future use and relaps e. These protective factors in clude general higher educational attainment, socio-economic status, social capit al, greater bonds to c onventional society, among many others. The differences in the prevalen ce and quantities generally consumed between college students and high school graduates that do not go to college (thus excluding drop outs, who are differentially at a hi gher risk for heavy drug use a nd abuse) suppor t this notion 4 The total number of violations for all reporting institutio ns for on-campus and off-campus violations was 25,058 in 2002, 21,699 in 2003, and 21,859 in 2004. Caution should be taken when assessing trends using these overall data as not every institution has jurisdictional outreach beyond their campus perimeters. 31

PAGE 32

(Johnston, OMalley, Bachman, & Schulenberg, 2007). Alcohol use is a notable exception. Students seem to matriculate into a culture th at readily promotes alcohol use and abuse as measured by drinking and binging rates.5 While the protective factors listed above often remain intact into the transition of hi gher education, substantial weaken ing of these factors may occur when students arrive at college. For exampl e, leaving home may weaken ones ties to conventional society in that a students bonds with his/her parents may change given the distance and newfound independence (Hirschi, 1969; Hirschi and Gottfredson, 1995). Alternately (or in tandem), students may differentially associate with other students that favor alcohol use and drinking heavily while distancing themselves from a wholesome home environment that frowned on alcohol (see Akers, 1998). While these protec tive factors typically re main intact, the joint weakening of this shielding combined with a cultu re ambivalent to alcohol use seems to create the perfect storm for exasperated alcohol problems on college campuses. However, current research notes an anomaly in the alcohol abuse lite rature pertaining to college students. In a recent journal articl e, Donald Misch (Misch, 2007) examines the phenomenon of natural recovery from alcohol abuse displayed by a su bstantial amount of college students as they progr ess through school and after they matriculate. Thus, while the culture of higher education itself may be a risk factor for excessive and dangerous drinking, Misch notes that researchers have been glossing over the potential protective factors that exist in that same environment that may help students natu rally reduce their levels of drinking. He was not only concerned with the factors that aid and abet college st udents into drinking habits, but highlighting the aging out of dangerous drinking behaviors. Accordingly, the substance use 5 See Appendix A. 32

PAGE 33

careers of college students and in particular, the desistance/ce ssation aspect of the life-course perspective is sorely unde rstudied and lacks development in current literature. A notable exception is the research done at the University of Wa shington on the natural history of alcohol use of college students (Bae r, Kivlahan, Blume, Mc Knight, & Marlatt, 2001; see also Schulenberg, Maggs, Long, Sher, Gotham, Baer, Kivlahan, Marlatt, & Zucker 2001). These researchers began their study by sending ou t a questionnaire to all incoming freshmen, class of 1990. Freshmen respondents who committed to the study were then screened for highrisk drinking patterns across the previous th ree years by using the Rutgers Alcohol Problem Index (White & Labouvie, 2008). Those determin ed high-risk were divided among a treatment group which would receive a one-time, indivi dualized prevention-base d intervention and a control group. The results coincide with what Misch (2007) predicted. Globally, students tend to drink more when beginning their college career relati ve to ones high school years. Yet, after freshmen year, students tend to drink less often that is until they reach their 21st birthday when the frequency of drinking occasions suddenly increas es. However, Baer and his colleagues discover that generally those in their advanced college years tend to drink less (in quantity) on each drinking occasion but drink more often so that the overall quantity increases over the freshmen and sophomore years. The high-risk students who received the inte rvention, however, drastic ally differed on their drinking habits relative to the ge neral sample and the controls. Those in the control group, for example, consistently drank more often and with greater quantities than did their global student body counterparts. But, more im portantly, a difference in the pattern of use was noted between the experimental group and the control group. The control groups frequency of drinking 33

PAGE 34

occasions remained relatively stable, but the qua ntity of use slowly diminished over time. The treatment group evidenced the same patterns wi th regard to drinki ng occasions, but the intervention had a significant impact in yielding additional reductions in the quantity of drinking beyond that displayed among the controls. Additi onally, Baer and his colleagues measured the self-reported unintentional conseq uences from alcohol use, and those in the treatment group showed considerable reductions in these ne gative outcomes relative to the controls. A final note on the life-course perspective on substance use and abuse as it relates to college students: just like developmental crim inology predicts a chronic offending, high-risk group and another that is noted by brief, often episodic delinquenc y, extant research on collegiate substance use and abuse aligns with this framew ork. The only exception to this analogy is the timing in which these events (crime versus problematic/delinquent alcohol consumption) generally occur in the lifespan. As alluded to in the University of Washington study on alcohol use over the course of the academic career, a small amount of substance users consistently consume the vast majority of intoxicating substa nces as a group relative to the wider population. For example, results from the most recent natio nally representative Co llege Alcohol Study find that 22.7% of the students in their study consum ed over two-thirds of the alcohol consumed by the entire sample (Weschler et al., 2002). The remaining students that do drink tend to do so moderately, but do experience epis odic problems direc tly related to their alcohol consumption. The present study will examine va rious renditions of labeling th eory to begin to frame why some students may continue to experience drugand alcoholrelated problems (measured by recidivism) after being processed and or sanctioned by an authorit y while other students may fare better depending on which authority (formal or in formal) processes them. These perspectives 34

PAGE 35

will be juxtaposed with deterrence theory as this perspective offers different solutions as to why students may continue to have problems. Reintegrative Shaming and Labeling Classic labeling theory suggests that efforts of social contro l can backfire. These efforts can produce institutional processe s that push some vulnerable indi viduals, especially novices, toward a criminal lifestyle. Offi cial reactions can alter concepts and identities, in part because they engender informal reactions from others an d because they entrench those labeled within a formal process that increases exposures to dense cr iminal elements (jails, prisons, etc.) rich with learning opportunities (Tannenbaum, 1938). The societal reaction itse lf may produce what Lemert (1951; 1972) called secondary deviance. Quite simply, when indi viduals are treated as drunks, as criminals, or as deviants, at some point, the labeled i ndividuals will gradually accept these labels and align their lifestyles to better suit what has become their newfound primary status through what is termed a self-ful filling prophesy. Labeli ng theorists argue that the deviant self-concept materiali zes in youth and this will in tu rn trigger deviant behavior. While many well-known criminologists have cont ributed to in this tradition (Becker, 1963; Erikson, 1966; Kitsuse, 1964; Schur, 1969), clas sical labeling theorys prominence has waned due to the lack empirical support and the internal complexities of researching its primary tenets, which critics argue remain too ambiguous (see Patternoster & Iovanni, 1996; Cullen & Agnew, 2006). A contemporary resurgence of the labeling pers pective has taken place in the form of novel applications of these classic th eories with innovative specifica tions that have the rigor to withstand empirical investiga tion. Both Bernburg and Krohn (2003) and Sweeten (2006) point out that this renaissance holds particular import when examin ing the impact of arrest and sanctions on youths and young adults. These res earchers draw attention to the process of 35

PAGE 36

cumulative continuity and couch it in a labeling perspective. That is, these researchers argue that the actions of the police and of the criminal justice system spur a chain reaction which creates barriers to legitimate opportunitie s and processes (e.g., employment and educational attainment). Furthermore, this process is mediated by ones location in the social structure those from disadvantaged backgrounds will have a more difficult time at overcoming these hurdles (see Sampson & Laub, 1997). In this case, future deviant behavior arises from the restricted access to legitimate opportunities, not from a deviant self-concept. Bernburg and Krohn (2003) and Sweeten (2006) find support for this in high scho ol students: students th at were arrested and processed by the criminal justice system were mo re likely to drop out of high school and to be unemployed in early adulthood, particularly if these individuals la ck the shielding provided by a higher position in our social stru cture. Figure 2-4 presents a depiction of Bernburg and Krohns theorized process of events that leads to further offending for ensnared youth. Figure 2-4: Bernburg and Krohns Hypothesized Effects of Offi cial Intervention in Adolescence on Crime in Early Adulthood A careful look at the processes proffered by Bernburg and Krohn in Figure 2-4 suggests that extrapolating this model to college students is possible with a shift in thinking when defining structural disadvantage. Expa nding structural disadvantage to include concepts broader than economic disadvantage makes this possible. For example, college students experience a 36

PAGE 37

spectrum of changes when transitioning to hi gher education which can yield structural disadvantage broadly defined: lo ss/alteration of social support network, loss of rights to privacy due to the demands of communal (dorm) living, and an unfamiliarity with available resources such as extracurricular activities, cultural even ts/venues, and volunteering opportunities. While many students are able to reintegr ate into the social structure w ith a smooth transition to higher education, others are less adept in accomplishi ng this. Thus, the missing factor from Bernburg and Krohns model is the life-course transition that shapes differential structural disadvantage across college students. This t ype of disadvantage should work in the same manner in that it has an interactive influence on the e ffect an official intervention ha s on educational attainment and recidivism. While the model provides an excellent resource in the aid of explaining the effect of official interaction on these two outcomes, it does not offer a clear understanding of how different types of official intervention fact or in to alter the process. Some alternative solutions are found with Br aithwaites reintegrative shaming approach (1989; 1992) which also leads the labeling renaissance into the ne w era. Spring-boarding from the concept of classical labeli ng, Braithwaite holds that the so cietal reaction to crime and deviance may paradoxically impact offenders negativ ely and make matters worse. He contends that reacting to crime and devi ance entails shaming. If shaming is done in a stigmatizing manner, crime and deviance will be amplified. Shaming is a natural reaction, whether form al (e.g., the courts) or informal (e.g., a disapproving mother), to behaviors which are not condoned. These reactions are designed both to elicit feelings of remorse from the targeted offender and to generate condemnation from those affected by with the incident or even from the surrounding community at large. For example, a juvenile who is caught vandalizing school prope rty may be sternly disciplined by school 37

PAGE 38

officials, generating a phone call to the childs family. In turn, childs teacher and parents both communicate their disapproval of the vandalism, t hus reinforcing the administrations initial act of shaming. As others in the school and co mmunity learn of the incident, they, too, may condemn the vandalism. Braithwaite (1989) argues, along with classic labeling theorists, that shaming which occurs in a stigmatizing manner has the potential to amplif y deviance and/or criminal behavior. On the other hand, if the shame were to occur in a reintegrative manner, the offender would be guided back to conventionality and respectability while steered away from the criminal models and opportunities that may await him or her if handled in the traditional, retributive model of justice. Ideally, criminal justice with recompense jointly combined w ith a truly penitent offender will lead to the best outcome one in which the o ffender is reintegrated with conventional society and the parties negatively affected by the offe nders actions are satisfied with the judicial outcome. This functional possibility of shaming was the missing link previous renditions of the labeling theory lacked. Could it be that some college students are more likely to receive functional reintegrative shaming than others, whic h would then predict th at a portion of students will likely re-offend at a lesser rate than others? What if labeling theory is wrong? Accordi ng to labeling theory, th e hypothesized impacts of official reactions like arre st and court sanctions contrast s with the expectations from proponents of deterrence. Deterren ce theory, on the other hand, predic ts that the risk of criminal justice processing and punishment should serv e to secure compliance (Becker, 1969; Cornish and Clark, 1986; Tunnel, 1992). For those who have already been caught and punished, the sanction serves as a deterrent so they will know th e costs of engage in the proscribed behavior and refrain from doing it again (specific deterrence ). Punishing offenders serves as a model to 38

PAGE 39

others as well; other learn of the threat of punishme nt so they, too, will not want to take the risk (general deterrence). Taking th is one step further, Stafford and Warr (2006) suggest a reconceptualization of the terms specific and genera l deterrence and shift their thinking to examine the impact of the deterrent effect of direct and indirect experien ce with punishment and punishment avoidance Either way, this perspective suggest s offenders make a rational choice to commit deviant and criminal acts with considerati on to a cost-benefit analysis of sorts. These tenets are at odds with the labeli ng perspective if the very pro cesses that are supposed to deter individuals from the commission of deviant a nd criminal acts actually backfire and enhance criminality, specific deterrence is a complete fa ilure. The great criminological debate that surrounds these two conflic ting perspectives is th at there is robust evid ence that supports the claims made by each (see Bernburg & Krohn, 2003; Piquero & Paternoster, 1998; Piquero & Pogarsky, 2002; Pratt, Cullen, Blevins, Daigle, & Madensen, 2006; Sweeten, 2006), particularly the revitalized versions of each perspective (e.g., Braithwaite, 1989; Braithwaite, 1992; Staffard & Warr, 2006). Shaming in a Culture of Ambivalence Leading one of the most expansive resear ch inquiries on alcohol on college campuses, Henry Weschler and his colleagues have described a culture of al cohol that exists on campuses across the nation. In a 2002 book called Dying to Drink Weschler and Wuethrich tap into some of the traditions at institutions of higher education that speak vo lumes on this subject. Several examples illustrate the ambivalence that exis ts about college drinking; some tradition are seemingly harmless (and may even be tied to wort hwhile goals), but may also hold the potential for harmful and negative consequences. Cupid W eek, a fraternity fundraising event in which the most timid fraternity brother is chosen to drink until intoxicated and dress up as Cupid. Cupid is then taken across campus to kiss willing girls, for each, he earns a donation for the American 39

PAGE 40

Heart Association. The tradition of drinking yo ur age occurs when one reaches the legal age to drink alcohol (or the night be fore) and attempts to drink 21 s hots is a particularly harmful tradition because of the very heavy drinking it involves (Weschler and Wuethrich, 2002). Substances other than alcohol are celebrated on college campuses and illustrate ambivalence about how to react to college student use. For example, at the University of Colorado at Boulder, students have held an annual pot smoking festival on April, 20th to celebrate the cult -phenomenon of 6 at Ferrand Field, a centralized public field on campus. The festivities are peppered with students openl y consuming marijuana from joints, pipes, onehitters,7 and laced-foods while assemble d to listen to live music, promote the legalization of marijuana, and socialize. On April 20th, 2006, campus police posted several signs along the perimeter warning students that anyone using ma rijuana would be photographed and prosecuted. Not heeding the warning, a crowd of 2,500 crosse d the fields boundary and participated in the festival. The following day, the police websit e contained the images of 150 individuals (primarily students) actively smoking, offering a $50 award for the identifica tion of each student. The mixed public reaction was telling some argued that the police over-extended their boundaries when they posted these images on th eir website (with bold IDENTIFIED stamps on each identified offender) while others comme nded the police for seek ing out blatant law violators, heating up the debate in the local newspapers. 6 The story of 420 is that of contemporary marijuana-user folklore. Karen Halnon (Halnon, 2005) has described it as a universal code that all marijuana-users relate to without even knowing the true basis for it. 420 can mean the time of day users toke up, a symbol between two or more users that they are pot friendly, and actually gives users an identity. Its imbedded in marijuana-using rituals, written on a milieu of products (t-shirts, cups, stickers, posters, and much more), can be a hip apartment number or address at college, and all people turned-on to the culture knows the power of 420. Halnons article itself, originally printed in High Times, has become part of the phenomenon: it has been reprinted or referred to on many pro-marijuana blogs and websites. 7 A one-hitter can be described as a compact, often covert, marijuana smoking device. It typically holds a very small amount of marijuana, which gives rise to its name. A specific example of this type of device is called a bat, which is often made out aluminum, and is stored in its dug out. The dug out is an inconspicuous storage box that both holds small quantities of marijuana and its paired bat. 40

PAGE 41

Figure 2-5: University of Colorado at Boulde r 420 at Farrand Field Police Surveillance. Source: University of Colorado at Boulde r Police website (n o longer available): http://www.colorado.edu/police/ Two years later, the event expanded to a larg er field on campus and the crowds swelled to over 10,000 while the police resigned to take a safety monitoring approach instead of aggressively enforcing drug laws. The Public Information Officer was quoted in a local newspaper, explaining his reasoni ng: We cant do the same thing year after year, [Commander Brad] Wiesley said hours before Sundays smoki ng began. So I doubt well do anything like the pictures. Theres no way our 12 to 15 officers are going to be ab le to deal with a crowd of 10,000. We just cant do strong enforcement when were outnumbered 700 or 800 to one 41

PAGE 42

(Miller, 2008). Thus, af ter turning on the sprinklers in one field during one year, and actively sanctioning students in another, the police deferred to the students and the communitys ambivalence. If the current climate of the campus culture is characterized by ambivalence, then it is important to note that the shaming component of formal or informal sanctions as predicted by Braithwaites theory of Reintegrative Shaming may be nullified, regardless of how the administration or the local criminal justice system process the majority student offenders. If the underlying assumptions of shaming require 1) invoking remorse within the offender, and possibly, but not absolutely 2) garnering condemnation from othe rs who learn of the offenders actions, the underlying sanctioning process may beco me broken if the offender is not remorseful (except, perhaps, of being caught) and/or if he or she perceives little to no condemnation for the delinquent act(s) committed that provoked official attention. One may contend that the process of arrest itself arguably has the pote ntial to elicit stigmatization due to the nature of this reaction; however, the typical police response with minor criminal infractions such as misdemeanor marijuana possession8 is to issue a written citation of arre st (known as a notice to appear). This process lacks handcuffs, being pla ced in the backseat of a polic e cruiser, and the subsequent transport to the local jail. Stigmatization in these circumstances seems to be as minimal as possible, which seems to be in line with Bra ithwaite. Yet, without any shame, specifically reintegrative shame, the culture of ambivalence w ill flourish. The reaction may make a differencediscretion over how to handle dri nking, underage drinking, marijuana consumption, et cetera may mean very different social re actions in a university town where much of the activity is in th e city or county. 8 Defined by possessing less than 20 grams of cannabis in the State of Florida. 42

PAGE 43

This phenomenon may not hold for all offenses however. As mentioned previously, drunk driving is considerably stigma tized (see Grasmick, Bursik, and Arneklev, 1993) despite the largely accepted patterns of drinking that enable drivers to slip into that state. In that vein, any acts or crimes considered serio us or being caught with substa nces considered dangerous or dealing drugs may have different quantities of shame for which the quality of shame, whether reintegrative or stigmatizing, can take shape. This added component to Braithwaites model of reintegrative shaming may be key in adapting hi s theory to college students, and especially collegiate alcohol and drug use. On the other hand, Piquero and Paternoster (1998) found Stafford and Warrs reconceptualization of deterrence th eory helpful in explaining drivers intentions of driving while intoxicated. Using a sample of over 1,600 rando mly selected drivers from across the United States, these researchers found that both pers onal and vicarious experiences affected these intentions. Additionally, being punished for a DUI and having successfully avoided capture and punishment for this crime in the past altered dr ivers intentions to drive drunk. Two concepts were found to be of particular import to this st udy: substantial reductions in the intentions of driving drunk among both those who reported high levels of the certainty of punishment and those who had strong moral beliefs that prohibit th is behavior. That is, respondents in the study that have acknowledge a certain ty in punishment for drunk driv ing tended to express reduced intentions to drink and drive while thos e that acknowledge a good chance of avoiding punishment tended to express increased intentions to drink and drive. These intentions were found to be more resolute for respondents with personal experience with punishment or avoiding punishment for driving drunk. For example, if someone drove intoxicated 100 times and was caught once, this individual may perceive a hi gh chance of avoiding punish ment for his behavior 43

PAGE 44

and tended to report increased intentions to drink and drive relative to someone who drove intoxicated 10 times and was caught two of t hose times. Vicarious experiences, knowing of people that were punished or a voided punished for drunk driving, has the same impact but to a lesser degree. If petty offenses beneath this level (underage drinking, public intoxication, open beverage container violations) ar e not taken seriously by a large segment of the population or the stigma attached to these offenses is weak, nonexist ent, or non-reintegrative, the lesser types of drug and alcohol offenses may flourish and thrive. One Journalists Perspective at the Study Institution In the fall of 2008, New York Times reporter Kevin Sack visited Gainesville, Florida on the weekend the University of Florida Gators foot ball team were slated to face the Ole Miss (University of Mississippi) Rebels at Ben Hill Griffi n Stadium. At the behest of his editors, Sack came to the college town to investigate what has been declared the number one party school by the Princeton Review (Princeton Review, 2008; Sack, 2008a). His article (2008b) begins with an interview with a fraternity brother that is notic eably drunk at 10:30 in th e morning of the football game a signal (eye opener alcohol consumpti on) to health practitioners of an underlying substance abuse or dependence pathology when in consideration with othe r factors (APA, 2000). This sets the reader up for the shocking state of the alcohol problem in Gainesville. In the lengthy piece, Sack interviews various underg raduates, local offici als, and university administrators to support the cl aim that in Gainesville, binge drinking remains ritualized behavior for many of the 51,000 students, even as admission to the university has become increasingly selective (2008b). This immersi on in a sea of alcohol is in spite of a multidimensional affront on underage and binge dr inking in the community and on campus in recent years. In this time frame, key univers ity and community decision-makers took various strides such as stricter enforcement of the lo cal liquor laws, an aggres sive social marketing 44

PAGE 45

campaign (see Kotler & Roberto, 1989; Wechsler Nelson, Lee, Seibring, Lewis, & Keeling, 2003; Zimmerman, 1997), offering alternative activ ities on Friday nights, and enhanced late night transportation that targeted areas with a high density of bars and clubs and high density student living areas (Glassman, 2008). Figure 2-6: Binge Drin king Rates. Source: University of Florida Gatorwell website: http://www.shcc.ufl.edu/gatorwell/ The aggressive approach may seem to be pa ying off by some metrics: the rate of binge drinking measured with an annual survey (using the Core Alcohol and Drug Survey) administered by the Student Health Care Cent er has fallen since 2004, most noticeably between years 2004 and 2005 (Gatorwell, 2008; see Figure 2-6). Using a more robust methodology and capturing a better response rate, a survey s upported by a large gran t uncovered that the percentage of undergraduates that reported binge drinking in the pa st two weeks was estimated to be 35.9% in the spring of 2007. This compares w ith the Student Health Care Centers estimate of 44.4% for the fall of 2007 and its consistently over 40 percent rates that span over a decade prior. Sacks qualitative assessment suggests, ho wever, that the culture of ambivalence seems to be thriving as discussed earlier despite any response from the uni versity or the local community. It seems that the campus self-report surveys may be missing the mark especially when considering the fallacies of comp aring cross-sectional data over time. The lull in the rate of 45

PAGE 46

binge drinking during the 2004-2005 academic year may be largely contributed to measurement artifact. Triangulation is key to reduce this pot ential noise and to incr ease the reliability of problem assessment. Summary Over twenty six years have passed sin ce Lender and Martin (1982) published their historical analysis on alcohol which noted a marked ambivale nce about alcohol and how to handle alcohol problems. Getting loaded as a college experience seems to be normalized, as evidenced by the volumes of research on the topic. To date, the vast majority of this research is based on self-report students performed on a crosssectional basis. The present study proposes to extend the available literature with official data collected longitudinally. In regards to timing, what seems to be operating in the lives of student s upon matriculation is a ch ange in the level of protection inherent in their so cial capital, socio-economic st atus, and bonds to conventional society while simultaneously entering an atmosphere that promotes alcohol use. Students bonds with their conventional families may weaken, social networks inevitably change, and many witness substantial freedom and independence formally unknown to them. However, for the present study, the exact mechanis m of pathology during this se nsitive transition period is secondary to the raw impact of official processing and sanctioning during this time. Specifically, this research seeks to unders tand the role of official processing and subsequent sanctioning (or lack thereof) in a ffecting the likelihood of reoffending and altering the academic performance of college students. These effects have been framed in two primary, and conflicting, ways. Labeling theo ry predicts an increase in the likelihood of reoffending and a decrease in academic performance as stude nts are handled through a process which is stigmatizing. Deterrence theory, on the other ha nd, would predict an increase in the likelihood of offending when students perceive low levels of certainty and severity of punishment. These 46

PAGE 47

classical theories have been re vitalized in recent years that of fer substantial advantages. In particular, Braithwaites (1989) th eory of reintegrative shaming offers a framework that is more adaptable to studying the cultu ral ambivalence of alcohol. Fundamentally, Braithwaite acknowledges that when shame occurs in a stigmatizing manner, there is an increased likelihood of an amplification of deviance and/or criminal behavior. The extension to classi cal labeling is that Braithwaite al so proposes that if this shame were to occur in a manner that is reintegrative, offenders will reduce their levels of offending. The working hypothesis is that if college st udents were processed and sanctioned through a system that is inherently less stigmatizing by na ture (e.g., the university sy stem), there should be marked reduction in the levels of reoffending and an increase in academic performance. Yet, when this shame occurs in a culture of ambi valence, this may nullify the shaming component tied with processing and sanctioni ng offenders. This alternate hypothesis predicts that offenders processed/sanctioned by either system will share similar outcomes regardless of any qualitative differences in the procedural experience. On the other hand, the reformulation of deterrence theory (Stafford and War, 2006) predicts that students who are puni shed by either the criminal just ice or university system for drug and alcohol offenses will have an increased likelihood of reoffending relative to students who avoid sanctions. As specified, this theory predicts no differences in outcomes across differential processing only diffe rential sanctioning. Extending th e alternate hypothesis listed in the previous paragraph, the working hypothesis in this scenario is th at level of sanctioning (e.g., none, mild, severe) will explai n the differences in outcomes for college students. While students may be embedded in a culture that encourages alcohol use and abuse, they should not be impervious to the effects of punishment. Yet, if students avoid punish ment, the culture of 47

PAGE 48

48 ambivalence may continue to exasperate problems in their college experience. As Piquero and Paternoster (1998) find with thei r respondents of drivers, persona l experience with punishment and punishment avoidance is the strongest pred ictor of future intentions of wrongdoing when taking moral beliefs out of the equation. Since all of these students were caught for drugand alcoholrelated offenses, eliminating this co ncept does not produce theoretical problems. What happens if an arrest and its subsequent sanctioning for alcohol and drug offenses fail to have an impact on future offending? Will these events trigger a decrement in student academic performance or no change at all? Wo uld going through the motions of traditional law enforcement and judicial process be benefici al to the university and to the surrounding community? If there were an alternative, less punitive and stigmatizing process available through the university, would that wo rk better? Do any sanctions ma tter for these crimes, or will the culture predict the outcome as many are sugge sting? The present study will now turn to assess these questions in its subsequent chapters.

PAGE 49

CHAPTER 3 PRELIMINARY ANALYSIS: STUDENTS A NONSTUDENTS IN A COLLEGE TOWN Introduction The purpose of this dissertation is to learn mo re about the role of alcohol and drug arrests on the academic performance and potential of re-arrest or re-referral of university students. The study is conducted in two parts. In the first phase, college student arrests are examined in context of the wider community arrest patterns in which a large public univ ersity resides. To achieve this goal, the present chapter analyzes a population of individual s arrested in Alachua County, Florida between 2003 and 2007 subdivide d into college students and community members. These results will be used to info rm the second phase of this dissertation which focuses on the impact of sanctions on academic performance and recidivism. Study Population The study population was derived in part of from a wider University research project initiated by Student Legal Servi ces. Its interest was in lear ning more about students who are arrested to see how that populati on was similar to or different from the students Student Legal Services actually serves. Toward this aim, each law enforcement agency that has jurisdiction in Alachua County was approached by the principle investigator to provide information of all arrests, notice to appear citations, and swor n complaints for the years 2003-2007 excluding minors. Of these agencies, the Alachua Count y Sheriffs Office (ASO), Gainesville Police Department (GPD), University of Florida Po lice Department (UFPD), and the Division of Alcoholic Beverages and Tobacco (ABT) responded. According to the Florida Department of Law Enforcement, these agencies are responsible for greater than 90% of the arrests in the 49

PAGE 50

county each year (2008).11 This police data contained the date of arrest, time of arrest, location of offense(s) and arrest, list of offenses/charges, name of a rrestee, police identified race and gender, and date of birth. Included in these police logs are all stude nts and local residents arrested in Alachua County that ar e 18 years of age and older. Arre sted visitors are also included in these logs, but criminological wisdom suggests that their particip ation in the local arrest rates is negligible.12 University students were isolated from non-student community members by cross-referencing names and dates of birth of arrestees with the Office of the Registrars and the Dean of Students Office databases. An analysis of student arre stees versus arrested community members is presented in this ch apter which will examine the differences in these subpopulations (if any) in the 18-24 age range in Alachua County. It will also provide information about how this study population compares with other studies. Descriptive Analysis: Comparing Student versus Non-Student Populations To date, no criminological study has examined arre st data in which nested college students were identified to explore the possibility of any differences in criminal activity relative to the wider (non-student) community. Since the county-wide population of arrestees13 was obtained from the primary local law enforcement agencies, and the university student subset was identified, it is possible to l ook for different patterns of offe nding between university students 11 All arresting agencies in the county include the Ga inesville Police Department (6,917/53.2%), Alachua County Sheriffs Office (4,853 / 37.3%), Univ ersity of Florida Police Department (562 / 4.3%), Alachua Police Department (480 / 3.7%), High Springs Police Department (126/ 1%), Santa Fe Police Department (67/ <1%), Alachua Division of Law Enforcement (5 / <1%), Florida Highway Patrol Gainesville (1 / <1%), Waldo Police Department, and Alachua Florida Game Commission (Number of Uniform Crime Report Index crimes for 2007 / % out of total Index crimes for 2007. Thus, the repor ting agencies were responsible for abou t 94.8% of all Index crimes for 2007.) 12 A common thread in criminal justice education, theore tical criminology, and law enforcement practice is the general acceptance of the tenets of routine activities (see Cohen & Felson, 2006). A main idea of the perspective is that offenders will commit the major ity of their crimes in th e areas in and around the pl aces they live, work, and frequent if given the opportunity. 13 Excluding Florida Highway Patrol, Santa Fe Community College Police Department, High Springs Police Department, Waldo Police Department, Alachua Division of Law Enforcement, and the Alachua Florida Game Commission. 50

PAGE 51

and college-aged non-students in the county. A bove and beyond examining raw occurrences and general differences, an overall assessment of the types of crimes these gr oups commit violent, non-violent, drug, and alcohol offenses would be helpful to examine whether the groups share similar patterns of behavior or present a need fo r separate strategies of law enforcement due to significant differences. To accomplish this, the arrestees primary offense e.g., their most serious offense as guided by the FBI Crime Indexs hierarchy rule is rec oded into one of these four categoriesalcohol, drug, violent, or nonviolen t. See Table 3-1 for list of example crimes included in these mutually exclusive categories. In this exercise, the focus is on the differe nces in the patterns of drug and alcohol offending between students and the wider communit y. By and large, the literature suggests an overabundance of alcohol-related ar rests of college students for th e following reasons: 1) college students have consistently self-reported higher rates of active drinking (consuming alcohol within the last 30 days) as well as higher amounts of drinking relative to their peers who do not matriculate into higher education over the span of several deca des and by different (large, wellfunded) studies (Johnston et al., 2007), 2) st udents have increasi ngly reported negative consequences directly due to their consumpti on of alcohol in recent years (Dowdall, 2007; Kinney, 2006; Mustaine & Tewksbury, 2007), and 3) many students are embedded in a culture that promotes problem drinking regardless of the known dangers of alcohol abuse (Weschler & Wuethrich, 2002). Furthermore, public atten tion has focused on problematic drinking (e.g., binge and underage drinking) by college stude nts in recent years, which forced many administrators and local law enforcement agencies to take action. Since st udents exhibit some of the highest levels of drinking (pro blem drinking in particular) re lative to other so cial statuses and they have been targeted for interventions, st udents may be expected to have higher rates of 51

PAGE 52

arrest relative to the wider community. To make the comparison, raw occurrences must be computed into rates by taking the total number of each type of offense and adjust it per 1,000 students and non-students for the study period.14 Arrest rate = (number of arrests / (average population for years 2003-2007)) 1000 In addition, an alternative ag ed-censored version of the general populations rates of offending will be included to hone in on the diffe rences between university student participation in criminal behavior from that of a matched group of the non-student population based on age. Obviously the university student populations age is mean-centered in early adulthood while the wider populations mean will inevit ably be greater. This alternative focus will assist in determining whether or not young adults at a stat e university share simila r criminal patterns to those in the same age group who are not attending this institution in the surrounding area.15 16 14 The population estimates for Alachua Co unty during the years of the present study were provided to the primary investigator by the Bureau of Economic and Business Research (BEBR) at the University of Florida. The student body population facts and demographics for these years were gathered from the Universitys Office of Institutional and Planning website (http://www.ir.ufl.edu ) as determined by the total number of enrolled students per fall of each academic year. Student population figures by race and age (jointly) were not readily accessible, but were provided by special request of th e principle investigator. 15 In the case of the age censored models, the Bureau of Economic and Business Research provided population estimates for all 18-24 year olds as we ll as estimates of gender and race for th is age range. Upon consultation with BEBR, creating an average of the population estimates for 2003-2007 would provide solid estimates of the population in the county for this time frame. The Universi tys Office of Institutional Research and Planning also provides total numbers of enrollment by age stratified by gender on their website, however a special data request was made to this office to provide figures by race speci fically for students aged 1824 during the study period. 16 A layer of complexity was added to this comparison to determine the levels of which violent and non-violent offenses have co-morbidities with drugand alcoholrelate d offenses. This exercise examined any mention of drugand alcoholrelated offenses in the list of charges ava ilable for each arrestee. A duly noted limitation of this analysis, then, was the issue of unreported illicit drugand alcoholrelated activity if the arrestee was not charged with this type of offense. For example, a student that ge ts into an altercation at a local bar and is subsequently arrested for battery may not have charged with public into xication or a similar alcohol-related offense. The same reporting problem may occur with disorderly conduct offens es. However, the analysis does retain the ability to compare across groups in the aggregate as the manner of which law enforcement applies charges to arrestees is congruent between non-students and university students. Af ter examination, it was determined that the differences between the rates of drugand alcohol-related arrests as determined by primary charge alone versus any mention of drugand alcohol-related charges in eith er primary charge or s econdary charges were neg ligible. For example, among the student population, 97 additional cases would be included if any mentions of drugor alcohol-related offense appear in either primary or secondary charges. Th is results in an overall rate increase of two per thousand. The rate differences among the non-student population were strikingly similar. 52

PAGE 53

Presentation and Discussion of Findings The findings of this descriptive exercise can be found in Tables 3-2 and 3-3. The overall pattern of arrest, without contro lling for age, found in Table 3-2 s upports extant research. When observing the percentages of arrests by type w ithin each group, the vast majority (65.38%) of college students who were arrested from 2003-2007 were, in fact, arrested for alcohol-related offenses. Interestingly, the percentages of arrests by type for non-students are evenly split among alcoholand drugrelated offenses as well as violent offenses (about 18% each). By and large, individuals in this gr oup tended to be arrested for non-violent offenses in which the primary charge is not related to alcohol or dr ugs (43.46%). A better metric to compare across these two groups would be to calculate the rate of the occurrence of each arrest by type per 1,000 individuals in each group. This, too, is provide d in Table 3-2. Overall, about 39 per thousand non-students were arrested for an alcohol-related offense while stude nts were arrested at almost double the rate (about 81 per thousand students). In regards to non-violent offenses, about 89 per thousand non-students were arrest ed, about three times the rate of student arrests (28 per thousand) for this type of offense. Table 3-3 presents data for students and community members aged 18-24. The differences in the rate in which students were arrested for alcohol-related offenses co mpared to non-students virtually disappear when focusing on this age range. The change is stark when controlling for age, the rate of arrests for al cohol-related offenses is about 93 per thousand students compared to about 86 per thousand non-students. This roughly makes these two groups equivalent in their offending patterns when focused on alcohol-related cr imes. This result is contrary to what the extant literature based on self-re port would predict, but is in ag reement on the direction of the differences between these groups. If individuals aged 18-24 who do not matriculate into higher learning tend to report statistically significant re ductions in harmful dri nking practices compared 53

PAGE 54

with those that do attend colleges and universi ties, these differences should translate into differences in official statistics. That is, if the harmful drinking practices documented among students lead to arrest as high lighted in the robust research on the topic (see Johnston et al., 2007; Weschler et al., 2002), non-students who are thought to engage in these practices at a lower rate should have lower arrest rates for alcohol-related offenses. In Alachua County, nonstudents present with only sligh tly lower rates of arrest for al cohol-related crimes. Perhaps the drinking experiences between students and non-students are not so great as thought. Alternatively, maybe enforcem ent patterns differ for students from non-students. Current research (again, based on self-reports ) also suggests individuals aged 18-24 who do not matriculate into higher education should also show slightly higher ra tes of arrest for drugrelated offenses (see Johnston et al., 2007). That is, individual s aged 18-24 who do not attend a college or university tend to report more drug use across most categories of substances relative to college students. In these data non-students (38.44 per thousand) were arrested at just over six times the rate of students (6.35 per thousand) for drug-related offenses while controlling for age. The difference here seems to be primarily driven by a large number of cocaine arrests in the community, but it is interesting to note that there is no dearth in marijuana or illicit prescription drug arrests in the present data. College st udents use is typically on par with their noncollegiate peers within these categories while students are less likely to be cocaine users but not by a wide margin (Johnston et al., 2007). For this reason it seems the wide division between these two groups cannot be fully explained by the le vels of cocaine arrests in the community, but the direction of the difference is as expected. Th is is especially the case when realizing the most abundant drug-related arrest is for marijuana offe nses. According to se lf-reports both students and non-students share equivalent rates of mar ijuana use, yet in these Alachua County data non54

PAGE 55

students are arrested to a greater extent for marijuana. Inasmuch as the police agencies were inconsistent (both within and be tween agency) in their coding of drug-related offenses with the exclusion of marijuana-related arrests, cau tion needs to be taken when making these comparisons. For example, an agency that arre sts a larger concentrati on of non-students relative to other agencies consistently teased out cocain e-related offenses in their data rather than categorizing them as a general controlled substan ce offense, while an agency that arrested a healthy mixture of both student s and non-students remained in consistent in its reporting practices. Regardless of what seems to be the driving force for the differences between student and non-student arrest for drug-related crimes, the initial premise still stands existing literature on the prevalence of drug use does not suggest such wide divergence in th e arrest rates between students and non-students as found in these Alachua County data. Further refining this descriptiv e exercise, Table 3-3 presents the arrest data for only 18-24 year olds by gender and by race. The overall tre nd seems to agree with the existing research in that men compared with women tend to have elevated alcohol-related problems in this case, being arrested for an alcohol-re lated offense. Interestingly, the differences between men and women shift when observing the differences in rates within each group. That is, male nonstudents exhibited almost twice the rate of arrests per thousand compared with their female counterparts, whereas student males had experienced elevated rates of arrest compared with their female counterparts, but not near ly as much in magnitude. It seems that the group membership driving these differences belongs with the student status. Remarkably no differences exist between the rates of arre sts of male students versus non-student s in respect to alcohol offenses when controlling for age. Thus, being enrolled in higher education seems to be an aggravating factor for alcohol problems (gauged by arrest) for young women, but not for young men. These 55

PAGE 56

results are striking in light of the extant literature available on the topic (see Johnston et al., 2007; Weschler et al., 2002) in that the lack of perceived differences among males is counterintuitive to what is exp ected. In fact, college student males are considered to be among the highest at risk for alcohol-re lated problems in early adulthood. Possible Explanations of Discrepant Findings The discovery of similar rates of arrest for alcoho l-related offenses between students and non-students particularly is striking relative to what is predicted by the current research available on the topic. The discrepancy between these local arrest data and national self-report data can be couched in two primary, yet nonexc lusive, ways. First, the area of study could truly be classified as a college town. It is isolated from any metropolises by about two hours, and offers a limited set of recreational resources to youths and young adults relative to what is available in metropolitan life. In addition to this, the area of study contains tw o districts with a particularly high-density of bars and nightclubs (without the richness of alterna tive recreational activities that major metropolises may offer). This atmosphere, co mbined with an isolated college culture that encourages alcohol use (and some would argue, misuse) may prove to promote these behaviors in teens and young adults who live in th e area and are not attending college. For example, every year during the fall term, football season drives mu ch of the activities in the town on Saturdays. Law enforcement from all available agencies are called upon to keep the peace between rival fans as well as ensure th e well-being of fans, citizens, the university, and surrounding businesses from the swell of visitors and the high con centration of students and local residents mustering around university property. Tailgating before the game is a common occurrence and part of the univers ity culture a part of that cu lture is alcohol consumption. Local law enforcement strategies for these days sh ift to a lax enforcement of the letter of the law to accommodate the requirements for crowd control to some degree. Regardless, it is 56

PAGE 57

inescapable that alcohol-related arrests are elevated during this period due to a culture that encourages the outdoor consumption of alcohol with in large crowds. While this may not impact non-students as much as it does students, it is likely that a sizable proportion of the local residents and visitors participate in such activit ies. This example is only one aspect of the college town culture that may bleed over to the community. Figure 3-1 depicts the levels of alcohol ar rest over time for students and non-students (censoring for age 18-24). It is apparent that the patterns of arrest be tween the two groups are closely tied together in that th e resulting outcomes appear to be mi rror images of each other, only being different in their order of magnitude. In particular, the summer and fall months of each year exhibit elevated alcohol-re lated arrests. These months correspond to the time in which beginning freshman arrive at the university, to foot ball season, and to rush for fraternities and sororities. Of these three cultural events, football season is likely to be tied with both student and non-student alcohol related arrests. For this reason, Figure 3-2 depicts the levels of alcohol arrest ove r time for both groups while excluding the arrests that occur on foot ball home game-days. While the two groups become discordant in their pattern s of alcohol-related arre st over time, there still appears to be some similarity between them. For example, bot h groups display similar patterns of arrest for the summer and fall months of years 2003 and 2007. Because of the limited time line for comparison, conclusions need to remain tentativ e. The results, however, are suggestive. The results may be due to law enforcement strategies (e.g., crackdowns), but it is interesting to note that while accounting for football season, some simila rities in the patterns of arrest between these two groups still exist. It a ppears that football season and th e college town culture have independent effects on non-st udents alcohol problems (a s measured by arrest). 57

PAGE 58

Another possible explanation can be couched in a social inequality argument. Since many of the students who attend the uni versity come from privileged or middle class backgrounds, the discrepancy may be evidence of a loca l criminal justice system that is institutionally biased against lower class citizens. This is not to say that individual actors of th e system are racist or elitist, but the activities of the local police fo rces (excluding the University of Florida Police Department) focus on areas entrenched with high rates of crime and areas in which drinking behavior may spill over into public venues, beco ming visible to law enforcement (especially if they are patrolling those areas more). These areas, in particular, have a higher density of impoverished residents a large percentage of which are minorities. This argument may be somewhat limited in explaining the discrepancy between students and non -students in terms of alcohol-related offenses because the concentration of these offenses occur in two primary areas the areas of high bar density as described above.17 At least one of these areas, however, is located fairly close to low inco me, high crime areas of the city. However, this second argument does hold pr omise as a possible explanation for the differences between drug-related arrests for students and non-st udents. The poverty-stricken areas included in the current study are wrought w ith drug problems that are more likely to be visible by the police and ex clusively targeted by police administ rators. The college students at the university may use illicit drugs, particularly marijuana, to a notable extent; yet, they seem not consume illicit drugs in a manner that is as detectible by law enforcement in their normal routines or use in places where they are at higher risk to get caught. A large study of students conducted at the university in the sp ring semester of 2008 (Haberman, Asal, Brady, Cavanaugh, Emmere, Kendell, Miller, Pritz, Weiler, & Zhang, 2008) 17 In a forthcoming article by Matt Nobles, Kathleen Fox, a nd David Khey, these data were geospacially analyzed to understand the special patterns of student versus non-student arrest. 58

PAGE 59

that utilized self-report su rveys found that 12% of stude nts respondents reported smoking marijuana in the last 30 days. A smaller, annual survey of students conducted at the university (Gatorwell Health Promotion Se rvices, 2008) suggests this percen tage may be higher 18% of respondents reporting using marijuan a in the last 30 days. Thes e two studies used differing methodologies, yet both estimate that about 1 in 5 to 1 in 10 students activ ely use illicit drugs. The arrest data for drug-related offenses fo r students shows that only 7 or 8 per thousand students actually get into legal trouble because of it over the course of five years. Students certainly seem more likely to evade an arrest for drug offenses as it is implausible to suggest that non-students are five times more lik ely to use illicit substances (w hich equals the difference in the rates of arrest for students and non-stude nts for a drug-related offense in the county). A Missing Link: Jointly Controlli ng for Age, Gender, and Race Table 3-4 combines race and gender to breakdown the arrest comparisons. It examines the rates of arrest by race and gende r jointly for 18-24 year olds. This further refinement uncovers that indeed three factors are driving the va riability in the rates of alcohol arrests for these different groups: student status, gender, and race. To the extent that arrest is an alcohol-related problem, the results are predicted by extant litera ture (see Johnston et al ., 2007; Weschler et al., 2002). White college student males surpass all ot her groups in rates of alcohol problems as measured by arrest for alcohol-related offenses. In fact, while accounting for race, this analysis has uncovered that white male and female students ar e arrested at a rate over ten times as much as non-white male and female students during th e study period. Remarkably these differences are not as pronounced when observing the rates in the non-student population. At first glance, race (being non-white) seems to be more important as a protective factor for students relative to non-students or to the counter, race (being wh ite) seems to be more important as an aggravating factor for students relative to non-students. Evid ence of this protective factor 59

PAGE 60

argument can be seen in the difference in rate s between non-white students and non-students in these data, non-white non-students are arrested at an elevated rate compared with their student counterparts. Furthermore, these differences are more pronounced for non-white females as being female adds another layer of shieldi ng, so to speak, inhere nt in their gender. The patterns in differences in the rates of drug-related offenses are more complicated. Overall, the patterns of arrest fo r these offenses are in agreement with this literature in that within racial groups males are more likely than females to experience problems with illicit drugs, including arrest. One key departure, however warrants examination: within the non-student arrest group aged 18-24, a large discrepancy exists across the rates of whites and non-whites arrested for drug-related offenses. These data pr esent that non-white students get arrested at a much elevated rate than do whites while in the current literature, non-wh ites in this age group typically report similar illicit drug use (if not slightly less) compared with their white counterparts (Johnston et al., 2007; SAMHSA, 2008) These findings have been replicated across many waves of data, across decades, and across samples. Yet, in essence, this use/arrest rate discrepancy does not exist when examining th e arrest rate difference within college students across race. In fact, the rate difference is in concurrence with the robus t extant literature on the topic white students were arrest ed at a slightly elevated rate relative to non-white students while controlling for age. In essence, the important difference is that for white males, the drugrelated arrest rate is on ly slightly elevated for non-students, but for non-white males (particularly black males), the arrest rate skyrockets among non-students leaving a wi de difference between non-white male non-students and non-white male students. Furthermore, the percent increase for black female non-students is just as striking. 60

PAGE 61

A key explanation here would be the racial differences in cocaine arrests in the community. This racial differe ntial will not impact the student population since cocaine arrests are extremely rare for students in Alachua Count y. However, as previously mentioned, a large number of cocaine arrests were made over th e five year span of the present study and a disproportionate level of minoritie s were arrested for cocaine. This seems to be a driving force for community differences in dr ug-related arrests across race. Summary After an initial examination, it appears that race, gender, student status, and age have coalescing impacts on predicting the likelihood of arrest of indivi duals based on group membership. White male students (aged 18-24) a ppear to be at the hi ghest risk for alcoholrelated arrest, but the al cohol-related arrest rate for white male students are only a little higher than that for their non-student counterparts. The reverse is true for non-white males. Non-white male students have lower rates of alcohol-relate d arrests than non-white male non-students. The university seems to be protective for nonwhite males even as it ag gravates arrest prospects for white males. Unexpectedly, the student status may be affecting white female students more than white males. The difference in alcohol-related arrests between white female students and their non-student counterparts is actually greater than is true of white males. If the university aggravates the problem, it seems to do so most among white females. Non-white female students trend in the same way as nonwhite male students; the university provides some protection for nonwhite female students. Their rate of alcohol-rela ted arrest is also lower than that of non-white female non-students. Given all of the media attention highlighting problematic drinking on college campuses, perhaps our focus has been misguided since there is reasonable evidence to suggest that this i ssue is affecting college student s but is affecting white and nonwhite students differently. 61

PAGE 62

On face value, it appears that the university cu lture may exasperate the problems predicted by being white students but protect nonwhite students. When mi norities matriculate into higher education, the rates of arrest of these individuals are a fraction of the rates for whites (in or out of the university) and much less than for nonwhite s ages 18-24 who do not enter the university. The magnitude of this difference is remarkable at minimum, minority students are arrested at a rate of about 5 times less than minorities in the community of the same age depending on the type of offense. Overall, the rate differen ce between these two groups is of a magnitude of twenty-eight. These students seem to be able to resist committing criminal offenses to a considerable degree. The following chapter begins to assess the imp act of arrest and subsequent sanctions on student re-offending and academic performance to determine if this is in fact the case. This assessment will take into careful consideration th e effects of gender and race since these factors were determined to have substantial influence at a basic level of analysis. Specifically, the next chapter will begin to examine the differences of being processed by the criminal justice system and by the internal university sanctioning syst em for analogous drug and alcohol offenses. 62

PAGE 63

Figure 3-1: Alcohol-Related O ffenses in Alachua County, 2003-2007 Figure 3-2: Alcohol-Related O ffenses in Alachua County, C ontrolling for Football Home Games, 2003-2007 63

PAGE 64

64 Table 3-1: List of Offenses Included in Alcohol, Drug, Violent, and Non-Violent Categories Type List of Offenses Alcohol Underage Possession of Alcohol Open Beverage Container Selling / Purchasing Alcohol for Underage Person Open House Party Public Intoxication Drug Possession of Marijuana (< 20 grams, 20 grams), Cocaine, and Controlled Substances Trafficking in Drugs (includes cultivation of marijuana, trafficking in cocaine and prescription drugs) Prescription Fraud Possession of Drug Paraphernalia Violent Murder Sexual Battery Assault and Aggravated Assault Battery and Aggravated Battery Robbery Non-Violent Fraud (includes drivers license fraud, excludes prescription fraud) Theft / Larceny, Grand Theft, Grand Theft Auto, and Shoplifting Eluding the Police / Resisting Arrest without Violence Trespassing Destruction of Property / Vandalism Burglary

PAGE 65

Table 3-2: Arrests in Alachua C ounty, Florida by Agency and Type ASO GPD UFPD ABT Total Pop. Est. % Within Group Total Rate per 1,000 N Arrests Age < 25 Pop. Est. Age < 25 Rate per 1000 <25 Student Alcohol 220 3101 651 95 4067 65.38 80.84 3528 93.83 Drug 105 181 108 1 395 6.35 7.85 326 8.68 Violent 69 267 39 0 375 6.03 7.45 228 6.07 Non-Violent 179 785 419 1 1384 22.25 27.51 1240 33.01 Total 573 4334 1217 97 6221 (50,308) 123.66 5322 (37,591) 141.58 Non-Student Alcohol 1313 6969 629 332 9243 18.77 38.52 4935 86.19 Drug 3138 5836 222 27 9223 18.73 38.44 3911 68.45 Violent 4228 4474 86 1 8789 17.84 36.63 3502 61.29 Non-Violent 10171 10221 1006 8 21406 43.46 89.22 9484 165.99 Total 19423 27518 1943 368 49252 (239,915) 205.29 21732 (57,137) 380.35 65

PAGE 66

66 Table 3-3: Arrests and Arrest Rates per 1,000 Individuals Aged 18-24 for All Reporting Agencies in Alachua County, Florida in 2 0032008 by Gender and Race. N Female Arrests Rate per 1,000 Females N Male Arrests Rate per 1,000 Males N NonWhite Arrests Rate per 1,000 NonWhites N White Arrests Rate per 1,000 Whites Student Alcohol 1568 76.23 1956 114.90 113 6.52 3411 134.89 Drug 51 2.48 275 16.15 43 2.48 283 11.19 Violent 62 3.01 166 9.75 47 2.71 181 7.16 Non-Violent 276 13.42 958 56.27 168 9.70 1066 42.15 Total 1957 95.15 3355 197.07 371 21.43 4941 195.39 Non-Student Alcohol 1732 58.42 3180 115.67 430 32.99 4485 101.69 Drug 535 18.07 3273 119.06 2214 169.86 1597 36.21 Violent 1008 34.00 2494 90.72 2147 164.82 1355 30.72 Non-Violent 2209 74.51 7273 264.56 5378 412.61 4114 93.28 Total 5484 201.71 16220 590.01 10169 780.19 11551 261.91

PAGE 67

Table 3-4: Arrests and Arrest Rates per 1,000 Individuals Aged 18-24 for All Reporting Agencies in Alachua County, Florida in 2003-2007 by Gender and Race Jointly. N White Female Arrests Rate per 1,000 White Females N White Male Arrests Rate per 1,000 White Males N NonWhite Female Arrests Rate per 1,000 NonWhite Females N NonWhite Male Arrests Rate per 1,000 NonWhite Males Student Alcohol 1522 111.22 1883 162.29 46 6.68 73 13.47 Drug 45 3.29 235 20.25 6 0.87 38 7.01 Violent 37 2.70 144 12.41 25 3.63 22 4.06 Non-Violent 216 15.78 846 72.91 72 10.46 112 20.66 Total 1820 132.99 3108 267.86 149 21.65 245 45.19 Non-Student Alcohol 1638 71.83 2841 133.38 89 13.01 341 55.07 Drug 312 13.68 1285 60.33 222 32.45 1990 321.38 Violent 362 15.87 1003 47.09 646 94.42 1491 240.79 Non-Violent 906 39.73 3203 150.38 906 188.54 4066 656.65 Total 8332 141.12 3218 391.17 2247 328.41 7888 1273.90 67

PAGE 68

CHAPTER 4 ASSESSING SIMILARITIES AND DIFFERE NCES AMONG OFFENDING STUDENTS ARRESTED BY LOCAL AUTHORITIES OR REFERRED TO JUDICIAL AFFAIRS Introduction Recall that the study population of the previous chapter was derived by approaching each law enforcement agency that has jurisdiction in Alachua County to provide information of all arrests, notice to appear citations, and swor n complaints for the years 2003-2007 excluding minors. This police data contained the date of arrest, time of arrest, lo cation of offense(s) and arrest, list of offenses/charges, name of arrest ee, police identified race and gender, and date of birth. Included in these police logs are all students and local residents arrested in Alachua County that are 18 years of age and older in which students were isolated from community members through an electronic process by the Re gistrars Office and th e Dean of Students Office. The study population for the subsequent tw o chapters are limited to those students designated as beginning freshman in the Summer or Fall terms of 2004 (N = 338) who had been arrested sometime between 2003-2007. At this po int, currently availabl e arrest data for 2008 (January through August) were included as in dicated from district court records via the assistance of Student Legal Services. These da ta were further refined to examine only those students with an alcohol or drug related offense in either their arrest or university records as collected by the Dean of Students Office (N = 292). The justification for choosing students beginn ing their undergraduate career in the year 2004 is based on the availability of complete academic data from the Office of the Registrar for years 2004 December 2008. The majority of the students identified by the Registrar as being a student prior to 2004 lack the data necessary to perform the analyses proposed by the current project. Also, by choosing students from the 2004 entering class, the pr inciple investigator 68

PAGE 69

retains the ability to determine whether students st ay on track to graduate the University within the standard four-year term. If these students were to remain on track, the vast majority should have graduated by May 2008 using a conservative estimate or by August 2008 at the latest. Focusing on only those arrestees entering the university in the summer and fall allows the principle investigator to track a c ohort of entering college students over time, the vast majority of whom completed high school the previous spring. The University actively encourages students to begin the academic careers in the summer term (N = 129) to acclimate themselves to college and campus life, and their experi ences at the University may di ffer, but their previous cohort experiences should be similar in na ture (Khey, Fox, & Lanza-Kaduce, 2008).18 Furthermore, to engage the research ques tions found in the present study, the principle investigator sought to identify a group of similarly situated st udents which did not receive the same treatment (e.g., arrest and sanctioning) as the study populati on. To meet this need, the principle investigator identified a subset of st udents who were referred to judicial affairs by university officials (typica lly by the Housing Office) without an o fficial arrest or criminal justice processing. These records were included with th e University of Florid a Police Departments (UFPD) daily arrest log and coded separately from 2003 to 2005. The remaining years (20062008) of data were made available by the Dean of Students Office as UFPD no longer was the primary source of this information. Data from various sources were obtained for these students too (N= 351). This comparison group will beco me relevant when analyzing the impact of official processing and sanctio ns on students academic caree rs and their likelihood of reoffending. 18 The arrest data from 2003 and the first quarter of 2004 were retained for analysis to determine if any prospective students were arrested before matriculation at age of majority. Only one student was arrested before matriculating into the university while 18 years or older (February 2004). This student was retained in subsequent analyses where applicable. 69

PAGE 70

The Data In addition to the data on arrest and referrals to university judicial affairs, information was collected from several other sources. Through an automated process, the Registrar generated student information such as student identificati on numbers, self-identif ied race, ethnicity and gender, academic progress before and after the arrest event, involvement in NCAA athletics, high school identification, and stan dardized test scores. The id entified student numbers were then given to the Office of Sorority and Frater nity Affairs, the Dean of Students Office, and Student Legal Services to (respectively) primar ily append arrestees 1) social greek affiliation (yes / no), 2) University imposed sanctions both due to the arrest event( s) included in the police data and from disciplinary action not handled by the criminal justice system, and 3) outcomes in criminal court. Additional data (see Table 4-1) was al so obtained from these sources to assist the principle investigator in the present study. All identifiers or potential identifiers were subsequently removed to protect the students pr ivacy and to comply with FERPA regulations. A comprehensive list of variable s and their sources can be found in Table 4-1. More detail about their operationalization is provide d in various sections below. Table 4-1: Variables and Sources of Data Variable Description Police or University Authority Date of Arrest/Referral Date individual was arrested by police agency or referred to judicial affairs Time of Arrest/Incident Time of day indivi dual was arrested by police agency or if incident Primary Offense/violation Using FBI hier archy rules, most serious offense or universit y violatio n Secondary Offenses Other charges or referral reasons, if applicable Date of Birth Date of birth of offender Race Police/university iden tified race of arrestee Gender Police/university identified gender 70

PAGE 71

Table 4-1: Continued. Variable Description Registrar Office Self-Reported Race Race reported by student when applying to school College Affiliated college at any given term Major Chosen major at any given term Term Course Load Total credit hours taken during term Term Grade Points Sum of grades in numerical form* Term GPA Calculated by taki ng grade points / course load Cumulative GPA Calculated by averaging all term GPAs # of Withdrawn Classes Count of withdrawn classes by term # of Failed Courses Count of faile d courses (grade = E) by term High School GPA Weighted GPA (6.0 s cale) as recorded at graduation Standardized Test Score Test score as required for admission Standardized Test Code Code for test taken (SAT or ACT) Athletics Status Denotes part icipation in NCAA sports Greek & Sorority Affairs Greeks Status Denotes active/inactiv e students, alumni, & termination House Identification Numerical identi ty of fraternity or sorority Dean of Students Office Primary Offense Offense listed by Judicial Affairs Adjudication Guilty or not guilty Action Taken Terms of sanctions Circuit Court Adjudication Guilty, not guilty, nolle prosequi, nolo contendere Action Taken Terms of sanctions As a baseline comparison for both the arrest ee group and the group of students solely processed by the university for analogous offenses, academic performance metrics at the department-, college-, and institutional-level were made available by the Office of Institutional Planning and Research. This included average GPA by college per calendar year, average number of withdrawn and failed courses by depart ment per calendar year, and retention rates by academic year per beginning freshmen cohort. An important assumption is made regarding these 71

PAGE 72

72 metrics: each of these figures can legitimately serve as a proxy for the academic performance of a third group students not id entified as being arrested or referred by the present study.19 The justification for this assumption is two-fold. First, there is stability in these data over time with very little variance from year-t o-year. In fact, due to the na ture of the increasing standards required for admission, these metrics show slow movement towards superior achievement for each measure. Additionally, students arrested and referred only make up a small percentage of the total student population. If arrested and/or referred students shoul d in fact under-perform their peers in regards to these academic performance metrics, th e global measures should only be slightly impacted as the weight of the group of student s that were not arre sted or referred to judicial affairs should signif icantly mask the smaller group. Figure 4-1 depicts the study population embedded in the wider cohort. A Profile of Freshman, 2004 The concentration of the present study cen ters on the entry class of 2004. Both the subpopulation of arrested students and that of students referred for alcohol violations can be compared with the entry class on basic admissions information. From information available on the Office of Institutional Planning and Research s website and the data provided across the institution, it also appears that the incoming cohort is remarkably homogeneous in regards to their stellar high school academic achievement. 19 Arrested students typically were presented with a written citation of arrest (also known as a notice to appear) by the arresting officer on the date of th e offense. Within a short period, thes e arrestees were sent a letter from the district attorney stating that they will defer prosecution upon the acceptance of the terms found within the letter. These terms are typically a small fine or community service for a first offense and notification of revocation of these terms if one were to reoffend within a year or less. If these terms were met by the end of that time, the charges will be dropped and the only thing that may be publicly available is information on the arrest event itself. Students referred to judicial affairs, on the ot her hand, typically received a letter from the Dean of Student s Office requesting a formal meeting to discuss the violation in question. The likely outcome for a first offense is a written letter or reprimand and an online course that seeks to address the problem behavior indicated by the violation. Rarely do referred students avoid adjudication as the burden of proof for student code violations of this caliber is low.

PAGE 73

73 Figure 4-1: Visual Repres entation of Arrest and Referral Populati ons as Embedded in their Wider Cohort.

PAGE 74

Neither the study sample (comprised of students w ho were arrested at some point while attending the university) nor the comparison sample (comprised of students solely referred to judicial affairs for a drugor alcohol-re lated offense) differ from the incoming class overall on any available achievement metric when entering the university (see Table 4-2) Overall, both these samples fit within the middle 50% of the class as identified by the Office of Institutional Planning and Research: Middle 50% of the Class: High School GPA 3.8 4.3 (academic classes only; weighted GPA) SAT total of 1200 1380 ACT composite of 26 30 Of Those Admitted: 79 percent were in the top 10 perc ent of their high school class. 82 percent took 20 or more academic classes in high school. 67 percent took 21 or more academic classes in high school. 50 percent took 22 or more academic classes in high school. 38 percent took 23 or more academic classes in high school. This certainly makes the pr ediction of group membership difficult, if not impossible. In fact, at a descriptive level, it appears that predict an arrest or a referral using any criteria available to the primary investigator upon entry into undergradu ate education will be difficult. The probable exceptions will be race and sex: anticipating th at whites and males will be overrepresented among those arrested or referred to judicial affairs. The principal investigator checked the juvenile records of thos e in the arrest group to learn about their past criminality. It was limited. Of the 4,856 individual student arrestees identified by filtering local arrest data through the univers itys registrar databases, only 333 of these students had been processed by the Florida juvenile justice system prior to matriculation. The entering class of the 2004 2005 academic year had roughly the same proportion of students 74

PAGE 75

with prior arrest or referral: 18 of 292 arrested students had a juvenile record (6 percent). Roughly half of these students recidivate beyond thei r first arrest while enro lled at the university, none of them drop out or are dismissed for poor academic performance, and seven graduate on time. Of these 18 students, half of them got into serious trouble (either more serious misdemeanors such as a DUI or felonies) as juven iles. Ironically the students that got into less serious trouble as juveniles tended to get into more serious trouble according to the arrest data obtained while they were in college. Yet their peers seem to get into serious trouble and repeat trouble at roughly the same rates. Since these offenders tend to bl end in with their peers that avoid the criminal justice system as juvenile s combined with the observation that juvenile offenses for the arrested student population is relatively a rare event, it stands that having a juvenile record does not seem to have much to offer as an indicator that makes these students qualitatively different relativ e to their co-offending peers.20 Although the arrest and referral groups were si milar to each other and to the entering class overall on a variety of educational and demogr aphic variables at the point of entry, some differences among the groups emerge during their undergraduate careers. Table 4-2 examines differences in joining a fraternity and sororit y, time to completion, average number of failed and withdrawn courses, and GPA. In this example, GPA differences are captured by taking an individuals term GPA and subtracting their wider colleges mean GPA averaged over the last 14 available terms. These term GPA differences are then added together to observe an individuals overall performance relative to the college(s) they were enrolled in over his/her academic career. The means of this summary statistic averaged across the study and comparison samples as a whole are presented in Table 4-2. 20 These data were limited to those students who were a rrested as adults. The referral population did not get examined by the Department of Juvenile Justice due to their internal IRB review. 75

PAGE 76

Table 4-2: Overall 2004 Freshman Profile in Re lation to Those of the Arrestee and Referral Subpopulations Freshman Profile Arrestee Profile Referral Profile Applications 23,595 Admitted 11,911 Total 6,750 (2,300 Summer / 4,450 Fall) 292 (129 Summer / 169 Fall) 351 (113 Summer / 238 Fall) White (Non-Hispanic) 67.1% 78.4% 81.9% Male 43.2% 61.2% 67.8% High School GPA (Mean) 3.95 3.87 3.93 SAT (Mean) 1260 1270 1286 ACT (Mean) 27 27 29 AP or IB Students 85% 71.2% 76.2% Greek Life Participation 28% 48.3% 42.5% Average Withdrawn Courses 2.36 3.54 2.80 Average Failed Courses 0.60 0.76 0.99 Average GPA Difference -0.14 -0.28 On-Time Completers 53.5% (estimated) 49.3% 55.4% Recidivism (Excluding Juvenile Arrest) 47.6% 10.0% Any Juvenile Arrest 18 (6.2%) Dropouts/Dismissal (Total N) 2/1 (3) 29/18 (47) At least at a basic (descriptiv e/univariate/bivariate) level, the study and comparison groups begin to show complex differences worth describing further. The most impressive difference is the attrition of undergraduates due to dropout or dismissal that have been arrested at least once versus those just referred to judi cial affairs, whether once or se veral times, but never arrested. 76

PAGE 77

The literature reviewed in a previous chapter woul d predict undergraduates with repeat problems to present with poorer outcomes across the board; th is is not present in this summary data. Over 47 undergraduates (13.4%) that have been referred to judicial affairs fo r a drugor alcoholrelated offense drop out or are di smissed by the University or poor academic performance. This is a stark contrast to only three undergradua tes (1%) who dropout or are dismissed from the arrestee group. On the other hand, the arrestee group has a higher proportion of students who habitually get into trouble whether arrested or re ferred to judicial affairs. Almost half of the undergraduates in that have been arrested some time in their academic careers either get arrested two or more times or have been referred to judici al affairs and arrested in dependently at least one time each. This compares with only 10% recidivism of students exclusively referred to judicial affairs. These findings are echoed through the othe r performance metrics listed in Table 4-2: the referral subpopulation tends to under-perform the study group in regards to GPA (t = 2.64, p = 0.008) and number of courses failed (t = 1.59, p = 0.112) as well as the wider university on these metrics. When withholding students w ho dropped out or were dismissed, however, the arrestee and referral groups show no significan t differences on these metrics except for withdrawn courses: the arrest ee group withdrew from an average of 3.48 courses while the referral group withdrew from an average of 2.62 (t = 3.482, p = 0.001). Two other departures that be g further description are the marked overrepresentation of fraternity and sorority members among both the arrestee and referral groups and the differences in on-time graduation. Both the arrestee a nd comparison groups contain almost double the amount of fraternity and sorority members relative to the levels of social greek membership for the 2004 first-year cohort. This is consistent with the existing literat ure that finds that fraternities and sororities enhance the likeli hood of substance use and abuse, alcohol in particular. These 77

PAGE 78

data suggest that social greek membership is a risk factor for an arrest or referral for drugand alcohol-related offenses. It is al so important to note the high ma gnitude in which fraternity and sorority members are officially sanctione d for drugand alco hol-related offenses. On the other hand, an unexpected departure ex ists when observing group differences in ontime graduation among the arrestee group, referral group, and the wider 2004 freshman class. Relative to the university average, the indivi duals in the arrestee gr oup are more likely to experience delay in graduation but seem to be on track for a 5or 6-year undergraduate completion. Conversely, the referral group was found to be more successful in graduating ontime, with completion rates that outperform the entire institution if one were to exclude the dropouts and dismissals from this groups ranks (65.4% graduate within 4 academic years). Thus a dichotomy exists: it seems that students who are referred to judicial affairs and avoid the criminal justice system seem to either do extremel y poorly or tend to be mo re likely to graduate on time relative to the average university student. These two groups have different outcomes, yet there does not appear to be any explainable differences in group membership at this descript ive level. The histogra m presentation of these students average GPA differences over their academic careers relative to their wider college found in Figure 4-2 may provide some initial an swers. The graph on the left depicts the distribution of average differences in GPA for students relative to their resident colleges in the referral-only subpopulation. Interestingly, it seem s that two peaks are observed one mean centered at about 1.5 grade point s below students respective college average GPA and one centered at 0 (no differences relative to student s wider college average GPA). This bimodal pattern in the data is not present for the sample of students arrested at some point in their academic careers. There is also a more impre ssive left tail on the re ferral group distribution 78

PAGE 79

representing a cluster of students that are considerably underperforming their peers. These individuals have GPAs between 2 and 3 grade po ints below the average GPA for their college. This level of underperformance is also not f ound in the subpopulation of arrested students. Figure 4-2: Comparison of Aver age GPA Differences. (Referra l subpopulation is on the left; Arrest subpopulation on the right) Perhaps the students at the left tail of the referral group are very much at risk for dropping out or being dismissed for poor ac ademic performance. This is statistically suggested by the means comparison of these GPA differences men tioned in the previous paragraphs. When comparing the mean GPA difference between th ese two groups while including dropouts and dismissals, the difference was statistically si gnificant. This significance disappeared when excluding these dismissals and dropouts. The visual representation of this difference is impressive: Figure 4-3 depicts a hi stogram that is virtually identic al to the arrest subpopulation distribution except a slightly more positive skew suggesting slightly better academic performance as a group. It may be possible that individuals arrested or referred to judicial affairs in their first term (or very early in their academic career) experi ence exasperated academic problems at a critical 79

PAGE 80

time just as they are making the break from home. Due to the different contexts in which students may find themselves in the referral grou p versus the arrest group, perhaps differences exist in the timing of first offense between thes e groups. As early as st udents arrive at the university and move into their dorm rooms, they begin to be monitored by officials at the division of housing. Their visibility and th e guardianship enhance th e prospect of their alcohol/drug behaviors being s een by university officials. Figure 4-3: Average GPA Differences (Relative to Each Students Resi dent College) of the Referral Subpopulation with Dr opouts and Dismissals Removed. The minority of students who do not move into dorms their freshman year may frequent new acquaintances dorm rooms which are actively mon itored by housing officials. This translates into immediate supervision when students begi n their experience with higher education. Alternatively, arrest may be more likely when students begin to feel comfortable with the surrounding community and start expanding where they choose to recreate. Possibly early 80

PAGE 81

referral or arrest is a warning sign of maladjustmen t. Figure 4-4 depicts the timing of first arrest and referral. Figure 4-4: Depiction of Timi ng of First Offense for Both Referral and Arrestee Groups As predicted, for those in the referral group, the referral to judicial affa irs is likely to occur in the first two terms of enrollment. For those in the arrest group, the fi rst arrests or referrals seem to range over a longer time span. It is important to note here that th e arrest group consists of any undergraduate who has any arrest while in residence at the unive rsity regardless of referral to judicial affairs. Some of those in the arrest group al so have referrals to judicial affairs that did not lead to a formal arrest. There are 66 undergraduates in the arrestee group whose first offense resulted in a referral to judicial affairs without any arrest, and 52 of those 66 referral s occurred within the first two terms of enrollment at the university. Yet thes e 52 offenders do not show signs of academic problems in contrast to some of those in the referral-only group described in the previous paragraphs. These 52 students who were referred early and whose records included an arrest 81

PAGE 82

perform well. Their mean grade point average re lative to their college(s) of residence is only slightly lower (0.12 points), their average number of course withdrawals is 3.71, and the average number of failed courses is 0.65. In f act, these averages do not differ much from the wider group of 98 undergraduates with both an arrest and a referral to judicial affairs regardless of the timing of these events. The mean group GP A difference for these st udents relative to their college(s) of residence is 0.17, the average number of courses withdrawn is 3.93, and the average number of failed courses is 0.83. Overall, there seems to be three phenomena co-occurring in these data that will need further analytical refinement to explain: 1) a segment of undergraduates exclusively referred to judicial affairs seems to display poor academic performance and are at higher risk for dropping out or dismissal from the university; 2) the remai nder of the undergraduates exclusively referred to judicial affairs seem to outperform undergraduates with an arre st and even fare better than predicted by university performance metric averag es; 3) students with an arrest, regardless of whether they are referred to judici al affairs for the offenses for which they are arrested or for independent offenses, seem to under-perform the wider university but are able to resist dropping out or dismissal. At this point, it is possible to suggest a few possibilities for why these events may be occurring. For instance, the su ccess of the students found in the comparison group may be due to a therapeutic and reintegra tive response from the dean of students at the university. The undergraduates found in the arrestee sample, on th e other hand, may be e xperiencing significant declines in their academic performance and progre ss due to stigmatic and/or ineffective criminal justice processing. Qualitatively, this asserts that the typical re sponse of the university system, which most times consist of a meeting with the dean of students, a lett er of reprimand, and a 82

PAGE 83

subsequent task such as community service and/or an alcohol education course, produces positive outcomes relative to the typical criminal justice system response, which most times consist of a deferred prosecution of criminal charges and fines or community service. It could also be that the qualita tive difference in the experience of being processed between these two systems may produce significant depart ures in academic performance and progress. In regards to the fact that the university response is privat e, one-on-one, and not viewed as criminal, a sanctioned student may react differently to the criminal justice response which is public, batch processed, and has the potential of creating a criminal record. This suggests that there will be differences in academic performance metrics preand postprocessing in these two manners. Alternatively, the context in which students are likely to belong to these two groups can lend to a substantially differen t interpretation. To be exclusively captured in the comparison group suggests that ones offending patterns are rele gated, by and large, to private settings (e.g., dorm rooms). Individuals with an arrest, on the other hand, have overtly broken the law primarily in public regardless if this is done so on university propert y. Thus the comparison group may be comprised of students that choose to be deviant, but not do so overtly where (seemingly) the likelihood of getti ng in trouble is higher in public Ironically, it seems that the likelihood of getting into trouble is greater at th e university as exhibited by the greater numbers of referred students relative to the number of students arrested. Nevertheless, these students may comprise a group of good students that make poor decisions and would normally abstain from criminal activity. If this were the case, the o fficial university reaction may serve to sure up these students resolve in abstaining from deviance and continue being good students in this situation, preand post-differences in performan ce metrics may not significantly change after a student is processed thro ugh the university system. 83

PAGE 84

The sub-group in the referralonly subpopulation that dropped out or were dismissed seem to be consistently bad students who are maladjusted to university life. This maladjustment may be expressed with alcohol and drug use and mi suse. Since early problems tend to be picked up by the university relative to law enforcement, group membership seems to be driven by student culture in that stude nts generally begin offending in dorm rooms and expand their comfort zone, so to speak, as time goes on. Along these lines, this scenario should be identified by the existence of no significant di fferences in performance metrics or worsening performance before and after a referral and its a ffiliated sanctions. These students should show poor performance from when starti ng higher education to the point in which they dropout or are dismissed. Continuing this line of reasoning, arrested stude nts are more likely to venture out into public and overtly break the law by choosing to ac quire and use fake or borrowed ids, gaining access to alcohol or illicit drugs and carrying/using them on city st reets or within cars, and/or by frequenting bars in which they have a better likelihood of drinking underage by some means. On face value, these students seem to be socially well-adjusted as they tend to engage in these activities in groups of friends but their scholastic achieve ment may wane due to their recreational choices and thei r inability to delay gratification. The evidence of impulsivity lies in the high rates of recidivism incurred within th is group and the reduced ability to graduate ontime. Differences in performance metrics should th en be better explained by levels of recidivism rather than the timing of the arrest and sanctioning for th ese individuals. Another outcome of interest concerns trends in grade point averages. The mean GPA differences relative to students resident college over their academic career as well as the difference in these metrics before and after bei ng exclusively processed by either the criminal 84

PAGE 85

justice system or judicial affairs can be examin ed. To make clear comparisons, this assessment separates those students with an experience with both systems. The results are opposite to what would be expected overall when one group has a higher concentration of repeat offenders relative to the other. Indeed, the results are far from what is expected bu t not at first glance. The violators who were solely processed by the judicial affairs system and completed their degree on time tended to have higher grade point aver ages than did those in their resident college over the span of their academic careers (mean di fference in GPA = 0.11). Similarly, arrestees who completed their degrees on time and were processed by the criminal justice system also had higher GPAs than did students generally in thei r respective colleges (mean difference in GPA = 0.12). Those that did not complete their degr ee on time expressed the opposite outcome across both groups: those processed by judici al affairs got worse grades th an did those in their college generally (mean difference in GPA = -0.41) as di d those processed throug h the criminal justice system (mean difference in GPA = -0.31) A more telling pattern emerges when observing the change in grades for the semesters before and after the students were first proce ssed by either system. The calculation of these differences are compiled in this manner: 1) the difference of a students performance relative to average GPA for their resident college(s) is cap tured for each semester, 2) the mean of these differences is calculated for the term(s) before and after a students first referral or arrest, ignoring the term in which the refe rral or arrest occurs, 3) the pr ereferral or arrest mean is subtracted from the postreferral or arrest mean to observe the change in academic performance relative to ones resident college(s) before and after a student first ge ts processed by either system. Any student that was arre sted or referred in the first or last term of attendance was 85

PAGE 86

removed from the analyses that utilized this change in GPA va riable since the preor postchange was not available to analyze. For those in the referral-only group (i.e., proc essed only judicial affa irs), the change for individuals who complete their de gree on time is slightly negative (change in GPA difference = 0.126). For individuals who comp lete their degree on time and are processed by the criminal justice system, the change in grades is slightly positive (change in GPA difference = 0.117). It is interesting to note that the individuals who co mpleted their degree on time and were processed by judicial affairs seemed to be earning higher GPAs relative to their college before being processed by the dean of students office, yet this level of performan ce virtually disappeared afterwards (see Table 4-3). Averaged across th e terms after processing, these students GPAs were on a whole only slightly above their reside nt colleges GPA average. On the other hand, individuals who completed thei r degree on time and were processed by the criminal justice system actually performed better after criminal ju stice processing. These differences seem to be most extreme for students who have not completed their degree on time. For these students that are proce ssed by the criminal justice system, there does not appear to be any changes in GPA differences before and after being arrested and processed (change in GPA difference = -0.036). Yet for st udents who are processed by judicial affairs and fail to complete their degree on time, the outcome appears to be quite bleak. These students average a change in GPA performance equal to a half of a lette r grade (-0.501) below the average GPA of students resident colleges after they are processed by judicial affairs. This impact on GPA occurs despite the observation that more fr equent and serious offending occurs within the group of arrested and not among those solely referre d to judicial affairs. There seems to be something inherent within judicial affairs proc essing that is associated with two extremes: 86

PAGE 87

students who are processed by judicial affairs fo r drug and alcohol offenses tend to graduate on time more frequently than those who are pro cessed by the criminal justice system; however, those who fail to do so experience exasperated decr ements in their academic performance that is just not found with students who also fail to graduate on time but are processed by the criminal justice system. Table 4-3: Changes in the Difference in Indi vidual Student GPA Relative to their Resident College Before and After First Arrest or Referral Group GPA Difference Before GPA Difference After Change Referral-only (N = 332) Complete degree on-time 0.185 0.057 0.128 Defer on-time completion 0.028 0.529 0.501 Dropouts/Dismissals 0.645 1.799 1.154 Arrest (N = 166) Complete degree on-time 0.102 0.175 + 0.073 Defer on-time completion 0.261 0.297 0.036 Mixed Both Arrest and Referral (N = 126) Judicial affairs 1s t processed (Completers) 0.331 0.086 0.245 Judicial affairs 1s t processed (Defer on-time completion) 0.273 0.351 0.078 Criminal justice 1s t processed (Completers) 0.049 0.086 + 0.037 Criminal justice 1s t processed (Defer on-time completion) 0.007 0.305 0.312 These findings are further supported by examin ing changes in GPA performance metrics for students who have experience with both syst ems (mixed group). Extreme caution must be 87

PAGE 88

taken here however, as the sample sizes of th ese groups do not offer mu ch analytical power. Accordingly, only directionality will be emphasized in presenting these results. The change in grades is negative in three of the four rows under Mixed Group in Table 4-3. Only those who are arrested and complete college show a positiv e change in grades after their offense. The next chapter of the present study will be gin to explore these phenomena with greater analytical power to begin to unravel these group differences as well as explain the impact of arrest and punishment on a students academic career and likelihood of being re-arrested or referred again. In particular, the unexpected results from the descriptive analyses included in this chapter will be critically dissected to investigate differential outcomes in the arrest and referral groups while controlling for other factors. The present chapter has identified that student s overall GPA performance relative to the average in their resident colleges over the span of their student careers does not tend to vary across students who are processed by either syst em. This performance metric does, however, vary across students that complete their degree on -time versus those that do not. On the other hand, the change in this measure before and after a students first re ferral or arrest suggests that individuals processed by judicial affairs experience a reduction in their academic performance versus those processed by the criminal justice system which tend to experience either no change or a positive change. These differen ces occur in spite of the fact th at students with an arrest are more likely than those referred to judicial affairs to recidiva te and to commit more serious offenses both of these would predict more se vere punishment as well. Furthermore, since academic performance seems to be closely associ ated with whether a student can manage to finish their undergraduate degree s on-time, it is important to me ntion that a higher concentration 88

PAGE 89

of individuals succeeded when pr ocessed through the university system when ignoring those that drop out or are dismissed for poor academic performance. Those who dropped out or were dismissed ca me almost exclusively from the group of students who were referred only to judicial affairs. A likel y reason for this phenomenon seems to be that these students fail to acclimate to uni versity life from the moment they begin their academic careers. They are referred early in th eir career and exhibit poor academic performance, with their grades getting worse after the referral. Students who are arrested did not drop out and were not dismissed, yet those that do not complete their degree on-time seem to under-perform relataive to their resident college average GPAs ( both before and after their first arrest). These data suggest that students who are jointly processed by both systems have the best outcomes in regards to academic performance, comple ting undergraduate study on-time, and reducing recidivism. This hypothesis will be put to the test with multivariat e analysis in the next chapter. A final note regarding the student s found in these data: athletes are largely absent from the ranks of offending students. Upon conferring with agents of the University Athletic Association, athletes are processed and sancti oned through a separate system and its data were not available to this researcher. It is interesting to remar k, however, that this disciplinary system does not prevent athletes from being arrested. Regardless, it would be unwise to speculate the reason why athletes are underrepresented in these data since the principle investigator is blind to the vast majority of incidents involving these students. This factor is theref ore dropped from further consideration. Summary An overview of the results of the descriptive analysis offers some unique challenges that multivariate modeling may be able to resolve. In particular, it seems that judicial affairs processing and sanctioning may actually be tied with detrimental impacts on academic 89

PAGE 90

90 performance while retaining the pow er to reduce recidivism relati ve criminal justice processing and sanctioning. Unexpectedly, it seems that students who are lik ely to dropout or be dismissed for poor academic performance are almost exclus ively processed/sanctioned by the university system alone. In this circumstance, only multivariate analysis can determine whether the university process or levied sanctioning has any impact on dropout/dismissal independent of other factors. Individuals pro cessed/sanctioned through the university system also seem to more likely to complete their undergradu ate education on time. Recidivism (either a subsequent arrest or referral) seems to be more heavily concentr ated among individuals in the arrest group. It could be that getting into hab itual trouble helps to explain the differences found in on-time completion at the bivariate level instead of the cr iminal justice process or the sanctions students receive from the local circuit court. If this were the case, the analytical models focusing on recidivism as an outcome will determine if higher levels of recidivism are indeed predicted by criminal justice processing and sanctioning wh ile holding all other f actors constant. The following chapter will assess these basic relationships with greater analytical refinement to begin to unravel the complexities identified by th e descriptive analysis presented earlier.

PAGE 91

CHAPTER 5 MULTIVARIATE ANALYSIS OF ARRESTEE AND REFERRAL GROUPS: A FOCUS ON SANCTIONS Research Questions and Strategies The Impact of Different Sanctions A major thrust of this dissertation is to exam ine the impact of different sanctions for those who were arrested or referred for alcohol or drug-related offenses while a university student. Previous scholarship focuses attention on two di fferent features of what might be happening the sanctions and the cultural context of substance use on American college campuses. Some criminological research and theory pl aces an emphasis on the sanctioning itself. Deterrence formulations argue that the sanctions can deter unwanted behavior (which may also increase desirable behavior). From that perspec tive, more severe sanctions can be expected to predict less recidivism and be tter academic performance. On the other hand, labeling perspectives s uggest that sanctioning can backfire. For example, from Braithwaite, the expectation is that more stigmatizing sanctions will lead to subsequent problems, especially if they are not accompanied by reintegrative measures. As guided by this perspective, the working hypothesis is that the academic performance indicators and recidivism of students who receive probation /suspension and/or formal court sanctions and criminal records will be negatively affected relati ve to students who receive informal sanctions levied by the university. Several more specific hypotheses can be advanced regarding the impact on academic performance. Students who receive more stigmatizing sanctions will (as compared with those who receive other sanctions) (1) be less likely to complete a degree in four years (measured dichotomously), (2) be more likely to drop out or be dismi ssed by the university for poor academic performance, (3) have a lower grade point average for course work taken after their arrest (calculated re lative to the college in which a stude nt belongs, as described later), and 91

PAGE 92

(4) have a lower grade point average overall relati ve to their peers (compu tation described later). They will also have a higher risk of recidivism (measured continuously representing the sum of arrests and judicial affairs referrals). The research on substance use on college campu ses raises other issues. The impacts of sanctions may play out differe ntly given the university a nd student cultures surrounding substance use, especially alcohol use. Certainly that literatu re has identified some risk and protective factors that should be considered in the analysis. Thus, educational outcomes and recidivism may not be affected by the sanctions so much as by ot her factors that include social greek membership and sex. A series of OLS and logistic regressions will be performed in this chapter to examine these issues. Dependent Variables Four primary dependent variables have been identified to measure academic performance for analytical purposes: (1) overall GPA, (2) GP A change, (3) dropout/dismissal, and (4) current student status (those who are still students did not complete on time). For these analyses, GPA is measured relative to a students resident college (in which the va st majority of students do not have an arrest or a referral) for each term. GPA change is a calculated difference and will be created for each student in the study population as well as the comparison subset of students who were only referred to judici al affairs and avoided arre st. If students were to change colleges, the calculation will be adjusted to the average GPA for the new college. These differences will be assessed (1) across students entire academic career by taking the mean of these GPA differences for all terms in which the student is enrolled in coursework, and (2) before and after the semester of the students first arrest/referral and averag ed over the previous and remaining semesters, respectively. Any student that was e ither first arrested or referred in their first or last term at the university were removed from consideration from the analyses that utilize this variable since 92

PAGE 93

they lack the requisite preor postmean GPA comparison. Both dropout/dismissal and current student status are measured dichot omously, 1 signifying dropout/dismissal and the status of being a current student respectively. The fi nal dependant variable analyzed in the present study, recidivism, is a count of the of fenses across the span of students academic career as identifie d by this study. Independent and Control Variables The primary independent variables of intere st are the sanctioning alternatives these students receive. To indicate group membership, three dichotomous variables are constructed to separate which students received sanctions from j udicial affairs alone, from circuit court alone, and from both judicial affairs and circuit court. A fourth group, for referenc e, is left out of the analytical models to indicate how these three groups fare re lative to those that were not sanctioned at all (those found not guilty or had th eir cases dropped). Alternatively, the study also examines the type of processing students receive across their academic career for drugand/or alcohol-related offenses. This ignores the impact sanctioning and solely examines the impact of going through the motions required by each system (or both systems combined) to resolve a case against a student. Doing so truly hones in on tr ue value of the sanction applied independent of the process. Three dichotomous variables are co mpiled that identify students solely processed by judicial affairs, solely by the criminal justice system, or processed by both systems at some point in time. Other variables in consideration are compile d as follows: (1) fraternity and sorority membership at any point in students academic careers is indicated dichotomously, with 1 signifying membership, (2) the impact of re-arrest/r e-referral is controlled for by utilizing a count of offenses accumulated by a student, (3) the term of first offense is an integer signifying the term a student first offends, counting only the terms in which students are enrolled, (4) high 93

PAGE 94

school GPA is the raw cumulative weighted GP A, which can exceed 4.00 by factoring in honors or advanced placement/Internationa l Baccalaureate coursework, and (5) enrollment in a five-year undergraduate program, with 1 signifying enrollment Serious offending is taken into account by defining a serious offense as one of the followi ng: driving unde r the influence, felony drug possession / intent to sell, and fe lony false identification / fraud. A count of serious offenses is utilized to control for the impact this type of offending may have on the outcomes. Lastly, gender (1 = male) and race (1 = white, nonhispanic) are measured dichotomously. Introduction to Multivariate Analysis As described in the previous chapter, some unique differences exist between the group of arrested students compared with students who we re only referred to judi cial affairs and were never formally processed through the criminal ju stice system. This ignores a key group of students that were processed by bot h systems. The present set of analyses found in this chapter not only adds analytical value by controlling for variables that ma y be correlated with the various outcomes identified by this study, but it also allows for these student s to be compared with their peers to determine how th eir outcomes may differ. To review, the following outcomes will be assessed using both ordinary least squares (OLS) regression (for continuous dependent variable outcomes) a nd logistic regression (for dichotomous dependent variable outcomes): co mpleting undergraduate study on-time (current student (1) / graduate (0)), dropout or dismissal from the univers ity (yes (1) / no (0)), overall GPA differences relative to students resident college averaged over their academic careers (continuous), change in mean GPA difference fo r the terms before and after students first arrest/referral (continuous), a nd recidivism (continuous, zero-censored). Of these outcomes, it seems the most imperative to sort out on-time undergraduate completion as it was determined in 94

PAGE 95

the previous chapter that this correlated with di fferences in academic performance as well as the change in academic performance at first intervention. Predicting On-Time Completion Logistic regression was employed to evaluate th e available variables as possible predictors of graduating on-time with an undergraduate de gree from the university among drug and alcohol offenders. Most important for the present study, differences in the system in which students are processed for their offenses (whether through th e criminal justice syst em or the university system) are tested to determine if this pro cessing has any impacts on degree completion while controlling for other variables. The results of this analysis can be found in Table 5-1. Each of the models presented utilizes the full sample of students while removing students with missing data, or more importantly, students who have dropped out or were dismissed from the university for poor academic performance. Obviously the latter gr oup should not be assessed in estimates of ontime completion as they will neve r achieve this outcome. Dismissals and dropouts will be handled separately in a forthcoming set of anal yses. The full model (model 5) is impacted by case deletion due to missing data since high school GPA is not available for all students in the study. A very small subset of students (and/or th eir parents) choose altern atives to traditional secondary education in which this metric e ither does not exist or goes unreported to the university. For example, student s that are home schooled or atte nd some charter schools may fit into this category. The variables that significantly predict on-tim e graduation seem to remain statistically significant across the partial mode ls (models 1 4) through the full model. The partial models signify how a particular type of official proc essing for drug and alcohol offenders (relative to 95

PAGE 96

students processed in a different manner) impact s students likelihood of being a current student in Fall 2008, one full semester after these students should have graduated. Table 5-1: Logistic Regression Estimates of the Effect of Official Processing on On-Time Graduation1 Independent Variables 1 N = 570 2 N = 570 3 N = 570 4 N = 570 5 N = 550 Race 0.89 0.92 0.90 0.89 0.95 Gender 2.69*** 2.54*** 2.44*** 2.68*** 2.57*** IB or AP Student 0.51*** 0.51*** 0.51*** 0.51** 0.55** Greek Affiliation 1.06 1.05 1.08 1.07 1.01 Term of First Offense 1.08 1.10* 1.13** 1.08 1.07 Number of Offenses 0.91 1.13 0.93 0.87 0.88 Sum of Serious Offenses 1.40 1.36 1.43 1.41 1.21 Five-Year Program 5.11*** 5.23*** 5.15*** 5.09*** 5.33*** Judicial Affairs Only 0.51*** Reference Reference Arrest Only 1.51 1.86** 1.87** Both Judicial and Arrest 1.64 2.21** 1.96* High School GPA 0.75 R2 0.17 0.16 0.16 0.17 0.18 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. There are three types of official processing asse ssed by this analysis: 1) judicial affairs only, comprising of students solely referred to judici al affairs whether on ce or several times by authorities other than any police agency, ther eby avoiding any contact with and processing by the criminal justice system, 2) arrest only, comprising of students so lely arrested by law enforcement agencies other than the university police department whether once or several times and subsequently processed by the local criminal justice system and 3) both judicial 96

PAGE 97

affairs and arrest, comprising of students that were either arrested by the university police department, whose practice is to automatically refe r these students to judicial affairs, or are referred to judicial affairs by another authority and arrested by an agency other than the university police department independently with no regard to the or der of these events. According to the results of this analysis ones gender, having previous advanced placement or International Baccala ureate coursework, and the type of official processing one receives have a statistically significant impacts on a students likelihood of being a current student during Fall 2008 while controlling for other variables. In the study sample, males are 2.69 times more likely to be current students, ad vance placement or International Baccalaureate students are about half (0.51) as likely to be current students, and both those processed by the criminal justice system alone or processed by the criminal justice and university system jointly at some point in their academic career are about two times as likely to be current students relative to students that were processed by judicial affairs alone. Importa ntly, these models control for students that are enroll ed in programs that are designated as 5-year undergraduate programs. The likelihood of these students to be enrolled in Fall 2008 is almost absolute, as demarcated by the magnitude of the effect size of this vari able (odds ratio = 5.33, p < 0.001). The absence of statistical significance for race, fr aternity and sorority membership, term of first offense, number of offenses, number of serious offenses, and high school GPA is also revealing. Controlling for all other variables, these factors do not seem to predict current st udent status. This should not come as a surprise, however, as these variables we re not significantly correlated to the dependent variable at the bivariate level in the first place. The next approach taken in the current an alysis focuses on the impact of the first intervention a student receives on cu rrent student status rather than the overall type of official 97

PAGE 98

processing a student receives as a result their offending over their academic career. That is, relative to receiving no punishment at all, the following analysis examines the impact of which authority (university or criminal justice) is dol ing out the punishment as well as what level of punishment (mild or severe) is given. In these data, level of punishment can easily be separated into mild and severe. In the university system, the standard mild or light punishment consists of a written reprimand by the dean of students in conjunction with a subset of optional disciplinary devices such as an drugs and alcohol education cour se (online; popularly assigned), community service (rarely assign ed), counseling (rarely assigned ), and a written assignment (rarely assigned). More severe punishment s consist of a conduct probation or semester suspension in conjunction with more liberal use of the disciplinary devices listed previously. For the criminal justice system, the standard mild or light punishment consists of a deferred prosecution in conjunction with a fine or community service. More severe punishments consist of unsupervised probation, fines, and community service (which may also serve to reduce the amount fined, but not to zero). The result of this analysis is remarkably similar to the previous one specifying the impact of official processing on current student status. The advance in knowledge the present model gives is the finding that receiving a mild punishme nt for ones first offense by either judicial affairs or the criminal justice system predicts a reduction in the odds of being a current student in Fall 2008. This impact can be calculated by observ ing the odds ratio of the effect of punishment adjusting for the odds ratio of the effect of the level of punishment. In general, it appears that the reduction in th e odds of being a current student as predicted by receiving a mild punishment for ones first of fense is greater for those solely punished by judicial affairs relative to thos e punished solely by the criminal justice system. The scenarios 98

PAGE 99

that predict the highest likelihood of current st udent status remain w ith students who do not receive any punishment for their first offense and with students who receive severe punishments, particularly those punished by the criminal justice system. Table 5-2: Logistic Regression Estimates of th e Effect of Official Intervention on On-Time Graduation1 (N = 570) Independent Variables Exp(B) Race 0.95 Gender 2.48*** IB or AP Student 0.49*** Greek Affiliation 1.09 Term of First Offense 1.07 Number of Offenses 1.02 Sum of Serious Offenses 1.21 Five-Year Program 5.44*** No Punishment (First Offense) Reference2 Judicial Affairs Punishment Only (First Offense) 1.28*** Criminal Justice Punishment Only (First Offense) 1.73* Joint Punishment (First Offense) 1.81 Mild Punishment (First Offense) 0.49* Severe Punishment (First Offense) 0.86 R2 0.17 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. 2 Serves as a reference category for both level of punishment (e.g., mild or severe) as well as type of punishment (e.g., judicial a ffairs or criminal justice) On the other hand, individuals who receive both a judicial affairs and criminal justice punishment are not predicted to experience an in crease in the odds of being a current student (as 99

PAGE 100

noted by statistical insi gnificance) while contro lling for other factors ev en though its odds ratio was 1.81. These students may receive a combination of mild and severe punishment that work in opposite directions, inflating th e standard error estimate and render the relationship to be statistically insignificant. Predicting Dropout or Dismissal Of the outcomes under review by the current study, dropping out or dismissal for poor academic performance is the paramount worst outcome for the students in the sample. As presented in the descriptive chapter, it appears that this outcome predominantly appears in the portion of the sample that was solely referred to j udicial affairs for drugand/or alcoholrelated offenses. For this reason, the next set of anal yses focuses on this subsample to predict this outcome while removing cases with missing data due tied to the high school GPA variable (as described earlier) when this va riable is applied. Logistic regression was employed to produce these results which can be found in Table 5-3. Without any consideration of an intervention by judicial affairs (models 1 and 2), a students race, fraternity or so rority membership, number of o ffenses, and number of serious offenses all have an initial impact (model 1) on the likelihood of dropping out or dismissal. When high school GPA is taken into consideratio n (model 2), however, this predictor eliminates the effect of race and number of offenses. Th e resulting model can be interpreted as follows: while controlling for other factors, students belonging to a fraternity or sorority are about half as likely to drop out or be dismissed (odds ratio = 0.46, p < 0.05), those with a serious offense are more likely (3.51 times per each serious offense; odds ratio = 3.51, p < 0.01) to drop out or be dismissed, and those with a lower high school GP A are more likely (4.35 times per 1 unit drop in high school GPA; odds ratio = 1 / 0.23, p < 0.01) to drop out or be dismissed from the university. 100

PAGE 101

Table 5-3: Logistic Regression Estimates of the Effect of Official In tervention on Dropout or Dismissal1 Independent Variables Model 1 N = 329 Model 2 N = 318 Model 3 N = 329 Model 4 N = 318 Race 0.50 0.60 0.49 0.60 Gender 1.90 1.77 1.88 1.77 IB or AP Student 0.65 0.82 0.65 0.82 Greek Affiliation 0.52 0.46* 0.54 0.47* Term of First Offense 0.94 0.94 0.91 0.93 Number of Offenses 1.85* 1.64 1.85* 1.62 Sum of Serious Offenses 3.54** 3.51** 3.56** 3.63** Mild Punishment (First Offense) 0.64 0.76 Severe Punishment (First Offense) 0.77 0.76 High School GPA 0.23** 0.24** R2 0.12 0.18 0.13 0.18 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The impact of the first intervention by judicial affairs is offered in models 3 and 4 of Table 5.3. These models are virtually identical as the previous two, indicating that any action taken by judicial affairs including non-action does not even the slight est impact on the likelihood of dropout or dismissal when controlling for other factors. Poorer high school academic performance and increased amount of serious offending during ones ac ademic career alone predict dropout or dismissal while membership in a fraternity or sorority can mitigate ones likelihood of this outcome. Predicting Recidivism Recall from previous discussion that many st udents in the present study reoffend at some point in their academic career, whether this mean s being re-arrested, referred back to judicial 101

PAGE 102

affairs, or crossing over to the sister system rath er than revisiting the one that first processed the student. The following set of analyses focus on th e predictors of recidivism, particularly how judicial affairs processing compares with cr iminal justice system processing in reducing recidivism. Additionally, these analyses examine th e impact of the intervention (or lack thereof) each system applies to an offenders first viol ation either in isolation or in tandem while controlling for other variables. Linear (ordin ary least squares) regression was employed to accomplish these goals and these results are found in Tables 5-4, 5-5, and 5-6.21 The initial examination into predicting reci divism reveals that ge nder, having previous advanced placement or Interna tional Baccalaureate coursework, fraternity or sorority membership, the term of a students first offense, and the type of official processing one receives has a significant impact on recidivism. However, when factoring in high school GPA, advanced placement or International Baccalaureate status loses its significance. The remaining results can be interpreted as follows: males and fraternity brothers and sorority si sters are predicted to reoffend to a greater extent as we ll as students that o ffend earlier in their academic career and/or have a lower high school GPA while controlling fo r other factors. The type of official processing, on the other hand, is more difficult to interpret. On face va lue, it appears that students who have an arrest at any point in th eir academic career relative to students processed solely by judicial affairs are more likely to reoffend, especially when processed by both systems whether jointly or separately at some point durin g the study period. However, the very fact that this latter group contains indi viduals who are independently pr ocessed by both systems thereby biases this group by absorbing the wider samples recidivists, skewing the results. 21 Alternatively, both Poisson and negative binomial regressi on were applied to this research question since the distribution of the dependent variable was skewed favori ng less recidivism. The data favorably fitted linear regression compared to these alternative approaches as determined by post-hoc goodness of fit tests and log likelihood estimates where applicable. 102

PAGE 103

Table 5-4: Ordinary Least Squares Estimates of the Effect of Official Processing on Recidivism1 Independent Variables 1 (N = 620) 2 (N = 599) Race 0.02 0.02 Gender 0.13*** 0.11*** IB or AP Student 0.06* 0.04 Greek Affiliation 0.05 0.05 Term of First Offense 0.16*** 0.17*** Dropout or Dismissal 0.03 0.02 Current Student 0.01 0.01 Five-Year Program 0.04 0.03 Judicial Affairs Only Reference Reference Arrest Only 0.13*** 0.13*** Both Judicial and Arrest 0.69*** 0.69*** High School GPA 0.01** R2 0.49 0.50 1 Standardized Beta values reporte d. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The subsequent analyses take this bias into consideration by utilizing a split-model approach students who are proce ssed solely by either judicial affairs or the criminal justice system are analyzed separately from students who have contact with both systems at some point in their academic careers. In Table 5-5, models 1 and 2 focus on the impact of the first official intervention (or lack thereof) on recidivism for students who have had contact with both systems while models 3 and 4 examine this impact for those solely processed by either judicial affairs or the criminal justice system. 103

PAGE 104

Table 5-5: Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism1 Independent Variables 1 (N = 124) 2 (N = 113) 3 (N = 496) 4 (N = 477) Race 0.03 0.05 0.02 0.02 Gender 0.35*** 0.31*** 0.05 0.04 IB or AP Student 0.12 0.05 0.01 0.01 Greek Affiliation 0.07 0.06 0.04 0.03 Term of First Offense 0.15 0.15 0.16** 0.19*** Received No Punishment (First Offense) Removed Removed Reference2 Reference2 Received Mild Punishment (First Offense) Reference Reference 0.27*** 0.30*** Received Severe Punishment (First Offense) 0.15 0.23* 0.14** 0.19*** Dropout or Dismissal Constant Constant 0.002 0.003 Current Student 0.01 0.01 0.02 0.003 Five-Year Program 0.16 0.16 0.02 0.02 Judicial Affairs Punishment Only (First Offense) Reference Reference 0.60*** 0.61*** Criminal Justice Punishment Only (First Offense) 0.15 0.13 0.23*** 0.26*** Joint Punishment (First Offense) 0.12 0.13 High School GPA 0.25** 0.03 R2 0.23 0.28 0.36 0.39 1 Standardized Beta values reporte d. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. 2 Serves as a reference category for both level of punishment (e.g., mild or severe) as well as type of punishment (e.g., judicial a ffairs or criminal justice) Model 1 reveals that students who receive their first interven tion from the criminal justice or jointly from both systems actually tend to re offend less than those di sciplined by judicial 104

PAGE 105

affairs, especially when receiving a severe punishment while controlling for other factors.22 When placing high school GPA into the analysis, as shown in model 2, the differential impacts of the authority that doles out th e intervention lose their statistic al significance but the level of sanctioning remains significant. Yet given the sample size (N = 113) and the degrees of freedom used to specify this model, it is interesting to note that the direction and magnitude (favoring a reduction in recidivism) of the impact of thes e interventions relative to judicial affairs intervention remain similar and are approaching si gnificance at the p = 0.10 level. What remains important in predicting recidivism is similar to what was previously found and can be interpreted as follows: males are predicted to reoffend to a greater extent as well as students that offend earlier in their academic career and/or have a lower high school GPA while controlling for other factors. Fraternity and sorority membership does not seem to carry over as a predictor of recidivism when taking the first intervention a student receives into consideration. Another difference found is that individual s who are enrolled in five-year undergraduate programs tend to be less likely to reoffend after respecifying the an alytical models to focus on intervention rather than official process. Lastly, as alluded to earlier, a reduction in the level of recidivism is predicted when a student receives a severe punis hment at their first intervention relative to receiving a mild punishment. For individuals solely processed by judicial affairs or the criminal justice system separately (models 3 and 4), notable differences exist wh en predicting recidivism What seems most important is the statistical insignificance of both gender a nd high school GPA. The other divergence in these models is the insignifican ce of enrollment in a five-year undergraduate program, but this comes as a lesser surprise since this variable is insignificant in model 1. What 22 In order for this model to compile properly in the sta tistical package used for anal ysis, students who received no punishment (N = 4) had to be dropped from this analytical model. 105

PAGE 106

remains are models that mirror those for students who have had contact with both judicial affairs and the criminal justice system. There are enough cases receiving no punish ment that they could constitute the reference category fo r models 3 and 4. In regards to the system doling out the first intervention (relative to no intervention), the cr iminal justice system seems neither to aggravate nor mitigate the levels of recidivism when it intervenes while judicial affairs intervention seems to increase the level of recidivism (regardless of the level of punishment). In fact, it appears that when the university system intervenes, students ar e likely to reoffend at least once more holding all other factors constant. The last point to cons ider is that the term in which a student first offends retains its significance in predicting reci divism those who offend earlier tend to experience increased levels of recidivism. Taking this investigation one step further, th e next set of analyses takes into account whether students had a juvenile ar rest record with the Florida Depa rtment of Juvenile Justice in that prior arrest has been found to be one of th e most robust predictors of future offending in criminological literature. These da ta are only available for students with an adult arrest record. Thus these analytical models utilize the subsam ple of students that have had contact with both judicial affairs and the criminal justice syst em. These results are found in Table 5-6. The results of this analysis are difficult to inte rpret. It seems that juvenile arrest does not predict adult recidivism when controlling for other factors but does at the bivariate level. Looking deeper into this issue, th e number of juvenile offenders (N = 10) in this subsample is not optimal to detect effect size as a result of j uvenile arrest. Neverthele ss, the significance level of this predictor does approach the p = 0.10 level, suggesting th at a larger sample size may resolve this issue. 106

PAGE 107

Table 5-6: Ordinary Least Squares Estimates of the Effect of Official In tervention on Recidivism for Arrestees While Factoring in Juvenile Arrest1 Independent Variables 1 N = 117 2 N = 107 Race 0.01 0.03 Gender 0.33*** 0.30*** IB or AP Student 0.05 0.02 Greek Affiliation 0.06 0.13 Term of First Offense 0.14 0.18 Received No Punishment (First Offense) Removed Removed Received Mild Punishment (First Offense) Reference Reference Received Severe Punishment (First Offense) 0.18* 0.20* Dropout or Dismissal Constant Constant Current Student 0.02 0.07 Five-Year Program 0.16 0.17 Judicial Affairs Punishment Only (First Offense) Reference Reference Criminal Justice Punishment Only (First Offense) 0.12 0.16 Joint Punishment (First Offense) 0.10 0.16 High School GPA 0.21* 0.13 Juvenile Arrest 0.10 Removed R2 0.27 0.24 1 Standardized Beta values reporte d. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. Additional evidence does exist in favor of this interpretation of thes e data: when removing students with a prior juvenile ar rest record as displayed in model 2 of Table 5.6, the results 107

PAGE 108

mirror that of model 1 in Table 5-5, high sc hool GPA loses its statistical significance and criminal justice and joint-system intervention (rela tive to judicial affairs) regain their mitigating influence on recidivism. The correlation of juve nile arrest and high school GPA seem to mask the impact of first intervention. This may also be interpreted with not able caution that students with a juvenile arrest r ecord (which enter the university with a lower high school GPA) may be resistant to the mitigating effects of the criminal justice system or joint-system intervention in regards to reducing levels of recidivism. At this point, it is difficult to disentangle the independent effects of juvenile arrest and high school GPA given the rarity of juvenile arrest in these data. However, juvenile arrest has the potential to be a particularly useful predictor of offending (and reoffending) in colle ge student samples given that these effects were felt when only ten juvenile arrestees were co ntrolled for in this analysis. Predicting Overall Mean GPA Differences a nd Change in Mean GPA Differences Before and After First Intervention The next set of analyses examines the impact of type of official processing as well as the first intervention (if any) appl ied to students on two outcomes: 1) the difference of students GPA relative to their resident colleges mean GP A averaged across the span of their academic careers, and 2) the difference in this metric for the terms before and after a students first offense. It should be noted here that th e sample size is reduced when i nvestigating the second outcome as some students were arrested/referred in their firs t semester of attendance, thus lacking available data for the change computation. These mode ls also remove students who dropout or are dismissed for poor academic performance for the same reasons as previously discussed. The results of these analyses are found in Table 5-7. 108

PAGE 109

Table 5-7: Ordinary Least Squares Estimates of the Effect of Of ficial Processing and Intervention on Overall GPA Difference (Models 1 & 2) and Change in GPA Difference Before and After Fi rst Intervention (Models 3 & 4) 1 Independent Variables 1 (N = 548) 2 (N = 548) 3 (N = 356) 4 (N = 356) Race 0.09** 0.09** 0.03 0.03 Gender 0.01 0.01 0.08 0.09 IB or AP Student 0.03 0.03 0.03 0.04 Greek Affiliation 0.07 0.07 0.04 0.04 Term of First Offense 0.07 0.06 0.15** 0.14** Number of Offenses 0.10* 0.10 0.34 0.01 Received No Punishment (First Offense) Reference Reference2 Reference Reference2 Received Mild Punishment (First Offense) 0.01 0.04 0.03 0.13 Received Severe Punishment (First Offense) 0.03 0.05 0.08 0.01 Current Student 0.46*** 0.46*** 0.14** 0.14** Five-Year Program 0.17*** 0.17*** 0.02 0.02 Judicial Affairs Only Reference Reference Arrest Only 0.07 0.19** Both Judicial and Arrest 0.09 0.12 Judicial Affairs Punishment Only (First Offense) 0.05 0.06 Criminal Justice Punishment Only (First Offense) 0.06 0.22*** Joint Punishment (First Offense) 0.07 0.03 High School GPA 0.25*** 0.25*** 0.10 0.11 R2 0.32 0.34 0.12 0.09 1 Unstandardized Beta values reported. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. 2 Serves as a reference category for both level of punishment (e.g., mild or severe) as well as type of punishment (e.g., judicial a ffairs or criminal justice) 109

PAGE 110

Models 1 and 2 focus on grades over the course of an academic In regards to overall GPA difference, race, term of first offense, student s enrolled in a five-year undergraduate program, and high school GPA have independent impacts on this outcome while holding all others constant and while controlling for current student status. In part icular, whites, students enrolled in a five-year undergraduate program, and those with better high school GPAs tend to perform better than their resident colle ges on a whole across their academ ic career. Offending earlier in ones academic career, on the othe r hand, predicts a reduction in overall academic performance. In regards to the official pro cessing students experience due to their drug and alcohol offenses (model 1), it seems that criminal justice or join t-system processing offers a slight improvement in overall GPA differences relative to those proce ssed by judicial affairs al one. In regards to the first intervention (or lack thereof) students receive (model 2), neither criminal justice nor jointsystem intervention seem to have an impact on this outcome. This finding also stands when examining the level of punishment doled out by th ese authorities. Regardless of who does the punishing and the severity of punishment applied to students, overall academic performance seems unfettered. The change in this metric centered at term of first offense offers some challenges. In particular, this analysis finds that criminal justice processing predicts a positive change in academic performance after interv ention relative to judicial affairs processing (model 3). This effect carries over to the actual intervention applie d by the criminal justice system for a students first offense without regard to the severity of the punishment issued th e intervention seemed to have a positive impact on students academic performance across subsequent semesters (model 4). Of particular interest, thes e analyses find that students who scholastically performed better in high school are more negatively impacted and exhibit lower academic achievement in the 110

PAGE 111

semesters after their first offense relative to students with lower high school GPAs holding all other factors constant. For this outcome, race does not retain its st atistically significant impact, meaning that whites and non-whites do not experience differe nces in the change in their academic performance in the terms after their first offense. However, it seems that white students with a drug and/or alcohol offense slightly perform be tter over their academic careers relative to the non-whites in the sample. Likewise, students th at are enrolled in a five-year undergraduate program in the sample perform better over their academic careers than those who are not, but do not experience any differences in the change in academic performance in the terms after first offense. The number of arrests or referrals also works in the same manner lower numbers predict better overall performance, but no differen ce in the change in performance. However, the logic is reversed for individuals that offend earlier in their academic career: earlier offending predicts a reduction in the cha nge of academic performance in the terms following the first offense but fails to reliably predict a re duction in overall academic performance. Summary The key findings of the present study are as fo llows. First, in conjunction with mild punishment for the first offense, students sanctio ned by the university system alone express an increase in odds of completing their undergra duate education on-time re lative to students who were sanctioned by the criminal justice system alone, specifically when mild sanctions are applied. Students who received no punishment or severe punishments, particularly when levied by the criminal justice system, express a decreas e in their odds of ontime graduation. While holding all other factors constant (in particular, enrollment in a 5-year degree program and entering the university with college credit), it seems that being male is the strongest predictor of deferring on-time graduation. Diffe rential processing and sanction serve to either exasperate or 111

PAGE 112

mitigate the likelihood of on-time graduation for male s as it is very rare to find females with serious offenses (that elicit severe punishment). Second, dropout or dismissal for poor academic performance is not evidenced to be related to judicial affairs processing or sanctioning. In stead, high school weight ed GPA and number of serious offenses seem to be the strongest predictors of this outcome. Individuals with lower high school weighted GPAs and with higher levels of serious offending in college seem to be most likely to dropout or be dismissed by the university. Of particular interest, fraternity and sorority membership decreases the odds of dropping out or being dismissed in ha lf while holding all other factors constant. Third, split-model OLS regression revealed th at increases in recidivism are tied to university processing and sanctioning, regardless of the level of sanction (mild or severe) applied. Interestingly, this anal ysis also detected that males are substantially more likely to offend again cross-system; females, on the other hand, seem to have an equal offending pattern to males, but tend to only offend in one system (either exclusively on-campus and not detected by university police or exclusivel y in the greater city). Fourth, regardless to the level of sanctioning (mild or severe) applied, students that were sanctioned by the criminal justice system alone tended to experience a positive change in their GPA performance relative to their resident college average GPA after their first arrest relative to those sanctioned solely by the university system. Furthermor e, students who entered the university with better high school weighted GP As expressed a marked decrease in the GPA performance relative to their resident college av erage GPA after their first offense the better the high school GPA, the larger decrease in college academic performance gauged by GPA performance after the semester the offense occurs Overall academic performance over the span 112

PAGE 113

113 of students college experience does not seem to be affected by the authority that levies the sanction nor the level of sanctioning (mild or se vere) applied. High school GPA, on the other hand, plays a critical role in predicting academic performance at the university. This suggests that being processed for drugand alcoholrelate d offenses levels the playing field, so to speak, such that stellar high school students re gress to the mean academic performance while those who enter the university on the lower end of the academic performance spectrum seem to remain close to the mean over the span of their college careers. It seems that criminal justice sanctions may mitigate this decrement in performance for all students relative to receiving no sanctions, university sanctions, or sanctions from both the criminal justice and university systems. The following chapter will expand on these findi ngs by utilizing the theoretical framework presented earlier. Does the university disciplin ary system follow Braithwaites reintegrative shaming? In general, does either labeling or deterrence perspective fit these results? How do they speak to the culture of ambivalence? More importantly, these results will be interpreted in a manner that can inform policy to optimize the best outcomes for college students that experience an arrest or referral in the future.

PAGE 114

CHAPTER 6 DISCUSSION OF FINDINGS, POLICY IMPLICATIONS, AND CONCLUSIONS Introduction Arriving back full circle to a remark made by a prominent criminology within a comprehensive review of official intervention and its efficacy presented to the United States Congress in 1997, the present study finds that th e effects of police [or the formal criminal justice system] on crime are complex, and often surprising (Sherman et al., 1997: 8-39). For example, the college students in this study experienced no substantial differences in the likelihood of re-offending subsequent to their first arrest regardless of whether they received an additional criminal justice sanction or not. If th e university were to handle these cases, however, it seems that students actually s eem to experience an increased likelihood of re-offending relative to receiving no punishment alone (e.g., only experiencing the process of being referred to the university). There is an exception to this tren d: male students proces sed by both the university and criminal justice system at so me point in their academic career s are substantially more likely to re-offend unless sanctioned by the criminal ju stice system or jointly by the university and criminal justice systems. Female students who were processed by both the university and the criminal justice system and received an additional formal sanction actually were less likely to reoffend than those who were not sanctioned. Whether the goal of the intervention of an arres t/referral and its subse quent sanctions is to be therapeutic or to prevent future offending (or both), the results of this study suggest that the local authorities may be missing the mark. In some ways, particular interventions may be making matters worse. A cultural ambivalence to alcohol and alcohol problems may be intervening here to some degree so th e discussion turns to explaining why. 114

PAGE 115

Labeling, Deterrence, and Modeling Success and Failure The theoretical perspectives presented to frame this study predict different outcomes regarding the impact of sancti ons on offending students. None predict the outcomes found in these data. For example, the extension of labeling found in Bernburg and Krohn (2003) and Sweeten (2006) lead to expectati ons that students arrested and pr ocessed by the criminal justice system will be more likely to re-offend and drop out of school, particularly if students lack the shielding provided by a higher positi on in our social structure. Th is adaptation of labeling theory does not seem to apply to college students arrested for drugand alcoholrelated offenses. In the study population, being arrested and sanctioned by the criminal justice system had no substantial impact on future offending. In regards to educati onal attainment (which w ill lead to employment opportunities), being arrested and sanctioned by the criminal just ice system does not seem to harm students chances of comp leting their degree and in some circumstances may improve students academic performance (as measured by a positive change in GPA relative to a students wider college across the terms after a his/her firs t offense). Additionally, the social structural component proffered by Bernburg and Krohn (2003) and Sweeten (2006) does not seem to play a role in either recidivism or educational attain ment. If one assumes that a college students disadvantage in the social structure incurre d upon matriculation improves over time (e.g., students regain social support ne tworks, identify private housing and move off-campus, discover local resources and volunteering opportunities), this model woul d suggest early offending would increase recidivism and reduce educational atta inment. While early offending is tied to reoffending in these data, educationa l attainment seems to be part icularly impacted when students are processed and sanctioned later on in their ac ademic careers. Perhaps offending early should not be framed in this manner at all it could be that these individuals just have more time to be caught and captured in these data. However, th ese concepts were not directly assessed by the 115

PAGE 116

present study; future study shoul d examine social structural di sadvantage in the transition to higher education to make light of these issues. Alternatively, Braithwaites re integrative shaming perspec tive (1989; 1992) suggests the potential importance of differences across the authorities that proc ess and sanction students. The working hypothesis of the current study is that if college students are processed and sanctioned through a system that is inherently less stigmatiz ing by nature (e.g., the univ ersity system), there should be a marked reduction in the levels of reoffending and an increase in academic performance. A qualitative assessment of the judicial affairs system at the university system reveals that this disciplinary system appears to have reintegrative qualities: the process is private (privileged record compared to public record) and informal (meeting with the dean of students or an agent of her office compared to public hearin g). Yet, rather than experiencing a therapeutic effect, students that were pro cessed and sanctioned through the j udicial affairs system actually experienced amplification in recidivism while not benefiting from the predicted improvement in academic performance when processed by the criminal justice system (regardless of sanction). Could it be that the university system applies sanctions in a stigmatizing nature? This is doubtful. The standard practice for students processed for minor offenses tend to include meeting with the dean of students or an agent of judicial affairs, discussing the event, receiving a written letter of reprimand, and attendi ng an online alcohol education cour se. Recidivists (known to the university system), driving under the influence offenders, and serious drug or alcohol offenders (e.g., intoxication to the point of requiring medical attention, posse ssion of drugs with intent to sell, felony false identification), on the other hand, receive a conduct probation or suspension, but the rest of the process is much the same. They may attend an online alcohol education 116

PAGE 117

course, perform community serv ice, pay restitution for any da mages done, and/or receive counseling. The standard crim inal justice processing for minor offenses (e.g., underage possession of alcohol) typically come s in the form of a letter from the state attorneys office for first offenses describing conditions of deferred prosecution. The term s are usually a small fine or community service and a clean record for the subse quent six months one year in exchange for a nolle prosequi At times, students choose to attend a misdemeanor court hearing, which is batchprocessed and becomes a part of the public record. More serious offenders (e.g., drunk drivers) almost always spend the night in jail and are most likely senten ced to probation at a public hearing in either misdemeanor or felony courts. If these data were to pick up on any problematic outcomes due to stigmatization, certainly criminal justice processing and severe sanctions would be the most likely candidates for poor outcomes. Labeling and reintegrative shaming cannot explain the results found by the current study well: students who received what would be considered the most stigmatizing of processing and sanctioning those who spent the night in jail, had to post bail/bond, were required to have a public hearing, and who ended up with a criminal record were just as likely to re-offend as those who received mild or no criminal justice punishment but less likely to re-offend than th ose processed by the university system. Those who were arrested and formally sanction ed in the criminal justice system also showed marked academic performance im provements compared with those whose referrals were dealt with only in the university system. Reintegrative shaming does not fit these data well; interpreting the effects of a culture of ambivalence on potentially stigmatizing processes and sanctions is difficult. Insomuch that certain crimes like driving under the influence are substantially stigmatized in society (see Grasmick, Bursik, and Arneklev, 1993), the culture of ambivalence should only impact the potential stigmatization of petty dr ug and alcohol offenses. This research finding runs counter to the hypothesis that stigma predicts offense amp lification and academic performance decrements. The deterrence framework does not seem fi t these data either. Students who have direct experience in punishment avoidance (e.g., arrested but not receiving sanctions) 117

PAGE 118

were as likely to re-offend as those who were processed and received any level of sanction (e.g., mild or severe) from the criminal justice system Furthermore, the students who were sanctioned by the university system experienced amplificationa finding that is completely out of line with deterrence. Piquero and Paternoster (1998) discovered that punishment and punishment avoidance is the strongest predictor of future intentions of wrongdoing (in this ca se, drinking and driving) when taking moral beliefs out of the equation among a sample of drivers across the na tion. In these data, neither punishment nor punishment avoidance seem to make a difference. The culture of ambivalence seems to be at pl ay for these students. Sanctions levied by either criminal justice or university system fail to reduce levels of recidivism while sometimes backfiring and increasing the likelihood of re-o ffending. Yet, stigmatization does not seem to play a role in determining when amplification occurs. The root of this problem remains a mystery. It could be that the perception of the university sanc tioning system is not taken as seriously as the criminal justice system; but even if it were on an equal plane, making improvement in reducing recidivism seems to be out of reach. A silver lin ing to this conundrum is that marked improvements in academic performance occurred for offenders processed through the criminal justice system regardless of sanction. Disciplinary Systems Imbedded in a Culture of Ambivalence First and foremost, the present study detects a substantial amount of drugand alcoholrelated offending among the college population. Across the years of study (2003 2007), almost 5,000 students were arrested many more than once for any criminal offense. Of this amount, the vast majority of student offenders were ci ted for drugand alcoholrelated offenses. The sheer amount of offending and the amount of re-offending, in particular, strongly suggest 118

PAGE 119

nonchalant attitudes towards these types of criminality. When cri tically analyzed, the criminal justice and university disciplinary systems seem to be impotent in preventing future offending. In fact, the judicial affairs system at the university seems to be amplifying a students likelihood of re-offending regardless of whet her or not a student receives punishment for their offending over and above the official processing. Another point embracing this pe rspective is the finding that, regardless of recidivism, the vast majority of the students in these data are poised to graduate or have successfully graduated. For them, there is little evidence of a delay in graduation or a disruption in their academic careers. When student offenders experience delays in completing their degree (most likely male students who did not enter the univer sity with college credit), the de lays are not out of step with the wider university aver ages in the timing of graduation. However, a minority of students do experience detrimental changes in their academic performance af ter official processing and sanctioning. These differences seem to shift th ese students to regress to the mean academic performance for their colleges ove r the span of their academic car eer: a students weighted high school GPA predicted his/her overa ll performance at the university as well as the change in academic performance after his/her first offense. These two functions were reflexive: the better a students high school grades, the better he/she pe rformed at the university overall but the worse he/she performed in the terms subsequent to first arrest or referral relative to previous terms. Males with lower weighted high school GPAs wh o did not enter the univ ersity with college credit are at the highest risk of experiencing a drop in academic performance after a first offense as well as poorer overall performance. Notably, it seems like students are very succes sful in continuing their substance using behaviors despite intervention while co ntinuing on an academic performance trajectory that is correlated to independent factors such as high school grades and gender. Evidence exists that supports the no tion that students are likely to re-offend despite negative consequences (experiencing a drop in their university GPA and being sanctioned for their first offenses). 119

PAGE 120

If the negative consequences of offici al intervention and academic performance decrements fail to reduce the likelihood of re-o ffending, the premise of a cultural ambivalence seems difficult to refute. Additional evidence ex ists when examining the arrest records of the non-student community members. When honing in on alcohol arrests (in which an alcohol violation was the primary offense), non-students are arrested at the sa me rate as students when controlling for age. Thus, college students are not particularly problematic in relation to the wider population when observing official statis tics; the extant literature on student problems may not examine simila r problems of non-students. The college students analyzed in this study ma y have distinctive alc oholand drugrelated problems; however, 18-24 year olds in the comm unity seem to resemble university students when examining alcohol-related o fficial statistics. While the le vels of binge or heavy drinking across these two groups are not presently assesse d, these differences are of reduced importance when considering the wider implica tions of this finding. Diving deep er into the patte rns of arrest in the greater city area, one uncovers that the vast majority of law enforcement efforts (and alcohol offending) occurs in two hospitality distri cts. These districts midtown and downtown are dense with restaurants and bars (bars in particular) that stay open until 2 a.m. and are in close proximity to the university campus. Since university students and non-students (controlling for age) tend to be arrested at e qual rates, meaning th at a roughly proportiona te amount of offending exists across these groups, making an assump tion that an equal pr oportion students and nonstudents (relative to their respectiv e populations) frequent these dist ricts seems within reason. If this were the case, both stude nts and non-students about equally immerse themselves in environments that place them at an increased risk for assault, sexual harassment/assault, theft, and many other forms of victimization (Dowda ll, 2008; Jennings, Gover, & Pudrzynska, 2007; Kinney 2007). While increased levels alcohol consumption are correlated with risk of victimization, placing oneself in that environm ent while drinking to some degree (remember, 120

PAGE 121

everyone in the present data is an offender) makes up some proportion of that risk. Thus, arguing the difference of consuming three alcoholic beverages versus five in one sitting may not yield substantive differences when observing alc ohol-related problems in this manner. It would be more beneficial to observe level of intoxi cation while patrons frequent these districts (e.g., routine activities) instead of brightline criteria (e.g., 4 or 5 drinks in one sitting) as defined in the binge drinking literature. Football season aside, students and non-student s exhibit a strikingly similar pattern of alcohol-related arrest over time. This suggests a shared/universal cultu ral experience to some extent. Considering that fraternity and sorori ty members make up about half of the students arrested, it is hard to imagine individuals not enrolled in a 4-y ear institution of about the same age sharing the same cultural experiences. Conve rsely, the evidence presen ted in this study leans towards favoring an overarching culture of ambivale nce to alcohol rather than the existence of campus cultural influences that unilaterally in flate alcohol-related inci dents within the nonstudent population of the same age. While these forces are at work as strongly suggested when observing arrests during football season across these tw o groups, it seems that the similar patterns of arrest for students and non-students co-exist despite differential cultures of students and non-students (albeit anecdotal at this poi nt). This leads to the conclusion that: Alcohol-related offending (and re-offending ) seems to be tied to a cultural ambivalence toward alcohol for students a nd non-students alike. While the study university and area may be unique in many re spects, it should not be surprising if these results could be replicated in cities without a Carnegie-1 institution or in major metropolises. It also should not be surpri sing if these findings could be replicated in areas with large universities with liberal to moderate campus/community alcohol policies, which are isolated from major metr opolises, that have a large fraternity and sorority base, and have areas of a high density of bars. 121

PAGE 122

Policy Implications As currently operating, any increase in enforcem ent will not serve to reduce the levels of alcoholor drugrelated offending among college students. In fact, there is evidence that suggests that increased enforcement may backfire There are distinct advantages, however, by handling these types of offenses through the criminal justice system rather than the university disciplinary system. For first offenders that would typically be solely referre d to judicial affairs, it may be beneficial to investigate an altern ate sanctioning system th at joins university and criminal justice resources. While joint proces sing and sanctioning (hav ing the two systems run parallel) does not seem to add any benefit to st udents academic performance, it still offers a reduction in this amplification effect found among students disciplined by the university. These data suggest an extensive review of the efficacy of the university disciplinary system. This review is especially important sinc e this authority is much more likely to identify potential problem students earlier in their academic career s and serves as a first line of defense to prevent future problems. It could be that some sort of criminal justic e component needs to be added to the process to create a perception of ge tting into real trouble. For this to be examined, a self-report study of the underlying diffe rences between the per ceptions of these two systems should be performed. Furthermore, the university disciplinary system is poised to readily make an impact on the prevention of dropping out or dismissal for poor academic performance. Regular follow-ups of students th at are underperforming their college mean GPA by a half of a grade poin t at the time of first offense and have a below median weighted high school GPA to increase this opportunity. These students do not seem to be acclimating to the university and do not respond to sanctioning. It is imperative to identify an alternative program for these individuals to improve their odds of su ccess as nothing works for them at this point. 122

PAGE 123

Perhaps referral to community college with an open admissions policy after completing two years of community college coursework would prove beneficial. Sharing intelligence across domains does not seem to offer much potential for optimizing outcomes. That is, if one system (either the crim inal justice system or the university system) had the knowledge of prior offenses that were proces sed by the system, the result may likely be to increase the level of sanction leviedsomething that did not a ffect subsequent arrests or referrals. For example, if the criminal justice sy stem knew that an arrestee had a prior record for a similar offense at the university, it would be less likely to offer a deferred prosecution. Recall that level of punishment does not seem to o ffer any relief in risk of recidivism for most offenders. This potentially costly database of shared knowledge holds little promise of reducing alcoholor drug-related offenses among students. Dowdalls (2008) review of the current thinking on preventing problem drinking and alcohol-related crime is informative. First and foremost, Dowdall reminds readers that educational efforts alone fail to make any differe nces. A successful appr oach is one that is multi-dimensional and coordinated across many laye rs of a local community and the state (less importantly, the federal government). Next, he reviews the National Institute on Alcohol Abuse and Alcoholisms (NIAAA) meta-analysis of evidence-based strate gies to reduce alcohol-related crime (See National Institute on Alcohol Abuse and Alcoholism, 2002). The successful community/ environmentalbased strategies (l abeled Tier 2) are as follows: (1) increased enforcement of minimum legal drinking age laws (2) implementation, increased publicity, and enforcement of all other laws to reduce alcohol-impaired driving, (3 ) restrictions on alcohol retail outlet density, (4) increased pric e and excise taxes on alcoholic beverages, (5) responsible beverage service policies in social and commercial settings, and (6) the formation of a campus 123

PAGE 124

and community coalition involving all major stake holders. Of these strategies listed, the study community/university has utilized numbers 1, 2, and 6, yet have not seen any results (as measured by official statistics) its concerted efforts. Of particular interest is nu mber 1: increased enforcement of minimum legal drinking age (MLDA) laws. Dowdalls previous research (2 006) as well as other prominent scholars (see Wagenaar and Wolfson, 1994; Wagenaar and Toomey, 2002) finds a substantial inverse relationship to MLDA and probl em drinking and alcohol-relate d problems. However, the present study predicts little su ccess, perhaps even increased pr oblems if policymakers choose to target students (or underage dr inkers, generally). Alternativ ely, Wagenaar and Wolfson (1994) suggest increased enforcement efforts to be di rected at those providi ng the alcohol to the underage. Yet in this circumstance, if policymak ers were to target the high density bar areas, displacement may also make matters worse by incr easing recidivism as noted previously. Thus, the NIAAA chief community-based strategy may actually be creating unintended consequences, or at the minimum, be expending valuable resources with little gain. At this point, it is easy to predict that any resources expended to address this problem will be met with minimal gains. For example, it is interesting to not e that any displacement that may occur due to target hardening may increase student drug and alcohol offending if these changes shift substance use patterns to the residential areas or make no substantial change s if violations became more difficult to detect due to wider use in private residences (therefore remain unpunished). It is quite the catch-22. Factors Tied to Drugand AlcoholRelated Offending and Re-offending These data confirm existing knowledge about the risk factors associated with problematic alcohol use, but only to a certain extent. Fraternity and sorority membership, in particular, predicts a substantial in crease in the likelihood of arrest or referral for drugor alcoholrelated offenses. However, this increased likelihood of offending does not carry through to predict re124

PAGE 125

offending When observing the characteristics of students upon their first offense, the only substantial predictor of future offending is gender males are much more likely than females to re-offend overall. Yet, in general, male students are only slightly more likely than females to offend in the first place. In regards to race, wh ite students are much more likely to offend than non-whites; yet, whites and non-white students re-offe nd at equal rates. What characteristics to fraternity brother offenders and regular male st udent offenders have in common? What of the rare minority student that offends that also is not part of the Gr eek system? It seems that once a student embraces the culture and gets caught, th eir characteristics do no t seem to matter in predicting re-offending. Limitations As previously mentioned, the findings of th e present study should be replicated to be generalizable to similarly situated college towns w ith a large university. Insofar that the criminal justice processing and sanctioning in these towns should be very similar, the affects of these interventions should be parallel. However, the standard practices among university sanctioning systems may vary substantially across universities. The findings of this study should be able to be replicated if the university sanctioning system at the replication site handles students in a similar manner. For certain, the results of this study can be extrapolated to other cohorts at the study university as there is no evidence to s uggest that subsequent cohorts will differ substantially on factors of consequence. Extrapolating these results to any other contex t is not suggested. Further evaluation of the same type (examining official statistics longitudinally) is proposed to examine whether processing/sanctioning for drugand alcoholrelate d offending differs across types of authorities in other contexts. 125

PAGE 126

Future Research The present study takes a different approach to examine a contemporary problem that exists on college campuses across the United States by utilizing official data. One particular issue that arises that will need further evaluation in future research is potential selection effects in these data. There is a possibi lity that an uncontrolled variable or group of variables may be helpful in explaining the reason why the present study did not find any positive outcome after criminal justice or university intervention for alco hol and drug offenses. Of specific interest are details about offenders at the time of entry into higher education future research should try to bring in more detail about students prior life hist ories at this point. These data should include information on students familial history (specifi cally family drinking problems, delinquency and arrest, opinions on drinking and drug use, et cetera), preexisting substance use behaviors (including alcohol and observi ng the age at first use for a ny substance consumed), peer influences in high school, and other factor s such as low self-control, socio-economic information, and opinions on drug and alcohol use before arriving on campus. Well-funded replications should also examine activated genetic markers as influence on life outcomes. The present research has revealed that previous contac ts with the criminal justice system may not aid researchers and administrators in predicting problems in early a dulthood. However, some other these other factors may provide better indicators of potential problem stude nts as they arrive on campus and help determine if students with different background ch aracteristics respond differently to sanctions. Methodological challenges will be abundant whil e trying to tease out the selection effects described above. One way to begin to control for these influences is to utilize propensity score matching (Rubin & Rosenbaum, 1983). If the bac kground history of students is available to analysts, propensity score matching can assign a conditional probability to each student gauging 126

PAGE 127

their likelihood of receiving a treatment or inte rvention. As long as the intervention is independent of the outcome, the selection effect can be controlled with the inclusion of this conditional probability. Alternatively, experiment al designs can be employed to determine the differential impact of a variety of treatment pa rticularly within unive rsity sanctioning systems as these systems can be readily flexible relative to the criminal justice system. In these designs, students can be randomly assigned to different types of sanctions and followed over time to determine treatment effects. Meta-analyses (Carey, Scott-Sheldon, Carey, & DeMartini, 2007; Vasilaki, Hosier, & Cox, 2006) indicate some program s seem to be beneficial using these kinds of results to identify a kind of sanction or interven tion that should be most effective in contrast to present practices on college campuses. For ex ample, the individuali zed prevention-based intervention applied that University of Wa shington (Baer, Kivlahan, Blume, McKnight, & Marlatt, 2001; see also Schulenberg, Maggs, L ong, Sher, Gotham, Baer, Kivlahan, Marlatt, & Zucker 2001) may yield significant reductions in drinking and its related problems. The analyses presented earlier identify a high-risk group that is likely to reoffend but eventually graduate. The vast majority of th ese students are privileged, white, and remarkably intelligent who are resilient and manage to achie ve their goals (e.g., graduation) regardless of sanctions. Yet, there is a subset of stude nts who seem doomed for problems (e.g., dropout or dismissal from the university) that need further investigation, whic h is consistent with previous research (see Weschler & Wuethrich, 2002). Bo th the correlates of offending/reoffending and poor academic performance need to be reexamined to begin to address these substance use related problems on campus. Future research should also examine the l ong-term consequences of delinquency in the collegiate years. This incl udes following up with students beyond graduation, dismissal, and 127

PAGE 128

dropping out. At this point, we do not know if the criminal justice system has distinct advantages over a university sanc tioning system years after gra duation. Students need to be reevaluated after graduation to observe residual effects of sanctioning, if they exist. For example, it may be that the impact of criminal justice sanctions remains latent until one enters the workforce. Researchers should also begi n to focus on the career model to better frame research in this area. If college (e.g., the campus culture, collegiate peers, local drinking rituals, et cetera) negatively impacted th ese students lives, administrators should arm themselves with the predictors of these potent ial problems to assist students through this tumultuous growth period in their lives. The life-course perspective can aptly frame the transitions into higher education and post-graduation life an d all of the years in between. Lastly, the findings of the pr esent study need to be further contrasted with the over abundance of self-report data that exists on collegiate alcohol and drug use. In particular, the reasons why the findings of existi ng self-report data do not translat e into the results found in the analysis of official data need further explanation. Replication in other college towns is necessary to determine were the gaps in these analyses are. These replications should vary the size of the host university and town, and also look into the variables that predict alcohol/substance use problems. These factors include the density of alcohol outlets, density of student living, residential versus commuter campuses, local at titudes and policies to wards alcohol/substance use, and average drink price. Campuses in comm unities that are state capitols (such as Virginia Commonwealth University, Ohio State Universit y, University of South Carolina, and Florida State University) may differ due to the level of politics that exist in these towns and should be carefully controlled for in any replication. Of particular interest is a better description of the cultural ambivalence outlined in the present study. Does this ambivalence project outward from 128

PAGE 129

129 campus cultures that promote alcohol and drug consumption, from host communities, or do these cultures feed off of each other yielding amplified problems for both students and community members aged 18-24? Conclusions Interventions levied on students for alcohola nd drugrelated offenses in the form of sanctions from either the criminal justice syst em or a university disciplinary system fail to produce the desired outcomes expected by these au thorities. Other than a slight increase in academic performance for students sanctioned by th e criminal justice system, the effects of each system are null or may actually backfire. Th ese data strongly suggest that a culture of ambivalence toward alcohol exists that impedes the sanctioning pro cess regardless of the severity of the punishment as proscribed by law. Thus, any crackdowns or increased prosecution of these types of offenses may prove to offer little resu lts other than slight improvements for students academic careers. The majority of theoretical frameworks that fo cus on procedural justice also fail to predict the outcomes presented in this study. These stude nts appear to be impervious to procedural stigmatization and the forces of deterrence for th ese types of offenses. Future research should concentrate on the cultural factor s that predict offending and recidivism within this population. For example, fraternity and sorority membersh ip predicts offending, but not recidivism. The same occurs across race: being white predicts offending, but not re-offending. Finally, understanding the reasons why the university disciplinary system predicts amplification in the likel ihood of recidivism is of great importa nce. If the underlying factors that are associated with this amplification can be iden tified, the university stands a better chance in seeing a reduction in the levels of recidivism since this system tends to catch offenders earlier in their academic careers.

PAGE 130

APPENDIX A DESCRIBING A CULTURE OF SUBSTANCE USE A Culture of Substance Use Products such as coffee, tea, and soda, cigarettes, cigars, and smoke less tobacco, and beer, wine, and liquor have long been intertwined with American culture at different levels throughout the countrys history. Take, for example, th e entanglement of smokeless tobacco and Major League Baseball (MLB). At the point in time in which professional baseball began in the 1840s, chewing was overwhelmingly the primary method of consuming tobacco products in the United States (Orleans, Connolly, & Wo rkman, 1993). In spite of the popular consciousness coming to believe that spitting spread tube rculosis, then the leading cause of death in the nation, the habit was still popular with ball players while the na tion had progressively switched to a safer alternative cigarettes. Many of these players reported to use chewing tobacco to keep their mouths moist in ballparks with poor air and dusty conditions a nd continued to use despite the potential for consequences. Over 150 years later, a substa ntial amount of players continue to use smokeless tobacco despite the modernization of ballp arks and thus without the poor air quality the forefathers of baseball were trying to combat. On ly, in recent times, slight differences in the form of the habit occurred since the advent of snuff and pouches (s o-called bandits). Us ing self-report surveys from 1998 to 2003, Severson, Klein, Lichtensen, and Orleans (2005) discovered that 24.8% of minor league players used smokeless tobacco pr oducts in the previous 7 days in 2003, down from 31.7% in 1998 a substantial, short-term victory for public health officials and the administration of the minor leagues. Yet, th ese researchers found that 36% of major league players reported using in the previous 7 days in 2003 compared to 35.9% in 1998. 130

PAGE 131

Without a doubt chewing tobacco has always been prevalent in baseball and deeply imbedded in its culture at rates that eclipse the general publics usage of this and similar products; yet, the active promotion and marketing (primarily) of snu ff in the late 1970s and early 1980s have greatly exasperated the amount of sm okeless tobacco use among professional players (Severson, Klein, Lichtensein, Kaufman, & Orl eans, 2005). Regardless of the larger and successful societal anti-tobacco movement, and even some internal pressures on players, coaching staff, and league administrators, shifts in tobacco use among professional baseball players are slow to come. The ha bit is well established in the inst itution of baseball to the point in which use among current players is seemingly se lf-sustaining while the pressures to initiate use among young players remain high. This is well after the dangers of t obacco use have been disseminated among the general public and the fates of Babe Ruth and Bill Tuttle sealed. When substances are so tightly tied to a culture (or subculture), effectively addressing the problem without addressing the underlying culture can prove disappointing to public health officials, researchers, and policy-makers. This sort of tale could be told of many products that are currently marketed to consumers or have been in the past. Of these products listed earlier, alcoholic beverages have the poten tial to cause a trifecta of impairment, interpersonal strife and violence and significant health maladies but have dominated, and now monopolizes, as the nations accep table and legitimized inebriant in social settings (Kenney, 2006). The nations youth, howeve r, are excluded from this norm, in theory, until their 21st birthday. Yet the Americ an culture tolerates some levels of alcohol use among our youth, particularly in private settings, and the prev alence of teenaged consumption partially reflects this notion. As remarked by Wechsler and Wuethrich (2002): 131

PAGE 132

Some parents oppose their children drinking hard liquor but dont mind beer. One student told us about her fathers re sponse to alcohol: On prom night I went to my friends house, his parents were not home. Dad wasnt sure he was going to let me go; he was concerned about drugs, sex, and drinking. I told him th at I could promise two of the things, but couldnt tell him there wasnt going to be beer. He said that was the least of his worries. (243) In a time when so many noxious influences, subs tances, and situations exist for children and young adults growing up, some beer or other soft substances with friends may not seem so bad relative to all the other possibilities available to our youth. In fact, to some, it may be considered a rite of passage that connects generations w ith a bond that has been shared by many previous generations that have preceded us. An imbalance of positive depictions of alcohol consumption, and to a lesser degree, soft drug use, by teens and young adults in American films and similar popular references in popular music (across genres) consumed by youth is prima facie evidence of the foremost contradiction: American society recognizes and tolerates some le vels of consumption as being ingrained in its culture yet it remains essentia lly punitive when youths are a rrested for alcohol and drug violations. A simple overview of popular culture speaks volumes For example, cult classic films and television series with hipster beauties of both sexes and jesti ng and jovial accounts of substance use continue to reap substantial profit each year. Weeds a Showtime series that follows an attractive, young, widowed mother who deals marijuana in Southern California to sustain her familys posh lifestyle, is entering it s third season with ra ve reviews and healthy DVD sales for previous seasons. In April of 2008, the first season of Weeds ranked third in Netflix Top 25 of Television Series rentals behind The Sopranos: Season 1 and 24: Season 1 (Netflix, 2008) That 70s Show which follows a clique of teens through their stoney highschool years and shortly beyond, lasted eight seasons and conti nues in syndication. Movies such as Friday Harold and Kumar Go to White Castle Half Baked American Pie Dazed and 132

PAGE 133

133 Confused and Jay and Silent Bob Strike Back depict substance use (alc ohol and soft-drug use) in a lighthearted manner, attracting millions that can relate. Ho llywood continues to embrace the young adult years and the follies of adolescence a nd post-adolescence, with drugs and alcohol conceivably inseparable. Many celebrities are astutely picking up on this and marketing th eir own intoxicating concoctions: shock rocker Marilyn Manson pr oduces his own brand of absinthe named Mansinthe, a winner of the 2008 San Francisco Wo rld Spirits Competition; hair band giant and tequila enthusiast Sammy Hagar manufactures a line of fine tequilas from 100% blue agave called Cabo Wabo, bearing the same name as his night club in Cabo San Lucas, Mexico; Madonna, Bon Jovi, Barbara Streisand, Celine Di on, Lil Jon, KISS, Francis Coppola, and George Lucas all have their own wine labels, an d the Rolling Stones are producing a special kind of pinot noir called ice wine in 2008 labeled Symphony for the Devil. As with smokeless tobacco, alcoholic beverages and other drugs have a long cultural history in America. Yet it seems that, unlike tobacco, alcohol and particul ar soft drugs are cu rrently portrayed as mainstream and acceptable especi ally among the youth of America.

PAGE 134

APPENDIX B PERCEPTION OF STUDENTS A Self-Report Study In the fall of 2006 and spring of 2007, Jennings Khey, Miller, and Lanza-Kaduce (2007) replicated a study primarily used as a teaching de vice for an undergraduate course in research methods in criminology. The study utilizes a convenience sample of undergraduates at the university and asks these respondents a battery of questions regarding their perceptions of crime and victimization as well as their own attitudes on these subjects. Relevant to the present study, several questions were posed to respondents asking them to reveal their opinions on and own use of drugs and alcohol at the unive rsity. Figure 4-7 summarizes the pe rtinent questions of interest. Broadly speaking, students remained either neutra l or agree with the statements drunkenness is a campus problem ( = 3.20, = 1.165) and binge drinking is a campus problem ( = 3.30, = 1.147). Interestingly, student responses on the st atement I usually drink when I go out were more widely disbursed. The two modal response s are strongly disagree and agree, however the mean response centers on neutral ( = 2.93, = 1.470). One can begin to see that there are a good amount of abstainers in this group of undergraduates as well as active drinkers. In fact, 10.7% of the respondents in this study reported neve r drinking alcohol in their lifetime, and of the remaining respondents, 64.7% were considered active drinkers (as measured by use in the last 30 days). Problem drinking, as proxy measur ed by responses to this statement: When I go out and drink alcohol, I drink until I am intoxicated, tended to center between disagree and neutral responses ( = 2.43, = 1.412). This seemed to be due to a large influx of abstainers in the sample, but the fact remains that a good prop ortion of students responded agree or strongly agree when posed to assess this statement. The preponderance of this evidence suggests that students are diverse in their as sessment of alcohol as a campus problem and their own drinking 134

PAGE 135

135 habits and both range widely. In terms of students assessment s of campus problems, drinking ranks highly: students tended to di sagree with the statements I am afraid of being robbed ( = 2.04, = 1.118), I am afraid of having my car stolen ( = 2.15, = 1.403), and I am afraid of being assaulted on campus ( = 2.10, = 1.116). Figure B-1: Summary of Students Attitudes and Perceptions of Drug and Alcohol Use at the Study University.

PAGE 136

APPENDIX C MULTIVARIATE TABLES IN CLUDING STANDARD ERRORS Table C-1: Logistic Regression Estimates of the Effect of Official Processing on On-Time Graduation1 Independent Variables 1 N = 570 2 N = 570 3 N = 570 4 N = 570 5 N = 550 Race 0.89 (0.246) 0.92 (0.245) 0.90 (0.244) 0.89 (0.246) 0.95 (0.254) Gender 2.69*** (0.204) 2.54*** (0.202) 2.44*** (0.199) 2.68*** (0.204) 2.57*** (0.210) IB or AP Student 0.51*** (0.208) 0.51*** (0.207) 0.51*** (0.207) 0.51** (0.208) 0.55** (0.220) Greek Affiliation 1.06 (0.185) 1.05 (0.184) 1.08 (0.184) 1.07 (0.185) 1.01 (0.192) Term of First Offense 1.08 (0.046) 1.10* (0.045) 1.13** (0.042) 1.08 (0.046) 1.07 (0.048) Number of Offenses 0.91 (0.136) 1.13 (0.120) 0.93 (0.161) 0.87 (0.164) 0.88 (0.174) Sum of Serious Offenses 1.40 (0.262) 1.36 (0.263) 1.43 (0.261) 1.41 (0.263) 1.21 (0.279) Five-Year Program 5.11*** (0.405) 5.23*** (0.402) 5.15*** (0.402) 5.09*** (0.405) 5.33*** (0.414) Judicial Affairs Only 0.51*** (0.220) Reference Reference Arrest Only 1.51 (0.220) 1.86** (0.237) 1.87** (0.245) Both Judicial and Arrest 1.64 (0.300) 2.21** (0.322) 1.96* (0.334) High School GPA 0.75 (0.270) R2 0.17 0.16 0.16 0.17 0.18 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard error is found below the reported Exp(B) values. 136

PAGE 137

Table C-2: Logistic Regression Estimates of the Effect of Of ficial Intervention on On-Time Graduation1 (N = 570) Independent Variables Exp(B) Race 0.95 (0.247) Gender 2.48*** (0.206) IB or AP Student 0.49*** (0.210) Greek Affiliation 1.09 (0.187) Term of First Offense 1.07 (0.046) Number of Offenses 1.02 (0.160) Sum of Serious Offenses 1.21 (0.277) Five-Year Program 5.44*** (0.408) No Punishment (First Offense) Reference2 Judicial Affairs Punishment Only (First Offense) 1.28*** (0.367) Criminal Justice Punishment Only (First Offense) 1.73* (0.408) Joint Punishment (First Offense) 1.81 (0.408) Mild Punishment (First Offense) 0.49* (0.348) Severe Punishment (First Offense) 0.86 (0.412) R2 0.17 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard error is found below the reported Exp(B) values. 2 Serves as a reference category for both level of punishment (e.g., mild or severe) as well as type of punishment (e.g., judicial a ffairs or criminal justice) 137

PAGE 138

Table C-3: Logistic Regression Estimates of the Effect of Of ficial Intervention on Dropout or Dismissal1 Independent Variables Model 1 N = 329 Model 2 N = 318 Model 3 N = 329 Model 4 N = 318 Race 0.50 (0.390) 0.60 (0.411) 0.49 (0.391) 0.60 (0.412) Gender 1.90 (0.405) 1.77 (0.414) 1.88 (0.410) 1.77 (0.418) IB or AP Student 0.65 (0.369) 0.82 (0.394) 0.65 (0.372) 0.82 (0.395) Greek Affiliation 0.52 (0.367) 0.46* (0.382) 0.54 (0.370) 0.47* (0.386) Term of First Offense 0.94 (0.099) 0.94 (0.099) 0.91 (0.106) 0.93 (0.106) Number of Offenses 1.85* (0.312) 1.64 (0.331) 1.85* (0.314) 1.62 (0.332) Sum of Serious Offenses 3.54** (0.485) 3.51** (0.500) 3.56** (0.498) 3.63** (0.518) Mild Punishment (First Offense) 0.64 (0.533) 0.76 (0.558) Severe Punishment (First Offense) 0.77 (0.650) 0.76 (0.692) High School GPA 0.23** (0.472) 0.24** (0.475) R2 0.12 0.18 0.13 0.18 1 Exp(B) values reported, representing the odds ratio of current student status. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard error is found below the reported Exp(B) values. 138

PAGE 139

Table C-4: Ordinary Least Square s Estimates of the Effect of Official Processing on Recidivism1 Independent Variables 1 (N = 620) 2 (N = 599) Race 0.04 (0.061) 0.05 (0.061) Gender 0.21*** (0.051) 0.19*** (0.051) IB or AP Student 0.12* (0.054) 0.08 (0.05) Greek Affiliation 0.08 (0.054) 0.08 (0.047) Term of First Offense 0.06*** (0.012) 0.06*** (0.012) Dropout or Dismissal 0.09 (0.091) 0.06 (0.092) Current Student 0.02 (0.052) 0.02 (0.051) Five-Year Program 0.11 (0.096) 0.09 (0.095) Judicial Affairs Only Reference Reference Arrest Only 0.24*** (0.061) 0.23*** (0.061) Both Judicial and Arrest 1.38*** (0.062) 1.36*** (0.062) High School GPA 0.19** (0.065) R2 0.49 0.50 1 Unstandardized Beta values reported. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard e rror is found below the reported Exp(B) values. 139

PAGE 140

Table C-5: Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism1 Independent Variables 1 (N = 124) 2 (N = 113) 3 (N = 496) 4 (N = 477) Race 0.09 (0.217) 0.12 (0.224) 0.02 (0.044) 0.02 (0.042) Gender 0.78*** (0.189) 0.66*** (0.190) 0.05 (0.037) 0.04 (0.035) IB or AP Student 0.25 (0.178) 0.10 (0.185) 0.01 (0.040) 0.02 (0.039) Greek Affiliation 0.14 (0.163) 0.11 (0.166) 0.04 (0.034) 0.03 (0.033) Term of First Offense 0.11*** (0.041) 0.07 (0.043) 0.03** (0.009) 0.04*** (0.008) Received No Punishment (First Offense) Removed Removed Reference2 Reference2 Received Mild Punishment (First Offense) Reference Reference 0.29*** (0.059) 0.31*** (0.057) Received Severe Punishment (First Offense) 0.38* (0.217) 0.55* (0.213) 0.19** (0.071) 0.25*** (0.067) Dropout or Dismissal Constant Constant 0.003 (0.060) 0.005 (0.058) Current Student 0.03 (0.171) 0.03 (0.172) 0.02 (0.038) 0.002 (0.036) Five-Year Program 0.54 (0.295) 0.49 (0.291) 0.04 (0.072) 0.04 (0.069) Judicial Affairs Punishment Only (First Offense) Reference Reference 1.09*** (0.069) 1.06*** (0.066) Criminal Justice Punishment Only (First Offense) 0.38* (0.217) 0.32 (0.213) 0.24*** (0.044) 0.25*** (0.042) Joint Punishment (First Offense) 0.25* (0.181) 0.25 (0.179) High School GPA 0.61** (0.213) 0.03 (0.046) R2 0.23 0.28 0.36 0.39 1 Unstandardized Beta values reported. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard e rror is found below the reported Exp(B) values. 2 Serves as a reference category for both level of punishment (e.g., mild or severe) as well as type of punishment (e.g., judicial a ffairs or criminal justice) 140

PAGE 141

Table C-6: Ordinary Least Squares Estimates of the Effect of Official Intervention on Recidivism for Arrestees While Factoring in Juvenile Arrest1 Independent Variables 1 N = 117 2 N = 107 Race 0.01 (0.226) 0.07 (0.236) Gender 0.72*** (0.192) 0.61*** (0.192) IB or AP Student 0.10 (0.189) 0.05 (0.188) Greek Affiliation 0.11 (0.168) 0.23 (0.170) Term of First Offense 0.07 (0.043) 0.08 (0.042) Received No Punishment (First Offense) Removed Removed Received Mild Punishment (First Offense) Reference Reference Received Severe Punishment (First Offense) 0.44* (0.217) 0.44* (0.209) Dropout or Dismissal Constant Constant Current Student 0.04 (0.177) 0.13 (0.178) Five-Year Program 0.52 (0.298) 0.50 (0.291) Judicial Affairs Punishment Only (First Offense) Reference Reference Criminal Justice Punishment Only (First Offense) 0.29 (0.221) 0.35 (0.213) Joint Punishment (First Offense) 0.20 (0.185) 0.31 (0.181) High School GPA 0.52* (0.222) 0.33 (0.225) Juvenile Arrest 0.37 (0.336) Removed R2 0.27 0.24 1 Unstandardized Beta values reported. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard e rror is found below the reported Exp(B) values. 141

PAGE 142

Table C-7: Ordinary Least Squares Estimates of the Effect of Official Processing and Intervention on Overall GPA Difference (Models 1 & 2) and Change in GPA Difference Before and After Fi rst Intervention (Models 3 & 4) 1 Independent Variables 1 (N = 548) 2 (N = 548) 3 (N = 356) 4 (N = 356) Race 0.13** (0.051) 0.13** (0.051) 0.06 (0.098) 0.05 (0.097) Gender 0.01 (0.042) 0.01 (0.042) 0.12 (0.086) 0.13 (0.086) IB or AP Student 0.03 (0.046) 0.03 (0.046) 0.06 (0.093) 0.06 (0.093) Greek Affiliation 0.07 (0.039) 0.07 (0.039) 0.07 (0.079) 0.06 (0.079) Term of First Offense 0.02 (0.010) 0.01 (0.010) 0.06** (0.021) 0.05** (0.021) Number of Offenses 0.07* (0.034) 0.06 (0.032) 0.03 (0.071) 0.01 (0.064) Received Mild Punishment (First Offense) 0.02 (0.067) 0.05 (0.073) 0.04 (0.122) 0.22 (0.139) Received Severe Punishment (First Offense) 0.04 (0.081) 0.08 (0.085) 0.16 (0.151) 0.02 (0.160) Current Student 0.50*** (0.040) 0.49*** (0.040) 0.21** (0.081) 0.21** (0.081) Five-Year Program 0.34*** (0.075) 0.33*** (0.075) 0.06 (0.147) 0.05 (0.147) Arrest Only 0.07 (0.049) 0.29** (0.096) Both Judicial and Arrest 0.12 (0.068) 0.22 (0.142) Judicial Affairs Punishment Only (First Offense) 0.08 (0.075) 0.15 (0.159) Criminal Justice Punishment Only (First Offense) 0.06 (0.052) 0.34*** (0.102) Joint Punishment (First Offense) 0.13 (0.085) 0.09 (0.177) High School GPA 0.35*** (0.054) 0.36*** (0.055) 0.19 (0.109) 0.21 (0.109) R2 0.32 0.34 0.12 0.09 1 Unstandardized Beta values reported. Significance values are as follows: p < 0.10; *, p < 0.05; **, p < 0.01; ***, p < 0.001. The standard e rror is found below the reported Exp(B) values. No punishment is the refe rence category for all models. 142

PAGE 143

LIST OF REFERENCES Akers, R. L. (1998). Social learni ng and social structure: A genera l theory of crime and deviance. Boston: Northeastern University Press. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (DSM-IV-TR) Ann Arbor: American Ps ychiatric Association. Anglin, M. D., Hser, Y. I., & Grella, C. E. (1997). Drug addiction and treatment careers among clients in the Drug Abuse Treat ment Outcome Study (DATOS). Psychology of Addictive Behaviors, 11 (4), 308-323. Baer, J. S., Kivlahan, D. R., Blume, A. W., McKnight, P., & Marlatt, G. A. (2001). Brief intervention for heavy-drinking college student s: 4-Year follow-up a nd natural history. American Journal of Public Health, 91 (8), 1310-1316. Becker, G. S. (1968). Crime and punishment: An economic approach. Journal of Political Economy 76: 169-217. Becker, H. S. (1963). Outsiders: Studies in th e sociology of deviance New York: The Free Press. Bernburg, J. G., & Krohn, M. D. (2003). Labeling, lif e changes, and adult crime: The direct and indirect effects of official intervention in adolescence on crime in early adulthood. Criminology 41 (4), 1287-1318. Blumstein, A., Cohen, J., Rot h, J.A., Visher, C.A. (1986). Criminal careers and Career criminals. Washington D.C.: National Academies Press. Braithwaite, J. (2002). Restorative justice and responsive regulation New York: Oxford University Press. Braithwaite, J. (1989). Crime, shame, and reintegration. Cambridge, UK: Cambridge University Press. Carey, K. B., Scott-Sheldon, L. A. J., Carey, M. P., & DeMartini, K. S. (2007). Individual-level interventions to reduce co llege student drinking: A meta-analytic review. Addictive Behaviors, 32 (11), 2469-2494. Caspi, A., & Moffitt, T. E. (1993). The continuity of maladaptive behavior: From description to understanding in the study of antisocial behavi or. In D. Cicchetti, & D. Cohen (Eds.), Manual in the development of psychopathology. New York: John Wiley & Sons. Critcher, C. (2008). Moral panics and newspaper c overage of binge drinking. In F. Franklin (Ed.) Pulling newspapers apart New York: Routledge. Core Institute. (2008). Retrieved April 1, 2008, from Core Institute Description: http://www.siu.edu/~coreinst/ 143

PAGE 144

Cornish, D. B. & Clarke, R. V. (1986). The reasoning criminal New York: Springer-Verlag. Cullen, F. T., & Agnew, R. (2006). Criminological theory: Past to present 3rd Edition. Los Angeles: Roxbury. Dowdall, G. W. (2006). How public alcohol policy shapes prevention. In R. J. Chapman (Ed.). When they drink: Practitioner views and lessons learned on preventing high-risk collegiate drinking Glassboro, NJ: Rowan University. Dowdall, G. W. (2007). The role of alcohol abuse in college st udent victimization. In B. S. Fisher & J. J. Sloan, III (Eds.). Campus crime: Legal, social, and policy perspectives Springfield, IL: Charles C. Thomas. Elder, G. H., Jr. (1994). Time, human agency, and social change: Perspectives on the life course. Social Psychology Quarterly 57 (1), 4-15. Elder, G. H., Jr. (1985). Perspectives in the life-course. In G. H. Elder (Ed.). Life-course dynamics. Ithaca, NY: Cornell University Press. Elder, G. H., Jr. (1975). Age di fferentiation and the life course. Annual Review of Sociology 1, 165-190. Erikson, K. T. (1966). Wayward puritans: A study in the sociology of deviance. New York: John Wiley. Esbensen, F. A., Thornberry, T. P., & Huizi nga, D. (1991). Arrest and delinquency. In D. Huizinga, R. Loeber, & T. P. Thornberry (Eds.), Urban delinquency and substance abuse. Washington, D.C.: U.S. Department of Justice. Farrington, D. P. (2006). Building developmental a nd life-course theories of offending. In F.T. Cullen, J.P. Wright, & K.R. Blevins (Eds.), Taking stock: The stat us of criminological theory. Advances in crimi nological theory, volume 15 New Bruswick, NJ: Transaction Publishers. Farrington, D. P. (1979). Longitudi nal research on crime and delinquency. In N. Morris, & M. Tonry (Eds.), Crime and Justice (Vol. 1, pp. 289-348). Chicago: University of Chicago Press. Farrington, D. P. (1977). The effects of public labeling. British Journal of Criminology, 17, 117125. Farrington, D. P., & Tarling, D. P. (Eds.). (1985). Prediction in Criminology. New York: SUNY Press. Farrington, D. P., Gallagher, B., Morley, L., St. Ledger, R. J., & West, D. J. (1986). Unemployment, school leaving, and crime. British Journal of Criminlogy, 26 355-356. 144

PAGE 145

Florida Department of Educa tion. (2008). Florida school indicat ors report. Retrieved June 1, 2008: http://data.fldoe.org/fsir/ Florida Department of Law Enforcement. (2008). Uniform crime reports. Retrieved October 1, 2008: http://www.fdle.state.fl.us/publications/ Francis, B., Soothill, K., & Pique ro, A. R. (2007). Estimation issu es and generational changes in modeling criminal career length. Crime & Delinquency, 53 (1), 84-105. Gatorwell Health Promotion Services. (2008). Alcohol, Tobacco and Other Drug Statistics. Retrieved on October 1, 2008 from: http: //www.shcc.ufl.edu/gatorwell/atod/atod.shtml. Glueck, S., & Glueck, E. (1930). 500 Criminal careers. New York: Knopf. Gold, M., & Mattick, H. W. (1974). Experiment in the streets: The Chicago Youth Development Project. Final report. Springfield, VA: National T echnical Information Sevice. Gold, M., & Williams, J. (1970). National study of the aftermath of apprehension. Prospectus, 3, 3-12. Gottfredson, M. R. & Hirschi, T. General theory of crime. Stanford: Stanford University Press. Grasmick, H. G., Bursik, Jr., R. J., & Arneklev, B. J. (1993). Reduction in drunk driving as a response to increased threats of sham e, embarrassment, and legal sanctions. Criminology, 31 (1), 41-67. Haberman, S., Asal, N., Brady, B., Cavanaugh, K., Emmere, J., Kendell, A., Miller, D., Pritz, S., Weiler, R., Zhang, Y. (2008). Healt hy Gators 2010. Retrieved on October 10, 2008 from: http://healthygators.hhp.ufl.edu/doc s/Student Survey/2008 HG Survey Report Public 10-10-08.pdf Halnon, K. B. (2005). The power of 420: Transforming the universal code into a collective consciousness for stoners. Retrieved May 28, 2008, from High Times: http://hightimes.com/entertainment/ht_admin/1326 Hirschi, T. (1969). Causes of delinquency Berkeley: University of California Press. Hirschi, T. & Gottfredson, M. R. Control theory and the life-course perspective. Studies on crime and crime prevention, 4 (2): 131-142. Hser, Y. I., Anglin, M. D., Grella, C., Longs hore, D., & Prendergast, M. L. (1997). Drug treatment careers: A conceptual fram ework and existing research findings. Journal of Substance Abuse Treatment, 14 (6), 543-558. Hser, Y. I., Grella, C., Chou, C.P., Anglin, M. D. (1998). Relationships between drug treatment outcomes: Findings from the national Drug Abuse Treatment Outcome Study (DATOS). Evaluation Review 22 (4), 496-519. 145

PAGE 146

Hser, Y. I., Longshore, D., & Anglin, M. D. (2007). A conceptual framework for understanding drug use trajectories. Evaluation Review, 31 (6), 515-547. Huizinga, D., & Esbensen, F. A. (1992). An arresting view of juvenile justice. School Safety, 3, 15-17. Huizinga, D., Esbensen, F. A., & Weiher, A. W. (1996). The impact of arrest on subsquent delinquent behavior. In R. Loeber, D. Huizinga, & T. P. Thornberry (Eds.), Annual report of the program of research on the causes and correlates of delinquency. Washington, D.C.: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention. Institute of Medicine. (2004). Reducing underage drinking: A collective responsibility. Washington, D.C.: The National Academies Press. Jennings, W. G., Gover, A. R., & Pudrzynska, D. (2007). Are institutions of higher education safe? A descriptive study of campus safety i ssues and self-reported campus victimization among male and female college students. Journal of Criminal Justice Education 18 (2), 191-208. Jennings, W. G., Khey, D. N., & Miller, B. L. (2007). Students perceptions and attitudes towards crime and victimization [Computer file]. Gainesville, Florida. Johnston, L. D., O'Malley, P. M., Bachma n, J. G., & Schulenberg, J. E. (2007). Monitoring the future: National survey resu lts on drug use, 1975-2006: Volume II, College students and adults ages 19-45. Bethesda, MD: National Institute on Drug Abuse. Jones, B. L., & Nagin, D. S. (2006). Advances in group-based traj ectory Modeling and a SAS procedure for estimating them Retrieved May 18, 2008, from http://www.andrew.cmu.edu/user/bjones/ref2.pdf Jones, B. L., Nagin, D. S., & Roeder, K. (2001) A SAS procedure based on mixture models for estimating developmental trajectories. Sociological Methods and Research, 29, 374-393. Khey, D. N., Fox, K. A., & Lanza-Kaduce, L. (2008). University of Florida comprehensive student crime study [Computer f ile]. Gainesville, Florida. Kinney, J. (2006). Loosening the grip: A handbook of alcohol information 8th Edition. New York: McGraw-Hill. Kitsuse, J. I. (1964). Societal reaction to devian t behavior: Problems of theory and method. In H. S. Becker (Ed.). The Other Side. New York: The Free Press. Klein, M. (1986). Labeling theory and delinquency policy: An empirical test. Criminal Justice and Behavior, 13 47-79. Kotler, P. & Roberto, E. L. (1989). Social marketing: Strategies for changing public behavior New York: Free Press. 146

PAGE 147

Kurlychek, M. C., Brame, R., & Bushway, S. D. (2006). Scarlet letters and recidivism: Does an old criminal record predict future offending. Criminology & Public Policy, 5 (3), 483504. Laub, J. L., & Sampson, R.J. (2003). Shared beginnings, divergent li ves: Delinquent boys to age 70. Cambridge, MA: Harvard University Press. Lemert, E. M. (1972). Human deviance, social probl ems, and social control 2nd Edition. Englewood Cliffs, NJ: Prentice Hall. Lemert, E. M. (1951). Social pa thology: A systematic approach to the theory of sociopathic behavior. New York: The Free Press. Lender, M. E., & Martin, J. K. (1982). Drinking in America. New York: The Free Press. Marciniak, E. (1994). Community policing of dom estic violence: Neighborhood differences in the effects of arrest. College Park: University of Maryland. McCabe, S. E., Schulenberg, J. E., Johnston, L. D., O'Malley, P. M., Bachman, J. G., & Kloska, D. D. (2005). Selection and socialization e ffects of fraternities and sororities on US college student substance use. A mu lti-cohort national longitudinal study. Addiction, 100 (4), 512-524. Miller, V. (2008). CU's 4/20 pot smoke-out draws crowd of 10,000: Police issue zero tickets during annual marijuana celebration. Retrieved May 28, 2008, from Daily Camera: http://www.dailycamera.com/news/2008/ap r/20/cus-420-pot-smoke-out-draws-10000/ Misch, D. (2007). "Natural recovery" from Alcohol Abuse Among College Students. Journal of American College Health, 55 (4), 215-218. Moffitt, T. E., Caspi, A., Harrington, H, & Milne, B.J. (1994). Males on the life-course-persistent and adolescent-limited pathways: Follow-up at age 26 years. Development and Psychopathology, 14, 179-207. Moffitt, T. E. (1993). Adolescent-limited and lif e-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100 674-701. Mustaine, E. E. & Tewksbury, R. (2007). The rou tine activities and crimin al victimization of students: Lifestyle and related factors. In B. S. Fisher & J. J. Sloan, III (Eds.). Campus crime: Legal, social, and policy perspectives Springfield, IL: Charles C. Thomas. Nagin, D. S. (1999). Analyzing developmental tr ajectories: A semi-par ametric, group-based approach. Psychological Methods, 4 139-177. Nagin, D. S., & Tremblay, R. E. (2005). Develo pmental trajectory groups: Fact or useful statistical fiction. Criminology, 43 (4), 873-904. 147

PAGE 148

Nagin, D. S., & Tremblay, R. E. (2001). Analyzin g developmental trajecto ries of distinct but related behaviors: A group-based method. Psychological Methods, 6 18-34. National Institute on Alcohol Abuse and Alcoholism. Test force on college drinking: Final report of the panel on prevention and treatment How to reduce high-risk college drinking. Rockville, MD. National Institute on Drug Abuse. (2008). Addict ion is a chronic disease. Retrieved May 21, 2008, from http://www.nida.nih.gov/about/wel come/aboutdrugabuse/chronicdisease/ National Security PAC. (1987). Willie Horton (Bush). Retrieved May 21, 2008, from The Living Room Candidate: Presidential Ca mpaign Commercials. 1952-2004: http://livingroomcandidate.movingima ge.us/election/index/index.php?ad_id=944 Netflix. (2008). Retrieved April 1, 2008, from Netflix Top 100: http://www.netflix.com Orleans, C. T., Connolly, G. N., & Workman, S. (1993). Beat the smokeless habit: A game plan for success. Bethesda, MD: National Institutes of Health. Patternoster, R., & Iovanni, L. (1996). The labeling perspective and delinquency: An elaboration of the theory and an explanation of the evidence. In P. Cordella, & L. Siegel (Eds.), Readings in contemporary criminological theory (pp. 171-190). Lebanon, NH: UPNE. Piquero, A. R., Moffitt, T. E., & Wright, B. E. (2007). Self-contro l and criminal career dimensions. Journal of Contemporary Criminal Justice, 23 (1), 72-89. Piquero, A. R., & Patternoster, R. (1998). An application of Stafford and Warrs reconceptualization of deterr ence to drinking and driving. Journal of Research in Crime and Delinquency, 35 (1), 3-39. Piquero, A. R., & Pogarsky, G. (2002). Beyond St afford and Warrs reconceptualization of deterrence: Personal and vicar ious experiences, impulsiv ity, and offending behavior. Journal of Research in Crime and Delinquency, 39 (2): 153-186. Pratt, T. C., Cullen, F. T., Blevins, K. R., Daigle, L. E., & Madensen, T. D. (2006). The empirical status of deterrence theory: A meta-ana lysis. In F. T. Cullen, J. P. Wright, & K. R. Blevins (Eds.), Taking stock: The status of crim inological theory Advances in criminological theory, Volume 15 New Brunswick, NJ: Transaction. Presley, C., Cheng, Y., & Pimentel, E. (2004) Alcohol and drugs on American college campuses: A report to college campuse s. Carbondale: Core Institute. Princeton Review, The. (2008). Parties. Retrieved December 1, 2008, from The Princeton Review: http://www.princetonreview. com/bestcolleges/parties.aspx Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrica 70, 41. 148

PAGE 149

Sack, K. (2008a). Personal Communication. Sack, K. (2008b). At the legal limit Retrieved December 1, 2008, from the New York Times: http://www.nytimes.com/2008/11/02/e ducation/edlife/alcohol.html Sampson, R.J. & Laub, J.H. (2005). A life-cour se view of the development of crime. American Annals of the American Academy of Political and Social Sciences 602: 12-45. Sampson, R.J. & Laub, J.H. (1993). Crime in the making Cambridge, MA: Harvard University Press. Scherer, M. (2008). A Willie Horton hit on Obama. Retrieved May 21, 2008, from TIME Magazine Online: http://www.time .com/time/politics/artical/0,855,1733873,00.html Schulenberg, J., Maggs, J., Long, S. W., Sher, K. J ., Gotham, H. J., Baer, J. S., Kivlahan, D.R., Alan Marlatt, G., & Zucker, R.A. (2001). The problem of college drinking: Insights from a developmental perspective. Alcoholism: Clinical & Ex perimental Research, 25 (3), 473-477. Schur, E. M. (1969). Reactions to deviance: A critical assessment. American Journal of Sociology 75, 309-322. Severson, H. H., Klein, K., Lichtensein, E., Ka ufman, N., & Orleans, C. T. (2005). Smokeless tobacco use among professional baseball pl ayers: survey results, 1998 to 2003. Tobacco Control, 14, 31-36. Sherman, L. W., Gottfredson, D., MacKenzie, D ., Eck, J., Reuter, P., & Bushway, S. (1997). Preventing crime: What works, what doesn't, what's promising. Washington, D.C.: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. Smith, D., & Gartin, P. R. (1989). Specifying sp ecific deterrence: The influence of arrest on future criminal activity. American Sociological Review, 54 (1), 94-106. Sprott, J. B., Jenkins, J. M., & Doob, A. N. (2000). Early offending: Unde rstanding the risk and protective factors of delinquency Hull, Quebec: Human Resources Development Canada. Stafford, M. C., & Warr, M. (1993). A reconceptu alization of general and specific deterrence. Journal of Research in Crime and Delinquency, 30 (2): 123-128. Substance Abuse and Mental Health Services Ad ministration, Office of Applied Studies. (2008). Results from the 2007 national survey on drug use and health: National findings (NSDUH Series H-34, DHHS Publication No. SMA 08-4343). Rockville, MD. Surette, R. (2007). Media, crime, and criminal justice: Images, realit ies, and policies (3rd Edition). Belmont, CA: Thomson Wadworth. Sweeten, G. (2006). Who will graduate? Disruption of high school education by arrest and court involvement. Justice Quarterly 23 (4), 462-480. 149

PAGE 150

Tannenbaum, F. (1938). Crime and the community New York: Columbia University Press. Tibbetts, S. G. & Piquero, A. R. (1999). The influence of gender, low birth weight, and disadvantaged environment in predicting early onset of offending: A test of Moffitts interactional hypothesis. Criminology 37 (4), 843-878. Tonry, M., Ohlin, L. E., & Farrington, D. P. (1991). Human development and criminal behavior: New ways of advancing knowledge. New York: Springer-Verlag. Tunnel, K. D. (1992). Choosing crime Chicago: Nelson-Hall. University of Florida. (2008). When to appl y to UF. Retrieved on September 10, 2008 from The University of Florida: http://www.admissions.ufl.edu/ugrad/frapply.html United States Department of Education. ( 2008). Campus security. Retrieved on May 1, 2008, from: http://www.ed.gov/admins/lead/safety/campus.html. Vasilaki, E. I., Hosier, S. G., & Cox, W. M. ( 2006). The efficacy of motivational interviewing as a brief intervention for excessive drinking: A meta-analytic review. Alcohol and Alcoholism 41(3), 328-335. Wagenaar, A. C. & Toomey, T. L. (2002). Eff ects of minimum drinking age laws: Review and analyses of the literature from 1960 to 2000. Journal of Studies on Alcohol 14, 206-225. Wagenaar, A. C. & Wolfson, M. (1994). Enforcemen t of the legal minimum drinking age in the United States. Journal of Public Health Policy, 15 (1): 37-53. Weschler, H., & Wuethrich, B. (2002). Dying to drink: Confronting bi nge drinking on college campuses. New York: Rodale. Weschler, H., Kuo, M., Nelson, T. F., & Lee, H. (2002). Trends in college binge drinking during a period of increased prevention efforts. Find ings from four Harvard School of Public Health College Alcohol Study surveys: 1993-2001. Journal of American College Health, 50 (5), 203-221. Weschler, H., Lee, J. E., Kuo, M., Lee, H. (2000). College binge drinking in the 1990s: A continuing problem. Results of the Harvard School of Public Health 1999 College Alcohol Study. Journal of American College Health, 48 (5): 199-210. Weschler, H., Nelson, T. F., Lee, E. L., Seibring, M., Lewis, C., & Keeling, R. P. (2003). Perception and reality: A national eval uation of social norms marketing. Journal of Studies on Alcohol 64 (4), 484-494. White, H. R., Labouvie, E. Rutgers Alcohol Problem Index. Retrieved on May 15, 2008, from http://alcoholstudies.rutge rs.edu/etiology/rapi.html. Wolfgang, M., Figlio, R., & Sellin, T. (1972). Delinquency in a birth cohort. Chicago: University of Chicago Press. 150

PAGE 151

151 Zimmerman, R. (1997). Social marketing strategies for cam pus prevention of alcohol and other drug problems Newton, MA: The Higher Education Ce nter for Alcohol and Other Drug Prevention Education Development Center, Inc.

PAGE 152

BIOGRAPHICAL SKETCH Dave Khey holds two masters degrees from the University of Florida, one in Pharmacy and Pharmaceutical Sciences with a concentration in Forensic Drug Analysis, and the other in Sociology. This work fulfills his requirements for a Doctorate in Criminology, Law, and Society at the University of Florida. His most recent works include an assessment of internal controls across gender, and their role in academic dis honesty in undergraduates with Chris Gibson and Chris Schreck, several investigations into a psychoactive plant called Salvia divinorum and its use among undergraduates with Bryan Miller and Hayden Griffin, and the geospacial analysis of college student arrestees with Matt Nobles and Kathleen Fox. 152

PAGE 153

DRUG AND ALCOHOL ARRESTEES IN A COLLEGE TOWN: EXPLORING MULTIPLEAGENCY OFFICIAL DATA TO ASSESS THE IMPACT OF ARREST AND SANCTIONS ON CRIMINAL AND ACADEMIC CAREERS David Nicolaus Khey (352) 273-0189 Sociology and Criminology Lonn Lanza-Kaduce Criminology, Law, & Society May 2009 This dissertation assesses the impact of alc oholand drugoffenses on a students college career as well as examine if offenders are likely to get into further trouble. The reader will gain a better understanding of what fact ors are involved in the likelihood of reoffending for these crimes, and in turn, understand how they may begin to address reducing drugand alcoholrelated problems in college student populations. This di ssertation also fills in a much needed gap by looking at this issue by using officially collected arrest da ta instead of those produced by surveys. It is always best to approach a problem using a variety of data sources and methodologies for data collection, and this dissertation begins to fill this void.