USVI Public Records

Document text

A STUDY OF JUVENILE OFFENDERS in St Thomas and St Croix, USVI Submitted to: The Law Enforcement Planning Commission Government of the US Virgin Islands Prepared by: The Eastern Caribbean Center University of the Virgin Islands June 2003 A STUDY OF JUVENILE OFFENDERS in St Thomas and St Croix, USVI Submitted to: The Law Enforcement Planning Commission Government of the US Virgin Islands Prepared by: Frank L. Mills, PhD Director Annette Gumbs Marcy Mason, PhD Research Analyst III Research Asst. Professor of Statistics James Richardson Research Analyst I Eastern Caribbean Center University of the Virgin Islands June 2003 Acknowledgements The 1990s was a particularly stressful decade for Virgin Islands parents, school administrators and teachers. Reports in the media of the increasing level and frequency of violence among youth across the Territory shocked adult residents into the realization that the culture of juveniles was very different from that within which they were raised. Perhaps the low points were reached when it became evident that dangerous weapons, including firearms, were being smuggled by high school students on to the campuses, and when this deviant behavior ultimately led to homicide. Inevitably, many in the community looked to our elected leaders for answers to the questions concerning the estrangement of our youth. It was with this background that former Virgin Islands Senator Allie “Allison” Petrus—who campaigned on a platform of youth issues—invited the Director of the Eastern Caribbean Center (ECC) to his office to discuss the possibility of conducting a penetrating study of juvenile behaviors. It was not until funds to support this activity were identified by Ms Helene Smollett of the Law Enforcement Planning Commission (LEPC) that this study got underway. The first formal meeting to launch this effort was held at the LEPC offices in Sub-Base, St Thomas under the patronage of former Commissioner of Police Ramon Davila. Also present at that meeting was former Commissioner of the Department of Human Services Ms Catherine Mills. She was very instrumental in suggesting that analysis should be undertaken to flesh out an explanation why some youth become recidivist and others do not. When the decision was made to have data collected on youthful offenders from records within the police department, two officers stand out for their contribution in supervising this task. They are Sgt Merlin Christian of the St Thomas Juvenile Investigation Bureau, and Sgt Thomas Hannah of the St Croix Youth Investigation Bureau. ECC also acknowledges the endeavors of the many other officers on both islands who assisted with this aspect of the project. There were many juveniles for whom records existed in both the Police Department and the Department of Human Services, and it was considered critical to link the two. ECC had the unhesitating support of Human Services Commissioner Sedonie Halbert, and it was no less fortunate in having the enthusiastic encouragement and willingness to help of Ms Janet Turnbull-Krigger in the St Thomas office. Ms Carol Battuelo in the St Croix office was equally supportive as she supervised the collection of data. This report has benefited enormously from the input of several staff members of the Eastern Caribbean Center during the time it required to complete it. One of our former staff, Ms Carmen Rogers-Green, provided significant input in the early stages of the project, including survey instrument development, data collection, tabulation and the initial writing of the report. Mr Ramesh Srivastava’s statistical expertise proved invaluable in proofing and certifying many of the statistical tables that appear in the report. And even though Ms Marsha Penn’s tenure with the Center was not very long, her willing contribution in helping to bring the report to near completion is sincerely appreciated. I particularly wish to recognize the current members of staff who have endured my exhortations to get this report into the hands of those who desire to understand what the data on juvenile offenses are saying about our community. Long-term staff member Ms Annette Gumbs deserves applause for the valuable assistance she provided through all phases of project planning, data collection and processing, as well as in the development of the report. Our relatively new staff member Mr James Richardson is also complimented for introducing new energy into the project, especially through his skills in producing our first single integrated electronic report. I value no less the enthusiasm and assiduousness of our latest staff member, Dr Marcy Mason, who not only reviewed the entire draft report, but who contributed through her skills several improvements in formatting, presentation and interpretation. Finally, ECC could not have completed its work without the support of several institutional offices, including: the office of Vice Provost Dr Henry Smith, that of Financial Management, Human Resources and Accounting. I thank all of those mentioned above, as well as those I may have inadvertently overlooked, and hope that through this work we will all help to improve the quality of life for the youth in the Virgin Islands community. Frank L. Mills Director TABLE OF CONTENTS ACKNOWLEDGEMENTS ............................................................................................. iii LIST OF TABLES........................................................................................................... vii LIST OF FIGURES ......................................................................................................... x HIGHLIGHTS ................................................................................................................. xi I. INTRODUCTION 1.1 Background to the Study..................................................................................... 1 1.1.1 Historical and Social Change .................................................................. 1 1.1.1.1 Some Effects of Immigration ............................................................ 1 1.1.2 Risk Factors Affecting Juveniles ............................................................. 4 1.1.3 At-Risk Indicators in the US Virgin Islands ........................................... 5 1.1.4 Youth Crime Trends................................................................................. 5 1.1.5 US Virgin Islands Youth Rehabilitation Center ..................................... 7 1.1.6 Department of Human Services .............................................................. 9 1.1.7 Juvenile Units .......................................................................................... 10 1.2 Purpose of the LEPC Study ............................................................................... 12 1.3 Scope of the Study ............................................................................................. 13 1.4 Survey Goals ...................................................................................................... 14 II. PERSPECTIVES AND METHODOLOGY 2.1 Delinquency and Antisocial Behavior ................................................................ 15 2.1.1 Causes and Recommendations ................................................................ 15 2.1.2 Development, Persistence and Pathways of Early Problem Behaviors .. 15 2.1.3 Explanation of the Causes of Antisocial Behavior .................................. 17 2.1.4 Prenatal and Perinatal Factors ................................................................. 17 2.1.5 Hyperactivity ........................................................................................... 17 2.1.6 Intelligence and Attainment..................................................................... 18 2.1.7 Parental Supervision, Discipline, and Attitude ........................................ 18 2.1.8 Broken Homes.......................................................................................... 19 2.1.9 Large Family Size .................................................................................... 19 2.2 Methodology....................................................................................................... 20 2.2.1 Steps in Developing the Framework of the Study................................... 20 2.2.2 Data Collection Instrument ...................................................................... 21 2.2.3 Selection of the Population...................................................................... 21 2.2.4 Data Collection Procedures...................................................................... 21 2.2.5 Data Availability and Deficiencies .......................................................... 22 2.2.6 Data Analysis Procedures ........................................................................ 22 2.2.7 Organization of the Report....................................................................... 22 2.2.8 Limitations of the Study........................................................................... 22 III. PRESENTATION ANALYSIS St Thomas Data: Description and Analysis ........................................................... 25 3.1 Juvenile Investigation Bureau Data Variables.................................................... 25 3.1.1 St Thomas Data Description: Juvenile Investigation Bureau................. 25 3.1.2 St Thomas: Juvenile Investigation Bureau Data Relationships .............. 27 3.1.3 Predicting Recidivism with the St Thomas Juvenile Investigation Bureau Data.............................................................................................. 33 1.1.3.1 Approach ........................................................................................ 33 2.1.3.2 Methodology .................................................................................. 34 3.1.3.3 Definition of Variables ................................................................... 34 4.1.3.4 The Model ...................................................................................... 35 5.1.3.5 Estimates of the St Thomas Juvenile Investigation Bureau Data… 35 6.1.3.6 The Prediction of Recidivism ......................................................... 37 3.2 Department of Human Services Data Variables………………………………. 40 3.2.1 St Thomas Data Description: Department of Human Services ............... 41 3.2.2 St Thomas: Department of Human Services Data Relationships ............ 45 3.2.3 Predicting Recidivism with the St Thomas Department of Human Services Data ........................................................................................... 48 1.2.3.1 Estimates of the St Thomas Department of Human Services Data ................................................................................................. 48 3.2.3.2 The Prediction of Recidivism ......................................................... 52 St Croix: Data Description and Analysis .............................................................. 53 3.3 Youth Investigation Bureau Data Variables ....................................................... 53 3.3.1 St Croix Data Description: Youth Investigation Bureau......................... 53 3.3.2 St Croix: Youth Investigation Bureau Data Relationships ...................... 55 3.3.3 Predicting Recidivism with the St Croix Youth Investigation Bureau Data ......................................................................................................... 63 3.3.3.1 Definition of the Variables............................................................. 64 3.3.3.2 Estimates of the St Croix Youth Investigation Bureau Data ......... 64 3.3.3.3 The Prediction of Recidivism ........................................................ 65 3.4 Department of Human Services Data Variables ................................................. 68 3.4.1 St Croix Data Description: Department of Human Services .................. 68 3.4.2 St Croix: Department of Human Services Data Relationships ................ 72 3.4.3 Predicting Recidivism with the St Croix Department of Human Services Data............................................................................................ 80 1.4.3.1 Estimates of the St Croix Department of Human Services Data ................................................................................................. 80 3.4.3.2 The Prediction of Recidivism ......................................................... 80 3.5 Summary and Overview for the Territory........................................................... 84 3.5.1 Juvenile Units............................................................................................ 84 3.5.2 Department of Human Services ............................................................... 85 IV. IMPLICATIONS OF THE FINDINGS AND RECOMMENDATIONS ................ 89 REFERENCES ................................................................................................................ 91 APPENDIX...................................................................................................................... 92 vii LIST OF TABLES Table 1.1 Place of Birth by Year: 1980, 1990, 1995, 1997 ...................................................... 3 Table 1.2 Possible Indicators of Juvenile Delinquent Behavior ............................................... 5 Table 1.3 Three Population Types Served by the Youth Rehabilitation Center: FY 1999 ..... 8 Table 1.4 Characteristics of Youths Remanded at YRC FY 1999 .......................................... 8 Table 1.5 Juveniles Remanded Outside YRC FY 1999............................................................. 9 Table 3.1.1 Year of First Offense ............................................................................................. 25 Table 3.1.2 Type of First Offense ............................................................................................. 26 Table 3.1.3 Weapon Used in First Offense ............................................................................... 26 Table 3.1.4 Total Number of Offenses ..................................................................................... 26 Table 3.1.5 Total Number of Offenses by Place of Birth ......................................................... 27 Table 3.1.6 Type of First Offense by Place of Birth ................................................................. 27 Table 3.1.7 Type of Weapon Used by Place of Birth ............................................................... 28 Table 3.1.8 Duration Between First and Second Offense ......................................................... 28 Table 3.1.9 Age at First Offense by Sex ................................................................................... 29 Table 3.1.10 Place of Birth of Juvenile by Sex ........................................................................... 29 Table 3.1.11 Age at First Offense by Juvenile’s Place of Birth ................................................. 30 Table 3.1.12 Age at First Offense by Mother’s Place of Birth ................................................... 30 Table 3.1.13 Age at First Offense by Father’s Place of Birth ..................................................... 31 Table 3.1.14 Juvenile’s Race by Sex .......................................................................................... 31 Table 3.1.15 Juvenile’s Guardian by Juvenile’s Sex .................................................................. 32 Table 3.1.16 School Enrollment ................................................................................................. 32 Table 3.1.17 School Type ........................................................................................................... 32 Table 3.1.18 Mother’s Place of Birth by Sex of Juvenile ........................................................... 33 Table 3.1.19 Mother’s Employment by Sex of Juvenile ............................................................ 33 Table 3.1.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Thomas, Juvenile Investigation Bureau ............................................................. 36 Table 3.1.3.2 Predicted Probabilities in Logistic Regression: St Thomas, Juvenile Investigation Bureau .............................................................................................. 38 Table 3.2.1 Juvenile’s Sex ........................................................................................................ 41 Table 3.2.2 Immigration Status ................................................................................................. 41 Table 3.2.3 Age of Juvenile ...................................................................................................... 41 Table 3.2.4 Juvenile’s Place of Birth ........................................................................................ 42 Table 3.2.5 Juvenile’s Race ...................................................................................................... 42 Table 3.2.6 Household Status of Parents .................................................................................. 42 Table 3.2.7 Age of Parents ........................................................................................................ 43 Table 3.2.8 Place of Birth of Parents ........................................................................................ 43 Table 3.2.9 Marital Status of Parents ........................................................................................ 43 Table 3.2.10 Employment Status of Parents ............................................................................... 43 Table 3.2.11 Juvenile’s Guardian ............................................................................................... 44 Table 3.2.12 Marital Relationship of Natural Parents ................................................................ 44 Table 3.2.13 Siblings in the Household ...................................................................................... 44 Table 3.2.14 School Enrollment .................................................................................................. 44 viii Table 3.2.15 Current or Highest Grade Completed .................................................................... 45 Table 3.2.16 Juvenile’s Employment Status ............................................................................... 45 Table 3.2.17 Current Job ............................................................................................................. 45 Table 3.2.18 Number of Children in the Household ................................................................... 46 Table 3.2.19 Juvenile’s Children ............................................................................................... 46 Table 3.2.20 Juvenile Responsible for Childcare ....................................................................... 46 Table 3.2.21 Juvenile’s Psychological Assessment..................................................................... 46 Table 3.2.22 Juvenile’s Sexual Activity ..................................................................................... 47 Table 3.2.23 Juvenile Forced to have Sex .................................................................................. 47 Table 3.2.24 Juvenile’s Sex Partners .......................................................................................... 47 Table 3.2.25 Pre-Outcome Status ............................................................................................... 48 Table 3.2.26 Outcome Status ...................................................................................................... 49 Table 3.2.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Thomas, Department of Human Services Data ................................................. 50 Table 3.2.3.2 Predicted Probabilities in Logistic Regression: St Thomas, Department of Human Services Data ....................................................................................... 51 Table 3.3.1 Year of First Offense ............................................................................................. 54 Table 3.3.2 Type of First Offense.............................................................................................. 54 Table 3.3.3 Weapon Used in First Offense ............................................................................... 54 Table 3.3.4 Total Number of Offenses ..................................................................................... 55 Table 3.3.5 Age at First Offense by Sex ................................................................................... 55 Table 3.3.6 Juvenile’s Place of Birth ........................................................................................ 56 Table 3.3.7 Juvenile’s Place of Birth by Sex ............................................................................ 57 Table 3.3.8 Age at First Offense by Juvenile’s Place of Birth ................................................. 57 Table 3.3.9 Age at First Offense by Mother’s Place of Birth ................................................... 58 Table 3.3.10 Age at First Offense by Father’s Place of Birth ..................................................... 58 Table 3.3.11 Juvenile’s Race by Sex .......................................................................................... 59 Table 3.3.12 Juvenile’s Guardian by Sex ................................................................................... 59 Table 3.3.13 School Enrollment ................................................................................................. 60 Table 3.3.14 School Type ........................................................................................................... 60 Table 3.3.15 Mother’s Place of Birth by Sex of Juvenile ........................................................... 60 Table 3.3.16 Mother’s Employment by Sex of Juvenile ............................................................ 61 Table 3.3.17 Father’s Place of Birth by Sex of Juvenile ............................................................ 61 Table 3.3.18 Father’s Employment by Sex of Juvenile .............................................................. 62 Table 3.3.19 Recidivism and Relatives Charged with Offenses ................................................. 62 Table 3.3.20 Juvenile’s Age by Current or Highest Grade ......................................................... 63 Table 3.3.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Croix, Youth Investigation Bureau Data ........................................................................... 65 Table 3.3.3.2 Predicted Probabilities in Logistic Regression: St Croix, Youth Investigation Bureau Data............................................................................................................. 66 Table 3.4.1 Juvenile’s Sex ........................................................................................................ 68 Table 3.4.2 Immigration Status ................................................................................................. 68 Table 3.4.3 Age of Juvenile ...................................................................................................... 69 Table 3.4.4 Juvenile’s Place of Birth ........................................................................................ 69 Table 3.4.5 Juvenile’s Race ...................................................................................................... 69 Table 3.4.6 Household Status of Parents .................................................................................. 70 ix Table 3.4.7 Age of the Parents .................................................................................................. 70 Table 3.4.8 Marital Status of Parents ........................................................................................ 70 Table 3.4.9 Juvenile’s Guardian ............................................................................................... 71 Table 3.4.10 Marital Relationship of Natural Parents ................................................................ 71 Table 3.4.11 Siblings in the Household ...................................................................................... 71 Table 3.4.12 School Enrollment ................................................................................................. 72 Table 3.4.13 Current or Highest Grade Completed .................................................................... 72 Table 3.4.14 Juvenile’s Age by Current or Highest Grade ......................................................... 73 Table 3.4.15 Juvenile’s Employment Status ............................................................................... 74 Table 3.4.16 Current Job ............................................................................................................. 74 Table 3.4.17 Children in the Household ..................................................................................... 74 Table 3.4.18 Juvenile’s Children ................................................................................................ 74 Table 3.4.19 Juvenile Responsible for Childcare ....................................................................... 74 Table 3.4.20 Juvenile’s Psychological Assessment .................................................................... 75 Table 3.4.21 Juvenile’s Sexual Activity ..................................................................................... 75 Table 3.4.22 Forced Sex ............................................................................................................. 75 Table 3.4.23 Juvenile’s Sex Partners .......................................................................................... 76 Table 3.4.24 Pre-Outcome Status ............................................................................................... 76 Table 3.4.25 Type of Offense ..................................................................................................... 77 Table 3.4.26 Year of Offense ...................................................................................................... 78 Table 3.4.27 Outcome of Offense ............................................................................................... 79 Table 3.4.28 Biological Father Known ....................................................................................... 80 Table 3.4.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Croix, Department of Human Services Data...................................................................... 81 Table 3.4.3.2 Predicted Probabilities in Logistic Regression: St Croix, Department of Human Services Data ............................................................................................. 82 Table A1 Parent’s Place of Birth – St Thomas.……………………………………………… 92 Table A2 Type of Offense – St Thomas……………………………………………………... 92 Table A3 Type of Weapon – St Thomas…………………………………………………….. 93 Table A4 Year of Offense – St Thomas…………………………………………………….... 94 Table A5 Outcome of Offense – St Croix…………………………………………………… 96 x LIST OF FIGURES Figure 3.1.3.1 Predicted Chances of Recidivism: St Thomas, JIB .......................................... 39 Figure 3.1.3.2 First-time Arrests of Juveniles in St Thomas: JIB.............................................. 40 Figure 3.2.3.1 Predicted Chances of Recidivism: St Thomas, DHS ........................................ 52 Figure 3.3.3.1 Predicted Chances of Recidivism: St Croix, YIB ............................................ 67 Figure 3.3.3.2 Age at First Offense: St Croix YIB Data........................................................... 67 Figure 3.4.3.1 Predicted Chances of Recidivism: St Croix, DHS ............................................ 83 xi HIGHLIGHTS ST THOMAS Juvenile Investigation Bureau ? The largest number of recorded first-time juvenile offenses committed between 1980 and 1997 was in 1991 with a total of 298. ? About 24 percent of the total number of first time offenses was classified as incidences (which include run-away and missing minors). ? Out of every 10 juvenile offenders, 8 did not use a weapon in committing their first offense. ? About 36.3 percent of all juvenile offenders committed two or more offenses. ? Approximately 7 out of 10 juvenile offenders were born in the US Virgin Islands. ? Juvenile offenders born in Puerto Rico comprised the highest percentage—41.3—of repeat offenders among all ethnic groups. ? Juvenile offenders born in the US Virgin Islands had the lowest percentage of any group classified as not using a weapon during the commission of an offense. ? Most repeat offenses—21.2 percent—occurred between 12 to 24 months after the first offense. ? At the time of their first offense, one out of every five—or 20 percent of— juvenile offenders was 15 years old; most males—18.9 percent—committed their first offense at this age, while most females—22.5 percent—recorded their first offense at age 14. ? Approximately 50 percent of juvenile offenders born in the US Virgin Islands were between the ages of 10 and 14. ? About 48 percent of first-time offenders between the ages of 10 and 14 had mothers who were born in the US Virgin Islands, and 43.3 percent had fathers who were born in the US Virgin Islands. ? Ninety-six percent of juvenile offenders were black. xii ? Approximately 50 percent of juvenile offenders lived in a household with their mother, but without their natural father. ? Eighty-eight percent of juvenile offenders were enrolled in school at the time of their offense. ? About 91 percent of the offenders were attending junior high or high school at the time of their offense, while 9 percent were in elementary school. ? Among all offending juveniles, 40 percent were children of mothers born in the Eastern Caribbean. ? Seventy-three percent of mothers of juvenile offenders were employed at the time of their arrest. Predicting Recidivism with Juvenile Investigation Bureau Data ? For every increase in age by one year (between 7 and 17 years) the odds of being a repeat offender more than triples, but the chances of recidivism increase at a decreasing rate. ? Males are almost twice as likely as females to be repeat offenders. ? Offenders born in the US Virgin Islands or in the Eastern Caribbean are almost twice as likely to be recidivists as children born in the United States, whereas juveniles born in Latin countries are about one-and-a-half times as likely to be recidivists as United States born children. ? Juveniles with one offense who live with a mother and a step-father or with foster parents are nearly twice as likely to be repeat offenders as first-time offenders who live with their natural parents (singly or together), a father and step-mother (singly or together) or with relatives. ? First-time male juvenile offenders, between 8 and 15 years old, born in the US Virgin Islands and living with a mother and step-father or with foster parents, have a relatively high risk of being repeat offenders. ? Females most at risk of recidivism are those who are 12-years old, born in either the US Virgin Islands, in a Latin country or the Eastern Caribbean, and who live with their mother and step-father or with foster parents. xiii Department of Human Services ? Approximately 85 percent of all juvenile offenders were male. ? About 86 percent of offenders were citizens of the United States. ? Nearly 19 percent of juvenile offenders were 15 years old. ? Three out of four offenders were born in the US Virgin Islands. ? The majority of juvenile offenders—96.4 percent—were black. ? While a mother was present in the household of juvenile offenders 87.1 percent of the time, a father was present in the household only 30.6 percent of the time. ? The majority of the mothers of offenders—61.2 percent—fell within the 40-to-49- year age group, while the fathers were most represented in the 50 and over age group. ? Approximately half of the mothers of juvenile offenders were born in the Eastern Caribbean and about 60 percent of the fathers were born in the Eastern Caribbean. ? Forty-seven percent of the mothers of offenders were married and 51 percent of the fathers were married, although not necessarily to each other. ? About 74 percent of the mothers and fathers of juvenile offenders were employed. ? Nearly one out of every two juvenile offenders were living with their mother. ? Thirty-seven percent of the natural parents of offenders were never married to each other. ? About 23 percent of juvenile offenders lived in a household with either one or two siblings. ? Eighty-four percent of offenders were enrolled in school; however there was no certainty as to actual attendance and participation. ? The majority of juvenile offenders—86.7 percent—were employed. ? About sixty-two percent of juvenile offenders lived in a household with three or more children. xiv ? Ninety-six percent of juvenile offenders did not have any children of their own. ? Approximately 58 percent of juvenile offenders had been suspended from school. ? Twenty-six percent of offenders had not used drugs or alcohol. ? About 43 percent of juvenile offenders had trouble controlling their anger and 41 percent experienced frequent arguments and fights. ? Fifty-eight percent of juvenile offenders were sexually active. Predicting Recidivism with Department of Human Services Data ? STT/DHS data confirm the STT/JIB findings on recidivism: ? That each year after age seven the likelihood of a repeat offense more than triples; ? That males are over twice as likely as females to be charged with more than one offense; ? That juveniles born in the US Virgin Islands or the Eastern Caribbean are about three time as likely to be recidivist as those born in the United States; and ? That a first-time offender who lives with a mother and step-father or foster parents is more likely to be charged with at least one other offense than one who lives with natural parents (singly or together), a father and a step-mother (singly or together), or with relatives. ? Additionally, a first-time offender who admits to feelings of failure or worthlessness is twice more likely to be a recidivist than one who does not. ? Males between 8 and 16 years who were born in the US Virgin Islands and who have low self-esteem and live with a mother and step-father or foster parents are predicted to have a high chance of being a repeat offender. ? A 12-year old female born in the US Virgin Islands or the Eastern Caribbean who expresses feelings of failure or worthlessness and lives with a mother and step-father or foster parents has a high risk being a recidivist. xv ST CROIX Youth Investigative Bureau ? The largest number of recorded first-time juvenile offenses committed between 1980 and 1997 was in 1995 with a total of 753, and the second highest was 435 offenses in 1991. ? About 41 percent of first-time offenses were Part I and II felonies. ? The overwhelming majority of offenders—84.6 percent—committed only one offense; only 15.4 percent were repeat offenders. ? Two-thirds of the female offenders committed their first offense between the ages of 14 and 16. ? Two-thirds of the male offenders committed their first offense between the ages 15 and 17. ? Eighty percent of juvenile offenders were born in the US Virgin Islands. ? About 77 percent of the mothers of juvenile offenders were born in the U.S. Virgin Islands and approximately 69 percent of the fathers of juvenile offenders were born in the U.S. Virgin Islands. ? Approximately three out of every four or 75 percent of female offenders were black and about 77 percent of the male offenders were black. ? Only 3.7 percent of juvenile offenders were not enrolled in school. ? Nearly 76 percent of female offenders resided in homes where the mother only was the householder. ? Approximately four out of every five male offenders lived in homes where the mother only was the householder. ? About 71 percent of first-time offenders were attending high school. xvi Predicting Recidivism with Youth Investigation Bureau Data ? YIB data in St Croix suggest that: ? Males are nearly twice as likely to be repeat offenders than females; ? Offenders enrolled in school at the time of their first charge are between three and four times more likely to be recidivist than those who are not enrolled; ? Juveniles who commit their first offense while in Grade 12 have more than twice the likelihood of being charged again than those who are not in Grade 12. ? An offender who has a relative that has been previously charged with a criminal offense is almost seven times more likely to be a recidivist compared to one who does not have such a relative. Department of Human Services ? About 86 percent of juvenile offenders were male. ? Approximately 94 percent of juvenile offenders were classified as a citizen of the United States. ? Juvenile offenders were most represented in the 16 year age category. ? The majority of the offenders—84.6 percent—were born in US Virgin Islands. ? About 98 percent of the juvenile offenders were black. ? Only 27.8 percent of the offenders had a father as a householder, but 86.1 percent lived in homes with a mother as a householder. ? More than half of the mothers—51.1 percent—of juvenile offenders fell in the 30-to-39- year age group, while the majority of fathers were in the 40-to-49-year age group. ? The majority of juvenile offenders had mothers—41.1 percent—and fathers—57.3 percent—who were married, although not necessarily to each other. ? More than half of the juvenile offenders—51.8 percent—lived with their mother only at the time of their first offense. xvii ? Forty percent of the natural parents of juvenile offenders were never married to each other. ? The majority of offenders live in a household with less than three siblings. ? About 78 percent of the offenders were enrolled in school at the time of their first offense. ? The highest percentage of juvenile offenders—26.8 percent—was in Grade 7 at the time of their first offense. ? Only 15.7 percent of the juvenile offenders were employed. ? Nearly half of the offenders lived in a household with three or more children. ? Approximately three percent of offenders had children. ? About 2 out of every 10 offenders were responsible for childcare. ? About 36 percent of the juvenile offenders expressed feelings of failure or personal worthlessness. ? Nearly 50 percent of offenders admitted that they had used drugs or alcohol. ? The majority of offenders—66.5 percent— reported they had been suspended from school at some point, while only 4.9 percent had been expelled. ? About 29 percent of the juveniles offenders indicated that they had been hurt or physically abused. ? More than half—57.9 percent—of the juvenile offenders reported themselves to be sexually active. ? About 9 percent of the offenders had been forced to have sex. ? Approximately six percent of the offenders admitted that they did not know their biological father. xviii Predicting Recidivism with Department of Human Services Data St Croix/DHS data suggest that: ? Males are over twice as likely as females to be recidivists. ? A juvenile who lives with his or her married parents has less chances of being a repeat offender than those who do not. ? First-time offenders who harbor feelings of being a failure or being worthless have almost twice the risk of being charged again compared to those who do not harbor such negative feelings. ? Juveniles who use drugs or alcohol have odds of recidivism that are nearly three times higher than those who do not use these substances. ? Males between 8 and 16 years old who do not live in a home with married parents, take drugs or use alcohol and have low self-esteem are predicted to have a high risk of recidivism. 1.1 Background to the Study 1.1.1 Historical and Social Change Before the US Virgin Islands achieved its cur- rent status as a United States territory, it had gone through a series of political and adminis- trative changes that impacted the social compo- sition of the islands. The Spanish, Dutch, Eng- lish and French each played a role in early colonization before Denmark, the longest occu- pying colonial power, took possession from around 1665 until 1917 when the United States purchased the territory for strategic purposes. Political and economic changes since the 1950s have moved the islands from being solely of strategic military importance and a primary ag- ricultural society, to a more diversified econ- omy. Tourism has become the main economic activity on St Thomas/St John, while oil and refining and spirits manufactures are prime economic activities on St Croix. This more diversified economy encouraged in- migration from United States citizens wishing to experience the novelty of island life or from entrepreneurs seeking economic opportunities. Persons of nonresident status, willing to take advantage of employment opportunities, were brought in as bonded workers. Most emigrated from the Caribbean, especially the neighboring English-speaking Eastern Caribbean islands. It was difficult for families of these bonded work- ers to join them at any stage. These workers had no residential rights beyond the temporary status that a job assured. Later, the Immigration and Nationality Act of 1965 made provisions for the reunification of families. This Act opened the floodgates for thousands of immi- grants to take up residence in the Virgin Is- lands. Residents of African descent, who had been brought to the territory as slaves for plantation labor, comprised the largest sector of the popu- lation and remained from inception mainly at the base of the socio-economic pyramid. Entrepreneurs of Asian descent have been drawn to the commerce associated with tour- ism. These immigration trends, together with a steady and often heavy flow of immigrants from the rest of the Caribbean, have produced a population that is highly diverse, both ethni- cally and culturally. 1.1.2 Some Effects of Immigration The movement of large numbers of people with differing ethnic and socio-economic back- grounds into an already established community has always been marked with accompanying changes, resistance and social stresses as each group contributes its own uniqueness. The majority of US mainland migrants into the territory have been white, and of middle- income levels or above. These migrants tend to reside in particular suburbs that are more afflu- ent. This injection affected the social balance and tension in the society, though no recent psychological studies have been carried out to determine the impact of these effects. Migrants from the rest of the Caribbean, espe- cially the Eastern Caribbean, have been usually economically diverse, and settle in a variety of areas. Some migrants join family and may move into equivalent or better jobs and social positions as those from which they have left The majority of this group, however, seeks in- creased employment opportunities and hope for an improvement in the standard of living that access to the United States is expected to pro- I. INTRODUCTION 2 vide. Table 1.1 shows population changes in the US Virgin Islands between 1980 and 1997 based on country of origin. The percent change col- umn was calculated from 1980 and 1997 col- umns, since more representative data were available for these years. Over the 17-year pe- riod, the overall population increased by 16.9 percent. The number of persons born in the US Virgin Islands increased by 34.7 percent over the same period. The increase from the Other Caribbean region accounted for 42.7 percent. In recent years there has been low, or in some cases diminishing, immigration from some is- lands such as Puerto Rico, St Lucia and Trini- dad and Tobago. This is due possibly to the sig- nificant improvement in the economies of these islands. The British Virgin Islands, for exam- ple, has experienced a reversal of the migration trend in recent years with a movement of the children of former immigrants returning there in search of a better life. However, there con- tinues to be increased immigration from Carib- bean countries such as the Dominican Republic (591.4 percent) and Dominica (67.1 percent). The number of persons who became natural- ized citizens increased by 62.7 percent during this period. A number of social and economic characteris- tics are associated with immigrant populations, especially those migrating in large numbers under difficult conditions. The following is an attempt to outline some of these characteristics that may be applied to immigration patterns in the US Virgin Islands. 1. The most important factor in making a deci- sion on where to locate usually depends on affordability of housing. New immigrants tend to settle near urban centers in low- quality and often substandard housing, with easy access to affordable public transporta- tion. They may be attracted to areas where persons from their homeland already reside. Availability of public housing, often with less than desirable conditions, may also be an option. 2. Different morals and values from those ex- perienced in the immigrant’s homeland usu- ally prevail in the new environment and must be learned, often rapidly, in order to facilitate social adjustment in the adopted community. In the absence of a systematic or monitored introduction to the new ways of life, children are often confused and overwhelmed. They may be exposed to ex- periences and behavior formerly considered undesirable. In addition, more readily avail- able money, illegal drugs and weapons, as well as unsupervised opportunities for ex- perimentation, leave children exposed and vulnerable. At the same time, children born of newly-arrived immigrants may be seen as passports to helping older migrants es- tablish themselves as belonging. This is es- pecially seen in the case of immigrants who speak a foreign first language and possess weak communication skills in English. Un- fortunately, such children may not always be embraced quickly as individuals with full rights into the receiving community. 3. Employment opportunities for new immi- grants who are usually unskilled tend to be in jobs that local residents regard as inade- quate in wages or lacking in social status. Employment may also be in rapidly grow- ing sectors of the economy with an insuffi- ciently available domestic labor force. Wages are usually —minimum wage or be- low in the case of undocumented immi- grants. It may be necessary to take several of these low-paying jobs in order to boost wages. This situation exerts more pressure on families, who find they must leave their children on their own for long periods of time while working one or two jobs. This minimizes the time for socialization. Mi- INTRODUCTION 3 nors are also left up to their own devices regarding such behavior as the choice of television programs to watch or the group of friends to lime with. Family support sys- tems that might have been available in the home country become disrupted or nonexis- tent and services such as childcare may be unaffordable. In the new setting, guardians Percent Place of Birth and Citizenship Number Percent of Total Number Percent of Total Number Percent of Total Number Percent of Total Change 1980-97 Total Persons 96,569 100.0 101,809 100.0 109,677 100.0 112,863 100.0 16.9 Born in US VI 43,234 44.8 49,839 49.0 55,389 0.5 58,256 51.6 34.7 St Croix … … 24,991 24.5 26,375 24.0 26,948 23.9 7.8 St John … … 698 0.7 675 0.6 820 0.7 17.5 St Thomas … … 24,150 23.7 28,329 25.8 29,707 26.3 23.0 Born in Puerto Rico 4,993 5.2 3,974 3.9 4,468 4.1 4,800 4.3 -3.9 Born in US 11,964 12.4 14,246 14.0 12,689 11.6 11,814 10.5 -1.3 Born Abroad of US Parents 282 0.3 1,032 0.0 389 0.4 418 0.7 48.2 Naturalized Citizens—US 7,819 8.1 14,361 14.1 18,894 17.2 12,723 0.1 62.7 Not a Citizen of the US 22,912 23.7 17,464 17.2 17,848 16.3 15,194 0.1 -33.7 Other Caribbean 25,327 26.2 25,609 25.2 34,330 31.3 36,131 32.0 42.7 Anguilla 1,210 1.3 899 0.9 1,073 1.0 1,412 1.3 16.7 Antigua/Barbuda 4,939 5.1 4,398 4.3 5,192 4.7 6,382 5.7 29.2 British VI 3,315 3.4 2,665 2.6 4,007 3.7 2,990 2.6 -9.8 Dominica 2,653 2.7 3,219 3.2 4,344 4.0 4,433 3.9 67.1 Dominican Republic 580 0.6 1,157 1.1 2,553 2.3 4,010 3.6 591.4 Montserrat 705 0.7 623 0.6 621 0.6 811 0.7 15.0 St Kitts-Nevis 6,523 6.8 5,828 5.7 7,932 7.2 8,113 7.2 24.4 St Lucia 2,713 2.8 2,533 2.5 2,603 2.4 2,604 2.3 -4.0 Trinidad & Tobago 2,689 2.8 1,837 1.8 2,699 2.5 1,667 1.5 -38.0 Elsewhere/Not Reported 5,365 5.6 7,871 7.7 2,801 2.6 1,861 1.6 -65.3 Eastern Caribbean Center, University of the Virgin Islands, 1997 Consumer Expenditure Survey: Population and Housing. Table1.1 Place of Birth by Year: 1980, 1990, 1995, 1997 Source: Bureau of the Census, Detailed Population Characteristics: Virgin Islands of the United States (PC80 - 1-D55). Bureau of the Census, Detailed Cross-tabulations for the US Virgin Islands (CPH – L – 156). Eastern Caribbean Center, University of the Virgin Islands, 1995 Population and Housing Survey. 1980 1997 1990 1995 INTRODUCTION 4 or parents with feelings of guilt may also allow lax, permissive behavior not usually accepted in their homeland in order to com- pensate the neglected child. 4. Immigrant families are often split up with the breadwinner emigrating first, and then gradually arranging for the rest of the fam- ily members to join as socio-economic con- ditions improve or as immigration policy allows. This puts a strain on family ties and relationships and may lead to continued tensions. Separations often occur even after the family is physically reunited. 5. Systems and institutions adjust slowly to change caused by the influx of immigrants. Thus, they may lag behind in providing so- cial services to these families until needs are manifested in a series of crises such as incidents of violence. This is seen in the case of juvenile delinquent behavior. A heavy burden is placed on institutions pro- viding services such as education, health- care, law enforcement and social welfare. The inadequate planning, reluctance to ac- commodate the increase in numbers until it is unavoidable, guarantees an escalation in the rate of social problems experienced in the community. 6. Even after institutions implement policies and procedures to accommodate immi- grants, there is usually a period of adjust- ment on both sides before social acceptance and assimilation into the new community are achieved. The life of this period can vary and may last for decades. 1.1.3 Risk Factors Affecting Juveniles Researchers have identified a number of impor- tant insights into juvenile delinquent behavior. Youth involved in such behavior are usually from economically stressed families and com- munities, have histories of physical and sexual abuse, lack educational and vocational skills and are prone to become involved in alcohol and other forms of drug abuse. They may be failing or dropping out of school, and involved in unprotected sex. Youth agencies have, for decades, used assess- ment instruments in evaluating high-risk youth. The following are seven background and per- sonal characteristics used by social workers and counselors to predict high-risk behavior. The higher the level or occurrence of these crucial characteristics, the higher the child’s risk-level for delinquent behavior: 1. Age variable: The child enters the correctional system at a very young age. 2. Psychological variable: Rebelliousness and identification with non- conformity, together with a low level of self-esteem. 3. School performance variable: Low achievement levels, undesirable social behavior, and truancy. 4. Home Adjustment variable: Poor interaction between the child and par- ents and between the child and siblings, curfew violation, failure to respond to disci- pline and supervision, and running away. 5. Drug and alcohol use variable: The child starts to abuse drugs at an early age, there is abuse of serious drugs, parents have a history of alcohol and drug abuse. 6. Neighborhood variable: The child lives in a neighborhood charac- terized by disorganization, poverty, and multiple social problems. 7. Social Adjustment variable: The child’s friends are involved in problem INTRODUCTION 5 behavior such as delinquency, drug abuse, truancy and disruption in school. There are displays of sexual acting-out and associa tion with gangs. 1.1.4 At-Risk Indicators in the US Virgin Islands Table 1.2 provides a comparison between chil- dren under the age of 18 in the US Virgin Is- lands to those children in the United States on various possible indicators of juvenile delin- quent behavior. Examination of the different indicators suggests that for the most part chil- dren in the US Virgin Islands are exposed to a higher level of risk than children in the United States. While Table 1.2 refers to all children under the age of 18, the findings presented be- low are limited to juvenile offenders 7 to 17 years of age. 1.1.5 Youth Crime Trends US Youth Crime Trends: Between 1990 and 1994 the violent crime index for juveniles un- der 18 years old increased by 26 percent, repre- senting four times that for individuals age 18 and over (Bartollas, 1997). Over the same pe- riod, juveniles’ arrests for murder increased by 14.6 percent compared to 4.6 percent for adults. INTRODUCTION Table 1.2 Possible Indicators of Juvenile Delinquent Behavior Indicator Status of USVI Status of Nation Low birth-weight babies 8% 7% Infant deaths per 1,000 live births 11 7 Families with children <18 years old headed by a single parent 44% 27% Death of children 1 to 14 years old per 100,000 42 26 Children in poverty 41% 20% High school dropout rate of children 16 to 19 years old 22% 10% Death rate of children by accident, homicide or suicide 15 to 17 years old per 100,000 82 62 Teen birth rate females 15 to 17 years old per 1,000 39 34 High school dropouts not working 16 to 19 years old 5% 10% Violent crime rate: arrests of youth 10 to 17 years old per 100,000 310 471 Median income of families with children 0 to 18 years old $30,530 $39,700 Child abuse and neglect 0 to 18 years old per 1,000 8 NA Source: Community Foundation of the Virgin Islands. (2000). US Virgin Islands Kids Count Data Book 2000. St Thomas, VI: Author. 6 A 1992 study found the arrest rate for African- American youth was five times higher than the rate for white youths (Bartollas, 1997, p.33). The use of guns by juveniles in homicides in- creased from 64 percent to 78 percent between 1987 and 1991. Juveniles’ arrest for weapons law violations during this period also increased by 62 percent. Status of Crime in US Virgin Islands: Accord- ing to the 1990 Census of Population and Housing, the St Thomas/St John community had approximately 10,432 persons ages seven to 17, the age group that constitutes the juvenile category as designated by the police depart- ment. Crimes involving excessive violence, especially when committed in public places such as schools or in the full view of adults, usually cause alarm and public outcry within the com- munity. During the 1990s the US Virgin Islands seemed to be following a nationwide trend in an escala- tion of violence by juveniles carried out on fel- low juveniles, caretakers, teachers and others. These assaults took place in neighborhoods by gangs or on school premises. Violence so seemingly deliberate, brutal at times and even fatal, left communities searching for an under- standing of the cause of violence and a method of prevention. For almost every year from 1989 to 1997, sta- tistics from the Department of Public Safety show a consistent number of offenses commit- ted by juveniles in the St Thomas/St John dis- trict. These cover the range from trespassing and auto theft to rape and homicide. The community has been forced to ask itself a number of obvious, as well as not so obvious, questions. How many juveniles are involved in criminal behavior? Are there characteristics or profiles that may help to identify these juve- niles? Are the conditions to which these juve- niles are exposed in the home, school or com- munity contributing to delinquent behavior? Where these conditions exist, are they in the form of commission––the presence of some negatively influencing factors in the young per- sons’ lives or omission, the absence of positive factors? Handling Juvenile Delinquency: The definition of delinquency includes the portrayal of delin- quent events, the reform and punishment of de- linquents, and policy decisions about delin- quency that occur in a particular setting. This social setting in turn is shaped by five contexts: historical, legal, socio-cultural, economic and political. All are interwoven to affect the whole. Historically, a community’s past methods of handling juvenile delinquents influence its cur- rent perception and methods. This past ap- proach could well extend to an influence of his- toric factors that may have affected juvenile justice, rise and decline of beliefs and practices that govern cycles of delinquent behavior. Legal context is set by the role and jurisdiction assigned juvenile courts, as well as constitu- tional obligations to which courts will hold the community responsible for dealing with youth in trouble. Socio-cultural context will be determined by the effectiveness of societal institutions in in- teraction with the delinquent. This reflects the extent that the family, school and religious or- ganizations, different lifestyles, peer groups, and the process of urbanization all play in con- tributing to juvenile problem behavior. Economic status and the ability to succeed at economic goals affect youth and influence their behavior. This is especially the case in a society where success and winning are held at such a premium. INTRODUCTION 7 Politically, although local conditions may exert influence, national and institutional policy posi- tions also affect youth attitudes toward crime and shape levels of severity or leniency for punishment (Bartollas, 1997). 1.1.6 US Virgin Islands Youth Rehabilitation Center1 A thorough audit of the Youth Rehabilitation Center (YRC) was carried out by the Office of Insular Affairs (OIA). A study was also com- pleted by the Federal Department of Justice, Office of Juvenile Justice Delinquency Preven- tion. The studies identified concerns such as inadequate staffing, limited treatment and lim- ited education opportunities for youth housed at the YRC in St Croix. The original facility was designed to cater to 27 male adolescents. Between 1991 and 1994, however, the average number increased from 20, to 134. There were no policies and proce- dures to regulate the facility. One social worker comprised the treatment staff and there were no health services in place. In addition, no special- ized training was provided for those working with adolescent offenders. The Territory also experienced an increase, not only in the number of juvenile delinquents, but also in the levels of crime. This follows na- tional and international trends. The focus of the Department of Human Ser- vices (DHS) in regard to the YRC since 1994 has been to put policies and procedures into effect for the facility. DHS has worked to im- prove community-based sentencing options, residential programs and effective interventions for juveniles within the community. These in- clude early intervention programs such as Head Start, child care and after-school programs, youth promotion programs for teenagers and parenting that positively impacts children. All youth at YRC are ordered to the facility by the court. Although the Virgin Islands Code permits Persons in Need of Supervision (PINS) to be mixed with adjudicated youth, this is con- trary to federal regulation. The practice puts the availability of federal funds for the facility in jeopardy. Adjudicated youth should not be housed within sight or sound of an adult of- fender. That includes adolescents transferred to be tried as adults or sentenced as adults. In FY 1999 the facility housed 17 such adjudicated youth with no provisions made within the facil- ity for the required separation. In March 2000 out of 55 youths, 14 to 25 percent were adoles- cents transferred as adults, (seven for murder and four for rape). All youth at the YRC facility are adolescents ages 13 to 18. However, a judge may—under special circumstances—retain a youth at YRC until that person turned 19 years of age. This is mostly for serious offenses. There were four youth aged 18 by May, 2000. Although the fa- cility was designed to accommodate 27 adjudi- cated delinquents at a time, the level of occu- pancy averaged 40 to 50. This overcrowding was due to the presence of youth transferred as adults who continued to be housed in the facil- ity. The facility was never designed to house fe- males and never did until 1992 when two fe- males were placed there. Subsequently, from 1994 to 1999 there were 20 or more females remanded to the facility yearly with as many as 34 placed there in 1997. As of the report date, 56 youth were housed in the facility. Eleven of them—20 percent—were females. In FY 1999, the YRC served three population types: pre- trial juveniles who were in contempt of court orders, adjudicated juveniles and youth trans- ferred as adults (Table 1.3). 1The following is a paraphrased outline of a presentation by Mrs. Sidonie Halbert, Commissioner, Department of Human Services (DHS), Division of Children, Youth and Families, to the Virgin Islands Legislature on May 17, 2000. INTRODUCTION 8 In FY 1999, the youth population remanded at the YRC had the characteristics shown in Table 1.4. In addition to those 155 juveniles re- manded at the YRC, there were 411 who, though not remanded at YRC, were being ser- viced by the system (Table 1.5). Data from the 1995 survey indicated that there were 12,528 teenagers between ages 13 to 18 in the Territory. The 566 in the DHS system (155 remanded at YRC, 411 not incarcerated) ac- counted for 4.5 percent of that youth popula- tion. YRC Expansion: Plans are in progress for the improvement of the facility that would include the following: a. Expansion of the YRC by a construction project of three phases that will result in a 51 bed dormitory facility. b. Development of a Pre-release Center com- prising an eight-bed pre-release home. c. A total of four trailers for use as office and classroom space. d. Security fence to replace the fence damaged by Hurricane Marilyn in 1995. e. Staffing included 54 Full-Time (FT) DHS staff at YRC with an additional 3 FT and two Part Time (PT) teachers from the De- partment of Education. Services and Activities: Educational activities were provided by one FT teacher from DHS who works all year round and five teachers from the Department of Education (three FT; two PT). Activities included: educational as- sessment, GED preparation, tutoring, special education and vocational education in the areas of air-condition/refrigeration and upholstery. Recreation: YRC had a full-time recreational specialist who coordinated outdoor activities such as basketball and softball and various in- door activities such as board games and pool. Treatment Services: Two full-time licensed so- cial workers coordinated with the Intervention bilitation Center: FY 1999 Number Percent Total 155 100 Pre-Trial Juveniles in contempt Court Order 57 37 STT/STJ 21 37 STX 36 63 Adjudicated Juveniles 81 52 STT/STJ 38 47 STX 43 53 Youth transferred (as adults ) 17 11 STT/STJ 8 47 STX 9 53 Note: Youth transferred as adults (both pretrial and sentenced) Table 1.3 Three Population Types Served by the Youth Reha- Source: Department of Human Services (2000). Table 1.4 Characteristics of Youths Remanded at YRC FY 1999 Characteristics Total Percent Youths Remanded at YRC 155 100 First time admission 87 56 Single parent household 58 37 Tested positively for drug use upon admission 57 37 School dropouts at admission 34 30 Diagnosed learning disability 20 13 Living at home at time of arrest 81 52 Arrested for Violent Crime 46 27 INTRODUCTION 9 Unit for intake and discharge and provided all court work and counseling. An average 80 youth counseling sessions were provided each month, along with six court reports and 16 court hearings. Residential counselors also worked with juvenile residents, listening to their concerns and helping to develop alterna- tive ways to deal with anger, depression and conflicts. There is an agreement between DHS and the Virgin Islands Behavioral Services (VIBS) to provide health services and supple- mental mental health services to youth at YRC. Following are some of the areas for which ser- vices were contracted in FY 1999: Mental Health: psychiatric evaluations, consul- tations, psychotherapy sessions, emergency, psychological evaluations. General Health: physicals, medical histories, emergency visits, hospital referrals, nursing care, medication provision, laboratory ser- vices/tests, pharmaceutical supplies, screening. Special Concerns: There are a large number of youth who are testing positive for drugs, usu- ally marijuana, upon intake. Plans have been discussed to use monies available from the Law Enforcement Planning Commission (LEPC) to renovate a facility to house drug treatment pro- grams. There was a recognized need for full evaluations for all youth entering the facility that may have special needs. 1.1.7 Department of Human Services History: Aid to the poor and needy of the US Virgin Islands was first provided from trust funds established under the Danish colonial administration and administered by local “Charitable aid Councils” on each island. In 1922, under the United States Navy administra- tion, a Department of Public Welfare was es- tablished by order of the Naval Governor, and upon direction of an admiral’s aide. With the transition to a civilian administration, all wel- fare functions were administered in the Gover- nor’s office, with the first civilian Superinten- dent of Public Welfare appointed in 1931, and the first native appointed to that position in 1933. This person also coordinated services provided to the Virgin Islands under the Fed- eral Emergency Relief Act of 1933. In 1943 the Virgin Islands Welfare Act estab- lished the Department of Social Welfare for providing public social welfare services on a territorial basis. It was a legal base for the ex- tension of federal social security programs to the US Virgin Islands. Federal child health and welfare services were extended to the territory in 1947; public assistance in 1950; old age and survivors insurance benefits in 1951; Medicaid benefits in 1966; the Food Stamp program in 1974; and the Child Support Enforcement pro- gram in 1975. A local medical assistance pro- gram, which was operated by the Department of Social Welfare since 1961, was absorbed by Title XIX—Medicaid—and transferred to the Department of Health in 1966. Vocational reha- bilitation services were transferred from the Department of Education to Social Welfare in 1969, and the Child Support Enforcement pro- gram was transferred from Social Welfare in the Department of Law in 1983. The Government Reorganization and Consoli- dation Act of 1987, VI Act 5265, established Table 1.5 Juveniles Remanded Outside YRC FY 1999 Juveniles Remanded Outside YRC Total Percent Total 411 100 Adjudicated Juveniles 337 82 STT/STJ 215 64 STX 122 36 PINS 74 18 STT/STJ 60 81 STX 14 23 INTRODUCTION 10 the Department of Human Services (DHS) to provide all services for youth, children, handi- capped, elderly and adults and families in the low-income bracket. Programs were consoli- dated from the former Department of Social Welfare, Office of Community Services, Virgin Islands Community Action Agency, Youth Ser- vices Administration, Commission on Youth, Commission on Aging, Commission on the Handicapped and the Developmental Disabili- ties Council, all of which were abolished by the Act. In addition, three programs for the elderly were transferred from other departments. Mission: The mission of the Department of Hu- man Services is to provide a sensitive, efficient, responsive human services delivery system to meet the needs of low-income persons, the eld- erly, children and families, and the disabled in the Territory. Through six program divisions, the department provides corrective, supportive and preventive programs, as well as advocacy to identify human needs. DHS develops inno- vative ways to break the cycle of dependency, and help its clients become self-sufficient, con- tributing members of the Virgin Islands’ com- munities. Vision: The DHS envisions its emergence as a department that empowers individuals and families toward self-sufficiency through a seamless delivery system of services. 1.1.8 Juvenile Units Background: The Juvenile Unit of the St Thomas/St John Police Department was es- tablished in February, 1969. The Juvenile Unit of the St Croix Police Department was estab- lished on December 1, 1969. A senior detective with extensive experience in juvenile crime was assigned as juvenile officer to head the unit. The unit offers many varied services to the community and its citizens. In accordance with the Code of Laws of the Territory of the US Virgin Islands, a juvenile is a person under 18 years of age. Mission: It shall be the mission of the Juvenile Unit to: a) Be on the lookout for potential delinquents and for conditions that cause delinquent be- havior. b) Work with other agencies to correct environ- mental conditions that may foster delin- quency. c) Use preventive patrol where it has potential for helping to control delinquency. d) Investigate delinquency problems that lead to juvenile offenses. This includes apprehen- sion and prosecution of adults involved in these cases. e) Investigate and prosecute all complaints of child abuse from the home when conditions warrant. f) Detect, and take into custody, juvenile of- fenders for both criminal and status offenses. g) Follow up on juvenile offenders, using refer- ral and other dispositions as appropriate. Duties: It is the duty and responsibility of the Juvenile Unit to: a) Interview all juvenile offenders and victims in accordance with the law. b) Investigate all criminal and status offenses involving juveniles. c) Determine which juvenile offenders will be placed on diversion programs and which ju- veniles will be referred to Family Court. d) Prepare court complaints and provide assis- tance to the Attorney General at hearings and cases involving juveniles before Family Court. e) Investigate and prepare for prosecution, when applicable, all cases of child abuse or neglect. The unit may also render assistance in child abuse or neglect cases on the island when requested by other agencies and ap- proved by the Head of the Juvenile Unit. f) Prepare Uniform Crime Reports (UCR) and INTRODUCTION 11 ensure proper data concerning missing or runaway juveniles. g) Conduct in-service training on juvenile laws and procedures for members of the depart- ment. h) Assist the Bureau of Criminal Investigation in criminal cases when the juvenile caseload permits. i) Conduct juvenile workshops; give lectures and talks to adults and students on all phases of juvenile law, crimes and abuses of the child. These programs will include visual aids and a realistic approach to the growing juvenile problems in our community and na- tion. j) Plan, organize, sponsor, or co-sponsor sum- mer youth information, recreational and safety programs, i.e. Police Movie Club, Op- eration Shipmate and bicycle rodeos. Such police information programs will teach the youth safety and responsibility in the home, family and community. k) Seek out and identify needy children and families who deserve food and toys fur- nished by military or civilian groups for dis- tribution during the holiday season. This is an on-going yearly program. l) Work closely with school officials and teachers to deter and suppress juvenile delin- quency, crimes, and disorders on school campuses. Duties and Responsibilities of the Commander: In addition to the general authority duties and responsibilities of command and supervisory personnel enumerated in Section 3.3.4 of this chapter, it is the responsibility of the Com- mander of the Juvenile Bureau to: a) Develop a program of delinquency preven- tion intended to eliminate factors that induce criminal tendencies and conditions that pro- mote criminal activities among juveniles. The commander shall promote such a pro- gram by enlisting the aid of the public, inter- ested agencies and other divisions and bu- reaus of the Department. The commander shall coordinate the functions of his bureau with that of the Territorial Court, Youth Ser- vices Administration, Department of Wel- fare and other organizations concerned with the social welfare of juveniles. b) Properly investigate and process all cases of delinquency coming to the bureau’s atten- tion. In addition, the commander should have members of his bureau patrol all places where delinquency is likely to occur, to de- tect any violations of law which involve ju- veniles and which may contribute to vice or delinquency. Also, the commander should ensure that bureau members assist members of other divisions, bureaus, sections and units, as well as community agencies and organizations involved in the prevention of juvenile delinquency and youth crime. c) Accept all juveniles taken into custody by any member of the Department; establish an adequate system of recording complaints, contacts, arrests, investigations and deposi- tions involving juveniles; and maintain a complete and accurate file of all cases re- ferred to the Juvenile Investigation Bureau for action, indicating its status. d) Collect data, discover and chart trends, stimulate juvenile protective measures, and work to improve general community condi- tions. The commander should work to elimi- nate those factors contributing to delin- quency. Also, the commander may interpret functions of the Department as a whole, and the Juvenile Bureau in particular, in the sup- pression of delinquency, when ever the com- mander or any member of the bureau has occasion to speak before any civil, educa- tional or religious groups. e) Have a thorough knowledge of the laws regulating the handling of juveniles, and in- struct and ensure the obedience of all such laws by members of the bureau. f) Report to the Deputy Chief of Police all im- portant activities of the bureau, and keep the deputy advised of the youth crime situation in the district, the effectiveness of the bu- INTRODUCTION 12 reau, morale, disciplinary problems or needs of the bureau. Types of Juvenile Offenses There are two types of offenses peculiar to ju- veniles: a) Criminal Offenses: The criminal offense is a crime against persons and/or property and is the same offense as the adult criminal of- fenses. Upon request, the juvenile, after re- ceiving a waiver hearing before a Family Court Judge, may be referred to the General Sessions Court. b) Status offense: the status offense is a non- criminal offense such as a runaway or tru- ancy. 1.2 Purpose of the LEPC Study In the Memorandum of Understanding (MOU) between the Virgin Islands Law Enforcement Planning Commission (LEPC) and the Univer- sity of the Virgin Islands (UVI) the following were outlined as the purpose of the study: a) Develop a statistical profile of delinquent youth, with the primary aim of adopting a proactive approach to the prevention and reduction of the escalating level of crime among youth; b) Collect data on delinquent youth between 1987 and 1997 from the files in the Virgin Islands Department of Human Services (DHS) and in the Juvenile Investigation Bu- reau of the Virgin Islands Police Department (JIB); c) Develop associated attributes of delinquents; d) Determine what are the significant character- istics which explain juvenile delinquency; e) Measure the degree of association between abused juveniles and delinquency; f) Establish characteristics that distinguish a first-time offender from a repeat offender; g) Measure the attributes of first-time and re- peat offenders h) Determine the time frame between the first and second arrest for repeat offenders; i) Measure the effectiveness of intervention services; and j) Develop a prognostic evaluation model that will assist caseworkers in determining the likelihood of evolving from a first-time to a repeat offender. LEPC was charged with developing profile in- formation on young offenders in the Territory. Through a coordinated and cooperative effort between the three working parties involved— UVI, Department of Human Services and Juve- nile Investigation Bureau—expertise and per- sonnel from UVI were commissioned to con- duct the study and gather the needed profile information. Services Performed by UVI: UVI was commis- sioned to perform the following services in ful- fillment of the purpose of the MOU. a) The design of two instruments to collect spe- cific data from the files of JIB and a sample selected case files from DHS, primary data items; b) The preparation of a computer screen that will be used in keying the data from the in- struments; c) Editing recorded data in the instruments be- fore data entry takes place; d) Generation of descriptive statistics of delin- quents and associated characteristics from the data; e) Identification of appropriate statistical ana- lytic techniques for the range of analyses to which the data will be subjected; f) Generation of appropriate tables for presen- tation; g) Application of statistical models and inter- pretation of output; and h) Preparation and printing of a final report for presentation. INTRODUCTION 13 1.3 Scope of the Study Among other things, the proposed study will seek to determine the impact of four fundamen- tal socioeconomic issues on juvenile delin- quency: a) the composition of the household in which children live—such as the number of par- ents, grandparents and other relatives, and the number of siblings; b) the educational attainment of the parents; c) the labor force behavior of the parents, espe- cially changes in employment of mothers; and d) poverty status of the household. Initially, an estimate of 4,500 case files at JIB and 5,000 at DHS spanning the 11-year period of 1987 to 1997 were believed to exist. At first it was anticipated that a scientific sample would be necessary. On closer examination, however, it was discovered that the estimates of available case files were highly exaggerated. The deci- sion was then taken to utilize all the cases available at JIB and match them with the corre- sponding cases from DHS. Two survey-type instruments were developed to record data from each selected arrest or de- linquent case. These instruments sought to gather data based on the following categories: From the Juvenile Investigation Bureau: Sex Age Date of birth Place of birth Race Birthplace of parent/guardian or caretaker School enrollment Education (highest level) Employment status of guardian Date of crime/incident Type of crime/incident Type of weapon used Status (outcome/disposition). From Department of Human Services: General Information: Sex Age Date of birth Place of birth Race Date of crime/incident Type of crime/incident Type of weapon used Status (outcome/disposition) Immigration status Mother in the household Father in the household Number of siblings in the household Education (highest level) Occupation Marital status Number of children Responsible for caring for children Type of live-in program, if any Head of household Marital relationship of natural parents Education of primary caretaker(s) Income (gross per month). Psychological Adjustment: Self esteem Frequent arguments and fights with oth- ers Age at first contact with police Use of drugs/alcohol Use of drugs/alcohol before or during school Sexual activity Number of different sex partners in past year Physically abused Case files of juveniles in both JIB and DHS are confidential and may be seen only by sworn staff members. Volunteer in-house staff mem- bers were compensated to do the data recording after their regular work hours. The data instru- ment for each case was edited to ensure that items of information were assigned numeric INTRODUCTION 14 codes. These codes were entered into an Inte- grated Microcomputer Programs Systems (IMPS) application for tabulation. 1.4 Survey Goals With a growing number of youth becoming in- volved in delinquent behavior in the US Virgin Islands, it became imperative that some under- standing of factors contributing to this problem be investigated. Although extensive studies have been con- ducted on the US mainland that point to cause and effect of juvenile delinquent behavior, to date no study with an in-depth scientific ap- proach has tackled this problem in the US Vir- gin Islands. This study serves as a first in gath- ering available data on juveniles that have passed through the correctional/counseling sys- tems. The data has been processed in a form that will facilitate analysis. By examining the data for correlations between variables, some conclusions can be suggested regarding the fac- tors that contribute to juvenile delinquent be- havior. It is anticipated that a number of indicators will be established and a framework developed on which to build an ongoing analysis of the prob- lems of juvenile delinquents. It is hoped that suggestions may be offered to assist the institu- tions working with youth in general and juve- nile delinquents in particular, to minimize the problem of delinquent behavior. It is expected that data and information pro- duced from this study will be used by legisla- tors with the resolve to identify resources to underwrite necessary preventive programs. The data can also be used by executive officers to develop policy geared to the reduction and pre- vention of delinquent behavior. It is also antici- pated that information would be released to make the public more aware of the factors play- ing a role in delinquent behavior of the terri- tory’s youth. It is expected that this first effort would also assist the JIB and DHS in both dis- tricts to fine tune their instruments of data col- lection. In so doing they can develop more effi- cient and comprehensive instruments for the purpose of the data collection. It is hoped that this study will form the basis for present analy- sis of the data gathered on juvenile delinquents, as well as provide a benchmark for subsequent comparison and study. INTRODUCTION 15 2.1 Delinquency and Antisocial Behavior 2.1.1 Causes and Recommendations In the section that follows, consideration is given to the extent to which various forms of delinquent behavior can be predicted from ear- lier conduct problems, the relationships among these behaviors, and implications of this knowl- edge for intervention focused on delinquency reduction. 2.1.2 Development, Persistence and Path- ways of Early Problem Behaviors 2 It is relatively simple to fall prey to the notion that juvenile crime is the manifestation of a sin- gle latent construct such as low self-control. If this is so, what then are the causes of, and the constraints on, the operation of low self- control? Ineffective child rearing—that is, fail- ure to set clear expectations for behavior, fail- ure to monitor children, and excessively severe and inconsistent disciplinary practices—clearly contributes to delinquent behavior. One theory suggests that differences in opportunities for personal gain may also be a cause of crime. Three aspects of the developmental continuity of delinquency will be discussed: different manifestations of the problems from childhood to adulthood, the persistence of problem behav- ior, and the impact of behavioral catalysts on the continuity or discontinuity of behavioral problems. The basic premise in the development sequence in problem behavior is that youngsters of dif- ferent ages have different capabilities and be- havioral repertoires for the expression of prob- lematic behavior. A rough approximation of a development ordering of problematic behaviors from early childhood to late adolescence, ac- cording to the earliest age at which particular behaviors manifest themselves, might conform to the following: a difficult temperament, hy- peractivity, overt conduct problems and aggres- siveness, withdrawal, poor peer relationships, academic problems, covert or concealing of conduct problems, association with deviant peers, delinquency leading to arrest and recidi- vism. A difficult temperament is one of the earliest manifestations of a problem behavior. The syndrome of hyperactivity—with its associated attention deficit problems—is usually not dis- cernible before age two or three. As soon as the child begins to walk, this opens up many new avenues of mischief. When children learn to speak, verbal aggression may be added to their behavioral repertoire. Social withdrawal may also become apparent during preschool years. The availability of peers in kindergarten may generate new problems or reinforce the aggressive problems that may already have be- come apparent. During elementary school, concealing problem behaviors such as truancy, theft and association with deviant peers may emerge. In junior high or high school, young- sters may begin to associate with deviant peers or begin to commit minor crimes. For the ma- jority of children this will be a passing phe- nomenon, but for others it will constitute a tran- sition to the frequent commission of more seri- ous delinquent activity. Contacts with police tend to accelerate during early adolescence, and so does recidivist offending. Studies of the continuity of child behavior indi- cate that certain behaviors function as catalysts in that other behaviors are prone to persist II. PERSPECTIVES AND METHODOLOGY 2Loeber, 1996. 16 when the catalyst is present; conversely, when the catalyst is absent, the problematic behavior tends to become less likely over time. Two ex- amples are hyperactivity and substance use. There is considerable evidence that hyperactiv- ity—and associated impulsivity and attention problems—and disruptive behaviors often oc- cur simultaneously. Related questions that may be posed are: To what extent is hyperactivity a catalyst for development of disruptive behav- ior? Is hyperactivity directly associated with later serious outcomes, or is the relationship conditional on the emergence of early and less serious forms of disruptive behavior? Longitu- dinal studies elsewhere of 6- to 16-year-old hy- peractive boys until age 16 have shown that those who continued to be hyperactive were far more likely to show antisocial or substance use disorder than those whose attention deficit dis- order abated. It has also been reported that hy- peractivity and conduct problems at age 8 to 10 each predicted chronic offending. It is clear that most of the evidence indicates the catalytic role of hyperactivity in the devel- opment of serious disruptive behavior in child- hood and adolescence. It should be understood that hyperactivity is often accompanied by im- pulsivity and attention problems, and that probably the impulsive element in hyperactivity is of most relevance for the associated onset of disruptive behaviors during childhood and ado- lescence. The following developmental se- quence is plausible. Hyperactivity may stimu- late an early onset of disobedience, conduct problems, and delinquency. The risk of serious conduct problems and delinquency may be- come more acute when physical aggression ac- companies disobedience at an early age. Whether a decrease in hyperactivity is associ- ated with a decrease in problem behavior is less certain. Studies have shown that the more serious the substance use, the higher the likelihood that individuals will engage in serious forms of de- linquency. One longitudinal analysis of boys from age 13 to 16 reported that the onset of drug use—marijuana or hard drugs—and deal- ing in adolescent males, was associated with an increase in person-related offenses and carrying a concealed weapon. If drug involvement and delinquency are intertwined, does this also mean that a decrease in drug use is followed by a decrease in delinquent activities? There is evidence for this: when individuals began us- ing hard drugs less frequently, their criminal involvement also decreased. In brief, hyperactivity and drug involvement may act as catalysts in the development of dis- ruptive behaviors, but the evidence is not alto- gether complete. Additional studies need to demonstrate that catalysts operate independ- ently from third factors such as impulsivity, aggression, or peer influences. It is common practice in criminology to classify youngsters on the basis of their first offense, but it can be argued that classification on the basis of a single behavior is rather limited. There is obviously more than one pathway to crime. The concept of a pathway allows indi- viduals with varying degrees of deviance to be placed on one or more developmental trajecto- ries. Three pathways have been identified: first, the overt pathway, which represents an escalation from minor aggression—such as an- noying others, bullying—to physical fighting, and eventually to violence as robbery, rape, etc.); second, the covert pathway, consisting first of minor covert acts—like shoplifting and frequent lying—then property damage and then more serious forms of theft, such as breaking and entering; and third, the authority conflict pathway, which had as its first step stubborn behavior, followed by serious disobedience and defiance, and finally by authority avoidance before the age of 12 in the form of staying out late at night or truancy. Typically, the onset of authority conflict preceded the onset of overt or PERSPECTIVES AND METHODOLOGY 17 covert behavior. In brief, the evidence points to multiple path- ways rather than a single pathway in the devel- opment of disruptive problem behaviors. The knowledge of pathways is important because it can help identify, at an early age, juveniles who appear to be at highest risk for later maladjust- ment and can distinguish them from youth with minor and/or more transient problem behavior. Pathways can also aid in the specification of the type of problem behaviors that certain risk groups of youth are likely to experience next. Once pathways have been formulated, risk and causal variables can explain why some youth do not initiate problem behavior at all, why oth- ers get into less serious problem behavior tem- porarily and do not progress, and a proportion of youth escalate in problem behavior and be- come seriously affected. The following section deals with some basic theoretical ideas that have been advanced to explain the causes of juvenile crime. Offend- ing is a type of behavior, similar in many re- spects to other types of antisocial behavior. Hence, the theories, methods and knowledge of other types of antisocial behavior can be ap- plied to the study of crime. 2.1.3 Explanation of the Causes of Antisocial Behavior Human behavior is, to a considerable degree, consistent and predictable. It has been argued that, in general, the antisocial child tends to be- come the antisocial teenager and the antisocial adult, just as the antisocial adult then tends to produce another antisocial child. This section begins by examining some of the most impor- tant risk factors that have been identified with offending and antisocial behavior.3 These in- clude family influences such as poor parental management and techniques, childhood antiso- cial behavior, offending by parents and sib- lings, low intelligence and educational attain- ment, separation from parents, prenatal and prenatal factors such as low birth weight, indi- vidual difference factors such as high impulsiv- ity, socioeconomic deprivation, peer influences such as having delinquent friends, school fac- tors, community characteristics such as social disorganization, and situational factors (Farrington, 1996). 2.1.4 Prenatal and Perinatal Factors In the Caribbean, early childbearing, or teenage pregnancy, predicts many undesirable out- comes for the children, including low school attainment, antisocial school behavior, sub- stance use, and early sexual intercourse. The children of teenage mothers are also more likely to become offenders. Studies in the US and England have found that teenage mothers were associated with low-income families, wel- fare support, absent biological fathers, used poor child-rearing methods, and that their chil- dren were characterized by low school attain- ment and delinquency. Additional research has shown that smoking during pregnancy was as- sociated with low birth weight, small height, and poor performance in school. Excessive alcohol consumption during pregnancy pre- dicted hyperactivity, low intelligence, and speech disorders in children. 2.1.5 Hyperactivity Hyperactivity is an important construct that predicts later delinquency. It is associated with restlessness, impulsivity, and a short attention span that has been termed the hyperactivity- impulsivity-attention deficit or HIA syndrome. Related concepts include a poor ability to defer gratification and a short future time perspec- tive. It was pointed out above that hyperactiv- ity and conduct problems between 8 and 10 3Risk factors are prior factors that increase the risk of occurrence of events such as the onset, frequency, persistence or duration of offending. PERSPECTIVES AND METHODOLOGY 18 years reliably predicted chronic offending, with the proviso that hyperactivity by itself is not necessarily associated with disruptive behavior. 2.1.6 Intelligence and Attainment Empirical work has shown that intelligence at age 4 significantly predicted the number of ar- rests up to age 27. Results of a nonverbal intel- ligence test in Cambridge showed that twice as many boys scoring 90 or less at ages 8 to 10 were convicted as juveniles than of the remain- der. Low nonverbal intelligence was highly correlated with low verbal intelligence and with low school attainment, and all of these meas- ures predicted juvenile convictions to much the same extent. In addition to their poor school performance, delinquents tended to be frequent truants, to leave school at the earliest possible age and to take no school examinations. Low nonverbal intelligence was especially characteristic of the juvenile recidivists and those first convicted at the earliest ages. It has also been argued that high intelligence is a pro- tective factor against offending for children from high-risk backgrounds. Delinquents often do better on nonverbal performance tests than on verbal tests. The association between school failure and offending has been demon- strated consistently in longitudinal surveys. They showed how the probability of offending increased according to the presence of truancy, school failure, and low socioeconomic status. A more plausible explanatory factor underlying the link between intelligence and offending is the ability to manipulate abstract concepts. Youngsters who are poor at this tend to do badly on intelligence tests, and they also tend to commit offenses, mainly because of their poor ability to foresee the consequences of their offending and to appreciate the feelings of vic- tims (i.e., their low empathy). One researcher argues that working-class, poorer parents tend to live for the present and to have little thought for the future, and tend to talk in terms of the concrete rather than the abstract. A lack of concern for the future is also linked to the con- cept of impulsivity. 2.1.7 Parental Supervision, Discipline, and Attitude An exhaustive review of family factors as pre- dictors of juvenile conduct problems and delin- quency found that poor parental supervision or monitoring, erratic or harsh parental discipline, parental disharmony, parental rejection of the child, and low parental involvement in the child’s activities (as well as antisocial parents and large family size) were all important pre- dictors of offending. A study conducted in Boston reported that poor parental supervision was the best predictor of both violent and prop- erty crimes. Parental aggressiveness—which included harsh discipline, shading into child abuse at the extreme—and parental conflict were significant precursors of violent crimes, while the mother’s attitude—passive or reject- ing—was a significant precursor of property crimes. Other studies also found that poor su- pervision and discipline were consistently re- lated to later offending, and that hostile and re- jecting mothers when children were age 5 pre- dicted their children’s frequent use of drugs at age 18. It was also found that poor parental supervision was the most important correlate of delin- quency for girls, and that it was the second most important for boys—after delinquent friends. The Boston study also reported that harsh or erratic parental discipline, cruel, pas- sive, or neglecting parental attitude, poor super- vision, and parental conflict predicted later ju- venile convictions. Generally, the presence of any of these adverse family background fea- tures doubled the risk of later juvenile convic- tion. Poor supervision, erratic or inconsistent discipline, and conflict between parents are all conducive to offending by children. It is im- PERSPECTIVES AND METHODOLOGY 19 portant for parents to explain to children why they are being punished, so that they can iden- tify precisely the behavior that is disapproved. One research study indicated a significant inter- generational transmission of aggressive and violent behavior from parents to children. Children who were physically abused up to age 11 were significantly likely to become violent offenders in the next 15 years. 2.1.8 Broken Homes The theory that broken homes cause delin- quency was inspired by psychoanalytic ideas in 1951. The research found that delinquents were significantly more likely than comparison children to have suffered a complete and pro- longed separation from their mothers during the first five years of life. It was argued that mother love in infancy and childhood was just as important for mental health as were vitamins and proteins for physical health. It was thought that it was essential that a child should experi- ence a warm, loving, and continuous relation- ship with a mother figure. If a child suffered a prolonged period of maternal deprivation dur- ing the first five years of life, this would have irreversible negative effects, including delin- quency. It is important to distinguish separation from a biological or operative parent after a loving re- lationship has been built up, attributable to dif- ferent causes such as death or parental dishar- mony, from the complete lack of contact with a biological parent. It was found that the preva- lence of offending was high for boys reared in broken homes without affectionate mothers and for those reared in united homes characterized by parental conflict, irrespective of whether they had affectionate mothers. The prevalence of offending was low for those reared in united homes without conflict and—importantly— equally low for boys from broken homes with affectionate mothers. These findings suggest that it is not so much the broken home that is criminogenic as the parental conflict that causes it. They also indicate that a loving mother might in some sense be able to compen- sate for the loss of a father. 2.1.9 Large Family Size Many studies in developed countries show that large families predict delinquency. There are many possible reasons a large number of sib- lings might increase the risk of delinquency. Generally, as the number of children in a fam- ily increases, not only does the amount of pa- rental attention that can be given to each child decrease, but the household will tend to be- come more overcrowded, possibly leading to increases in frustration, irritation and conflict. However, one is guided by our knowledge that in the US Virgin Islands there have been many large (extended) families with loving parents in which delinquency never reared its head. A few additional comments concerning inhibi- tions against offending are offered. First, one researcher emphasized differences in parental child rearing behavior as the major source of differences in criminal tendencies. Middle- class parents are more likely to explain to chil- dren why they were being punished and to be concerned with long-term character building and the inculcation of general moral principles. In contrast, working-class parents supervised their children less closely and were more incon- sistent in their use of discipline. Middle-class parents in general tend to use love-oriented dis- cipline, relying on withdrawal of love as the main sanction, whereas working-class parents used much more physical punishment. The re- searcher contended that working-class children committed more crimes because working-class parents used less effective methods of sociali- zation. Second, two modern criminologists suggested that people differed in their underlying criminal tendency, and that whether a person chose to PERSPECTIVES AND METHODOLOGY 20 commit a crime in any situation depended on whether the expected benefits of offending out- weighed the expected costs. The benefits of offending, including material gain, peer ap- proval, and sexual gratification, tended to be contemporaneous with the crime. In contrast, many of the costs of offending, such as the risk of being caught and punished and the possible loss of reputation or employment, were uncer- tain and long-delayed. Other costs, such as pangs of conscience (or guilt), disapproval by onlookers, and retaliation by the victim, were more immediate. The key is the extent to which people’s behavior is influenced by im- mediate as opposed to delayed consequences. 2.2 Methodology 2.2.1 Steps in Developing the Framework of the Study A series of discussion meetings were convened with representatives from the Eastern Carib- bean Center of UVI, the Juvenile Investigation Bureau (JIB) and the Department of Human Services (DHS). Through these meetings, a clearer understanding was gained of the meth- ods, procedures and guidelines each depart- ment/division utilized in processing juveniles entering the system. It was learned that youngsters 7 to 17 years were considered juveniles and processed as such. In rare cases, children as young as five years appeared on file; however, these proved to be Persons in Need of Service (PINS), vic- tims of neglect or crime, or children lost or abandoned; these were eventually excluded from the study sample. Forms used by employees in each division to record information concerning juveniles admit- ted into the system for delinquent behavior formed the basis in compiling the data- gathering instrument. These forms included the Contact Card used by JIB as well as several forms from the DHS that provided information on general, personal, psychological, educa- tional, and legal aspects of the juvenile history. The normal process allowed an officer with the JIB to arrest a juvenile age seven to 17 who was involved in offensive behavior and to place on file a Contact Card containing a detailed description of the offender and the crime perpe- trated. The juvenile is then referred to DHS with a copy of the information gathered by JIB. DHS conducts a thorough assessment of the juvenile and makes a decision about the needs of each juvenile, as well as the affected com- munity. Two survey instruments were compiled by the Eastern Caribbean Center in collaboration with JIB and DHS. Since all juveniles must first pass through JIB, that became the starting point of data-gathering. To ensure confidentiality and avoid disclosure to unauthorized persons, names of juvenile delinquents were adhered to the data-gathering instrument by a removable label to facilitate identification by JIB and DHS data gatherers. Each case was then identified by a serial number consisting of four ordinate dig- its and a first offense date for identification. When JIB employees completed the recording of data for all cases they were passed on to DHS where the process was continued with in- formation from files of juvenile delinquents. Name-identifying labels were removed from the data-gathering instrument before delivery to the Eastern Caribbean Center for processing. It was discovered that the total number of cases existing on file at JIB did not match total cases at DHS. This was due at least in part to the ex- istence of some cases where juveniles with first offenses that were considered minor in nature were placed on file at JIB, counseled and re- leased into the care of family members but not passed on to DHS for further processing. PERSPECTIVES AND METHODOLOGY 21 Data gathering commenced at JIB in the St Thomas-St John district and data were re- corded for all cases on file for the years 1987 to 1997 inclusive. When all cases were recorded from both JIB and DHS, the data-gathering in- struments were edited manually, and numeric codes were inserted in preparation for data en- try. An Integrated Micro-computer Processing Systems (IMPS) software program was used to produce data dictionaries and data application screens. 2.2.2 The Data Collection Instrument Two data-collection instruments were devel- oped: the Contact Card and Arrest Record and the Department of Human Services Form to meet the special requirements of data available at JIB and DHS, respectively. Each instrument sought to capture information on each juvenile related to the first offense. The Contact Card and Arrest Record consisted of 10 items that required basic identification, background, family information as well as ar- rest data and offenses. Arrest data and offenses included total number of offenses recorded throughout the juvenile’s contact with JIB, as well as arrest date and offense, type of crime and weapon used for up to six incidents: namely, the first three and the last three of- fenses. Forms for both districts carried the same items with the exception of an additional item on past involvement of family members in criminal activities for the St Croix Contact Card and Arrest Record The DHS data collection instrument consisted of 29 items that sought information on the juve- nile’s ethnicity, educational progress, political status, social background, economic status, and psychological adjustment. Information was also gathered on the status and outcome of the six offenses detailed in the JIB Contact Card and Arrest Record. In processing juvenile delinquents in the dis- trict of St Croix, JIB and DHS did not carry out the same type of collaboration as seen in the district of St Thomas/St John. The St Croix dis- trict did not track juveniles across the two divi- sions, as did St Thomas/St John. The district of St Croix also added an item to test the extent to which a juvenile offender was acquainted with their biological father. 2.2.3 Selection of the Population In the St Thomas/St John district, the popula- tion consisted of all juveniles that were on JIB files dated 1987 to 1997. Some of the earlier offenses of these cases extended to years prior to 1987 These data records were then matched at DHS, and filled out for each case by JIB personnel using available data. Both data forms—without personal identifying information—were then handed over to the Eastern Caribbean Center. Overall, there was no strict consistency in the process of data- gathering between St Thomas/St John and St Croix districts. In St Croix, there was no match between JIB and DHS in the data-gathering process, since there were no systems in place to track each case from one division to the other. 2.2.4 Data Collection Procedures In both divisions, employees who pledged con- fidentiality and were familiar with the forms used to store data on juveniles’ case files, were employed temporarily as data gatherers as well as supervisors. Preliminary briefings and train- ing sessions were conducted with employees to ascertain the need for accuracy and integrity in the data gathering process. Guidelines for fill- ing out the data collection instruments were prepared to ensure uniformity and clarity of the PERSPECTIVES AND METHODOLOGY 22 process. These included a description of simple steps in the data-gathering process. In some cases, suggestions were given as to where cer- tain pieces of data could be located in the divi- sion’s forms. The designated supervisor in each division was responsible for overall quality of the job, to be confirmed by thorough checks, before delivery of the finished product to the Eastern Caribbean Center. In the St Thomas/St John district, data- gathering instruments completed by JIB were turned over to the supervisor at DHS for con- tinuation of the data-gathering process. This ensured that the names of juvenile offenders remained confidential. On St Croix after com- pletion, the data-gathering instruments were forwarded to the Eastern Caribbean Center in- dependently. 2.2.5 Data Availability and Deficiencies The sources from which data were gathered were the manual case files kept by juvenile- arresting officers at JIB and social workers at DHS. The forms used by these persons to re- cord information were designed to serve the needs of these agencies, while processing juve- niles through various stages of the system. Data used in this study were therefore limited to what was available and the data gathering in- struments could only take advantage of data already present on the forms. In addition, whereas the basic information on identification of juveniles was routinely filled in, there seemed to have been diligence in cap- turing data on areas such as social background, educational levels, ethnicity and psychological adjustment. A significant number of cases had items with missing data. 2.2.6 Data Analysis Procedures Data entered into the IMPS Integrated Micro- computer Processing System storage files were examined for mistakes or incongruence by tabulating frequency tables and making adjust- ments where necessary. New frequencies and cross tabulations were produced and key items or indicators were looked at to identify pres- ence of relationships. For example, household type was compared with number and type of juvenile delinquent; place of birth of juvenile delinquent and number of offenders. 2.2.7 Organization of the Report In general, background information in this re- port is presented for the Territory with specific references to districts. There is comparative section on the Territory which also indicates differences between districts. However, the sec- tion on Survey Findings is presented by district, s i n c e t h e d a t a c o l l e c t e d f o r St Thomas/St John were not the same variables as those for St Croix. Whereas St Thomas/St John could track cases at JIB with their counterparts at DHS, this was not possible on St Croix. In this report each data set for St Thomas/St John and St Croix is each pre- sented and analyzed separately by district. 2.2.8 Limitations of the Study Ideally, a study that seeks to develop profiles or characteristics that could be used as red light or risk identifiers in the prevention of juvenile de- linquency would be more comprehensive if the attributes of delinquents, as well as those of non-delinquents, were examined and a com- parison made of the results. Since this study focuses only on the delinquents, it is difficult to make categorical statements as to possible cause and effect without an adequate available control group. PERSPECTIVES AND METHODOLOGY 23 The population of the study is restricted to the files that were available in JIB and DHS divi- sions at the time of data-gathering. In the St Thomas/St John district a number of files were identified for which data were recorded at JIB, but did not have corresponding matches at DHS. Cross tabulation and analysis of data for certain items in both sets of data were limited to items with cases presented in both sets of data. At times there were inadequate or insufficient data available on division forms used as source information to fill out data collection instru- ments for the study. For some items, a signifi- cant number of cases had missing data, having been left vacant by persons originally or per- sons handling that particular case. In the district of St Croix, on the other hand, since JIB and DHS data were each gathered without a cross reference for cases, data could only be processed in separate JIB and DHS batches. No direct connection could be made between these groups of data. It was therefore not possible to process the St Croix data to track relationships between variables as in- depth as that carried out for the St Thomas/St John district. On the higher level, since the method of data- gathering differed in each district, it was not possible to combine the data sets for the two districts and look at relationships as they oc- curred for the territory as a whole. These differences in data gathering occurred due to different forms and methods of data col- lection and storage within the two districts, as well as a difference in the relationship and sup- port systems practiced within districts by JIB and DHS. PERSPECTIVES AND METHODOLOGY 24 25 St Thomas Data: Description and Analysis 3.1 Juvenile Investigation Bureau Data Variables Data collection commenced at Juvenile Investi- gation Bureau (JIB) in the St Thomas/St John district.3 Data were recorded for all cases on file for the years 1987 to 1997. The data collec- tion instrument developed was the Contact Card and Arrest Report (CCAR). Data were recorded for each arrest and the following char- acteristics were covered in the CCAR: Sex Age Date of birth Place of birth Race/ethnic origin Person with whom living Birthplace of mother and father Enrollment status Education (highest level) Employment status of guardian Date of crime/incident Type of crime/incident Type of weapon used 3.1.1 St Thomas Data Description: Juvenile Investigation Bureau Year of First Offense—There were 2,348 cases of first offenses in the JIB data set (Table 3.1.1). On the average there were about 198.7 first offenses annually between 1987 and 1997. There was a gradual increase annually from 1987 to 1991 with the exception of 1990. The period 1992 to 1997 was mixed: there were de- clines in 1992 and 1993, an upsurge in 1994, declines again in 1995 and 1996, and an in- crease in 1997. The smallest number of of- fenses was 104 in 1996, and the highest was 298 in 1991. This suggests a wide range in the annual number of offenses. It is not certain whether this disparity was due to a difference in record keeping or whether it was real. Type of First Offense—In Table 3.1.2, Part I Felony accounted for 21.8 percent of first of- fenses. Part I Felony includes robbery, feloni- ous assault, burglary, grand larceny, homicide, rape and auto theft. Part II Felony accounted for 11.5 percent of all offenses. These include possession of con- III. PRESENTATION ANALYSIS Table 3.1.1 Year of First Offense Year of Offense Number Percent Total 2348 100.0 Before 1980 4 0.2 1980 1 0.0 1981 8 0.3 1982 13 0.6 1983 9 0.4 1984 15 0.6 1985 9 0.4 1986 22 0.9 1987 180 7.7 1988 219 9.3 1989 259 11.0 1990 220 9.4 1991 298 12.7 1992 243 10.3 1993 178 7.6 1994 204 8.7 1995 135 5.7 1996 104 4.4 1997 146 6.2 After 1997 81 3.4 3All future references to the St Thomas/St John district are shortened to St Thomas. 26 trolled substances, unlicensed firearms, stolen property, unauthorized use of vehicle, fraud, and forgery. Misdemeanor accounted for 13.2 percent, which includes aggravated assault and battery, and petit larceny. Incidences accounted for 23.5 percent, and this category includes run- away minors and missing minors. “Others” ac- counted for 30.0 percent. Weapon Used in First Offense—The weapon categories were gun, knife, hand, bottle, penis, none and other. In 78.4 percent of the cases no weapon was used (Table 3.1.3). Hand was used in 6.4 percent of the cases, followed by knife (3.9 percent), gun (2.7 percent), bottle (1 per- cent), and penis (0.5 percent). Total Number of Offenses—The majority—or about close to two-thirds (63.7 percent)—of the juveniles had only one offense. About one in six, or 17.1 percent, had two offenses, and 7.3 percent had 3 offenses. The percentages for more than three and more than six were 8.2 percent and 3.1 percent, respectively (Table 3.1.4). Table 3.1.2 Type of First Offense Year of Offense Number Percent Total 2335 100.0 Part I Felony 509 21.8 Part II Felony 268 11.5 Misdemeanor 309 13.2 Incidence 548 23.5 Other 701 30.0 use of vehicle, Fraud, Forgery. Misdemeanor: Aggravated assault and battery, Petit larceny. Incidences: Run-away minor, missing minor. Part I Felony: Robbery, Felonious assault, Burglary, Grand larceny, Homicide, Rape, Auto theft. Part II Felony: Possession of controlled substance/ unlicensed firearm/ stolen Table 3.1.3 Weapon Used in First Offense Weapon Number Percent Total 2332 100.0 Gun 64 2.7 Knife 90 3.9 Hand 149 6.4 Bottle 23 1.0 Penis 11 0.5 None 1828 78.4 Other 167 7.2 Table 3.1.4 Total Number of Offenses Offense Number Percent Total 2340 100.0 1 1491 63.7 2 401 17.1 3 170 7.3 4 92 3.9 5 58 2.5 6 43 1.8 7 11 0.5 8 13 0.6 9 9 0.4 10 11 0.5 11 16 0.7 12 6 0.3 13 3 0.1 14 2 0.1 15 6 0.3 17 3 0.1 19 1 0.0 22 1 0.0 24 1 0.0 27 1 0.0 34 1 0.0 PRESENTATION ANALYSIS 27 3.1.2 St Thomas: Juvenile Investigation Bureau Data Relationships Total Number of Offenses by Place of Birth—Table 3.1.5 represents the total number of offenses cross-classified by place of birth. Total numbers of offenses were grouped in two categories: one offense, and two or more of- fenses. In the case of juveniles born in the United States or other countries, three-fourths more of them had only one offense; less than one quarter of them had two or more offenses. For juveniles having birthplaces in the Virgin Islands (VI), Puerto Rico (PR) and the Eastern Caribbean (EC), the percentages for two or more offenses were higher—37.6 percent, 41.3 percent and 36.0 percent, respectively. Type of First Offense by Juvenile’s Place of Birth—The data in Table 3.1.6 cross-classify 'type of first offense by place of birth'. In the Virgin Islands, Part I Felony accounted for 21.7 percent of the cases (Table 3.1.2). In those cases where the birthplace was Puerto Rico, or the Eastern Caribbean, it accounted for 24.4 and 24.2 percent, respectively. Both of these were above the overall percentage for the Vir- gin Islands. Juveniles with a US birthplace had 15.2 percent of their offenses in the Part II Felony category, which was much higher than that—11.0 per- cent—for the Territory overall. Puerto Ricans had the lowest percentage of misdemeanor cases—4.4 percent—while native-born Virgin Islanders had the highest—14.3 percent. Virgin Islanders had the lowest percentage—21.1 per- cent—of Incidences. Type of Weapon Used by Place of Birth— Table 3.1.7 reveals that the overwhelming ma- jority of the offenses—77.3 percent—were not Table 3.1.5 Total Number of Offenses by Place of Birth Offenses Number Percent Number Percent Number Percent Number Percent Number Percent Total 1667 100.0 46 100.0 165 100.0 422 100.0 34 100.0 1 Offense 1040 62.4 27 58.7 125 75.8 270 64.0 27 79.4 2 or more 627 37.6 19 41.3 40 24.2 152 36.0 7 20.6 Other Place of Birth Virgin Islands Puerto Rico United States Eastern Caribbean Table 3.1.6 Type of First Offense by Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 1668 100.0 45 100.0 165 100.0 422 100.0 34 100.0 Part I Felony 362 21.7 11 24.4 30 18.2 102 24.2 3 8.8 Part II Felony 184 11.0 4 8.9 25 15.2 51 12.1 4 11.8 Misdemeanor 238 14.3 2 4.4 16 9.7 50 11.8 3 8.8 Incidences 352 21.1 15 33.3 53 32.1 114 27.0 13 38.2 Other 532 31.9 13 28.9 41 24.8 105 24.9 11 32.4 Place of Birth Virgin Islands Puerto Rico United States Eastern Caribbean Other Type of Offense PRESENTATION ANALYSIS 28 committed with weapons. In cases where a weapon was used by juveniles born in the US Virgin Islands, the most commonly used ones were by Hand (7.0 percent), Knife (4.1 per- cent), and Gun (2.9 percent). Duration Between First and Second Of- fense—The mode—or the category with the largest number of occurrences—for the dura- tion between the first and second offense was “One to less than two years.” One out of every five repeat offense took place between 12 and 24 months after the first offense (Table 3.1.8). Age at First Offense by Sex—Table 3.1.9 in- dicates juvenile offense levels were relatively very low between the ages of seven and nine— two percent or below. The number increased to 3.2 percent for 10-year-olds. Offense rates con- tinued to increase with age, climbing to 12 per- cent for 13-year-olds, and remaining high until age 17. The peak of activity occurred at 15, the age at which 19.5 percent of juveniles commit- ted their first offense. The data suggest a posi- tive association between age and the number of juveniles committing their first offense. When the data are broken down by gender, it is observed that a pattern of increased activity oc- curred for both sexes between the ages 13 to 17. The largest proportion of males committed their first offense between the ages 13 and 17, while this occurred for females between ages 13 to 16. By age 17 females had dropped to 9.9 percent of cases having first offenses, while the males were still relatively active, with 13.1 per- cent committing their first-time offense at this age. These data would seem to suggest that fe- males abandoned their delinquent activity at an Table 3.1.7 Type of Weapon Used by Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 1676 100.0 46 100.0 165 100.0 422 100.0 34 100.0 Gun 49 2.9 0 0.0 2 1.2 12 2.8 1 2.9 Knife 69 4.1 1 2.2 3 1.8 16 3.8 1 2.9 Hand 117 7.0 4 8.7 9 5.5 18 4.3 1 2.9 Bottle 16 1.0 0 0.0 1 0.6 6 1.4 0 0.0 Penis 10 0.6 0 0.0 0 0.0 1 0.2 0 0.0 None 1295 77.3 38 82.6 138 83.6 328 77.7 27 79.4 Other 120 7.2 3 6.5 12 7.3 41 9.7 4 11.8 Weapon Place of Birth Virgin Islands Puerto Rico United States Eastern Caribbean Other Table 3.1.8 Duration Between First and Second Offense Duration Number Percent Total 824 100.0 < 1 Month 104 12.6 1 to < 2 Months 75 9.1 2 to < 3 Months 43 5.2 3 to < 6 Months 103 12.5 6 to < 12 Months 155 18.8 1 to < 2 Yrs 175 21.2 2 to < 3 Yrs 75 9.1 3 to < 4 Yrs 50 6.1 4 to < 5 Yrs 18 2.2 5 Yrs and Over 26 3.2 PRESENTATION ANALYSIS 29 earlier age than did males. This observation coincides with other child development con- cepts that assign higher levels of physical and social development to females at an earlier age than to males. Place of Birth of Juveniles by Sex—Table 3.1.10 indicates that the majority of juvenile offenders were born in the US Virgin Islands— 73.3 percent. There were 12.4 percent of juvenile offenders born on an Eastern Caribbean Island and 7.2 percent of juvenile delinquents who were born in the US. Female offenders were slightly less likely to by born in the US Virgin Islands or and Eastern Caribbean Island and slightly more likely to be born in the US than male offenders. Age at First Offense by Place of Birth—In Table 3.1.11 juveniles from all places of birth under the age of 10 appeared to have a low par- ticipation in delinquent behavior compared to the older age groups. Significant increases in the rate occurred between ages 10 to 14 for all Table 3.1.9 Age at First Offense by Sex Number Percent Number Percent Number Percent Total 2282 100.0 1633 100.0 649 100.0 Age 7 22 1.0 18 1.1 4 0.6 Age 8 37 1.6 32 2.0 5 0.8 Age 9 46 2.0 40 2.4 6 0.9 Age 10 73 3.2 66 4.0 7 1.1 Age 11 108 4.7 91 5.6 17 2.6 Age 12 177 7.8 127 7.8 50 7.7 Age 13 274 12.0 193 11.8 81 12.5 Age 14 430 18.8 284 17.4 146 22.5 Age 15 446 19.5 308 18.9 138 21.3 Age 16 391 17.1 260 15.9 131 20.2 Age 17 278 12.2 214 13.1 64 9.9 Male Female Total Age Table 3.1.10 Place of Birth of Juvenile by Sex Number Percent Number Percent Number Percent Total 2288 100.0 1627 100.0 640 100.0 United States 165 7.2 106 6.5 57 8.9 US Virgin Islands 1676 73.3 1202 73.9 460 71.9 Eastern Caribbean 283 12.4 206 12.7 73 11.4 Other Caribbean 90 3.9 62 3.8 28 4.4 Puerto Rico 46 2.0 31 1.9 14 2.2 Other 28 1.2 20 1.2 8 1.3 Female Place of Birth Total Male PRESENTATION ANALYSIS 30 groups. The peak age group for which first time delinquent activity occurred for all juveniles, regardless of place of birth, was between ages 10 and 14, with a falling off at the “Over 16” age category. Age at First Offense by Mother’s Place of Birth—In Table 3.1.12 the same pattern ap- pears as for Table 3.1.11 above. First-time de- linquent activity remained low for juveniles under age 10 for all place of birth groups, but increased significantly among the 10- to 14- year-old age group. There was a higher rate of first-time offenses in the 10 to 14 age group— 50 percent—in the case of mothers born in Puerto Rico, compared with the other place of birth groups. This would seem to suggest that a greater proportion of juveniles who have a mother born in Puerto Rico become involved in delinquent activity at an early age. Age at First Offense by Father’s Place of Birth—The frequencies among the age catego- ries in Table 3.1.13 for place-of-birth of moth- ers are not widely different from those in 3.1.12 for place-of-birth of fathers. However, while the percentages in the age 10 to 14 group are higher for mothers that were born in the Virgin Islands and Puerto Rico, the frequency in the same age category is higher for fathers born in the US than for fathers born elsewhere. Juvenile’s Race by Sex—Table 3.1.14 shows that of the total number of juveniles who com- mitted a first offense, 96.0 percent were black, while 1.7 percent were white. Very small per- centages were classified as Asian and Other— 0.2 and 2.0 percent, respectively. These per- centages were generally consistent for males and females. Table 3.1.11 Age at First Offense by Juvenile's Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 1642 100.0 45 100.0 160 100.0 422 100.0 34 100.0 Under 10 yrs 93 5.7 1 2.2 4 2.5 16 3.8 1 2.9 10 to 14 yrs 769 46.8 19 42.2 80 50.0 183 43.4 17 50.0 15 to 16 yrs 587 35.7 20 44.4 58 36.3 167 39.6 10 29.4 Over 16 yrs 193 11.8 5 11.1 18 11.3 56 13.3 6 17.6 Age Place of Birth Virgin Islands Puerto Rico United States Eastern Caribbean Other Table 3.1.12 Age at First Offense by Mother's Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 688 100.0 52 100.0 105 100.0 1167 100.0 32 100.0 Under 10 yrs 39 5.7 2 3.8 12 11.4 87 7.5 1 3.1 10 to 14 yrs 332 48.3 26 50.0 44 41.9 490 42.0 10 31.3 15 to 16 yrs 251 36.5 17 32.7 35 33.3 433 37.1 15 46.9 Over 16 yrs 66 9.6 7 13.5 14 13.3 157 13.5 6 18.8 Puerto Rico United States Eastern Caribbean Other Age Mother's Place of Birth Virgin Islands PRESENTATION ANALYSIS 31 In a 1995 survey of the general population of St Thomas and St John, among 5- to 19-year- olds, blacks accounted for 89.4 percent of the total, while whites accounted for 4.3 percent, and those of other race represented 6.3 percent. Although the age categories do not match ex- actly—the age group of the data set is 7 to 17— this gives a broad general comparison of the situation. Juvenile’s Guardian by Sex of Juvenile—The data in Table 3.1.15 show the majority of juve- nile delinquents—54.8 percent—lived with their mother alone. The second highest percent- age—21.1 percent—lived with both mother and father. The remainder of the guardian catego- ries ranged between 0 and 10.4 percent. Pro- portionally, eight times as many lived with “mother and stepfather” as opposed to “father and stepmother”—3.3 percent to 0.4 percent, respectively. When comparisons are made by gender, both male and female categories of juvenile offend- ers followed similar patterns within similar ranges. All of the remaining guardian catego- ries experienced relatively low proportions of the total number of offenders. One can assume with some confidence that this is quite likely a reflection of the representation of these groups in the general Virgin Islands’ population. Namely, the assumption is that the majority of children in this age group live with mothers alone, and the second largest group live with both natural parents. School Enrollment at Time of Arrest—Table 3.1.16 indicates that only a small percentage of Table 3.1.13 Age at First Offense by Father's Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 478 100.0 45 100.0 110 100.0 1141 100.0 30 100.0 Under 10 yrs 51 10.7 2 4.4 11 10.0 75 6.6 1 3.3 10 to 14 yrs 207 43.3 19 42.2 55 50.0 487 42.7 11 36.7 15 to 16 yrs 168 35.1 17 37.8 31 28.2 423 37.1 12 40.0 Over 16 yrs 52 10.9 7 15.6 13 11.8 156 13.7 6 20.0 United States Eastern Caribbean Age Father's Place of Birth Virgin Islands Other Puerto Rico Table 3.1.14 Juvenile's Race by Sex Total Male Female Number Percent Number Percent Number Percent Total 2346 100.0 1666 100.0 657 100.0 Black 2252 96.0 1602 96.2 629 95.7 White 39 1.7 26 1.6 13 2.0 Asian 4 0.2 2 0.1 2 0.3 Other 47 2.0 33 2.0 13 2.0 Race PRESENTATION ANALYSIS 32 offenders—12 percent—were not enrolled in school at the time of arrest. This suggests that being enrolled in school is not a deterrent to youngsters committing offenses. A more com- prehensive picture might be gained if data were available for analysis on levels of school and class attendance, and even achievement meas- ures. School Type at Time of Arrest—Table 3.1.17 shows data on the school level at which juve- nile delinquents were enrolled at the time of their first offense. These data seem to correlate with previous sets of data that reflected age and first offense. The data show that a relatively low percentage of juveniles committed first of- fenses in elementary school—9.3 percent— increasing to 45.1 percent in Jr. high school, and 45.7 in high school. Mother’s Place of Birth by Juvenile’s Sex— Table 3.1.18 indicates that the majority of juve- nile offenders—39.8 percent—were children of mothers born in the Eastern Caribbean. The second highest proportion—35 percent—were children whose mothers were natives of the US Virgin Islands. The percentage of juvenile de- linquents with mothers born in the other Carib- bean islands or island group was 14.9 percent. Mother’s Employment by Sex of Juvenile— Table 3.1.19 shows that 72.9 percent of juve- nile offenders had mothers who were em- Table 3.1.15 Juvenile's Guardian by Juvenile's Sex Number Percent Number Percent Number Percent Total 2,346 100.0 1,666 100.0 657 100.0 Mother & Father 496 21.1 366 22.0 123 18.7 Mother 1,285 54.8 914 54.9 359 54.6 Father 137 5.8 105 6.3 31 4.7 Mother and Step Father 77 3.3 63 3.8 14 2.1 Father and Step Mother 9 0.4 5 0.3 4 0.6 Step Mother 3 0.1 2 0.1 1 0.2 Step Father 5 0.2 3 0.2 2 0.3 Foster Parents 23 1.0 8 0.5 15 2.3 Relatives 245 10.4 157 9.4 85 12.9 Other 46 2.0 32 1.9 14 2.1 Female Guardian Total Male Table 3.1.16 School Enrollment Status Number Percent Total 972 100.0 Enrolled 855 88.0 Not Enrolled 117 12.0 Table 3.1.17 School Type School Number Percent Total 799 100.0 Elementary 74 9.3 Jr. High 360 45.1 High 365 45.7 PRESENTATION ANALYSIS 33 ployed. Male offenders were slightly more likely to have employment mothers—74.0 percent—than were females—70.2 percent. The finding that most of the mothers of juve- nile offenders were employed may suggest that employed mothers were away from the home for extended hours, especially at times when children returned home from school. These data do not indicate if any of the employed mothers held more than one job or worked overtime. These factors would potentially add to a mother-child separation, and leave unsuper- vised children exposed and vulnerable to in- volvement in delinquent behavior. 3.1.3 Predicting Recidivism with the St Thomas Juvenile Investigation Bureau Data 3.1.3.1 Approach One of the primary purposes of the study iden- tified above was to adopt a proactive approach to the study of juvenile delinquent behavior. A persistent issue that is of significance to the De- partment of Human Services (DHS) relates to the task of trying to determine if a first-time offender will commit additional violations that will repeatedly result in his or her arrest. The data from the Contact and Arrest Reports show that during the study period, about 36 percent of all arrested juveniles had been arrested more than once. Of the total, about 12 percent of the juveniles had been arrested 4 or more times. It is the relatively high percent of repeat offenses that argues for closer study of this phenome- non. It is evident why DHS has a primary interest in knowing what are the factors that might explain why some first-time offenders—male or fe- male—never commit another offense to pre- cipitate an arrest, while other youth persist in patterns of delinquent behavior that cause them Table 3.1.18 Mother's Place of Birth by Sex of Juvenile Number Percent Number Percent Number Percent Total 2,014 100.0 1,438 100.0 559 100.0 United states 108 5.4 76 5.3 32 5.7 US Virgin Islands 705 35.0 501 34.8 198 35.4 Eastern Caribbean 801 39.8 577 40.1 218 39.0 Other Anglo Caribbean 300 14.9 214 14.9 84 15.0 Puerto Rico 53 2.6 36 2.5 14 2.5 Other 47 2.3 34 2.4 13 2.3 Total Male Female Place of Birth Table 3.1.19 Mother's Employment by Sex of Juvenile Number Percent Number Percent Number Percent Total 1,979 100.0 1,435 100.0 527 100.0 Employed 1,443 72.9 1,062 74.0 370 70.2 Unemployed 536 27.1 373 26.0 157 29.8 Female Status Total Male PRESENTATION ANALYSIS 34 to be arrested several times. A keen under- standing of this phenomenon becomes even more critical because DHS social caseworkers realize that juvenile repeat offenders often ma- ture into adult criminals. The question to be resolved might be stated another way: what is the combination of char- acteristics that might help to predict if a first- time offender might become a recidivist or re- peat offender, that is, one showing a tendency to relapse into a previous mode of delinquent behavior? Statistical methods provide one technique known as logit modeling that can help to pro- vide answers to the question just posed. In the sections below, this statistical method will be applied to some of the demographic character- istics of juveniles that were recorded in the sur- vey data, age, place of birth, race, place of birth of parents, school enrollment status and grade at time of arrest. These data were collected from both the Juvenile Investigation Bureau (JIB) and DHS offices in St Thomas/St John and in St Croix. [Note: Readers who are not interested in the statistical methodology or procedure of logisti- cal analysis may skip sections 3.1.3.2 to 3.1.3.4]. 3.1.3.2 Methodology Although logit modeling or logistic regression is a powerful statistical analytic technique, it may be presented in relatively simple terms for the non-statistical reader. An elementary de- scription of the logit model is that it is one that objectively examines in a numerical way the relationship between a dependent variable and one or a set of independent characteristics. In this case, the method determines the chances of a first-time offender becoming a recidivist, and it also helps to measure the strength of the im- pact of the demographic characteristics on the chance of being a repeat offender. Further, it also enables the analyst to determine the order of the importance of the characteristics that are significant in their influence upon reverting to offensive behavior. A concept that is commonly used in this kind of analysis is that of odds. This term is commonly used by gamblers and is easily understood by most persons. The odds indicate the relative number of chances of falling into one of two categories on some characteristic of interest. For example, if in a population of couples liv- ing together—married and unmarried—the chance of divorce or break-up is 0.4 and the chance of remaining married or living together is 0.6, then the odds of getting divorced or splitting up would be 0.4/0.6 = 0.67. This indi- cates that couples are two-thirds as likely to break-up as they are to stay together. 3.1.3.3 Definition of Variables Dependent Variable: The primary outcome variable or characteristic of interest for each individual in the data set is a dichotomous vari- able (consisting of two groups). A juvenile was classified as a one-time offender if he or she were recorded as having been arrested one and only one time. The juvenile was classified as a recidivist if he or she were arrested two or more times within the period of study. The one-time offender is the referent group. Independent or Explanatory Variable: These variables or characteristics are included in the study because they have been determined by prior preliminary analysis to exert relatively strong influence on the outcome variable. These include age, sex and ethnicity of the ju- venile, and the primary person with whom the youth lives. Other variables that were consid- ered, but not included in the final predictive model were place of birth of the mother and father, whether the mother and father were presently occupied, school enrollment status of PRESENTATION ANALYSIS 35 the youth at the time of first arrest, and grade level at the time of first arrest. Additionally, the type of offense and the type of weapon used were also included in a preliminary test of vari- ables that might exert statistically significant impact on the outcome variable. Even though logic suggested that certain demographic vari- ables should be included in such a model, only those that met statistical significance were re- tained in the final estimating equation. i) Age: The variable age is simply the age of the juvenile at the time of first arrest. Using the actual age instead of a set of age categories affords an analysis of the changing chances or odds of recidivism as he juvenile moves across the age spectrum that is measured—7 to 17 years. ii) Age Squared: This is a theoretical vari- able that captures the essence of the re- lationship between age and recidivism in that as juveniles mature, the chances of additional offenses level off. That is, the chances of recidivism decrease with age at a decreasing rate. (Age squared is therefore age multiplied by age.) iii) Gender: This variable is dichotomous and refers to the sex of the juvenile. Female is the referent category. iv) Ethnicity: Juveniles were classified into four categories based on the juvenile’s place of birth at the time of first arrest. The groups were: born in the Virgin Islands, born in Puerto Rico or other Hispanic country, born in the Eastern Caribbean, and born in the Mainland United States. This latter is the referent group or the group with which the oth- ers were compared. v) Guardian: The primary person with whom the juvenile was living at time of first arrest was grouped into one of two classifications. One was “mother and step father” or “Foster parents”. All other juveniles fell into the category “Other parent”. This latter was the ref- erent group. 3.1.3.4 The Model Let it be assumed that Ri is recidivism or repeat offending, Ai is the age of each juvenile, ASi is age squared, Si is sex or gender, Ei is ethnicity and Gi is primary guardian. It may further be assumed that pi is the chance—or conditional probability—that recidivism is high, and 1 – pi that the chance that it is low, given Ai, ASi, Si, Ei, and Gi. The logistic regression model for the log odds of a recidivist is: Log(p/(1-p)) = log Ri = a + ß1(Ai) + ß2(ASi) + ß3(Si) + ß4(Ei) + ß5(Gi) where log (p/(1-p)) is simply the log odds of becoming a recidivist, given the explanatory variables, and the betas represent the change in the log odds due to unit increments in the val- ues of the predictors (Menard, 1995). In sim- pler terms, the model can be written as: Prob(event) = 1 / (1 + e-Z) where Z is the familiar linear combination Z = B0 + B1X1 + B2X2 + … + BpXp. 3.1.3.5 Estimates of the St Thomas Juvenile Investigation Bureau Data Numerical results from estimation of the logis- tic regression model are shown in Table 3.1.3.1. While the total number of cases ini- tially included in the analysis was 2,346 of- fenders, only 2,275 were used to estimate the results shown in the table. Seventy-one of the cases had at least one item of information miss- ing, and had to be excluded. At the bottom of the main part of the table is a row that indicates the Likelihood ratio chi- PRESENTATION ANALYSIS 36 square = 196.1936. This is a test of the hy- pothesis that all of the explanatory variables have coefficients of 0. Since the associated p- value is less than 0.01, it is concluded that the hypothesis is substantiated and that at least one of the coefficients for the independent variables in column (1) is not 0. It was defined above that the odds of an event occurring are the ratio of the chance that it will occur to the chance that it will not occur. When a parameter estimate—as shown in Table 3.1.3.1—is positive, the odds ratio will be greater than one, which means that the odds of the occurrence of the event are increased. If the parameter estimate is negative, the odds ratio will be less than 1, and the odds of the event are decreased. The variable Age has a positive sign on the pa- rameter estimate—shown in the second column of Table 3.1.3.1. This positive coefficient for Age indicates that the likelihood of recidivism increases with Age. The negative sign associ- ated with the variable Age Squared means that the chances of repeated offenses increase with age at a decreasing rate. Thus, having been ar- rested once, the odds of relapsing into offensive behavior with a one-year increase in age are about three times as likely as not being arrested again. This information is derived from the adjusted odds ratio of 3.23 for Age shown in the last column. This means that for a one-year increase in age, the odds of becoming a recidi- vist are more than three times that of not be- coming a repeat of offender. The table shows that the negative estimate for Age Squared— 0.0521—produces an odds ratio that is slightly PRESENTATION ANALYSIS Table 3.1.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Thomas, Juvenile Investigation Bureau Data Variable Parameter Estimate (1) Standard Error (2) Pr> Chi Square (3) Adjusted Odds Ratio (4) Intercept -6.5287 1.2110 <0.0001 … Age 1.1726 0.1871 <0.0001 3.230 Age Squared -0.0521 0.0072 <0.0001 0.949 Sex (1=Male) 0.3220 0.0525 <0.0001 1.904 Ethnicity USVI Borna 0.2044 0.0941 0.0298 1.879 Puerto Rico or Dominican Republic Borna 0.0083 0.2045 0.9675 1.545 Eastern Caribbean Born a 0.2138 0.1157 0.0647 1.897 Guardian (1=Mother and Stepfather or Foster Parents)b 0.3201 0.1068 0.0027 1.897 Likelihood Ratio Chi Square = 196.1936 df = 7 Pr > Chi Square < 0.0001 a Compared to US Born. b Compared to living with mother and father, mother only, father only, father and step mother, step mother only, step father only, foster parents, or relatives. Notes: N=2275. Odds ratios greater than 1.0 indicate greater odds of recidivism. 37 less than 1.0. Because the relationship between age and recidivism is nonlinear, this variable helps to dampen the effect of age as the juve- niles become older. The estimate for Sex is positive with an odds ratio of 1.904. This indicates that the chance of being a recidivist, with other variables held constant, is about twice as high among males as it is for the base case, or for females. Four ethnic groups were included in the model, and are explained in the lower half of Table 3.1.3.1. Three ethnic types are compared to the base case of US born. The odds ratio for born in the US Virgin Islands and born in the East- ern Caribbean are 1.879 and 1.897, respec- tively. The odds ratio for juveniles born in Puerto Rico or the Dominican Republic is smaller than those of the other two, with a value of 1.545. It is noted, however, that the column headed ‘Pr > Chi Square’ shows that this variable greatly exceeds the 0.05 level of significance, and is therefore not considered statistically substantive. This is probably due to the relatively small number of observations that fell into this category when compared to the others. A first-time juvenile offender born in either the US Virgin Islands or in the Eastern Caribbean, with an odds ratio of 1.879 and 1.897, respec- tively, is almost twice as likely to become a recidivist as a stateside-born first-time of- fender. A Hispanic juvenile first-time offender that is one born in either Puerto Rico or the Do- minican Republic, is only slightly less likely than either Virgin Islands-born juveniles or Eastern Caribbean youth to become a repeat offender. The guardian or the primary person with whom the offender was living at the time of first arrest was classified into two groups. The first group included a juvenile who was living with his or her mother and a stepfather, or with foster par- ents. The second group consisted of all other guardians, primarily mother and father together or mother alone, father alone, father and step- mother, stepmother, stepfather, relatives or other kinds of living arrangements. The odds ratio for the variable Guardian of 1.897 indicates that the odds of being a recidi- vist, with other variables held constant, are about 90 percent more for first-time offenders living with a mother and a stepfather or with foster parents, than it is for a first-time offender in some other kind of living arrangement. 3.1.3.6 The Prediction of Recidivism The logit model can be used effectively to pre- dict or forecast the probability or the chances that a juvenile first-time offender will become a repeat offender. If a juvenile has predicted chances of offense status that are less than 0.5, it is said that his or her chances of recidivating are 0. If the prediction is greater than 0.5, the chances of becoming a recidivist are 1.0, i.e., it is highly likely that the juvenile will be a re- cidivist. Similarly, if predicted odds are greater than 1.0, then the probability or chances of the outcome will be, or correspond to a probability, greater than 0.5. And if the predicted odds are less than 1.0, then the probability will be less than 0.5. Therefore, in cases where the predic- tion of the odds is less than 1.0, or the probabil- ity is less than 0.5, it can be said that the juve- nile will likely not have another arrest. The data in Table 3.1.3.2 present various com- binations of offender characteristics and the associated chances of recidivism. The table shows the parameter estimates of the variables taken from Table 3.1.3.1, as well as various juvenile profiles. The predicted chances of the first profile shown, A1, are those of an offender who is eight years old, is male, was born in the US Virgin Islands, and who lives with a mother and stepfather, or with foster parents. Column (10) of the table shows that the predicted odds PRESENTATION ANALYSIS 38 PRESENTATION ANALYSIS Table 3.1.3.2 Predicted Probabilities in Logistic Regression: St Thomas, Juvenile Investigation Bureau Data Proba- Coefficients: a AgeFO AgeSq Sex Ethnicity1 Ethnicity2 Ethnicity3 Guardian Odds bility (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Parameter Estimates: -6.5287 1.1726 -0.0521 0.3220 0.2044 0.0083 0.2138 0.3201 Juvenile Profiles: A1 8 64 1 1 0 0 1 1.439 0.590 A2 9 81 1 1 0 0 1 1.918 0.657 A3 10 100 1 1 0 0 1 2.302 0.697 A4 11 121 1 1 0 0 1 2.490 0.713 A5 12 144 1 1 0 0 1 2.427 0.708 A6 13 169 1 1 0 0 1 2.131 0.681 A7 14 196 1 1 0 0 1 1.686 0.628 A8 15 225 1 1 0 0 1 1.202 0.546 A9 16 256 1 1 0 0 1 0.772 0.436 A10 17 289 1 1 0 0 1 0.447 0.309 A11 15 225 1 1 0 0 0 0.873 0.466 A12 12 144 0 1 0 0 1 1.759 0.638 A13 12 144 1 0 1 0 1 1.995 0.666 A14 12 144 0 0 0 1 1 1.775 0.640 A15 12 144 0 1 0 0 1 1.759 0.638 A16 12 144 0 0 0 1 1 1.775 0.640 A17 15 225 1 1 0 0 1 1.202 0.546 A18 15 225 0 0 1 0 1 0.716 0.417 A19 15 225 1 0 1 0 1 0.988 0.497 A20 15 225 1 1 0 0 0 0.873 0.466 A21 15 225 0 0 0 1 0 0.639 0.390 A22 10 100 0 1 0 0 0 1.211 0.548 Variables: AgeFO = Age of juvenile at first offense. AgeSq = The square of the age of juvenile at first offense. Sex = Sex of the juvenile (1 = male, 0 = female). Ethnicity1 = Virgin Island Born. Ethnicity2 = Puerto Rican Born. Ethnicity3 = Eastern Caribbean Born. Ethnicity4 = US Mainland Born. Guardian = Person with whom living (1 = mother and stepfather or foster parents, 0 = other guardian). 39 corresponding to this combination of character- istics are greater than 1.0—1.439—and there- fore have a prediction of recidivism of 0.590. Because this value is greater than 0.5, this may be interpreted that an 8-year-old male offender, born in the Virgin Islands and living with his mother and stepfather or with foster parents is highly likely to become a repeat offender. Profiles A1 to A10 show outcomes for juve- niles with identical characteristics except for age. Thus it is noted that as age increases for the foregoing combination of attributes, the chances of recidivating increase at a decreasing rate until age 11 (for profiles A1 to A4). This is illustrated graphically in Figure 3.1.3.1. It shows that the chances of repeat offenses in- crease at a decreasing rate until the chances peak with A4 at age 11, and then decrease at an increasing rate to age 17 thereafter. While the attributes of A8 indicate that this 15-year-old juvenile will be a recidivist, the chances are slightly less than that of a 14-year-old. How- ever, by the time the juvenile male with the as- sociated characteristics reaches age 16, his pre- dicted odds drop below 1.0 and his probability is less than 0.5, thus indicating he will not be a repeat offender. Juvenile profile A11—of a 15-year-old US Vir- gin Islands born male living with either both parents, or a mother alone, or a father, or a fa- ther and stepmother, a stepmother alone, a step- father, or with relatives—has predicted odds of less than 1.0 and a probability of 0.466. This juvenile is therefore predicted to be a non- repeat offender, and is not likely to be arrested again up to age 17. The profile of juvenile A12 is that of a 12-year-old female offender born in the US Virgin Islands who lives with her mother and stepfather or with foster parents. Her chances are predicted to be about 64 per- cent, and this indicates that she is highly likely to commit at least one other offense and be ar- A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 Juvenile Profile 0.0 0.2 0.4 0.6 Probability or Chances Figure 3.1.3.1 Predicted Chances of Recidivism: St Thomas JIB For a USVI-Born Male Living With Mother and Stepfather Note: See text for profile characteristics PRESENTATION ANALYSIS 40 rested before age 18. Other profiles in the table show similar combination of characteristics, and the odds and probabilities can be inter- preted similarly (Figure 3.1.3.2). 3.2 Department of Human Services Data Variables The following data items were collected from case files at the St Thomas Department of Hu- man Services (DHS): Sex Age Date of birth Place of birth Race Date of crime/incident Type of crime/incident Type of weapon used Status (outcome/disposition) Immigration status Mother in the household Father in the household Number of siblings in the household Education (highest level) Occupation Marital status Number of children Responsible for caring for children Type of live-in program, if any Head of household Marital relationship of natural parents Education of primary caretaker(s) Income (gross per month) Psychological Adjustment Self esteem Frequent arguments and fights with others Age at first contact with police Use of drugs/alcohol Use of drugs/alcohol before or during school Sexually active Number of different sex partners in past year Physical abuse PRESENTATION ANALYSIS 7 8 9 10 11 12 13 14 15 16 17 Age 0 5 10 15 20 Percent Figure 3.1.3.2 First-time Arrests of Juveniles in St Thomas Juvenile Investigation Bureau Data Male Female 41 There were 707 records in the St Thomas dis- trict DHS data set. As reported earlier these consisted of all juveniles who were on DHS files during the period 1987 to 1997. Many of the tables below will reflect differing totals be- cause of missing data. 3.2.1 St Thomas Data Description: Depart- ment of Human Services Juvenile’s Sex—According to Table 3.2.1 a large proportion of the total juvenile offend- ers—84.5 percent were males compared to 15.5 percent who were females. The age category here refers to 7- to 17-year olds: as a compari- son, the 5-to-19-year-olds in the total St Thomas/St John population had a 50.4 to 49.6 percent male-female ratio (as estimated in the 1995 Population and Housing Survey). This over representation of males involved in delin- quent behavior is a common and well-known phenomenon, though the reasons for this occur- rence are still not as clear. Immigration Status—In Table 3.2.2 US citi- zens made up the overwhelming majority of juvenile offenders—86.2 percent—whereas juveniles of resident alien status accounted for just 13.8 percent. The closest corresponding age group in the US Virgin Islands general population for which data are available is 5 to 19; US citizens and resident aliens or perma- nent residents in this group made up 88.4 and 10.8 percent, respectively (1995 Population and Housing Survey). Juvenile’s Age—Delinquent activity remained at a low rate between ages 7 to 9 (Table 3.2.3). At age 13 there was a relative picking up of activity and this increased dramatically be- tween ages 14 to 17 with a peak at ages 14 and 15. Although delinquent activity at age 17 re- mained in the high-level category, the begin- ning of a falling off can be observed. Juvenile’s Place of Birth—The data in Table 3.2.4 show that 74 percent of juveniles commit- ting first-time offenses were born in the US Virgin Islands, 4.3 percent on the US mainland, 17.2 percent in Eastern Caribbean islands, and 2.6 percent in Puerto Rico. Juvenile’s Race—Table 3.2.5 shows that 96.4 percent of juveniles involved in delinquent be- Table 3.2.1 Juvenile's Sex Number Percent Total 691 100.0 Male 584 84.5 Female 107 15.5 Sex Juvenile Table 3.2.2 Immigration Status Number Percent Total 623 100.0 US citizen 537 86.2 Resident Alien 86 13.8 Immigration Status Juvenile Table 3.2.3 Age of Juvenile Number Percent Total 693 100.0 7-9 35 5.1 10-12 102 14.7 13 80 11.5 14 123 17.7 15 129 18.6 16 118 17.0 17 106 15.3 Juvenile Age in Years PRESENTATION ANALYSIS 42 havior were black and 1.1 percent were white. Data on the age group 5 to 19 in the general population (1995 Population and Housing Sur- vey) indicated that 89.4 percent of St Thomas/St John’s population was black, 4.3 percent white, and 6.3 percent fell in the ‘Other’ category. These figures suggest that a disproportionately larger number of blacks be- came involved in delinquent behavior while whites and other groups were underrepresented based on their size in the general population. Household Status of Parents—Of the cases in Table 3.2.6 reporting data on the presence of parents in the household from which juvenile delinquents reside, 87.1 percent of mothers lived in the household compared to 30.6 per- cent of fathers. This supports a general concept in discourse on the causes of juvenile delin- quency. It suggests that in households where either parent is absent, the tendency is to foster a less supportive and stable environment that permits a higher level of risk to juveniles be- coming delinquent. Unfortunately, data are not available on the presence of mothers and fa- thers in the household for the general popula- tion. Age of Parents—In Table 3.2.7 fathers and mothers of juvenile delinquents appear to have generally fallen in different age categories. The majority of mothers—61.2 percent—fell in the 40 to 49 age group. The greater proportions of fathers (over half) were in the “50 years and above” age group. Fathers tended to be some- what older than mothers. Birthplace of Parents—The data in Table 3.2.8 show that only 35.7 percent of the moth- ers and 28.5 percent of the fathers were born in the US Virgin Islands. The percentage for the US mainland-born was 5.2 and 6.1 percent for mothers and fathers, respectively. The majority of the parents were born in the Eastern Carib- bean. The leading islands were St Kitts and Ne- vis with 18.6 and 17.6 percent, respectively if the data on place of birth of parents in the case files are compared on general population, a no- table feature emerges. In the general popula- PRESENTATION ANALYSIS Table 3.2.5 Juvenile's Race Number Percent Total 700 100.0 Black 675 96.4 White 8 1.1 Asian 2 0.3 Others 15 2.1 Juvenile Race Table 3.2.6 Household Status of Parents Number Percent Number Percent Total 635 100.0 604 100.0 In the household 553 87.1 185 30.6 Not in the household 82 12.9 419 69.4 Mother Father Household Presence Table 3.2.4 Juvenile's Place of Birth Number Percent Total 690 100.0 United States 30 4.3 Virgin Islands 514 74.5 Puerto Rico 18 2.6 Eastern Caribbean 119 17.2 Others 9 1.3 Juvenile Place of Birth 43 tion (1995 Population and Housing Survey) mothers and fathers born in St Kitts and Nevis were only 12.5 and 11.7 percent, respectively. In the case of Antigua and Barbuda, in the gen- eral population, percentages were 7.7 and 9.2 for mothers and fathers, respectively; whereas among juveniles in the data set, they were 10.6 and 11.8, respectively. Marital Status of Parents—Table 3.2.9 shows data on marital status for parents of those juve- nile delinquents. Available data show that in the highest category, 46.8 percent of mothers and 50.7 percent of fathers were married at the time of the juvenile’s first offense, (“Now Mar- ried”), though not necessarily to each other. The next two categories for mothers were “Never Married,” 28.2 and “Divorced” 12.3 percent. (Here again, divorce not necessarily between the two parents of the juvenile). For fathers, the next two categories were the same: “Never Married” was 24.3 percent and “Divorced”, 12.7 percent. A higher “Separated” rate, 9.4 percent, existed for fathers of juveniles than it did for mothers of juveniles, 8.5 percent. Employment Status of Parents—Of the cases in Table 3.2.10 reporting data on the employ- ment status of parents, 74.0 percent of mothers were employed compared to 73.5 percent of fathers. Juvenile’s Guardian—Of the reported cases of juvenile offenders that had information on the person or persons with whom they lived, 48.5 percent lived with their mother, 4.9 per- cent with their father, 23.4 percent with mother and father, 8.4 percent with relatives and 10.5 percent with a mother and a stepfather (Table 3.2.11). These data suggest that more juvenile delinquents who lived with their mother alone ended up becoming delinquent more than juve- niles in the other categories. With almost half of the juvenile cases living in households with a mother alone, the association between these two variables in the data is evident. Table 3.2.8 Place of Birth of Parents Number Percent Number Percent Total 614 100.0 555 100.0 United States 32 5.2 34 6.1 Virgin Islands 219 35.7 158 28.5 Puerto Rico 17 2.8 11 2.0 Eastern Caribbean 320 52.1 325 58.6 Others 26 4.2 27 4.9 Place of Birth Mother Father Table 3.2.7 Age of Parents Number Percent Number Percent Total 237 100.0 105 100.0 14-19 2 0.8 0 0.0 20-29 5 2.1 0 0.0 30-39 85 35.9 12 11.4 40-49 145 61.2 40 38.1 50 & above 0 0.0 53 50.5 Age in years Mother Father Table 3.2.9 Marital Status of Parents Number Percent Number Percent Total 365 100.0 276 100.0 Now married 171 46.8 140 50.7 Widowed 15 4.1 8 2.9 Divorced 45 12.3 35 12.7 Separated 31 8.5 26 9.4 Never married 103 28.2 67 24.3 Marital Status Mother Father Table 3.2.10 Employment Status of Parents Number Percent Number Percent Total 592 100.0 431 100.0 Employed 438 74.0 317 73.5 Unemployed 154 26.0 114 26.5 Employment Status Mother Father PRESENTATION ANALYSIS 44 Marital Relationship of Natural Parents— Data on current marital status of natural parents highlight categories for the natural parents of juvenile delinquents. At time of the first of- fense “Never Married" and “Divorced” catego- ries represented 36.6 and 12.4 percent, respec- tively, making up about half of the juvenile de- linquents. The proportion for the category “Married and Living Together” was 28.7 per- cent. The categories “One Parent Deceased” and “Separated” accounted for 12.1 and 10.2 percent respectively (Table 3.2.12). Siblings in the Household—Data in Table 3.2.13 indicate that juvenile offenders were most represented in households with one or two siblings present. These two categories com- bined represented 45.2 percent of all juvenile offenders. In households with three siblings, 21.6 percent of offenders were represented. In contrast, only 10.1 percent of juvenile offend- ers resided in households with four siblings and only 4.2 percent and 4.8 percent of juveniles lived in households with five or six or more siblings, respectively. School Enrollment—Table 3.2.14 indicates that 84.4 percent of juvenile delinquents were enrolled in school. Unfortunately, data indicat- ing actual attendance at school and class par- ticipation were not part of the data set. Al- though the majority of juveniles involved in delinquent behavior were enrolled in school, that does not mean they attended school, par- ticipated in and benefited from class activities. Table 3.2.12 Marital Relationship of Natural Parents Current status Number Percent Total 421 100.0 Married and living together 121 28.7 Separated 43 10.2 Divorced 52 12.4 Never married 154 36.6 One parent deceased 51 12.1 Both parents deceased 0 0.0 PRESENTATION ANALYSIS Table 3.2.13 Siblings in the Household Number of siblings Number Percent Total 523 100.0 No siblings 74 14.1 One siblings 118 22.6 Two siblings 118 22.6 Three siblings 113 21.6 Four siblings 53 10.1 Five siblings 22 4.2 Six or more siblings 25 4.8 Note: Median number of siblings in the household is 2. Table 3.2.14 School Enrollment Enrollment Status Number Percent Total 546 100.0 Enrolled 461 84.4 Not Enrolled 85 15.6 Table 3.2.11 Juvenile's Guardian Guardian Number Percent Total 629 100.0 Mother and Father 147 23.4 Mother 305 48.5 Father 31 4.9 Mother and Stepfather 66 10.5 Father and Stepmother 6 1.0 Foster parents 4 0.6 Relatives 53 8.4 Other 17 2.7 45 It is important to note that 15.6 percent were not enrolled in school, although the law stipu- lates that children below the age of 16 must be enrolled. Attendance at school and class is nei- ther enforced nor easily enforceable. Current or Highest Grade Completed— Table 3.2.15 indicates that juveniles involved in delinquent activities were represented across Grades 1 through 12. Children in Grades one through six were involved at low levels of juve- nile delinquent behavior except Grade two which was surprisingly high—30.6 percent. The peak was between juveniles in Grades seven through nine, with a range between 15.2 and 15.5 percent of the juvenile delinquents taking part. The proportion of juveniles in- volved in delinquent behavior began to de- crease at Grade 10 and continued through to the higher grades. It should be remembered that these data represented the educational status at time of first offense and cannot predict the edu- cational progress after that offense. 3.2.2 St Thomas: Department of Human Services Data Relationships Juvenile’s Employment Status—Table 3.2.16 indicates that the majority of juveniles involved in delinquent behavior were unemployed—86.7 percent. Here again it could help to have fig- ures showing the proportion of this age group of juveniles who are employed and unem- ployed in the general population in order to get a clearer picture. Current Job—Table 3.2.17 suggests that at the time of the first offense, the juvenile delin- quents were employed in a variety of job types and no particular activity seems to have had a monopoly on their time. It is assumed that these are part-time jobs but there is no way of know- ing weekly or monthly duration. Another point to remember is that only one out of 10 reported current jobs due to the fact that majority of them (86.7 percent) reported unemployed. Table 3.2.15 Current or Highest Grade Completed Grade Number Percent Total 689 99.6 Grade 1 1 0.1 Grade 2 212 30.6 Grade 3 0 0.0 Grade 4 10 1.4 Grade 5 7 1.0 Grade 6 20 2.9 Grade 7 105 15.2 Grade 8 111 16.0 Grade 9 107 15.5 Grade 10 59 8.5 Grade 11 34 4.9 Grade 12 23 3.3 Table 3.2.16 Juvenile's Employment Status Currently Working Number Percent Total 429 100.0 Employed 57 13.3 Unemployed 372 86.7 Table 3.2.17 Current Job Job Number Percent Total 68 100.0 Restaurant or Hotel 11 16.2 Construction 6 8.8 Store clerk 5 7.4 Mechanic shop 5 7.4 Gas station attendant 3 4.4 Other 38 55.9 PRESENTATION ANALYSIS 46 Children in the Household—Table 3.2.18 seems to suggest that households with one, two, three or four children were more likely to have delinquent juveniles—between 18.6 and 22.9 percent—than households with five or more children. It is possible that the presence of more children provide examples of alternate behavior and persons with whom youngsters may associ- ate and communicate with. It also may limit negative outside influences on their behavior and activities. Juvenile’s Children—Table 3.2.19 shows that a small proportion of juvenile delinquents—4.4 percent—had children of their own. Juveniles who were responsible for childcare—Table 3.2.20 show a slightly lower proportion—3.5 percent—than those with children. This respon- sibility did not necessarily refer to children owned by the juvenile delinquent, and could have been a younger sibling or relative. Presence of Risk Indicators—Each of the variables in Table 3.2.21 holds some signifi- cance for the juvenile’s psychological well- being. The presence of one or more of these factors could contribute to the undermining of PRESENTATION ANALYSIS Table 3.2.19 Juvenile's Children Parental Status Number Percent Total 405 100.0 Children 18 4.4 No Children 387 95.6 Table 3.2.20 Juvenile Responsible Childcare Number Percent Total 396 100.0 Yes 14 3.5 No 382 96.5 for Childcare Table 3.2.18 Number of Children in the Household Children Number Percent Total 485 100.0 0 5 1.0 1 90 18.6 2 88 18.1 3 111 22.9 4 95 19.6 5 36 7.4 6-10 53 10.9 11-14 4 0.8 15 or more 3 0.6 Table 3.2.21 Juvenile's Psychological Assessment Total Number Percent Number Percent Felt like a failure or worthless 171 44 25.7 127 74.3 Had trouble controlling anger 196 84 42.9 112 57.1 Had frequent arguments and fights 186 76 40.9 110 59.1 Ever been suspended from school 213 123 57.7 90 42.3 Ever been expelled from school 174 40 23.0 134 77.0 Used drugs or alcohol 167 43 25.7 124 74.3 Getting hurt (physically abused) 124 16 12.9 108 87.1 Yes No 47 the juvenile’s self esteem and self worth. The influence of any one or more of these variables at any level usually suggests negative impacts on behavior. There were relatively high per- centages in areas such as had trouble control- ling anger (42.9 percent), had frequent argu- ments and fights( 40.9 percent), ever been sus- pended from school (57.7 percent), used drugs or alcohol (25.7 percent), felt like a failure or worthless (25.7 percent), getting hurt (physically abused)—(12.9 percent), ever been expelled from school (23 percent). Juvenile’s Sexual Activity—Table 3.2.22 indi- cates that 56.7 percent of the first time offend- ers were sexually active. This variable is often identified as an indicator associated with delin- quent behavior. Forced Sex—Table 3.2.23 indicates that 6.5 percent of the juveniles with first offenses were forced to have sex. Looked at along with the previous table on Juvenile’s Sexual Activity, the data suggest that a large proportion of juve- niles opt to participate in sexual activity freely. Juvenile’s Sex Partners—In Table 3.2.24 the number of juveniles (32) for whom data are available here is relatively low. About 60 per- cent reported one or two sex partners but 9.4 percent reported six or more sex partners. The low number of cases for this variable was probably due to the reluctance of juveniles to be forthcoming with data that would obviously be considered very sensitive. Pre-outcome Status—Of the six offenses listed in Table 3.2.25 the number of juveniles involved increasingly decreased in number. The proportion of cases going to trial rather than plea agreement was lower for Offenses 1 and 2. The pattern was reversed for Offense 3 onwards, the proportion of cases going to trial, rather than plea agreement became close to 60 percent. Outcome of Offense—Table 3.2.26 gives a breakdown of the outcome of cases that went through the JIB and DHS systems for the six offenses being tracked. This was the final proc- essing of juvenile delinquents and represented the final sentencing stage. For the first and sec- ond the most common form of treatment is community service ranging between 21.3 to 31.0 percent of the cases. For cases ranging three to six offenses, how- ever, the treatment in the majority of instances changed to sentencing for a period at the Youth Rehabilitation Center (YRC)—between 31.3 to 39.5 percent of cases. It should be noted that Table 3.2.22 Juvenile's Sexual Activity Sexual Activity Number Percent Total 164 100.0 Sexually Active 93 56.7 Not Sexually Active 71 43.3 Table 3.2.23 Juvenile Forced to Have Sex Number Percent Total 123 100.0 Forced 8 6.5 Not Forced 115 93.5 Table 3.2.24 Juvenile's Sex Partners Sex Partners Number Percent Total 32 100.0 1 14 43.8 2 5 15.6 3 5 15.6 4 2 6.3 5 3 9.4 6-10 1 3.1 11 or more 2 6.3 PRESENTATION ANALYSIS 48 for this population of juvenile delinquents for whom the literature indicates there is high marijuana usage, drug rehabilitation is listed as a treatment only in Offenses 1, 2, 3 and 4 and that at an extremely low rate. 3.2.3 Predicting Recidivism with the St Thomas Department of Human Services Data This discussion focuses on forecasting the chances of recidivism, or the chances that a ju- venile will be arrested again after committing a first offense. While most of the features of this section are similar to those of 3.1.3.1— particularly the approach, the methodology and the model—there are two primary differences. First, the data set is confined to only those ju- veniles whose cases are on file at DHS for the period 1987 to 1997. The number of observa- tions in this data set—701—is less than half of that of JIB. Second, in addition to the variables employed in the model in section 3.1.3, there is an additional variable that was recorded in DHS case files, and that is the response to the question, Has … felt like a failure or worth- less? 3.2.3.1 Estimates of the St Thomas Depart- ment of Human Services Data The number of cases on which the DHS St Thomas analysis is based one is 686 out of a total of 701. Table 3.2.3.1 shows the results of the estimation of the logistic regression model. The likelihood ratio chi-square reported has a probability value less than 0.01, indicating that at least one of the coefficients for the explana- tory variables is not 0. A glance at the values of the parameter estimates of the explanatory variables in column (1) are not largely different from those in Table 3.1.3.1, because in many respects they are measuring characteristics of some of the same juvenile offenders. The standard errors that are associated with the parameter estimates are shown in column (2). Column (3) shows that the variables Age, Age squared and Sex are statistically significant. For example, the chances that they are not signifi- cant when one treats them as such are less than 1 in 10,000. The level of significance of US Virgin Island born category hovers around 0.05, with Puerto Rico or Dominican Republic Born category showing no statistical signifi- cance. The relatively small size of the number PRESENTATION ANALYSIS Table 3.2.25 Pre-Outcome Status Number Percent Number Percent Number Percent Total 366 100.0 130 100 70 100 Trial 170 46.4 62 47.7 41 58.6 Plea Agreement 196 53.6 68 52.3 29 41.4 Table 3.2.25 Pre-Outcome Status: Continued Number Percent Number Percent Number Percent Total 36 100 26 100 25 100 Trial 21 58.3 15 57.7 16 64.0 Plea Agreement 15 41.7 11 42.3 9 36.0 Offense 4 Offense 5 Pre-Outcome Status Pre-Outcome Status Offense 6 Offense 1 Offense 2 Offense 3 49 Table 3.2.26 Outcome Status Number Percent Number Percent Number Percent Total 632 100.0 211 100.0 112 100.0 Crisis Stabilization Center 10 1.6 2 0.9 0 0.0 Youth Rehabilitation Center 62 9.8 38 18.0 35 31.3 Boys/Girls Group Home 0 0.0 1 0.5 0 0.0 Transfer to Correctional Facility 1 0.2 2 0.9 5 4.5 Drug Rehabilitation Center 10 1.6 7 3.3 1 0.9 Psychological evaluation 18 2.8 5 2.4 2 1.8 Dismised 154 24.4 30 14.2 10 8.9 Left jurisdiction 8 1.3 1 0.5 0 0.0 Case not prosecuted 37 5.9 7 3.3 3 2.7 Case pending 6 0.9 6 2.8 0 0.0 Community service 196 31.0 45 21.3 27 24.1 Counseling 25 4.0 12 5.7 5 4.5 restit = $2,000 8 1.3 1 0.5 1 0.9 Attend school regularly 1 0.2 0 0.0 0 0.0 Academic prog 0 0.0 0 0.0 1 0.9 read/subm bk rep 0 0.0 1 0.5 0 0.0 curf/prnt/princ 10 1.6 6 2.8 3 2.7 rpmnd/prin/j off 11 1.7 3 1.4 2 1.8 other 75 11.9 44 20.9 17 15.2 Table 3.2.26 Outcome Status: Continued Number Percent Number Percent Number Percent Total 61 100.0 43 100.0 38 100.0 Crisis Stabilization Center 1 1.6 1 2.3 0 0.0 Youth Rehabilitation Center 22 36.1 17 39.5 15 39.5 Boys/Girls Group Home 0 0.0 0 0.0 0 0.0 Transfer to Correctional Facility 2 3.3 0 0.0 3 7.9 Drug Rehabilitation Center 1 1.6 0 0.0 0 0.0 Psychological evaluation 2 3.3 1 2.3 2 5.3 Dismised 9 14.8 5 11.6 3 7.9 Left jurisdiction 0 0.0 0 0.0 2 5.3 Case not prosecuted 2 3.3 1 2.3 0 0.0 Case pending 1 1.6 0 0.0 0 0.0 Community service 10 16.4 12 27.9 6 15.8 Counseling 0 0.0 1 2.3 2 5.3 restit = $2,000 0 0.0 0 0.0 0 0.0 Attend school regularly 0 0.0 0 0.0 0 0.0 Academic prog 0 0.0 0 0.0 0 0.0 read/subm bk rep 0 0.0 0 0.0 0 0.0 curf/prnt/princ 1 1.6 2 4.7 1 2.6 rpmnd/prin/j off 1 1.6 0 0.0 0 0.0 other 9 14.8 3 7.0 4 10.5 Offense 1 Offense 2 Offense 3 Outcome Status Offense 4 Offense 5 Offense 6 Outcome Status PRESENTATION ANALYSIS 50 of cases for this variable probably explains its low level of significance. Guardian is statisti- cally significant, and so is self-esteem, with both levels of significance well below 0.01. As was mentioned above in section 3.1.3, a positive parameter estimate produces an odds ratio greater than 1.0, and a negative odd ratio results in an odds ratio below 1.0. The ad- justed odds ratio for Age of 3.334 in column (4) leads one to the same conclusion as that in Table 3.1.3.1, that is, with a one-year increase in age, the chances of committing at least one other offense and being arrested, are more than three times as likely as not committing another offense, all other things being equal. The odds ratio for Sex of 2.435 indicates that males are about 2.5 times more likely than fe- males to commit at least one other offense for which they will be arrested. It is notable that this ratio is somewhat higher than that for males in the JIB data set. The odds ratios for the Eastern Caribbean Ethnicity variable— 3.027—suggest that offenders from the East- ern Caribbean are slightly more likely to be repeat offenders than VI-born offenders— 2.794. Both of these groups are about three times more likely to be repeat offenders than US-born offenders. The variable for Guardian does not appear to exert as much influence in the DHS data set as evidenced by the odds ratio of 1.081 compared PRESENTATION ANALYSIS Table 3.2.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Thomas, Department of Human Services Data Variable Parameter Estimate (1) Standard Error (2) Pr> Chi Square (3) Adjusted Odds Ratio (4) Intercept -5.1437 2.2758 <0.0001 … Age 1.2042 0.3443 <0.0001 3.334 Age Squared -0.5880 0.0130 <0.0001 0.943 Sex (1=Male) 0.4449 0.1154 <0.0001 2.435 Ethnicity USVI Borna 0.5192 0.1903 0.0493 2.794 Puerto Rico or Dominican Republic Borna -0.6105 0.3819 0.9896 0.903 Eastern Caribbean Borna 0.5995 0.2274 0.0554 3.027 Guardian (1=Mother and Stepfather or Foster Parents)b 0.0389 0.1421 0.0058 1.081 Self-Esteem 0.3502 0.1890 <0.0001 2.015 Likelihood Ratio Chi Square = 141.2858 df = 8 Pr > Chi Square < 0.0001 a Compared to US Born. b Compared to living with mother and father, mother only, father only, father and step mother, step mother only, step father only, foster parents, or relatives. Notes: N=686. Odds ratios greater than 1.0 indicate greater odds of recidivism. 51 Self- Proba- Coefficients: a AgeFO AgeSq Sex Ethnicity1 Ethnicity2 Ethnicity3 Guardian Esteem Odds bility (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) Parameter Estimates: -5.1437 1.2042 -0.0588 0.4449 0.5192 -0.6105 0.5995 0.0389 0.3502 Juvenile Profiles: B1 8 64 1 1 0 0 1 1 8.004 0.889 B2 9 81 1 1 0 0 1 1 9.821 0.908 B3 10 100 1 1 0 0 1 1 10.713 0.915 B4 11 121 1 1 0 0 1 1 10.391 0.912 B5 12 144 1 1 0 0 1 1 8.959 0.900 B6 13 169 1 1 0 0 1 1 6.868 0.873 B7 14 196 1 1 0 0 1 1 4.681 0.824 B8 15 225 1 1 0 0 1 1 2.836 0.739 B9 16 256 1 1 0 0 1 1 1.528 0.604 B10 17 289 1 1 0 0 1 1 0.732 0.423 B11 15 225 1 1 0 0 0 0 1.922 0.658 B12 12 144 0 1 0 0 1 1 5.742 0.852 B13 12 144 1 0 1 0 1 1 2.895 0.743 B14 12 144 0 0 0 1 1 1 6.222 0.862 B15 13 169 0 1 0 0 1 1 4.402 0.815 B16 14 196 0 0 0 1 1 1 3.251 0.765 B17 15 225 0 1 0 0 1 1 1.818 0.645 B18 15 225 0 0 1 0 1 1 0.587 0.370 B19 15 225 1 0 1 0 1 1 0.916 0.478 B20 15 225 1 1 0 0 0 0 1.922 0.658 B21 15 225 0 0 0 1 0 0 1.335 0.572 B22 10 100 0 1 0 0 0 0 4.653 0.823 Variables: AgeFO = Age of juvenile at first offense. AgeSq = The square of the age of the juvenile at first offense. Sex = Sex of the juvenile (1 = male, 0 = female). Ethnicity1 = Virgin Islander. Ethnicity2 = Hispanic. Ethnicity3 = Eastern Caribbean-born. Ethnicity4 = US Mainland-born. Guardian = Person with whom living (1 = mother and stepfather or foster parents, 0 = other guardian). Self-esteem = Self-esteem of the juvenile, coded 1 if person does not feel like a failure, otherwise 0. Table 3.2.3.2 Predicted Probabilities in Logistic Regression: St Thomas, Department of Human Services Data PRESENTATION ANALYSIS 52 to 1.897 in the JIB data. The odds ratio for Self-esteem of 2.015, a variable not included in the JIB data, indicates clearly that first-time offenders who report that they experience low esteem or feel worthless are twice as likely to be a recidivist as an offender who expresses self-assurance and feels self-confident. 3.2.3.2 The Prediction of Recidivism As in Table 3.1.3.2, Table 3.2.3.2 show several combinations of characteristics of juveniles that might help to determine whether a juvenile will be a recidivist or not. The set of variables is the same as that in the former table with the addition of the Self-Esteem variable. In a simi- lar way, the set of juvenile profiles is repeated so as to allow comparison with the JIB data. Table 3.2.3.2 illustrates that because the pa- rameter estimates of the coefficients in the DHS data are in general close to those of the JIB data, the odds and the associated probabili- ties are not very different. Hence, in the first 10 profiles of juveniles between age eight and age 17, male US Virgin Islands-born first-time offenders living with a mother and stepfather or foster parents and having low self-esteem are predicted to be repeat offenders. This combi- nation of demographic and social characteris- tics predicts recidivists for all ages consistently up to age 16. It is not until the youthful of- fender with these attributes reaches age 17 that the chances of recidivism are below 0.5, hence predicted to remain a one-time offender. (See section 3.1.3.6 for an explanation of interpreta- tion of the prediction probabilities). PRESENTATION ANALYSIS B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 Juvenile Profile 0.00 0.15 0.30 0.45 0.60 0.75 0.90 Probability or Chances Figure 3.2.3.1 Predicted Chances of Recidivism: St Thomas DHS For a USVI-Born Male Living With his Mother and a Stepson or Foster Parents Note: See text for profile characteristics 53 St Croix Data: Description and Analysis 3.3 Youth Investigation Bureau Data Variables The initial assumption of the researchers was that the same data collection instrument—the Contact Card and Arrest Report—that was utilized with the Investigation Bureau in St Thomas could also be used with the St Croix Youth Investigation Bureau data. This was not entirely the case as the St Croix office deemed it valuable to consider an additional variable for which data were immediately available. The data variables were as follows: Sex Age Date of birth Place of birth Race/ethnic origin Person with whom living Birthplace of mother and father School enrollment status Education (highest level) Employment status of guardian Date of crime/incident Type of crime/incident Type of weapon used Relative charged with criminal offense Degree of criminal involvement of relative The number of cases that occurred before 1987 were expected to be small since they repre- sented only those juveniles that committed of- fenses before the period of study but were still active after 1987. Data between 1987 and 1997, on the other hand, should have reflected all cases that committed first offenses during those years. It appears very likely that the number of cases used in data gathering particularly at YIB on St Croix, did not constitute the entire popula- tion of cases for the years stipulated and as an- ticipated in the parameters of the study. This is evident especially when observing Year of First Offense data. This would mean that all the YIB data would be affected by this absence of a full quota of cases even when it is not immediately obvious in the results. It is not possible to esti- mate or devise the extent to which the data pre- sented here do not represent the complete data set. The actual data gathering process, though guided by ECC, was carried out by YIB em- ployees in order to comply with confidentiality agreements. There are concerns as to whether this data set represents enough of the total population and what possible effect the inclusion of additional, but absent cases, would have had on the out- come of variables. Thought should be given also to how these factors restrict or facilitate the drawing of conclusions using this data. Following is an attempt to describe the data that are presented in frequency tables, followed by a section of cross tabulation and analysis based on the items outlined in the data gather- ing forms. Data are set out first for YIB, fol- lowed by those for DHS. An attempt is made to provide background information and brief de- scriptions toward a clearer appreciation of the data. 3.3.1 St Croix Data Description: Youth Investigation Bureau Year of First Offense—The period from which cases were chosen for this study was 1987 to 1997. All cases that were active during that time should have been included during the data gathering process. In some cases first and subsequent offenses had been committed before 1987 and these years and offenses were in- cluded in the data set as well. The data in Table 3.3.1 suggest a very uneven pattern in entry of juveniles as first time of- fenders into the YIB system. These data sug- gest that in 1989 and the period before had few recorded juvenile first offenses—a total of PRESENTATION ANALYSIS 54 three during the period before 1980 and in 1989—that required intervention by the YIB system. This pattern appears extremely unusual and might likely be a reflection of administra- tive recordkeeping up to the time of data gath- ering, and not necessarily an accurate account of the cases that were recorded by the YIB sys- tem during those years. The cluster of years beginning in 1990 and ex- tending to 1996 show juveniles as relatively active in committing first offenses. During that period, however, the data suggest a steep dip of two years where the percent of offenders fell to one and below—(1993, 1994). This is followed by a period of increased activity that peaked at 41.1 percent in 1995, followed by a drop to 7.5 percent in 1996. Once again, it is both an inter- esting and unusual phenomenon that first-time delinquent behavior would drop drastically to as low as zero in 1997 after such an intense pe- riod of activity, and suggests that this might be a reflection of a deficiency in administrative recordkeeping. No serious significance should therefore be placed on the distribution of first offenses over time using this data set. Type of First Offense—Table 3.3.2 lists first- time offenses under very broad categories. Part 1 Felony accounted for the largest single of- fense type—28.7 percent—and included crimes such as robbery, felonious assault, burglary, and grand larceny. Part II Felony—11.8 per- cent—covered crimes such as possession of a controlled substance, possession of unlicensed firearm, possession of stolen property, and un- authorized use of vehicle. A large proportion— 41.0 percent—of offense types fell under the category other, compiled from a number of un- defined offenses. Weapon Used in First Offense—According to Table 3.3.3, the highest proportion of cases fell in the other category, implying a number of un- defined objects that may be deemed weapons were used by juveniles while carrying out de- PRESENTATION ANALYSIS Table 3.3.3 Weapon Used in First Offense Weapon Juvenile Percent Total 634 100.0 Gun 101 15.9 Knife 50 7.9 Hand 187 29.5 Bottle 8 1.3 Penis 41 6.5 None 8 1.3 Other 239 37.7 Table 3.3.2 Type of First Offense Type of Offense Juvenile Percent Total 1824 100.0 Part I Felony 523 28.7 Part II Felony 215 11.8 Misdemeanor 187 10.3 Incidence 152 8.3 Other 747 41.0 Table 3.3.1 Year of First Offense Year of Offense Juvenile Percent Total 1821 100.0 Before 1980 1 0.1 1980 to 1988 1 0.1 1989 1 0.1 1990 102 5.6 1991 435 23.9 1992 361 19.8 1993 18 1.0 1994 13 0.7 1995 753 41.4 1996 136 7.5 55 linquent activities. Hand is the next highest category—29.5 percent—suggesting some amount of unplanned, spontaneous involve- ment. This is followed by gun and knife—15.9 and 7.9 percent, respectively. Each suggests some level of deliberate and planned armed ac- tivity. Penis—6.5 percent—implies the com- mitting of sexual offenses. Total Number of Offenses—Table 3.3.4 indi- cates that the vast majority of juveniles—84.6 percent—entered the system by committing a first offense and exited the system, never again committing an offense that led to arrest as a juvenile. Repeat offenders that committed two offenses before exiting the system made up 10 percent of the cases. The percent of juveniles committing three or more offenses was rela- tively small—5.4 percent. 3.3.2 St Croix: Youth Investigation Bureau Data Relationships Age at First Offense by Sex—Table 3.3.5 in- dicates juvenile delinquency levels were very low—below 1 percent—between ages 7 to 9 and increased to 2 percent for 10-year-olds. De- linquency rates continued to increase with age, climbing to 10.3 percent for 13-year-olds and remaining high until age 17. The peak of activ- ity occurred at age 15, with 22.3 percent of ju- veniles committing their first offense at this age. Table 3.3.4 Total Number of Offenses No. Of Offenses Juvenile Percent Total 1824 100.0 1 1544 84.6 2 183 10.0 3 49 2.7 4 18 1.0 5 4 0.2 6 17 0.9 7 3 0.2 8 2 0.1 9 0 0.0 10 3 0.2 11 0 0.0 12 1 0.1 Table 3.3.5 Age at First Offense by Sex Number Percent Number Percent Number Percent Total 1797 100.0 1285 100.0 511 100.0 Age 7 10 0.6 6 0.5 4 0.8 Age 8 12 0.7 9 0.7 3 0.6 Age 9 15 0.8 10 0.8 5 1.0 Age 10 36 2.0 33 2.6 3 0.6 Age 11 52 2.9 35 2.7 17 3.3 Age 12 83 4.6 52 4.0 30 5.9 Age 13 185 10.3 123 9.6 62 12.1 Age 14 324 18.0 205 16.0 119 23.3 Age 15 400 22.3 282 21.9 118 23.1 Age 16 368 20.5 263 20.5 105 20.5 Age 17 312 17.4 267 20.8 45 8.8 Male Female Age at First Offense Total PRESENTATION ANALYSIS 56 When the data are broken down by gender, it is observed that a pattern of increased activity oc- curred for both sexes between ages 13 to 17. The largest percent of males committed their first offense between ages 15 and 17, while the corresponding period for females was between ages 14 to 16. By age 17, females had dropped to 8.8 percent of cases having a first offense, while the males were still relatively active, with 20.8 percent committing a first-time offense. These data suggest that females became in- volved in delinquent activity at a younger age than males. Females also abandoned their de- linquent activity at an earlier age than did males. This observation coincides with other child development concepts that assign higher levels of physical and social development to females at an earlier age than to males. How- ever, given the uncertainty surrounding the completeness of the data set being used here, it would be inadvisable to apply such conclusions on any broad scale. Additional data could change the outcome of these variables. Juvenile’s Place of Birth—The data in Table 3.3.6 permit a comparison between the percent- age of juvenile offenders by ethnic origin and the actual percentage in the population at large. The data show that the percentage of juvenile offenders—7-to-17-year olds—born in the US Virgin Islands was 79.9 percent. This figure compares with 75.8 percent that made up the 5- to-19-year olds that were counted in the 1990 Census of Population. These data suggest, pro- portionately, that slightly more Virgin Islanders were among the offenders than their numbers in the population would suggest. The percentage of offenders born in Puerto Rico, Antigua and Barbuda and St Lucia were also slightly higher than their percentage in the total population would suggest. Juvenile’s Place of Birth by Sex—Table 3.3.7 shows males and females having similar per- centages of juvenile delinquents based on the total delinquents for each category. Here again, for both males and females, persons born in the US Virgin Islands constituted by far the largest proportion of delinquents—79.7 and 80.4 per- cent respectively. The next highest category was made up of those born in the US—6.6 percent males and 8.3 percent females. The remainder of the juve- nile delinquents born in other Caribbean islands or group of islands ranged between 1.1 and 2.6 percent, males and 1.0 and 3.3 percent, fe- males. Age at First Offense by Juvenile’s Place of Birth—In Table 3.3.8 suggests that juveniles from all ethnic groups under the age of 10 had a relatively low rate of committing their first offense, compared to the other age categories. Significant increases in the rate occurred be- tween ages 10 to 13 for all ethnic groups but with a noticeable difference in percentage for those born in Puerto Rico where it rose to 29.5 percent, which was a much higher rate than for similarly aged juveniles who were represented in the other ethnic categories. The peak age group for which first-time delinquent activity across all ethnic groups was between 14 to 16 year-olds with a falling off at the over 16 age category. Juveniles of Puerto Rican or Domini- PRESENTATION ANALYSIS Table 3.3.6 Juvenile's Place of Birth Place of Birth Number Percent Total 1825 100.0 US Virgin Islands 1458 79.9 Puerto Rico 48 2.6 United States 129 7.1 Antigua or Barbuda 36 2.0 Dominica 22 1.2 St Kitts or Nevis 32 1.8 St Lucia 42 2.3 Other Eastern Caribbean 39 2.1 Other 19 1.0 57 can Republic descent were especially likely to not have committed their first offense when over 16 years old—8.2 percent—when com- pared to those juveniles in the other place of birth categories—16.4 to 50.0 percent. Age at First Offense by Mother’s Place of Birth—In Table 3.3.9, a similar pattern appears as in Table 3.3.8 above. First-time delinquent activity remained low for juveniles under 10 years across all ethic groups, but increased sig- nificantly among the 10 to 13 age groups. There was a peaking of delinquent activity for juveniles with in the 14 to 16 age category for all ethnic groups. The trend that seems to follow juvenile delin- quents born in Puerto Rico was apparent here in the case of juveniles with mothers of Puerto Rican or Dominican Republic descent. For in- stance, juveniles with mothers who were born in Puerto Rico or Dominican Republic had a higher rate of a first offense in the 10 to 13 age group—30.2 percent—than juveniles with mothers born elsewhere—16.7 to 16.9 percent. This would seem to suggest that a greater pro- portion of juveniles with a mother from a Puerto Rican or Dominican Republic back- ground become involved in delinquent activity between ages 10 to 13. The relative fall off in first time offenses by juveniles over age 16 was Table 3.3.7 Juvenile's Place of Birth by Sex Number Percent Number Percent Number Percent Total 1825 100.0 1303 100.0 521 100.0 US Virgin Islands 1458 79.9 1038 79.7 419 80.4 Puerto Rico 48 2.6 31 2.4 17 3.3 United States 129 7.1 86 6.6 43 8.3 Antigua or Barbuda 36 2.0 28 2.1 8 1.5 Dominica 22 1.2 17 1.3 5 1.0 St Kitts or Nevis 32 1.8 26 2.0 6 1.2 St Lucia 42 2.3 34 2.6 8 1.5 Eastern Caribbean 39 2.1 29 2.2 10 1.9 Other 19 1.0 14 1.1 5 1.0 Place of Birth Total Male Female Table 3.3.8 Age at First Offense by Juvenile's Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 1431 100.0 61 100.0 128 100.0 169 100.0 6 100.0 Under 10 yrs 30 2.1 2 3.3 4 3.1 1 0.6 0 0.0 10 to 13 yrs 286 20.0 18 29.5 24 18.8 26 15.4 1 16.7 14 to 16 yrs 863 60.3 36 59.0 79 61.7 111 65.7 2 33.3 Over 16 yrs 252 17.6 5 8.2 21 16.4 31 18.3 3 50.0 United States Eastern Caribbean Other Dominican Republic Age at First Offense Place of Birth Puerto Rico or Virgin Islands PRESENTATION ANALYSIS 58 especially apparent for those juveniles with mothers of Puerto Rican or Dominican Repub- lic descent—10.5 percent. In contrast, juve- niles beyond the age of 16 with mothers who were not born in Puerto Rico or the Dominican Republic had rates of a first-time offense that ranged from 13.3 percent to 50 percent. Age at First Offense by Father’s Place of Birth—It is of interest to note that the pattern apparent in the total population of juvenile de- linquents and those with fathers born in Puerto Rico or Dominican Republic in Table 3.3.10 is similar to that found in two previous tables de- picting Age at First Offense by Place of Birth Group, and Age at First Offense by Mother’s Place of Birth. The involvement in delinquent activity became apparent in the 10 to 13 age group for all juveniles across all place of birth categories. The rate of first time offenses also peaked at the 14 to 16 age category for all juve- niles, regardless of their father’s place of birth. Juvenile’s Race by Sex—Table 3.3.11 shows that of the total number of juvenile delinquents who committed first offenses, 76.8 percent were black, 1.7 percent white, a small percent was Asian, and an undefined Other was 21.4 percent. These proportions were generally con- sistent for males and females. Slightly more males than females—77.4 percent versus 75.1 percent—were among black offenders. PRESENTATION ANALYSIS Table 3.3.9 Age at First Offense by Mother's Place of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 1355 100.0 86 100.0 90 100.0 219 100.0 6 100.0 Under 10 yrs 26 1.9 2 2.3 2 2.2 4 1.8 0 0.0 10 to 13 yrs 268 19.8 26 30.2 15 16.7 37 16.9 1 16.7 14 to 16 yrs 823 60.7 49 57.0 61 67.8 136 62.1 2 33.3 Over 16 yrs 238 17.6 9 10.5 12 13.3 42 19.2 3 50.0 Puerto Rico or Virgin Islands United States Eastern Caribbean Mother's Place of Birth Age at First Offense Other Dominican Republic Table 3.3.10 Age at First Offense by Father's Pace of Birth Number Percent Number Percent Number Percent Number Percent Number Percent Total 728 100.0 88 100.0 57 100.0 187 100.0 4 100.0 Under 10 yrs 15 2.1 3 3.4 1 1.8 3 1.6 0 0.0 10 to 13 yrs 142 19.5 24 27.3 12 21.1 28 15.0 1 25.0 14 to 16 yrs 445 61.1 55 62.5 37 64.9 118 63.1 0 0.0 Over 16 yrs 126 17.3 6 6.8 7 12.3 38 20.3 3 75.0 Father's Place of Birth Age at First Offense Puerto Rico or Other Virgin Islands Dominican Republic United States Eastern Caribbean 59 Juvenile’s Guardian by Sex of Juvenile—The data in Table 3.3.12 suggests that the majority of juvenile offenders—78.7 percent—lived with their mother; the second highest percent— 18.3 percent—lived with their mother and fa- ther. The remainder of the guardian categories ranged between 0 to 1.3 percent. When looked at by gender, both male and fe- male categories of juvenile delinquents fol- lowed similar patterns within similar ranges. Female offenders were slightly more like to live with both their mother and father than were male offenders—21.3 percent and 17.0 percent, respectively. All the remaining of guardian categories experienced relatively low propor- tions of the total number of juvenile delin- quents. One can assume with some confidence that this is quite likely a reflection of the repre- sentation of these groups in the general Virgin Islands’ population, namely, the majority of children in this age group live with Mothers alone and the second largest group with Mother and Father. However, a categorical statement cannot be made on this matter since there is no corresponding data available for comparison in the general population data on St Croix. Table 3.3.11 Juvenile's Race by Sex Number Percent Number Percent Number Percent Total 1819 100.0 1299 100.0 519 100.0 Black 1397 76.8 1006 77.4 390 75.1 White 31 1.7 22 1.7 9 1.7 Asian 1 0.1 1 0.1 0 0.0 Other 390 21.4 270 20.8 120 23.1 Race Total Male Female Table 3.3.12 Juvenile's Guardian by Sex Number Percent Number Percent Number Percent Total 1726 100.0 1236 100.0 489 100.0 Mother and Father 315 18.3 210 17.0 104 21.3 Mother 1359 78.7 989 80.0 370 75.7 Father 12 0.7 10 0.8 2 0.4 Mother and Step Father 3 0.2 1 0.1 2 0.4 Father and Step Mother 1 0.1 0 0.0 1 0.2 Step Mother 0 0.0 0 0.0 0 0.0 Step Father 0 0.0 0 0.0 0 0.0 Foster Parents 1 0.1 0 0.0 1 0.2 Relatives 23 1.3 19 1.5 4 0.8 Other 12 0.7 7 0.6 5 1.0 Guardian Total Male Female PRESENTATION ANALYSIS 60 School Enrollment at Time of Arrest—Table 3.3.13 indicates that only a small proportion of juvenile delinquents—3.7 percent—were not enrolled in school at the time of arrest. This suggests that being enrolled in school is not a deterrent to youngsters committing offenses. A more comprehensive picture might be gained if data were available for analysis on levels of school and class attendance, and even achieve- ment measures. School Type at Time of Arrest—Table 3.3.14 shows data on school type at which juvenile delinquents were enrolled at the time of first offense. These data seem to correlate with pre- vious sets of data that reflected age and first offense and show that a relatively low percent- age of juvenile delinquents committed their first offense in elementary school—5.5 per- cent—increasing to 23.3 percent in Jr. high school and rising to 71.1 percent by high school. Mother’s Place of Birth by Juvenile’s Sex— Table 3.3.15 reflects a similar pattern encoun- tered in juvenile’s place of birth. The majority of juvenile delinquents, 77.3 percent, came from mothers born in the US Virgin Islands. The second highest proportion was from the United States—5.1 percent. Puerto Rico and St Lucia each had around four percent. The pro- portion of juvenile delinquents with mothers born in other Caribbean islands or island group fell to approximately between one and three percent. PRESENTATION ANALYSIS Table 3.3.13 School Enrollment Status Number Percent Total 1789 100.0 Enrolled 1722 96.3 Not Enrolled 67 3.7 Table 3.3.14 School Type School Number Percent Total 1684 100.0 Elementary 93 5.5 Jr. High 393 23.3 High 1198 71.1 Table 3.3.15 Mother's Place of Birth by Sex of Juvenile Number Percent Number Percent Number Percent Total 1785 100.0 1277 100.0 507 100.0 Virgin Islands 1380 77.3 986 77.2 394 77.7 Puerto Rico 72 4.0 48 3.8 24 4.7 United States 91 5.1 56 4.4 34 6.7 Antigua or Barbuda 38 2.1 28 2.2 10 2.0 Dominica 27 1.5 21 1.6 6 1.2 St Kitts or Nevis 36 2.0 28 2.2 8 1.6 St Lucia 74 4.1 58 4.5 16 3.2 Eastern Caribbean 46 2.6 36 2.8 10 2.0 Other 21 1.2 16 1.3 5 1.0 Place of Birth Total Male Female 61 Mother’s Employment by Sex of Juvenile— Table 3.3.16 shows that only a relatively small proportion, 5.5 percent, of the total juvenile delinquents were of mothers who were unem- ployed. The percentage of female offenders with mothers who were employed was slightly higher than it was for male offenders. This may suggest that employed mothers were away from the home for extended hours especially at times when children returned home from school. Data do not indicate if any of those mothers held more than one job or worked overtime, factors that would obviously have added to the mother- child separation and left unsupervised children exposed and vulnerable to involvement in de- linquent behavior. Father’s Place of Birth by Juvenile’s Sex— Table 3.3.17 indicates that 68.8 percent of juve- nile offenders had fathers who were born in the Virgin Islands; those born of Puerto Rican fa- thers made up 7.2 percent and offenders with St Lucian fathers made up 6.7 percent. Female offenders were slightly more likely to have fa- thers who were born in the US Virgin Islands and Puerto Rico than male offenders. Father’s Employment by Juvenile’s Sex— The employment statistics of fathers of juvenile offenders at time of first offense follows closely that of mother's employment statistics (Table 3.3.18). A small percentage of fathers— 5.4 percent—were unemployed. Unemploy- Table 3.3.16 Mother's Employment by Sex of Juvenile Number Percent Number Percent Number Percent Total 1398 100.0 1000 100.0 397 100.0 Employed 1321 94.5 939 93.9 381 96.0 Unemployed 77 5.5 61 6.1 16 4.0 Employment Status Total Male Female Table 3.3.17 Father's Place of Birth by Sex of Juvenile Number Percent Number Percent Number Percent Total 1079 100.0 757 100.0 321 100.0 Virgin Islands 742 68.8 508 67.1 233 72.6 Puerto Rico 78 7.2 50 6.6 28 8.7 United States 57 5.3 43 5.7 14 4.4 Antigua or Barbuda 28 2.6 21 2.8 7 2.2 Dominica 17 1.6 12 1.6 5 1.6 St Kitts or Nevis 37 3.4 29 3.8 8 2.5 St Lucia 72 6.7 59 7.8 13 4.0 Eastern Caribbean 34 3.2 25 3.3 9 2.8 Other 14 1.3 10 1.3 4 1.2 Place of Birth Total Male Female PRESENTATION ANALYSIS 62 ment was slightly higher for the fathers of male juvenile offenders—5.6 percent compared to the fathers of female offenders—5.0 percent. Relationship Between Relatives With Previ- ous Criminal Offenses and Recidivism— Officers of the YIB in St Croix expressed the view that, based on their experience, there was an observable link between the number of re- peat offenders and homes with relatives that had previously been charged with criminal of- fenses. The data in Table 3.3.19 were compiled to investigate this assumed relationship. From the data in the table, it can be observed that of the total 1,782 youth charged with of- fenses, 505 or 28.3 percent of them came from homes in which relatives had been previously charged with criminal offenses. The table highlights that of the 505 homes with previ- ously charged relatives, 35.4 percent of them were recidivists compared to the 7.7 percent recidivists that came from homes in which there were no relatives charged with offenses. The data in the table were further subjected to more rigorous statistical analysis to test the proposition of a significant relationship be- tween households with relatives that had previ- ously been charged and recidivism among youth. The test does strongly support the view that there is a very strong relationship between the rate of recidivism and households with rela- tives that have been charged with criminal ac- tivities. The p-value at the bottom of the table indicates that this conclusion is likely to be in- correct in less than 1 in 10,000 times. Juvenile’s Age by Current or Highest Grade—Table 3.3.20 suggests that juveniles offenders tended to be in grade levels consis- tent with their age. There were a few excep- tions. For instance, there were several 14 and 15 year olds who had not reached Grade 8 yet and some 16 and 17 year olds who were at a grade level less than 10. The existence of some juveniles being at a grade level much PRESENTATION ANALYSIS Table 3.3.18 Father's Employment by Sex of Juvenile Number Percent Number Percent Number Percent Total 719 100.0 517 100.0 201 100.0 Employed 680 94.6 488 94.4 191 95.0 Unemployed 39 5.4 29 5.6 10 5.0 Employment Status Total Male Female 63 lower than would be expected given their age may indicate the propensity for such juveniles to be underachievers. However, since social promotion was practiced throughout the public school system of the US Virgin Islands at the period referenced by the study, it is difficult to assess whether underachievement as indicated by grade level was a factor influencing behav- ior. 3.3.3 Predicting Recidivism with the Youth Investigation Bureau Data The methods and procedures utilized in this section and similar to those that were applied to the St Thomas JIB and DHS data above. (See Sections 3.1.3.1, 3.1.3.2, and 3.1.3.4 for a thor- ough discussion of the approach, methodology and recidivism model.) There is a notable difference between the data of the Youth Investigation Bureau (YIB) in St Croix and the JIB data in St Thomas. Some of the variables that were significant in the St Thomas model showed no statistically sig- nificant association with recidivism, while a new set of variables were clearly very good predictors of repeat arrests. For example, it was shown above empirically that age is a very strong predictor of recidivism in St Thomas in both the JIB and DHS data sets, but in the St Croix data set, it is a very poor predictor. Similarly, ethnicity and the role of the parents or guardians fail to exert any statistically strong Table 3.3.20 Juvenile's Age by Current or Highest Grade Number Percent Number Percent Number Percent Number Percent Total 1670 100.0 133 100.0 112 100.0 237 100 7-9 36 2.2 33 24.8 0 0.0 1 0.4 10-12 165 9.9 95 71.4 58 51.8 7 3.0 13 178 10.7 2 1.5 50 44.6 115 48.5 14 311 18.6 2 1.5 2 1.8 101 42.6 15 379 22.7 0 0.0 2 1.8 12 5.1 16 347 20.8 0 0.0 0 0.0 0 0.0 17 254 15.2 1 0.8 0 0.0 1 0.4 Total Grade 1-6 Grade 7 Grade 8 Age in years Table 3.3.20 Juvenile's Age by Current or Highest Grade: Continued Number Percent Number Percent Number Percent Number Percent Total 358 100.0 450 100.0 311 100.0 69 100.0 7-9 0 0.0 0 0.0 2 0.6 0 0.0 10-12 4 1.1 0 0.0 1 0.3 0 0.0 13 8 2.2 3 0.7 0 0.0 0 0.0 14 196 54.7 7 1.6 0 0.0 3 4.3 15 134 37.4 228 50.7 3 1.0 0 0.0 16 14 3.9 203 45.1 129 41.5 1 1.4 17 2 0.6 9 2.0 176 56.6 65 94.2 Age in years Grade 12 Grade 9 Grade 10 Grade 11 PRESENTATION ANALYSIS 64 relationship with repeat offense behavior. The immediate suggestion is that both sets of juvenile come from significantly different populations or social milieus, and that attempts to prescribe similar explanations for self-same behaviors could be enormously misleading. The variables included in the St Croix models reflect such differences. 3.3.3.1 Definition of Variables Dependent Variable: As in the previous cases for St Thomas, the outcome or dependent vari- able of interest is recidivism. Each juvenile was classified in one of two discrete categories. If the youth were charged with more than one offense, (s)he was classified as a repeat of- fender or recidivist. The other classification comprised youth who were charged with a sin- gle offense. This latter group was made the referent group. Independent or Explanatory Variables: Ide- ally, it would have been most valuable to com- pare the impact of the same variables on St Croix as were used for St Thomas, but the dif- ferences in data precluded this. For example, preliminary analysis showed considerable in- consistency with the age distribution of the YIB data, hence they were excluded from the analy- sis. This is a significant loss since theory does point to age as critical in this prediction model. Additionally, one new variable that relates to the involvement of relatives of the juvenile in criminal activity appears to be an essential pre- dictor. i) Gender: As in the previous models, this variable is dichotomous and refers to the sex of the juvenile. Female is the referent category. ii) School Enrollment: This variable indi- cates whether the juvenile was enrolled in school at the time of the first offense. The referent group was “not enrolled in school at the time of first offense”. iii) Grade: This variable is dichotomous and refers to whether the offender was in Grade 12 or in a grade level less than 12 when first charged with an offense. The referent group was Grade 12. iv) Relative Charged: This dichotomous variable classified each youth as to whether (s)he had a relative who had been charged with a crime previously. The referent group was those who did not have relatives that had been previ- ously charged with a criminal offense. 3.3.3.2 Estimates of the St Croix Youth Investigation Bureau Data The total number of cases that were recorded in the study period was 1,827. In some instances, data items were missing and some cases had to be dropped from a particular analysis. In the application of the logistic regression model, 1,796 valid cases were used. Four variables in the YIB data set show strong association with repeat offenses. The sex of the arrested youth remains a reliable predictor. En- rollment in school at the time of the first arrest is also a good indicator of the type of subse- quent behavior that might be expected from first-time arrested youth. A particularly strong predictor is related to whether the first-time arrested youth had rela- tives that had been charged with criminal of- fenses at any time previously. Data analysis in 3.2.2 (Table 3.3.19) illustrated that there is a statistically significant relationship between deviant behaviors in youth when a relative has been known to have been charged with criminal offenses in the past. The odds ratio as a meas- ure of the strength of the relationship confirms that a first-time offender with relatives charged with criminal offenses is seven times more likely to be a repeat offender than one who PRESENTATION ANALYSIS 65 does not have relatives charged with such of- fenses. The data in Table 3.3.3.1 are interpreted simi- larly to those in 3.1.3.1 and 3.2.3.1 above. The odds ratio of 1.970 for Sex denotes that the pre- dicted odds of recidivism for a first-time male offender are about twice the odds for a female offender or 97 percent higher than the odds for a female first-time offender. The adjusted odds ratio of a first-time offender who is enrolled in school is 3.595. This means that the odds of an offender who was not en- rolled in school at the time of his or her first arrest becoming a repeat offender are at least three and one-half times the odds of someone who is enrolled. Equivalently, it can be said that the odds of repeat arrests of a juvenile who was enrolled at school are only about 28 per- cent of the odds of someone who was not en- rolled in school. The grade level of a juvenile is also an indica- tor of what to anticipate in a youth’s behavior. The odds ratio for Grade in Table 3.3.3.1 is 2.375. This represents the predicted odds of recidivism of a youth who was in a grade below 12 at the time of first arrest. The odds of re- cidivism for youngsters who are below Grade 12 are more than twice—2.375—as high as those who committed their first offense in Grade 12. In other words, the odds of recidi- vism for juveniles in Grade 12 are only 42 per- cent of the odds of a youth who had not reached Grade 12. It is to be recalled that in section 3.3.3 it was shown that there was a very strong association between repeat offense behavior and the pres- ence in the lives of these youth of relatives who have been charged with criminal offenses (Table 3.3.19). This relationship is maintained at about the same level even in the presence of other explanatory variables. The table shows that the adjusted odds ratio for the Relatives Charged variable is 6.908. This value asserts that the predicted odds for repeat arrests from a first-time male offender with relatives who have been charged with criminal behavior in the past are about seven times the odds of a first-time offender without relatives having such a criminal history. Table 3.3.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Croix, Youth Investigation Bureau Data Variable Parameter Estimate (1) Standard Error (2) Pr> Chi Square (3) Adjusted Odds Ratio (4) Intercept -1.5531 0.3361 <0.0001 … Sex (1=Male) 0.3389 0.0937 0.0003 1.970 School Enrollment (1=Enrolled) 0.6398 0.2513 0.0109 3.595 Grade (1=Less than Grade 12) 0.4324 0.2265 0.0562 2.375 Relative Charged (1=Yes) 0.9664 0.0757 0.0001 6.908 Likelihood Ratio Chi Square = 204.1609 df = 4 Pr > Chi Square < 0.0001 Notes: N=1796. Odds ratios greater than 1.0 indicate greater odds of recidivism. PRESENTATION ANALYSIS 66 3.3.3.3 The Prediction of Recidivism The probabilities in Table 3.3.3.2 indicate the chances of recidivism or non-delinquent behav- ior—using YIB St Croix data—for a given combination of the variables under study. (See also Figures 3.3.3.1 and 3.3.3.2). Again, for odds—in column (6)—greater than 1.0 or for probabilities—in column (7)—greater than 0.5, the combination of characteristics predicts that the juvenile will commit offenses for which he or she will be arrested. And for odds less than 1.0 or probabilities less than 0.5, the prediction is that a first-time offender will not be arrested as a juvenile again. The profile for juvenile C1 in the table is inter- preted as follows. A first-time male offender who is not enrolled in school at the time of his arrest, who has not reached Grade 12, and who has relatives that had been charged with crimi- nal offenses has odds of 2.281 and probability of 0.695, thus indicating that this youth has a high probability of being a recidivist. Profile C2 is that of a female who is enrolled in school in Grade 12 with relatives charged for commit- ting criminal offenses is predicted to be a one- time offender and not be a recidivist. The youth that is represented by profile C11 is that of a female with one offense, who is enrolled in Grade 12 at the time of arrest, and who does PRESENTATION ANALYSIS Table 3.3.3.2 Predicted Probabilities in Logistic Regression: St Croix, Youth Investigation Bureau Data Coefficients: a Sex Enrollment Grade Rel. Charged Odds Probability (1) (2) (3) (4) (5) (6) (7) Parameter Estimates: -1.5531 0.3389 0.6398 0.4324 0.9664 Juvenile Profiles: C1 1 1 1 1 2.281 0.695 C2 0 1 1 1 1.625 0.619 C3 1 1 0 1 1.480 0.597 C4 1 0 1 1 1.203 0.546 C5 0 1 0 1 1.055 0.513 C6 1 1 1 0 0.868 0.465 C7 0 0 1 1 0.857 0.462 C8 1 1 0 0 0.563 0.360 C9 0 0 0 1 0.556 0.357 C10 1 0 1 0 0.458 0.314 C11 0 1 0 0 0.401 0.286 C12 0 0 0 0 0.212 0.175 Variables: Sex = Sex of the juvenile (1 = male, 0 = female). Enrollment = Coded 1 if not enrolled in school at first arrest, otherwise 0. Grade = Coded 1 if school grade is less than 12th grade, otherwise 0. Rel. Charged = Coded 1 if a relative was charged with a criminal offense, otherwise 0. 67 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 Juvenile Profile 0.00 0.15 0.30 0.45 0.60 0.75 Probability or Chances Figure 3.3.3.1 Predicted Changes of Recidivism: St Croix YIB For Various Combinations of Youth Characteristics Note: See text for profile characteristics 7 8 9 10 11 12 13 14 15 16 17 Age 0 5 10 15 20 Percent Figure 3.3.3.2 Age at First Offense: St Croix Youth Investigation Bureau Data Male Female PRESENTATION ANALYSIS 68 not have relatives charged with any criminal offense. A female offenders with these charac- teristics is predicted to have a relatively low chance of being a repeat offender. It is notable that for most variable combinations that include the ‘Relatives Charged’ variable, the prediction is that the juvenile will be a repeat offender. This is consistent with all of the foregoing analyses that include this particular variable. 3.4 Department of Human Services Data Variables The following data items were collected from case files at the St Croix Department of Human Services (DHS): Sex Immigration Status Age Place of Birth Race Household Status of Parents Age of Parents Marital Status of Parents Guardian Marital Relationship of Natural Parents Number of Siblings in the Household School Enrollment Employment Status Juvenile Children of Juvenile Psychological Assessment of Juvenile Sexual Activity of Juvenile Pre-outcome Status of Juvenile Type of Offense Year of Offense Outcome of Offense Biological Father Known by Juvenile There were 888 records in the St Croix district DHS data set. These records consist of all ju- veniles who were on DHS files during the pe- riod 1987 to 1997. As a result of missing data, many of the tables below will reflect differing totals. 3.4.1 St Croix Data Description: Department of Human Services Juvenile’s Sex—According to Table 3.4.1 a large proportion of the total juvenile offend- ers—85.9 percent—were males compared to 14.1 percent who were females. The age cate- gory here refers to seven-to 17-year-olds: as a comparison, the five-to 19-year-olds in the total St Croix population had a 49.9 to 50.1 male- female ratio (1995 Population and Housing Survey). This over representation of males in- volved in delinquent behavior is a common and well-known phenomenon, though the reasons for the occurrence are still not as clear. Immigration Status—In Table 3.4.2 US citi- zens made up the overwhelming majority of juvenile offenders—94.1 percent—whereas juveniles of resident alien status accounted for just 5.9 percent. The closest corresponding age group in the St Croix general population for which data are available is five to 19; US citi- zens and resident aliens or permanent residents in this group made up 88.4 and 10.8 percent, respectively (1995 Population and Housing Survey). PRESENTATION ANALYSIS Table 3.4.1 Juvenile's Sex Number Percent Total 885 100.0 Male 760 85.9 Female 125 14.1 Sex Juvenile Table 3.4.2 Immigration Status Number Percent Total 717 100.0 US Citizen 675 94.1 Resident Alien 42 5.9 Juvenile Immigration Status 69 Juvenile’s Age—The data in Table 3.4.3, Juve- niles’ Age suggest that delinquent activity re- mained at a low rate between ages seven to 12. At age 13 there was a relative increase of activ- ity and this delinquent activity increased more dramatically between ages 14 to 17 with a peaking at ages 15 and 16. Although delinquent activity at age 17 remained relatively high, the beginning of a falling off is observed. Juvenile’s Place of Birth—The data in Table 3.4.4 indicate that 84.6 percent of juveniles committing first-time offense were born in the US Virgin Islands, 5.0 percent in the US and a combined 6.7 percent for other Caribbean ar- eas, except Puerto Rico with 1.9 percent. In comparison, figures for five- to 19-year-olds in the general population of St Croix for which data is available show persons born in St Croix account for 74.6 percent of the population. Per- sons born in Puerto Rico were 1.9 percent and in the matching Caribbean areas combined 6.7 percent. The US mainland accounted for 9.9 percent (1995 Population and Housing Survey). It must be noted that these figures represent the entire US Virgin Islands population, and that these groups are not necessarily spread uni- formly across the islands. Juvenile’s Race—Table 3.4.5 indicates that 98.1 percent of juveniles involved in delinquent behavior were black and 1.1 percent were white. Data on the age group five to 19 in the general population (1995 Population and Hous- ing Survey) indicated that 72.4 percent of St Croix’s population was black, 5.7 percent white, and 21.8 percent fell in the other cate- gory. These figures suggest that a dispropor- tionately larger number of blacks became in- volved in delinquent behavior, while whites and other groups were underrepresented based on their size in the St Croix population. Household Status of Parents—Of the cases in Table 3.4.6 reporting data on the presence of parents in the household from which juvenile delinquents reside, 86.1 percent of mothers lived in the household compared to 27.8 per- Table 3.4.3 Age of Juvenile Number Percent Total 860 100.0 7-9 3 0.3 10-12 31 3.6 13 57 6.6 14 151 17.6 15 202 23.5 16 219 25.5 17 197 22.9 Juvenile Age in years Table 3.4.4 Juvenile's Place of Birth Number Percent Total 732 100.0 United States 37 5.0 Virgin Islands 619 84.6 Puerto Rico 13 1.8 St Kitts or Nevis 7 0.9 Antigua or Barbuda 10 1.4 Dominica 8 1.1 Other Eastern Caribbean 24 3.3 Other 14 1.9 Sex Juvenile Table 3.4.5 Juvenile's Race Number Percent Total 639 100.0 Black 627 98.1 White 7 1.1 Asian 5 0.8 Sex Juvenile PRESENTATION ANALYSIS 70 cent of fathers. This supports a general concept in discourse on the causes of juvenile delin- quency that suggest households where either parent is absent may foster a less supportive and stable environment that permits higher risk to juveniles becoming delinquent. Unfortu- nately, data are not available on the presence of mothers and fathers in the household for the general population of St Croix. Age of Parents—In Table 3.4.7, fathers and mothers of juvenile delinquents appear to be differently represented in the various age cate- gories. The majority of mothers—51.1 per- cent—fell in the 30 to 39 age group, whereas fathers were most represented in the 40 to 49 age category. Thus, fathers tended to be some- what older than mothers. Marital Status of Parents—Table 3.4.8 shows data on the marital status for parents of those juvenile delinquents. Available data show that in the highest category, 41.4 percent of the mothers of juvenile offenders and 57.3 percent of the fathers of juvenile offenders were mar- ried at the time of the juvenile’s first offense, (now married), though not necessarily to each other. The next two highest categories for mothers were “Never Married,” 26.7, and “Divorced,” 18.0 percent. (Here again, divorce is not necessarily between the two natural par- ents of the juvenile). For fathers, the next two highest categories were “Divorced,” 21.9 and “Separated,” 11.1 percent. PRESENTATION ANALYSIS Table 3.4.6 Household Status of Parents Number Percent Number Percent Total 806 100.0 723 100.0 In the Household 694 86.1 201 27.8 Not in the Household 112 13.9 522 72.2 Father Household Status Mother Table 3.4.7 Age of Parents Number Percent Number Percent Total 454 100.0 213 100.0 14-19 1 0.2 0 0.0 20-29 2 0.4 0 0.0 30-39 232 51.1 55 25.8 40-49 183 40.3 102 47.9 50 & above 36 7.9 56 26.3 Mother Father Age in Years Table 3.4.8 Marital Status of Parents Number Percent Number Percent Total 589 100.0 361 100.0 Now Married 244 41.4 207 57.3 Widowed 37 6.3 3 0.8 Divorced 106 18.0 79 21.9 Separated 45 7.6 40 11.1 Never Married 157 26.7 32 8.9 Mother Father Marital Status 71 Juvenile’s Guardian—In Table 3.4.9 of the reported cases of juvenile delinquents that had information on the person or persons they were currently living with, 51.8 percent were living with their mother, 4.2 percent with their father, 16.3 percent with mother and father, 12.5 per- cent with their relatives, and 10.0 percent with their mother and stepfather. These data suggest that the majority of juvenile delinquents lived with their mother alone. It should be noted that the category of single female was the highest represented category for guardian in house- holds of the general St Croix population. To get a more in-depth understanding of the sig- nificance of these data, it would have been de- sirable to have data on this same variable for the general St Croix community. Marital Relationship of Natural Parents— Data on the current marital relationship of natu- ral parents highlight categories for the natural parents of juvenile delinquents at time of first offense. "Never Married” and “Divorced” cate- gories represented 40.0 and 21.5 percent re- spectively, representing a combined total of 61.5 percent. The proportion for the category “Married and Living Together” was 20.9 per- cent. The fact that categories like “Both Parents Deceased” accounted for a low rate of juvenile delinquents might be due to an equally small occurrence of such cases in the general St Croix community (Table 3.4.10). Siblings in the Household—Data in Table 3.4.11 indicate that the lower the number of siblings that were in the household, the greater the representation of juvenile delinquents. For example, in households with three or fewer sib- lings, the range of juvenile delinquents was 16.7 to 20.5 percent of the total. In those households with four to six or more siblings, 5.6 and 9.2 percent of the juveniles were repre- sented. This may imply that the more siblings present within families the more opportunities Table 3.4.9 Juvenile's Guardian Guardian Number Percent Total 874 100.0 Mother and Father 142 16.3 Mother 453 51.8 Father 37 4.2 Mother and Stepfather 87 10.0 Father and Stepmother 17 2.0 Stepmother 2 0.2 Stepfather 1 0.1 Foster Parents 8 0.9 Relatives 110 12.5 Other 17 2.0 Table 3.4.10 Marital Relationship of Natural Parents Status Number Percent Total 628 100.0 Married and living together 131 20.9 Separated 49 7.8 Divorced 135 21.5 Never Married 251 40.0 One parent deceased 60 9.5 Both parents deceased 2 0.3 Table 3.4.11 Siblings in the Household Siblings Number Percent Total 606 100.0 No siblings 101 16.7 One sibling 123 20.3 Two siblings 124 20.5 Three siblings 117 19.3 Four siblings 56 9.2 Five siblings 34 5.6 Six or more siblings 51 8.4 PRESENTATION ANALYSIS 72 there are for interaction with large numbers of children and the less likely juveniles are to en- gage in delinquent behavior. School Enrollment—Table 3.4.12 indicates that the majority of juvenile delinquents were enrolled in school. Unfortunately, data indicat- ing actual attendance at school and class par- ticipation were not part of the data set. Al- though 77.6 percent of juveniles involved in delinquent behavior were enrolled in school, that does not mean they attended school, par- ticipated in and benefited from class activities. It is important to note that 22.4 percent were not enrolled in school although the law stipu- lates that children below the age of 16 must be enrolled in school. Attendance at school and class is neither enforced nor easily enforceable. Current or Highest Grade Completed— Table 3.4.13 indicates that juveniles involved in delinquent activities were represented across Grades 1 through 12. However, juvenile of- fenders were only limitedly reprepresented in Grades one through five. There was an in- crease in representation in Grades six to nine with a range between 14.9 and 26.8 percent of the juvenile offenders. The proportion of juve- niles involved in delinquent behavior began to decrease around Grade 10 and continued to do so to the higher grades. It should be remem- bered that these data represented the educa- tional status at time of first offense and cannot predict the educational progress after that of- fense. 3.4.2 St Croix: Department of Human Services Data Relationships Juvenile’s Age by Current or Highest Grade Completed—Table 3.4.14 presents the age of juvenile delinquents together with the highest grade achieved at that time. It could be as- sumed that on average the upper age limit for Grades one through six would be 13 years, though it is not unusual to find children above or below that age. Juvenile delinquents who are in Grades one to six and between ages 14 to 17 amounted to 62.4 percent of the total. In Grade seven, the 15- to 17-year-olds accounted for 56.4 percent, and 51 percent in Grade eight. Grade nine had 20.6 percent of delinquents who were 17 years old. Children develop at different levels, however, but the sense of underachievement can be very severe and lead to activity that could draw at- tention to the low or underachiever. This trend of a juvenile at a certain age being in a lower grade than expected is of concern, especially since the policy of social promotion existed during this time in public schools. PRESENTATION ANALYSIS Table 3.4.12 School Enrollment Enrollment Status Number Percent Total 704 100.0 Enrolled 546 77.6 Not Enrolled 158 22.4 Table 3.4.13 Current or Highest Grade Completed Grade Number Percent Total 586 100.0 Grade 1 2 0.3 Grade 2 1 0.2 Grade 3 4 0.7 Grade 4 2 0.3 Grade 5 9 1.5 Grade 6 87 14.9 Grade 7 157 26.8 Grade 8 106 18.1 Grade 9 102 17.4 Grade 10 53 9.0 Grade 11 42 7.2 Grade 12 21 3.6 73 The reverse—a juvenile in a grade above that expected for the average age or overachiev- ers—was largely absent. For example, there was a lack of juvenile offenders 13 years or younger in Grade nine or higher. Juveniles in Grades 11 and 12, as would be expected, tended to be 16 and 17 years old—80.9 and 95.2percent, respectively. Juvenile’s Employment Status—Table 3.4.15 shows that the majority of juveniles involved in delinquent behavior were unemployed—84.3 percent. This could suggest a confirmation of the old adage of mischief and idle hands. Here again, it could help to have figures showing the proportion of this age group of juveniles who are employed and unemployed in the general population in order to get a clearer picture. Current Job—Table 3.4.16 suggests that at time of first offense the juvenile delinquents were employed in a variety of job types and no particular activity seems to have had a monop- oly on their time. It is assumed that these are part-time jobs but it is not known if these are of weekly or monthly duration. Table 3.4.14 Juvenile's Age by Current or Highest Grade Number Percent Number Percent Number Percent Number Percent Total 860 100.0 104 100.0 156 100.0 105 100 7-9 3 0.3 2 1.9 0 0.0 0 0.0 10-12 31 3.6 21 20.2 5 3.2 1 1.0 13 57 6.6 16 15.4 20 12.8 12 11.4 14 151 17.6 26 25.0 43 27.6 32 30.5 15 202 23.5 18 17.3 41 26.3 29 27.6 16 219 25.5 17 16.3 28 17.9 23 21.9 17 197 22.9 4 3.8 19 12.2 8 7.6 Table 3.4.14 Juvenile's Age by Current or Highest Grade: Continued Number Percent Number Percent Number Percent Number Percent Total 102 100.0 102 100.0 42 100.0 21 100.0 7-9 0 0 0 0.0 0 0.0 0 0.0 10-12 1 1.0 1 1.0 0 0.0 0 0.0 13 0 0.0 0 0.0 0 0.0 0 0.0 14 17 16.7 17 16.7 1 2.4 1 4.8 15 24 23.5 24 23.5 7 16.7 0 0.0 16 39 38.2 39 38.2 10 23.8 5 23.8 17 21 20.6 21 20.6 24 57.1 15 71.4 Grade 8 Grade 9 Grade 10 Grade 11 Grade 12 Total Grade 1-6 Grade 7 Age in years Age in years PRESENTATION ANALYSIS 74 Children in the Household—Table 3.4.17 seems to suggest that households with one, two, three or no children were more likely to have delinquent juveniles—between 14.0 and 20.3 percent—than households with four or more children. It is possible that the presence of more children provide examples of alternate behav- ior, persons with whom youngsters may associ- ate, communicate and so limit negative outside influences on their behavior and activities. Juvenile’s Children—Table 3.4.18 shows that a small proportion of juvenile delinquents—2.6 percent had children of their own. Juvenile Responsible for Childcare—Table 3.4.19 suggests that 4.6 percent of juvenile of- fenders were responsible for childcare. This responsibility did not necessarily refer to the offspring of juvenile offenders, and could have been a younger sibling or relative. Presence of Risk Indicators—Each of the in- dicators in Table 3.4.20 holds some signifi- cance for the juvenile’s psychological well be- ing. The presence of these factors contributes to the undermining of the juvenile’s self esteem, self-worth and a healthy sense of being in con- trol of events affecting one’s life. The influence of any one or more of these vari- ables at any level usually suggests negative connotations for behavior. Juvenile offenders had relatively high percentages in the following areas: trouble controlling anger, 62.9; frequent arguments and fights, 58.8 suspension from PRESENTATION ANALYSIS Table 3.4.16 Current Job Job Number Percent Total 92 100.1 Restaurant/Hotel 4 4.4 Construction 6 6.5 Store Clerk 0 0.0 Mechanic Shop 4 4.4 Gas Station Attendant 4 4.4 Other 74 80.4 Table 3.4.15 Juvenile's Employment Status Employment Status Number Percent Total 630 100.0 Employed 99 15.7 Unemployed 531 84.3 Table 3.4.17 Children in the Household Siblings Number Percent Total 600 100.0 No children 84 14.0 One child 118 19.7 Two children 112 18.7 Three children 122 20.3 Four children 59 9.8 Five children 50 8.3 6 to 10 children 54 9.0 11 to 14 children 0 0.0 15 & over 1 0.2 T a b l e 3 . 4 . 1 8 J u v e n i l e ' s C h i l d r e n N u m b e r P e r c e n t T o t a l 6 0 7 1 0 0 . 0 C h i l d r e n 1 6 2.6 N o C h i l d r e n 5 9 1 9 7 . 4 Table 3.4.19 Juvenile Responsible for Childcare Childcare Number Percent Total 592 100.0 Responsible 27 4.6 Not Responsible 565 95.4 75 school, 66.5; use of drugs or alcohol, 48.5; feelings of failure or worthless, 36.4; getting hurt (physically abused), 28.8. The factor with the lowest representation of juvenile offenders was expulsion from school, 4.9 percent, which may imply that delinquent activities are not consistently curtailed in school or that the rules and regulations are not that severe. Juvenile’s Sexual Activity—Table 3.4.21 indi- cates that more than half—57.9 percent—of the first-time offenders were sexually active. This variable is often identified as an indicator of association with delinquent behavior. Forced Sex—Table 3.4.22 indicates that 9.4 percent of the juveniles with first offenses were forced to have sex. When these data are exam- ined along with the previous table, Juvenile’s Sexual Activity, the data suggest that a large proportion of juveniles opt to participate in sex- ual activity freely. It must be remembered, however, that a juvenile’s perception of free choice may not meet all psychological stan- dards and youngsters may be under subtle pres- sures and constraints that propel them into negative sexual activity. Juvenile’s Sex Partners—In Table 3.4.23 the number and percentage of juveniles for whom data are available here is relatively low, given that 124 offenders had indicated they were sexually active. These data may not be very reliable. Pre-outcome Status—Of the six offenses listed in Table 3.4.24, the number of juveniles involved increasingly decreased in number, ex- cept for Offense 6 with an increase above the offenses immediately preceding. The propor- tion of cases going to trial rather than plea agreement was relatively low for all offenses. As the cases spread over the six offenses, there does not seem to be any definite pattern indi- cating an increase or decrease in the proportion Table 3.4.20 Juvenile's Psychological Assessment Number Percent Number Percent Felt like a failure or worthless 448 163 36.4 285 63.6 Had trouble controlling anger 498 313 62.9 185 37.1 Had frequent arguments & fights 495 291 58.8 204 41.2 Ever been suspended from school 460 306 66.5 154 33.5 Ever been expelled from school 465 23 4.9 442 95.1 Used drugs or alcohol 415 199 48.0 216 52.0 Getting hurt (physically abused) 347 100 28.8 247 71.2 Total Yes No Table 3.4.21 Juvenile's Sexual Activity Sexual Activity Number Percent Total 214 100.0 Sexually Active 124 57.9 Not Sexually Active 90 42.1 Table 3.4.22 Forced Sex Number Percent Total 223 100.0 Forced to have sex 21 9.4 Not forced to have sex 202 90.6 PRESENTATION ANALYSIS 76 of cases that went to trial as against the propor- tion that utilized a plea agreement. Type of Offense—Table 3.4.25 lists the type of offenses committed by juvenile delinquents for a total of six offenses. Felonious Assaults, Other Part I Felony and Other Part II Felony account for the highest number of cases among offenses. Overall, the type of offense or crime varied from one offense to another. Whereas offenses committed by juvenile delinquents in the early offenses seemed to be spread widely, offenders with four or more offenses seemed to have nar- rowed down their scope to activities such as robbery, felonious assault, burglary, possession of a firearm and Other Part II Felony. It is pos- sible that these delinquents are not just experi- menting but have settled into a pattern of more serious activities. Year of Offense—Table 3.4.26 indicates the year that juvenile delinquents committed their first through sixth offenses. Although the time period of the study examined cases current be- tween 1987 and 1997, a number of cases had first offenses occurring before the study’s time frame and the cases remained active during the period covered by the study. Cases were on record from as early as 1980. It should be noted that since juvenile delinquent behavior is recorded for 10 years between ages 7 to 17, the year 1997 did not necessarily indi- cate the end of the case history of all cases cov- ered in the study. Only those delinquents with a first offense in 1987 or before could safely be regarded as having completed their history of delinquent behavior. Outcome of Offense—Table 3.4.27 gives a breakdown of the outcome of cases that went through the YIB and DHS systems for the six offenses being tracked. This was the final proc- PRESENTATION ANALYSIS Table 3.4.23 Juvenile's Sex Partners Sexual Partners Number Percent Total 17 100.0 1 15 88.2 2 1 5.9 3 1 5.9 4 0 0.0 5 0 0.0 6 to 10 0 0.0 10 and over 0 0.0 Table 3.4.24 Pre-Outcome Status Number Percent Number Percent Number Percent Total 507 100.0 110 100 51 100 Trial 33 6.5 6 5.5 5 9.8 Plea Agreement 474 93.5 104 94.5 46 90.2 Table 3.4.24 Pre-Outcome Status: Continued Number Percent Number Percent Number Percent Total 6 100 6 100 21 100 Trial 1 16.7 0 0.0 3 14.3 Plea Agreement 5 83.3 6 100.0 18 85.7 Status Offense 4 Offense 5 Offense 6 Status Offense 1 Offense 2 Offense 3 77 Table 3.4.25 Type of Offense Number Percent Number Percent Number Percent Total 888 100.0 196 100.0 87 100.0 Robbery 71 8.0 16 8.2 10 11.5 Felonious Assaults 159 17.9 30 15.3 11 12.6 Burglary 101 11.4 17 8.7 7 8.0 Grand Larceny 79 8.9 19 9.7 7 8.0 Other Part I felony 126 14.2 32 16.3 13 14.9 Possession of controlled substance 44 5.0 12 6.1 9 10.3 Possession of unlicensed firearm 37 4.2 11 5.6 7 8.0 Possession of stolen property 19 2.1 3 1.5 2 2.3 Unauthorized use of a vehicle 2 0.2 2 1.0 0 0.0 Other Part II felony 45 5.1 24 12.2 9 10.3 Aggravated assault and battery 31 3.5 5 2.6 2 2.3 Petit Larceny 12 1.4 2 1.0 0 0.0 Run away minor 1 0.1 0 0.0 0 0.0 Missing minor 3 0.3 0 0.0 0 0.0 Other 158 17.8 23 11.7 10 11.5 Table 3.4.25 Type of Offense: Continued Number Percent Number Percent Number Percent Total 8 100.0 12 100.0 40 100.0 Robbery 1 12.5 1 8.3 9 22.5 Felonious Assaults 0 0.0 3 25.0 7 17.5 Burglary 1 12.5 0 0.0 2 5.0 Grand Larceny 1 12.5 0 0.0 2 5.0 Other Part I felony 0 0.0 1 8.3 4 10.0 Possession of controlled substance 0 0.0 1 8.3 5 12.5 Possession of unlicensed firearm 1 12.5 0 0.0 2 5.0 Possession of stolen property 0 0.0 0 0.0 0 0.0 Unauthorized use of a vehicle 0 0.0 0 0.0 0 0.0 Other Part II felony 3 37.5 1 8.3 6 15.0 Aggravated assault and battery 0 0.0 0 0.0 1 2.5 Petit Larceny 0 0.0 0 0.0 0 0.0 Run away minor 0 0.0 0 0.0 0 0.0 Missing minor 0 0.0 0 0.0 0 0.0 Other 1 12.5 5 41.7 2 5.0 Offense type Offense type Offense 5 Offense 6 Offense 1 Offense 2 Offense 3 Offense 4 PRESENTATION ANALYSIS 78 Table 3.4.26 Year of Offense Number Percent Number Percent Number Percent Total 882 100.0 193 100.0 85 100.0 1980 1 0.1 0 0.0 0 0.0 1981 0 0.0 0 0.0 0 0.0 1982 1 0.1 0 0.0 0 0.0 1983 1 0.1 0 0.0 0 0.0 1984 1 0.1 0 0.0 0 0.0 1985 1 0.1 0 0.0 0 0.0 1986 3 0.3 1 0.5 1 1.2 1987 79 9.0 7 3.6 1 1.2 1988 94 10.7 22 11.4 11 12.9 1989 59 6.7 23 11.9 6 7.1 1990 78 8.8 11 5.7 7 8.2 1991 66 7.5 28 14.5 8 9.4 1992 81 9.2 20 10.4 10 11.8 1993 51 5.8 9 4.7 5 5.9 1994 91 10.3 15 7.8 7 8.2 1995 104 11.8 24 12.4 11 12.9 1996 102 11.6 14 7.3 13 15.3 1997 67 7.6 12 6.2 2 2.4 1998 2 0.2 7 3.6 3 3.5 Table 3.4.26 Year of Offense: Continued Number Percent Number Percent Number Percent Total 8 100.0 11 100 35 100.0 1980 0 0.0 0 0.0 0 0.0 1981 0 0.0 0 0.0 0 0.0 1982 0 0.0 0 0.0 0 0.0 1983 0 0.0 0 0.0 0 0.0 1984 0 0.0 0 0.0 0 0.0 1985 0 0.0 0 0.0 0 0.0 1986 0 0.0 0 0.0 0 0.0 1987 0 0.0 0 0.0 0 0.0 1988 1 12.5 1 9.1 3 8.6 1989 0 0.0 0 0.0 4 11.4 1990 1 12.5 1 9.1 5 14.3 1991 1 12.5 2 18.2 3 8.6 1992 1 12.5 1 9.1 3 8.6 1993 1 12.5 1 9.1 1 2.9 1994 1 12.5 0 0.0 4 11.4 1995 0 0.0 2 18.2 5 14.3 1996 0 0.0 2 18.2 3 8.6 1997 1 12.5 1 9.1 3 8.6 1998 1 12.5 0 0.0 1 2.9 Offense Year Offense Year Offense 3 Offense 4 Offense 5 Offense 6 Offense 1 Offense 2 PRESENTATION ANALYSIS 79 Table 3.4.27 Outcome of Offense Number Percent Number Percent Number Percent Total 888 100.0 196 100.0 85 100.0 Crisis Stabilization Center 5 0.6 0 0.0 0 0.0 Youth Rehabilitiation Center 43 4.8 37 18.9 20 23.5 Boys/Girls Group Home 3 0.3 0 0.0 1 1.2 Transfer to Correctional Facility 4 0.5 2 1.0 1 1.2 Drug Rehabilitation Center 2 0.2 1 0.5 0 0.0 Psychological Evaluation 11 1.2 3 1.5 1 1.2 Dismissed 187 21.1 47 24.0 19 22.4 Left Jurisdiction 0 0.0 0 0.0 0 0.0 Case not prosecuted 70 7.9 7 3.6 0 0.0 Case pending 0 0.0 0 0.0 2 2.4 Community Service 3 0.3 1 0.5 0 0.0 Counseling 2 0.2 0 0.0 1 1.2 Restitution = $2000 10 1.1 0 0.0 0 0.0 Attend School Regularly 0 0.0 0 0.0 0 0.0 Academic Program 0 0.0 0 0.0 0 0.0 Read/Submit book report 0 0.0 0 0.0 0 0.0 Curfew/prnt/princ 2 0.2 0 0.0 0 0.0 reprimand/prin/j off 0 0.0 0 0.0 0 0.0 probation 328 36.9 69 35.2 27 31.8 Other 218 24.5 29 14.8 13 15.3 Table 3.4.27 Outcome of Offense: Continued Number Percent Number Percent Number Percent Total 9 100.0 12 100.0 38 100.0 Crisis Stabilization Center 0 0 0 0.0 0 0.0 Youth Rehabilitiation Center 2 22.2 5 41.7 13 34.2 Boys/Girls Group Home 0 0.0 0 0.0 0 0.0 Transfer to Correctional Facility 0 0.0 0 0.0 1 2.6 Drug Rehabilitation Center 0 0.0 1 8.3 0 0.0 Psychological Evaluation 0 0.0 0 0.0 0 0.0 Dismissed 1 11.1 2 16.7 12 31.6 Left Jurisdiction 0 0.0 0 0.0 0 0.0 Case not prosecuted 1 11.1 0 0.0 0 0.0 Case pending 0 0.0 1 8.3 2 5.3 Community Service 0 0.0 0 0.0 0 0.0 Counseling 0 0.0 0 0.0 0 0.0 Restitution = $2000 0 0.0 0 0.0 0 0.0 Attend School Regularly 0 0.0 0 0.0 0 0.0 Academic Program 0 0.0 0 0.0 0 0.0 Read/Submit book report 0 0.0 0 0.0 0 0.0 Curfew/prnt/princ 0 0.0 0 0.0 0 0.0 reprimand/prin/j off 0 0.0 0 0.0 0 0.0 probation 3 33.3 3 25.0 5 13.2 Other 2 22.2 0 0.0 5 13.2 Offense 4 Offense 5 Offense 6 Outcome Outcome Offense 1 Offense 2 Offense 3 PRESENTATION ANALYSIS 80 essing of juvenile delinquents and represented the final sentencing stage. Between offenses one and four the most common form of treat- ment is probation ranging between 32.2 to 36.9 percent of cases. For cases with five and six offenses, however, the treatment in the majority of instances changed to sentencing for a period at the Youth Rehabilitation Center (YRC)— between 22.2 to 41.7 percent of cases. It should be noted that for this population of juvenile delinquents for whom the literature indicates there is high marijuana usage, drug rehabilitation is listed as a treatment only in Offenses 1, 2 and 5 and that at an extremely low rate. Biological Father Known—Table 3.4.28 was inserted into the St Croix data gathering form to measure the extent to which juvenile delin- quents did not know the identity of their fa- thers. In discussions among DHS social work- ers, it was their perception that this factor would rate high. However, these data indicate that an overwhelming majority of juvenile de- linquents did have knowledge of their fathers. Perhaps closer examination should be given to the level and types of bonds and relationships existing between father and juvenile delin- quents, as a more important determining factor of delinquent behavior. 3.4.3 Predicting Recidivism with the St Croix Department of Human Services Data In this analysis, five predictor variables that showed statistically significant association with the outcome variable, recidivism were included in the model. The estimation of the predictor variables was severely hampered by the wide- spread occurrences of missing data values. While there were 888 cases in the original data set of St Croix’s Department of Human Ser- vices’ data, only 361 valid observations could be used in deriving the estimates. This indi- cates that only 361 cases had non-missing val- ues for each of the five variables that were in- cluded in the model. The five predictor vari- ables were Age, Sex, Status of Natural Parents, Feeling of Failure, and Use of Drugs. 3.4.3.1 Estimates of the St Croix Department of Human Services Data The likelihood ratio chi-square in Table 3.4.3.1 suggests that at least one of the variables in the model is significant. The table indicates that while four of the variables were statistically significant—Sex, Feelings of Failure and Use of Drugs—two of them failed to reach the tra- ditional level of statistical significance of 0.05. However, because the variable Age has shown itself to be such a reliably strong predictor in all other models included in this study, it was re- tained even though its significance level of 0.0612 is just slightly beyond the usual 0.05 level. When the variables Age-Squared and Age were included in the model, both showed no statistical significance.4 Even more surpris- ing is the failure of the Status of Natural Par- ents variable to reach the significance level of 0.05. Even though a 2 x 2 table cross-classified this variable with recidivism and showed them PRESENTATION ANALYSIS 4A second possible explanation of the relatively poor predictive strength of the Age variable is that the data reflect some degree of contamination. The data analysts were unable to verify the data as they had no direct access to the juveniles’ case files. Table 3.4.28 Biological Father Known Knowledge of Father Number Percent Total 652 100.0 Known 609 93.4 Unknown 43 6.6 81 to be significantly related, this variable was barely below the 0.2 level of significance— 0.1811. Because of the strong association shown in the 2 x 2 case, this variable was also retained. It would appear that the high attrition of cases mentioned above is probably the pri- mary reason for the relatively poor showing of these two variables in terms of statistical sig- nificance. The adjusted odds ratio for age of 0.857 may be interpreted as follows. For each one-year in- crease in age, the odds of a repeat offense are decreased by a factor of 0.857 or eventuates in a 100 x (0.857 – 1) = -14.3 percent change, i.e., 14.3 percent reduction in the odds of the juve- nile becoming a repeat offender. The odds ra- tio for Sex of 2.481 again indicates that the odds of males being repeat offenders are about 2.5 times higher than the odds for females. First-time offenders who live with their natural parents in a married state and living together have odds that are reduced by 37.2 percent of those juveniles who live with a parent or par- ents in some other kind of relationship. This may be explained in other words by saying that first-time offenders who do not live with their natural parents in a married union have odds of being recidivists by a factor that is—1/0.628 = 1.591—about 1.6 times higher than the odds for offenders who live with their natural parents in a married relationship. Juvenile offenders who indicated they harbored feelings of being a failure or of being worthless had predicted odds of being repeat offenders of 1.799. In other words, the odds of recidivism for a first-time offender who experienced these feelings of failure and worthlessness are about 80 percent higher than the odds for other juve- nile offenders who do not have such feelings. The final variable is one that derives from re- sponses to the question concerning the use of drugs or alcohol before or during school. The adjusted odds ratio shown in table 3.4.3.1 of 2.851 is the highest among all the variables. Hence, the odds of recidivism for juveniles who indicated they used alcohol or drugs be- fore or during school was nearly three times as Table 3.4.3.1 Parameter Estimates in Logistic Regression Predicting Recidivism: St Croix, Department of Human Services Data Variable Parameter Estimate (1) Standard Error (2) Pr> Chi Square (3) Adjusted Odds Ratio (4) Intercept 1.3319 1.2519 0.2874 … Age -0.1545 0.0825 0.0612 0.857 Sex (1=Male) 0.4543 0.2086 0.0295 2.481 Parents (1=Married and Living Together) -0.2322 0.1736 0.1811 0.628 Feelings of Failure (1=Yes) 0.2935 0.1267 0.0205 1.799 Drug Use (1=Yes) 0.5238 0.1223 <0.0001 2.851 Likelihood Ratio Chi Square = 48.8380 df = 5 Pr > Chi Square < 0.0001 Notes: N=361. Odds ratios greater than 1.0 indicate greater odds of recidivism. PRESENTATION ANALYSIS 82 Table 3.4.3.2 Predicted Probabilities in Logistic Regression: St Croix, Department of Human Services Data Proba- Coefficients: a Agefo Sex Parents Failure Drugs Odds bility (1) (2) (3) (4) (5) (6) (7) (8) Parameter Estimates: 1.3319 -0.1545 0.4543 -0.2322 0.2935 0.5238 Juvenile Profiles: D1 8 1 0 1 1 3.926 0.797 D2 9 1 0 1 1 3.364 0.771 D3 10 1 0 1 1 2.882 0.742 D4 11 1 0 1 1 2.469 0.712 D5 12 1 0 1 1 2.116 0.679 D6 13 1 0 1 1 1.813 0.645 D7 14 1 0 1 1 1.553 0.608 D8 15 1 0 1 1 1.331 0.571 D9 16 1 0 1 1 1.141 0.533 D10 17 1 0 1 1 0.977 0.494 D11 15 1 1 1 1 1.055 0.513 D12 10 0 1 0 1 1.082 0.520 D13 12 0 0 1 1 1.343 0.573 D14 12 0 1 1 1 1.065 0.516 D15 12 1 0 1 0 1.253 0.556 D16 12 1 0 0 0 0.934 0.483 D17 15 1 1 1 0 0.625 0.385 D18 15 1 0 0 1 0.993 0.498 D19 15 1 1 0 1 0.787 0.440 D20 15 1 0 1 0 0.788 0.441 D21 15 0 0 0 1 0.630 0.387 D22 10 1 0 1 0 1.707 0.631 Variables: Agefo = Age of juvenile at first offense. Sex = Sex of the juvenile (1 = male, 0 = female). Parents = Natural parents of juvenile (1 = married and living together, otherwise 0). Failure = Feeling like a failure or worthless (1 = failure, otherwise 0). Drugs = Use of drugs or alcohol before or during school (1 = use, otherwise 0). PRESENTATION ANALYSIS 83 high as the odds for those juveniles who indi- cated they did not. Or it might be said that this latter group has odds of being recidivists of only about 35 percent of the odds of those us- ing drugs or alcohol in school. 3.4.3.2 The Prediction of Recidivism Table 3.4.3.2 shows the predicted chances of becoming a recidivist for a combination of stated levels of the variables in the model. The first 10 profiles are those of juveniles for each successive age from 8 to 17 with the other vari- ables in the model—Sex, Status of Natural Par- ents, Feeling of Failure, and Use of Drugs— remaining constant throughout the age range (See Figure 3.4.3.1). It is noted that for these predictions shown in column (8), there is a gradual decrease from a high of 0.797 to a low of 0.494. The first profile may be interpreted to say that an 8-year-old first-time male offender who is living with parents that are not married and living together, who thinks of himself as a failure or worthless and who used drugs or al- cohol before or during school is predicted to be an offender who will be arrested at least one more time before he reaches the age of 18. The data show that this combination of characteris- tics predicts recidivism for all ages up to 16 years. It is not until the juvenile reaches 17 years in age that his prediction falls below 0.50, therefore indicating that he will likely not com- mit another offense before he is 18 years old. Profile D11 suggests that a 15-year-old-male living with his natural parents who are married but who has low self esteem, and has used drugs or alcohol at school is still predicted to be a repeat offender before age 18. Profiles D1 to D22 in the table present other combinations of characteristics and their predictions of repeat or non-repeat of offenses that result in arrests. D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Juvenile Profile 0.00 0.15 0.30 0.45 0.60 0.75 Probability or Chances Figure 3.4.3.1 Predicted Chances of Recidivism: St Croix DHS For a Male Using Drugs, With Feelings of Failure and not Living With Married Parents Note: See text for profile characteristics PRESENTATION ANALYSIS 84 3.5 Summary and Overview for the Territory While looking at data results for both JIB/YIB and DHS it should be borne in mind that each item did not have the same rate of response. Response rate depended on whether or not a piece of data was gathered at the time a juve- nile was processed into the system. As a result, with each new variable, the total tended to change. 3.5.1 Juvenile Units Background Characteristics The rate of involvement of juveniles engaged in offensive behavior differed based on age, gen- der, racial composition, time and type of of- fense. A relatively small percentage of total first-time offenders went on to become recidi- vists or repeat offenders (17.4 St Thomas, 10.0 St Croix) and an extremely small percent went on to commit multiple crimes, leading to con- finement at the Youth Rehabilitation Center. On both islands there was very little involve- ment in delinquent behavior up to the 7- to 10 age bracket but this involvement climbed with increasing age to peak around age 15. A com- parison of gender suggested a much higher male than female participation in delinquent activity and over longer periods of time. Blacks were over-represented in juvenile offen- sive activity, while whites and other races were underrepresented. However, the overrepresen- tation for blacks was more acute on the island of St Thomas. This is more likely, reflecting social and income differences in the popula- tions than a racial factor. The age at first offense, when looked at by eth- nic groups showed, in general, that there were consistent and similar patterns for the juvenile delinquents of a particular ethnic group as well as for juvenile delinquents of mothers and fa- thers from these ethnic groups. For both is- lands there was a low rate of juvenile delin- quency below age 10, showing a gradual in- crease with age, a peaking among the 11- to 14- year-olds, and a falling off at age 16 plus. The only ethnic group that seemed to have shown a difference in this trend was Puerto Rico, where a relatively higher rate of activity occurred around ages 10 to 13 than for the other ethnic groups. This tapered off at an age earlier than that for other ethnic groups as well. Although the percentage of juvenile delin- quents who were not enrolled in school was low on both islands, the percentage for St Thomas more than doubled that of St Croix. However, the law requires that children of school age, 16 years and under, be enrolled in school, and enrollment did not seem to play a significant part in juvenile behavior. Problems were likely to arise when children failed to at- tend school where they could participate mean- ingfully in class activities. It should be recalled that indications are that most delinquent behav- ior occur after school hours when more chil- dren tend to be unsupervised. Greater focus could be placed on gathering data that represent that period of juvenile life. The general pattern of low delinquency rates in primary school, with an obvious increase with age and grade into middle and high school, held true for both St Croix and St Thomas. Within that pattern another interesting pattern showed St Thomas with approximately twice the rate of juvenile delinquents in primary school as did St Croix. Middle and high school juvenile de- linquent rates on St Thomas were about the same, about two fifths of the total. For St Croix, middle school was a little over one- fifth, while high school had the majority of first-time juvenile delinquents—over 70 per- cent. The reason for these marked differences across the two islands is not clear. PRESENTATION ANALYSIS 85 Family and Community Characteristics The majority of juvenile delinquents lived with mother alone, though the proportions differed significantly when looked at between islands. Over two-thirds on St Croix and a little more than half on St Thomas lived with mother alone. However, a higher proportion of juvenile delinquents on St Thomas, about one-fifth, lived with both father and mother than did on St Croix. Other guardian types were minimally represented, corresponding, it is believed, to their marginal presence in the communities. This pattern remained true across gender. In St Thomas, juveniles seemed to have been overrepresented for persons born in the Eastern Caribbean and underrepresented for those born in the US and US Virgin Islands. The reverse occurred for St Croix where juvenile delin- quents were over-represented for persons born in the US Virgin Islands and underrepresented for the Eastern Caribbean. Place of birth seemed to have an influence on delinquent behavior from a number of perspec- tives. On both islands, the majority of juvenile delinquents came from mothers and fathers born in the US Virgin Islands. However, for those with parents born in the Eastern Carib- bean and other Caribbean islands there was a disproportionate number of juvenile offenders relative to their size in the community. On both islands, the majority of mothers and fathers were employed. However, the unem- ployment rate for mothers of juveniles on St Thomas was almost five times more than that for mothers of St Croix’s juvenile delin- quents. On both islands, the rate was slightly higher for the mothers of male juvenile offend- ers and slightly lower for female offenders. On St Croix, the unemployment rate for fathers of juvenile delinquents was similar to that for mothers. Offenses Committed The majority of juvenile offenders in the study on both islands showed no evidence of recidi- vism; that is, the juveniles who committed one offense were processed through the system but were never arrested for further offenses during the period of the study. The recidivism rate for St Croix was about six percent lower than that for St Thomas. It should be remembered, how- ever, that the period that the study spanned al- lowed for juveniles not yet 17 years of age to commit offenses outside the terminating date of the study so that such offenses would have been outside the scope of this study. St Thomas recorded over two-thirds of the male juvenile delinquents who had committed of- fenses having used no weapon; this rate for St Croix is less than two percent. Use of weap- ons, including hand, bottle, knife, gun and pe- nis that suggested involvement in aggressive behavior accounted for approximately 15 per- cent in St Thomas; while on St Croix, this cate- gory represented over 60 percent of juvenile delinquents. 3.5.2 Department of Human Services Background Characteristics On both islands the proportion of male juvenile delinquents—approximately 85 to 86 percent— exceeded that for females—approximately 14 to 16 percent. This was dramatically dispropor- tionate to the actual population proportions in the wider community where approximate male/female ratios tend to be closer to 50/50 one way or the other. This trend is common in most communities for delinquent behavior gen- erally, and for juvenile delinquent behavior in particular. In regard to age, the data showed that juvenile delinquents displayed the same pattern of be- PRESENTATION ANALYSIS 86 havior on both islands. There was low involve- ment between ages seven to 12 with a slight increase around age 13, and a dramatic increase between the ages of 14 and 17. The years of highest activity differed with St Croix, peaking at ages 15 and 16 and St Thomas at 14 to 15. On both islands, the black race was over repre- sented in juvenile delinquent behavior but at different rates. St Croix’s rate of over- representation was 27 and St Thomas 7 percent. The white race, on the other hand, had the same rate of delinquent behavior on both islands and was underrepresented 4 to 5 times. This result is likely affected by broader social and eco- nomic influences. There was a higher rate of juvenile offenders who were citizens involved in juvenile activity on the island of St Croix than on St Thomas. The Resident Alien category on St Thomas had a juvenile delinquent rate more than twice that for St Croix. In all cases, except resident aliens on St Croix, delinquent rates were higher than that population group’s representation in the community. For both islands, the vast majority of juvenile delinquents were enrolled in school. St Thomas had approximately 7 percent higher enrollment than did St Croix. It is obvious that enrollment in school is not a deterrent factor in juvenile’s involvement in delinquent behavior. There were no data available on attendance, class par- ticipation or involvement in extra curricula ac- tivities. On both St Thomas and St Croix there were very few juvenile offenders between Grades one through five, except a surprisingly high incidence in Grade two for St Thomas. Then there was a dramatic increase over Grades six through nine. Although there was still relatively high activity by Grade 10, the rate had begun to fall and continued to do so up to Grade 12. Employment rates for juvenile delinquents were low for both islands with St Croix having a little over two percent higher than St Thomas. There was no specific reference regarding the location of employment activities and whether monetary compensation had to be a necessary part of the definition of employment. The ab- sence of data for the juvenile population in the broader communities made it difficult to deter- mine whether these rates are representative. On the surface however, it can be said that a large proportion of juvenile delinquents had a lot of unproductive time on their hands that could have contributed to involvement in criminal activities. There were no particular activities that stood out on either island as a means of employment. Since the unemployment rate was so high, those employed represented a small proportion of the juvenile delinquent population. The juvenile delinquent rate was more than 10 percent higher on St Croix than St Thomas for persons of US Virgin Islands birth. For those born in the Eastern Caribbean, the rate for St Thomas more than doubled that for St Croix. On both islands, the rates were close for those who were both US mainland- and Puerto Rico- born. Family and Community Characteristics The rate at which mothers were present in the household was similar for both islands, repre- senting the vast majority. The presence of fa- thers, though slightly higher for St Thomas, revealed rates that were relatively low on both islands. Mothers of juvenile delinquents in the 40 to 49 age group accounted for the largest proportion on St Thomas. On St Croix, the highest rate fell in the 30 to 39 age group. On both islands fa- thers appeared relatively older than mothers, falling between the 40 to 49 age group on PRESENTATION ANALYSIS 87 St Croix and the over 50 on St Thomas. On both islands the “Now Married” category for mothers of juvenile offenders was over 40 percent compared to fathers where more than half were in the “Now Married” category. The next highest categories were “Never Married” and “Divorced” on St Thomas, making up 40.7 and 47.0 percent for mothers and fathers re- spectively. On St Croix these categories to- gether made up 44.7 percent for mothers, whereas, the “Never Married” and “Separated” for fathers totaled 43 percent. Approximately one-half of the natural parents of juvenile delinquents on St Thomas were “Divorced” or “Never Married,” whereas on St Croix those categories together were over 10 percent higher. Those natural parents who were “Married” and “Living Together” made up ap- proximately 29 percent in St Thomas and 21 percent in St Croix. Almost half of the juvenile offenders in St Thomas and a little over half in St Croix were currently living with mothers, compared to a little over 4 percent living with fathers on both islands. Those living with both mother and father made up a little over one-fifth for St Thomas and somewhat less for St Croix. According to data for both islands, the more siblings in a household, the lower the represen- tation of juvenile delinquents. In households with four to six or more siblings present the rate of representation of juvenile offenders ranged between 4 to 10 percent. Households with between zero to three siblings accounted for approximately 14 to 27 percent of the juve- nile delinquents. As in the case of the “siblings-in-household” variable there seemed to have been a clear trend with children in household for both is- lands. The smaller the number of children in the household, the higher the representation of juvenile delinquents, and the higher the number of children in the household, the lower the rep- resentation of juvenile delinquents. It is likely that this factor does not operate alone but works in conjunction with other factors. The data appear to suggest that the more individuals available for communication and interaction at home, the less a juvenile is likely to go outside seeking dubious adventure. A question concerning knowledge of fathers by juvenile delinquents was posited only on the island of St Croix. The results showed that the vast majority of juvenile delinquents knew who were their biological fathers. The problem sited by the DHS probably centers around the type of bond or relationship that exists between the ju- venile delinquents and their fathers, despite that knowledge. Data were not available to measure this factor. Presence of Risk Factors On St Croix 2.6 percent of juvenile offenders, and 4.4 percent of offenders on St Thomas were reported as having children of their own. Although the rate on St Thomas was almost twice that of St Croix, both proportions were relatively low. At the same time, less than 5.0 percent of re- ported juvenile delinquent cases on both islands were responsible for child care. This childcare responsibility was not necessarily for their own offspring. It could have been of siblings or rela- tives. In all risk indicators, St Croix showed a higher rate of occurrence than did St Thomas except in the “Expelled From School” group where St Thomas had a rate almost 20 percent more than that for St Croix On both St Croix and St Thomas the rate of ju- veniles involved in sexual activity was a little over half the reported cases. PRESENTATION ANALYSIS 88 St Croix had approximately twice the rate of juvenile delinquents reporting to be sexually active as did St Thomas. Of the juvenile delinquents for whom data was available, St Croix had 88.2 percent with one sex partner and 5.9 percent with two and three partners. This compared with 60 percent with one sex partner and 10 percent with six or more sex partners on St Thomas. Outcome of Offenses Both islands seemed to have had a low trial and a high plea-bargain rate for all cases. There were several outcomes for each offense and outcomes varied with offense on both islands. On St Croix the most common outcome for Of- fenses 1 through 4 was Probation followed by Dismissal. This changed to sentencing to the Youth Rehabilitation Center as the first out- come for Offenses 5 and 6 and Dismissal or Probation remained the second highest out- comes. On St Thomas Offenses 1 and 2 had Commu- nity Service as the most common outcome fol- lowed by Dismissal or sentencing to Youth Re- habilitation Center. Offenses 3 to 6 had Youth Rehabilitation Center as the number one out- come followed by Community Service. PRESENTATION ANALYSIS 89 Any substantive methods that are aimed at pre- venting or treating antisocial behavior should be based on empirically validated theories about causes. The major emphasis should be on the early prevention of offending. Serious antisocial behavior might be viewed as a chronic disease that requires continuous moni- toring and intervention over the life course. The following recommendations are aimed at prevention, or reduction, of offensive behaviors in juveniles. First, delinquency prevention may begin with pregnancy and infancy. Programs may be insti- tuted to alleviate problems in pregnancy and infancy by encouraging visits to prenatal clinics designed to help mothers during pregnancy and during the first two years of life. Results else- where have shown that home visits during pregnancy led to teenage mothers having heav- ier babies, and women who had previously smoked decreased their smoking and had fewer pre-term deliveries. The postnatal visits to clinics caused a decreased in recorded physical abuse and neglect during the first two years in life. This latter result is important because of the common observation that being physically abused or neglected as a child predicts later violent offending. Second, hyperactivity and impulsivity might be altered using techniques termed cognitive- behavioral interpersonal skills training. Delin- quents can be taught the cognitive skills in which they are deficient, and this could lead to a decrease in their offending. This kind of pro- gram should aim to modify the impulsive, ego- centric thinking of delinquents, to teach them to stop and think before acting, to consider the consequences of their behavior, to conceptual- ize alternative ways of solving interpersonal problems, and to consider the impact of their behavior on other people, especially their vic- tims. This includes social skills training and social perspective training in recognizing and understanding other people’s feelings. Third, if low intelligence and school failure are causes of offending, then any program that leads to an increase in school success should lead to a decrease in offending. Fourth, if poor parental supervision and erratic child-rearing behavior are causes of delin- quency, it seems likely that parent training might succeed in reducing offending. Parents of antisocial children tend to be deficient in their methods of child rearing, often using more punishment but failing to make it contingent on the child’s behavior. Fifth, if socioeconomic deprivation causes of- fending, then providing increased economic resources for the more deprived families should lead to a decrease in offending by their chil- dren. Sixth, if having delinquent friends causes of- fending, then any program that reduces their influence or increases the influence of pro- social friends could have a reductive effect on offending. Seventh, if one accepts that today’s children are tomorrow’s parents, voters, workers and com- munity leaders, then there ought to be a sus- tained commitment to a set of coherent family policies that are relevant to today’s families and that are dedicated to the premise that families should be supported in meeting the needs of IV. IMPLICATIONS OF THE FINDINGS AND RECOMMENDATIONS 90 their children. To this end, it is proposed that a set of children’s indicators—like percent of children in living arrangements, percent of chil- dren in poverty, infant health risk indicators, age of teen births, child abuse and neglect, and juvenile crime rates—require statistical data collection on a regular basis. Finally, it can be said that any measure that re- duces crime in USVI will probably also reduce alcohol abuse, drunk driving, drug abuse, sex- ual promiscuity, family violence, truancy, school failure, unemployment, marital dishar- mony, and divorce. It is clear that problem children tend to grow up into problem adults, and that problem adults tend to produce more problem children. Are these then not enough solid reasons to make a sustained commitment to return these islands to their former status of being nearly crime-free? 91 REFERENCES Bartolas, C. (1997). Juvenile Delinquency, (Fourth Ed.). MA: Allyn and Bacon. Bureau of the Census, Detailed Population Characteristics: Virgin Islands of the United States (PC80-1-D55). Bureau of the Census, Detailed Cross Tabulation for the U.S. Virgin Islands (CPH-L-156). Community Foundation of the Virgin Islands. (2002). US Virgin Islands Kids Count Data Book 2002. St Thomas, Virgin Islands: Author. Department of Human Services. (May 2000). Youth Rehabilitation Center Facilities and Problems Paper Presented at a Senate hearing. Charlotte Amalie, St. Thomas. Eastern Caribbean Center, University of the Virgin Islands. (August 1997). 1995 Population and Housing Survey. Eastern Caribbean Center, University of the Virgin Islands. (August 2000). 1997 Consumer Ex- penditure Survey: Population and Housing. Farrington, D. D. (1996). The Explanation and Prevention of Youthful offending. In J.D. Hawkins, (Ed.), Delinquency and Crime: Current Theories, (pp. 68-148). New York, NY: Cambridge University Press. Loeber, Rolf. (1996). Development Continuity, Change, and Pathways in Male Juvenile Problem Behaviors and Delinquency. In J.D. Hawkins, (Ed.). Delinquency and Crime: Current Theories, (pp. 1 -27). New York NY: Cambridge University Press. Menard, S. (1995). Applied Logistic Regression Analysis. Sage University Paper Series on Quantitative Applications in the Social Sciences, (pp. 07-106). Thousand Oaks, CA: Sage. 92 APPENDIX Table A1 Parent's Place of Birth, St Thomas Mother Father 1995 1995 Place of Birth Number Percent Number Percent MPOB Percent FPOB Percent Total 2,044 100.0 1,804 100.0 100.0 100.0 US Virgin Islands 688 33.7 478 26.5 25.8 27.0 Puerto Rico 52 2.5 45 2.5 2.6 2.0 United States 105 5.1 110 6.1 10.8 10.5 Antigua or Barbuda 202 9.9 198 11.0 7.0 8.6 Dominica 168 8.2 164 9.1 7.5 7.4 St Kitts or Nevis 357 17.5 341 18.9 14.6 13.7 Other Caribbean 443 21.7 439 24.3 27.6 26.4 Other 29 1.4 29 1.6 4 4.3 Table A2 Type of Offense, St Thomas Offense 1 Offense 2 Offense 3 Type Number Percent Number Percent Number Percent Total 2,335 100.0 836 100.0 433 100.0 Part I Felony 509 21.8 236 28.2 132 30.5 Part II Felony 268 11.5 112 13.4 57 13.2 Misdemeanor 309 13.2 110 13.2 41 9.5 Incidences 548 23.5 179 21.4 98 22.6 Other 701 30.0 199 23.8 105 24.2 Table A2 Type of Offense, St Thomas: Continued Offense 4 Offense 5 Offense 6 Type Number Percent Number Percent Number Percent Total 117 100.0 173 100.0 266 100.0 Part I Felony 28 23.9 47 27.2 96 36.1 Part II Felony 13 11.1 28 16.2 32 12.0 Misdemeanor 17 14.5 21 12.1 33 12.4 Incidences 30 25.6 38 22.0 40 15.0 Other 29 24.8 39 22.5 65 24.4 93 Table A3 Type of Weapon, St Thomas Offense 1 Offense 2 Offense 3 Type Number Percent Number Percent Number Percent Total 2,332 100.0 832 100.0 434 100.0 Gun 64 2.7 45 5.4 27 6.2 Knife 90 3.9 45 5.4 20 4.6 Hand 149 6.4 48 5.8 21 4.8 Bottle 23 1.0 15 1.8 5 1.2 Penis 11 0.5 4 0.5 2 0.5 None 1,828 78.4 597 71.8 325 74.9 Other 167 7.2 78 9.4 34 7.8 Table A3 Type of Weapon, St Thomas: Continued Offense 4 Offense 5 Offense 6 Type Number Percent Number Percent Number Percent Total 116 100.0 174 100.0 269 100.0 Gun 3 2.6 14 8.0 23 8.6 Knife 0 0.0 6 3.4 12 4.5 Hand 7 6.0 8 4.6 16 5.9 Bottle 2 1.7 2 1.1 5 1.9 Penis 0 0.0 0 0.0 0 0.0 None 99 85.3 131 75.3 191 71.0 Other 5 4.3 13 7.5 22 8.2 94 Table A4 Year of Offense, St Thomas Offense 1 Offense 2 Offense 3 Offense Year Number Percent Number Percent Number Percent Total 2342 100.0 838 100.0 437 100.0 1980 1 - 1 0.1 1 0.2 1981 8 0.3 3 0.4 1 0.2 1982 13 0.6 0 0.0 1 0.2 1983 9 0.4 6 0.7 3 0.7 1984 15 0.6 7 0.8 3 0.7 1985 9 0.4 8 1.0 3 0.7 1986 22 0.9 12 1.4 8 1.8 1987 180 7.7 54 6.4 28 6.4 1988 219 9.4 80 9.5 32 7.3 1989 259 11.1 87 10.4 51 11.7 1990 220 9.4 68 8.1 43 9.8 1991 296 12.6 103 12.3 63 14.4 1992 243 10.4 98 11.7 49 11.2 1993 178 7.6 81 9.7 37 8.5 1994 204 8.7 73 8.7 44 10.1 1995 135 5.8 44 5.3 19 4.3 1996 104 4.4 42 5.0 16 3.7 1997 146 6.2 45 5.4 23 5.3 1998 81 3.5 26 3.1 12 2.7 95 Table A4 Year of Offense, St Thomas: Continued Offense 4 Offense 5 Offense 6 Offense Year Number Percent Number Percent Number Percent Total 119 100.0 176 100.0 272 100.0 1980 0 0.0 0 0.0 0 0.0 1981 0 0.0 0 0.0 0 0.0 1982 0 0.0 0 0.0 0 0.0 1983 0 0.0 0 0.0 0 0.0 1984 0 0.0 0 0.0 0 0.0 1985 0 0.0 0 0.0 0 0.0 1986 6 5.0 4 2.3 0 0.0 1987 11 9.2 14 8.0 14 5.1 1988 13 10.9 16 9.1 18 6.6 1989 17 14.3 24 13.6 30 11.0 1990 7 5.9 11 6.3 23 8.5 1991 22 18.5 34 19.3 34 12.5 1992 12 10.1 19 10.8 43 15.8 1993 10 8.4 13 7.4 30 11.0 1994 9 7.6 13 7.4 16 5.9 1995 5 4.2 12 6.8 21 7.7 1996 3 2.5 5 2.8 16 5.9 1997 3 2.5 9 5.1 18 6.6 1998 1 0.8 2 1.1 9 3.3 96 Table A5 Outcome of Offense, St Croix Offense 1 Offense 2 Offense 3 Outcome Number Percent Number Percent Number Percent Total 888 100.0 196 100.0 85 100.0 Crisis Stabilization Center 5 0.6 0 0.0 0 0.0 Youth Rehabilitation Center 43 4.8 37 18.9 20 23.5 Boys/Girls Group Home 3 0.3 0 0.0 1 1.2 Transfer to Correctional Facility 4 0.5 2 1.0 1 1.2 Drug Rehabilitation Center 2 0.2 1 0.5 0 0.0 Psychological Evaluation 11 1.2 3 1.5 1 1.2 Dismissed 187 21.1 47 24.0 19 22.4 Left Jurisdiction 0 0.0 0 0.0 0 0.0 Case not prosecuted 70 7.9 7 3.6 0 0.0 Case pending 0 0.0 0 0.0 2 2.4 Community Service 3 0.3 1 0.5 0 0.0 Counseling 2 0.2 0 0.0 1 1.2 Restitution = $2000 10 1.1 0 0.0 0 0.0 Attend School Regularly 0 0.0 0 0.0 0 0.0 Academic Program 0 0.0 0 0.0 0 0.0 Read/Submit book report 0 0.0 0 0.0 0 0.0 Curfew/prnt/princ 2 0.2 0 0.0 0 0.0 reprimand/prin/j off 0 0.0 0 0.0 0 0.0 Probation 328 36.9 69 35.2 27 31.8 Other 218 24.5 29 14.8 13 15.3 97 Table A5 Outcome of Offense, St Croix: Continued Offense 4 Offense 5 Offense 6 Outcome Number Percent Number Percent Number Percent Total 9 100.0 12 100.0 38 100.0 Crisis Stabilization Center 0 0 0 0.0 0 0.0 Youth Rehabilitation Center 2 22.2 5 41.7 13 34.2 Boys/Girls Group Home 0 0.0 0 0.0 0 0.0 Transfer to Correctional Facility 0 0.0 0 0.0 1 2.6 Drug Rehabilitation Center 0 0.0 1 8.3 0 0.0 Psychological Evaluation 0 0.0 0 0.0 0 0.0 Dismissed 1 11.1 2 16.7 12 31.6 Left Jurisdiction 0 0.0 0 0.0 0 0.0 Case not prosecuted 1 11.1 0 0.0 0 0.0 Case pending 0 0.0 1 8.3 2 5.3 Community Service 0 0.0 0 0.0 0 0.0 Counseling 0 0.0 0 0.0 0 0.0 Restitution = $2000 0 0.0 0 0.0 0 0.0 Attend School Regularly 0 0.0 0 0.0 0 0.0 Academic Program 0 0.0 0 0.0 0 0.0 Read/Submit book report 0 0.0 0 0.0 0 0.0 Curfew/prnt/princ 0 0.0 0 0.0 0 0.0 reprimand/prin/j off 0 0.0 0 0.0 0 0.0 Probation 3 33.3 3 25.0 5 13.2 Other 2 22.2 0 0.0 5 13.2 98