<%BANNER%>

Studying Patterns and Correlates of Recidivism


PAGE 1

STUDYING PATTERNS AND CO RRELATES OF RECIDIVISM ACROSS RACE By ROHALD ARDWAN MENESES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Rohald Ardwan Meneses

PAGE 3

ACKNOWLEDGMENTS I would like to thank my committee members, Dr. Alex Piquero and Dr. Marian Borg, for all their time and effort in helping me to accomplish this project. Special thanks go to the chair of my committee, Dr. Alex Piquero, for all his guidance and help throughout my thesis. Also, I thank my parents Maria and Johnny, and my brothers Johann and Eelhard. I would also like to thank the following for all their love and support: Dr. Lonn Lanza-Kaduce, Vicente Cuka, Jessica Malmad, Ivo Cuka, Miguel Molina, Fernando Molina, Romeo Meneses, Salvatore Stelluto, Christina Anton, Wendy Colon, Christian Monje, Carola Sabja, Roxana Otasevic, Maria Jose Guzman and Jeimmy Arredondo. iii

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi ABSTRACT......................................................................................................................vii CHAPTER 1 INTRODUCTION........................................................................................................1 2 REVIEW OF LITERATURE.......................................................................................5 Juvenile Delinquency....................................................................................................5 Recidivism....................................................................................................................6 Theoretical Analysis.....................................................................................................8 Analysis of the Factors that are Related to Recidivism among Juvenile Offenders...12 Cognitive Ability........................................................................................................13 Age..............................................................................................................................14 Seriousness of Offense...............................................................................................15 Alcohol and Drug Abuse............................................................................................16 Family Background....................................................................................................17 Race............................................................................................................................18 Current Focus..............................................................................................................18 3 DATA.........................................................................................................................21 Variables.....................................................................................................................23 Independent Variables.........................................................................................23 Dependent Variables...........................................................................................24 Research Questions.....................................................................................................24 Analytic Plan..............................................................................................................24 4 FINDINGS..................................................................................................................26 Results for One-Way ANOVA...................................................................................26 Results of the Survival Analysis.................................................................................27 iv

PAGE 5

5 DISCUSSION AND CONCLUSIONS......................................................................32 LIST OF REFERENCES...................................................................................................35 BIOGRAPHICAL SKETCH.............................................................................................41 v

PAGE 6

LIST OF TABLES Table page 1 Descriptive Statistics: One-Way ANOVA analysis.................................................27 2 Cox Proportional Hazard Models Predicting Hispanic Juveniles Parole Success/Failure.........................................................................................................29 3 Cox Proportional Hazard Models Predicting African American Juveniles Parole Success/Failure.........................................................................................................30 4 Cox Proportional Hazard Models Predicting White Americans Juveniles Parole Success/Failure.........................................................................................................31 5 Cox Regression Hazard Model: Split-Race Analysis..............................................31 vi

PAGE 7

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts STUDYING PATTERNS AND CORRELATES OF RECIDIVISM ACROSS RACE By Rohald Ardwan Meneses May, 2003 Chair: Alex Piquero Cochair: Marian J. Borg Major Department: Sociology The purpose of this study was to evaluate a model of life course criminal offending (recidivism) that included cognitive ability, history of alcohol and drug misuse, marital status of parents and parolees and age of parolee at reception. Analyses were conducted on a sample of 4,065 juvenile parolee offenders who were committed to the California Youth Authority in 1964 and 1965, with a 20-year follow-up of arrest data. Results suggest that 1) failure rates were similar across race, 2) time to failure varied across race with White Americans failing sooner, and 3) the risk factors were similarly unrelated to recidivism across race. vii

PAGE 8

CHAPTER 1 INTRODUCTION The prevalence of crime in the United States is an issue that concerns the American population and those who try to explain it. In addition to the costs to the victims and their families, crime disrupts the ties among and within neighborhoods and its members and might have longer lasting effects on society. Criminal activity can start anywhere, any place, and any time. Some of the people in the community are not aware of how violent their community could really be. Furthermore, the extraordinary high rates of violent crimes in the United States are disrupting the communities. Crime has become an important issue in the United States. Although crime decreased in the 1990s, recent increases in crime have made legislators and lawmakers design new and tough laws (Blumstein and Wallman, 2000). Adult crime has always been an issue in almost every state in the United States. Many studies have identified the possible explanations and subsequent policy recommendations regarding adult crime. However, the new generation of citizens, i.e. juveniles, have started committing more and more crimes in the past ten years (William, DiIulio, and Walters, 1996). The crime rate increase does not stop just with the prosecution and incarceration of adult criminals; it also has to do with the nature of the crime and who commits the crime. Juveniles have always been responsible for a disproportionate share of violent crime, but the problem has worsened since the mid-1980s (Office of Juvenile Justice and Delinquency Prevention, 1992). Around the world, crimes committed by juveniles are a major concern. In the United States, some individuals warn of a crime wave of epidemic 1

PAGE 9

2 proportions, as changing demographics will increase the number of offense-prone juveniles (Snyder and Sickmund, 1999). What makes the difference between juveniles and adults is that juveniles are assumed to be less responsible for their deviant behavior. Juveniles are less culpable and more responsive to positive behavioral change than adult offenders (Yablonsky, 2000). One of the most important issues facing the criminal justice system is the prediction of juvenile delinquency (Wierson and Forehand, 1995). Every year, almost 3 million teenagers in the United States fall victim to crimes, almost 2 million of these cases involve violence, and some end in far more tragic results where juveniles die in situations where firearms are involved (Baron, 2001). Evidence indicates that juveniles are at comparatively high risk of becoming offenders. Studies have indicated that a gradual increase in the juvenile crime rate beginning in 1960 and continuing through the 1990s (Bureau of Justice Statistics, 1994), leading to more than 100,000 youth incarcerated yearly in the United States (Office of Juvenile Justice and Delinquency Prevention, 1999), and that high crime communities tend to be low in economic status, with economic status being measured in terms of such variables as income and poverty (Lerner and Galambos, 1998), unemployment, welfare, occupation, education, inequality, owner-occupied dwellings, and substandard housing among others. High crime communities also tend to be large in size and high in population density, overcrowding, residential mobility (particularly poor communities), and percentage non-White (Agnew, 1999). It is not only the risk of becoming an offender that is of concern but also the fact that the juveniles re-engage in criminal activity over time. The longer the juvenile keeps

PAGE 10

3 out of trouble the less he/she will be seen throughout the criminal justice system. The main goal of reentry into society is to have returned an individual who has discharged his/her legal obligation to society by serving his/her sentence and has demonstrated an ability to live by societys rules. To achieve this goal, the primary objective for the offender and the criminal justice system is to prevent the repetition of antisocial behavior, and this process should begin at sentencing and continue throughout the period of release. And when the goal of reintegration has been met, the moment should be officially acknowledged so that the offenders new life can begin (National Institute of Justice, 2000). To accurately predict the recidivism of juvenile offenders, researchers need information on a variety of sources such as psychological education, prior criminal behavior, progress in the institutional program and family involvement, social adjustment, personal characteristics, social and economic environment, and the ethnic characteristics, among others. However, the more important objective for the juvenile system is to identify juvenile offenders before they become chronic offenders. Since prior research tends to only provide descriptive data and partial direction on the causes and patters in problems of recidivism, and the increase of seriousness of crimes committed by young offenders, the purpose of this study is to explore, in depth both patterns of recidivism among a sample of juveniles and the possible variables associated with recidivism among White Americans, African Americans and Hispanics. In the next section, I present a review of the literature that will explain some of the variables that have been examined in prior research that relate to recidivism, general concepts associated with recidivism, and I will try to fill the gap in knowledge about

PAGE 11

4 recidivism across different ethnicities. To examine this issue, this project uses a sample of youth released from the California Youth Authority and followed for 4 years.

PAGE 12

CHAPTER 2 REVIEW OF LITERATURE Juvenile Delinquency Juvenile delinquency refers to any number of behaviors performed by young people that are violations of laws applicable to young peoples behavior (Kaplan, 1984). Crime patterns and their consequences have been studied, but less importance has been given to juvenile crime. According to James Alan Fox (1994) juvenile delinquents are not only violent teens maturing into even more violent young adults, but also they are being succeeded by a new and larger group of violent teenagers (Minerbrook, 1994). Increasingly, juvenile courts have required offenders to compensate their victims (Butts and Snyder, 1992). Legislators attempt to restore the victims losses by restorative justice, community service, or time in jail. Unfortunately, at times, these programs do not advocate the steps toward crime-free lives. Empirical research has tried to identify the risk factors that characterize the juvenile crime activity but research has been low to materialize on this front, especially across race. But is the recent offending increase equal for all ethnic-races? Among Whites and African Americans, the latter have higher crime rates than Whites (Jackson, 1997). Non-whites commit not only more crime, but also commit more serious crime than other groups (Jackson, 1997), and this is especially the case among African Americans. Furthermore, juvenile offenders with extensive prior records appear to have a high 5

PAGE 13

6 probability of reoffending (Visher, Lattimore and Linster, 1991), and this is especially the case among minorities. Recidivism In the study of recidivism, research on criminal careers is now considered the state of the art (Farrington, 1992). A criminal career is a sequence of offenses during a period of an individuals life, meaning that the criminal behavior is repeated several times and it is not an isolated incident. This repetitive criminal behavior is called recidivism and indicates the proportion of the population that becomes involved in criminal behavior, at what age the criminal behavior begins, how long the criminal career lasts, and the number of offenses typically committed during the course of the career (Farrington, 1992). It is also defined as an estimate of the percentage of released prisoners who commit another offense. Estimates of recidivism vary with the length of the follow up period and the measures selected (see Bureau of Justice Statistics, 1994). For example, in a recidivism study in Kentucky (Kentucky Department of Corrections, 2000), approximately 33% of the total of the inmates released in 1995 returned to prison within two years. The ideal approach for studying recidivism requires a longitudinal data (a follow-up of a period of time). Taxman and Piqueros (1998) analysis of drunk-driving recidivism conceptualized the dependent variable, recidivism, as whether or not recidivism occurred within the follow-up period, and the time until recidivism measured in days. Similarly, Lee (1980), conceptualized recidivism as a dynamic social process in which the time until recidivism, conceived as a failure or rearrest, is the dependent variable. Policymakers also consider recidivism an important measure of the corrections systems performance (Florida Department of Corrections, 2001).

PAGE 14

7 According to the Bureau of Justice Statistics (1994), there are three different measures of recidivism: rearrest, reconviction and reincarceration. Rearrest refers to any arrest for a felony or serious misdemeanor that was reported to the state identification bureau or the FBI after release from a correctional institute. It also can be understood as any counts of contact between the offender and the justice system without regard to whether the contact results in the offender actually re-entering the jurisdiction of the justice system. Reconviction refers to a conviction on at least one charge after the date of release from prison. Reincarceration refers to any return to prison or any admission to a local jail with a sentence for a new offense. The Texas Youth Commission reported in the National Recidivism Methods Study (1997), that 27 states use the fallowing recidivism measures: Rearrest, 6 states, 22%; Reincarceration, 5 states; 19%; Readjudication; 3 states; 11%; Recommitment, 11 states; 41%. Recidivism research has been used differently in adult and juvenile justice populations; in adult populations it has primarily been employed by correctional policy makers in order to create a probation or parole guideline based on determinations of which individuals are at risk for future offenses (Quist and Matshazi, 2000). Juvenile justice policy makers have used recidivism as an outcome measure in rehabilitation programs. In theory, the purpose of the juvenile justice system is to avoid incarceration of children and adolescents and emphasize rehabilitation (Bartol and Bartol, 1994). The population has directed criticism at the juvenile justice system because its courts are overcrowded, recidivism rates are high, and there is a lack of system efficacy (Harrison, Maupin and Mays, 2001). Moreover, repeated juvenile delinquency can lead to

PAGE 15

8 career paths in criminal activity in adulthood. Therefore, researchers explain the use of two types of actions to prevent this criminal activity: active sanctions that are class or seminars, community service assignments, participation in educational programs, and passive sanctions, verbal reprimands, written warning, and disciplinary probation (Kompalla and McCarthy, 2001), in order to prevent reincarceration. Furthermore, Olshak (1999) argued that active sanctions in particular are an alternative for reducing recidivism. Several studies have demonstrated that high-risk offenders can be identified by gathering information on a youth and parent interviews. However, gathering this information to identify juveniles as chronic offenders is limited because it is too costly to do so. Consequently researchers argue that looking at second-time offenders is more plausible to identify them as chronic offenders because they represent a small proportion of all juveniles brought to court, they recidivate more frequently, and likely, become chronic delinquents (Smith and Aloisi, 1999). Theoretical Analysis Theoretical accounts have been applied to understand recidivism. Two theories will be discussed in this section Hirschis Social Bonding Theory For most criminologists, Hirschis social bonding theory is referred to as control theory (Akers, 1999; Barlow and Kauzlarich, 2002). This theory proposes that the delinquent acts result when an individuals bond to society is weak or broken (Hirschi, 1969:16). The assumption of the theory is that deviant behavior is lessened when people have strong bonds with other members of society. It is an individuals ties to social institutions that inhibit or control them from committing deviant acts.

PAGE 16

9 The four main elements of this theory are: attachment, commitment, involvement and belief. Hirschi argues that the stronger these elements with parents, neighbors, adults and peers are, the more the individuals behavior is controlled and maintained in conformity, and the less likely the opportunity for the individual to commit crime. On the other hand, the weaker these elements are, the weaker the social bonds are, and the more likely the individual will commit crime. Another important aspect of this theory is that these elements are inter-correlated, meaning, the weakening of one will probably weaken the others (Akers, 1999; Yablonsky, 2000). Attachment refers to the close affection ties to others, admiration of them and identification with them. According to this element, if individuals are insensitive to others opinion, the more likely the individual is going to commit a crime, because they are less compelled by the norms (Akers, 1999). Hirschi argues that attachment to individuals such as parents is important in controlling delinquent tendencies and maintaining conformity. The important aspect is the attachment to people, not necessarily the character of the person that the individual is attached to: The more one respects or admires ones friends, the less likely one is to commit delinquent acts (1969:152). Commitment refers to the extent to which individuals have built up an investment in conventionality and can be jeopardized or lost if they engage in deviant behavior. People develop a stake in their education, reputation or building careers investments such as a job or family are in line according with this element. When people are faced with the decision to commit a criminal act, they weigh what they can lose and what they can gain if they commit a crime. Commitment refers to a more or less rational element in the decision to commit a crime.

PAGE 17

10 Involvement refers to an individuals involvement in conventional activities, e.g. leisure activities, spend time with parents, friends, athletic clubs, and participation in legitimate extracurricular activities. The more the individuals involvement in conventional activities, the less likely they are believed to get in trouble or commit crime. If the individual is busy doing other conventional things, they will not have time to commit crime. Belief refers to the importance of obeying the rules and the belief of legitimacy of conventional values and norms. Hirschi argues that this element does not refer to the belief about specific norms or laws; it is just the belief of general norms or laws. Hirschi argues that the less a person believes the individual should obey the rules, the more likely the individual is to violate them. Hirschi (1969) tested his control theory with a self-report survey administered to 4,000 high school students in California. He found that youth with more attachment to parents were less likely to show deviant behavior. In regard of commitment, Hirschi found that commitment to conventional values was related to good behavior. However, conventional activities lead to some increase in crime. Agnew (1985) explored this theory in a study of almost 1,900 male youths. The juveniles were first interviewed in the tenth grade and then again in the eleventh grade. For involvement, he found it to be stable over the two-year period. However, attachment to parents, school grade and commitment explained only 2% of the variation in delinquency. Agnew described this finding as the result of the maturation of the children and therefore the importance of the bonds may become diminished overtime. In another study, the attachment between the child and the parent was found to not be correlated

PAGE 18

11 with delinquency, and the involvement in school-related activities and delinquency were negatively related (Greenberg, 1999). Moffitts Developmental Theory According to Moffitt (1993), rates of prevalence and incidence of offending is higher in adolescence such that by age 17, we observe the highest rates of criminality at which point it drops when the adolescent reaches the adulthood. However, this does not mean that criminal behavior starts a few years before age of 17; instead for some offenders criminal activity emerges very early in life, age 7. Furthermore, the career activity of an adolescent changes over time. There is a group of adolescents that engage in antisocial behavior at every stage of the life-course, life-course-persistent, and a larger group of offenders who follow the well-known age-crime curve, or adolescence-limited. Among those persistent offenders, a small percentage of them are frequent crime offenders, meaning that the most persistent offenders are responsible for about 50% of known crimes. These high-rate offenders begin their criminal activity earlier and continue through time, have been arrested by police in the early teen years, which can be seen as an important variable in predicting long-term recidivistic offending (Moffitt, 1993). In her developmental taxonomy, Moffitt presents some of the causes of individuals antisocial behavior. She starts by pointing out factors that can be present at individuals birth or before birth. One of the common factors is the neuropsychological ([A]natomical structures and psychological processes within the nervous system [which] influence psychological characteristics such as temperament, behavioral development, cognitive abilities, or all three, Moffitt, pp. 677) function of the infant; this means that the neural development can be disrupted by the use of drugs or poor nutrition before birth. In addition internal factors can also contribute to neuropsychological risk, for

PAGE 19

12 example child abuse and neglect. This may explain who becomes a frequent offender and who does not. Moffitt also argues that verbal and executive functions are correlated with antisocial behavior of life-course-persistent offenders. In this sense, she identifies learning disabilities that cause problems with the expressive speech, writing skills, and reading skills among others. These learning disabilities are also caused by parental heritance or lack of remedial cognitive stimulation provided by the parents to their childrens below average cognitive capacities. In sum, Moffitt argues that criminal activity is caused by cutting off opportunities to practice prosocial behavior, due to discipline problems and academic failure and could lead to a life-course-persistent offending. On the other hand, those youth with lees likelihood of following life-course-persistent trajectories encounter motivation for crime when they enter to adolescence; during that period they mimic others behavior, but eventually they enter into adulthood and desist from crime. Analysis of the Factors that are Related to Recidivism among Juvenile Offenders Research has identified several factors that contribute to recidivism. Fendrich and Archer (1991) suggested that factors expected to influence recidivism among juvenile offenders include social and demographic characteristics such as age, commitment, race and gender, intellectual functioning, i.e., IQ achievement test scores (Duncan et al., 1995; Carcach and Leverett, 1999), family functioning and environment, prior criminal history, alcohol and drug abuse, and family background (Visher et al., 1991). Also, prior research consistently has shown that most recidivism occurs fairly swiftly following release (Winner et al., 1997). When time served in institutions was introduced to predict recidivism, in a study among youth serious offenders from the California Youth Authority in 1981, Visher,

PAGE 20

13 Lattimore, and Linster (1991) found that time served contributed significantly to recidivism, and was statistically significant only for those who had not recidivated in the first 36 weeks following release. In a similar study, Schmidt and Witte (1988) found that longer incarcerations were related to a greater risk of recidivism, as measured by time to re-incarceration. The same result was found in a study conducted by Myner, Santman, Cappelletty and Perlmutter (1998). In a sample of juveniles, these authors argued that the length of first incarceration was associated with a greater likelihood of recidivism (Myner et al., 1998). Cognitive Ability Cognitive ability refers to abstract thinking ability and IQ. Researchers argue that lower cognitive ability results in impulsivity, attention deficits and lack of abstract thinking (Moffitt, 1993). Furthermore, Moffitt (1993) argues that differences in cognitive ability exist between people that engage in criminal activity during their adolescence and those who engage in life-long criminal activity. On the other hand, individuals with higher cognitive ability are more likely to desist from illegal activities than those with lower cognitive ability. In sum, those with more significant academic deficits are more likely to commit a delinquent act again in their life. Recidivism has been found to be higher for persons who had not completed high school than among high school graduates. Delinquent behavior has been found to be associated with poor academic performance (Farrington, 1987) and school attendance is also predictive of recidivism (Cymbalisty et al., 1975). Consistent with Farringtons findings (1987), Myner et al. (1998) found, in a study among male juveniles convicted of criminal offenses in California, that lower GPAs are related to reoffending. Researchers reported that men who had a lower IQ scores were at greater risk for criminal activity

PAGE 21

14 than those who had higher scores. Individuals with lower IQ scores reported more delinquent acts than individuals with higher IQ scores. In another study, Haberman and Quinn (1986) found that in a sample of 759 previously incarcerated juvenile offenders only about 12% completed high school after returning to the community. Age When looking at age, studies have related the age at first arrest to recidivism. Furthermore, this relationship seems to be curvilinear, i.e. recidivism increases in the early teenage years and decreases after about age sixteen (Fendrich, Archer and Kennedy, 1998; Gruenewald and West, 1989). In a study by Lanza-Kaduce, Gartin, Dundes, Blount, Boulinois-Manning, Kuhns, Holten, Mahan, Taylor, Terry, Frank and Godshall (1990), they found that older inmates reoffended less frequently. The Bureau of Justice Statistics (1987a) found similar results showing that recidivism was inversely related to the age of the prisoner at time of release: the older the prisoner the lower the rate of recidivism. The same study documented that more than 75% of those age 17 or younger when released from prison were rearrested, compared with 40.3% of those age 45 and older. The peak age of onset (the beginning) of a criminal career was 14 (4.6% of first convictions), with a second peak at 17 (4.4%). The Kentucky Department of Corrections (2000) found similar results in that young offenders recidivated at higher rates than older individuals. In a sample of 5,509 Australian juvenile offenders, Carcach and Leverett (1999), found that juveniles between the ages of 14 and 17 years recidivate the most compared with the older ones. The authors suggest that this pattern is caused by the influence of delinquent peers. Moreover, those individuals who were first convicted at the earliest age (10 to 13) have been found to be the most persistent offenders. The average duration of criminal

PAGE 22

15 careers dropped sharply from those first convicted between the ages of 14 and 16 (8.2 years) to those first convicted between 17 and 20 (2.7 years). This finding suggests that males first convicted as juveniles were much more persistent offenders than those first convicted as adults (Farrington and Hawkins, 1991). In another study, the Florida Department of Corrections (2001) found that the average inmates probability of reoffending drops by 2.1% for each year older the inmate is at release and that younger offenders reoffend more than older offenders. Overall, more than three quarters (78.3%) of those convicted as juveniles were also convicted as adults, compared with less than a quarter (21%) of those not convicted as juveniles. Seriousness of Offense Several studies have documented that high rates of juvenile recidivism are found among chronic juvenile offenders and those who commit serious crimes (Bureau of Justice Statistics, 1987a). Recidivism is higher for those who have committed serious offenses (Kentucky Department of Corrections, 2000). Gruenewald and West (1989), in their study of recidivism among juvenile offenders, found 73% recidivism for serious and chronic offenders. Another study by Visher, Lattimore, and Linster (1991) found that chronic juvenile offenders released from California Youth Authority have an 88% recidivism rate. Duncan, Kennedy and Patrick (1995), in a study of 129 juveniles released from a training school in Florida, found that 52% of the juvenile offenders were rearrested within 6 months of their release. Similarly, the Bureau of Justice Statistics (1987a) found that 69% of juvenile delinquent parolees were rearrested for serious crimes within 6 years of their release. However, in a study of recidivism of 396 prison releases over the 12 months period following their dates of release from confinement, Lanza-Kaduce, Parker and Thomas

PAGE 23

16 (1999) found that sex offenders and violent offenders were less likely to re-offend. Recidivism is also more likely among property offenders than among violent offenders, which is likely to be due to the differences between parolees from private and public prisons (see also Craig and Budd, 1967). Similar results were obtained by the Florida Department of Corrections (2001) whose study of male parolees uncovered that released inmates whose primary offense was violent, were 32% less likely to reoffend. Carcach and Leverett (1999), found that juveniles dealt in court for violent and property offenses had similar levels of recidivism, and the most interesting finding was that these two groups of offenders, i.e. violent and property offenders, had higher rates of recidivism compared with those charged with drug offenses. Alcohol and Drug Abuse A drug is any chemical or chemicals that change the body function, and may alter its biological structure (Bureau of Justice Statistics, 1994). Alcohol and drug abuse is a global phenomenon. It has been said that the use of alcohol may increase hostile behavior (Reiss and Roth, 1993). Many of the crimes committed have their roots in alcohol and drug addictions, and this has lead to an increase in the United States prison population. Most directly, it is a crime to use, possess, manufacture or distribute drugs classified as having a potential for abuse (Bureau of Justice Statistics, 1994). Several studies have shown that alcohol and drug abuse predicts recidivism and it is a contributor to criminal behavior (Myner at al., 1998, Wierson and Forehand, 1995). Evidence of the association between drugs, alcohol, and violence has been well documented. For example, in a report by the Bureau of Justice Statistics (1993), almost 50% of prison inmates had used illegal drugs in the month before the offense for which they were incarcerated and about 30% were under the influence of drugs at the time of

PAGE 24

17 the crime. In sum, empirical research has consistently shown a relationship between alcohol use and aggressive behavior (Parker and Auerhahn, 1998; Taylor, 1983). In another study, Taxman and Piquero (1998) showed that alcohol use contributes to criminal offenses with a sample of Maryland offenders. Furthermore, Ge et al. (2001) argued that adolescents who smoke, drink and use illicit drugs are more likely to commit delinquent acts compared to those who do not smoke, drink or use illicit drugs. In a sample of 27,414 arrestees, Martin et al., (2001) found that almost 95% of the sample reported that they have tried alcohol at some point during their lives; approximately 52% of the sample had also consumed alcohol 72 hours prior to their arrest and almost 21% reported the use of alcohol at the time of their offenses. Family Background Family background can be defined as the sum of positive or negative factors, which include providing for basic needs, attention, marital status of the parents, sexual, psychological abuse, domestic violence between parents and employment of parents among others (Heilbrun et al., 2000). Adolescents that experience these stressors are considered at risk meaning that juveniles with these risk factors begin a trajectory of criminality, addiction and dependency early in life (Carr and Vandiver, 2001). Cohesion and good communication between parents and children is strongly associated with good adaptation. Analysis of youths family characteristics such as criminal behavior of parents or siblings, poor parenting, family conflict and disruption in family structure are characteristic of serious juvenile offenders (Visher et al., 1991). One longitudinal study showed that delinquency is associated with parent separation, implying that the marital status of the parents lead to juvenile delinquency (Patterson et al., 1989). Patterson et al.

PAGE 25

18 has also argued that the lack of family supervision is an important determinant of early onset delinquency, and research supports this hypothesis (Patterson et al., 1992). Race Regarding race, research suggests white releasees are less likely to reoffend than are releasees from other racial groups (see Lanza-Kaduce, Parker and Thomas, 1999; Bureau of Justice Statistics 1988, 1992). The recidivism rate is higher among African Americans and Hispanics, and it is increasing steadily (Kentucky Department of Corrections, 2000), with non-whites having slightly higher recidivism rates than whites, approximately 5 to 8% higher for each measure. Individuals of Hispanic-origin also had recidivism rates that were about 6% higher than those among non-Hispanics (Bureau of Justice Statistics, 1987b). On average, African Americans recidivate 43% more than non-African Americans. In sum, African Americans appear to recidivate at higher rates than White Americans (Florida Department of Corrections, 2001). The Bureau of Justice Statistics (1987a), in a report about recidivism of young parolees, found that recidivism rates were higher among African Americans: Hispanics had recidivism rates that were 5 to 12% less than African Americans, but higher rates than White Americans. In another study, Ge et al. (2001) found that African American juveniles were arrested more frequently than were [White Americans] European Americans. Similar results were found for Hispanics having more arrests than European Americans. Current Focus The primary purpose of this study is to identify the risk factors associated with: 1) White Americans; 2) African Americans; and 3) Hispanics in terms of their ability to predict recidivism. Risk factors analyzed represented information on demographic,

PAGE 26

19 familial, and behavioral characteristics available in records at the California Youth Authority (CYA) and Wenks (1990) data set. The present study also explored serious, and violent crimes committed by juveniles, and the factors that could be related to recidivism, age of juvenile offenders, cognitive abilities, alcohol and drug use, and marital status of parents and the juvenile offenders in attempt to add to the limited number of studies in the recidivism literature, especially across race. In order to achieve this goal, survival analysis will be used as the method of analysis. This study also uses two theories as a theoretical backdrop: Social Bonding Theory (Hirschi, 1969) and Developmental Theory (Moffitt, 1993). According to the former, delinquency takes place when social bonds are weak. In this study, marriage is used as the indicator of social bonds and it is expected to be negatively correlated with recidivism. Marital status of the parolees could be correlated with Hirschis Social Bonding Theory elements of achievement and commitment. The former refers to close affection ties to others and the latter to the investment that individuals have built up and can be jeopardized if they engage in criminal behavior. Parolees marital status could inhibit their antisocial behavior as a result the close ties with their wife and because they do not want to loose their family. The higher the attachment and commitment of the parolees with their wife, the lower the likelihood of reoffending, and the lower the attachment and commitment, the higher their antisocial behavior after being release. Moffitts Developmental theory predicts that serious and persistent antisocial behavior is caused by learning disabilities, among other influences. Some of the variables that I am going to use to predict recidivism across race have been mentioned by these two theories. Cognitive ability is one of the recidivism predictors that is going to be used in

PAGE 27

20 the model, and according to Moffitts theory, lower cognitive ability results in higher antisocial behavior and therefore in higher rates of criminal activity. In this sense, the lower the cognitive ability of the juvenile offenders, the higher the likelihood of recidivism, and the higher their cognitive ability, the lower the likelihood of recidivism. The scarcity of studies that examines the predictive power of recidivism across race makes this study a contribution to the literature. The importance of the topic lays in the lack of information and research about recidivism and the three major ethnic groups. Looking for answers about recidivism in the United States makes this topic relevant and a unique contribution.

PAGE 28

CHAPTER 3 DATA The data used here is part of a larger study designed to examine the criminal career patterns of violent juvenile offenders (Wenk, 1990). The data contain a total sample of 4,146 juvenile male offenders who were committed to the Reception Guidance Center at the Duel Vocational Institution in Tracy, California between 1964 and 1965. This institution was a facility operated by the California Youth Authority (CYA), which provides a setting for training, treatment, and education of serious, juvenile offenders. Two studies have used Wenks data to predict recidivism using different sets of variables. Donnellan, et al. (2000) and Ge et al. (2001) conducted both of the studies. Their first approach looked at the cognitive abilities among adolescent-limited and life-course persistent offenders among White Americans, African Americans and Hispanics. They argued that delinquent juveniles have lower scores on test of cognitive abilities. In order to gather information on adolescents cognitive abilities, Ge et al. looked at different tests; the California Achievement Test (CAT), California Test of Mental Maturity (CTTM), General Aptitude Test Battery (GATB) and Raven Test of Progressive Matrices. Tiegs and Clark (1951) designed the CAT or California Achievement Test that was used in California schools to determine academic achievement. It gauges information on academic domains of reading, arithmetic, language and overall academic achievement. The CTTM is a test designed by Sullivan et al., (1963) to test language and non-language 21

PAGE 29

22 intelligence. The general Aptitude Test Battery was designed by the U.S. Department of Labor. Its diverse test includes verbal, arithmetic, spatial, perceptual and general intellectual abilities. Finally, the Raven Test of Progressive Matrices was developed by Raven (1960) and measures general intelligence abilities. The researchers found that for White Americans, juvenile offenders had significantly higher scores in the cognitive tests than life-course-persistent offenders. Similar results were found for Hispanics; however, for African Americans, the results were opposite showing that there were no significant differences between adolescent-limited offenders and life-course-persistent offenders on tests of cognitive ability. The second study conducted by Ge et al., a year after their first one (2001) looked at 2,263 juvenile offenders. Variables such as age at first arrest, family environment, cognitive ability, and early behavioral problems were measured to predict the persistence of criminal offending into adulthood among males. Looking first at the cognitive ability analyses, the researchers found that cognitive ability was not significantly related to juvenile delinquency. In a second analysis, they turned to the question of the prediction of chronic offending; in this case, cognitive ability was a significant predictor of persistent offending after the offenders turned 18. When the researchers looked at alcohol and drug abuse, they found that its use influenced the amount of delinquency but not the timing of delinquency. In sum, Ge et al. found that low scores on cognitive ability were significantly associated with frequencies of arrest. Furthermore, involvement with alcohol and drug use was a predictor of adult arrest, suggesting that alcohol and drug abuse influence the criminal behavior of the juveniles. Importantly, neither of these two

PAGE 30

23 studies examined the inter-relationships between race and cognitive abilities within a survival framework. Variables Independent Variables Six independent variables will be used from Wenk's data in this research. The first variable, California Achievement Test, is used to determine academic achievement; it gauges information on academic domains of reading, arithmetic, language and overall academic achievement. This variable measures the scores of the California achievement total grade placement of the juvenile offenders. Marital status of the parolee is the second variable. This variable measures the marital status of the juvenile offenders. This variable will be recoded into two variables: not married (0) and married (1). The third variable that will be used in the research is marital status of natural parents. The variable will be coded into not married (0) and married (1). Alcohol and drug use are the fourth and fifth variables to be incorporated in the research. History of alcohol misuse measures either if the juvenile offenders have had moderate problems or severe problems with alcohol in general, which means, if they consumed alcohol on a daily basis. History of drug use dealt with use in general, and it measures if the juvenile offenders have used drugs on a daily basis. Age at reception is the sixth variable and measures the age of the juvenile offenders at reception. The mean for the juvenile offenders age at reception is 19.0 with a standard deviation of .97 (range: 16-24).

PAGE 31

24 Dependent Variables The dependent variables that will be used in the research are as follows: days to parole suspension measures the days to which the juvenile offenders have committed another crime and been suspended from parole. The variable parole outcome success measures if the juvenile offender, after being released from jail, has or has not recidivated (0=failure, 1=success). Research Questions There are two main questions to be answered in this research: 1. Are the survival recidivism patterns the same for White Americans, African Americans and Hispanics? And 2. Do the independent variables relating to recidivism differ among White Americans, African Americans and Hispanics? Analytic Plan As mentioned above, the survival analysis method will be used to examine the data. To analyze the timing of recidivism, Cox proportional hazard regression (Cox, 1972), a type of survival method, will be used. The method seeks to explain the differences in terms of both of the following criteria whether a reconviction occurred (i.e., failure), and the period of time until the reconviction. The Cox proportional hazard regression facilitates analysis because it tests the dependent variables and its effects separately while controlling for the other variables and it is a method for modeling time-to-event data en the presence of censored cases. The hazard model seeks to focus on the patterns of recidivism exhibited in the sample until data censoring. The hazard model provides the probability that a subject will fail in a period of time. It also provides more information than typical recidivism measures because it allows for un-covering heterogeneity among subjects with respect to the risk of recidivism. Importantly, it also does not make

PAGE 32

25 assumptions about the distribution of the underlying heterogeneity, and it identifies covariates that tend to be associated with the time to failure (Visher et al., 1991).

PAGE 33

CHAPTER 4 FINDINGS Results for One-Way ANOVA One-way analysis of variance (ANOVA) was performed on the test of the independent and dependent variables separately for each ethnic group. The idea was to conduct separate analyses for each ethnic group by controlling for the impact of ethnicity on test scores. Table 1 provides the results of the analysis for White Americans, Hispanics, and African Americans. Although Whites Americans, African Americans and Hispanics did not differ on parole outcome/success, there were mean differences on time to failure. For example, White Americans were more likely to fail sooner, in terms of days, than were African Americans by approximately forty days. On the other hand, Hispanics were less likely to fail sooner than African Americans by approximately one day. As shown in Table 1, Cognitive Ability was significantly different across ethnic groups, p< .000. White Americans had higher cognitive abilities than African Americans and Hispanics. History of Alcohol Misuse and History of Drug Misuse was also significantly different across ethnic groups, p< .000. This pattern suggests that Hispanics were more likely to be involved in the use of alcohol and drugs than White Americans and African Americans. However, African Americans were less likely to be involved in the use of alcohol and drugs than White Americans. 26

PAGE 34

27 Table 1 Descriptive Statistics: One-Way ANOVA analysis. Variable White African-American Hispanic Sig M SD M SD M SD Cognitive A. 6.63 12.79 5.15 11.69 4.9 13.23 w>b>h* Marital S. Parolee .13 .34 .12 .33 .16 .36 yes*** Marital S. Parents .48 .50 .49 .50 .56 .50 yes* History Alcohol .56 .72 .43 .63 .94 .79 h>w>b* History Drug .23 .63 .19 .55 .45 .82 h>w>b* Parole (1=succ) .61 .48 .61 .49 .61 .49 no Days parole susp. 251 168.93 294 185.13 295 199.66 w>b>h* ________________________________________________________________________ Notes: *p< .000, **p< .05, ***p< .10 Results of the Survival Analysis The outcome measures discussed focus on whether the juvenile offender has recidivated or not. The way to operationalize recidivism is by examining the length of the time until the new arrest. This measure of recidivism allows the researcher to examine desistance from criminal behavior and to explore the differences across race. To analyze the timing of recidivism, I used Cox proportional hazard regression, which is a method for modeling time-to-event data in the presence of censored cases. The results of the analyses are presented in Figure 1 (Survival function). Figure 1 presents the survival function, taking into account the variable days to parole suspension and parole outcome success. Juvenile offenders recidivate the most approximately after 480 days of being released. We can observe this same result if we look at the hazard rate, which is the estimate of the probability per unit time that a case that has survived to the beginning of an interval will experience an event in that interval. The hazard rate gets

PAGE 35

28 higher approximately at days 490 and 504. The median survival time for the juvenile offender data is 239.03 days. Survival FunctionDAYS TO PAROLE SUSPENSION10008006004002000-200Cum Survival1.21.0.8.6.4.20.0-.2 Figure 1 Survival Function Table 2 contains the results of the Cox proportional hazard regression for the sample of Hispanics. The first column is B (slope coefficient); the second column is the SE (standard error), the third is the Wald statistic (this statistic test the hypothesis that B equals 0 in the population). The negative coefficient means that a higher value on the independent variable leads to a lower likelihood of re-arrest, i.e. the effect of the variable is such that it increases the time until parole failure. For Table 1, Hispanics were less likely than non-Hispanics to incur parole failure. In other words, Hispanics had a longer time to failure than non-Hispanics.

PAGE 36

29 Table 2 Cox Proportional Hazard Models Predicting Hispanic Juveniles Parole Success/Failure Variable B SE Wald Race (1 = Hispanic) -.222 .075 8.872 Cognitive A 000 .003 .002 Marital S. Parolee -.035 .089 .154 Marital S. Parents -.003 .058 .002 History Alcohol .027 .040 .457 History Drug .020 .041 .022 Age par at reception -.002 .032 .002 Notes: Log-Likelihood = 14619.15; X2= 9.180; df= 7; p= .220 *p< .05 Table 3 contains the Cox Regression for the African American sample. The results suggest that African Americans were less likely than non-blacks to incur parole failure. In other words, African Americans had a longer time to failure than non-blacks. Table 4 provides the results for the Cox Regression pertaining to the sample of White Americans suggesting that White Americans were more likely than non-White Americans to incur in parole failure. The White American juvenile offenders had a lesser time to failure than non-White Americans. The results for the split-race sample, which are presented in Table 5, reveal that among White Americans none of the independent variables are significantly related to recidivism. The same results were found for Hispanics. In the case of the African American juvenile offenders, all but one variable was insignificant, p< .05. The increase

PAGE 37

30 in drug use increases the probability of parole failure for the sample of African American juvenile offenders. Table 3 Cox Proportional Hazard Models Predicting African American Juveniles Parole Success/Failure Variable B SE Wald Race (1 = African A.) -.159 .066 5.755* Cognitive A .001 .003 .038 Marital S. Parolee -.052 .089 .338 Marital S. Parents -.012 .058 .041 History Alcohol -.013 .040 .106 History Drug -.007 .041 .032 Age par at reception .007 .032 .042 Notes: Log-Likelihood = 14613.00; X2= 6.044; df= 7; p= .523 *p< .10

PAGE 38

31 Table 4 Cox Proportional Hazard Models Predicting White Americans Juveniles Parole Success/Failure Variable B SE Wald Race (1 =Whites) .274 .059 21.83* Cognitive A .000 .003 .005 Marital S. Parolee -.063 .089 .509 Marital S. Parents .000 .058 .000 History Alcohol .007 .039 .027 History Drug .006 .040 .023 Age par at reception .002 .032 .005 Notes: Log-Likelihood = 14597.05; X2= 22.23; df= 7; p= .002 *p < .05 Table 5 Cox Regression Hazard Model: Split-Race Analysis Variable White Hispanic African-American B SE Wald B SE Wald B SE Wald Cognitive A .004 .004 .834 -.008 .007 1.41 -.001 .006 .055 Marital S. Parolee -.197 .123 2.562 .120 .185 .423 .089 .185 .230 Marital S. Parents .063 .082 .581 -.214 .131 2.68 .051 .110 .215 History Alcohol -.027 .055 .237 .143 .088 2.60 -.009 .085 .011 History Drug -.011 .056 .036 -.060 .075 .640 .216 .099 4.76* Age par at reception -.034 .044 .583 .096 .070 1.86 -.009 .062 .021 Notes: *p < .05

PAGE 39

CHAPTER 5 DISCUSSION AND CONCLUSIONS The main purpose of this study was to identify the risk factors associated with White Americans, African Americans and Hispanics in predicting recidivism. This study is the first that I am aware of to compare three different ethnic races in predicting recidivism on a series of independent variables on a model of life-course criminal offending. The data was collected by Wenk (1990), and presents information on demographic, familial and behavioral characteristics in records at the California Youth Authority (CYQA). First, I examined if there was any difference across race in cognitive ability, history of drug use, history of alcohol use, parole outcome and days until reconviction. I then sought to study the influence of these correlates on parole success/failure and days to parole suspension across race. Finally, I sought to demonstrate in a split-race sample the correlation among the independent variables and the dependent variables with each ethnic group. I found differences on time to failure among White Americans, African Americans and Hispanics, whereas I did not find any differences on parole outcome/success. In this sample, White Americans were more likely to fail sooner than African Americans and Hispanics. When cognitive ability was taken into account, White Americans were more likely to have higher cognitive ability than their counter parts, but it was not related to recidivism. I found that Hispanics were more likely to be involved in the use of alcohol and illicit drugs than White Americans and African Americans. The latter were less likely 32

PAGE 40

33 to be involved in its illicit use compared with White Americans. I conducted additional analyses designed to predict recidivism across race. The results of these analyses revealed that juvenile offenders recidivated the most approximately after 480 days. Finally, looking at each ethnic group, I ran separate analyses to predict recidivism. The Cox Proportional Hazard Regression results for the sample of Hispanics showed that Hispanics were less likely to recidivate than non-Hispanics; for the sample of African Americans, they were less likely to recidivate than non-African Americans; and for the sample of White Americans, they were more likely to incur parole failure than non-White Americans. From this study, it is possible to assess different limitations. First, the records obtained by Wenk (1990) were gathered in 1963, this could suggest that the records obtained are outdated. Second, the data did not include information on offenders previous alcohol and drug use, nor did it include information on cognitive ability before release. Third, the data set only included information only for male offenders. The findings do not support any of the two theories used in this analysis. Both Social Bonding Theory (Hirschi, 1969) and Developmental Theory (Moffitt, 1993) did not show any significant relation between the variables. For example, marital status of the parolee was expected to be correlated with recidivism. However, the results showed non-significant effects. For Moffitts Developmental Theory, cognitive ability was believed to predict recidivism. In this study, White Americans had higher cognitive scores, but cognitive abilities were not related to recidivism. On the other hand, Hispanics had lower cognitive ability scores and were less likely to fail sooner than White Americans, but cognitive abilities were not directly related to recidivism.

PAGE 41

34 The null results from this study could be due to several factors. First, other different variables, such as peer relationships, should be included in the study to predict variation in the dependent variable. Second, the data set used in this study came from a select sample of serious offenders, and it should be interesting to look at a sample of non-serious offenders. The variables in the data set predicted continuing criminal activity. Finally, the juvenile criminal records were composed only from official records and further research should look at self-report records as well. Despite the limitations, these results could be considered fundamental for the study of recidivism across race. The scarcity of studies examining differences in recidivism across race makes this study an important contribution to the literature. It appears that in this particular sample cognitive ability, history of alcohol and drug use was not related to recidivism in neither of the ethnic groups, although differences were found across race in the timing of recidivism. Further research should continue studying issues related to recidivism as well as obtain different factors from those studied in this study to determine how various risk factors relate to recidivism generally, and across race in particular.

PAGE 42

LIST OF REFERENCES Agnew, R. (1985). Social Control Theory and Delinquency. Criminology 23:47-60. Agnew, R. (1999). A General Strain Theory of Community Differences in Crime Rates. Journal of Research in Crime and Delinquency, Vol. 36(2): 23-30. Akers, R. (1999). Criminological Theories: Introduction and Evaluations. Chicago, IL: Fitzroy Dearborn.. Barlow, H., and Kauzlarich, D. (2002). Introduction to Criminology. Upper Saddle River, New Jersey. Baron, S. (May 2001). Street Youth Labour Market Experiences and Crime. The Canadian Review of Sociology and Anthropology Vol. 38(2): 189-215. Bartol, C., and Bartol, A. (1994). Psychology Law: Research and Application. Belmont, CA. Benda, B., and Tollett, C., (1999). A Study of Recidivism of Serious and Persistent Offenders Among Adolescents. Journal of Criminal Justice, Vol. 27(2): 111-116. Blumstein, A., and Wallman, J. (2000). The Crime Drop in America. New York, NY: Cambridge University Press.. Bureau of Justice Statistics (1987a). Recidivism of Young Parolees. Special Report. Washington D.C.: U.S. Department of Justice. Bureau of Justice Statistics (1987b). Recidivism of Prisoners Released in 1983. Special Report. Washington D.C.: U.S. Department of Justice. Bureau of Justice Statistics (1988). Report for the Nation on Crime and Justice. Washington D.C.: U.S. Department of Justice. Bureau of Justice Statistics (1992). Pretrial Release of Felony Defendants, 1990. Washington D.C.: U.S. Department of Justice. Bureau of Justice Statistics (1993). Survey of State Prison Inmates. Washington D.C.: U.S. Department of Justice Bureau of Justice Statistics (1994). Fact Sheet: Drug-related Crime. Washington D.C.: U.S. Department of Justice. 35

PAGE 43

36 Butts, J., and Snyder, H. (1992). Restitution and Juvenile Recidivism. Washington D.C.: U.S. Department of Justice. Carcach, C., and Leverett, S. (1999). Recidivism among Juvenile Offenders: An Analysis of Time to Reappearance in Court. Research and Public Policy Series. Australian Institute of Criminology. Research and Public Policy Series, Vol. 17: 1-25. Carr, M., and Vandiver, T., (2001). Risk and Protective Factors among Youth Offenders. Adolescence, Vol. 36: 409-426. Cox, D. (1972). Regression Models and Life Tables. Journal of the Royal Statistical Society, Vol. 34: 187-200. Craig, M., and Budd, L. (1967/0. The Juvenile Offender: Recidivism and Companions. Crime and Delinquency, Vol. 13: 344-351. Cymbalisty, B., Schuck, S., and Dubeck, J. (1975). Achievement Level, Institutional Adjustment and Recidivism among Juvenile Delinquents. Journal of Community Psychology, Vol. 3: 289-294. Donnellan, M., Ge, X., and Wenk, E. (2000). Cognitive Abilities in Adolescent-limited and Life-course-persistent Criminal Offenders. Journal of Abnormal Psychology, Vol. 109(3): 396-402. Duncan, R., Kennedy, W., and Patrick, C. (1995). Four-factor Model of Recidivism in Male Juvenile Offenders. Journal of Clinical Child Psychology, Vol. 24(3): 250-257. Farrington, D. (1987). Early Precursors of Early Offending. In J. Q. Wilson and G. C. Loury (Eds.). From Children to Citizens, Vol. 3. Families, schools and Delinquency Prevention, pp. 27-50. New York, Springer-Verlag. Farrington, D. (1992). Criminal Career Research in the United Kingdom. British Journal of Criminology, Vol. 32: 521-536 Farrington, D., and Hawkins, D. (1991). Predicting Participation, Early Onset and Later Persistence in Officially Recorded Offending. Criminal Behavior and Mental Health, Vol. 1: 1-33. Federal Bureau of Investigation (1990). Crime in the United States, 1989. Washington D.C.; US Department of Justice. Fendrich, M., and Archer, M. (1998). Long-term Rearrest Rates in a Sample of Adjudicated Delinquents: Evaluation the Impact of Alternative Programs. The Prison Journal, Vol. 78(4): 360-389. Florida Department of Corrections. Factors Affecting Recidivism Rates. May 2001. Recidivism report. Tallahassee, Fl.

PAGE 44

37 Fox J., Maureen A. (2002). FBA for Children and Youth with Emotional-behavioral Disorders: Where we should go in the Twenty-first Century. Preventing School Failure Vol. 44(4): 140-141. Ge, X., Donnellan, M., and Wenk, E. (2001). The Development of Persistent Criminal Offending in Males. Criminal Justice and Behavior, Vol. 28(6):731-755. Greenberg, D. (1999). The Weak Strength of Social Control Theory. Crime and Delinquency 45(1): 66-81 Gruenewald, P., and West, B. (1989). Survival Models of Recidivism among Juvenile Delinquents. Journal of Quantitative Criminology, Vol. 5(3): 215-217. Haberman, M. and Quinn, L. (1986). The High School Re-entry Myth: A Follow-up Study of Juveniles Released from Two Correctional High Schools in Wisconsin. Journal of Correctional Education, Vol. 37, 114-117. Harrison, P., Maupin, J., and Mays, G. (2001) Teen Court: An Examination of Process and Outcomes. Crime and Delinquency, Vol. 47(2): 243-264. Heilbrun, K., Brock, W., Waite, D., Lanier, A., Schmidt, M., Witte, G., Keeney, M., Westendorf, M., Buinavert, L., and Shumate, M., (2000). Risk Factors for Juvenile Criminal Recidivism. Criminal Justice and Behavior, Vol. 27(3): 275-291. Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California Press. Jackson, K. L. (1997) Differences in the Background and Criminal Justice Characteristics of Young Black, White, and Hispanic Male Federal Prison Inmates. Journal of Black Studies, Vol.27(4): 494-509. Kaplan, H. (1984). Patterns of Juvenile Delinquency. Beverly Hills, CA: Sage Publications. Kentucky Department of Corrections (2000). Recidivism in Kentucky 1993-1995. Executive Summary. Frankfort, KY. Kompalla, S., and McCarthy, M. (2001). The Effect of Judicial Sanctions on Recidivism and Retention. College Student Journal, Vol. 35(2): 223-231. Lanza-Kaduce, L., Gartin, P., Dundes, L., William, R., Boulinois-Manning, S., Kuhns, J., Holten, G., Mahan, S., Taylor, D., Terry, C., Frank, H., and Godshall, B. (1990). Prison Utilization Study: Risk Assessment Techniques and Floridas Inmates, Vol. 1. Application to Males and Females, Tallahassee, Fl; Division of Economic and Demographic Research, Joint Legislative Management Committee of the Florida Legislature.

PAGE 45

38 Lanza-Kaduce, L., Parker, K., and Thomas, C. (1999). A Comparative Recidivism Analysis of Releasees from Private and Public Prisons. Crime and Delinquency, Vol. 45(1): 28-47. Lee, E. (1980). Statistical methods for survival data analysis. Lifetime Learning Publications. Lerner, R. and Galambos, N. (1998). Adolescent Development: Challenges and Opportunities for Research, Programs and Policies. Annual Review of Psychology, Vol. 49, 413-446. Maine Department of Corrections (1998). Recidivism. Division of Juvenile Services. Juvenile Recidivism Baseline Report. Charleston, ME. Martin, S., Bryant, K., and Fitzgerald, N. (2001). Self-reported Alcohol Use and Abuse by Arrestees in the 1998 Arrestee Drug Abuse Monitoring Program. Alcohol Research and Healthy, Vol. 25(1): 72-79. Minerbrook, S (1994). A Generation of Stone Killers. US News and World Report (January), pp. 33-37. Moffitt, T. (1993). Adolescence Limited and Life Course Persistent Antisocial Behavior: A Developmental Taxonomy. Psychological Review, 100, 674-701. Myner, J., Santman, J., Gordon, G.C., Permutter, B. (1998). Variables Related to Recidivism Among Juvenile Offenders. International Journal of Offender Therapy and Comparative Criminology, 42(1): 65-80. National Institute of Justice (2000). Sentencing and Corrections: But they all Come Back: Rethinking Prisoner Reentry. Washington D.C. US Department of Justice. Office of Juvenile Justice and Delinquency Prevention (1992). Restitution and juvenile recidivism. Special report. Washington D.C. US Department of Justice. Office of Juvenile Justice and Delinquency Prevention (1996). Juvenile Offenders and Victims. National Report. Washington, D.C. Office of Juvenile Justice and Delinquency Prevention (1999a). Juvenile Offenders and Victims. National Report. Washington, D.C. Office of Juvenile Justice and Delinquency Prevention (1999b). Juvenile Offenders in Residential Placement. Washington, D.C. Olshak, R. (1999). A Guide for Effective Sanctioning: From Theory to Practice. Normal, IL: Illinois State University. Parker, R., and Auerhahn, K. (1998). Alcohol, Drugs and Violence. Annual Review of Sociology, Vol. 24: 291-311.

PAGE 46

39 Patterson, G., DeBaryshe, B., and Ramsey, E. (1989). A Developmental Perspective on Antisocial Behavior. American Psychologist, Vol. 44: 329-335. Patterson, G., Crosby, L., and Vuchinich, S. (1992). Predicting Risk of Early Police Arrest. Journal of Quantitative Criminology, Vol. 8(4): 335-355. Quist, R., and Matshazi, D. (2000). The Child and Adolescent Functional Assessment Scale (CAFAS): A Dynamic Predictor of Juvenile Recidivism. Adolescence, Vol. 35(137): 181-192. Raven, J. (1960). Guide to the Standard Progressive Matrices. United Kingdom, London: H.K. Lewis. Reiss, A., and Roth, J. (1993). Understanding and Preventing Violence. National Research Council: Washington D.C. Schmidt, P., and Witte, A. (1988). Predicting Recidivism Using Survival Models. New York, NY: Springer-Verlag. Smith, R., and Aloisi, M. (1999). Prediction of Recidivism among Second Timers in the Juvenile System: Efficiency in Screening Chronic Offenders. American Journal of Criminal Justice, Vol. 23(2): 201-222. Snyder H., and Sickmund M. (1999). Juvenile Offenders and Victims: 1999 National Report. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Sullivan, E., Clark, W., and Tiegs, E. (1963). Examiners Manual for California Short-form Test of Mental Maturity. California Test Bureau. Taxman, F., and Piquero, A., (1998). On Preventing Drunk Driving Recidivism: An Examination of Rehabilitation and Punishment Approaches. Journal of Criminal Justice, Vol. 26(2): 129-143. Taylor, S. (1983). Alcohol and Human Physical Aggression. In Alcohol, Drug Abuse, Aggression. Springfield, IL. Texas Youth Commission. National Recidivism Method Study (1997). Recidivism. Austin, TX. Tiegs, E., and Clark, W. (1951). California Achievement Test: Manuals of Directions for Primary, Elementary, Intermediate, and Advanced Batteries. California Test Bureau. Visher, C., Lattimore, P., and Linster, R. (1991). Predicting Recidivism of Serious Youthful Offenders Using Survival Analyses. Criminology Vol. 29:329-66

PAGE 47

40 Wierson, M., and Forehand, R. (1995). Predicting Recidivism in Juvenile Delinquents: the Role of Mental Health Diagnoses and the Qualification of Conclusions by Race. Behavior Research Theory Vol. 33(1): 63-67. William, J., DiIulio, Jr. J., and Walters. J. (1996) Body Count: Moral Poverty and How to win Americas War against Crime and Drugs. Simon & Schuster, New York, NY. Winner, L., Lanza-Kaduce, L., Bishop, D., and Frazier, C. (1997). The Transfer of Juveniles to Criminal Court: Re-examining Recidivism over the Long Term. Crime and Delinquency, Vol. 43: 548-563. Yablonsky, L. (2000). Juvenile Delinquency: Into the 21st Century. Belmont, CA: Wadsworth. Wenk, E. Criminal Careers, Criminal Violence, and Substance Abuse in California, 1963-1983. Inter-University Consortium for Political and Social Research.

PAGE 48

BIOGRAPHICAL SKETCH With the constant support of my parents and brothers, I have aspired to seek a position in the law field, and had achieved this goal in my country of origin, Bolivia. As soon as I graduated from high school I began attending the Catholic Bolivian University (Universidad Catolica Boliviana) pursuing the law degree. My interest in law and especially criminology led me to continue the specialization in a foreign country. After five years of continued education and my interest on pursuing my goals, I decided to enroll at the prestigious University of Florida. I moved to Gainesville, Florida, with the aim of getting a Master in Arts at the University of Florida. I am planning to continue my education at the University of Florida and obtaining Ph.D. with a major in criminology and law. 41


Permanent Link: http://ufdc.ufl.edu/UFE0000745/00001

Material Information

Title: Studying Patterns and Correlates of Recidivism
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000745:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000745/00001

Material Information

Title: Studying Patterns and Correlates of Recidivism
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000745:00001


This item has the following downloads:


Full Text












STUDYING PATTERNS AND CORRELATES OF RECIDIVISM
ACROSS RACE















By

ROHALD ARDWAN MENESES


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Rohald Ardwan Meneses















ACKNOWLEDGMENTS


I would like to thank my committee members, Dr. Alex Piquero and Dr. Marian

Borg, for all their time and effort in helping me to accomplish this project. Special thanks

go to the chair of my committee, Dr. Alex Piquero, for all his guidance and help

throughout my thesis. Also, I thank my parents Maria and Johnny, and my brothers

Johann and Eelhard. I would also like to thank the following for all their love and

support: Dr. Lonn Lanza-Kaduce, Vicente Cuka, Jessica Malmad, Ivo Cuka, Miguel

Molina, Fernando Molina, Romeo Meneses, Salvatore Stelluto, Christina Anton, Wendy

Colon, Christian Monje, Carola Sabja, Roxana Otasevic, Maria Jose Guzman and

Jeimmy Arredondo.
















TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF TABLES ....................................................... ............ ....... ....... vi

ABSTRACT ........ .............. ............. .. ...... .......... .......... vii

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 REVIEW OF LITERATURE ......................................................... .............. 5

Juvenile D elinquency ......... ............................................................ ... ............. .5
R ecidivism ......................................................................... . 6
T theoretical A analysis ........................................... .... ... ....... .. ..... .... ......... ... 8
Analysis of the Factors that are Related to Recidivism among Juvenile Offenders... 12
C cognitive A ability .......................................................................13
A g e ........................................................................................1 4
Seriousness of Offense ............... ......................................... .. .. .............. 15
A alcohol and D rug A buse .......................................................................... 16
F am ily B background ............................................ .. .... .... ......... .. .. .... 17
R a c e ................................................................................. 1 8
C u rren t F o cu s ...................................... .............................................. 18

3 D A T A ............................................................................ .. 2 1

V ariables ......................................................... ................ ......... 23
Independent V ariables ............................................................ ............... 23
D dependent V ariables .................................................... ........ ....... ............24
R research Q u estion s............ .......................................................... .... .. .......24
A nalytic P lan ................................................................... 24

4 F IN D IN G S ................................................................................ 2 6

R results for O ne-W ay A N O V A ............................................................................. .... 26
R results of the Survival A nalysis........................................... .......................... 27










5 DISCU SSION AND CON CLU SION S ............................................. ....................32

L IST O F R E F E R E N C E S ...................................... .................................... ....................35

B IO G R A PH IC A L SK E T C H ...................................................................... ..................41






















































v
















LIST OF TABLES


Table page

1 Descriptive Statistics: One-Way ANOVA analysis..........................................27

2 Cox Proportional Hazard Models Predicting Hispanic Juveniles Parole
Success/Failure .................................... ............................... .........29

3 Cox Proportional Hazard Models Predicting African American Juveniles Parole
Success/Failure .................................... ............................... .........30

4 Cox Proportional Hazard Models Predicting White Americans Juveniles Parole
Success/Failure .................................... ............................... .........31

5 Cox Regression Hazard M odel: Split-Race Analysis ............................................31















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts

STUDYING PATTERNS AND CORRELATES OF RECIDIVISM
ACROSS RACE

By

Rohald Ardwan Meneses

May, 2003

Chair: Alex Piquero
Cochair: Marian J. Borg
Major Department: Sociology

The purpose of this study was to evaluate a model of life course criminal offending

(recidivism) that included cognitive ability, history of alcohol and drug misuse, marital

status of parents and parolees and age of parolee at reception. Analyses were conducted

on a sample of 4,065 juvenile parolee offenders who were committed to the California

Youth Authority in 1964 and 1965, with a 20-year follow-up of arrest data. Results

suggest that 1) failure rates were similar across race, 2) time to failure varied across race

with White Americans failing sooner, and 3) the risk factors were similarly unrelated to

recidivism across race.














CHAPTER 1
INTRODUCTION

The prevalence of crime in the United States is an issue that concerns the American

population and those who try to explain it. In addition to the costs to the victims and their

families, crime disrupts the ties among and within neighborhoods and its members and

might have longer lasting effects on society. Criminal activity can start anywhere, any

place, and any time. Some of the people in the community are not aware of how violent

their community could really be. Furthermore, the extraordinary high rates of violent

crimes in the United States are disrupting the communities.

Crime has become an important issue in the United States. Although crime

decreased in the 1990's, recent increases in crime have made legislators and lawmakers

design new and tough laws (Blumstein and Wallman, 2000). Adult crime has always

been an issue in almost every state in the United States. Many studies have identified the

possible explanations and subsequent policy recommendations regarding adult crime.

However, the new generation of citizens, i.e. juveniles, have started committing more and

more crimes in the past ten years (William, Dilulio, and Walters, 1996).

The crime rate increase does not stop just with the prosecution and incarceration of

adult criminals; it also has to do with the nature of the crime and who commits the crime.

Juveniles have always been responsible for a disproportionate share of violent crime, but

the problem has worsened since the mid-1980's (Office of Juvenile Justice and

Delinquency Prevention, 1992). Around the world, crimes committed by juveniles are a

major concern. In the United States, some individuals warn of a crime wave of epidemic









proportions, as changing demographics will increase the number of offense-prone

juveniles (Snyder and Sickmund, 1999). What makes the difference between juveniles

and adults is that juveniles are assumed to be less responsible for their deviant behavior.

Juveniles are less culpable and more responsive to positive behavioral change than adult

offenders (Yablonsky, 2000).

One of the most important issues facing the criminal justice system is the prediction

of juvenile delinquency (Wierson and Forehand, 1995). Every year, almost 3 million

teenagers in the United States fall victim to crimes, almost 2 million of these cases

involve violence, and some end in far more tragic results where juveniles die in situations

where firearms are involved (Baron, 2001).

Evidence indicates that juveniles are at comparatively high risk of becoming

offenders. Studies have indicated that a gradual increase in the juvenile crime rate

beginning in 1960 and continuing through the 1990's (Bureau of Justice Statistics, 1994),

leading to more than 100,000 youth incarcerated yearly in the United States (Office of

Juvenile Justice and Delinquency Prevention, 1999), and that high crime communities

tend to be low in economic status, with economic status being measured in terms of such

variables as income and poverty (Lerner and Galambos, 1998), unemployment, welfare,

occupation, education, inequality, owner-occupied dwellings, and substandard housing

among others. High crime communities also tend to be large in size and high in

population density, overcrowding, residential mobility (particularly poor communities),

and percentage non-White (Agnew, 1999).

It is not only the risk of becoming an offender that is of concern but also the fact

that the juveniles re-engage in criminal activity over time. The longer the juvenile keeps









out of trouble the less he/she will be seen throughout the criminal justice system. The

main goal of reentry into society is to have returned an individual who has discharged

his/her legal obligation to society by serving his/her sentence and has demonstrated an

ability to live by society's rules. To achieve this goal, the primary objective for the

offender and the criminal justice system is to prevent the repetition of antisocial behavior,

and this process should begin at sentencing and continue throughout the period of release.

And when the goal of reintegration has been met, the moment should be officially

acknowledged so that the offender's new life can begin (National Institute of Justice,

2000). To accurately predict the recidivism of juvenile offenders, researchers need

information on a variety of sources such as psychological education, prior criminal

behavior, progress in the institutional program and family involvement, social

adjustment, personal characteristics, social and economic environment, and the ethnic

characteristics, among others. However, the more important objective for the juvenile

system is to identify juvenile offenders before they become chronic offenders.

Since prior research tends to only provide descriptive data and partial direction on

the causes and patters in problems of recidivism, and the increase of seriousness of

crimes committed by young offenders, the purpose of this study is to explore, in depth

both patterns of recidivism among a sample of juveniles and the possible variables

associated with recidivism among White Americans, African Americans and Hispanics.

In the next section, I present a review of the literature that will explain some of the

variables that have been examined in prior research that relate to recidivism, general

concepts associated with recidivism, and I will try to fill the gap in knowledge about






4


recidivism across different ethnicities. To examine this issue, this project uses a sample

of youth released from the California Youth Authority and followed for 4 years.

















CHAPTER 2
REVIEW OF LITERATURE

Juvenile Delinquency

Juvenile delinquency refers to any number of behaviors performed by young people

that are violations of laws applicable to young people's behavior (Kaplan, 1984). Crime

patterns and their consequences have been studied, but less importance has been given to

juvenile crime. According to James Alan Fox (1994) juvenile delinquents are not only

violent teens maturing into even more violent young adults, but also they are being

succeeded by a new and larger group of violent teenagers (Minerbrook, 1994).

Increasingly, juvenile courts have required offenders to compensate their victims

(Butts and Snyder, 1992). Legislators attempt to restore the victim's losses by restorative

justice, community service, or time in jail. Unfortunately, at times, these programs do not

advocate the steps toward crime-free lives. Empirical research has tried to identify the

risk factors that characterize the juvenile crime activity but research has been low to

materialize on this front, especially across race.

But is the recent offending increase equal for all ethnic-races? Among Whites and

African Americans, the latter have higher crime rates than Whites (Jackson, 1997). Non-

whites commit not only more crime, but also commit more serious crime than other

groups (Jackson, 1997), and this is especially the case among African Americans.

Furthermore, juvenile offenders with extensive prior records appear to have a high









probability of reoffending (Visher, Lattimore and Linster, 1991), and this is especially the

case among minorities.

Recidivism

In the study of recidivism, research on criminal careers is now considered the state

of the art (Farrington, 1992). A criminal career is a sequence of offenses during a period

of an individual's life, meaning that the criminal behavior is repeated several times and it

is not an isolated incident. This repetitive criminal behavior is called recidivism and

indicates the proportion of the population that becomes involved in criminal behavior, at

what age the criminal behavior begins, how long the criminal career lasts, and the number

of offenses typically committed during the course of the career (Farrington, 1992). It is

also defined as an estimate of the percentage of released prisoners who commit another

offense. Estimates of recidivism vary with the length of the follow up period and the

measures selected (see Bureau of Justice Statistics, 1994). For example, in a recidivism

study in Kentucky (Kentucky Department of Corrections, 2000), approximately 33% of

the total of the inmates released in 1995 returned to prison within two years.

The ideal approach for studying recidivism requires a longitudinal data (a follow-

up of a period of time). Taxman and Piquero's (1998) analysis of drunk-driving

recidivism conceptualized the dependent variable, recidivism, as whether or not

recidivism occurred within the follow-up period, and the time until recidivism measured

in days. Similarly, Lee (1980), conceptualized recidivism as a dynamic social process in

which the time until recidivism, conceived as a failure or rearrest, is the dependent

variable. Policymakers also consider recidivism an important measure of the corrections

system's performance (Florida Department of Corrections, 2001).









According to the Bureau of Justice Statistics (1994), there are three different

measures of recidivism: rearrest, reconviction and reincarceration. Rearrest refers to any

arrest for a felony or serious misdemeanor that was reported to the state identification

bureau or the FBI after release from a correctional institute. It also can be understood as

any counts of contact between the offender and the justice system without regard to

whether the contact results in the offender actually re-entering the jurisdiction of the

justice system. Reconviction refers to a conviction on at least one charge after the date of

release from prison. Reincarceration refers to any return to prison or any admission to a

local jail with a sentence for a new offense. The Texas Youth Commission reported in the

National Recidivism Methods Study (1997), that 27 states use the fallowing recidivism

measures:

* Rearrest, 6 states, 22%;
* Reincarceration, 5 states; 19%;
* Readjudication; 3 states; 11%;
* Recommitment, 11 states; 41%.

Recidivism research has been used differently in adult and juvenile justice

populations; in adult populations it has primarily been employed by correctional policy

makers in order to create a probation or parole guideline based on determinations of

which individuals are at risk for future offenses (Quist and Matshazi, 2000).

Juvenile justice policy makers have used recidivism as an outcome measure in

rehabilitation programs. In theory, the purpose of the juvenile justice system is to avoid

incarceration of children and adolescents and emphasize rehabilitation (Bartol and Bartol,

1994). The population has directed criticism at the juvenile justice system because its

courts are overcrowded, recidivism rates are high, and there is a lack of system efficacy

(Harrison, Maupin and Mays, 2001). Moreover, repeated juvenile delinquency can lead to









career paths in criminal activity in adulthood. Therefore, researchers explain the use of

two types of actions to prevent this criminal activity: active sanctions that are class or

seminars, community service assignments, participation in educational programs, and

passive sanctions, verbal reprimands, written warning, and disciplinary probation

(Kompalla and McCarthy, 2001), in order to prevent reincarceration. Furthermore,

Olshak (1999) argued that active sanctions in particular are an alternative for reducing

recidivism.

Several studies have demonstrated that high-risk offenders can be identified by

gathering information on a youth and parent interviews. However, gathering this

information to identify juveniles as chronic offenders is limited because it is too costly to

do so. Consequently researchers argue that looking at second-time offenders is more

plausible to identify them as chronic offenders because they represent a small proportion

of all juveniles brought to court, they recidivate more frequently, and likely, become

chronic delinquents (Smith and Aloisi, 1999).

Theoretical Analysis

Theoretical accounts have been applied to understand recidivism. Two theories will

be discussed in this section

Hirschi's Social Bonding Theory

For most criminologists, Hirschi's social bonding theory is referred to as control

theory (Akers, 1999; Barlow and Kauzlarich, 2002). This theory proposes that the

"delinquent acts result when an individual's bond to society is weak or broken" (Hirschi,

1969:16). The assumption of the theory is that deviant behavior is lessened when people

have strong bonds with other members of society. It is an individual's ties to social

institutions that inhibit or control them from committing deviant acts.









The four main elements of this theory are: attachment, commitment, involvement

and belief. Hirschi argues that the stronger these elements with parents, neighbors, adults

and peers are, the more the individual's behavior is controlled and maintained in

conformity, and the less likely the opportunity for the individual to commit crime. On the

other hand, the weaker these elements are, the weaker the social bonds are, and the more

likely the individual will commit crime. Another important aspect of this theory is that

these elements are inter-correlated, meaning, the weakening of one will probably weaken

the others (Akers, 1999; Yablonsky, 2000).

Attachment refers to the close affection ties to others, admiration of them and

identification with them. According to this element, if individuals are insensitive to

others' opinion, the more likely the individual is going to commit a crime, because they

are less compelled by the norms (Akers, 1999). Hirschi argues that attachment to

individuals such as parents is important in controlling delinquent tendencies and

maintaining conformity. The important aspect is the attachment to people, not necessarily

the character of the person that the individual is attached to: "The more one respects or

admires one's friends, the less likely one is to commit delinquent acts" (1969:152).

Commitment refers to the extent to which individuals have built up an investment

in conventionality and can be jeopardized or lost if they engage in deviant behavior.

People develop a stake in their education, reputation or building careers investments such

as a job or family are in line according with this element. When people are faced with the

decision to commit a criminal act, they weigh what they can lose and what they can gain

if they commit a crime. Commitment refers to a more or less rational element in the

decision to commit a crime.









Involvement refers to an individual's involvement in conventional activities, e.g.

leisure activities, spend time with parents, friends, athletic clubs, and participation in

legitimate extracurricular activities. The more the individuals' involvement in

conventional activities, the less likely they are believed to get in trouble or commit crime.

If the individual is busy doing other conventional things, they will not have time to

commit crime.

Belief refers to the importance of obeying the rules and the belief of legitimacy of

conventional values and norms. Hirschi argues that this element does not refer to the

belief about specific norms or laws; it is just the belief of general norms or laws. Hirschi

argues that the less a person believes the individual should obey the rules, the more likely

the individual is to violate them.

Hirschi (1969) tested his control theory with a self-report survey administered to

4,000 high school students in California. He found that youth with more attachment to

parents were less likely to show deviant behavior. In regard of commitment, Hirschi

found that commitment to conventional values was related to good behavior. However,

conventional activities lead to some increase in crime.

Agnew (1985) explored this theory in a study of almost 1,900 male youths. The

juveniles were first interviewed in the tenth grade and then again in the eleventh grade.

For involvement, he found it to be stable over the two-year period. However, attachment

to parents, school grade and commitment explained only 2% of the variation in

delinquency. Agnew described this finding as the result of the maturation of the children

and therefore the importance of the bonds may become diminished overtime. In another

study, the attachment between the child and the parent was found to not be correlated









with delinquency, and the involvement in school-related activities and delinquency were

negatively related (Greenberg, 1999).

Moffitt's Developmental Theory

According to Moffitt (1993), rates of prevalence and incidence of offending is

higher in adolescence such that by age 17, we observe the highest rates of criminality at

which point it drops when the adolescent reaches the adulthood. However, this does not

mean that criminal behavior starts a few years before age of 17; instead for some

offenders criminal activity emerges very early in life, age 7. Furthermore, the career

activity of an adolescent changes over time. There is a group of adolescents that engage

in antisocial behavior at every stage of the life-course, life-course-persistent, and a larger

group of offenders who follow the well-known age-crime curve, or adolescence-limited.

Among those persistent offenders, a small percentage of them are frequent crime

offenders, meaning that the most persistent offenders are responsible for about 50% of

known crimes. These high-rate offenders begin their criminal activity earlier and continue

through time, have been arrested by police in the early teen years, which can be seen as

an important variable in predicting long-term recidivistic offending (Moffitt, 1993).

In her developmental taxonomy, Moffitt presents some of the causes of individuals'

antisocial behavior. She starts by pointing out factors that can be present at individuals'

birth or before birth. One of the common factors is the neuropsychological

("[A]natomical structures and psychological processes within the nervous system [which]

influence psychological characteristics such as temperament, behavioral development,

cognitive abilities, or all three", Moffitt, pp. 677) function of the infant; this means that

the neural development can be disrupted by the use of drugs or poor nutrition before

birth. In addition internal factors can also contribute to neuropsychological risk, for









example child abuse and neglect. This may explain who becomes a frequent offender and

who does not. Moffitt also argues that verbal and executive functions are correlated with

antisocial behavior of life-course-persistent offenders. In this sense, she identifies

learning disabilities that cause problems with the expressive speech, writing skills, and

reading skills among others. These learning disabilities are also caused by parental

heritance or lack of remedial cognitive stimulation provided by the parents to their

children's below average cognitive capacities.

In sum, Moffitt argues that criminal activity is caused by cutting off opportunities

to practice prosocial behavior, due to discipline problems and academic failure and could

lead to a life-course-persistent offending. On the other hand, those youth with lees

likelihood of following life-course-persistent trajectories encounter motivation for crime

when they enter to adolescence; during that period they mimic other's behavior, but

eventually they enter into adulthood and desist from crime.

Analysis of the Factors that are Related to Recidivism among Juvenile Offenders

Research has identified several factors that contribute to recidivism. Fendrich and

Archer (1991) suggested that factors expected to influence recidivism among juvenile

offenders include social and demographic characteristics such as age, commitment, race

and gender, intellectual functioning, i.e., IQ achievement test scores (Duncan et al., 1995;

Carcach and Leverett, 1999), family functioning and environment, prior criminal history,

alcohol and drug abuse, and family background (Visher et al., 1991). Also, prior research

consistently has shown that most recidivism occurs fairly swiftly following release

(Winner et al., 1997).

When time served in institutions was introduced to predict recidivism, in a study

among youth serious offenders from the California Youth Authority in 1981, Visher,









Lattimore, and Linster (1991) found that time served contributed significantly to

recidivism, and was statistically significant only for those who had not recidivated in the

first 36 weeks following release. In a similar study, Schmidt and Witte (1988) found that

longer incarcerations were related to a greater risk of recidivism, as measured by time to

re-incarceration. The same result was found in a study conducted by Myner, Santman,

Cappelletty and Perlmutter (1998). In a sample of juveniles, these authors argued that the

length of first incarceration was associated with a greater likelihood of recidivism (Myner

et al., 1998).

Cognitive Ability

Cognitive ability refers to abstract thinking ability and IQ. Researchers argue that

lower cognitive ability results in impulsivity, attention deficits and lack of abstract

thinking (Moffitt, 1993). Furthermore, Moffitt (1993) argues that differences in cognitive

ability exist between people that engage in criminal activity during their adolescence and

those who engage in life-long criminal activity. On the other hand, individuals with

higher cognitive ability are more likely to desist from illegal activities than those with

lower cognitive ability. In sum, those with more significant academic deficits are more

likely to commit a delinquent act again in their life.

Recidivism has been found to be higher for persons who had not completed high

school than among high school graduates. Delinquent behavior has been found to be

associated with poor academic performance (Farrington, 1987) and school attendance is

also predictive of recidivism (Cymbalisty et al., 1975). Consistent with Farrington's

findings (1987), Myner et al. (1998) found, in a study among male juveniles convicted of

criminal offenses in California, that lower GPA's are related to reoffending. Researchers

reported that men who had a lower IQ scores were at greater risk for criminal activity









than those who had higher scores. Individuals with lower IQ scores reported more

delinquent acts than individuals with higher IQ scores. In another study, Haberman and

Quinn (1986) found that in a sample of 759 previously incarcerated juvenile offenders

only about 12% completed high school after returning to the community.

Age

When looking at age, studies have related the age at first arrest to recidivism.

Furthermore, this relationship seems to be curvilinear, i.e. recidivism increases in the

early teenage years and decreases after about age sixteen (Fendrich, Archer and Kennedy,

1998; Gruenewald and West, 1989). In a study by Lanza-Kaduce, Gartin, Dundes,

Blount, Boulinois-Manning, Kuhns, Holten, Mahan, Taylor, Terry, Frank and Godshall

(1990), they found that older inmates reoffended less frequently. The Bureau of Justice

Statistics (1987a) found similar results showing that recidivism was inversely related to

the age of the prisoner at time of release: the older the prisoner the lower the rate of

recidivism. The same study documented that more than 75% of those age 17 or younger

when released from prison were rearrested, compared with 40.3% of those age 45 and

older. The peak age of onset (the beginning) of a criminal career was 14 (4.6% of first

convictions), with a second peak at 17 (4.4%). The Kentucky Department of Corrections

(2000) found similar results in that young offenders recidivated at higher rates than older

individuals. In a sample of 5,509 Australian juvenile offenders, Carcach and Leverett

(1999), found that juveniles between the ages of 14 and 17 years recidivate the most

compared with the older ones. The authors suggest that this pattern is caused by the

influence of delinquent peers.

Moreover, those individuals who were first convicted at the earliest age (10 to 13)

have been found to be the most persistent offenders. The average duration of criminal









careers dropped sharply from those first convicted between the ages of 14 and 16 (8.2

years) to those first convicted between 17 and 20 (2.7 years). This finding suggests that

males first convicted as juveniles were much more persistent offenders than those first

convicted as adults (Farrington and Hawkins, 1991). In another study, the Florida

Department of Corrections (2001) found that the average inmate's probability of

reoffending drops by 2.1% for each year older the inmate is at release and that younger

offenders reoffend more than older offenders. Overall, more than three quarters (78.3%)

of those convicted as juveniles were also convicted as adults, compared with less than a

quarter (21%) of those not convicted as juveniles.

Seriousness of Offense

Several studies have documented that high rates of juvenile recidivism are found

among chronic juvenile offenders and those who commit serious crimes (Bureau of

Justice Statistics, 1987a). Recidivism is higher for those who have committed serious

offenses (Kentucky Department of Corrections, 2000). Gruenewald and West (1989), in

their study of recidivism among juvenile offenders, found 73% recidivism for serious and

chronic offenders. Another study by Visher, Lattimore, and Linster (1991) found that

chronic juvenile offenders released from California Youth Authority have an 88%

recidivism rate. Duncan, Kennedy and Patrick (1995), in a study of 129 juveniles released

from a training school in Florida, found that 52% of the juvenile offenders were

rearrested within 6 months of their release. Similarly, the Bureau of Justice Statistics

(1987a) found that 69% of juvenile delinquent parolees were rearrested for serious crimes

within 6 years of their release.

However, in a study of recidivism of 396 prison releases over the 12 months period

following their dates of release from confinement, Lanza-Kaduce, Parker and Thomas









(1999) found that sex offenders and violent offenders were less likely to re-offend.

Recidivism is also more likely among property offenders than among violent offenders,

which is likely to be due to the differences between parolees from private and public

prisons (see also Craig and Budd, 1967). Similar results were obtained by the Florida

Department of Corrections (2001) whose study of male parolees uncovered that released

inmates whose primary offense was violent, were 32% less likely to reoffend. Carcach

and Leverett (1999), found that juveniles dealt in court for violent and property offenses

had similar levels of recidivism, and the most interesting finding was that these two

groups of offenders, i.e. violent and property offenders, had higher rates of recidivism

compared with those charged with drug offenses.

Alcohol and Drug Abuse

A drug is any chemical or chemicals that change the body function, and may alter

its biological structure (Bureau of Justice Statistics, 1994). Alcohol and drug abuse is a

global phenomenon. It has been said that the use of alcohol may increase hostile behavior

(Reiss and Roth, 1993). Many of the crimes committed have their roots in alcohol and

drug addictions, and this has lead to an increase in the United States' prison population.

Most directly, it is a crime to use, possess, manufacture or distribute drugs classified as

having a potential for abuse (Bureau of Justice Statistics, 1994).

Several studies have shown that alcohol and drug abuse predicts recidivism and it is

a contributor to criminal behavior (Myner at al., 1998, Wierson and Forehand, 1995).

Evidence of the association between drugs, alcohol, and violence has been well

documented. For example, in a report by the Bureau of Justice Statistics (1993), almost

50% of prison inmates had used illegal drugs in the month before the offense for which

they were incarcerated and about 30% were under the influence of drugs at the time of









the crime. In sum, empirical research has consistently shown a relationship between

alcohol use and aggressive behavior (Parker and Auerhahn, 1998; Taylor, 1983).

In another study, Taxman and Piquero (1998) showed that alcohol use contributes

to criminal offenses with a sample of Maryland offenders. Furthermore, Ge et al. (2001)

argued that adolescents who smoke, drink and use illicit drugs are more likely to commit

delinquent acts compared to those who do not smoke, drink or use illicit drugs. In a

sample of 27,414 arrestees, Martin et al., (2001) found that almost 95% of the sample

reported that they have tried alcohol at some point during their lives; approximately 52%

of the sample had also consumed alcohol 72 hours prior to their arrest and almost 21%

reported the use of alcohol at the time of their offenses.

Family Background

Family background can be defined as the sum of positive or negative factors, which

include providing for basic needs, attention, marital status of the parents, sexual,

psychological abuse, domestic violence between parents and employment of parents

among others (Heilbrun et al., 2000). Adolescents that experience these stressors are

considered "at risk" meaning that juveniles with these risk factors begin a trajectory of

criminality, addiction and dependency early in life (Carr and Vandiver, 2001). Cohesion

and good communication between parents and children is strongly associated with good

adaptation.

Analysis of youth's family characteristics such as criminal behavior of parents or

siblings, poor parenting, family conflict and disruption in family structure are

characteristic of serious juvenile offenders (Visher et al., 1991). One longitudinal study

showed that delinquency is associated with parent separation, implying that the marital

status of the parents lead to juvenile delinquency (Patterson et al., 1989). Patterson et al.









has also argued that the lack of family supervision is an important determinant of early

onset delinquency, and research supports this hypothesis (Patterson et al., 1992).

Race

Regarding race, research suggests white releases are less likely to reoffend than

are releases from other racial groups (see Lanza-Kaduce, Parker and Thomas, 1999;

Bureau of Justice Statistics 1988, 1992). The recidivism rate is higher among African

Americans and Hispanics, and it is increasing steadily (Kentucky Department of

Corrections, 2000), with non-whites having slightly higher recidivism rates than whites,

approximately 5 to 8% higher for each measure. Individuals of Hispanic-origin also had

recidivism rates that were about 6% higher than those among non-Hispanics (Bureau of

Justice Statistics, 1987b). On average, African Americans recidivate 43% more than non-

African Americans. In sum, African Americans appear to recidivate at higher rates than

White Americans (Florida Department of Corrections, 2001).

The Bureau of Justice Statistics (1987a), in a report about recidivism of young

parolees, found that recidivism rates were higher among African Americans: Hispanics

had recidivism rates that were 5 to 12% less than African Americans, but higher rates

than White Americans. In another study, Ge et al. (2001) found that African American

juveniles were arrested more frequently than were [White Americans] European

Americans. Similar results were found for Hispanics having more arrests than European

Americans.

Current Focus

The primary purpose of this study is to identify the risk factors associated with: 1)

White Americans; 2) African Americans; and 3) Hispanics in terms of their ability to

predict recidivism. Risk factors analyzed represented information on demographic,









familial, and behavioral characteristics available in records at the California Youth

Authority (CYA) and Wenk's (1990) data set. The present study also explored serious,

and violent crimes committed by juveniles, and the factors that could be related to

recidivism, age of juvenile offenders, cognitive abilities, alcohol and drug use, and

marital status of parents and the juvenile offenders in attempt to add to the limited

number of studies in the recidivism literature, especially across race. In order to achieve

this goal, survival analysis will be used as the method of analysis.

This study also uses two theories as a theoretical backdrop: Social Bonding Theory

(Hirschi, 1969) and Developmental Theory (Moffitt, 1993). According to the former,

delinquency takes place when social bonds are weak. In this study, marriage is used as

the indicator of social bonds and it is expected to be negatively correlated with

recidivism. Marital status of the parolees could be correlated with Hirschi's Social

Bonding Theory elements of achievement and commitment. The former refers to close

affection ties to others and the latter to the investment that individuals have built up and

can be jeopardized if they engage in criminal behavior. Parolees' marital status could

inhibit their antisocial behavior as a result the close ties with their wife and because they

do not want to loose their family. The higher the attachment and commitment of the

parolees with their wife, the lower the likelihood of reoffending, and the lower the

attachment and commitment, the higher their antisocial behavior after being release.

Moffitt's Developmental theory predicts that serious and persistent antisocial

behavior is caused by learning disabilities, among other influences. Some of the variables

that I am going to use to predict recidivism across race have been mentioned by these two

theories. Cognitive ability is one of the recidivism predictors that is going to be used in









the model, and according to Moffitt's theory, lower cognitive ability results in higher

antisocial behavior and therefore in higher rates of criminal activity. In this sense, the

lower the cognitive ability of the juvenile offenders, the higher the likelihood of

recidivism, and the higher their cognitive ability, the lower the likelihood of recidivism.

The scarcity of studies that examines the predictive power of recidivism across race

makes this study a contribution to the literature. The importance of the topic lays in the

lack of information and research about recidivism and the three major ethnic groups.

Looking for answers about recidivism in the United States makes this topic relevant and a

unique contribution.
















CHAPTER 3
DATA

The data used here is part of a larger study designed to examine the criminal career

patterns of violent juvenile offenders (Wenk, 1990). The data contain a total sample of

4,146 juvenile male offenders who were committed to the Reception Guidance Center at

the Duel Vocational Institution in Tracy, California between 1964 and 1965. This

institution was a facility operated by the California Youth Authority (CYA), which

provides a setting for training, treatment, and education of serious, juvenile offenders.

Two studies have used Wenk's data to predict recidivism using different sets of

variables. Donnellan, et al. (2000) and Ge et al. (2001) conducted both of the studies.

Their first approach looked at the cognitive abilities among adolescent-limited and life-

course persistent offenders among White Americans, African Americans and Hispanics.

They argued that delinquent juveniles have lower scores on test of cognitive abilities. In

order to gather information on adolescents' cognitive abilities, Ge et al. looked at

different tests; the California Achievement Test (CAT), California Test of Mental

Maturity (CTTM), General Aptitude Test Battery (GATB) and Raven Test of Progressive

Matrices.

Tiegs and Clark (1951) designed the CAT or California Achievement Test that was

used in California schools to determine academic achievement. It gauges information on

academic domains of reading, arithmetic, language and overall academic achievement.

The CTTM is a test designed by Sullivan et al., (1963) to test language and non-language









intelligence. The general Aptitude Test Battery was designed by the U.S. Department of

Labor. Its diverse test includes verbal, arithmetic, spatial, perceptual and general

intellectual abilities. Finally, the Raven Test of Progressive Matrices was developed by

Raven (1960) and measures general intelligence abilities. The researchers found that for

White Americans, juvenile offenders had significantly higher scores in the cognitive tests

than life-course-persistent offenders. Similar results were found for Hispanics; however,

for African Americans, the results were opposite showing that there were no significant

differences between adolescent-limited offenders and life-course-persistent offenders on

tests of cognitive ability.

The second study conducted by Ge et al., a year after their first one (2001) looked

at 2,263 juvenile offenders. Variables such as age at first arrest, family environment,

cognitive ability, and early behavioral problems were measured to predict the persistence

of criminal offending into adulthood among males. Looking first at the cognitive ability

analyses, the researchers found that cognitive ability was not significantly related to

juvenile delinquency. In a second analysis, they turned to the question of the prediction of

chronic offending; in this case, cognitive ability was a significant predictor of persistent

offending after the offenders turned 18. When the researchers looked at alcohol and drug

abuse, they found that its use influenced the amount of delinquency but not the timing of

delinquency. In sum, Ge et al. found that low scores on cognitive ability were

significantly associated with frequencies of arrest. Furthermore, involvement with

alcohol and drug use was a predictor of adult arrest, suggesting that alcohol and drug

abuse influence the criminal behavior of the juveniles. Importantly, neither of these two









studies examined the inter-relationships between race and cognitive abilities within a

survival framework.

Variables

Independent Variables

Six independent variables will be used from Wenk's data in this research. The first

variable, California Achievement Test, is used to determine academic achievement; it

gauges information on academic domains of reading, arithmetic, language and overall

academic achievement. This variable measures the scores of the California achievement

total grade placement of the juvenile offenders.

Marital status of the parolee is the second variable. This variable measures the

marital status of the juvenile offenders. This variable will be recorded into two variables:

not married (0) and married (1).

The third variable that will be used in the research is marital status of natural

parents. The variable will be coded into not married (0) and married (1).

Alcohol and drug use are the fourth and fifth variables to be incorporated in the

research. History of alcohol misuse measures either if the juvenile offenders have had

moderate problems or severe problems with alcohol in general, which means, if they

consumed alcohol on a daily basis. History of drug use dealt with use in general, and it

measures if the juvenile offenders have used drugs on a daily basis.

Age at reception is the sixth variable and measures the age of the juvenile offenders

at reception. The mean for the juvenile offenders age at reception is 19.0 with a standard

deviation of .97 (range: 16-24).









Dependent Variables

The dependent variables that will be used in the research are as follows: days to

parole suspension measures the days to which the juvenile offenders have committed

another crime and been suspended from parole. The variable parole outcome success

measures if the juvenile offender, after being released from jail, has or has not recidivated

(0=failure, successs.

Research Questions

There are two main questions to be answered in this research:

1. Are the survival recidivism patterns the same for White Americans, African
Americans and Hispanics? And

2. Do the independent variables relating to recidivism differ among White
Americans, African Americans and Hispanics?

Analytic Plan

As mentioned above, the survival analysis method will be used to examine the data.

To analyze the timing of recidivism, Cox proportional hazard regression (Cox, 1972), a

type of survival method, will be used. The method seeks to explain the differences in

terms of both of the following criteria whether a reconviction occurred (i.e., failure), and

the period of time until the reconviction. The Cox proportional hazard regression

facilitates analysis because it tests the dependent variables and its effects separately while

controlling for the other variables and it is a method for modeling time-to-event data en

the presence of censored cases. The hazard model seeks to focus on the patterns of

recidivism exhibited in the sample until data censoring. The hazard model provides the

probability that a subject will fail in a period of time. It also provides more information

than typical recidivism measures because it allows for un-covering heterogeneity among

subjects with respect to the risk of recidivism. Importantly, it also does not make






25


assumptions about the distribution of the underlying heterogeneity, and it identifies

covariates that tend to be associated with the time to failure (Visher et al., 1991).

















CHAPTER 4
FINDINGS

Results for One-Way ANOVA

One-way analysis of variance (ANOVA) was performed on the test of the

independent and dependent variables separately for each ethnic group. The idea was to

conduct separate analyses for each ethnic group by controlling for the impact of ethnicity

on test scores. Table 1 provides the results of the analysis for White Americans,

Hispanics, and African Americans. Although Whites Americans, African Americans and

Hispanics did not differ on parole outcome/success, there were mean differences on time

to failure. For example, White Americans were more likely to fail sooner, in terms of

days, than were African Americans by approximately forty days. On the other hand,

Hispanics were less likely to fail sooner than African Americans by approximately one

day. As shown in Table 1, Cognitive Ability was significantly different across ethnic

groups, p< .000. White Americans had higher cognitive abilities than African Americans

and Hispanics. History of Alcohol Misuse and History of Drug Misuse was also

significantly different across ethnic groups, p< .000. This pattern suggests that Hispanics

were more likely to be involved in the use of alcohol and drugs than White Americans

and African Americans. However, African Americans were less likely to be involved in

the use of alcohol and drugs than White Americans.









Table 1 Descriptive Statistics: One-Way ANOVA analysis.
Variable White African-American Hispanic Sig

M SD M SD M SD

Cognitive A. 6.63 12.79 5.15 11.69 4.9 13.23 w>b>h*

Marital S. Parolee .13 .34 .12 .33 .16 .36 yes***

Marital S. Parents .48 .50 .49 .50 .56 .50 yes*

History Alcohol .56 .72 .43 .63 .94 .79 h>w>b*

History Drug .23 .63 .19 .55 .45 .82 h>w>b*

Parole (l=succ) .61 .48 .61 .49 .61 .49 no

Days parole susp. 251 168.93 294 185.13 295 199.66 w>b>h*

Notes: *p< .000, **p< .05, ***p< .10

Results of the Survival Analysis

The outcome measures discussed focus on whether the juvenile offender has

recidivated or not. The way to operationalize recidivism is by examining the length of the

time until the new arrest. This measure of recidivism allows the researcher to examine

desistance from criminal behavior and to explore the differences across race. To analyze

the timing of recidivism, I used Cox proportional hazard regression, which is a method

for modeling time-to-event data in the presence of censored cases.

The results of the analyses are presented in Figure 1 (Survival function). Figure 1

presents the survival function, taking into account the variable days to parole suspension

and parole outcome success. Juvenile offenders recidivate the most approximately after

480 days of being released. We can observe this same result if we look at the hazard rate,

which is the estimate of the probability per unit time that a case that has survived to the

beginning of an interval will experience an event in that interval. The hazard rate gets









higher approximately at days 490 and 504. The median survival time for the juvenile

offender data is 239.03 days.


Survival Function
1.2


1.0


.8-


.6


.4


-F .2


(0 0.0
E
8 -.2
-200 200 400 600 800 1000

DAYS TO PAROLE SUSPENSION

Figure 1 Survival Function

Table 2 contains the results of the Cox proportional hazard regression for the

sample of Hispanics. The first column is B (slope coefficient); the second column is the

SE (standard error), the third is the Wald statistic (this statistic test the hypothesis that B

equals 0 in the population). The negative coefficient means that a higher value on the

independent variable leads to a lower likelihood of re-arrest, i.e. the effect of the variable

is such that it increases the time until parole failure. For Table 1, Hispanics were less

likely than non-Hispanics to incur parole failure. In other words, Hispanics had a longer

time to failure than non-Hispanics.









Table 2 Cox Proportional Hazard Models Predicting Hispanic Juveniles Parole
Success/Failure
Variable B SE Wald


Race (1 = Hispanic) -.222 .075 8.872

Cognitive A 000 .003 .002

Marital S. Parolee -.035 .089 .154

Marital S. Parents -.003 .058 .002

History Alcohol .027 .040 .457

History Drug .020 .041 .022

Age par at reception -.002 .032 .002
Notes: Log-Likelihood = 14619.15; X2= 9.180; df= 7; p= .220
*p< .05

Table 3 contains the Cox Regression for the African American sample. The results

suggest that African Americans were less likely than non-blacks to incur parole failure. In

other words, African Americans had a longer time to failure than non-blacks. Table 4

provides the results for the Cox Regression pertaining to the sample of White Americans

suggesting that White Americans were more likely than non-White Americans to incur in

parole failure. The White American juvenile offenders had a lesser time to failure than

non-White Americans.

The results for the split-race sample, which are presented in Table 5, reveal that

among White Americans none of the independent variables are significantly related to

recidivism. The same results were found for Hispanics. In the case of the African

American juvenile offenders, all but one variable was insignificant, p< .05. The increase






30


in drug use increases the probability of parole failure for the sample of African American

juvenile offenders.

Table 3 Cox Proportional Hazard Models Predicting African American Juveniles Parole
Success/Failure
Variable B SE Wald


Race (1 = African -.159 .066 5.755*
A.)

Cognitive A .001 .003 .038

Marital S. Parolee -.052 .089 .338

Marital S. Parents -.012 .058 .041

History Alcohol -.013 .040 .106

History Drug -.007 .041 .032

Age par at reception .007 .032 .042
Notes: Log-Likelihood = 14613.00; X2= 6.044; df= 7; p= .523
*p<.10









Table 4 Cox Proportional Hazard Models Predicting White Americans Juveniles Parole
Success/Failure
Variable B SE Wald

R~o / =Wht1 J;+, 74A n0Q 21.83"


Cognitive A

Marital S. Parolee


Marital S. Parents


History Alcohol


History Drug


.000


-.063


.000


.007


.006


.003

.089


.058


.039


.040


Age par at reception .002 .032
Notes: Log-Likelihood = 14597.05; X2= 22.23; df=
*p <.05


.005

.509


.000



.027

.023


.005


7; p= .002


Table 5 Cox Regression Hazard Model: Split-Race Analysis
Variable White Hispanic

B SE Wald B SE

Cognitive A .004 .004 .834 -.008 .007

Marital S. Parolee -.197 .123 2.562 .120 .185

Marital S. Parents .063 .082 .581 -.214 .131

History Alcohol -.027 .055 .237 .143 .088

History Drug -.011 .056 .036 -.060 .075

Age par at reception -.034 .044 .583 .096 .070

Notes: *p < .05


Aald

1.41

423

2.68

2.60

640

1.86


African-American

B SE Wald

-.001 .006 .055

.089 .185 .230

.051 .110 .215

-.009 .085 .011

.216 .099 4.76*

-.009 .062 .021















CHAPTER 5
DISCUSSION AND CONCLUSIONS

The main purpose of this study was to identify the risk factors associated with

White Americans, African Americans and Hispanics in predicting recidivism. This study

is the first that I am aware of to compare three different ethnic races in predicting

recidivism on a series of independent variables on a model of life-course criminal

offending. The data was collected by Wenk (1990), and presents information on

demographic, familial and behavioral characteristics in records at the California Youth

Authority (CYQA). First, I examined if there was any difference across race in cognitive

ability, history of drug use, history of alcohol use, parole outcome and days until

reconviction. I then sought to study the influence of these correlates on parole

success/failure and days to parole suspension across race. Finally, I sought to demonstrate

in a split-race sample the correlation among the independent variables and the dependent

variables with each ethnic group.

I found differences on time to failure among White Americans, African Americans

and Hispanics, whereas I did not find any differences on parole outcome/success. In this

sample, White Americans were more likely to fail sooner than African Americans and

Hispanics. When cognitive ability was taken into account, White Americans were more

likely to have higher cognitive ability than their counter parts, but it was not related to

recidivism. I found that Hispanics were more likely to be involved in the use of alcohol

and illicit drugs than White Americans and African Americans. The latter were less likely









to be involved in its illicit use compared with White Americans. I conducted additional

analyses designed to predict recidivism across race. The results of these analyses revealed

that juvenile offenders recidivated the most approximately after 480 days. Finally,

looking at each ethnic group, I ran separate analyses to predict recidivism. The Cox

Proportional Hazard Regression results for the sample of Hispanics showed that

Hispanics were less likely to recidivate than non-Hispanics; for the sample of African

Americans, they were less likely to recidivate than non-African Americans; and for the

sample of White Americans, they were more likely to incur parole failure than non-White

Americans.

From this study, it is possible to assess different limitations. First, the records

obtained by Wenk (1990) were gathered in 1963, this could suggest that the records

obtained are outdated. Second, the data did not include information on offender's

previous alcohol and drug use, nor did it include information on cognitive ability before

release. Third, the data set only included information only for male offenders.

The findings do not support any of the two theories used in this analysis. Both

Social Bonding Theory (Hirschi, 1969) and Developmental Theory (Moffitt, 1993) did

not show any significant relation between the variables. For example, marital status of the

parolee was expected to be correlated with recidivism. However, the results showed non-

significant effects. For Moffitt's Developmental Theory, cognitive ability was believed to

predict recidivism. In this study, White Americans had higher cognitive scores, but

cognitive abilities were not related to recidivism. On the other hand, Hispanics had lower

cognitive ability scores and were less likely to fail sooner than White Americans, but

cognitive abilities were not directly related to recidivism.









The "null" results from this study could be due to several factors. First, other

different variables, such as peer relationships, should be included in the study to predict

variation in the dependent variable. Second, the data set used in this study came from a

select sample of serious offenders, and it should be interesting to look at a sample of non-

serious offenders. The variables in the data set predicted continuing criminal activity.

Finally, the juvenile criminal records were composed only from official records and

further research should look at self-report records as well.

Despite the limitations, these results could be considered fundamental for the study

of recidivism across race. The scarcity of studies examining differences in recidivism

across race makes this study an important contribution to the literature. It appears that in

this particular sample cognitive ability, history of alcohol and drug use was not related to

recidivism in neither of the ethnic groups, although differences were found across race in

the timing of recidivism. Further research should continue studying issues related to

recidivism as well as obtain different factors from those studied in this study to determine

how various risk factors relate to recidivism generally, and across race in particular.















LIST OF REFERENCES

Agnew, R. (1985). Social Control Theory and Delinquency. Criminology 23:47-60.

Agnew, R. (1999). A General Strain Theory of Community Differences in Crime Rates.
Journal ofResearch in Crime and Delinquency, Vol. 36(2): 23-30.

Akers, R. (1999). Criminological Theories: Introduction and Evaluations. Chicago, IL:
Fitzroy Dearborn..

Barlow, H., and Kauzlarich, D. (2002). Introduction to Criminology. Upper Saddle River,
New Jersey.

Baron, S. (May 2001). Street Youth Labour Market Experiences and Crime. The
Canadian Review of Sociology and Anthropology Vol. 38(2): 189-215.

Bartol, C., and Bartol, A. (1994). Psychology Law: Research and Application. Belmont,
CA.

Benda, B., and Tollett, C., (1999). A Study of Recidivism of Serious and Persistent
Offenders Among Adolescents. Journal of Criminal Justice, Vol. 27(2): 111-116.

Blumstein, A., and Wallman, J. (2000). The Crime Drop in America. New York, NY:
Cambridge University Press..

Bureau of Justice Statistics (1987a). Recidivism of Young Parolees. Special Report.
Washington D.C.: U.S. Department of Justice.

Bureau of Justice Statistics (1987b). Recidivism of Prisoners Released in 1983. Special
Report. Washington D.C.: U.S. Department of Justice.

Bureau of Justice Statistics (1988). Report for the Nation on Crime and Justice.
Washington D.C.: U.S. Department of Justice.

Bureau of Justice Statistics (1992). Pretrial Release of Felony Defendants, 1990.
Washington D.C.: U.S. Department of Justice.

Bureau of Justice Statistics (1993). Survey of State Prison Inmates. Washington D.C.:
U.S. Department of Justice

Bureau of Justice Statistics (1994). Fact .\leet Drug-related Crime. Washington D.C.:
U.S. Department of Justice.









Butts, J., and Snyder, H. (1992). Restitution and Juvenile Recidivism. Washington D.C.:
U.S. Department of Justice.

Carcach, C., and Leverett, S. (1999). Recidivism among Juvenile Offenders: An Analysis
of Time to Reappearance in Court. Research and Public Policy Series. Australian
Institute of Criminology. Research and Public Policy Series, Vol. 17: 1-25.

Carr, M., and Vandiver, T., (2001). Risk and Protective Factors among Youth Offenders.
Adolescence, Vol. 36: 409-426.

Cox, D. (1972). Regression Models and Life Tables. Journal of the Royal Statistical
Society, Vol. 34: 187-200.

Craig, M., and Budd, L. (1967/0. The Juvenile Offender: Recidivism and Companions.
Crime andDelinquency, Vol. 13: 344-351.

Cymbalisty, B., Schuck, S., and Dubeck, J. (1975). Achievement Level, Institutional
Adjustment and Recidivism among Juvenile Delinquents. Journal of Community
Psychology, Vol. 3: 289-294.

Donnellan, M., Ge, X., and Wenk, E. (2000). Cognitive Abilities in Adolescent-limited
and Life-course-persistent Criminal Offenders. Journal ofAbnormal Psychology,
Vol. 109(3): 396-402.

Duncan, R., Kennedy, W., and Patrick, C. (1995). Four-factor Model of Recidivism in
Male Juvenile Offenders. Journal of Clinical Child Psychology, Vol. 24(3): 250-
257.

Farrington, D. (1987). Early Precursors of Early Offending. In J. Q. Wilson and G. C.
Loury (Eds.). From Children to Citizens, Vol. 3. Families, schools and
Delinquency Prevention, pp. 27-50. New York, Springer-Verlag.

Farrington, D. (1992). Criminal Career Research in the United Kingdom. British Journal
of Criminology, Vol. 32: 521-536

Farrington, D., and Hawkins, D. (1991). Predicting Participation, Early Onset and Later
Persistence in Officially Recorded Offending. Criminal Behavior andMental
Health, Vol. 1: 1-33.

Federal Bureau of Investigation (1990). Crime in the United States, 1989. Washington
D.C.; US Department of Justice.

Fendrich, M., and Archer, M. (1998). Long-term Rearrest Rates in a Sample of
Adjudicated Delinquents: Evaluation the Impact of Alternative Programs. The
Prison Journal, Vol. 78(4): 360-389.

Florida Department of Corrections. Factors Affecting Recidivism Rates. May 2001.
Recidivism report. Tallahassee, Fl.









Fox J., Maureen A. (2002). FBA for Children and Youth with Emotional-behavioral
Disorders: Where we should go in the Twenty-first Century. Preventing School
Failure Vol. 44(4): 140-141.

Ge, X., Donnellan, M., and Wenk, E. (2001). The Development of Persistent Criminal
Offending in Males. Criminal Justice and Behavior, Vol. 28(6):731-755.

Greenberg, D. (1999). The Weak Strength of Social Control Theory. Crime and
Delinquency 45(1): 66-81

Gruenewald, P., and West, B. (1989). Survival Models of Recidivism among Juvenile
Delinquents. Journal of Quantitative Criminology, Vol. 5(3): 215-217.

Haberman, M. and Quinn, L. (1986). The High School Re-entry Myth: A Follow-up
Study of Juveniles Released from Two Correctional High Schools in Wisconsin.
Journal of Correctional Education, Vol. 37, 114-117.

Harrison, P., Maupin, J., and Mays, G. (2001) Teen Court: An Examination of Process
and Outcomes. Crime and Delinquency, Vol. 47(2): 243-264.

Heilbrun, K., Brock, W., Waite, D., Lanier, A., Schmidt, M., Witte, G., Keeney, M.,
Westendorf, M., Buinavert, L., and Shumate, M., (2000). Risk Factors for Juvenile
Criminal Recidivism. Criminal Justice and Behavior, Vol. 27(3): 275-291.

Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California Press.

Jackson, K. L. (1997) Differences in the Background and Criminal Justice Characteristics
of Young Black, White, and Hispanic Male Federal Prison Inmates. Journal of
Black Studies, Vol.27(4): 494-509.

Kaplan, H. (1984). Patterns of Juvenile Delinquency. Beverly Hills, CA: Sage
Publications.

Kentucky Department of Corrections (2000). Recidivism in Kentucky 1993-1995.
Executive Summary. Frankfort, KY.

Kompalla, S., and McCarthy, M. (2001). The Effect of Judicial Sanctions on Recidivism
and Retention. College Student Journal, Vol. 35(2): 223-231.

Lanza-Kaduce, L., Gartin, P., Dundes, L., William, R., Boulinois-Manning, S., Kuhns, J.,
Holten, G., Mahan, S., Taylor, D., Terry, C., Frank, H., and Godshall, B. (1990).
Prison Utilization Study: Risk Assessment Techniques and Florida's Inmates, Vol.
1. Application to Males and Females, Tallahassee, Fl; Division of Economic and
Demographic Research, Joint Legislative Management Committee of the Florida
Legislature.









Lanza-Kaduce, L., Parker, K., and Thomas, C. (1999). A Comparative Recidivism
Analysis of Releasees from Private and Public Prisons. Crime and Delinquency,
Vol. 45(1): 28-47.

Lee, E. (1980). Statistical methods for survival data analysis. Lifetime Learning
Publications.

Lerner, R. and Galambos, N. (1998). Adolescent Development: Challenges and
Opportunities for Research, Programs and Policies. Annual Review ofPsychology,
Vol. 49, 413-446.

Maine Department of Corrections (1998). Recidivism. Division of Juvenile Services.
Juvenile Recidivism Baseline Report. Charleston, ME.

Martin, S., Bryant, K., and Fitzgerald, N. (2001). Self-reported Alcohol Use and Abuse
by Arrestees in the 1998 Arrestee Drug Abuse Monitoring Program. Alcohol
Research and Healthy, Vol. 25(1): 72-79.

Minerbrook, S (1994). A Generation of Stone Killers. US News and World Report
(January), pp. 33-37.

Moffitt, T. (1993). Adolescence Limited and Life Course Persistent Antisocial Behavior:
A Developmental Taxonomy. Psychological Review, 100, 674-701.

Myner, J., Santman, J., Gordon, G.C., Permutter, B. (1998). Variables Related to
Recidivism Among Juvenile Offenders. International Journal of Offender Therapy
and Comparative Criminology, 42(1): 65-80.

National Institute of Justice (2000). Sentencing and Corrections: But they all Come Back:
Rethinking Prisoner Reentry. Washington D.C. US Department of Justice.

Office of Juvenile Justice and Delinquency Prevention (1992). Restitution and juvenile
recidivism. Special report. Washington D.C. US Department of Justice.

Office of Juvenile Justice and Delinquency Prevention (1996). Juvenile Offenders and
Victims. National Report. Washington, D.C.

Office of Juvenile Justice and Delinquency Prevention (1999a). Juvenile Offenders and
Victims. National Report. Washington, D.C.

Office of Juvenile Justice and Delinquency Prevention (1999b). Juvenile Offenders in
Residential Placement. Washington, D.C.

Olshak, R. (1999). A Guide for Effective Sanctioning: From Theory to Practice. Normal,
IL: Illinois State University.

Parker, R., and Auerhahn, K. (1998). Alcohol, Drugs and Violence. Annual Review of
Sociology, Vol. 24: 291-311.









Patterson, G., DeBaryshe, B., and Ramsey, E. (1989). A Developmental Perspective on
Antisocial Behavior. American Psychologist, Vol. 44: 329-335.

Patterson, G., Crosby, L., and Vuchinich, S. (1992). Predicting Risk of Early Police
Arrest. Journal of Quantitative Criminology, Vol. 8(4): 335-355.

Quist, R., and Matshazi, D. (2000). The Child and Adolescent Functional Assessment
Scale (CAFAS): A Dynamic Predictor of Juvenile Recidivism. Adolescence, Vol.
35(137): 181-192.

Raven, J. (1960). Guide to the Standard Progressive Matrices. United Kingdom, London:
H.K. Lewis.

Reiss, A., and Roth, J. (1993). Understanding and Preventing Violence. National
Research Council: Washington D.C.

Schmidt, P., and Witte, A. (1988). Predicting Recidivism Using Survival Models. New
York, NY: Springer-Verlag.

Smith, R., and Aloisi, M. (1999). Prediction of Recidivism among "Second Timers" in
the Juvenile System: Efficiency in Screening Chronic Offenders. American Journal
of Criminal Justice, Vol. 23(2): 201-222.

Snyder H., and Sickmund M. (1999). Juvenile Offenders and Victims: 1999 National
Report. Washington, DC: Office of Juvenile Justice and Delinquency Prevention.

Sullivan, E., Clark, W., and Tiegs, E. (1963). Examiners Manual for California Short-
form Test of Mental Maturity. California Test Bureau.

Taxman, F., and Piquero, A., (1998). On Preventing Drunk Driving Recidivism: An
Examination of Rehabilitation and Punishment Approaches. Journal of Criminal
Justice, Vol. 26(2): 129-143.

Taylor, S. (1983). Alcohol and Human Physical Aggression. In Alcohol, Drug Abuse,
Aggression. Springfield, IL.

Texas Youth Commission. National Recidivism Method Study (1997). Recidivism.
Austin, TX.

Tiegs, E., and Clark, W. (1951). California Achievement Test: Manuals of Directions for
Primary, Elementary, Intermediate, and Advanced Batteries. California Test
Bureau.

Visher, C., Lattimore, P., and Linster, R. (1991). Predicting Recidivism of Serious
Youthful Offenders Using Survival Analyses. Criminology Vol. 29:329-66









Wierson, M., and Forehand, R. (1995). Predicting Recidivism in Juvenile Delinquents:
the Role of Mental Health Diagnoses and the Qualification of Conclusions by Race.
Behavior Research Theory Vol. 33(1): 63-67.

William, J., Dilulio, Jr. J., and Walters. J. (1996) Body Count: Moral Poverty and How to
win America's War against Crime and Drugs. Simon & Schuster, New York, NY.

Winner, L., Lanza-Kaduce, L., Bishop, D., and Frazier, C. (1997). The Transfer of
Juveniles to Criminal Court: Re-examining Recidivism over the Long Term. Crime
and Delinquency, Vol. 43: 548-563.

Yablonsky, L. (2000). Juvenile Delinquency: Into the 21st Century. Belmont, CA:
Wadsworth.

Wenk, E. Criminal Careers, Criminal Violence, and Substance Abuse in California, 1963-
1983. Inter-University Consortium for Political and Social Research.















BIOGRAPHICAL SKETCH

With the constant support of my parents and brothers, I have aspired to seek a

position in the law field, and had achieved this goal in my country of origin, Bolivia. As

soon as I graduated from high school I began attending the Catholic Bolivian University

(Universidad Catolica Boliviana) pursuing the law degree. My interest in law and

especially criminology led me to continue the specialization in a foreign country. After

five years of continued education and my interest on pursuing my goals, I decided to

enroll at the prestigious University of Florida. I moved to Gainesville, Florida, with the

aim of getting a Master in Arts at the University of Florida. I am planning to continue my

education at the University of Florida and obtaining Ph.D. with a major in criminology

and law.