Review of the Roots of Youth Violence: Literature Reviews

Volume 5

Deterrence Strategies4


In the last century, Canadians have witnessed a definite shift in how the youth justice system responds to youth crime, gradually moving from the more welfare-oriented policies of the Juvenile Delinquents Act (1908–1984) towards the increasingly harsher and more punitive policies of the Young Offenders Act (1984–2003) and the current Youth Criminal Justice Act (2003–present). Most often, the justifications for turning to harsher, more severe punishments lay within the tenets of rational choice and deterrence based theories.

The aim of this paper is to provide a comprehensive review of the literature and empirical research regarding the use of more punitive sentencing and general deterrence practices towards youth offenders. While this paper does not encompass all of the literature available, it does offer a relatively comprehensive review of many pertinent studies conducted regarding these topics and the issues pertaining to them.

In the second section, a brief synopsis of deterrence and rational choice theories will be offered. Section three focuses on psychosocial maturity as it relates to juveniles in terms of responsibility, rational decision-making and deterrence in general. Section four addresses issues regarding the transferring of youth offenders into the adult criminal justice system and/or correctional institutions. Section five highlights both specific and general deterrence research, focusing on recidivism rates as well as youths’ perceptions, attitudes, and experiences regarding system involvement. In section six, a focus will be placed upon “shock” deterrence methods aimed towards youth, specifically focusing on the use of boot camps and Scared Straight programs. Section seven will offer a brief summation and conclusion regarding the topics discussed in the previous sections

Rational Choice and Deterrence Theories

Originating from classical criminology, rational choice theory posits that decisions are made and actions are taken by individuals in the rational exercise of free will. Specifically, it is suggested that prior to a criminal act being committed, individuals will take into account the probable legal penalties, the potential gains produced by the crime and the likelihood that they will be caught. Additionally, it is posited that their calculation is based on their own experience with criminal punishment as well as their knowledge of legal sanctions and punishments meted out in the past for offences. As such, deterrence theorists maintain that, as a punishment increases in severity, so too should its efficacy in deterring future criminal acts. Additionally, it is further maintained that, in order to effectively deter, punishments must be swift and certain. That is, swift punishments are more likely to be associated with the crime committed while the efficacy of punishments is decreased if offenders do not perceive their implementation to be likely.

There are two ways in which deterrence is proposed to function. First, there is “specific deterrence,” where it is suggested that individuals who commit crimes and are apprehended and appropriately punished will be less likely to commit future crimes. Second, it is proposed that the punishment of those who commit crimes will serve as an example to potential offenders, proactively deterring them from engaging in criminal activities in the first place. This is known as “general deterrence.”

Over the years, the empirical evidence supporting the position that stricter laws and harsh punishments are effective ways of deterring youth crime, both generally and specifically, has been at best inconsistent (Baron and Kennedy, 1998; Doob and Cesaroni, 2004; Peterson-Badali, Ruck and Koegl, 2001; Teevan, 1976), with negative effects seemingly resulting in some cases (Bishop and Frazier, 2000; Fagan, Kupchick, and Liberman, 2003; Myers, 2001; Singer and McDowell, 1998).

Psychosocial Maturity

One of the primary reasons for the existence of a separate youth justice system is an understanding that youths in general do not possess a level of maturity that allows for them to be treated in the same manner as adults. Moreover, as Zimring and Fagan (2000: 423) point out, “transferring a fourteen-year-old defendant from one court to another does not add to his or her age or maturity.” While some early studies have suggested that judicial waivers were most likely to target property offences (Feld, 1989; Hamparian et al., 1982 as cited in Lemmon, Austin, Verrecchia and Fetzer, 2005), more recent studies indicate that there is a trend toward greater violence among juvenile exclusion cases (Myers, 2003b; Snyder et al., 2000). However, contrary to a seemingly widespread belief that youth crime, and specifically violent youth crime, is on the rise, Canada’s official crime statistics suggest that its levels have remained relatively stable over the last 20 years (Doob and Cesaroni, 2004). Moreover, while some would suggest that youths who commit certain types of serious and violent crimes need to be dealt with in harsher ways, such as adult-like sentencing, the fact remains that there is no evidence to suggest that committing such crimes are indicative of a youth offender’s being more mature than his or her age peers (Zimring and Fagan, 2000).

Despite existing evidence regarding the decreased culpability of youth offenders, it has been suggested that there are three primary reasons why policies relating to youth transfers are not more influenced by the developmental psychology literature (Zimring and Fagan, 2000, as cited in Doob and Cesaroni, 2004). First, it is suggested that even though there are a wide range of developmental justifications for excluding very young children from complete criminal liability, the pressure to transfer children under the age of 14 is relatively nonexistent; that is, they are largely seen as “just being kids.” Second, developmental changes are gradual and vary between individuals, making it somewhat difficult to relate a high-variance gradual change to a yes-or-no decision. Third, determining levels of responsibility is fundamentally a moral judgment, and moral judgments are often not informed by empirical evidence. As Doob and Cesaroni (2004) suggest, the transferring of youths into adult criminal courts is seemingly more influenced by politics than by empirical evidence.

Cauffmen and Steinberg (2000) recently examined the influence of three psychosocial factors (responsibility, perspective and temperance) on maturity of judgment and processes of decision-making among an overall sample of 1,015 middle school (grade eight), high school (grades 10 and 12), and college students (under 21 years and over 21 years) in Philadelphia, PA (Cauffmen and Steinberg, 2000). Responsibility scales assessed individuals’ feelings of internal control and the ability to make decisions without extreme reliance on others. Two aspects of perspective were assessed. First, the ability to see short- and long-term consequences, and second, how often participants take other people’s perspectives into account. Measurements of temperance assessed individuals’ tendencies to exhibit impulse control and self-restraint with regard to aggressive behaviours. Study participants completed assessments of their psychosocial maturity regarding the three domains and responded to an array of questions regarding would-be anti-social or risky behaviours (i.e., shoplifting, drug use, joy-riding in a stolen car). Additionally, anti-social/risky-behaviours scenarios were presented as either having definite consequences (i.e., being arrested), no consequences, or unknown consequences.

The results of the study show that individuals differ significantly in their psychosocial maturity, as well as in their anti-social decision-making, and that these variances occur largely as a function of age. Moreover, anti-social decision-making appears to be more strongly influenced by psychosocial maturity than by age, with individuals who exhibit higher levels of responsibility, perspective, and temperance displaying more mature decision-making than those with lower scores, independent of age. Adolescents, on average, scored worse than adults, and even young adults were found to make more socially responsible decisions in comparison with adolescents. The study also reflected that the most important transition point in psychosocial development and mature decision-making is between 16 and 19 years. Together, these findings are relevant to debates regarding legal borders between adolescence and adulthood and, specifically, the reasoning, or lack of it, behind exposing youth to the adult criminal justice system.

Youth Transfers

The primary purpose of judicial waivers and the imposing of adult sentences on youth offenders are to mete out more severe sanctions than are available in the juvenile courts, in the belief that a greater deterrence effect, both specifically and generally, will be achieved. However, evidence regarding the effect of judicial exclusions on recidivism rates is at best inconsistent (Greileich, Trager, and Chishalm, 1982; Lemmon, Austin, Verrecchia and Fetzer, 2005), with the results of a number of studies suggesting that they may in fact contribute to an increase in rates of recidivism among youth offenders (Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Myers, 2001; Singer and McDowell, 1998). Until the recent implementation of the Youth Criminal Justice Act in 2003, Canadian youth 14 years and older were eligible to be transferred out of the youth justice system into the adult courts. While today youth offenders in Canada cannot have their cases transferred into the adult criminal justice system, those convicted of juvenile offences can still be ordered to serve them within an adult correctional institution.

It has been proposed that prior to youth being waived into the adult criminal justice system and/or sentenced to an adult correctional facility, they often do not receive adequate individual assessments to account for the influence of such things as psychosocial maturity as well as socioeconomic and family circumstances—factors that can be helpful in assessing their ability to be held responsible for their actions (Myers, 2003b). Moreover, once these youth enter into adult correctional facilities, they are less likely to receive educational and cognitive-behavioural and/or therapeutic counselling services, and when they do receive them, they are not necessarily developed or administered with adolescents in mind and, thus, are perhaps less effective than those available in juvenile facilities. As well, youth serving time in adult correctional facilities are more likely to experience harsh treatment and abuse and to be subjected to violence from other inmates (Bishop and Frazier, 2000; Doob and Cesaroni, 2004; Myers, 2003a). Moreover, it has been suggested that adult correctional facilities act as training grounds for youth offenders by exposing them to more sophisticated and experienced adult criminals, resulting in the youth being better prepared for future criminality (Myers, 2003a). Almost invariably, youth who have served time in adult facilities will see their future prospects for education, employment and opportunities to productively contribute to society greatly reduced (Myers, 2003a; Paternoster and Iovanni, 1989). These and additional factors may help to explain, as previously mentioned, why some research has found that youth transfers increases rather than decreases the likelihood of future offending.

A recent study conducted by David Myers (2001) examined the effectiveness of a policy targeting violent youth offenders for prosecution in Pennsylvania’s adult criminal courts. Comparisons were made between 138 male youth offenders who were transferred to adult criminal court and 419 who were retained in juvenile court. The offenders, who ranged in age from 15 to 18 years, had been arrested for robbery, aggravated assault, or both, and had used a deadly weapon while committing their offences. All court dispositions had occurred in 1994, with the author collecting data in 1998 from face-to-face interviews with judges, public defenders, and probation officials, as well as court records and a juvenile justice research and training centre.

Results of the study found that the transferred offenders were significantly more likely to be re-arrested on a violent felony offence during a pre-dispositional time period than were youths retained in juvenile court. Moreover, the transferred youths were specifically more likely to be convicted of a target offence of robbery or aggravated assault than those released from secure custody before final disposition. Additionally, it was found that in comparison with retained youths, transferred offenders were more likely to be released from secure custody prior to final disposition of their cases, with the estimated probability of release for a waived youth being nearly 20 per cent higher than that for a retained youth. This finding contradicts the deterrence-based motives of youth transfers, as it appears, at least in terms of remaining in secure custody, that the retained youths received the harsher sanctions. Ultimately, the findings suggest that, despite the seemingly more harsh conditions associated with transfers, the affected youths were apparently not more deterred by their experiences in the adult criminal justice system. In fact, the evidence suggests that these experiences may in fact have served to increase the frequency and seriousness of future offending by those youth.

A more recent study conducted by the same author sought to further explore the relationship between judicial waivers and rates of recidivism while simultaneously accounting for the possibility of selection biases (Myers, 2003b). The author highlights the results of studies finding that transferred youth offenders exhibit greater recidivism rates than those retained in juvenile courts might be viewed as limited in that it may be the highest-risk youth; that is, those who are already most likely to recidivate, who are most likely to be selected for transfer into the adult criminal justice system. If such is the case, one might logically expect then to witness greater rates of recidivism among those transferred youth, independent of their experiences with the adult criminal justice system. Feld (1989) found that juveniles waived to adult court were more likely to be older, to be charged with serious offences, to have more severe prior records, to have been previously incarcerated in a juvenile institution, and to have been detained prior to waiver hearings, while other studies have found that minorities are overrepresented among juvenile exclusions (Mears, 2003; Walker, Spohn and Delone, 2004, both as cited in Lemmon, Austin, Verrecchia and Fetzer, 2005).

Data were examined regarding 494 violent male juvenile offenders located in Pennsylvania, between the ages of 15 and 18 years, who were arrested for robbery, aggravated assault, or both. Additionally, in all instances a deadly weapon was involved during the offence. Of the 494 total offenders, 79 were judicially waved to adult criminal court and 415 were retained in juvenile court. The timing, likelihood, and seriousness of their recidivism were analyzed while numerous variables were used to control for any influence they might have on the decision to transfer and on future criminal behaviour. These control variables included age, race (whites versus non-whites), location of processing (rural, urban, or suburban), school status (enrolled/graduated versus not enrolled), family status (living with two parents versus other living arrangements), type of weapon used in the offence (firearm versus other), prior offence history, age of first referral to juvenile court, total number of prior juvenile court referrals, adjudications, and placement in correctional facilities. Finally, to account for the seriousness of prior offending, juveniles who were previously adjudicated delinquent for one of a variety of violent felonies targeted by a recently enacted legislative waiver law in Pennsylvania were accounted for.

The main findings of the study are threefold. First, it was found that the transferred youth were significantly more likely to be re-arrested following final disposition than were their counterparts in juvenile court and that this held true both with and without the inclusion of statistical controls for selection bias. Specifically, results indicate that being waived to adult court more than double the simple odds of a post-dispositional arrest. Second, it was found that the transferred youth were re-arrested significantly more quickly at any point in time following disposition than those retained in youth courts and that they were also more frequent offenders. Additionally, independent of transfer status, conviction and incarceration length were both found to have a significant influence on how quickly youths were re-arrested. That is, youths who were convicted and served longer sentences were more likely to be re-arrested more quickly, regardless of whether they were retained in juvenile court or transferred. Finally, regarding selection biases, it was found that the only factors significantly related to the likelihood of transfer were age, with older youths more likely to be transferred, and prior record, with youths having a more extensive prior record being more likely to be transferred. Surprisingly, weapon type was negatively related to transfer likelihood; that is, youths who used a firearm while committing their offence were less likely to be transferred. The author suggests that this may be related to the inability to have controlled for victim injury, as it may be the case that youths using other weapon types were more likely to have caused injury to someone, a factor which may influence whether a transfer occurs or not.

A larger-scale study that was recently conducted similarly found negative outcomes as being associated with prosecuting and sentencing youth in adult criminal courts (Fagan, Kupchick and Liberman, 2003). The study’s authors examined data regarding the case outcomes and criminal histories of 2,382 youth, in three New Jersey counties, aged 15 to 16 years, who were charged between 1992 and 1993 in either juvenile or adult criminal courts with felony, assault, and/or burglary.

Similarly to the results of the study conducted by Myers (2003b), it was found that serious adolescent offenders prosecuted in the criminal court were more likely to be re-arrested more quickly and more often for violent, property, and weapons offences and that they were more quickly returned to incarceration. However, it was also found that adolescents retained in youth courts were more likely to be re-arrested specifically for drug offences. The authors contend that categorical exclusion of certain groups of youth offenders based solely on offence and age are ineffective as a specific deterrence in youth crime, suggesting that individual assessments and more judge-centred transfer policies can help to limit the number of youth subjected to criminal court prosecution and to the factors related to serving time in adult correctional facilities that may contribute to negative outcomes.

Youth Transfers and General Deterrence

According to deterrence theory, transferring youth offenders into criminal courts and sentencing them into adult correctional facilities should, ultimately, have a general deterrent effect on juvenile crime rates. That is, potential youth offenders, both those who have not yet engaged in delinquent activities and those who have, should theoretically be deterred from engaging in future delinquent activities from fear that, if caught, they will be subject to harsh adult-like sanctions

An early study conducted by Ruhland, Gold and Heckman (1982) questioned whether severity of punishment has a deterrent effect on juvenile crime by comparing Uniform Crime Report (UCR) arrest and National Survey of Youth (NSY) self-report data for youths between 16 and 17 years of age in jurisdictions where they are either defined as juveniles or adults. From a deterrence standpoint, it is hypothesized that youth defined as adults will have lower arrest rates and lower rates of self-reported offences than will youths defined as juveniles, as there should be a tendency to believe that being under adult rather than juvenile jurisdiction will expose them to more severe penalties. Status was classified according to whether adult status was reached at age 16, 17, or 18 in each of the years 1970 to 1976. The offences committed ranged from property-related (i.e., car theft, burglary) to violent (i.e., aggravated assault, murder).

Initial analyses found that in all seven years, the juveniles, either 16 or 17 years old, were more delinquent than their adult counterparts. Further analyses were then conducted to test for the deterrent effect of adult versus juvenile status by comparing 16-year-old juveniles from states where 16 was the maximum age for juvenile status with 16-year-old “adults” from states where the minimum age for adult status was 16 years, with the same method used regarding 17-year-old youth. Results indicated that the oldest juveniles under the juvenile law tended to appear more often in the juvenile records of their states than did their adult-age counterparts in other states. Specific results showed that for 16-year-old youth, the tendency to appear more often in the juvenile records of their states was reflected only within the processes of the juvenile justice system, but not by the behaviour itself, as measured by the self-report data. In contrast, this tendency was captured both by official juvenile justice system data as well by the youths’ self-reports among the 17-year-old youth. Thus, as the authors note, it appears that the deterrent effects of adult status was present to some degree only among the 17-year-old youth while it did not have a deterrent effect regarding 16-year-old youth.

A later published study similarly sought to measure the deterrent effect of a legislative waiver implemented in Idaho in 1981 on violent juvenile crime (Jensen and Metsger, 1994). Specifically, the legislation mandated the automatic waiver for youths aged 14 to 18 years who were accused of murder of any degree, as well as for those accused of attempted murder, robbery, forcible rape, mayhem (i.e., dismemberment or severe disfigurement), and assault or battery with the intent to commit any of the aforementioned offences. The authors employed a multiple time series design by examining agency records during the five-year periods of 1976–1980 (pre-statute) and 1982–1986 (post-statute). Analyses testing for the deterrent effects of the legislative waiver were conducted by comparing arrests for violent juvenile crime in Idaho with those in Montana and Wyoming, both before and after the statute implementation. The states of Montana and Wyoming were chosen as comparisons because, as the authors maintain, they are, first, deemed as being both demographically and economically similar to Idaho, and second, their remand procedures were much like those used in Idaho prior to the statute change in 1981, in that both used judicial waiver to remand youth charged with violent crimes to criminal court.

The results of the study did not provide support for the deterrent effects of the Idaho legislative waiver on rates of violent juvenile crime. Specifically, it was found that the average arrest rate in Idaho during the pre-statute period was 71.1 per 100,000 persons under the age of 18, a rate that significantly increased to 83.9 during the post-statute period, representing an 18 per cent increase. In comparison, Montana showed a change from 31.2 during the pre-statute period to 17.1 during the post-statute period, or a significant decrease of 45 per cent. Similarly, Wyoming exhibited a significant decrease of 14 per cent, with the average arrest rate of persons under 18 years of age for violent crime during the pre-statute period being 30.2 compared with 26.0 in the post-statute period. Based on these results, it appears that the legislative waiver law did not have a deterrent effect on rates of serious violent juvenile crime.

As the authors acknowledge, there are several possible explanations for the failure of the legislative waiver to deter violent juvenile crime in Idaho. First, it may be the case that prosecutors in Idaho were charging some youth offenders accused of serious violent crimes with lesser offences, resulting in their remaining in the juvenile court system. Second, youth simply may not have been aware of the fact that they would be tried in criminal court and possibly sentenced to prison for committing one of the violent offences specified by the statute. In support of this notion, the authors note that media coverage in Idaho regarding the enactment of the legislative waiver was almost non-existent. Third, the authors suggest that the young ages of juvenile defendants may result in a tendency for them to be sentenced less severely, subsequently decreasing any deterrent effect the legislative waiver may have had.

Youth Waivers and Sanction Certainty and Severity

As mentioned, one of the primary goals of judicial exclusion is to mete out presumably more severe sanctions than those available in the youth justice system. Above and beyond the debate over whether these more severe sanctions have a desirable effect on rates of recidivism is the issue of whether this goal is actually being achieved. That is, some evidence suggests that youth transferred into adult courts actually may not be more likely to be incarcerated (Lemmon, Austin, Verrecchia and Fetzer, 2005), and that when they are, they may in fact be serving shorter sentences than those retained in juvenile jurisdictions (Dawson, 1992; Rudman et al., 1986; Fritsch, Caeti and Hemmens, 1996).

Focusing on two dimensions of sentence severity, sentence length and actual time served, Fritsch, Caeti and Hemmens (1996) sought to determine if in fact juveniles waived to adult courts do receive more severe sanctions. The authors note that while previous studies have found that juveniles waived to adult court usually receive longer sentences than are available in juvenile court (Dawson, 1992; Rudman et al., 1986), these results may be misleading, as none of the studies investigated whether longer sentences actually equated to lengthier terms of incarceration.

Data were collected on all youth waived to adult court in Texas between the years 1981 to 1993 and sentenced to prison, resulting in a total sample size of 946. The most common offence was homicide (34.1 per cent) followed by robbery (25.1 per cent) and burglary (13.8 per cent), with the remaining offences ranging from drug-related to larceny. Demographic information, including age of the offenders, was not provided.

The average sentence length for youth waived to adult court and sentenced to prison was 21.9 years, with only 13.1 per cent of these youth receiving sentences of four years or less and almost 35 per cent receiving sentences over 20 years in length. Judging by these sentences, the authors do acknowledge, as other studies have found, that transferred youths do typically receive harsher sentences than those retained in the juvenile courts. However, when actual time served was taken into consideration, a different picture emerged. It was found that transferred youth rarely served more lengthy sentences than are available in juvenile court, serving an average of only 27 per cent of their sentences (i.e., less than six years of the average 21.9 year sentence).

A study conducted more recently investigated the effect that legal and extralegal factors might have on the statutory exclusions of youth offenders and, in doing so, similarly tested for the presence of two primary elements related to deterrence theory—certainty and severity of punishment (Lemmon, Austin, Verrecchia and Fetzer, 2005). Specifically, it was assessed whether juveniles processed in the adult court system receive greater certainty and severity of punishment compared with those processed in the juvenile system. The study looked at looked at 701 cases from four different U.S. states: Arizona, Pennsylvania, South Carolina and Utah. Of the 701 cases, 114 were dismissed at the preliminary hearing, leaving 198 cases being returned to juvenile court and 389 remaining in the adult judicial system. The average age of the youths at the time of their arrest was slightly over 16 years and the vast majority of the charged crimes were aggravated assault (50 per cent) or robbery (45 per cent), while approximately 96 per cent of the youth had used a weapon while committing the offence. Demographic data regarding the cases included race, gender, age and county where the incident occurred. Legal data were arranged into four categories: offence information, criminal court processing information, juvenile court processing information and juvenile court history. A variety of variables were used to examine certainty and severity of punishments in both court types, including whether cases were dismissed, whether a youth was found guilty and, if so, whether the youth was sentenced to incarceration.

Regarding certainty of punishment, it was found that there were no significant differences between the two groups of offenders. Specifically, whether youth were dealt with in the juvenile or adult courts made no difference on the likelihood that they would be convicted. In fact, it was found that the conviction rate was slightly higher in juvenile court (76 per cent) than in adult court (72 per cent), but this difference was not statistically significant.

Regarding severity of punishment, it was found that youth found guilty in adult court were significantly more likely to be incarcerated (87 per cent) than those found guilty in juvenile court (55 per cent). However, further analyses found that legal factors such as juvenile court history, seriousness of offence, and weapon were significantly related to incarceration in adult court.

When considered together, the results of these two studies suggest that the deterrence-based principles of punishment certainty and severity are not necessarily always being met when transferring youth into the criminal justice system. Seemingly, in some cases youth are not more likely to receive sentences than if they were to remain in the juvenile courts and the sentences they do receive are often not necessarily good indicators of the actual time they will serve. If such goals are not being met, we must then ask ourselves why we should continue to transfer youth offenders into adult courts and correctional institutions. Ultimately, these studies are limited in that they can only speak to punishment certainty and severity of transferred youths in select U.S. states during specific time periods. However, much of what occurs in the United States regarding youth justice is reflected within the Canadian context and so such studies are useful in informing Canadian policy-makers, stakeholders, and the public in general. Specifically, it would of interest to determine to what extent Canadian youth sentenced to adult correctional institutions are serving their sentences, as well as how this might vary according by province.

In sum, it seems that the deterrent effects of youth transfers are unclear, with negative effects being seen in a number of instances. Specifically, the results of some studies demonstrate that youth transferred into the adult criminal justice system may be at a higher risk to recidivate (Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Myers, 2001, 2003b), and to do so more quickly than retained youth (Fagan, Kupchick and Liberman, 2003; Myers, 2003b). Moreover, as the results of the studies conducted by Ruhland, Gold and Heckman (1982) and Jensen and Metsger (1994) suggest, policies pertaining to judicial waivers of youth offenders and adult-like sentencing do not appear to have a clear general effect on juvenile crime rates. As well, and in line with principles of developmental psychology, it appears that the deterrent effects of harsher, more adult-like sanctions are variant according to age (Cauffmen and Steinberg, 2000; Ruhland, Gold and Heckman, 1982), suggesting the need for individual assessments and judicial discretion regarding youth waivers. Finally, there is a reasonable likelihood that, above and beyond debates regarding the efficaciousness of youth waiver policies, in a certain number of cases the deterrent goals of punishment certainty and severity are not necessarily being met, leaving us to question the utility of continuing to employ such practices.

General Research on Youth Offending and Deterrence

Compared with other Western countries, including the United States, Canada tends to sentence a relatively high number of youth to custody, even though official statistics show that the majority of youth offending in Canada comes in the form of minor, typically property-related offences (Doob and Cesaroni, 2004). Perhaps not entirely unrelated then is the fact that Canada also tends to sentence a relatively high number of youth to custody for seemingly minor offences, despite evidence suggesting that most of the youths committing them likely are adolescence-limited offenders and will desist naturally (Doob and Cesaroni, 2004; Moffitt, 1993; Phillips and Votey Jr., 1987). Over the years, the empirical evidence supporting the position that stricter laws and harsh punishments are effective ways of deterring crime, both generally and specifically, has been at best inconsistent (Baron and Kennedy, 1998; Doob and Cesaroni, 2004; Peterson-Badali, Ruck and Koegl, 2001; Teevan, 1976), with negative effects being seen in a number of studies (Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Myers, 2001, 2003b; Pogarsky, Kim and Paternoster, 2005).youth to custody for seemingly minor offences, despite evidence suggesting that most of the youths committing them likely are adolescence-limited offenders and will desist naturally (Doob and Cesaroni, 2004; Moffitt, 1993; Phillips and Votey Jr., 1987). Over the years, the empirical evidence supporting the position that stricter laws and harsh punishments are effective ways of deterring crime, both generally and specifically, has been at best inconsistent (Baron and Kennedy, 1998; Doob and Cesaroni, 2004; Peterson-Badali, Ruck and Koegl, 2001; Teevan, 1976), with negative effects being seen in a number of studies (Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Myers, 2001, 2003b; Pogarsky, Kim and Paternoster, 2005).

Perceptions of Punishment Certainty and Severity

According to deterrence theorists, in order for laws and sanctions to act as effective deterrents, offenders and would-be offenders must first perceive that there is a reasonable likelihood of being caught committing a crime, and second, they must perceive that ensuing punishments will both be certain and severe enough to outweigh the potential proceeds garnered from the crime. An early study conducted in Canada by James Teevan (1976) examined the deterrent effects of punishment by collecting self-report data regarding punishment perceptions and personal history information regarding various forms of deviance for a sample of youth who frequented drop-in centres in London, Ontario, as well from residents in a boys’ and girls’ treatment centre. Specifically, data regarding the youths’ fear of punishment as well as the youths’ perceptions of punishment certainty and severity were collected in order to establish whether deviants commit crimes regardless of fear or perceived certainty of punishment. All of the youths were asked questions regarding the likelihood of being caught for committing specific offences and what the typical punishments for the offences were. Youths who had committed deviant acts were asked to indicate in percentages what they thought the chances were of being caught prior to committing the act. As well, these youths were asked how severe they perceived the likely penalty for the act to be (“nothing at all” to “very bad”). Finally, all of the youth were asked to rank fear of punishment as being either first, second, or third in importance as a reason for not engaging in deviant acts.

The results of the study demonstrated that perceived levels of punishment certainty are only weakly related to certain types of deviance, while the perceived level of punishment severity is even more weakly related to deviance levels than is punishment certainty. As such, the author concludes that youths who perceive a higher severity of punishment do not necessarily refrain from committing deviant acts and those who do attend to punishment severity do not necessarily engage in less deviance as a function of their also perceiving a level of punishment certainty.

A more recent Canadian study focused on youth offenders’ perceptions of the sentences that they had received in order to investigate the relationship between youth justice system involvement and perceptions of specific and general deterrence (Peterson-Badali, Ruck and Koegl, 2001). The authors interviewed 53 male youth offenders in Toronto, Ontario, between the ages of 13 and 17 years, who were convicted of offences ranging from minor thefts to armed robbery. Additionally, the study explored whether specific offender (age, race, and criminal history), offence (type of offence), and disposition characteristics (sentence length and type) were influential in predicting the youths’ perceptions of the specific and general deterrent value of their sentences.

The results of the study are mixed, with no clear evidence to support a deterrence approach for harsh sentencing. It was found that youths vary in how they perceive their sentences as deterrents and that offender, offence, and disposition characteristics were not able to explain this variance. Regarding specific deterrence, 58 per cent of the youth indicated that their sentence would deter them from committing future criminal acts, while 42 per cent said that their sentence would have no bearing whatsoever on whether they participated in criminal activities in the future. As a note, 40 per cent of the youth who did believe that their sentences would deter them from reoffending indicated that they had undergone personal changes as a result of experiences with counselling and anger management, as well as cognitive and behavioural focused sessions. In terms of general deterrence, 66 per cent of youth stated that their sentences would not deter their friends from committing the same offence. Moreover, while more than half of the youth believed that some form of criminal sanction would occur should they be caught committing their offences, more experienced repeat offenders indicated incarceration as a likely consequence. However, all but three of the youth indicated that they did not think they would be caught when they committed their offence. As the authors highlight, it seems that, ultimately, youths’ views regarding their sentences reflect both rehabilitative and harsher deterrent approaches. While there are some limitations to the study regarding issues of using retrospective self-report data as well as a having a relatively small sample size, the results do suggest that the role that disposition experiences play is one which warrants further empirical study.

The findings of the study conducted by Peterson-Badali, Ruck and Koegl (2001) suggest that multiple rather than singular approaches are perhaps the most effective way to respond to youth crime. Accordingly, a very recent Canadian study focused specifically on how five different sentencing models might influence serious and violent young offenders’ decisions to engage in future delinquent acts (Corrado, Cohemn, Glackman and Odgers, 2005). The first sentencing model was “deterrence,” understood as youths’ fear of certain and severe punishment. Next is “fairness,” defined as the degree to which youths perceive their sentences to be fair proportional to their understanding of the seriousness, criminality and culpability of their behaviour. The third sentencing model tested was “chronic offender lifestyle,” which speaks of the degree to which a youth offender has adopted offending as a lifestyle choice, as there is evidence that youth who do so do not employ the same sort of cognitive cost-benefit processes that other offenders might engage in to some degree (Baron and Kennedy, 1998). The fourth sentencing model is “special needs,” which suggests that youth who perceive that their subjective emotional and personal needs are being met are less likely to recidivate. The final sentencing model included is “procedural rights”; that is, the degree to which young offenders feel that their procedural rights have been protected.

The study’s sample consisted of 400 youth who were serving custodial sentences at one of four detention centres located in Vancouver, British Columbia, between April 1998 and December 1999. Of these 400 youth, 87 per cent had had at least one previous disposition. The sample was predominantly male (81 per cent) with an average age of 16 years. At the time of their offence, 47 per cent of the youth lived at home with one or more of their biological parents and 53 per cent were enrolled in school. Almost all of the youth reported relatively frequent drug and alcohol use (93 per cent), including cocaine (56 per cent), heroin (35 per cent), and crack cocaine (44 per cent). Additionally, 72 per cent of the youth reported having a family member with an alcohol problem and 60 per cent indicated the same regarding drugs. Finally, 70 per cent of the youth reported having a family member with a criminal record.

The authors report that the average youth in the sample had already spent 33 months on probation, 72 days in closed custody, and 47 days in open custody. Regarding offence profiles, the authors note that over 66 per cent of the youth were either currently serving a sentence for a violent offence or had previously been convicted for one.

The authors note that the 400 youth represent a 93 per cent response rate allowing for them to interview youths at all stages of the incarceration process, ranging from those who had just been adjudicated to those who were in the final stages of their sentences. Individual interviews ranged in length between 2 to 2-1/2 hours and were semi-structured in nature, assessing the youths’ attitudes, values, and experiences regarding a wide range of issues, including those pertaining to the aforementioned sentencing models as well as to a multitude of other variables (i.e., school, family, criminal history, social bonding, peers, restorative justice issues). Finally, the authors state that each youth’s individual file was coded in order to account for a wide range of information, including socio-demographic and mental health issues.

The results of the study indicate that no one sentencing model is primarily attended to by youth offenders regarding their decision to recidivate, but rather a variety of sentencing models are attended to. However, it was also found that a large portion of the young offenders in the sample did not attend to any of the models. Regarding deterrence, it was found that perceptions and attitudes regarding punishment severity were related to youths’ intentions to recidivate. However, the authors do acknowledge that because the majority of the youths who comprised the study’s sample were being housed in a maximum security facility, the construct of punishment severity was likely more salient. It was also found that the chronic offending lifestyle sentencing model was significantly related to intention to recidivate, indicating that situational and lifestyle factors are important things to consider in responding to youth crime. Fairness, procedural rights, and special needs factors also appear to play a role in young offenders’ decisions to recidivate; however, they do not do so to the degree that deterrence and lifestyle factors do. Ultimately though, the study found that more than half the youth in the sample did not attend to any of the sentence models in terms of their reports regarding engaging in future criminality.

Two recent and larger-scale studies using data from the U.S. National Youth Survey, a longitudinal study of delinquency, substance abuse, and other problem behaviours, similarly investigated youths’ perceptions of punishment certainty and severity, also showed somewhat mixed results. The first study measured perceptions of the U.S. criminal justice system among 4,621 male youths between 12 and 16 years of age (Lochner, 2003). The author found that young males who engage in crime without being arrested are less likely to believe that they will be arrested while committing future criminal acts in comparison with those who engage in crime and are arrested. Additionally, it was determined that youths with a lower perceived probability of arrest are more likely to engage in future criminal activities. In sum, and seemingly consistent with deterrence theory, following an arrest, youth were less likely to commit crime and more likely to increase their perception of the likelihood of arrest should they engage in future crime. However, it has been posited that the findings of this study are perhaps overstated in terms of the degree to which youth offenders employ a “rational choice” approach to offending, in that only 25 per cent of the tests measuring whether youth increased their perception of punishment certainty in response to being arrested reached statistical significance (Pogarsky, Kim and Paternoster, 2005).

The second study utilized similar NYS data, consisting of 1,247 youth between the ages of 11 and 17 years, and similarly investigated how perceptions of punishment certainty are related to offending experiences, with seemingly contrasting results (Pogarsky, Kim and Paternoster, 2005). First, in contrast with the findings of the study published by Lochner (2003), it was found that arrests had no effect on perceptions of punishment certainty; however, more experienced offenders were more likely to display decreased perceptions of both arrest and sanction probability. Second, peer offending corresponded with reductions in perceived certainty of punishment for stealing, but not for attacking, suggesting that crime decision-making processes may differ by crime type. Third, prior offending experience did not decrease the effects of offending experiences on risk perceptions, in some cases even enhancing them. Finally, it was found that “moral inhibition”; that is, perceptions of how wrong it is to commit certain deviant acts, reduced the degree to which offending experiences affected sanction risk perceptions. However, it was deemed that stronger moral inhibition levels were required to decrease the influence of violent offending experiences on perceived punishment certainty. These results suggest that even if youth offenders do engage in rational decision-making prior to committing offences, a position that is only weakly supported, they perhaps do so differentially as a function of crime-type among other things. Moreover, sanctions that include programs aimed at increasing moral inhibition may be more effective at decreasing rates of recidivism for a variety of youth offenders.

Lifestyle and Economic Factors

It has been put forth that the majority of serious and violent youth crime is committed by a relatively small number of “high-risk” offenders (Baron and Kennedy, 1995; Doob and Cesaroni, 2004; Phillips and Votey Jr., 1987). Moreover, it has been suggested that many of these chronic offenders come from the street youth population and that they are perhaps less likely to fear formal punishments in comparison with other youth offenders (Baron, 1995). As such, studies focusing on better understanding how specific groups of high-risk offenders perceive threats of formal punishment can perhaps be useful in more effectively decreasing rates of recidivism among these high-risk populations.

Baron and Kennedy (1998) examined how the effects of the threat of formal punishment on the criminal behaviour of homeless male street youth might be influenced by their living conditions and other qualities of their lifestyle such as poverty, homelessness, substance abuse, missing normative constraints, and association with criminal peers. The authors conducted interviews with 125 male street youth under the age of 24 in Edmonton, Alberta, asking them how often in the past year they had committed serious property crimes (i.e., breaking into a car/house) and/or a serious assault (i.e., attack with a weapon, causing injuries likely needing medical attention). All of the youth in the sample had participated in various criminal activities in the past year, with some being more heavily involved than others. To measure punishment certainty, respondents were asked how likely it was they would be caught for specific crimes, and in order to assess punishment severity perceptions, they were asked the degree to which punishments for specific crimes would cause problems within their lives.

The study found that, generally, fear of punishments was diminished by elevated levels of poverty, drug use, association with criminal peers, and missing normative constraints. Results further showed that while many street youths fear legal sanctions to some degree, more serious offenders do not. Additionally, perceptions of the impact of punishments and the likelihood of future offending differed by offender and crime type. Specifically, those who thought it reasonably likely they might be caught for property crime, and those who indicated that the ensuing punishments would cause a problem in their lives, were less likely to commit these crimes. In contrast, street youths’ perception of how likely it was that they might be caught committing a serious assault had no bearing on whether they had engaged or would engage in serious violent behaviour. Moreover, for these serious violent offenders, perceptions of the impact of punishments were not related to their own level of violence. As the authors note, those street youth engaging in serious violent crimes do not appear to engage in any form of rational decision-making prior to committing such crimes, but rather they are swayed by impulse and emotion. Furthermore, it is suggested that the best way to increase threats of punishment for these high-risk youth may be to provide various forms of economic resources, drug rehabilitation, and integration into conventional peer groups, and that simply imposing harsher penalties will only “clog up” the penal institutions while doing very little, if anything, to affect actual crime rates (Baron and Kennedy, 1998).

While the preceding study directed its focus towards how lifestyle factors, such as poverty, might affect crime involvement among a high-risk sample of street youth, a study conducted much earlier study similarly found that economic opportunities similarly played a role in whether a large and nationally representative sample of youths engaged in criminal activity (Phillips and Votey Jr., 1987). The study utilized data collected by the previously described U.S. National Youth Survey (NYS) regarding 6,398 male and 6,288 female youth between the ages of 14 and 24 years of age. Data examined for the purposes of the study included comprehensive information regarding educational and employment experience, family situation, sources of earned and supplemental income, involvement with property, personal, and drug crimes, and contacts with police, courts, and correctional institutions. The study’s authors sought to investigate if a rational process of choice may be influenced by both deterrence and economic opportunities, perhaps ultimately leading to three sub-groups regarding youth offenders: innocents (a large group who never engage in deviant activities), desisters (those youth who engage in deviant activities but eventually cease to do so), and persisters (a small group of recidivists who are responsible for much of the serious youth crime).

The results of the study were mixed, in that there was some support for deterrence and rational choice models, but also in that economic opportunities provided a mediating effect. Specifically, based on the youths’ responses regarding thefts, reported income from crime, and number of police stops, it was generally found that the experience of a first and sometimes second apprehension tends to reduce future criminal behaviour. However, it was further determined that the presence of legitimate economic opportunities also tend to decrease the likelihood of engaging in criminal activity. While the results of the study showed that apprehension and the threat of punishment did play a limited role in deterring some forms of criminal behaviour, it was further found that the availability of alternative, legitimate sources of income was also related to a decrease in criminal activity.

System Involvement, Correctional Experiences and Deterrence

There have been a number of empirical studies conducted over the years assessing the impacts that experiences within correctional environments have on youth offenders and the likelihood of their engaging in future criminal activity. Generally, there has been either little or, at best, mixed empirical evidence to support the notion that longer and/or harsher sentences are efficacious in deterring youth crime (Abrams, 2006; Doob and Cesaroni, 2004; Greilich, Trager and Chishalm, 1982; Schneider, 1990).

For instance, a relatively early study conducted by Greilich, Trager and Chishalm (1982) focused on the effectiveness of correctional placements as a deterrent of future criminality among youth in Massachusetts who had been convicted for committing various serious violent offences, including murder, rape, arson, and assault and battery. The results of the study indicate that while youth offenders do respond differentially to environments according to restrictiveness, this effect was greatly influenced by the number of prior offences a youth had committed.

In partial support of deterrence models, it was found that first-time offenders responded more positively to restrictive settings as compared with those who had multiple prior offences, with those placed into secure settings exhibiting the greatest decreases in subsequent offending rates. However, in contrast to the principles of deterrence, youth who had multiple prior offences were more responsive to a less restrictive setting. The authors posit that for those youth who had no prior experience with correctional sanctions, a sort of “shock” effect of incarceration was produced, which served to deter them from engaging in future delinquency. In contrast, they suggest that youth who had prior experience with total incapacitation were likely less shocked and intimidated. Just as importantly, it was found that the length of stay was not related to reducing the likelihood of whether the youth did or did not engage in future violent offending; that is, shorter sentences were just as effective as longer ones.

A more recently conducted study similarly sought to examine how male youth offenders experience the treatment or deterrence aspects of institutional confinement (Abrams, 2006). Ethnographic data was collected from 19 male youth offenders ranging in age from 14 to 18 years who were placed in one of two residential correctional facilities located in Minnesota. The first facility, Wildwood House, adopted a blended rehabilitation approach and dealt with youth who had committed relatively minor and non-violent offences, housing them for a period of four to six months as well as including a three-month aftercare program. Treatments included individual, group, and family counselling, cognitive-behavioural therapy, and substance abuse treatment. While the second facility, Cottage Grove, did incorporate some treatment components, it used a harsher, more restrictive approach and dealt with violent and repeat offenders serving minimum nine-month sentences. In comparison with Wildwood House, Cottage Grove was extremely strict and had a lengthy and detailed set of rules with a clear set of punishments for infractions. Additionally, Cottage Grove is a highly self-contained facility, resulting in youth never leaving the premises during their confinement.

Data was collected through three primary techniques. First, participant observations offered a sense of the how the youth were treated, as well as specifics regarding the environment, staff and youth interactions, and how youth responded to treatment. Second, in-depth interviews were conducted to assess the youths’ experiences and perceptions of their confinement and treatment. Finally, the author reviewed facility records in order to gain a better idea of program goals and the case histories of the youth.

Results of the study found that, first, many of the youth did not necessarily understand the purposes of their treatment and group activities, feeling that staff did not adequately explain treatment meanings and relevance to their individual situations. However, the author notes that the less-strict and family-focused program at Wildwood did appear to offer more room for positive relationships with adults. Second, it was found that secure confinement did not appear to have a deterrent effect, and that this was especially true for youth who were used to leading chaotic lifestyles and/or were accustomed to institutional confinement. As has been noted in previous research (Baron, 1998), for youth coming from chaotic lifestyles, confinement may simply hold little shock value and, in some cases it may be seen as a welcome break from the chaos that they deal with on the streets. While this study is limited, in that the two facilities included housed two different populations of offenders in terms of offences committed, it does still offer some insights into how youth offenders in general perceive treatment and secure confinement.

A longitudinal study using data collected through the Iowa Youth and Families Project (IYFP) was recently conducted to assess how criminal justice system involvement might affect the likelihood that youth will continue to engage in deviant behaviour (Johnson, Simon and Conger, 2004). Seven waves of data across the second decade of life from early (junior high) to late adolescence (high school) and into adulthood were examined regarding 153 male youth who had all had early criminal justice system contact. In order to tap into family processes, individual family member characteristics, and socio-economic circumstances of the parents and a sibling within four years of the age of the youth (an inclusion criterion), all provided information for the study. All of the families were white and had an average size of five members, with the average income of the families being just under $30,000US per year. Data collected by the IYFP included information regarding criminal behaviour, deviant peer association, and criminal justice system involvement from early adolescence to adulthood.

The main results of the study are twofold. First, continuity in offending across waves was revealed, with partial continuity regarding deviant peer association. Specifically, deviant peer association was found to be related at grade 10 and early adulthood, effects that were not seen during the youngest adolescent years. The authors note too that prior deviant peer association was not found to be related to later deviant peer association until after justice system involvement in Grades 8 and 9, which might suggest that formal sanction played a role in the forging of deviant relationships. Additionally, strong correlations between deviant peer association and crime were found at grade seven, grade 10, and three years after high school.

Most importantly, it was found that early criminal justice system involvement was positively and significantly related to future delinquent behaviour as well as to deviant peer association. As such, a rational choice/deterrent argument was not supported, as prior system involvement was not found to be inversely related to later criminal behaviour. Delinquency during grade seven was positively related to system involvement during grades eight and nine, and delinquency during grade 10 was positively related to system involvement during grade 12 and the first year after high school. Most notably, system involvement during Grades 8 and 9 was positively related to delinquency during grade 10, and the same relationship was found between grade 12 and the first year after high school.

In sum, in contrast with theories of deterrence, there is evidence to suggest that youths who perceive or experience higher severity of punishment are not necessarily deterred from committing deviant acts (Abrams, 2006; Greilich, Trager and Chishalm, 1982; Johnson, Simon and Conger, 2004; Teevan,1976). Additionally, it seems that fear of punishment is diminished with a variety of lifestyle factors, including poverty, elevated drug and alcohol use, and missing normative constraints (Baron, 1998). Moreover, deviant peer association appears to quite often be influential on later delinquent behaviour (Baron, 1998; Johnson, Simons and Conger, 2004). System involvement itself may in fact contribute to “labelling” effects, which can result in reducing opportunities for youth offenders to access legitimate and productive members of society. Positive effects may be most likely achieved by focusing on individual and family counselling, drug and alcohol interventions, anger management, and cognitive-behavioural sessions aimed at increasing self-esteem, moral inhibition, and feelings of responsibility (Abrams, 2006; Pogarsky, Kim and Paternoster, 2005; Teevan, 1976). Finally, there does not appear to be simply one overarching approach that is most effective in responding to youth crime, but rather, multiple and blended approaches including both rehabilitative and deterrent goals are perhaps most likely to achieve positive results (Corrado, Cohemn, Glackman and Odgers, 2005).

“Shock” Approaches: Scared Straight and Boot Camps

Advocates of Scared Straight and similar types of “shock” programs believe that by exposing existing and potential youth offenders to realistic consequences of crime through visits to correctional facilities and interactions with prisoners the likelihood that they will engage in future criminal activities can be greatly decreased. However, while these programs have enjoyed a high level of support throughout the years, both from policy-makers and the general public, there is consistent empirical evidence demonstrating that they are not effective in achieving their intended goals of reducing youth crime and that, in fact, they might actually serve to increase the likelihood that youth will engage in future criminal activities (Cook and Spirrison, 1992; Finckenauer, 1982; Petrosino, Turpin-Petrosino and Buehler, 2003; Petrosino, Turpin-Petrosino and Finckenauer, 2000). Thus, in most cases, doing nothing at all would have better than exposing youth to these programs. Accordingly, the popularity and longevity of these programs has been suggested to be due to their basic “get-tough” premise, the fact that they fit well with common notions of how to prevent and reduce crime, their low cost, and the idea that they are a way for incarcerated offenders to effectively contribute to society, rather than to the outcomes they produce (Petrosino, Petrosino-Turpin and Beuhler, 2003).

In 1976 the first Scared Straight program began in New Jersey’s Rahway State Prison and subsequently gained widespread attention through a nationally televised documentary, which aired in 1979 and was awarded the Oscar for Best Documentary Feature of the Year. The purpose of the program was to literally scare youth away from delinquency by bringing them into the prison “where menacing inmates (serving life sentences) subjected them to threats, intimidation, emotional shock, and verbal bullying” (Greenwood, 2006: 101). The film proclaimed that sixteen of seventeen delinquents who had participated in the program had remained “crime free” during a three-month follow-up period, equating this to a 94 per cent success rate (Finckenauer, 1982). Subsequently, the Scared Straight programs very quickly caught on and have since been replicated both domestically and internationally. However, what the documentary did not highlight was that the youths involved in the program were not necessarily at a high risk for delinquency in the first place, with seven of the youth not having any prior delinquency records (Greenwood, 2006).

A recently conducted meta-analysis examined the effects of nine Scared Straight-like programs implemented in the United States between the years 1967 and 1992 in order to test for possible deterrent effects (Petrosino, Turpin-Petrosino and Buehler, 2003). Criteria for a study being included in the analyses included the presence of a no-treatment control group as well as at least one outcome measure of “post-visit” behaviour, components that many assessments regarding these programs have failed to include. The follow-up periods across the studies were diverse, including measurements at three, six, nine, 12, and 24 months. The average age of the participants in each study ranged from 15 to 17 years and most of the youth had already come into contact with the juvenile justice system. The vast majority of the youth involved were male, with only one study including female participants. Additionally, the authors posit that the samples were racially diverse, with percentages of youth who were white ranging from 36 to 84 per cent. In total, the meta-analysis accounts for 946 juveniles and young adults.

The results of the meta-analysis indicate that Scared Straight and other “juvenile awareness” programs are not effective in deterring youth from engaging in future criminality. In fact, it was found that, in most cases, the programs tended to contribute to negative outcomes, with youth exposed to them being more likely to recidivate, and that this pattern generally held, to varying degrees, across the span of follow-up times. For example, one study conducted in Michigan found that 43 per cent of the youths who participated in the Scared Straight program recidivated, compared with only 17 per cent in the control group (Michigan Department of Corrections, 1967, as cited in Petrosino, Turpin-Petrosino and Buehler, 2003). The only study that did report positive results found that, at six months, the control group had less new court intakes (39 per cent) compared with the Scared Straight group (41 per cent). However, the reverse was seen by nine and 12 months (Orchowsky and Taylor, 1981, as cited in Petrosino, Turpin-Petrosino and Buehler, 2003).

Farrington and Welsh (2005) also conducted a review of randomized experiments, assessing two Scared Straight programs. The first program was conducted in 1983 and included 108 male youth offenders between the ages of 14 and 18 years (Lewis, 1983, as cited in Farrington and Welsh, 2005). The second program was conducted in 1992 and included 176 male youth offenders with an average age of 15 years (Cook and Spirrison, 1992). Both studies included a comparison control group.

The results of the analysis found that at 12-month follow-up periods, the youth exposed to the Scared Straight programs were more likely to be arrested than their control group counterparts. For instance, in the study conducted by Lewis (1983), over 80 per cent of the Scared Straight youth had been arrested at the one-year period, compared with 67 per cent of the control group youth.

In sum, there is consistent empirical evidence to suggest that Scared Straight and similar “juvenile awareness” programs have no deterrent effects and in many cases may contribute to increasing the likelihood that youth will engage in future criminal activities (Cook and Spirrison, 1992; Finckenauer, 1982; Petrosino, Turpin-Petrosino and Buehler, 2003; Petrosino, Turpin-Petrosino and Finckenauer, 2000). However, despite these widespread and consistent findings, such programs are still used, largely because of their “common sense” approaches and low administrative costs.

Boot Camps

Correctional boot camps, a form of “shock” incarceration, were first used in the United States during the early 1980s with adult offender populations. Today, however, boot camp models can be found in a number of countries, including Canada, as well as in both adult and juvenile settings. Boot camps differ from traditional correctional facilities in that they are based on military models and are typically reserved for use among non-serious and youthful offenders, with sentence lengths being relatively short in length (Greenwood, 2006). That they are typically aimed at such offender populations may not necessarily be commonsensical considering that, as previously highlighted, many young offenders who commit relatively minor offences are likely to “age out” of their delinquent behaviours naturally (Doob and Cesaroni, 2004; Moffitt 1993).

While boot camps dealing with adult and juvenile offenders do typically differ somewhat in quality, with juvenile camps placing less emphasis on hard labour and often incorporating therapeutic and educational components, they do share the core elements that advocates see as being integral to their efficacy as a deterrent to future offending (Mackenzie, Wilson and Kider, 2001). All boot camps typically incorporate rigorous daily schedules focused upon physical training, drill and ceremony, arising early in the mornings, resulting in participants having virtually no unstructured free time during the day. Correctional officers within the camps typically have military titles by which inmates are to address them, and both inmates and correctional staff inmates usually don military-like uniforms. When inmates break rules or misbehave, the punishments are, in alignment with theories of deterrence, swift and certain, and most often physical in nature. Most boot camps also incorporate elaborate graduation ceremonies in which family members and other contacts from the participants’ outside environments are invited to attend. Additionally, some boot camp models do also include aftercare or reentry components to aid with reintegration into the community.

In terms of their effects on rates of recidivism, a great deal of the empirical assessment has largely found no difference between boot camps and traditional correctional approaches (Bottcher and Ezell, 2005; Kempinen and Kurlychek, 2003; Mackenzie, Wilson and Kider, 2001), and, in some cases, negative effects have been seen, with boot camp participants being more likely to commit future offences (Farrington and Welsh, 2005; Wright and Mays, 1998). However, as mentioned, some boot camp programs have also incorporated aftercare or re-entry programs aimed at helping participants successfully make the transition back into society, a component that has achieved some seemingly positive effects (Kurlychek and Kempinen, 2006; Mackenzie, Wilson and Kider, 2001). As with the Scared Straight programs, the popularity and widespread use of boot camps is suggested to be more related to their “common sense” approach and relatively low operating costs rather than to the effects they achieve (Cullen, Blevins, Trager and Gendreau, 2005). As a note, it has been asserted that a great deal of the evaluative research regarding boot camps is limited, in that there are often scant comparative data as well as program implementation issues (Botcher and Ezell, 2005).

A recent study conducted by Bottcher and Ezell (2005) used an experimental design to examine the California Youth Authority’s (CYA) juvenile boot camp and intensive parole program, LEAD (an acronym alluding to expected participant outcomes— leadership, esteem, ability, and discipline). Two LEAD sites were examined, both incorporating the typical rigorous components described previously. In addition, both sites included group counselling sessions and substance abuse programs. The study utilized comparison group data generated by CYA researchers regarding youth assigned to traditional institutions for the impact evaluation of the LEAD program. Long-term official follow-up arrest data, excluding parole violations, was examined for both the LEAD and control groups for an average time period of 7.5 years. As a note, youths who dropped out of the LEAD program before completion (slightly over 25 per cent) were included in the study’s analyses. The study’s total sample size consisted of 632 youth offenders, with 348 being in the LEAD group and 284 in the control group. All of the youths were male, with the average age being 17.5 years, and both groups were posited as being racially diverse, made up primarily of white, Latino, and Black youth. Initial commitment offences ranged from drug and minor offences to assaults, with the majority (70 per cent) being property-related. Finally, the majority of youth in both the control (83 per cent) and LEAD (83.4 per cent) groups were first-time offenders.

The main findings of the study showed that the LEAD boot camp program had no significant effect on recidivism rates in comparison with more traditional correctional settings. Youth in both groups displayed similar levels in terms of average time to first arrest as well as overall arrest rates during the first year and all available years following release to parole (Bottcher and Ezell, 2005).

A relatively recent and comprehensive meta-analysis examining 29 juvenile and young adult boot camp assessment studies was conducted in order to determine what effects, if any, boot camps might have on rates of recidivism (Mackenzie, Wilson and Kider, 2001). Ultimately, the meta-analysis tested for recidivism effects across 44 boot camp versus comparison group samples. As was the case in the previous study, technical parole violations were excluded from extracted measures of recidivism, offering a more conservative testing for effects on rates of recidivism. The authors’ criteria for including a study in the analyses included 1) a need for the program to incorporate a militaristic environment; 2) the presence of a comparison control group either receiving community supervision or serving time in a more traditional correctional facility; 3) inclusion of participants who were convicted or adjudicated; and 4) the inclusion of post-program measures of criminal activity, such as arrest or conviction.

While results of the meta-analysis were mixed, the authors ultimately determined that there were no overall significant differences between boot camp participants and comparison samples (Mackenzie, Wilson and Kider, 2001). Specifically, 12 of the 29 studies deemed that there were no significant differences between offenders exposed to boot camps versus those experiencing more traditional forms of sanctions. However, in eight of the 29 studies, it was found that individuals in comparison groups actually displayed significantly lower rates of recidivism, suggesting that, at least in these cases, boot camps may have increased the likelihood of future offending. In contrast though, nine of the 29 studies found that those individuals involved in boot camp programs did exhibit significantly lower rates of recidivism. Finally, it was noted that “the only program characteristic exhibiting a strong effectiveness of the boot camp programs was the presence of an aftercare treatment component for the adult programs” (Mackenzie, Wilson and Kider, 2005: 135). The authors suggest that such components may therefore be useful in reducing recidivism rates among adult offender populations.

Ultimately, the validity and generalizability of any meta-analytic findings is reliant upon the quality of the studies of which it is composed and, as such, this study is not without its limitations. Using a five-point qualitative methodological rating scale, the authors report that 19 of the 44 evaluations (43 per cent) were judged to be methodologically solid (score of 4), while eight of the evaluations were deemed as being of poor methodological quality (score of 2). The remaining 17 evaluations were of reasonable methodological quality (score of 3). The most notable methodological shortcomings had to do with a lack of random assignment (39 of 44 evaluations) and no statistical controls being used during analyses (27 of 44 evaluations).

A relatively large-scale study conducted by Wright and Mays (1998) highlights the possibility that boot camps can actually serve to increase rates of recidivism. The study examined recidivism rates among 1,937 non-violent first-time male offenders in Oklahoma who were either placed in a boot camp (n = 560), served traditional prison sentences (n = 802), or were placed on probation (n = 575). The offenders ranged in age from 16 to 47 years; however, over 90 per cent of them were between the ages of 19 and 26 years. The authors controlled for the influence of such variables as age, race and type of offence on recidivism. Offence types were all non-violent, with the majority of them (72 per cent) being property offences and the remaining offences being drug-related.

The results of the study found that youth offenders placed in boot camp settings were significantly more likely to recidivate upon release in comparison with youth who served time in traditional prison sentences or those on probation, a general effect that has also been found in other studies (Cook and Spirrison, 1992; Farrington and Welsh, 2005; Wright and Mays, 1998). Additionally, within the boot camp group, it was found that those who had been sentenced for property offences were more likely to recidivate than drug offenders

As mentioned, it has been found that incorporating aftercare or re-entry components into boot camp programs, focused on helping offenders reintegrate into the community upon release, may be somewhat effective in lowering offender recidivism rates (Josi and Seachrest, 1999; Kurycheck and Kempinen, 2006; Mackenzie, Wilson, and Kider, 2001). For example, a recent study specifically evaluated the residential aftercare component of the Queanna six-month Motivational Boot Camp, located in Queanna, PA (Kurycheck and Kempinen, 2006). An experimental group was exposed to 90 days of residential aftercare involving such things as cognitive-behavioural therapy, substance abuse counselling, employment preparation, and, in most cases, individual treatment plans. It is noted, though, that not everyone experienced the same aftercare.

The study sample consisted of 720 offenders who graduated from the Motivational Boot Camp Program between April 2001 and December 2002. While all of the offenders in the sample (N = 720) graduated from the boot camp, 383 of them did so prior to a mandatory policy change implementing the 90-day residential aftercare component, with the remaining 337 offenders having been exposed to this aftercare component. The average age of the offenders at admission for both groups was 25 years, with the majority being male (over 95 per cent). Additionally, for each group the average number of prior arrests was three, while the average age at first arrest was 18 years. The two groups were not found to differ significantly on any other major demographic variables, including sex (over 95 per cent male), race, marital status, education or current offence (mostly drug offences). In order to test for possible effects of the aftercare component, post-graduation recidivism measurements were conducted for a two-year follow-up period.

The results of the study indicate that offenders exposed to the 90-day aftercare component had significantly lower rates of recidivism at the six-month, one-year, and two-year time periods (Kurlychek and Kempinen, 2006). Subsequently, the authors suggest that because “the two year re-arrest rate for the experimental (aftercare) group was approximately equal to the one-year re-arrest rate for the control (no aftercare) group,” aftercare seemingly both reduces the risk of recidivism and lengthens the time between release and re-arrest for those who do recidivate. As always, there are study limitations to be considered. First, this is an evaluation of only one aftercare program and, although the findings are consistent with previous ones regarding aftercare components (Brown et al., 2001; Hiller, Knight and Simpson, 1999; Mackenzie, Wilson and Kider, 2001), it is difficult to generalize them to boot camps or aftercare programs in general. Second, this study does not speak to what types of aftercare programs are most effective, only finding that some form of aftercare appears to be better than none at all. Finally, and most pertinent to the goals of this report, these findings suggest only that aftercare components may be effective when dealing with adult offender populations, while it remains unclear how they may fare when dealing with juvenile ones. However, it is not unreasonable to think that similar or even more positive effects might be seen when exposing juvenile offenders to aftercare programs, as it has generally been found that, compared with adults, juveniles are more receptive to the types of rehabilitative components generally found in aftercare and re-entry programs.

In sum, the majority of evidence regarding boot camps suggests that they generally have either no effect on subsequent offending rates (Bottcher and Ezell, 2005; Kempinen and Kurlychek, 2003; Mackenzie, Wilson and Kider, 2001) or, in some cases, that they may actually lead to increased rates of recidivism (Farrington and Welsh, 2005; Wright and Mays, 1998). However, positive effects may perhaps be achieved by incorporating aftercare programs (Kurlychek and Kempinen, 2006; Mackenzie, Wilson and Kider, 2001), although it is currently unclear what the most effective aftercare programs might look like or how they might be best tailored for increasing their effects among juveniles and specific types of offenders. As such, further empirical research regarding aftercare might help to better distinguish among these programs. Finally, as noted in discussing the meta-analysis conducted by Mackenzie, Wilson and Kider (2001), much of the existing research regarding boot camps is hampered by methodological issues such as a lack of random assignment and statistical controls.


The empirical evidence regarding harsh deterrence-based approaches towards youth crime is, for the most part, inconsistent, with negative outcomes being seen in a number of different studies (Baron, 1998; Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Johnson, Simons and Conger, 2004; Mackenzie, Wilson and Kider, 2001; Myers, 2001, 2003b).

Regarding youth waiver policies, the results of some studies demonstrate that youth transferred into the adult criminal justice system may be at a higher risk to recidivate (Bishop and Frazier, 2000; Fagan, Kupchick and Liberman, 2003; Myers 2001, 2003b), and to do so more quickly than retained youth (Fagan, Kupchick and Liberman, 2003; Myers, 2003b). Additionally, there is evidence to suggest that such policies have little or no effect on juvenile crime rates in general (Jensen and Metsger, 1994; Ruhland, Gold and Heckman, 1982) and that their goals of harsher punishment certainty and severity are perhaps not even being met (Dawson, 1992; Rudman et al., 1986; Fritsch, Caeti and Hemmens, 1996; Lemmon, Austin, Verrecchia, and Fetzer, 2005). It appears too that the deterrent effects of harsher, more adult-like sanctions fluctuate with age (Cauffmen and Steinberg, 2000; Ruhland, Gold and Heckman, 1982), suggesting the need for individual assessments and judicial discretion regarding youth waivers.

It has been found that youth who perceive or experience higher severity of punishment are not necessarily deterred from committing deviant acts (Abrams, 2006; Greilich, Trager and Chishalm, 1982; Johnson, Simon and Conger, 2004; Teevan, 1976). Moreover, there is evidence to suggest that fear of punishment is diminished with a variety of lifestyle factors, including poverty, elevated drug and alcohol use, missing normative constraints, and deviant peer associations (Baron, 1998; Johnson, Simons and Conger, 2004).

Among the least-supported deterrence-based approaches are “shock” programs such as Scared Straight and boot camps, with many studies finding that they contribute to the likelihood of future offending (Bottcher and Ezell, 2005; Farrington and Welsh, 2005; Finckenauer, 1982; Kempinen and Kurlychek, 2003; Petrosino, Turpin-Petrosino and Buehler, 2003; Wright and Mays, 1998).

In terms of post-confinement, the most positive effects may be achieved by incorporating aftercare and re-entry programs (Kurlychek and Kempinen, 2006; Mackenzie, Wilson and Kider, 2001). Additionally, incorporating individual and family counselling, drug and alcohol interventions, anger management, and cognitive-behavioural sessions aimed at increasing self-esteem, moral inhibition, and feelings of responsibility into youth offender approaches has been found to produce positive effects (Abrams, 2006; Pogarsky, Kim and Paternoster, 2005; Teevan, 1976). Ultimately however, there does not appear to be one approach that is most effective in responding to youth crime. Rather, employing multiple and blended strategies incorporating both rehabilitative and deterrent goals is most likely to achieve the most desirable results.


Abrams, L.S. (2006). Listening to juvenile offenders: Can residential treatment prevent recidivism? Child and Adolescent Social Work Journal, 23 (1), 61–85

Baron, S. W. (1995). Serious Offenders. In James Creechan and Robert A. Silverman (Eds.), Canadian Delinquency (pp.135–148). Toronto: Prentice Hall.

Baron, S. W. and L. Kennedy. (1998). Deterrence and homeless male street youths. Canadian Journal of Criminology, 40, 26–60.

Bishop, D. M. and C. Frazier. (2000). Consequences of transfer. In J. Fagan and F. E. Zimring (Eds.), The Changing Borders of Juvenile Justice: Transfer of Adolescents to the Criminal Court (pp. 227–276). Chicago: University of Chicago Press.

Bottcher, J. and M. E. Ezell. (2005). Examining the effectiveness of boot camps: A randomized experiment with a long-term follow up. Journal of Research in Crime and Delinquency, 42 (3), 309–332.

Brown, B. S., K. E. O’Brady, D. Battjes, E. V. Farrel, N. P. Smith and D. N. Nurco. (2001). Effectiveness of a stand-alone aftercare program for drug-involved offenders. Journal of Substance Abuse Treatment, 21, 185–192.

Cauffmen, E. and L. Steinberg. (2000). (Im)maturity of judgment in adolescence: Why adolescents may be less culpable than adults. Behavioral Sciences and the Law, 18, 741–760.

Cook, D.D. and C.L. Spirrison. (1992). Effects of a prisoner-operated delinquency deterrence program: Mississippirsquos Project Aware. Journal of Offender Rehabilitation, 17, 89–99.

Corrado, R. R., I. M. Cohen, W. Glackman and C. Odgers. (2005). Serious and violent young offenders’ decisions to recidivate: An assessment of five sentencing models. Crime and Delinquency, 49(2), 179–200.

Cullen, F. T., K. R. Blevins, J. S. Trager and P. Gendreau. (2005). The rise and fall of boot camps: A case study in common-sense corrections. In Brent B. Benda and Nathaniel J. Pollone (Eds.), Rehabilitation Issues, Problems, and Prospects in Boot Camp (pp. 53–70). New York: Haworth Press.

Dawson, R. (1992). An empirical study of kent style juvenile transfers to criminal court. St. Mary’s Law Journal, 23, 975–1054.

Doob, A. N. and C. Cesaroni. (2004). Responding to Youth Crime in Canada. Toronto: University of Toronto Press.

Fagan, J., A. Kupchick and A. Liberman. (2003). Be Careful What You Wish For: The Comparative Impacts of Juvenile Versus Criminal Court Sanctions on Recidivism Among Adolescent Felony Offenders. New York: Columbia University.

Farrington, D. P. and B. C. Welsh. (2005). Randomized experiments in criminology: What have we learned in the last two decades? Journal of Experimental Criminology, 1, 9–38.

Feld, B. C. (1989). Bad law makes hard cases: Reflections on teen-aged axe murderers, judicial activism, and legislative default. Law and Inequality: A Journal of Theory and Practice, 8, 1–101.

Finckenauer, J. O. (1982). Scared Straight and the Panacea Phenomenon. Englewoood Cliffs, NJ: Prentice Hall.

Fritsch, E. J., T. J. Caeti and C. Hemmens. (1996). Spare the needle but not the punishment: The incarceration of waived youth in Texas prisons. Crime and Delinquency, 42 (4), 593–609.

Greenwood, P. W. (2006). Changing Lives: Delinquency Prevention as Crime-Control Policy. Chicago: The University of Chicago Press.

Greilich, W. J., B. Trager and J. Chishalm. (1982). Punishing Juveniles: What Works? Boston: Massachusetts Criminal Justice Committee.

Hiller, M. L., K. Knight and D. D. Simpson. (1999). Prison-based substance abuse treatment, residential aftercare, and recidivism. Addiction, 94, 833–842.

Jensen, E. L. and L. K. Metsger. (1994). A test of the deterrent effect of legislative waiver on violent juvenile crime. Crime and Delinquency, 40(1), 96–104.

Johnson, L. M., R. L. Simons and D. D. Conger. (2004). Criminal justice system involvement and continuity of youth crime: A longitudinal analysis. Youth and Society, 36(1), 3–29.

Josi, D. A. and D. K. Seachrest. (1999). A pragmatic approach to parole aftercare: Evaluation of a community reintegration program for high-risk youthful offenders. Justice Quarterly, 16, 51–80.

Kempinen, C. and M. C. Kurlychek. (2003). An outcome evaluation of Pennsylvania’s boot camp: Does rehabilitative programming within a disciplinary setting reduce recidivism? Crime and Delinquency, 4, 581–602.

Kurlychek, M. and C. Kempinen. (2006). Beyond boot camp: The impact of aftercare on offender reentry. Criminology and Public Policy, 5 (2), 363–388.

Lemmon, J. H., T. L. Austin, P. J. Verrecchia and M. Fetzer. (2005). The effect of legal and extralegal factors on statutory exclusion of juvenile offenders. Youth Violence and Juvenile Justice, 3 (3), 214–234.

Lochner, L. (2003). Individual Perceptions of the Criminal Justice System. Cambridge, MA: National Bureau of Economic Research.

Mackenzie, D. L., D. B. Wilson and S. B. Kider. (2001). Effects of correctional boot camps on offending. The Annals of the American Academy of Political and Social Sciences, 578, 126–143.

Moffitt, T. E. (1993). Adolescence-limited and life-course persistent antisocial behaviour: A developmental taxonomy. Psychological Review, 100(4), 675–701.

Myers, D. L. (2001). Excluding Violent Youths From Juvenile Court: The Effectiveness of Legislative Waiver. New York: LFB Scholarly Publishing LLC.

Myers, D. L. (2003a). Adult crime, adult time: Punishing violent youths in the adult criminal justice system. Youth Violence and Juvenile Justice, 1, 173–197.

Myers, D. L. (2003b). The recidivism of violent youths in juvenile and adult court: A consideration of selection bias. Youth Violence and Juvenile Justice: An Interdisciplinary Journal, 1, 79–101.

Paternoster, R. and L. Iovanni. (1989). The labeling perspective and delinquency: An elaboration of the theory and assessment of the evidence. Justice Quarterly, 6(3), 359–394.

Peterson-Badali, M., M. D. Ruck and C. J. Koegl. (2001). Youth court dispositions: perceptions of Canadian juvenile offenders. International Journal of Offender Therapy and Comparative Criminology, 45(5), 593–605.

Petrosino, A., C. Turpin-Petrosino and J. Buehler. (2003). Scared straight and other juvenile awareness programs for preventing juvenile delinquency: A systematic review of the randomized experimental evidence. Annals of the American Academy of Political and Social Science, 589, 41–60.

Petrosino, A., C. Turpin-Petrsino and J. O. Finckenauer. (2000). Well-meaning programs can have harmful effects! Lessons from experiments in Scared Straight and other like programs. Crime and Delinquency, 46, 354–379.

Phillips, L. and H. L. Votey, Jr. (1987). The influence of police interventions and alternative sources on the dynamic process of choosing crime as a career. Journal of Quantitative Criminology, 3, 251–273.

Pogarsky, G., K. Kim and R. Paternoster. (2005). Perceptual change in the national youth survey: Lessons for deterrence theory and offender decision-making. Justice Quarterly, 22(1), 1–29.

Rudman, C., E. Hartstone, J. Fagan and M. Moore. (1986). Violent youth in adult court: Process and punishment. Crime and Delinquency, 32, 75–96.

Ruhland, D. J., M. Gold and R. J. Hekman. (1982). Deterring juvenile crime: Age of Jurisdiction. Youth and Society, 13(3), 353–375.

Schneider, A. L. (1990). Deterrence and Juvenile Crime: Results from a National Policy Experiment. New York: Springer-Verlag.

Singer, S. I. and D. McDowell. (1998). Criminalizing delinquency: The deterrent effects of the New York Juvenile Offender Law. Law and Society Review, 22(3), 521–36.

Teevan, J. J. (1976). Deterrent effects of punishment for breaking and entering and theft. Canadian Journal of Criminology and Corrections, 19(2), 46–68.

Wright, D. T. and G. L. Mays. (1998). Correctional boot camps, attitudes, and recidivism: The Oklahoma experience. Journal of Offender Rehabilitation, 28, 71–87.

Zimring, F. E. and J. Fagan. (2000). Transfer policy and law reform. In J. Fagan and F.E. Zimring (Eds.), The Changing Borders of Juvenile Justice: Transfer of Adolescents to the Criminal Court (pp. 407–424). Chicago: University of Chicago Press.

4 This chapter was written with the assistance of James Dorion, MA, Centre of Criminology, University of Toronto.


Volume 1. Findings, Analysis and Conclusions

Volume 2. Executive Summary

Volume 3. Community Perspectives Report

Volume 4. Research Papers

Volume 5. Literature Reviews