Adult Punishment for Juvenile Offenders: Does it Reduce Crime?

Richard Redding. Handbook of Children, Culture, and Violence. Editor: Nancy E Dowd, Dorothy G Singer, Robin Fretwell Wilson. Sage Publication. 2008.

The decade prior to 1994 saw a significant increase in violent juvenile crime, high-profile cases of serious and violent crimes committed by juveniles and young adults, and the resulting perception that America was experiencing a juvenile crime wave unlike anything in its history (Zimring, 1998). Based on the projected growth in the juvenile population during the early 21st century, some predicted a coming storm of youth violence (Welch, Fenwick, & Roberts, 1997) and the emergence of young “super-predators” (Dilulio, 1995).

The public, and perhaps even more so policy makers, demanded action. There was a rough consensus among legislators that the juvenile court was too lenient, that serious offenders were beyond rehabilitation and must be incarcerated to ensure public safety, and that juveniles were as culpable for their crimes as adults (Redding, 1997). Thus, states passed legal reforms designed to “get tough” on juvenile crime. The most significant change was states’ revision of their transfer laws to expand the type of offenses and offenders eligible for transfer from the juvenile court to the adult court for trial and sentencing. Changes also occurred at the federal level with the passage of the Violent Crime Control and Law Enforcement Act of 1994, which allowed the transfer of 13-year-olds who committed crimes with firearms on federal property. Congressman Bill McCollum, a key sponsor of the federal legislation, said that “in America today, no population poses a greater threat to public safety than juvenile criminals” (Lacayo & Donnelly, 1997, p. 26).

This chapter discusses the research on the general and specific deterrent effects of transferring juveniles for trial in adult criminal court, identifies gaps in our knowledge base that require further research, discusses the circumstances under which effective deterrence may be achieved, and examines whether there are effective alternatives for achieving deterrence other than adult sanctions for serious juvenile offenders. As a backdrop to this analysis, this chapter first examines the role of public opinion in shaping the “get tough” policies, and how policy makers have misunderstood and perceived support for those policies.

The Role of Public Opinion

During the 1980s and 1990s, the public appeared to support the new approach to juvenile crime. Voters passed state propositions allowing more juveniles to be tried as adults (Beresford, 2000). Consider California’s Proposition 21, enacted by voter initiative, which lowered the age for transfer from 16 to 14 and shifted discretion for making transfer decisions from juvenile court judges to prosecutors (Gang Violence and Juvenile Crime Prevention Act of 1998). The 1993 Gallup Poll showed that 73% of respondents were in favor of trying violent juveniles as adults, and influential public officials like Los Angeles County District Attorney Gil Garcetti proclaimed, “we need to throw out our entire juvenile justice system” (Redding, 1997, p. 712). Alfred Regnery, administrator of the Office of Juvenile Justice and Delinquency Prevention in the Reagan administration, argued that juvenile offenders were “getting away with murder,” that juvenile offenders “are criminals who happen to be young, not children who happen to commit crimes” (Regnery, 1985, p. 65).

Perhaps due in part to media hype, much of the public continues to believe that the juvenile crime rate remains high (Shepherd, 1999), reflecting an alarmist reaction to crime generally (Welch et al., 1997). But the public is misinformed about the juvenile crime problem. Since 1993, the decrease in juvenile crime has been three times greater than the decrease in adult crime (Snyder, 2004). Because of this decline, juvenile crime rates are now comparable to what they were in 1980 (Blumstein, 2001), before the policy shift toward increased transfer and other punitive responses to juvenile offenders. The juvenile arrest rate for violent crime, after peaking in 1994, has now reached its lowest level since 1980 and is 29% lower than it was in 1993. There has been a 64% decrease in the number of juveniles arrested for murder (Snyder, 2004), likely due to the declining crack cocaine market and tougher anti-firearms laws and policing programs (Butts & Travis, 2002; Zimnng, 1998).

Despite erroneous public perceptions about juvenile crime rates, however, the public does not favor abandoning the rehabilitative ideal of juvenile justice in favor of wholesale punitive responses to juvenile offenders. While recent polls show that between 58 and 91% of Americans favor (depending on the type of crime and poll) trying violent juvenile offenders as adults (The American Enterprise, 2001; Schwartz, Guo, & Kerbs, 1993; Wu, 2000), there is far less support for imposing adult sentences (Schwartz etal., 1993). Most still believe in the efficacy of the juvenile justice system and want it strengthened, favor early intervention and prevention programs along with rehabilitation over punishment for juvenile offenders, would reserve incarceration only for the most serious and violent offenders, want juvenile offenders tried as adults to receive rehabilitative treatment, and strongly disagree with the confinement of juveniles in adult prisons (Moon, Sundt, Cullen, & Wright, 2000; Schiraldi & Soler, 1998). Polls that do report high public support for punitive policies typically ask omnibus questions about juvenile crime (e.g., asking whether respondents favor “punishing violent juvenile offenders as adults”). More nuanced survey questions that provide respondents with information about particular cases or sentencing options (e.g., the background of a particular offender, rehabilitative programs available) reveal considerably less support for punitive sentencing policies (Stalans & Henry, 1994). While the public favors incarceration for those serious and violent offenders who most threaten public safety, they do not favor the wholesale imprisonment of young, nonviolent, or first-time offenders (Sundt, 1999). But policy makers consistently overestimate the public’s support for punitive policies (Latessa, 2004).

Legislative Changes in Transfer Laws and Their Impact

Transfer laws (also called waiver or certification laws), which transfer juveniles from the juvenile court to adult criminal court for trial and sentencing, exist in every state (Griffin, 2003; Redding, 1997). During the last 20 years, states have revised their transfer laws to lower the minimum age for transfer, increase the number of transferable offenses, expand prosecutorial discretion while reducing judicial discretion in transfer decision making, and expand the reach of laws requiring that certain juvenile offenders be automatically tried as adults (Redding, 2003; see Fagan & Zimring, 2000). In 1979, for example, only 14 states had “automatic” transfer laws, but by 1995, twenty-one states had such statutes, with 31 states having these laws by 2003 (Steiner & Hemmens, 2003). In addition, 13 states lowered the age at which juvenile court jurisdiction ends, to age 15 or 16 (Sanborn, 2003).

As a result of the legislative changes, the number of youth convicted of felonies in criminal courts and incarcerated in adult facilities has increased (Redding, 2003). The number reached a peak in the mid-1990s and has declined since (Puzzanchera, 2003), due in part to the decrease in juvenile crime. Despite the legislative changes in transfer laws, transfer remains relatively uncommon; less than 1% of all juvenile court cases are transferred (Puzzanchera, 2003).

According to the most recent data available, an estimated 5,600 youth were committed to state adult prisons in 1999, representing 2% of all new prison commitments. These numbers reflect a 26% decrease in youth commitments since the peak year of 1995, but nonetheless an overall increase of 70% between 1985 and 2000 (Sickmund, 2004). These youth are overwhelmingly male (96%) and most were 17 years old at the time of their commitment; 36% were White and 57% were African American. Sixty-two percent were incarcerated for person offenses, 22% for property offenses, 5% for public orders offenses (e.g., weapons possession), and 1% for drug offenses (Sickmund, 2004). Seventy-eight percent were released before their 21st birthday and 95% were released before their 25th birthday, with an average of about 2 years, 8 months served.

Prior research (before about 1990) was inconclusive on whether juveniles sentenced in adult criminal court received more severe sentences than those sentenced in the juvenile court for similar crimes (Kupchik, Fagan, & Liberman, 2003). But several more methodologically sophisticated studies confirm that, in recent years, transferred juveniles do receive tougher sentences (Kupchik et al., 2003; Kurlychek & Johnson, 2004). Data from some states indicates that they may even receive more severe sentences than adults convicted of the same crime (Kurlychek & Johnson, 2004; Virginia Department of Criminal Justice Services, 1996). Indeed, the nationwide policy shift toward transferring juveniles to the criminal court is based partly on the assumption that more punitive, adult sentences will follow, and that these sentences will act as a general or specific deterrent to juvenile crime.

In terms of specific deterrence—that is, whether trying and sentencing juvenile offenders as adults decreases the likelihood that they will recidivate—seven recent large-scale studies in various jurisdictions have all found higher recidivism rates among juveniles convicted for violent offenses and sentenced as adults, when compared to similar juvenile offenders tried in the juvenile court. On the other hand, it is unclear whether transfer affects recidivism among nonviolent property offenders (Redding, 2003; Redding & Mrozoski, 2005). With respect to general deterrence—that is, whether transfer laws deter would-be juvenile offenders—the picture is considerably less clear, because there are only three systematic empirical studies, and the studies, which produced conflicting findings, were conducted 15 to 20 years ago (Redding, 2003; Redding & Mrozoski, 2005). The issues of general and specific deterrence are discussed in the following sections.

General Deterrence: Do Transfer Laws Deter and Prevent Juvenile Crime?

Two well-designed studies conducted in the 1980s found that transfer laws did not reduce juvenile crime. On the contrary, Jensen and Metsger’s (1994) time-series analysis found a 13% increase in arrest rates for violent juvenile crime in Idaho after the state implemented its automatic transfer law. In a similar analysis, Singer and McDowall (1988) found that a New York state law that automatically sent violent juvenile offenders to criminal court (by lowering the age for criminal court jurisdiction) had no deterrent effect, even though the law was widely applied and publicized in the media. In addition, brochures were sent to public schools announcing the law and the risks juveniles faced, and juvenile court judges warned youth about the risks of committing violent offenses (S. Singer, 2004, personal communication). The limited evidence available at the time suggested that juveniles in New York were aware of the automatic transfer law (Singer & McDowall, 1988).

On the other hand, the results of a multi-state economic analysis for the years 1978 to 1993 suggests that trying juveniles as adults may have moderate deterrent effects (Levitt, 1998). The study found a 25% decrease in violent juvenile crime and a 10 to 15% decrease in property crime committed by juveniles in states that lowered the jurisdictional age for criminal court from 18 to 17. The greatest decreases in crime were found in states having the greatest disparity in punishment severity between the criminal and juvenile courts. These data suggest the deterrent effect of criminal court sanctions. The researcher concluded that, “the estimated decrease in crime associated with incarcerating an additional juvenile is at least as large as the corresponding reduction in crime for adults” (Levitt, 1998, p. 1181). The same study, however, found no relationship between the punitiveness of juvenile court sanctions and later criminal offending in adulthood.

Levitt’s (1998) aggregate analysis of crime rates across states differs substantially from the methodology used by Jensen and Metsger (1994) and Singer and McDowall (1988), which used careful offender case comparisons and quasi-experimental controls to study crime rates in a particular state. In addition, the Levitt study specifically examined the effects of criminal court jurisdiction (when youth reached the age of majority), rather than the effects of transfer laws per se. Unlike knowing that one could be tried as an adult for crimes committed while a juvenile, which most juveniles do not seem to realize (as discussed below), “it is probably well known that dramatically greater penalties for all offenses are imposed once a juvenile reaches the age of majority” (Robinson & Darley, 2004, p. 177).

Data from some communities also suggest that transfer laws deter juvenile crime. In Jacksonville, Florida, the juvenile arrest rate decreased 30% and the juvenile violent crime rate decreased 44% between 1993 and 1994, after the local prosecutor instituted aggressive policies to prosecute serious juvenile offenders in criminal court (Bennett, Dilulio, & Walters, 1996).

Only a few studies have interviewed juvenile offenders, however. Before the widespread expansion of transfer laws, Glassner, Ksander, Berg, and Johnson (1983) reported the results of interviews with a small number of juvenile offenders in New York, who said they had decided to stop offending once they reached the age at which they knew they could be tried as adults. A recent small-scale study interviewed 37 juvenile offenders who had been transferred to criminal court, for armed robbery or murder in Georgia. The study examined their knowledge and perceptions of transfer laws and criminal sanctions (Redding & Fuller, 2004). Georgia had undertaken a public awareness campaign to inform juveniles about the state’s new automatic transfer law. Nonetheless, juveniles reported being unaware of the transfer law; only 30% knew that juveniles who committed serious crimes could be tried as adults. Even among those who knew about the law, none expected that it would be enforced against them for the serious crime they committed. On the contrary, many thought they would only get slap-on-the-wrist sentences from the juvenile court. These results are consistent with those in a recent Canadian study finding that many juvenile offenders did not think that they would receive a serious punishment if caught (Peterson-Badali, Ruck, & Koegl, 2001).

How Might Transfer Laws Have Deterrent Effects on Juvenile Crime?

There are likely two explanations for juvenile offenders’ inaccurate perceptions about the risk of being tried as adults. First, juveniles’ psychosocial immaturity (Beckman, 2004; Scott, Reppucci, & Woolard, 1995; Steinberg & Cauffman, 1996) may make them less likely to perceive accurately the likelihood of apprehension and serious punishment. Second, the relatively mild sanctions the juveniles had previously received from the juvenile court may have communicated the wrong message about the consequences of committing crimes as a juvenile. As one juvenile interviewed by Redding and Fuller (2004) explained, “[Being tried as an adult] showed me it’s not a game anymore. Before, I thought that since I’m a juvenile I could do just about anything and just get six months if I got caught” (p. 39). The juvenile justice system may fail to provide meaningful sanctions until it is too late. “How is an offender supposed to judge which ‘last chance to go straight— is really his last? He is likely to keep testing the system until it lands on him hard…. [Thus], every detected non-trivial violation of law ought to lead to some nontrivial deprivation of liberty” (Kleiman, 1999, p. 13). And as one Los Angeles Assistant District Attorney said, “You talk to youngsters … and they tell you, repeatedly, that they got away with so much—that they commit crimes, but aren’t arrested, and if they are arrested, when they are brought into [juvenile] court, nothing happens” (Michaelis, 2001, p. 309).

An initial light sanction by the juvenile court followed by ever-increasing punishment severity for subsequent offenses, up to and including criminal sentences, may have counterproductive effects.

… Judges often do not send youthful offenders to prison because the experience may increase their future likelihood of committing criminal offenses…. However, from the deterrence perspective, it may bring about the “hardening to punishment” effect observed in animals, in which an escalating series of punishments, if it begins at a level that is ineffective in controlling the initial transgression, simply conditions the person to tolerate the increasing punishments, without reducing the rate of transgressions. (Robinson & Darley, 2004, p. 187)

If true, this offers an especially compelling rationale for ensuring that the initial sanctions applied by the juvenile court have enough bite. Nonetheless, the studies (e.g., Minor, Hartmann, & Terry, 1997; Wooldredge, 1988) are mixed on whether the initial actions by the juvenile court (diversion vs. adjudication and sentencing) impact recidivism, at least vis-à-vis first-time or relatively non-serious offenders.

The juvenile offenders interviewed by Redding and Fuller (2004) indicated that being tried as adults taught them, apparently for the first time, that their criminal behavior had real consequences. The challenge would be how to deliver this “wake-up call” without inflicting the permanently disfiguring (Zimring, 2000) and counter-rehabilitative effects of the criminal justice system. Scared straight programs, shock incarceration programs, and boot camps have all proved ineffective in reducing recidivism in juvenile offenders (Finckenauer & Gavin, 1999). But knowing they could be tried and sentenced as adults, juvenile offenders say, may have prevented them from committing the crime (Redding & Fuller, 2004). We cannot know whether their introspections are accurate. A recent study with serious juvenile offenders found a correlation between their self-reported likelihood of committing a future offense and the number of offenses they committed after their release (Corrado, Cohen, Glackman, & Odgers, 2003), mirroring a similar study with adult offenders (Burnett, 2000). The Redding and Fuller (2004) study, though conducted on a small sample in only one jurisdiction, provides some limited evidence that juvenile offenders may calibrate their behavior as a function of the perceived likelihood of receiving an adult punishment.

Some have argued that the recent decline in crime rates is due to “get tough” policies (see Bennett et al., 1996; Scheidegger & Rushford, 1999). There has been relatively little research on deterrence with respect to juveniles, and the results have been mixed, with some studies finding deterrence effects for certainty and severity of punishment and others finding no such effects or even negative effects (Corrado et al., 2003). Corrado et al.’s (2003) recent study with serious juvenile offenders incarcerated in a maximum-security facility found a negative relationship between intent to reoffend and sentence severity, with evidence that offenders made “some explicit calculations about the advantages and disadvantages of committing future crimes” (p. 197).

Given the paucity of research on juvenile crime deterrence, this chapter turns now to research with adults that is relevant to the question of whether we should expect transfer laws to deter juvenile crime. A comprehensive analysis of the extant research on criminal deterrence conducted in 1998 by the Institute of Criminology at Cambridge University concluded: “The studies plainly suggest that when potential offenders are made aware of substantial risks of being punished, many of them are induced to desist” (Von Hirsch, Bottoms, Burney, & Wikstrom, 1999, p. 47). The perceived certainty of punishment (e.g., being apprehended and tried as an adult) appears to affect crime rates. It is unclear, however, whether the severity of punishment (e.g., receiving a substantial adult sentence) affects crime rates, though the limited evidence available suggests that it does not. Von Hirsch et al. (1999) speculate that this is because potential offenders typically have much more information about certainty of apprehension than sentence severity; studies show that offenders and the public generally know little about potential sentences and tend to greatly underestimate their severity (Robinson & Darley, 2004; Von Hirsch et al., 1999). Moreover, punishment is an uncertain future event that offenders tend to discount (whereas the short-term rewards of crime are salient), and relatively small or large changes in penalties may not be calibrated to offenders’ thresholds for offending. At the same time, “future contingent costs may be discounted less, if their magnitude is sufficiently great and their likelihood of being incurred increases. Severe sentencing policies thus might possibly have an impact if coupled with much higher probabilities of conviction” (Von Hirsch et al., 1999, p. 48). Thus, one might suppose that transfer laws would serve a deterrent function if would-be juvenile offenders are made aware of such laws and if the laws are widely implemented, with convictions resulting in significant adult sentences.

But in order for criminal sanctions to have deterrent effects, potential offenders must (1) believe that there is a reasonable likelihood of getting caught, (2) know that the likelihood of a conviction and receiving a substantial sentence is significant or has increased, (3) believe that the penalty will be applied to them if caught, and (4) consider the risk of the penalty when deciding whether to offend (Von Hirsch et al., 1999). Moreover, the perceived costs of obeying the law must outweigh the perceived benefits of offending. Robinson and Darley (2004) argue that, for a variety of reasons, such conditions rarely are present in the real world. Youths’ psychosocial immaturity (e.g., impulsivity, risk-taking proclivity, short-term time perspective, limited ability to foresee future consequences, limited life experience and metacognitive skills) (Beckman, 2004; Scott, Reppucci, & Woolard, 1995; Steinberg & Cauffman, 1996) could make this rational-choice model of deterrence, which assumes that perceived consequences influence decisions about committing crime, less applicable to juvenile offenders (Schneider & Ervin, 1990). Consider each of these necessary preconditions for successful deterrence in the context of juvenile offending. A law cannot act as a deterrent if the targeted population is unaware that the law exists or does not believe it will be enforced. Recall Redding and Fuller’s (2004) finding that few violent juvenile offenders knew they could be tried as adults, none thought it would happen to them, and few thought they would face serious punishment. Moreover, few reported thinking about the possibility of getting caught when they committed the offense. It seems that offenders generally underestimate the risk that they will be caught, thinking instead that they will avoid the mistakes that ensnarled others (Robinson & Darley, 2004). Juveniles’ psychosocial immaturity makes it even less likely that they will perceive a significant risk of being convicted and sentenced as an adult.

Substantial further research is urgently needed to examine whether transfer laws have (or could have, given the appropriate conditions) the general deterrent effect of preventing juvenile crime. In particular, it is important to examine whether juveniles are aware of transfer laws, whether this awareness deters delinquent behavior, and whether they believe the laws will be enforced against them. In conjunction with such research, there is a need for better-designed and well-targeted public-awareness campaigns on the state and local levels designed to make would-be juvenile offenders aware of the consequences of serious and violent crime (Redding & Fuller, 2004), and for rigorous evaluations of their effectiveness. Such campaigns have proved effective in reducing adult crime in some contexts (e.g., Johnson & Bowers, 2003). Unfortunately, however, the few public-awareness campaigns instituted to inform juveniles about transfer laws have been of fairly limited scope and duration and of questionable effectiveness in targeting the population at risk (Redding & Fuller, 2004).

Specific Deterrence: Do Transfer Laws Decrease Offenders’ Recidivism?

Seven large-scale studies indicate that youth tried in adult criminal court for violent crimes have greater recidivism rates after release than those tried in juvenile court, though it is unclear whether transfer affects recidivism for property offenders.

Fagan (1996) examined the recidivism rates of 800 randomly selected 15- and 16-year-old juvenile offenders charged with robbery or burglary. Controlling for eight variables (prior offenses, offense severity, race, gender, age at first offense, case length, sentence length, and court), this natural experiment compared offenders charged in New Jersey’s juvenile courts with offenders charged in New York’s criminal courts under that state’s automatic transfer law. Both geographical areas shared similar demographic, socioeconomic, sociolegal, and crime-indictor characteristics. Thus, the study provides a direct comparison of recidivism rates as a function of whether cases are processed in juvenile or criminal court, without many of the sample-selection problems inherent in studies comparing cases within a single jurisdiction where prosecutors or judges decide which cases to transfer. Youth who had committed robbery and were sentenced in adult criminal court had a higher post-release recidivism rate than those tried in juvenile court, but the recidivism rates for burglary offenders tried in criminal and juvenile courts were similar. The findings on robbery offenders suggest that criminal court processing, irrespective of whether youth are incarcerated in juvenile or adult facilities, produces a higher recidivism rate. This finding is emphasized by the parallel finding that youth sentenced to probation in criminal court had a substantially higher recidivism rate than those incarcerated in the juvenile justice system (see also Mason & Chang, 2001).

Bishop and colleagues (Bishop, Frazier, Lanza-Kaduce, & Winner, 1996) compared the one-year recidivism rate of 2,738 juvenile offenders transferred to criminal court in Florida with a matched sample of 2,738 juvenile offenders who had not been transferred. Florida relies almost exclusively on transfer by prosecutors, whose transfer decisions are largely offense driven and made soon after arrest, before gaining access to information about the youth’s background. Therefore, it is less likely that the youth retained in the juvenile justice system had lower recidivism rates due to selection factors (Bishop & Frazier, 2000). Nonetheless, this study cannot completely rule out possible selection effects in some (and perhaps a significant number) of cases. The study, which controlled for seven variables (race, gender, age, most serious prior offense, number of referrals to juvenile court, number of charges, and most serious charge), found that the rearrest rates were higher (30 versus 19%) and the time until reoffending shorter (135 versus 227 days) for the transferred youth across seven offense types (ranging from violent felonies to minor misd meanors). Following the same Florida offenders 6 years after this initial study, Winner and colleagues (Winner, Lanza-Kaduce, Bishop, & Frazier, 1997) also found higher recidivism rates among those transferred to criminal courts, with the exception of property felons.

Controlling for demographic and offense-related variables (e.g., age of onset of offending, prior offenses, use of a firearm), Myers (2001) examined the recidivism rates of 557 violent juvenile offenders in Pennsylvania. Youth who were judicially transferred to criminal court were rearrested more quickly upon their return to the community than youth who were retained in the juvenile justice system during the same period. However, transferred youth who were incarcerated for longer periods had a lower recidivism rate upon release than those incarcerated for shorter periods. Similarly, Podkopacz and Feld (1996) compared transferred with non-transferred juvenile offenders in Minnesota, and found higher recidivism rates among those transferred.

The above five studies involved all three types of transfer laws (automatic, judicial, and prosecutorial), used fairly large sample sizes (557 to 5,476), employed different methodologies (natural experiment and matched groups), and were conducted in five different jurisdictions (Florida, New Jersey, New York, Minnesota, Pennsylvania). Yet they each had significant methodological limitations, primarily the inability to control completely for possible differential selection effects (vis-à-vis juveniles’ amenability to treatment and recidivism risk) between those cases retained in the juvenile court versus those that were transferred.

But armed with two very recent large-scale studies that better control for possible selection effects, we can now conclude with much greater confidence that transfer generally does increase recidivism, though remaining methodological limitations still do not allow for definitive conclusions. Fagan, Kupchik and Liberman’s (2003) recent finding of greater recidivism for transferred juveniles (charged with robbery, burglary, or assault) replicates Fagan’s (1996) previous study but with a larger data set (2,400 juveniles) and methodology that better controls for important variables relating to possible selection effects. As in the previous study, by controlling for sentence lengths, the study showed that criminal court processing per se (rather than differential sentences between the juvenile and criminal courts) increased recidivism. Similarly, Lanza-Kaduce, Frazier, Lane, and Bishop’s (2000) recent follow-up study to the Bishop et al. (1996) Florida recidivism study also replicated the previous findings of higher recidivism rates for transferred juveniles as a function of criminal court processing per se (rather than differential sentences), using better matching techniques to control for possible selection effects, more extensive recidivism data, and data drawn from six Florida judicial circuits in rural and urban jurisdictions.

Why Do Juveniles Tried as Adults Have Higher Recidivism Rates?

What explains the higher recidivism rates for violent juvenile offenders tried in criminal court? The stigmatization and other negative effects of labeling juveniles as convicted felons, the sense of resentment and injustice juveniles feel about being tried as adults, the decreased focus on rehabilitation and family support in the adult system, and the learning of criminal mores and behavior from adult criminals have all been singled out as possible reasons for the increased recidivism (see Bazemore & Umbriet, 1995; Thomas & Bishop, 1984; Winner et al, 1997). Moreover, a felony conviction also usually results in the loss of a number of civil rights and privileges (see Redding, 2003), further reducing the opportunities for employment and community réintégration.

Juveniles’ sense of injustice at criminal court processing may cause them to react defiantly through reoffending and only harden their concept of themselves as “criminals” (see Thomas & Bishop, 1984; Winner et al., 1997). “The concept of fairness appears to be an important variable in an individual’s perception of sentence severity and its subsequent relationship to future recidivism” (Corrado et al., 2003, p. 183). And, conduct-disordered adolescents, it seems, already have a sense of having been dealt an unfair hand by authority figures (Chamberlain, 1998). Bishop and Frazier (2000) interviewed 95 serious and chronic juvenile offenders in Florida who had been transferred to the criminal justice system and incarcerated in correctional facilities. Many of the juveniles felt a strong sense of injustice and resentment about being tried as adults:

Many experience the court process not so much as a condemnation of their behavior as a condemnation of them. Unlike the juvenile court, the criminal court failed to communicate that young offenders retain some fundamental worth. What the youths generally heard was that they were being punished not only because their behavior was bad but also because they were personifications of their behavior. Far from viewing the criminal court and its officers as legitimate, the juvenile offenders we interviewed saw them more often as duplicitous and manipulative, malevolent in intent, and indifferent to their needs. It was common for them to experience a sense of injustice and, then, to condemn the condemners. (Bishop & Frazier, 2000, p. 263)

These findings are consistent with those of Redding and Fuller (2004) who found that juveniles tried as adults clearly did not perceive transfer laws as being fair and just. Many felt that their juvenile status and immaturity dictated that they should be tried as juveniles, despite the serious crime they had committed. They also did not understand what the law was attempting to accomplish by trying them as adults and felt that they were somehow being treated differently than other similarly situated juveniles. Both perceptions contributed to their sense of unfairness.

An especially compelling explanation for the increased recidivism is the greatly reduced opportunities for meaningful rehabilitation in the adult criminal justice system and the hardening of youth who serve time in adult prisons. Bishop and Frazier’s (2000) recent study vividly portrays the differences between juvenile and adult correctional facilities. “Despite the punitive rhetoric” of juvenile justice in Florida (alongside Texas, the state having the highest per capita number of juveniles tried in criminal court), they found that the juvenile correctional institutions were treatment oriented and adhered to therapeutic models of rehabilitation (Bishop & Frazier, 2000, p. 255). Juveniles in these facilities had positive feelings about the staff, whom they felt cared about them and taught them appropriate behaviors. In contrast, Florida prisons were clearly custodial in nature (Annino, 2000), and the juveniles in adult prisons reported that much of their time was spent learning criminal behavior from the inmates and responding to pressure to prove how tough they were. They also were much more fearful of being victimized than they had been in juvenile facilities; more than 30% had been assaulted or had witnessed assaults by prison staff. Notably, most of the juveniles incarcerated in juvenile facilities felt confident that they would not reoffend after release, often crediting the staff with helping them make this positive change. But only a third of the juveniles in adult prisons said that they would not reoffend. In sum, “compared to the criminal justice system, the juvenile system seems to be more reintegrative in practice and effect” (Bishop & Frazier, 2000, p. 265). This is well described by Forst, Fagan, and Vivona’s (1989) study, which found that unlike adult prisons, counseling in juvenile facilities was provided by line staff as part of their regular duties. Youth in juvenile facilities gave higher marks than youth in adult facilities to the available treatment and case management services, which youth in detention described as helpful in providing counseling, obtaining needed services, encouraging participation in programs, teaching the consequences of rule-breaking, and deepening their understanding of their problems.

Adult prisons are unlikely to provide an environment conducive to rehabilitation, for either adult or juvenile offenders. A prisoner’s experience is “from the outset, an experience of being violently dominated, and it is colored from the beginning by the fear of being violently treated” (Cover, 1986, p. 1608). Force, intimidation, and threat from prison gangs are the norm, as are overcrowded and starkly inadequate living conditions and the significant physical and psychological stresses of prison life. As one federal court explained, modern prison life “may press the outer bounds of what most humans can psychologically tolerate” (Madrid v. Gomez, 1995, p. 1267). Beyer (1997) paints a bleak picture of life in adult prison for juveniles, who are at greater risk for suicide and physical and sexual abuse from older inmates. As compared with juvenile facilities, juveniles incarcerated in adult prison are 8 times more likely to commit suicide, 5 times more likely to be sexually assaulted, and almost twice as likely to be attacked with a weapon by inmates and beaten by staff (Beyer, 1997). One study found that 10% of youth held in adult prisons reported being raped or sexually assaulted, 10 times higher than the rate in juvenile facilities (Ziedenberg & Schiraldi, 1997). Because juveniles in adult prisons are exposed to a criminal culture in which inmates commit crimes against each other, these institutions may socialize delinquent juveniles into true career criminals. Violent juvenile offenders who were interviewed in an older study about their life in prison (Eisikovits & Baizerman, 1983) reported that their daily survival required finding ways to fit into the inmate culture, dealing with difficult and authoritarian relationships with adult inmates, and adjusting to the institution by accepting violence as a part of daily life and, thus, becoming even more violent.

Redding and Fuller (2004) found that juveniles whose jail or prison experiences were worse than they had expected, and those who reported witnessing or experiencing violence while incarcerated, were less likely to say that their incarceration would deter them from committing crimes in the future. This finding raises the possibility that incarceration in adult facilities may have brutalizing effects on juveniles, which may partly account for the increased recidivism among juveniles incarcerated in adult facilities. The term “brutalization effect” describes the finding that homicide rates in a state often increase after an execution (Bowers, 1998), perhaps because executions model and communicate that violence is an acceptable and psychologically cathartic alternative. Likewise, juveniles’ brutal experiences in adult prison may teach the wrong lessons about the acceptability and psychological benefits of criminal conduct, particularly violent crime, while also contributing to their sense of being treated unfairly. Further research is needed on this important issue.

Psychological research and theory on the effects of punishment also suggest that harsher punishments for serious juvenile offenders, in the form of lengthy sentences in adult prisons, may have counter-deterrence effects. In what will surely become an important contribution to the deterrence literature, Robinson and Darley (2004) offer an intriguing analysis of the psychological literature on punishment, suggesting that lengthier prison sentences may have less of a specific deterrence effect than shorter sentences. Offenders gradually become desensitized to incarceration, which loses much of its initial aversive bite by the end of the lengthy prison term. Thus, a shorter sentence “will be experienced as much more aversive than a much longer sentence that is equally aversive at the beginning but less so at the end.” The experience of the incarceration just before the offender is released is what matters most psychologically when the offender subsequently calculates whether to reoffend (Robinson & Darley, 2004, p.190). Moreover, because time passes much slower for juveniles than for adults, a juvenile will experience a prison term as lasting much longer, in psychological terms, than will an adult.

With increasing numbers of juveniles being incarcerated in adult facilities, research is urgently needed on the effects of such incarceration on juveniles’ psychological and behavioral functioning and on effective, developmentally appropriate programming for juveniles in these facilities (Redding, 2003). Perhaps the most important challenge for future research is to determine what features of criminal court processing increase recidivism, an important question for policy making. For example, are there changes that could be made in the criminal court processing of juveniles to make it less detrimental? In what ways should the juvenile justice system be on guard against those features of the criminal justice system that serve to increase recidivism? How can states’ blended sentencing systems, which allow the juvenile courts to impose adult sentences in certain cases (see Redding & Howell, 2000), incorporate the best features of both the juvenile and criminal justice systems while avoiding the inadvertent negative effects of criminal justice system processing?

Implications for Legal Policy and Practice

Effective Responses to Serious and Violent Juvenile Offenders

It is important that the juvenile court’s response to first-time juvenile offenders be calibrated so as to have sufficient bite without being too punitive. Overly punitive punishments, like trial and sentencing in the adult criminal court, may have the unintended effect of delaying desistence from crime and promoting life-course criminality (Scott, 2000). Yet in Florida, for example, 43% of the 1,100 juveniles incarcerated in adult prisons for offenses committed when they were 15 years old or younger had not previously been committed to a juvenile justice program (Annino, 2000). Thus, the juvenile justice system was never given a chance to rehabilitate these youth before they were transferred to the adult system. Consider also that many of these youth were accomplices (and not the primary perpetrator) to violent offenses committed by older juveniles or that they intended to commit a property crime but unintentionally committed a violent crime (Annino, 2000).

Florida is not unique in transferring first-time serious offenders to the criminal court. Transfer laws, particularly automatic transfer laws, target these offenders even though they do not pose the greatest recidivism risk or threat to community safety (Bishop, 2004; Redding, 1997). First-time offenders typically do not reoffend—“the probability of violence following violence is especially rare” (Bishop, 2004, p. 637). Rather, it is the chronicity of offending instead of the seriousness of the first offense that predicts recidivism and the offender’s risk for committing another violent offense (Bishop, 2004; Piquero, 2000; Redding, 1997). Thus, transfer should be based primarily on the offender’s characteristics and offending history, not the seriousness of the charged offense. (First-time offenders who commit particularly serious or violent crimes may, however, warrant transfer on retributive grounds.) Transfer should be discretionary, rather than automatic, with transfer decisions made by juvenile court judges based on statutory guidelines that direct the court to consider factors relating to the offender’s psychosocial maturity, competence to stand trial in the criminal court, potential for rehabilitation, and recidivism risk (see Redding, 1997).

A small number of repeat offenders (about 8–10%) are responsible for most of the serious or violent offenses (between 60–80%) committed by juveniles, and are the offenders most likely to become the “career criminals” (Loeber, Farrington, & Waschbusch, 1998). The juveniles at risk for becoming chronic offenders can be reliably identified upon their first contact with the juvenile justice system because they have a unique constellation of risk factors that includes criminal involvement at a very early age and typically family problems, problems in school, substance abuse, and gang involvement or running away from home (Schumacher & Kurtz, 1999). Orange County, California, implemented a comprehensive early intervention program—called “the 8% solution”—for the 8 to 10% of offenders who are responsible for most of the serious juvenile crime, when they first come into contact with the juvenile justice system. The program provides an array of services that target the individual, family, school, and peer group risk factors that contribute to their offending. Initial evaluations have shown promising results in reducing recidivism (U.S. Department of Justice, 2001). An important agenda item for future research is the development of standardized risk assessment and classification instruments, such as the Structured Assessment of Violence Risk in Youth (SAVRY) (Borum, Barte, & Forth, 2002), that could be used to help to identify those few youth who should be transferred due to their high recidivism risk and relatively low rehabilitative potential. Unfortunately, however, a small percentage of offenders will require criminal sanctions. Since juvenile sanctions cannot be imposed past the age of 21, a criminal sentence is needed in some cases to ensure community safety, and this is the overriding reason for transfer (Redding, 1997). Such sentences may be achieved through transfer, or through the blended sentencing options now available in many states, which allow juvenile courts to impose limited adult sentences (see Redding & Howell, 2000).

The Role for Advocacy

There is no out-of-control juvenile crime problem, an informed public does not support the widespread adjudication and sentencing of (even) serious and violent juvenile offenders as adults, the public still supports the rehabilitative goals of the juvenile justice system, and research does not support the efficacy of punitive juvenile justice policies. Convincing policy makers of these realities will require vigorous and sustained efforts by juvenile justice researchers and advocates. Importantly, policy makers must be persuaded that less punitive, rehabilitative responses to juvenile crime yield measurable benefits in terms of reduced crime and recidivism rates, as well as reduced justice system costs (Latessa, 2004). As a leading criminological researcher and advocate says, “I have found very few policy makers unwilling to at least listen to the empirical research when you frame it within the context of public protection” (Latessa, 2004, p. 549). Our knowledge about the “causes” (risk factors) of delinquency is strong, the effectiveness of evidence-based juvenile prevention and rehabilitative programs has been clearly demonstrated in rigorous evaluations, and evidence continues to mount on the counter-rehabilitative effects of adult sanctions (see Redding, Goldstein, & Heilbrun, 2005).

Sometimes policy makers do respond. For example, following hearings in 2003 on evidence-based correctional programs, the Oregon legislature passed legislation requiring that 75% of the spending of the Youth Authority and Commission on Children and Families be allocated to evidence-based programs by the year 2009 (Latessa, 2004)! In Florida, which has had some of the most punitive juvenile justice policies in the nation, the number of juveniles transferred decreased by two-thirds between 1996 and 2003 (while the total number of juvenile court cases decreased by only 9%), apparently due to research disseminated to policy makers showing the counter-deterrent effects of transfer (Bishop, 2004). In addition, Florida has undertaken vigorous efforts to promote and institute evidence-based programming (Latessa, 2004) and has expanded the number of placements within the juvenile justice system for serious offenders (Bishop, 2004). In the last several years, some states have even reduced the scope of transfer laws to make fewer juvenile offenders eligible for discretionary transfer (Bishop, 2004; Griffin, 2003).

But at other times it can be a hard sell. When it comes to crime control, “everyone is an expert” (Latessa, 2004, p. 551) with a strong intuitive sense of what works (e.g., tougher laws and punishment), even though their intuitions often run contrary to the findings of empirical research. In my experience working with juvenile justice policy makers and practitioners, I have found that they are fairly impressed by results from treatment programs such as Multisystemic Therapy (MST) showing sizeable reductions in recidivism for serious and violent juvenile offenders (Henggeler, Schoenwald, Borduin, Rowlands, & Cunningham, 1998), and by research showing the superior effectiveness of community-based treatment (Sheidow & Henggeler, 2005). It also is not difficult to persuade policy makers that incarcerating juveniles in adult prisons will fail to rehabilitate them.

Yet many policy makers still favor transfer laws on the theory that they will deter juveniles from committing crime in the first place or that such laws enhance public safety in the near term by incapacitating, through lengthier adult sentences, serious juvenile offenders. Many also are unpersuaded by studies showing the counter-deterrent effects of transfer, because of the studies’ inherent methodological limitations (i.e., the inability to fully control for selection effects) and the somewhat counterintuitive nature of the findings (i.e., that tougher penalties have no deterrent effect). In addition, some policy makers are skeptical of research by social scientists, perceiving it to be shaped by a liberal political mindset (Redding, 2001) that is overly sympathetic to offenders. For example, in an analysis of the New York state legislature’s debate on the death penalty during the years 1977 to 1995, Galliher and Galliher (2002) show how many legislators were unpersuaded by the social science research on deterrence because they relied instead on their own commonsense assumptions, distrusted statistical analysis or social science, or distrusted the social scientists. Consider the following statements made by state senators during the floor debates:

  • It simply defies all common sense and all my knowledge of human nature to argue that a penalty of death does not act as a deterrent.
  • These studies basically are the work of criminologists or social scientists; and why they are called scientists, I don’t know.
  • Almost all of the death penalty studies were done by people who started out opposing the death penalty and wanted in effect to find out how to oppose it through the deterrence argument. (Galliher & Galliher, 2002, p. 328)

It is a mistake to argue that transfer is never appropriate or that punishment is inconsistent with rehabilitation. Those who so advocate lose their credibility with policy makers and the public, who will never accept such propositions because of “the punitive necessity of transfer” (Zimring, 2000). A small number of chronic, juvenile offenders who have not responded to previous extensive efforts at rehabilitation in the juvenile justice system do warrant criminal sanctions.

Attorneys who represent juvenile offenders at risk for transfer also play a critical role. Good legal advocacy is one of the most practical and effective ways to improve the quality of justice and programming afforded to youthful offenders. Lawyers and juvenile justice professionals must be trained on the relevant juvenile mental health and forensic issues, and, most importantly, on the rehabilitative options available or potentially available. In arguing against transfer, lawyers must be equipped to provide the court with specific and detailed recommendations for effective dispositional alternatives that will ensure community safety while providing meaningful rehabilitation. Because juvenile court transfer recommendations may be influenced by the local availability of treatment options (Mulvey & Reppucci, 1988), the emphasis should be on whether the juvenile is amenable to rehabilitation and what kinds of programs could rehabilitate the youth, not whether the juvenile is amenable given locally available resources. “A finding of amenability places some pressure on the courts to provide adequate treatment to youth who are amenable to treatment” (Salekin, 2002, p. 67). Similarly, Shridharan et al. (2004) recommend that states require localities to have in place certain services before implementing a statewide transfer law.

On a systemwide level, it is important to educate local prosecutors and juvenile court judges about the counter-deterrent and counter-rehabilitative effects of transfer and about the effectiveness of community-based rehabilitation programs. There is evidence that programs aimed at educating judges and other juvenile justice professionals can have substantial positive effects in reducing the number of juveniles receiving adult sanctions. The Miami-Dade County Public Defender’s Office developed the Juvenile Sentencing Advocacy Project (JSAP), a highly effective program that has produced a 350% increase in the number of transferred cases receiving a juvenile (rather than an adult) sanction (Mason, 2000).


The juvenile justice system was created over 100 years ago “to save young people from the savagery of the criminal courts and prisons,” by removing juveniles from the criminogenic influences of the criminal justice system while providing rehabilitative interventions (Zimring, 2000, p. 248). It appears that the founders of the juvenile court largely got it right. The available evidence, while not definitive, strongly suggests that transferring juveniles to the criminal court increases the recidivism rate. Moreover, although transfer has produced the intended effect of imposing lengthier sentences on serious juvenile offenders, the psychological literature on punishment suggests that shorter sentences may actually be experienced as more punitive (and thus, be a greater deterrent) than longer sentences. Policy makers must weigh the relatively short-term benefits of incapacitation resulting from transfer and imprisonment against the long-term costs of criminal justice system processing in terms of increased recidivism, which in turn would contribute to higher crime rates. If, however, transfer laws deter juvenile crime, then some of these offenders would not have offended in the first place. But based on current theorizing and the limited empirical research available, the weight of the evidence suggests that transfer laws, at least as currently implemented, likely have minimal general deterrent effects.

Thus, we can tentatively conclude that trying and sentencing juveniles as adults does not further the penal goals for which it was intended, particularly the goal of specific deterrence. Perhaps more importantly, however, most juvenile offenders are probably not deserving of adult punishment. As recently acknowledged by the U.S. Supreme Court in Roper v. Simmons (2005), which held that the death penalty for juveniles is unconstitutional, the limited life experience and psychosocial and brain immaturity of juveniles—particularly of juvenile offenders (see Redding, 1997)— lessens their culpability. Punishment that is proportional to the offender’s culpability should be at the heart of the justice system.