Line Drawing

by Jeffrey A. Butts and John K. Roman
February 2014

Preparation of this report was supported with funding from the David Rockefeller Fund, Elias Foundation, and The Prospect Hill Foundation. The authors wish to acknowledge the support of the New York Juvenile Justice Initiative (NYJJI), a coalition of philanthropic organizations and donors dedicated to improving the youth justice system in New York, and especially thank the NYJJI Chair, Ms. Penny Fujiko Willgerodt, and members of the NYJJI Executive Committee.


In his 2014 State of the State address, New York Governor Andrew Cuomo announced the formation of a state commission to produce a plan for raising the age at which juveniles are charged as adults in New York courts. New York was one of two states where all 16 year-olds accused of violating the law are automatically handled in criminal (adult) court. Governor Cuomo directed the new Commission on Youth, Public Safety & Justice to develop an effective strategy for changing the policy that currently sets the lower boundary for criminal court jurisdiction at age 16. He described the policy as “outdated.” The ultimate goal of the newly formed Commission, according to the Governor, was to create a “roadmap” for reforming the justice system to “promote youth success and ensure public safety.”

During 2014, the Commission was directed to:

  • Develop a plan to raise the age of criminal responsibility, including proposing concrete recommendations to protect public safety with regard to the small number of young violent offenders. As with the entire plan, these recommendations will be informed by the science of what works and other relevant factors to reduce recidivism and maintain public safety;
  • Make other specific recommendations as to how New York’s juvenile and criminal justice systems can better serve youth, improve outcomes, and protect communities; and
  • Ensure that for the small percentage of youth who engage repeatedly in violent or other harmful behavior, protecting communities and preventing victimization remain the top priority.

The following report was designed to inform the Commission’s efforts by examining the reasons for changing the age of criminal jurisdiction and by reviewing the implications of such a change. The report examines the relationship of jurisdictional age to serious crime and it reviews the experiences of states that have previously changed their jurisdictional age laws. Next, the report addresses the cost considerations involved in these policy changes and it describes the types of detailed cost-benefit analyses that New York would need to undertake to project their effects on shifting court caseloads and the number of youth likely to be placed in various supervision programs and placement settings.

Anticipating the consequences of a policy shift is always difficult, but New York can benefit from the experiences of other states. In the 1990s, several states lowered the age of criminal responsibility in a failed attempt to be “tough on crime.” A review of the record suggests that placing young people in adult courts does not, by itself, reduce crime. In two of the three states that lowered the age of criminal responsibility during the 1990s, youth violence trends in the next ten years were actually worse than the national average.

Violent crime has been falling across the United States for nearly 20 years. The pattern of these changes is not strongly correlated with policies that place larger numbers of youth in criminal court. For example, total violent crime across the state of New York dropped substantially in recent decades, but the amount of change in New Jersey was nearly the same, and New Jersey juvenile courts retain original jurisdiction over all offenses by youth under age 18.

After considering the reasons and likely effects of processing youth in either the juvenile or criminal court system, this report draws a number of conclusions:

  • Developmental and behavioral science suggest that adolescents who violate the law belong in a court system that is designed to respond to the unique characteristics of adolescents, and in this regard juvenile courts are more appropriate than the existing criminal court process.
  • Youth may be affected more negatively by contact with the current criminal justice system than are older, adult defendants.
    Nearly every state in the country has experimented with crime policies that “criminalized delinquency” by lowering the age of criminal court jurisdiction and/or transferring more youth to criminal court.
  • The weight of the research evidence suggests that these experiments failed, as changes in violent crime are not strongly correlated with the use of policies that place young people in criminal court rather than juvenile or family court.
  • If placing young people in adult courts does not advance crime reduction goals, the only remaining argument against raising the age of criminal jurisdiction is the cost and complexity of implementation, which is not a sufficient justification for the potential harm being done to youth in the existing criminal court process.
  • New York should raise the minimum age of criminal responsibility at least to age 18, and it should consider modifying practices and policies for adolescents over age 18 as well because the best science suggests that adolescent development does not end on a person’s 18th birthday.
  • Any reform efforts should include a process for estimating the administrative and fiscal implications of changes in policy.



As of 2014, New York and North Carolina were the only two states where youth as young as age 16 were handled by default in criminal court. The juvenile court (or family court) in those two states retained original jurisdiction over law violations for youth only through age 15. All cases involving 16- and 17-year-olds began in criminal court.

Until they are well above the age of 18, often in their early 20s, many adolescents and young adults lack the full cognitive abilities required for mature judgment. This raises basic questions about whether handling adolescents in a strictly retributive court process is ever appropriate. Adolescents may not be competent to understand the impact of court proceedings (Steinberg and Schwartz 2000). Juveniles are more easily distracted in court, more likely to exhibit poor demeanor, to become bored with the court process, and to make decisions based solely on immediate concerns, all of which could affect court outcomes (Tobey et al. 2000). Psychologists argue that developmental status should be a relevant factor in determining adjudicative competence (Grisso et al. 2003) and in decisions regarding leniency toward youth and the transfer of juveniles to criminal court (Brink 2004; Grisso 1996).

An expert panel assembled by the National Academies of Science in 2012–2013 reviewed the implications of neuroscience and behavioral science for youth justice. In their final report, the members of the panel asserted the following (some citations omitted from original):

Current empirical evidence from the behavioral sciences suggests that adolescents differ from adults and children in three important ways that lead to differences in behavior. First, adolescents lack mature capacity for self-regulation in emotionally charged contexts, relative to adults and children. Second, adolescents have a heightened sensitivity to proximal external influences, such as peer pressure and immediate incentives, relative to adults. Third, adolescents show less ability to make judgments and decisions that require future orientation. The combination of these three cognitive patterns accounts for the tendency of adolescents to prefer and to engage in risky behaviors that have a high probability of immediate reward but in parallel can lead to harm to self or to others. The preference for risky behaviors rises by a third of a standard deviation between ages 10 and 16, and then it declines by a half standard deviation by age 26. … One can conclude from the body of behavioral and brain studies that adolescents clearly differ from adults in crucial ways that suggest the need for a different response from the justice system. One can also conclude that age 18 does not suddenly mark complete transition to adulthood (Bonnie et al. 2013: 91).


For very good reasons related to their emotional and intellectual development, children and adolescents are held to a different legal standard when they violate the criminal law. They are handled less formally and they are more likely to receive services and supports rather than punishments and confinement.

The differential response to childhood criminality is an established legal principle. Setting the operational legal boundaries between children, youth, and adults is complicated and contentious. It has been so for centuries. It is easy to agree that a specialized juvenile court is the proper forum for handling cases involving law violations by young people, but where exactly should states draw the line between juvenile and adult status?

Every court system in the United States responds to the illegal acts of young people based at least in part on age. When a 5-year-old child steals a playmate’s bicycle, courts are reluctant to see the act as a crime. It is a family matter, or at most something for child welfare authorities to consider. When a 15-year-old steals a bicycle, he or she is more likely to be charged with “delinquency” – an act against the state’s juvenile code that is not actually a crime and thus carries lesser penalties. On the other hand, a 25-year-old who steals a bicycle is likely to be charged with the crime of theft, which is a violation of the state’s criminal code.

The principle of diminished culpability due to age is clear. Implementation is complex. Lawmakers have struggled for more than a century to draw the line between delinquency and crime in ways that respect our understanding of childhood development as well as our expectations for public safety and the costs of public policy.

Where exactly should the line be drawn? Under what circumstances should the legal system be permitted to move the line, either for an individual case or an entire class of cases? Who should make such decisions, and what are the long-term implications for youth development, public safety, and the cost-effectiveness of the justice system?

The Problem

Many historical accounts trace the origins of juvenile justice in the United States to 1899 when the first juvenile court opened in Chicago, but the idea to hold children to a lower standard of criminal responsibility has a much longer heritage (Mennel 1973). The founding principles of American law are often derived from English Common Law — the system of legal precedents formed by centuries of individual court judgments. Under Common Law since at least the 13th Century, English courts were empowered to exempt children from otherwise deserved sentences by determining that they were “too young for punishment” (Watkins 1998: 15).

Fixing the age of criminal responsibility in each case was difficult in practice, partly because judges in the pre-industrial era could not be sure of a child’s age. Birth registrations were not customary in England until the 17th century and not required until the 19th century, making exact age distinctions problematic and forcing courts to rely on other factors.

Lenience was not always granted to children, especially if a court learned that a child tried to conceal his or her crime — a critical legal distinction at the time (Watkins 1998). In 1338, for example, an English court hanged a child under age seven because he had attempted to hide from authorities after causing the death of a playmate. The court ruled that the child’s effort to avoid detection proved that he knew the difference between right and wrong, which justified the death sentence. Another English court in the 1400s sentenced a child under ten years of age to death because the boy attributed his blood stained clothes to a nosebleed, proving that he knew it was wrong to have killed another child.

By the 16th Century, English courts had settled on the idea that youth age 14 or older were fully responsible for their criminal behavior. Children under age seven were incapable of criminal responsibility, while those between ages seven and 13 were presumed not responsible — a presumption that could be reversed with evidence of intent. Without clear boundaries between the criminal responsibility of children and adults, courts were left to interpret the law on a case by case basis, sometimes with disturbing results. In 1835, after reviewing all the factors involved in a case of stealing, one English court imposed the death sentence on a nine-year-old child (Polier 1989: 38).


The Apparent Solution

With every horrific story of the excessive punishments sometimes inflicted upon children in the criminal justice system, pressures mounted for a new approach — especially in the United States. Court systems in the U.S. began to experiment with innovative approaches to juvenile crime long before the founding of the first separate juvenile court in 1899.

By the mid-1800s, American courts had already explored various methods of diminishing the severity of punishments to which children and adolescents could be subjected (Tanenhaus 2004). New York courts, for example, began to hold trials for young defendants on separate days from those of adults at least 30 years before the founding of the first juvenile court in Chicago (Watkins 1998).

Reformers wanted more; they demanded a court system and a legal code that was entirely separate from each state’s criminal code. The motivations of 19th Century reformers were not as simple as they may seem today. Some of those who advocated the establishment of juvenile courts in the late 1800s were at least as interested in crime control as they were in child welfare. Police, prosecutors, and judges went along with social reformers because they realized that this new juvenile court system would be able to intervene more quickly and more easily, without the bureaucracy and procedural protections that restricted the operations of criminal courts (Platt 1977; Rothman 1980; Schlossman 1977).

Criminal courts were designed to deliver deterrence and retribution. Their goal was to determine an offender’s guilt as expeditiously as possible and to impose a proportionate punishment. Detailed investigations of the offender’s background, character, and circumstances were unnecessary. Only the facts of a case mattered in criminal court. Transparency and fairness were important, but the fundamental purposes of the process were deterrence and retribution. Mistakes happened, of course, but criminal defendants could appeal to a higher court if they believed the process was flawed.

Juvenile courts, on the other hand, were charged with rehabilitation. Their basic goal was to identify and resolve the underlying reasons for a child’s misbehavior. Juvenile court judges were free to intervene in whatever fashion they deemed appropriate based upon the youth’s family situation, school performance, peer relationships, or anything else that seemed relevant. Juvenile courts and juvenile justice agencies were designed to focus on the offender, and not on the offense (Feld 1987). The disposition of each case was to address the unique circumstances of each youth rather than simply matching sentences to offenses.

The flexibility to fulfill this mission was provided by requiring fewer procedural protections. Officially, at least, juvenile courts were not intended to punish children. They were designed to help young people overcome their problems and develop into law-abiding citizens. Young offenders were adjudicated for acts of delinquency rather than being convicted of crimes. Until the U.S. Supreme Court began to order greater due process for juveniles, few children in juvenile court were even represented by attorneys (see In Re Gault 1967).



The deceptively straightforward demarcation between juvenile and criminal law never really worked in practice. From the very beginning of juvenile courts, state policymakers wanted to exempt certain classes of children from the protections of the juvenile code, sometimes regardless of age. When the District of Columbia enacted its first juvenile court law in 1906, it limited juvenile jurisdiction to misdemeanors. Felony charges filed against DC youth were to remain in criminal court (U.S. Senate 1927: 47). Under the 1909 New York juvenile code, children as young as age 12 who were charged with crimes punishable by life imprisonment could be tried and convicted as adults (Woods 1980: 5).

As the juvenile court model spread to nearly every state in the country between 1899 and 1930, each state legislature had to decide for itself which children to exclude from the new court. It was the same issue that bedeviled English courts for centuries prior to the founding of the American juvenile court — when should young people be presumptively responsible for their illegal behaviors, and how should the law make exceptions to this general rule? Which children belong in a separate juvenile court, and which do not?

Lawmakers tinkered with the age of juvenile jurisdiction throughout the 20th Century. States enacted complicated and often sweeping provisions that exempted certain cases from juvenile court or transferred them to criminal court, and they began to do so almost immediately after founding the new courts.

Many states began to expand their provisions for criminal court transfer as early as the 1910s and 20s, including Arkansas, California, Colorado, Florida, Georgia, Kentucky, North Carolina, Ohio, Oregon, and Tennessee. Others expanded transfers as early as the 1940s, such as Delaware, Indiana, Maryland, Michigan, Nevada, New Hampshire, New Mexico, Rhode Island, South Carolina, and Utah (Feld 1987).

Until the 1980s, most criminal court transfers had to be reviewed and approved by juvenile court judges on a case by case basis. In other words, a juvenile court judge needed to “waive” jurisdiction. In the last two decades of the 20th century, states began to rely even more on exclusion policies that moved entire classes of youth to criminal court without the involvement of a juvenile court judge or indeed any judge.

main5By the early 21st century, non-judicial transfer mechanisms accounted for the majority of criminal court transfers (Snyder and Sickmund 2006). Non-judicial mechanisms include prosecutor discretion (i.e. when charges are filed originally in criminal rather than juvenile court) and legislative exclusion (i.e. when all cases meeting certain criteria begin automatically under criminal court jurisdiction). For example, New York’s “Juvenile Offender” (J.O.) law of 1978 excluded children as young as age 13 from juvenile court if they were charged with serious offenses, and it mandated specific terms of confinement (Singer and McDowall 1988).

States do not report the volume of all transfers and exclusions in a consistent way, which means that it is not possible to know with certainty just how many youth are affected each year. The best available estimate from the U.S. Department of Justice, however, is that more than ten thousand youth otherwise eligible for juvenile court processing are transferred to criminal court each year by legislative exclusion, prosecutorial discretion, and judicial waiver combined (Griffin et al. 2011).

Ten thousand sounds like a big number until one considers the final legal pathway that sends young people under age 18 into criminal court. The single policy that exposes more youth to criminal court processing than any other is the statutory definition of the minimum age of criminal court jurisdiction, or the youngest age at which the juvenile court loses its original jurisdiction over cases involving youth.

As of 2014, New York was one of two states that defined the lower boundary of criminal jurisdiction as age 16. In eight states, criminal jurisdiction began at age 17, while in the District of Columbia and the remaining 40 states criminal court jurisdiction began at age 18. Juvenile (or family) courts in New York lost original jurisdiction over criminal law violations after age 15 and all offenses by youth ages 16 and 17 began in criminal court. In addition, 16- and 17-year-olds in New York were routinely detained and incarcerated in adult jails and prisons.

When state legislators redraw the jurisdictional line by defining the minimum age of original criminal court jurisdiction at age 16 or 17, they affect the lives of many more juveniles than all criminal court transfers by judges and prosecutors combined. In 2011, Griffin and his colleagues at the National Center for Juvenile Justice estimated that criminal (adult) courts nationwide were likely handling 190,000 cases per year involving youth under age 18, and 175,000 of those were due to statutory age exclusions (Griffin et al. 2011: 21).

The NCJJ estimate was based on two factors: 1) the population of 16- and 17-year-old youth known to be residing in states that set the minimum age of criminal court jurisdiction below age 18, and 2) the predicted rate of law violations and court referrals among 16- and 17-year-olds in those states based on knowledge about arrests and court processing among different groups by age, sex, race, and population density.

Griffin and his colleagues created their estimate with data from 2007. Since then, three states changed their laws to move the criminal jurisdiction boundary back up to age 18, including Connecticut (2012), Illinois (2014), and Massachusetts (2013). Because of these changes, the youth population still affected by upper age exclusions is likely to be 36 percent smaller than it was in 2007 (1.4 versus 2.2 million), according to recent population estimates for these states from the U.S. Census Bureau.

main6bMoreover, according to the Bureau of Justice Statistics, the arrest rate of 16- and 17-year-olds dropped at least 25 percent since 2007 (Snyder and Mulako-Wangota 2014). Thus, the national number of 16- and 17-year-olds still excluded from juvenile courts by their state’s jurisdictional age limits may now be 84,000 (75% x 64% x 175,000) — which is still more than seven times the estimated 11,000 transferred by all other means (i.e., 75% of the estimate of 14,000 calculated by Griffin et al. in 2011).

All forms of exclusion, therefore, may still affect as many as 95,000 cases each year. The central question is, “what exactly do we get from these policies?” Are states placing 95,000 young people in criminal court each year because the policy is a good investment in public safety?


When lawmakers or court officials place large numbers of youth in criminal court rather than family court, they often simply assume that their actions improve public safety. They assume that the criminal court process and the sanctions available through the criminal court have stronger deterrent effects compared with those available in the juvenile system, either by changing the behavior of transferred youth directly (“specific deterrence”) or by influencing other youth not to offend in the first place (“general deterrence”).

Researchers have tried for several decades to find clear evidence for the deterrent effect of using criminal court sanctions rather than juvenile court sanctions for young offenders. To date, only one or two studies have managed to tease even a slight effect from the available data (e.g., Levitt 1998). In nearly all existing studies, researchers find either no evidence of a differential effect on subsequent offending, or they find negative effects — i.e. criminal court handling is associated with more rather than less crime (see Mulvey and Schubert 2012).

There are a number of ways New York policymakers could test the conventional assumption that moving youth into criminal court improves public safety. One way would be to examine long-term trajectories in violent crime between states according to where they process most young offenders.

If placing larger numbers of young people in criminal court enhances public safety, there should be a clear pattern in serious crime that favors states with lower ages of juvenile court jurisdiction. Violent crime trends should be consistently worse in states that keep more youth in juvenile court. Trends should be better in states that exclude all 17-year-olds from juvenile court, and they should be even better in states that exclude all 16-year-olds as well.

An analysis of violent crime trends in the six largest Northeastern states between 1980 and 2010 does not show the expected pattern. The percentage drop in violent crime over 30 years is more pronounced in New York, and this may appear at first to support the hypothesis that excluding more youth from juvenile court leads to better public safety outcomes. The decline in New Jersey, however, is nearly as steep as the decline in New York, and New Jersey extends juvenile court jurisdiction through age 17.

Similarly, while the pattern appears to be worse in Pennsylvania, a state that places 16- and 17-year-olds under original juvenile jurisdiction, crime trends in the remaining states do not support the hypothesis. Maryland, Massachusetts, and Connecticut represent all three variations in upper age of juvenile court jurisdiction — or at least they did until upper age laws changed in Connecticut (2010-2012) and Massachusetts (2013). Until 2010, Massachusetts excluded 17-year-olds from juvenile court and Connecticut excluded 16-year-olds and 17-year-olds, while Maryland retained all 16- and 17-year-olds.

If the public safety hypothesis was accurate, a comparison of violent crime trends in these states would show different patterns. Yet, the percentage change in violent crime over 30 years is similar in all three states. Each state experienced the nationwide rise and fall of crime, but by 2010 levels of violent crime dropped to almost exactly the same degree in all three states — about 80 percent of 1980 levels.

Thus, differences in the upper age of juvenile court jurisdiction did not produce consistent effects on public safety as measured by the change in total violent crime.


Of course, it may be too much to expect that differences in the minimum age of criminal court jurisdiction in a state would affect total levels of violent crime. If so, another way to test the effect of a change in policy would be to compare youth crime in states before and after a policy shift.

The policy changes in Connecticut, Illinois, and Massachusetts are too recent to conduct a fair test of their impact on crime, but three other states changed their laws during the 1990s: New Hampshire, Wisconsin, and Wyoming. In all three instances, laws were changed to move more youth into criminal court. Wyoming enacted a law in 1993 to move offenses by 18-year-olds under the original jurisdiction of the criminal court instead of juvenile court. In 1996, both New Hampshire and Wisconsin moved 17-year-olds out of juvenile court and into criminal court. If the public safety hypothesis is correct, an analysis of crime trends in those states should reveal a consistent pattern of improving crime rates.

Again, the trends in serious crime do not support the idea that more criminal court handling for youth improves public safety. If moving more youth into criminal court had a strong, positive effect on public safety, an analysis of the three states should show them all beating national trends during the crime decline of the 1990s and early 2000s. In fact, however, two of the three states (Wyoming and Wisconsin) experienced violent youth crime trends after they changed their jurisdictional age laws that were actually worse than the national average. A 2-to-1 split across three different states showing negative effects from lowering the minimum age of criminal court jurisdiction is not exactly strong evidence to suggest that placing more youth in criminal court reduces serious crime or improves public safety.


main10As a crime control policy, placing more young people in criminal court appears to symbolize toughness more than it actually delivers toughness, and that symbol may have a high price. Sending more juveniles to adult court may not result in significantly lower crime or less recidivism among young offenders, but it likely means longer pre-trial delays, more pre-trial incarceration with few services to address youth problems, greater population management problems in prisons and jails, and greater exposure of youth to adult inmates. Fagan reached this conclusion nearly 20 years ago after an extensive review of the long-term outcomes for youth tried in criminal court compared with those retained in juvenile court:

By neither public safety nor punishment (or just deserts) standards can claims be made that the criminal justice system affords greater accountability for adolescent felony offenders or protection for the public. If criminalization is intended to instill accountability, its effects are diluted by the lengthier case processing time. If it is intended to protect the public by making incarceration more certain and terms lengthier, it fails also on this count. While these processes may have symbolic value to the public, they seem to offer little substantive advantage in the legal response to adolescent crimes. It is only for the earlier accumulation of a criminal record, leading to lengthier terms and more severe punishments for subsequent offenses, that there is a marginal gain in the relocation of adolescent crimes to the criminal court (Fagan, 1996: 100-101).

For these reasons, state policymakers increasingly agree that the juvenile court is the proper forum for handling cases involving law violations by young offenders, and for now they appear to be settling on age 18 as the most appropriate minimum age boundary for original criminal court jurisdiction. Before New York joins the other 40 states that place offenders younger than age 18 under the jurisdiction of family courts, lawmakers will have to consider a range of administrative and programmatic issues. One of the key issues involved in such policies is anticipating costs and benefits.



Cost always plays a central role in policy proposals to raise the age of juvenile jurisdiction, and three issues tend to dominate debate about what costs and benefits to count. The choices made by analysts as they address these three issues can lead to extremely different conclusions about whether moving juveniles into the juvenile system is cost-effect or cost-prohibitive.

First, some analyses, such as the official fiscal analysis published by the Connecticut government during the recent debate in that state, include only administrative costs and savings to the government. This narrow perspective ignores the broader benefits to be realized from improvements in public safety. For instance, many studies suggest that youth processed in the juvenile justice system experience less recidivism than they would have had in the adult system. Including the savings from reduced victimization would be crucial to a full understanding of a proposal to raise the age. Moreover, and as detailed in the Vera Institute’s recent analysis of a potential change in North Carolina policy (described below), preventing youth from being labeled as felony offenders in criminal court would likely improve their future earnings. Whether all of those additional benefits should be captured in a cost-benefit analysis is debatable, but not including any of them yields a biased estimate of the real effects of the policy change.

Second, researchers must forecast the size of the newly expanded population likely to enter the juvenile justice system after a policy change to raise the age. On one hand, there are likely to be added population pressures from creating a larger and older age group with eligibility for juvenile processing. On the other hand, systems always adjust to new circumstances. Some of the juveniles adjudicated delinquent under the old rules (particularly the youngest youth) may not be convicted when older youth begin to enter the juvenile system. As described below, the State of Illinois originally expected to see increasing juvenile court caseloads after raising the age, but the opposite happened. Caseloads continued to fall after the change in law. A cost analysis should not simply count the number of 16- and 17-year-olds being handled in New York criminal courts and then assume that the same number of youth will appear in family courts after raising the age. Doing so would overstate the costs of the policy change.

Third, a cost analysis in New York should weigh carefully any assumptions about how a change in the age of jurisdiction will lead to changes in the number and types of facilities needed to house adjudicated youth. Incorporating the construction of new facilities in an analysis could generate costs that exceed benefits. This was a powerful determinant of the original (and mistaken) models used in Connecticut. Such calculations are based on the simple assumption that all youth held in a correctional facility when processed by the adult system will be mechanistically housed in secure confinement for the same length of time in the juvenile system. This ignores substantial differences in correctional philosophy and treatment resources available in the two systems. A New York study should pay careful attention to differences in case processing, placements, and duration of custody between the two systems.

Prior Research

Comprehensive and accurate investigations of these issues require substantial investments in research and analysis. The following discussion describes the most recent research on the costs and benefits of raising the age of jurisdiction, the strengths and weaknesses of existing studies, and the implications of choices made by researchers in conducting the studies. The report describes the data necessary to conduct a rigorous study if the goal is to ensure that data, rather than a researcher’s choices, drive the findings. Finally, the report discusses the type of trade-offs that are typically necessary given real-world data limitations in modeling complex policy shifts in justice systems.

Three states recently studied the costs and benefits of raising the age of jurisdiction. Researchers estimated the effects of such policy changes in Connecticut (Roman 2006), North Carolina (Henrichson and Levshin 2011) and Illinois (Kollman 2013). Of the three studies, the Vera Institute’s effort for North Carolina was the most adequately funded. While the study period was relatively brief, it is the best example of a successful project conducted within practical time and resource constraints.

The Connecticut report was generated with few resources, a very short study period, and it relied on publicly available data alone. While it conformed to many of the recommended analysis strategies, the limits on time, data and money, meant that the report had less impact on the policy debate. The Illinois study was conducted by an appointed commission and was concerned with a variety of matters beyond the cost benefit analysis, and thus the economic analysis itself was quite limited, although it did underscore the fact that raising the age does not always increase the juvenile justice population.

North Carolina

logo_veraIn 2011, the North Carolina Assembly created a Youth Accounting Planning Task Force to study the question of raising the minimum age of criminal court jurisdiction. The Task Force commissioned the Vera Institute of Justice to study the costs and benefits of a policy shift. The Vera report (Henrichson and Levshin 2011) calculated the costs and benefits of raising the minimum age for misdemeanants and low-level felons to age 18 and returning 16- and 17-year-olds to juvenile court (assuming that more serious felons would remain in the criminal system). The benefit-cost analysis included fiscal cost increases and cost savings (benefits) to the government, benefits to victims from reductions in the number of recidivism incidents, and both costs and benefits to youth involved in the system (mainly increases in future earnings that would be achieved by allowing youth to avoid the stigma of a criminal conviction). The report estimated that raising the jurisdictional age would return about $50 million in benefits on an ongoing, annual basis (about a 40 percent reduction from current costs).

The Vera report highlighted key assumptions made in the study. In doing so, the authors revealed many of the challenges involved in this type of research. For instance, the analysis assumed that the number of youth arrested for offenses which qualified for juvenile justice handling would not change with the modification in jurisdiction. However, research generally shows that the probability of system contact for youth subject to juvenile versus criminal court handling is not equivalent. Especially for minor or less serious crimes, the juvenile system is often more likely than the adult system to process and adjudicate, even for youth charged with similar offenses (Kupchik et al. 2003). Thus, the Vera analysis could underestimate fiscal costs to the extent that there is net-widening as youth move into the juvenile system.

The Vera study also assumed that 16- and 17-year-olds would experience a 10 percent reduction in recidivism if they were processed in the juvenile rather than the criminal system. While there is supportive research on the question of re-offending when youth are handled in juvenile versus adult courts (e.g., Fagan 1995; 1996), there is no clear consensus on the expected rate of reduction (though we note that this assumption does not appear to be a large driver of cost savings). The Vera study also assumed that 16- and 17-year-olds would progress through the newly expanded juvenile justice system at the same rate as younger youth (i.e. the percentage of youth arrests likely to result in adjudication was not expected to vary by age). There are several plausible reasons why this may not occur, notably that the influx of older youth may effectively crowd younger youth out of the system, which would result in additional cost savings. National juvenile court statistics suggest that the probability of formal prosecution and adjudication among delinquency cases could vary by age and offense, and these variations should be included in long-term estimates (Sickmund et al. 2013).

Two additional assumptions in the Vera study were particularly important. First, the analysis assumed that by returning original jurisdiction over 16- and 17-year-olds to the juvenile system, youth would avoid the stigma associated with an adult conviction, and thus would have improved employment prospects, increasing both lifetime wages and government tax receipts. While there is certainly evidence to support such an assumption, these earnings represented a substantial portion of the total benefit identified by the Vera analysis, and it is not clear how the probability of later contact with the criminal justice system was incorporated into the analysis. Second, the authors explicitly noted that no capital costs were included in their study — costs that would be associated with future construction of courtrooms and secure confinement facilities if the change in law were to result in more rather than fewer justice contacts for youth. This issue was a prominent concern in the Connecticut study. It may be fair to assume no such costs, of course, but such assumptions must be examined.


urbanheader09In 2009, Connecticut lawmakers signed a bill that raised the minimum age of criminal jurisdiction in the state from age 16 to 18, effective in 2012. In early 2006, the Urban Institute conducted an analysis of the costs and benefits of the state’s initial proposal to raise the age (Roman 2006). The Urban Institute approached the issue in a manner similar to that of the 2011 Vera study. First, the study identified the number of 16- and 17-year-olds processed annually by the criminal system for various types of crime. Next, using the same assumption as the Vera study that older youth would progress through the system with the same probabilities as youth currently in the system, the study modeled the probability of proceeding from one event (e.g., arrest) to later events (e.g., judicial handling) through disposition and placement (e.g., probation, incarceration, training school, and parole). Daily costs and the average number of days per service type or placement setting were used to estimate costs in the new juvenile system. The study relied on a parallel process to estimate savings to the adult system. For future years, the study assumed that processing in the juvenile system would reduce recidivism by 29 percent (cf. Fagan 1995). Thus future costs for the juvenile system were reduced by 29 percent. The Urban Institute study also considered costs to victims from reduced recidivism, using estimates of the value of harm from criminal victimization taken from Rajkumar and French (1997).

The study estimated that with no new capital spending, benefits would exceed costs by a ratio of $3: $1. The Urban Institute study did not consider potential costs of new capital spending, on the grounds that the literature showed recidivism reductions would likely offset the new population. The Connecticut Fiscal Office, when scoring the proposed legislation, did not consider the potential reduction in recidivism, but did include capital costs for new facilities. This issue remained contentious throughout the legislative process in Connecticut, with costs often being cited for the two-year delay in final implementation.


ILReportIn 2013, the Illinois Juvenile Justice Commission released a report describing, among other matters, the costs and benefits of a 2010 law to return misdemeanor 17-year-olds to juvenile jurisdiction (Kollman 2013). The study used national data from the National Center for Juvenile Justice (NCCJ) to predict expected changes in juvenile justice caseloads, assuming that juvenile court processing would conform to national trends. The report expected the policy shift would result in substantial increases in juvenile caseloads, but a subsequent analysis of Illinois data actually showed reductions. The report did not address this important issue: whether the declines following national trends were simply a reflection of those trends, or if they were a result from the change in the upper age law.

Estimating Costs and Benefits in New York

As this review of past research shows, there are a number of decisions that need to be addressed in modeling the costs and benefits of raising the minimum age of criminal court jurisdiction in New York. Some decisions, like the choices in the Connecticut and North Carolina studies to exclude capital costs, are based on interpretation of past studies or data limitations. Other choices, such as the decision to include future earnings in the North Carolina study but not in the Connecticut study, are more arbitrary. Other issues, such as the choice made in the Illinois study to exclude national trends as a possible explanation for the decline in juvenile arrests and referrals, are more fundamental.

These choices could profoundly affect any New York analysis. Excluding capital costs reduced expected costs in Connecticut, and the decision to include future earnings increased expected benefits in North Carolina. Both decisions were supported by prior research, and both studies included a sensitivity analysis, basically a replication of the study without that assumption in place. The Illinois choice to attribute any declines in arrests to the changes in law and not to larger national trends may well have introduced bias into the analysis, where the real cause of an effect is not identified.

A New York study will present other challenges, of course, including data availability. Modeling a change in New York’s law will require at least three types of data to be collected:

  1. To understand the fiscal implications of raising the age, data must be collected about current costs associated with processing youth through the juvenile justice system.
  2. Next, equivalent data about the costs associated with processing same-aged youth in the adult system must be collected and analyzed;
  3. Finally, data on recidivism outcomes must be used to estimate the relative effects of processing youth in either the juvenile or adult system.

New Costs

The main challenge of understanding the costs of moving youth from the adult system to the juvenile system is that we cannot observe the experience of the same youth in both systems. First, since youth cannot be simultaneously processed in both systems, establishing an equivalent comparison group is difficult. The most obvious choice is to use the experience of those nearest in age currently processed in the juvenile system to estimate the expected costs for the incoming youth. To the extent that there are evidence-based reasons to believe that the experience of older youth newly under juvenile jurisdiction will have different experiences, those should be incorporated into the analysis as well. (We note a potentially biased, but intuitively appealing alternative would be to compare youth processed in the juvenile justice system with same-aged youth who were transferred to the adult system. This would compare a relatively less delinquent cohort of youth with a group of more serious offenders.)

main15The most rigorous analysis will rely on individual-level data rather than aggregated data. Individual data tracks the experience of each youth in the juvenile system (to estimate what new costs older youth will incur) and the adult criminal justice system (to estimate what savings the adult system will experience). Thus, these data would track the experience of individual youth as they progress through the system. For youth in the juvenile system, this would include data on whether they progressed from one stage of processing to the next, and if placed, how long the placement lasted.

Individual-level information has two substantial advantages over aggregate data. First, it allows researchers to distinguish prevalence (how many different youth enter the system) from incidence (how many different arrests lead a youth into the system). If a small number of youth generate a disproportionate number of arrests, then changes in their processing may lead to enhanced cost savings. Second, individual data show variation, where most youth receive relatively short placements and a few receive much longer placements. Again, a change in practice for such youth would lead to cost savings.

Next, as was done in the Connecticut report, the processing of youth currently in the juvenile system must be documented. The goal is to observe as many different “cost” events as possible. For instance, after a juvenile is arrested, he or she may be released to home or held in detention. At an initial court hearing, the youth may proceed to probation, which may assign either diversion or court referral. A court adjudication leads to a disposition which may be a placement (i.e. confinement), probation, or various other types of placement with social service agencies. Or, the case may be dismissed or otherwise not lead to any placement. Placements may include a day placement, placement in a private facility or placement in a state operated facility. The case may also be dropped during this process such that far fewer youth are placed than are arrested. Each progression along this chain of events introduces a different set of costs, and relying on data to observe as many different costs as possible will improve the accuracy of cost estimates.

Benefits to the Criminal (Adult) System

The criminal justice system benefits conversely to new juvenile costs. Each youth not processed in the adult system accrues some costs savings. Thus, the estimation process must be repeated in reverse. Individual-level data should be collected from the adult system and each stage of case processing should be observed. There are two important caveats to this otherwise straightforward analysis.

First, it is important to observe case processing for as many different categories of crime as possible. It is likely that the processing of youth charged with assault, for instance, differs between the adult and juvenile systems. Savings in one system are not necessarily a mirror of costs in the other system. Youth charged with assault may have large costs in the adult system (and thus their removal has larger savings), but they may be less costly in the juvenile system. Although the benefits of savings from not processing youth in the adult criminal justice system may be generated from adult data about those types of cases, the expected new costs of juvenile system processing should be generated from data about similar cases already in the juvenile system.

The second caveat is that youth may experience altered criminal offending over their life course as a result of which system they experience. Beyond the fiscal costs, changes in recidivism rates and changes in the severity of re-offending will have cost consequences for the juvenile justice system in future years. If, as estimated by Fagan (1995) youth in the juvenile system experience 29 percent less recidivism, then those reductions in re-offending should be “priced” into the analysis. As youth recidivism is reduced, the costs of justice system processing should fall as well.


Other Benefits

As all three prior studies noted, a wide range of potential costs and benefits may accrue from a change in jurisdictional age. Cost-benefit analysts often talk about policy changes as being similar to a rock thrown into a pond, where consequences ripple outward over time. The first ripple may be fiscal costs to affected systems. The next ripple could be costs associated with changes in re-offending behavior. This has certain consequences for private citizens whose potential victimizations were prevented. The next ripple may be the change in non-delinquent (or non-criminal) behavior of youth, such as future labor market participation, as noted in the North Carolina study. Moreover, as in the Illinois study, better outcomes for youth entail positive consequences for their families, friends and communities as well.

Ultimately the decision about how far to extend the benefits calculation is a policy choice. The general rule in cost-benefit analysis is that as consequences occur farther away from the direct event under study, the analysis should treat their linkage to that event as more fragile. That said, the goal of the justice system is to protect communities, not to limit spending, so focusing exclusively on fiscal savings understates the relevant goals. Regardless of whether the benefits of a change in jurisdiction are extended merely to averted victimizations, or to future earnings and beyond, these benefits should be the subject of a sensitivity analysis to determine how much they contribute to the overall costs and benefits.

Recommendations for a Cost Analysis

This review of the reasons and approaches for raising the minimum age of criminal court jurisdiction leads to the following recommendations for New York lawmakers as they plan a change in the state’s justice system.

Rely on Individual Data

Individual-level data are necessary for estimating the extent to which outcomes vary around an expected average. Many youth receive relatively short placements while a few youth receive very long placements. The costs of a change in jurisdiction can be substantially affected by policy recommendations that limit the length of placement through diversion or home-based treatments. In the Connecticut study, only publicly available data were available and researchers had to rely on data such as the average number of days in each type of placement by crime type. As a result, there was no way to observe variation around the means.

Differentiate Between Policy Choices and Research Choices

The decision about how far to extend an analysis beyond a narrow study of fiscal cost is a policy choice that should be included in a proposed scope of work. Policy choices are those where there is some reason to support one approach versus another in conducing a cost study (such as including future earnings or including costs beyond the justice system such as victimization harms). Research decisions by contrast are decisions about how to make unbiased estimates using the best available data within constraints determined by policy choices. This would include decisions about how to use the experience of current youth in the justice system (which are likely the only viable comparison group) to model the expected experiences of older youth coming under juvenile jurisdiction.

Always Include Sensitivity Analysis

Where policy choices are made in the presence of uncertainty (e.g., where there is not a clear research precedent), investigations of costs and benefits should include sensitivity analyses that model key assumptions under every plausible scenario. This increases transparency and facilitates more informed debate. For example, when costs and benefits are estimated using an analysis that assumes a 29 percent reduction in recidivism outcomes, other analyses should test the sensitivity of those results. What if the reduction in recidivism turned out to be 20 percent or 40 percent? Would this alter the policy recommendation?


This report reviews evidence in support of raising the minimum age of criminal court jurisdiction in New York, describes the critical role economic consequences often play in such policy debates, and suggests how lawmakers could move forward by projecting the costs and benefits of reform in a transparent and objective way. Raising the minimum age of criminal court jurisdiction is not a radical idea. Enacting such a policy change in New York would bring the state’s practices in line with those of most other states, and studies suggest that the costs of expanding the juvenile system would be at least matched by the benefits to be gained from reducing criminal court processing for youth. New York can learn from the experiences of other states and proceed with this reform based upon a rigorous review of policy and practice for justice-involved youth and a careful analysis of the costs and administrative details involved in changing the law.



Bonnie, Richard J., Robert L. Johnson, Betty M. Chemers, and Julie A. Schuck (Editors) (2013). Reforming juvenile justice: A developmental approach. Washington, DC: National Research Council of the National Academies.

Brink, David O. (2004). Immaturity, normative competence, and juvenile transfer: How (not) to punish minors for major crimes. Texas Law Review, 82(6): 1555-1585.

Fagan, Jeffrey (1995). Separating the men from the boys: The comparative advantage of juvenile versus criminal court sanctions on recidivism among adolescent felony offenders. In A Sourcebook: Serious, Violent, and Chronic Juvenile Offenders, edited by James Howell, Barry Krisberg, J. David Hawkins, and John Wilson. Thousand Oaks, CA: Sage Publications, Inc.

Fagan, Jeffrey (1996). The comparative advantage of juvenile versus criminal court sanctions on recidivism among adolescent felony offenders. Law & Policy, 18: 77-113.

Feld, Barry C. (1987). The juvenile court meets the principle of the offense: Legislative changes in juvenile waiver statutes. Journal of Criminal Law & Criminology, 78: 471-533.

Griffin, Patrick, Sean Addie, Benjamin Adams, and Kathy Firestine (2011). Trying youth as adults: An analysis of state transfer laws and reporting. Washington, DC: Office of Juvenile Justice and Delinquency Prevention, U.S. Department of Justice. [NCJ 232434]

Grisso, Thomas (1996). Society’s retributive response to juvenile violence: A developmental perspective. Law and Human Behavior, 20: 229–247.

Grisso, Thomas, Lawrence Steinberg, Jennifer Woolard, Elizabeth Cauffman, Elizabeth Scott, S. Graham, F. Lexcen, N.D. Reppucci and Robert Schwartz (2003). Juveniles’ competence to stand trial: A comparison of adolescents’ and adults’ capacities as trial defendants. Law and Human Behavior, 27: 333-363.

Henrichson, Christian and Valerie Levshin (2011). Cost-benefit analysis of raising the age of juvenile jurisdiction in North Carolina. New York, NY: Vera Institute of Justice.

Kollman, Stephanie (2013). Raising the age of juvenile court jurisdiction: The future of 17-year-olds in the Illinois’ justice system. Illinois Juvenile Justice Commission.

Kuchik, Aaron, Jeffrey Fagan and Akiva Liberman (2003). Punishment, proportionality, and jurisdictional transfer of adolescent offenders: A test of the leniency gap hypothesis. Stanford Law & Policy Review, 14(1): 57–83.

Levitt, Steven D. (1998). Juvenile crime and punishment. Journal of Political Economy, 106(6): 1156–1185.

Mennel, Robert M. (1973). Thorns and thistles: Juvenile delinquents in the United States 1825-1940. Hanover, NH: The University Press of New England.

Mulvey, Edward P. and Carol A. Schubert (2012). Transfer of juveniles to adult court: Effects of a broad policy in one court. Juvenile Justice Bulletin. Washington, DC: Office of Juvenile Justice and Delinquency Prevention, U.S. Department of Justice. [NCJ 232932]

Platt, Anthony M. (1977). The child savers: The invention of delinquency. Chicago: University of Chicago Press.

Rajkumar, Andrew S. and Michael T. French (1997). Drug abuse, crime costs, and the economic benefits of treatment. Journal of Quantitative Criminology, 13(3): 291-323.

Polier, Justine Wise (1989). Juvenile justice in double jeopardy: The distanced community and vengeful retribution. Hillsdale, NJ: Lawrence Erlbaum Associates.

Roman, John (2006). The economic impact of raising the age of juvenile jurisdiction in Connecticut. Testimony before the Judiciary and Appropriations Committees, Connecticut General Assembly. Washington, DC: Urban Institute.

Rothman, David J. (1980). Conscience and convenience: The asylum and its alternatives in progressive America. Glenview, IL: Scott, Foresman and Company.

Schlossman, Steven L. (1977). Love and the American delinquent. Chicago: University of Chicago Press.

Sickmund, Melissa, Anthony Sladky, and Wei Kang (2013). Easy access to Juvenile Court Statistics: 1985-2010. Online. Available:

Singer, Simon I. and David McDowall (1988). Criminalizing delinquency: The deterrent effects of the New York juvenile offender law. Law & Society Review, 22(3): 521-536.

Snyder, Howard N. and Melissa Sickmund (2006). Juvenile offenders and victims: 2006 national report. Washington, DC: Office of Juvenile Justice and Delinquency Prevention, U.S. Department of Justice.

Steinberg, Lawrence and Robert Schwartz (2000). Developmental psychology goes to court. In Thomas Grisso and Robert Schwartz (Editors), Youth on trial (pp. 9-31). Chicago, IL: University of Chicago Press.

Tanenhaus, David S. (2004). Juvenile justice in the making. New York: Oxford University Press.

Tobey, Ann, Thomas Grisso and Robert Schwartz (2000). Youths’ trial participation as seen by youths and their attorneys: An exploration of competence-based issues. In Thomas Grisso and Robert Schwartz (Editors), Youth on trial (pp. 225-242). Chicago, IL: University of Chicago Press.

United States Senate (1927). Message from the President of the United States, transmitting a letter from the Judge of the Juvenile Court of the District of Columbia, submitting a report covering the work of the juvenile court during the period from July 1, 1906 to June 30, 1926. Washington, DC: United States Government Printing Office. [Reprinted in Metropolitan America, New York, NY: Arno Press, 1974.]

Watkins, John C. Jr. (1998). The juvenile justice century. Durham, NC: Carolina Academic Press.

Woods, John P. (1980). New York’s Juvenile Offender Law: An overview and analysis. Fordham Urban Law Journal, 9(1): 1–50.


Copyright by the Research & Evaluation Center, John Jay College of Criminal Justice, City University of New York (CUNY).

Research & Evaluation Center
John Jay College of Criminal Justice
524 West 59th Street (BMW 605)
New York, NY 10019
Published online February 2014

Any views expressed here are those of the authors and do not necessarily reflect the policies of John Jay College, the City University of New York, the Urban Institute, or their funders and trustees.

Recommended Citation
Butts, Jeffrey A. and John K. Roman (2014). Line Drawing: Raising the Minimum Age of Criminal Court Jurisdiction in New York. New York, NY: Research & Evaluation Center, John Jay College of Criminal Justice, City University of New York.

Jeffrey A. Butts is director of the Research & Evaluation Center at John Jay College of Criminal Justice, City University of New York.
John K. Roman is senior fellow with the Justice Policy Center, Urban Institute, Washington, DC.