Moreover, the likelihood that juveniles will be prosecuted in criminal court, rather than juvenile court, varies according to race, gender, time and county in which they are prosecuted in, says Simon I. Singer, UB associate professor of sociology.

For example, juveniles arrested for certain offenses in Albany County are much more likely to be prosecuted in criminal court than those arrested for the same offenses in Erie County.

In a new book, "Recriminalizing Delinquency: Violent Juvenile Crime and Juvenile Justice Reform" (Cambridge University Press, 1996), Singer examines how one state-New York-attempted to control violent juvenile crime by redefining certain acts of delinquency as crimes, and delinquents as juvenile offenders. It begins with the brutal violence of a 15-year-old chronic delinquent and the subsequent passage of legislation that transferred jurisdiction over juveniles from juvenile court to criminal court.

The 1978 Juvenile Offender (JO) Law-also referred to as waiver legislation-lowered the age of criminal responsibility to 13 for murder and 14 for rape, robbery, assault and violent categories of burglary. The law also allowed criminal justice officials discretion in charging a juvenile as an adult. Juveniles who were considered amenable to treatment in the juvenile justice system could be removed to that system through a process known as reverse waiver.

Singer argues that the reasons for bringing juveniles into criminal court go beyond sensational acts of violence and the immediate concerns of elected officials to do something about violent juvenile crime. Recriminalization, he says, is a product of earlier juvenile justice reforms and modern-day political and organizational interests.

Although the book focuses specifically on New York State, Singer notes that other states have enacted their own versions of waiver legislation. And 73 percent of respondents in a 1993 USA Today/Gallup survey favored transferring juveniles to criminal court for violent crimes.

Singer studied the development, implementation and effect of the JO Law in New York by analyzing data collected by the state Division of Criminal Justice Services between Sept. 1, 1978, when the JO Law took effect, and May 30, 1985.

He found that the four largest counties in the state-all located within New York City-account for 85 percent of JO arrests. Brooklyn alone produced one-third of the JO arrests. Bronx officials charge juveniles on average with more serious offenses than do officials in the other New York City counties.

Erie County, which includes the state's second-largest city, Buffalo, recorded little more than 1 percent of the total JO arrests, although the county's population of nearly 1 million is comparable to other downstate counties where JO arrests are more common. In Albany County, officials seem to follow the letter of the JO Law by charging a larger proportion of juveniles for less serious JO offenses. In Albany, more juveniles are arrested as JOs than in Erie County, based on the counties' juvenile populations. The rate in Albany is 93 per 100,000 juveniles, compared to 35 per 100,000 in Erie County.

Singer interprets these variations as differences among jurisdictions as to what constitutes a JO offense. "Officials in Buffalo and Rochester appear more inclined to follow the letter of the JO Law only for the most serious, violent offense charges," he says. "In these Western New York counties, juveniles who commit the less-serious JO offenses are less likely to be arrested as juvenile offenders and are more often charged as juveniles."

This may be due, he says, to a more loosely coupled criminal justice system in which a difference in opinion and concerns among law enforcement officials may keep all but the most serious cases exclusively in the juvenile justice system.

Singer's analysis also found that the majority of cases of juveniles arrested for offenses for which they were criminally responsible are dismissed or removed to juvenile court. Only 25 percent of juveniles arrested as offenders were ultimately convicted in criminal court-10 percent if "youthful offender" (YO) status is eliminated. YO status includes the possibility of probation or a reduced term of imprisonment.

In addition, the analysis determined that the incarceration rate for JOs was linked to the temporal cost to a county of incarcerating the offenders. During 1979, the first full year under the JO Law, less than 4 percent of eligible juveniles were incarcerated. The incarceration rate increased until 1983, when it peaked at 21 percent. In 1984, a year after the cost-sharing formula for counties of jurisdiction kicked in-whereby counties gradually were required to increase the cost they paid for the incarceration of a JO from 12.5 percent to the 50 percent that it cost to incarcerate a juvenile delinquent through family court-the percentage of juveniles incarcerated started to drop. By the time the cost of incarceration increased to the full 50 percent in 1988, the rate of incarceration plunged to 6 percent of juvenile-designated felony arrests.

Is the JO Law "more bark than bite?" Yes and no, Singer says, noting that it all depends on the offense and offender. Juveniles who are charged with more serious offenses face a higher probability of conviction in criminal court and a lower probability of YO status, or having their cases removed to juvenile court or dismissed. Older juveniles, boys and minority youth are more often adjudicated as JOs.

Moreover, justice also appears to depend on jurisdiction and timing, he adds, with the manner in which the JO Law is implemented being different in various parts of the state.

Singer evaluated the impact of the JO Law on crime by comparing the number of monthly juvenile arrests both before and after the law was implemented. He found little evidence that the JO law produced a decline in arrests.

There are several explanations for this, Singer says. The law may be too weak to produce a measurable effect on crime patterns. In addition, juveniles may not have been deterred by even a marginal increase in the certainty and severity of punishment promised by the law.

Moreover, the loose fit between crime and punishment for juvenile offenders in criminal court runs contrary to an important principle in the theory of deterrence that stresses the certainty of punishment, he says. Although most juveniles appearing before juvenile courts have received relatively little punishment, the probability of punishment is "certainly not zero." The additional, harsher penalties of the JO Law may not have appeared significant enough to deter potential offenders, he says.

Singer recommends that states enact a more tightly coupled system of juvenile justice in which recriminalization is more narrowly focused and confined to the most serious violent offenses. A combination of legislative waiver-in which juveniles cannot be convicted in criminal court except for the most violent offenses, such as murder-and judicial waiver-which includes a wider range of offenses-would set limits as to the kinds of offenses for which juveniles would enter criminal courts. He doesn't believe, for example, that the current provision for automatic waiver for "C" felonies for 14-year-olds for crimes such as assault and burglary makes much sense.

At the same time, a tightly coupled system of juvenile and criminal justice would require that criminal court officials be aware of an adult's earlier juvenile court convictions for violent designated felonies.

"Juvenile justice has developed into a complex, bureaucratic system with all sorts of legal avenues and legal categories," Singer says. "With this, you have all sorts of bureaucratic interests and conflicts that develop independent of the needs of the child and the needs of the state. We need to have a more tightly coupled system of juvenile justice that is less susceptible to political whims and really takes to heart the best interests of the child and that of the community."