In Chicago, a 14-year-old girl is shot to death by an 11-year-old gang member who is in turn found dead a few days later, two bullets in the back of his head; his suspected killers are 14 and 16. In Fuquay-Varina, North Carolina, a 13-year-old boy is accused of beating a 22-year-old female neighbor with a mop handle and then raping her. In Somerset, Pennsylvania, a 14-year-old is charged with hammering nails into the heels of a younger boy.

Closer to home this past January in Detroit, 16-year-old Demetrius Anderson was stabbed to death by another teenager in front of dozens of classmates at Denby High School. That event came one day after another 16-year-old from nearby Pershing High School was beaten with a lead pipe by teenage gang members.

These hellish snapshots flesh out the disturbing statistics on juvenile crime rates: Between 1983 and 1992, reports the FBI, juvenile arrest rates for violent crime jumped 128 percent for murder and non-negligent manslaughter (versus 9 percent for adults); 95 percent for aggravated assault (versus 69 percent); and 25 percent for rape (versus 14 percent). Adding to the fear is the sense that things are only going to get worse.

As a Washington, D.C., Superior Court judge told Time, "Youngsters used to shoot each other in the body. Then in the head. Now they shoot each other in the face." No wonder citizens everywhere are clamoring for a crackdown on juvenile crime.

But it is far from clear whether the most politically popular reform, lowering the age at which offenders can be tried as adults, will have much, if any, effect on violent juvenile crime. Designed to let courts impose longer sentences on children (in Michigan, convicted juveniles cannot be incarcerated beyond their 21st birthday, regardless of the crime), it is inherently a rearguard action.

"The more draconian the sentence," writes UCLA criminologist James Q. Wilson, "the less (on the average) the chance of its being imposed; plea bargains see to that . . . . The most draconian sentences will . . . tend to fall on adult offenders nearing the end of their criminal careers and not on the young ones who are in their criminally most productive years."

What we need are policies that make young criminals responsible and accountable for their behavior at every turn. In the current system, youthful offenders are usually given a number of "free rides" or "diversions" before serious punishment is meted out.

As Peter W. Greenwood of the RAND Corporation notes, "studies of responses to punishment suggest that initial low levels of punishment and gradual escalation desensitize subjects and make them less likely to respond." Former Kent County Probate Court Judge Randall Hekman, now executive director of the Michigan Family Forum, agrees-- "If courts would respond decisively to offenders--even on the first offense--many potential offenders will be deterred from beginning careers in criminal activity. Furthermore, the certainty of punishment is probably a greater deterrent than the severity of punishment."

There is nothing particularly mysterious or elusive about reducing the incidence of juvenile crime. It will respond to the same thing that works with "adult" crime: raising the costs of criminal behavior, thereby making it less attractive as an alternative to lawful behavior.

Between 1950 and 1992, according to Justice Department statistics, the number of serious crimes per 100 people rose from 1.2 to 5.9, while the expected days in prison (the average sentence for serious offenses adjusted for the probability that offenders are actually sent to prison) dropped from 24 to 8. The only time during that period that serious crime rates (including juvenile ones) slowed was during the 1980s, when expected days in prison rose. Such consistent inverse correlations suggest that criminal behavior is ultimately a choice, not an inevitability.

It is particularly important for youthful offenders to learn that criminality is a choice, since long-term patterns of violent criminal behavior generally begin during the juvenile years and grow in intensity. It is rare when a murder, rape, or assault is a criminal's first crime. About one-third of all boys in the United States will be arrested before turning 18. While most will not be arrested again, for those who are, each successive arrest places them at higher and higher chances of being detained in the future, culminating in 90 percent probability for those with five or six arrests. University of Pennsylvania criminologist Marvin Wolfgang has identified the latter as the "chronic offenders," the 6 percent of boys who account for more than 50 percent of all arrests.

Each arrest, then, is an opportunity to deter the educable and incarcerate the incorrigible. A variety of well defined and sequenced sanctions, including individual mentoring, formal probation, at-home supervision, community service, non-secured group homes, and short-term incarceration in military-style boot camps or locked facilities, would allow the system to screen out salvageable offenders while providing long-term lock-up for those who remain a violent threat to society.

"If every community could deal quickly with offenders and foster an atmosphere of certainty of punishment," says Midland County Probate Court Judge Donna Morris, "we could make a difference." The reforms needed to make that happen come neither easily nor cheaply, but they are effective.